This

boob is dedicated to Henry M. Wallbrunn, beloved teacher, mentor, and

friend.

 The Gardener's and Farmer's Guide to Plant Breeding and Seed Saving

CAROL DEPPE

Acknowledgments ix

Introduction to the Original Edition Xi

Introduction to the Second Edition xiv How to Use This Book xvi

Part I: An Introduction to Plant Breeding and Seed Saving

1 Amateur Vegetable Breeding 3

Why every gardener should be a plant breeder. Stories of three amateurs and
what they've done.

2 How Much Space Do You Need? How Much Time? 17

Vegetable breeding can be done on any scale. You can do an elaborate

tomatobreeding project in a few half-gallon pots of soil, or a pea-breeding
project in a few feet of row. Some projects require only a year or two to
produce material that is an improvement over anything available
commercially.

3 Roles and Goals for Amateurs; Wish Lists and Wild Ideas 23

Breeding for flavor. Breeding for size, shape, color, earliness, cold or heat
resistance, disease resistance, regional adaptation, yield. Breeding as an
expression of individuality, for your tastes and needs. Breeding varieties that
do well under organic gardening or farming methods. Breeding new and
unusual crops. Discovering popbeans and tiny fast-cooking chickpeas
(garbanzos). Thinking small, thinking big, daring to dream.

4 Finding Germplasm 42

Obtaining germplasm and information about it. How to work with seed
companies, seed saving organizations, and plant science professionals. How to
gain access to and use the collections of the USDAARS National Plant
 Germplasm System.

5 Evaluating Germplasm and Experimental Material; Variety Trials and

Gardening Research 54

How to design, conduct, and evaluate garden trials. How to combine trials with
production of food and beauty. How to get good information with the least
amount of land and labor.

6 Genetics and Plant Parenthood 77

7 Sex and the Single Gene; Mendel's Genes 85

8 Modern Genes 93

9 The Genetic Basis of Seed Saving 111

Inbreeding and the genetic nature of inbreeding crop varieties. Saving seed
from inbreeders. Heirlooms. Outbreeding and the genetic nature of
outbreeding crop varieties. Inbreeding depression. Outcrossing and self-
incompatibility. Saving seed of outbreeders. Inbreeder or outbreeder - how you
can tell? Saving seed from hybrids. Making and breaking hybrids.

10 Plant Breeding Stories 126

Popbeans and purple peas. Perennial vegetable buckwheat and perennial

lettucesalsify. Power selection, power inbreeding, crosses, backcrosses, and
recurrent backcrossing. `Rainbow Inca' sweet corn. Tomatoes, squash, and
melons.

11 Bigger, Brighter, and More Beautiful 148

Creating polyploids. Chromosome doubling using colchicine. Breeding with

established polyploids.

12 Fun with Wide Crosses 152

Crosses between distant relatives within a species. Crosses between different

species. Creating entirely new crop species.
13 Happy Accidents 156

Taking advantage of new mutations, sports, bud sports, and accidental crosses.
More cold-hardy fava beans, bigger tomatoes, and giant top-setting onions.

14 Domesticating Wild Plants 163

15 Expanding Horizons 172

Table I: 801 Interesting Plants 175

Vegetables of the world and their wild relatives; edible plants that have the
potential for being developed into vegetables; fruits, nuts, and grains.
Scientific names, common names, families, and lifestyles. Basic breeding
systems, chromosome numbers, flowering patterns, flower types and
modifications, average cross-pollination frequency, major pollen vectors, and
incompatibility system information. Recommended isolation distances, seed
yields, location in the USDAARS National Plant Germplasm System, and
references.

Part II: Seed Saving Practice

16 An Introduction to Seed Saving 209

Why save seeds? Seed-saving overview. Saving seed from hybrids. Roles and
purposes.

17 Growing Seed 213

Preparation and planning, planting, tending.

18 Isolation 218

Traditional seed saving and isolation distances. Isolation distances for organic
farmers and gardeners. Isolation basics. Factors that affect the need for
isolation. Isolation distances, absolute and practical. The Basic Rule for
Everyday Seed Saving. Isolation tricks and methods.

19 How Many Plants? 232

 Inbreeders and outbreeders. Practical compromises.

20 Selection 237

Selection basics. Selection complexities, subtleties, and surprises. Selection for

the purpose of germplasm preservation. Evaluating a selection program. Does
selection always work?

21 Harvesting, Processing, and Storing Seed 243

Harvesting, threshing and cleaning. Dry and wet processing. Drying seed.
Protecting seed from insects and rodents. Storing seed.

Part III: Developing Crops for a Sustainable Future

22 Genetic Engineering and Genetically Modified Foods 261

Carol Deppe meets the FlavrSavr tomato. Standard plant breeding versus
genetic engineering. Everything in this book is illegal with genetically
engineered varieties. Genetic engineering and sustainable agriculture.

23 Conversations with a Squash 272

Why not just select? Choosing the right cross. The agroecological virtues of a
squash. The grand plan. Choosing the cytoplasm. The reality. The squash
speaks. Carol falls in love. Disaster and opportunity. Sandwich-slice. To
market, to market, to sell a new squash.

Appendix A. Breeding and Seed Saving for Eight Common Vegetables - An

Illustrated Guide 291

Tomato 291

Lettuce 296

Pea 299

Common Bean 304

Alliums (Onions) 309

Brassicas (Cabbage and Relatives) 311

Squash and Pumpkins 314

Corn (Maize) 317

Appendix B. Technical Aspects of Hand-Pollination and Performing Crosses;

Overcoming Incompatibility Barriers 322

Appendix C. USDAARS Plant Introduction Stations and Germplasm

Collections; Using GRIN (Germplasm Resources Information Network) 332

Appendix D. Addresses of Seed Saver Exchanges, Seed Companies, and

Organizations 335

Appendix E. Sources for Seed Saving, Plant Breeding, and Garden Research
Supplies 339

Appendix F. Statistical Predictions and Actuality 340

Glossary 341

Annotated Bibliography 347

Index 350

 ' welcome the opportunity to thank the following people for information,
assistance, and encouragement: Ken Allan, Jan Blum, Dan and Ruth Borman,
Rebecca Brown, Helen Deppe, Glenn Drowns, Miriam Ebert, Ewald Eliason,
lanto Evans, John Gale, Rich Hannan, Howard Haynes, Karl Huber, Alan and
Linda Kapuler, Damon Knight, Cy Dratzer, Rob Mabe, Rose Marie Nichols
McGee, Ralph McNees, Tim Peters, Forest Shomer, Anita Sullivan, Jeff and Joy
Taylor, Lee Wallingford, Kent Whealy, Kate Wilhelm, and Jimmy Williams.

I am indebted to Roger Swain, science editor at Horticulture, who shepherded

this book in its initial stages, when it was only a magazine article with ambitions.
I am especially grateful to Catherine Crawford and Christina Ward, my editors at
Little, Brown and Company, for their enthusiasm, support, and guidance. I also
thank Barbara Jatkola for her excellent copyediting, Jan Blum for use of three of
her photos, and Abigail Rorer for her illustrations in Appendix B.

Finally, I acknowledge and thank Richard Balkin, my agent.

Chapter 1 of this book originally appeared, in somewhat different form, in

Horticulture.

Acknowledgments to the Second Edition

I acknowledge and thank Christina Ward, who started as this book's editor in
the first edition and continued as its agent in the second. I am grateful to
everyone at Chelsea Green Publishing Company for their enthusiasm, skill, and
support for making a second, expanded edition of this book possible. Most
especially, I thank Ben Watson for his kind, gentle and skillful editing.

A few paragraphs of Section II (here and there) were first published by

National Gardening Magazine (Parts I and II of `Custom-Crafted Vegetables,' the
June/July and August issues, 1998).
Introduction to the Original Edition

LL OUR major food crops were originally developed by amateurs.

Until recently, all gardeners and farmers saved their own seed; all gardeners and
farmers were automatically amateur plant breeders - and amateur plant breeding
was the only kind of plant breeding there was.

This book has two major purposes. First, I want to encourage all of you
gardeners and farmers to rediscover the excitement, the pleasure, and the
rewards of developing your own vegetable varieties. Second, I want to show
amateurs how to breed plants more efficiently. Our ancestors often needed
generations to accomplish what a modern professional plant breeder does in just
a few years. But the major techniques modern professionals use are simple, and
they require no special equipment. Any gardener can do them.

This book extracts out of modern genetics exactly those methods needed for
practical small-scale plant breeding and presents them with a minimum of
theoretical trappings. I hope that this book will enable the modern amateur to
breed plants with very nearly the efficiency of the modern professional.

I feel I am ideally situated to write this book because I am both a professional

plant breeder and an amateur. I have one foot firmly planted in the soil of the
professionals because I am a geneticist. I have a Ph.D. in biology from Harvard
University with a specialty in genetics and more than twenty years of experience
as a working geneticist.

I have my other foot even more firmly planted in the soil of the amateur plant
breeder. No one pays me to breed vegetables. I do it just for fun. I learned plant
breeding on my own only after leaving academia and only after finding that I
could not get many vegetable varieties that I wanted for my garden. They didn't
exist. As a geneticist, I could see that much of what I wanted would be easy to
develop. But no one was doing it.

The primary focus of this book is on vegetables and their wild relatives and on
wild plants that have the potential for being bred into vegetables. The techniques
presented herein can easily be applied to breeding other plants, however. Those
interested primarily in plants other than vegetables should note that Appendix A,
"801 Interesting Plants," gives basic breeding information for most common
food plants of the world as well as for many uncommon ones.

This book is for all gardeners everywhere. It's for every gardener who has ever
dreamed of bigger, better, tastier, earlier, more productive, or more beautiful
vegetables. It's for every gardener who has ever been told, "You can't grow that
here," but who wants to anyway. It's for people who may not have a lot of space
or time, but who have some ideas - ideas about what they wish they could grow
that isn't available now.

This book is for every gardener who likes to grow unusual vegetables, unusual
varieties of vegetables - rarities. It's a lot of fun to discover a plant that does well
in your area and to introduce it to your friends and neighbors. And what could be
rarer than the plant that nobody in the world has ever grown before, because
you've just developed it?

This book is for every seed saver. Saving seed is the basic method of plant
breeding. If

you save seed, you are already a plant breeder, already practicing selection,
already choosing what genetic material to perpetuate. This book will help you do
it better.

In addition, it's a short step from saving seed to deliberately breeding plants.
Why limit yourself to preserving the achievements of the amateurs of the past?
They didn't. You can both preserve the fine old heirloom varieties and use them
to create new varieties of your own.

Growing and preserving heirloom and open-pollinated varieties combines well

with plant breeding. Heirloom varieties often serve as sources of breeding
material, ideas, and ideals. Breeding projects give you more excuses to grow and
evaluate heirlooms. After all, you don't know what you want to breed for until
you've tried a number of varieties and have some favorites.

You can grow and maintain heirlooms and open-pollinated varieties and use
them to develop your own hybrids and produce your own hybrid seed. You can
have hybrids whose parents you know, hybrids you yourself control. In addition,
you can use simple plant breeding methods to take any desirable commercial
hybrid and convert it into a stable open-pollinated variety.

This book is for every gardener or farmer who grows organically, who wants
plants that yield well without pesticides, herbicides, and chemical fertilizers. It's
for everyone who is interested in the sustainability of agriculture, who wants
varieties designed for use in sustainable agricultural systems. It's for everyone
who cares about strengthening the regional basis of gardening and farming, who
wants plants that are fine-tuned to produce well with minimal cost or labor in
specific localities and under specific conditions. The surest way to get such
varieties is to breed them yourself.

But above all this book is for every adventuresome gardener - every gardener
who likes to try things, who likes to experiment - who just likes to play around.

 elcome to the new, expanded second edition of Breed Your Own
Varieties. Vegetable

This book is in many ways two books written by two different people.

Part I, the core of the first edition, is full of the joy of first discovery. My
formal academic background was in genetics and molecular biology, not in plant
breeding. I used the process of writing the book to learn about plant breeding
myself.

Part I is set in the years of about 1990 to 1993, when the first edition was
written. The techniques of standard plant breeding were fully developed by then,
and haven't changed since. However, some of the seed companies I refer to in
Part I no longer exist. Others have merged or transmogrified. And seed people
have changed too - migrating from one company to another, starting their own
companies, and so on.

In reviewing Part I for the second edition, I found it impractical to try to

change the specifics about people and seed companies to reflect the realities of
the year 2000. So I left Part I largely intact in the second edition, allowing it to
represent people and institutions as they existed in the early nineties. I have
updated the appendixes, however, to reflect current information about seed
companies and seed people.

Parts II and III, written in spring of 2000, are the work of greater experience.
They reflect my current interests and situation - an increasing concern with the
overall sustainability of agriculture and the security of our food supply as well as
my own transition from amateur plant breeder to professional, and from
backyard gardener to farmer and food company.

In between the writing of the first and second editions of this book, the entire
world of plant breeding has changed almost beyond recognition. In 1993 there
was a serious dearth of university-based vegetable breeders working on projects
aimed at the public interest. Now there are almost none. Changes in patenting
laws have allowed publicly funded institutions to patent plant varieties created
with public money.

The result of these legal changes is a new generation of university-based

breeders who create mostly patented varieties and F, hybrids - who choose to
work on developing varieties resistant to herbicides, for example - instead of
varieties that taste better or grow better, that are more agroecologically adapted,
or that enhance the viability and independence of the farms in their regions
instead of destroying them.

In short, most university-based plant breeders now behave in exactly the same
way as agribusiness breeders. Profit, not the real needs of gardeners or farmers,
is their primary motivation. But what is most profitable for a corporation is not
necessarily what is most profitable, let alone most important, for farmers or
gardeners.

In addition, genetically engineered crops and foods were only on the horizon
in 1993. Now they are a reality. In the universities, the last generation of plant
breeders skilled in standard field methods has retired or is retiring. They are
being replaced with genetic engineers. Greater perspective would have added
genetic engineering to a full repertoire of standard plant breeding. Instead, we
now have universities full of "plant breeders" who only have one tool - genetic
engineering - and who are prepared to conduct only those projects that use the
genetic engineering approach.

There are many limitations as to what can be approached with genetic

engineering, however. Much of what is most important to home gardeners and
farmers is not very amenable to the genetic engineering approach. And almost
all of what matters in breeding plants for organic or sustainable agriculture is
best and most easily approached using standard plant breeding methods. In their
blind leap onto the genetic-engineering bandwagon, the universities and
multinational seed companies have virtually abandoned nearly all of the most
important problems.

Crop varieties incorporate the values of their creators. When you grow varieties
bred by others, you propagate their values along with their varieties. Today's
professional plant breeders - university and corporate - are breeding plants to
facilitate and serve the modern megafarm agribusiness pattern. These varieties
produce well in huge monocultures grown with massive doses of herbicides and
chemical fertilizers. Bred into the varieties are the values of their creators - that
more is always better, that monocultures are best, and that pollution,
biodiversity, and sustainability don't matter.

It is time for new patterns - new patterns for agriculture, and new patterns for
plant breeding. It's time for the rising up of a new generation of plant breeders
out of the very soil of our farms and gardens. It is time for farmers and gardeners
everywhere to take back our seeds, to rediscover seed saving, and to practice our
own plant breeding. It is time to breed plants based upon an entirely different set
of values.

 ►he information, concepts, and stories in Part I build consecutively.
Thus, it's best to read the chapters in Part I in the order presented.

Parts II and III assume familiarity with Part I, but don't depend upon it entirely.
I believe they stand alone well enough that many people will be able to read the
three parts independently. Chapters within each part build somewhat
consecutively, however.

Chapter 5 in Part I comprises a thorough but nonstatistical introduction to

doing variety trials, and to garden-and farm-based plant and crop research.

Table I and the appendixes constitute a reference book to plant breeding that
should be useful to both amateur and professional plant breeders. The
appendixes also comprise an advanced course in standard plant-breeding
methods. The appendixes fill the gap between the amateur and professional
level.

Table I gives basic breeding information on more than eight hundred plant
species. It is organized by scientific name. If you know only the common name
of a plant, look it up in the index. The index cross-references common and
scientific names.


 VERY GARDENER should be a plant breeder. Developing new
vegetables doesn't require a specialized education, a lot of land, or even a lot of
time. It can be done on any scale. It's enjoyable. It's deeply rewarding. You can
get useful new varieties much faster than you might suppose. And you can eat
your mistakes.

There has never been a better time to get involved in amateur vegetable
breeding. The seed saver exchanges that have emerged during the past decade
provide a rich source of raw materials for plant breeding. The smaller seed
companies, many also founded recently, are eager to help perpetuate and
distribute the creations of amateurs. And the professional plant breeders are all
busy elsewhere. They are engaged, almost exclusively, in developing
commercial varieties of vegeta bles - vegetables bred for uniformity and once-
over picking so they can be harvested by machines, for tough skin and hard flesh
so they aren't ruined by those machines, and for good storage and shipping
characteristics so they can be transported long distances. These are not usually
the qualities home gardeners need. Yet most of the new varieties that are released
annually with such fanfare are commercial cultivars.

Gardeners buy only small amounts of seed compared to commercial growers,

so seed of varieties that are best suited for gardeners is sold in only small
amounts. Large seed companies often can't afford to carry it. No one can make a
profit developing it. So no one is. If we gardeners want good new garden
varieties, we'll have to breed them ourselves. But this is as it should be.
Gardeners have been developing their own varieties for centuries. Besides, why
should we let the professionals have all the fun?

This chapter is an introduction to how to breed your own vegetable varieties

and focuses on three amateurs who have done it. None of them had any special
education in genetics or plant breeding. Yet each has produced good new garden
varieties - varieties that have been formally tested by a seed company and have
been introduced or scheduled for introduction; varieties that have been found
worthy of being listed and sold side by side with the best developed by anybody
anywhere.

Glenn Drowns was only sixteen when he started breeding plants, but he was
already an experienced and enthusiastic gardener. It all started when he was two
and a half, and his family planted some flower seeds. He was fascinated. By the
age of four he was crawling through the fence to help his neighbor, who had a
bigger garden. At eight he had a 500square-foot garden of his own and was
ordering his own seed. About then he developed a passion for vine crops. Each
year he ordered every squash variety he could find. By age eleven he was selling
all the produce his family didn't need and using the money to buy seed. By age
sixteen he was growing fifty different varieties of squash and ten of cucumbers.
He also tried about half a dozen of the shortest-season watermelons he could
find.

More than anything else, Glenn wanted a ripe watermelon. His family lived in
the extreme northern part of Idaho, where the growing season is short, cold, and
unpredictable. His melons wouldn't ripen. "I tried everything," he recalls. "I even
grew them inside little plastic tents all summer, just really baked them. But I
never got a ripe melon. The only fully ripe watermelons I had ever tasted came
from the stores - which didn't count."

Then Glenn took high school biology, and his class discussed crop
improvement and hybridization. "Wow, maybe I could get a ripe melon that
way," he thought. So he tried a cross. One of the parents was probably `Sugar
Baby'. The other was from a package of seed some friends had given him. The
package was labeled "watermelon." Glenn didn't keep much in the way of
records. He knew about record keeping, but he thought he was only playing
around.

Within four years, growing no more than a dozen plants per year, Glenn
Drowns produced a new stable variety - a variety that reliably produces similar
plants from seed to seed and year to year. `Blacktail Mountain' is a round, deep
green melon with very faint stripes. Vines grow out to form a plant about 10 feet
in diameter. The average melon is 8 inches across, weighs 8 to 10 pounds, and
has a rind about half an inch thick. The flesh is orangish red, crisp, and very
sweet.
 What is essential about `Blacktail Mountain' is that it's early - quite possibly
the earliest watermelon ever grown. It's about five days earlier than `New
Hampshire Midget', for example, one of the earliest watermelons. `Blacktail
Mountain' has been scheduled for introduction by Seeds Blum, of Boise, Idaho,
in 1993 or 1994.

Glenn, now twenty-seven and a high school science teacher with a degree in
biology, laughs when he describes the breeding program that led to `Blacktail
Mountain'. Glenn hadn't been able to find out anything at all about.how to do
crosses. "So I just sort of guessed," he says.

Watermelons, like most cucurbits, have separate male and female flowers.
Glenn noticed that some flowers had what looked like little fruits underneath and
others didn't. He figured the flower buds with fat bases must be females and the
ones with skinny bases must be males. He made that first cross exactly the way
he does watermelon and squash crosses today, eleven years and thousands of
hand-pollinations later.

Glenn starts by taping the male and female flower buds shut with masking tape
in the evening before the buds open for the first time. The timing is part
convenience and part necessity. Late afternoon of the day before opening also is
OK, but any earlier and the buds might still be growing fast enough to damage
themselves on the tape. The following day Glenn untapes the buds. If they are
ready to open that day, they will slowly expand after untaping. He plucks the
male flower and uses it to sprinkle pollen onto the stigma of the female flower.
(The stigma is at the top of the pistil.) Then he retapes the female flower with
fresh tape and labels it. He never untapes the female flower; it just shrivels at the
end of the developing fruit.

Glenn crossed the putative `Sugar Baby' to the unknown watermelon. Then he
saved the seed and planted it. The plants of that first generation, as he recalls,
were similar to the `Blacktail Mountain' of today. He selfpollinated each. "That
was 1978," Glenn remembers. "Nineteen seventy-eight was a gruesome

year." It was unusually cool, and his garden suffered a lot of deer damage. Only
a few plants survived. They gave him exactly one ripe watermelon. "Selection
was easy," Glenn says. "There was just one ripe melon. So I selected it."
 He planted the seed from that melon in 1979. The plants that resulted, his
second generation, looked pretty similar to those of 1978. He selfpollinated each
and got two or three ripe melons. He planted the seed from those in 1980, but the
vines were all frozen out on August 4. Fortunately, Glenn had known better than
to plant all his seed in one year. He replanted in 1981 and continued to inbreed -
that is, to self-pollinate each plant. By 1982 he was offering `Blacktail Mountain'
through the Seed Savers Exchange. It was already stable and was basically the
same variety it is today. But he grew it for a number of years before he felt
confident enough in its stability to offer it to a seed company.

Part of the fun of breeding your own varieties is the surprises. Glenn had
deliberately selected for a very early watermelon. Unexpectedly, `Blacktail
Mountain' also proved to have unusual keeping qualities. Glenn found that out
completely by accident. "There are long-storage melons, the Christmas types,"
Glenn says. "I'm surprised more people don't grow them. I always do, and I eat
my last watermelons in February. But the storage types all seem to have that
white rind. I've always thought of the green-skinned melons as an immediate
eating thing." Storage melons are harvested at just under ripe and finish ripening
during storage. Glenn usually had no reason to harvest `Blacktail Mountain' that
way. But Iowa, where Glenn lives and gardens now, intervened.

In 1988 an early heat wave and drought killed all Glenn's melon vines, so he
had to replant in late June. Then, the last week in August, came the torrential
rains. The melon field was in a low place. It flooded. Only the earliest half dozen
varieties had melons that were mature enough to harvest. Glenn grabbed those
and put them in his garage. Many were not quite ripe.

Several weeks later, Glenn opened one of his `Blacktail Mountain' melons to
look at the seed. To his surprise, the melon popped open. And it was just as crisp
and tasty as if it had been harvested in its prime and eaten immediately. The
melons from the other five varieties had long since turned to mush. Intrigued,
Glenn used the rest of his melons in a storage trial. He found they kept nicely for
up to two months. That's not as long as a winter-storage type, but it's unexpected
for a standard green-skinned type, and it means that Glenn can eat his last
`Blacktail Mountain' watermelons on Thanksgiving.

The development of `Blacktail Mountain' was unusually easy. The plants of the
first generation after Glenn's cross already had the characteristics he was looking
for. It's much more common for the desired type to show up only in a later
generation. In addition, there was little variation among the plants in subsequent
generations. That also was unexpected.

A more typical example is Glenn's development of `Mexigold' squash. He

started by crossing `Mexican Banana', a blue-green banana squash, to `Golden
Delicious', which is orange and heart shaped.

Glenn usually grows just a few plants of that first generation after a cross,
because the plants are usually all pretty similar. He plants as many of the second
generation as he has room for, though, often fifty or more. Each second-
generation plant is different from the others, each showing a different
combination of the characteristics of the original parents.

Glenn doesn't remember what the first generation of plants after the cross were
like. He selfpollinated them to obtain a second generation. That seed grew into
plants of all sorts, each with its own kind of squash. There were bananas and
hearts of all hues, pinkish and bluish and orangish. He selected for pink hearts.
Each year he selfpollinated plants that produced pink hearts and eliminated the
rest. After a number of years, all the plants produced pink hearts. Glenn had
"stabilized" the strain - had developed a new pure breeding variety.

`Mexigold' is a pink heart-shaped squash that weighs 10 to 12 pounds. It's very

productive. But like `Blacktail Mountain', `Mexigold' produced a surprise: it's
extremely early. Glenn had selected for squash shape and color, not earliness,
and neither of the two original parents were particularly early. Yet `Mexigold' is
consistently earlier than `Buttercup', one of the earliest squashes. In fact, the
only variety Glenn knows of that is earlier than `Mexigold' is 'Arikara', which
doesn't have good table quality. `Mexigold' will be introduced by Seeds Blum
when enough seed becomes available.

It's not always necessary to start a plant breeding project by doing a cross.
Glenn sometimes simply notices and uses genetic accidents. One vine on one of
his bush acorn squash plants, for example, had white fruit. It was a mutant. So
Glenn saved the seed from that plant and is inbreeding to develop a white bush
acorn line. He's selected over a number of generations and has mostly
whitefruited plants now, but occasionally the strain will still throw a plant that
produces green fruit. Glenn is still inbreeding and stabilizing the line.
You don't have to start out as a youngster to become an amateur plant breeder.
Nor must you necessarily take pains to do carefully controlled crosses. You
might not even have to do any inbreeding to stabilize new varieties. A lot
depends on the vegetable you're working with.

Ewald Eliason, age seventy-three, is a retired Minnesota dairy farmer. He

started breeding his own potatoes seventeen years ago. "I've always enjoyed
gardening," he says, "and potatoes are my favorite."

Potatoes are much easier to work with than squash or watermelons. For either
of those, Glenn Drowns has to control pollination to develop or maintain a
cultivar. Otherwise, bees would cross all his cultivars at random, and the seed
would not breed true. Maintaining a dozen melon or squash varieties entails
hand-pollinating each variety each year. But since potatoes are propagated from
tubers, not seed, all Ewald has to do to maintain a variety is plant its tubers. The
resulting plants are clones of and genetically identical to the mother plant,
whether insects are crosspollinating the flowers or not.

Ewald started breeding potatoes just by saving true seed from the plants in his
gar

den. The tuber pieces that are used as starts are called seed, so potato breeders
say "true seed" when they mean the product of sexual reproduction. Since potato
cultivars are maintained by vegetative propagation, they are usually not
genetically pure. When true seed is planted, the seedlings are individuals, each
genetically different from its parents and all others. When such a plant forms
tubers, it is an instant new variety.

Fertilized potato flowers form berries that resemble small green tomatoes.
Each contains about two hundred seeds. Ewald picks the ripe berries, puts them
in a food blender with water, and blends very briefly. The seeds sink, and the
pulp floats. He plants the seeds in pots in a cold frame and transplants to his
garden.

As Ewald's potato hobby grew, he accumulated more and more varieties.

When the Seed Savers Exchange started, he joined and obtained varieties -
including many heirlooms - from fellow members. (An heirloom variety is one
that has been handed down from generation to generation.) He received potatoes
from friends. He visited Minnesota's North Shore Experiment Station one
summer and obtained experimental material. His sister went to Finland one
summer and brought him back some true seed from European types. And he
started doing his own crosses.

Potatoes, unlike watermelons, have perfect flowers - flowers containing both

male and female parts. Potato flowers are usually pollinated by pollen from other
potato flowers, but they do self-pollinate to a significant extent. Professional
potato breeders emasculate the flowers when they do crosses - that is, they
remove the stamens before the pollen is shed.

Ewald, however, doesn't bother emasculating. He just takes a flower from the
variety he wants as the male parent and dusts its pollen onto the stigma of a
flower of the other variety. Sometimes the offspring will be from his cross, and
sometimes they won't. "I just dust the pollen on there and take what I get,"
Ewald says. "Most of life is like that, I've found." Whatever the seedlings are,
Ewald knows they're sure to be interesting.

Just sprinkling pollen on there and taking what you get is a perfectly valid
breeding method. When professionals do it, they call it a "fertile x fertile cross."
Dignified with that fancy name, it sounds much less cavalier. All it means,
though, is that the breeder felt the task of emasculating was not worth the effort.
In some vegetables, for example, the flowers are so tiny that emasculating is
very difficult or even impossible. For practical breeding work, it's usually not
necessary that all the offspring of an attempted cross really be hybrid, as long as
enough of them are.

Ewald plants fifty to one hundred seeds from each cross he does. Virtually all
the plants from his various crosses give him hills of usable potatoes. Some plants
don't yield well, however, and others are scabby. Most plants are pretty good, he
finds, but not all are worth increasing and maintaining. During the growing
season Ewald evaluates the potato plants - how vigorous they are, and how
disease free. At the end of the season he digs the potatoes and looks at them. He
likes potatoes that "grow clean" - that are scab free and blight resistant. He also
makes notes

about the yield. Then, during the winter, he evaluates the potatoes for keeping
and eating quality.
 Ewald grows twelve 100-foot rows of potatoes per year, but since he has about
three hundred varieties, he doesn't usually have very much of any of them. His
soil is sandy loam, and he has subsurface water. He doesn't have to irrigate, even
during droughts. He never sprays, either. He gardens organically. Presumably, he
is selecting for varieties that perform well under organic gardening methods.

Ewald's favorite variety - "so far," he says emphatically - is `Mesabi Gold',

which Seeds Blum plans to introduce when enough tubers are available. It's a
result of a cross between `Swedish Mandell', a yellow fingerling type, and
`Minnesota #30'. `Mesabi Gold' has yellow skin, yellow flesh, and good keeping
and eating quality, Ewald says. Yield is good, too - about seven or eight full-size
tubers per hill. The growing plant is distinctive. Unlike most potatoes, it has
upright stems about 30 inches high. (Both parents had the normal bushlike
form.) "This is like a throwback to a tomato," Ewald jokes. "It's more a starch-
type tomato."

`Blossom', another of Ewald's varieties, was introduced by Seeds Blum in

1989. It's one of Jan Bliim's favorites. "It's a beautiful potato," she says. "It has
red skin and pink flesh and is a very nice producer. And it has lots of big flowers
and looks just wonderful in the garden."

`Blossom' came from a random berry off a plant of `Poorlander', an heirloom

variety. `Blossom' has good keeping and eating quality, says Ewald, and yields
potatoes "of more than ordinary size." But unlike most potato plants, which
flower only on the ends of their stems, `Blossom' also tends to flower at the
nodes. Ewald thinks it could be used to breed an ornamental potato. "I'm
thinking it has the potential of being developed into a plant you could grow in
your flower garden," he says, "and at the same time, you could enjoy the
potatoes."

Ewald Eliason is still breeding more and more potatoes. "When you grow all
the various plants and dig them up, every one is a surprise," he says. "They're
like children in a family. You just don't know what you're going to get."

Like Glenn Drowns, Tim Peters became an amateur plant breeder while still in
high school. Glenn had the good fortune to start with a vegetable that was easy
to work with. Tim began with broccoli, a plant breeder's nightmare. And
although Glenn seems to have been born a gardener, Tim had it thrust upon him.
 At the age of fifteen, Tim was not particularly interested in gardening or in
plants. He liked animals and was learning how to breed fish. His father, though,
was an avid gardener. Then the family moved from California to Myrtle Creek, a
small town in the foothills of the Cascade Mountains in southwestern Oregon.
The soil was poor. The insects were terrible. There was no rain. After a couple of
years, Tim's dad took him to the garden and said, "It's your job."

Tim was not pleased. He wanted to spend his time with his fish. Instead he
hauled manure, weeded, and watered the garden by hand. "Boy, I hated that," he
recalls. "But after a while I decided that since I was going to be stuck there in the
garden all summer, I'd better start learning to like it. And after a while I did start
liking it some."

At the end of that summer Tim saved the seeds from a tomato plant some
friends had given him. It was those seeds that changed him. When he held them
in his hands, they seemed magical, mysterious, profound ... and he could hardly
wait for the next year to come.

The following year Tim made his first cross, a selfpollination of `Green Comet
Hybrid' broccoli. It was a disaster. He saved the seed and planted it, but the
plants hardly grew at all.

Broccoli, like many of the brassicas, has a genetic incompatibility mechanism

- a physiological mechanism that prevents selfpollination. The plants have
perfect flowers and are anatomically capable of pollinating themselves, but the
pollen doesn't usually fertilize well. In some cases it won't germinate at all on the
stigmas of flowers on the plant from which it was derived. In other cases the
pollen germinates but grows so slowly that it is usually beaten out by foreign
pollen. Even where self-pollen grows and achieves fertilization, seed may not
develop properly or, if it develops, may produce only weak, sickly seedlings.

Tim didn't give up. He had read in a magazine that the Japanese practice bud
pollination of brassicas. He didn't know what bud pollination was, but it sounded
like a good idea. So he started pollinating buds, and it worked. Only much later
did he find out that the incompatibility property of the stigma of brassicas
develops fairly late during flower formation, so that by pollinating in the bud
stage, he was bypassing the incompatibility system entirely.
 Later in his breeding project Tim needed to isolate broccoli plants to prevent
crosspollination. A professional would have had isolation farms where each
variety could be grown far from any others. Tim developed his own approach.
He put the plants in pots and loaded them onto his bicycle. Then he rode around
looking for isolated homes. When he found one, he knocked on the door with a
potted broccoli in his hands, explained what he was trying to do, and asked the
owners if they could help him out. Usually they did. Tim's early work with
broccoli didn't lead to any new introductions, but he learned a lot about breeding
plants.

After he graduated from high school, he received a plea for help from a friend
and fellow Seventh-Day Adventist in Africa, and off he went. He spent several
years setting up a farm for a mission school and breeding tropical corn. He
returned in 1980, during a "minidepression" in Oregon. There were no jobs.
There was also no money for college. Tim got a part-time job doing landscaping
and found a position as caretaker on a farm. The caretaking job included a trailer
for him to live in and a 4-acre field where he could grow and breed vegetables.
Oregon State University, just a few hours north, had a fine agricultural library.
"I'm going to learn everything I can from this field and that library," Tim
resolved. And he did.

Tim works with broccoli, tomatoes, corn, melons, mustard, cauliflower, and
many other plants. In 1988, Territorial Seed intro

duced the first Peters-bred vegetable variety to be released commercially. It is,

appropriately, a broccoli.

`Umpqua', named after the river in southwestern Oregon, is a very early, deep
green broccoli with especially large central heads. "In our trials," the Territorial
catalog declares proudly, "`Umpqua' has outperformed all other open-pollinated
varieties that we know of." And it's early - earlier than all other open-pollinated
varieties and as early as the earliest hybrids. (An open-pollinated variety will
come true from seed; a hybrid - the offspring of a cross between two true
breeding varieties - will not.)

It took more than just inbreeding from a single cross, however, to produce
`Umpqua'. Like most plants that have incompatibility mechanisms, broccoli
exhibits strong inbreeding depression. This means that the more genetically
uniform the plants are, the smaller and weaker they tend to be and the lower their
yield. Thus inbreeding, which is done to increase genetic uniformity, can be
counterproductive. Instead of working toward genetic uniformity, the breeder
tries to obtain uniformity for the important agricultural characteristics while
simultaneously maintaining enough genetic heterogeneity to produce vigorous
plants.

Tim started by crossing several hybrids with each other. Then he inbred the
progeny from each cross to develop lines and selected the type he wanted - very
early broccoli with medium height, big central heads, and good yield - within
each line. Next, he pooled all the lines and mass-selected - that is, he let all the
lines he had developed separately grow together and cross-pollinate. For several
generations thereafter, he selected for the type he wanted until his new open-
pollinated variety had become stable. It was a mess but a practical approach to
dealing with the plant that had become his priority.

Tomatoes are one of Tim's other favorites. Tomatoes are much easier to work
with than broccoli. Because tomatoes are almost totally selfpollinating, you don't
have to hand-pollinate to maintain a number of different varieties or to inbreed
from selected individuals. The plants inbreed themselves. They don't usually
show inbreeding depression, either. It's fairly easy to produce good inbred
varieties that perform as well as or better than the commercial hybrids.

With tomatoes Tim does a lot of what could be called "dehybridizing the
hybrids." "The professional breeders have put lots of good things into those
modern hybrids," Tim says. "I like those good things." So he tries to get the good
things out of the hybrids and into open-pollinated varieties. He might cross a
very disease-resistant hybrid, for example, with an heirloom variety that has
spectacular flavor, then inbreed and try to select a line that has the best
characteristics of both parents.

Tim often starts by saving seed from a good hybrid. The seed company has
already done the original cross between two different cultivars. Seed companies
usually keep the identities of the parents secret so that no one else can produce
the hybrid seeds. Customers often believe that they can't save hybrid seed, that
it's not viable. But you can save the seed from most hybrid vegetables.

The seed from Tim's hybrid broccoli failed not only because the parent was a
hybrid but

also because of the incompatibility system. The hybrid broccoli plants were
genetically identical to each other, like clones of one plant. The incompatibility
mechanism prevented the plants from pollinating themselves or from being
pollinated by genetically identical others. Tomatoes, corn, squash, melons,
cucumbers, and many other vegetables are self-compatible. Seed from their
hybrids is good; it just doesn't breed true. For a commercial grower, a row or
field of plants that are all individuals could be a disaster. But for the gardener, it
might represent opportunity.

For example, Tim has done a fair amount of work with `Early Girl Hybrid'
tomato, which does well in the Pacific Northwest. It's very early, it is resistant to
cracking and verticillium wilt, and it has good flavor. But the fruit isn't very big.
Tim saved and planted the seed from an `Early Girl' fruit. All the resulting plants
were pretty early, and all the tomatoes tasted good. There was a tremendous
range in fruit size, however, from almost cherry types to great big ones. He
saved the seed from plants with big tomatoes and has developed several lines
that have the virtues of `Early Girl' but are open-pollinated and bear big
beefsteak-type tomatoes.

Tim Peters had been an amateur vegetable breeder for more than a decade
when he read that Territorial Seed Company was looking for a vegetable breeder.
He could not imagine anything more wonderful than being paid to breed
vegetables. Although he had only a high school diploma, he screwed up his
courage and went to talk with the people at Territorial. He now handles the seed
company's trial and breeding program.

Glenn Drowns, Ewald Eliason, and Tim Peters all developed useful new
vegetable cultivars - ones that met their requirements better than anything else
available. You can do it, too. But where should you start? What kinds of things
are worth doing? Are you likely to make any money at it? Should you patent
your plant? And how do you get a new variety out where others can use it?

Many amateur breeders start by accident. They simply notice something unusual
and save the seed If you want to start deliberately, choose a vegetable you like to
eat; you could be eating a lot of it before you're done. The amateurs profiled in
this chapter all suggest that you limit yourself to one or two vegetables or
projects, but none of them did. You probably won't either, and it is perhaps just
as well. Not all projects work out.

John Gale, president of Stokes Seeds, has two suggestions based on his
experience with amateurs. First, he says, an unstable cross straight out of the
garden isn't useful, no matter how interesting the individual plant is. You have to
grow a new variety for several years and cull out the off types each generation.
Only after the variety is yielding all desired-type plants is it ready to offer to a
seed company.

Second, find out what is already available. There are tomatoes of almost every
color, for example - pink, orange, yellow, green, brown, purple, bronze, white.
They aren't popular, but they're certainly not new.

Rose Marie Nichols McGee, of Nichols Garden Nursery, concurs, and

mentions that she is especially unthrilled by the idea of a "new" white carrot.
"Carrots started white,"

she points out. "Breeders worked for a long time to get the beautiful orange
color. And wild carrots are white, and any accidental cross with them gives you a
white, so they turn up all the time. And anyway, we already carry white carrots."

There are a lot of things that Rose Marie does want. There has never been
much breeding work with herbs, she says. Almost anything that's different is
interesting. Foliage differences; variegation; color differences in plants, flowers,
or stems; differences in size, growing habit, or form; differences in flowering
characteristics or patterns; and differences in odor are all worth paying attention
to. Disease resistance also matters with herbs.

With vegetables, Rose Marie is interested in early maturity, compactness, and

disease resistance. If, for example, you have an eggplant that is doing nicely
when all the rest are dying from verticillium wilt, you might have something
there. "The sophisticated amateur is just real well positioned to notice these
things," Rose Marie says. "You may not know what the disease is, but you can
see one plant is behaving differently." (An Agricultural Extension Service agent
can probably help you identify the disease.)

"With vegetables," Rose Marie says, "we look for quality and also beauty.
Many people are gardening on small plots, and the attractiveness of their
vegetables as plants is an increasingly important part of the pleasure.

"Genetic accidents occur all the time," Rose Marie says. "They are not
uncommon. But what is important isn't just the plant. It's whether there is
someone around to notice - a person with a keen eye, with the ability to observe
and recognize what's special."

Seeds Blum carries many unusual vegetables for which no professional

breeding work is done. "Unless amateurs do it," Jan Blum notes, "it's not going
to be done." Is anyone interested in breeding sunroots (aka Jerusalem artichokes,
Helianthus tuberosa), for example, American groundnut (Apios americana), or
camases (Camassia species)? Camas is a native plant the American Indians used.
"There are lots of varieties," Jan says, "and they're good, and the plants are
beautiful. But very few people have played around with them."

Professionals usually spend most of their time working on crops that are
already important. Amateurs can afford to gamble. They need not justify their
efforts to superiors or grant agencies. Sunroots, groundnuts, and camases aren't
major vegetables now, but perhaps they could be if someone did the breeding
work.

With the common vegetables, Jan cares about flavor above all else. She also
loves vegetables that are used for particular dishes. "If you cherish a tomato
because it makes the very best lasagna, or whatever, that matters," she says.

Jan also has two suggestions for amateur plant breeders. First, like Rose Marie
Nichols McGee, she emphasizes that good things frequently come from
accidents - the spontaneous mutation, the accidental cross. "So many people,
when you talk about this sort of thing, will tell you about some wonderful thing
that showed up in the garden one year," Jan says. "But they didn't save the seed

from it. Sometimes I really ache when I hear about these things. You just know
you're not going to see that again." Make records of anything unusual in your
garden, Jan advises. Save the seed and grow some of it out next year.

Second, Jan emphasizes that selection, by itself without making crosses, is a

valid breeding method. People do crosses to generate the genetic heterogeneity
from which to select new types. But heirloom varieties often already contain a
lot of genetic heterogeneity. When they do, you can guide the variety toward
your own breeding goal just by culling out the plants you don't like each
generation and saving seed from the best. "Everyone who saves seed is a plant
breeder," Jan says. "They choose which genetic material to perpetuate."

What about selling seed of your new variety? And what about plant patenting?
Do you have to patent your new variety to be able to sell its seed? Can you
legally breed new varieties using a patented variety or a hybrid as one of the
parents? The answers to these questions depend partly on what country you are
in.

Breeding new garden vegetables is usually a matter of love, not money. In the
United States, plant variety protection for seed-propagated plants is provided for
by the U.S. Plant Variety Protection Act of 1970. This act isn't of much
relevance to those who develop home garden varieties, though. Licensing a
cultivar is too expensive - $2,000 for eighteen years. In most cases garden
varieties don't sell well enough to cover that, so even those developed by seed
companies are usu ally unprotected. Breeding them is not profitable, which is
one of the reasons amateur efforts are so important.

Some new vegetable varieties, especially those intended for commercial

production, may have enough profit potential to warrant patenting - if you
believe in it. Many people don't; they feel the patenting of plant varieties is
immoral. Germplasm, they feel, is the property of all humankind. I find the issue
more complicated than I might wish. An excellent coverage of all points of view
on plant patenting can be found in Jack Kloppenburg's Seeds and Sovereignty:
The Use and Control of Plant Genetic Resources (see bibliography).

Many amateur breeders do make money by growing and selling seed of their
new varieties without patenting them. Usually, the breeder simply serves as the
grower of the variety for the retail company that sells it. Breeders actually are
being paid for the growing, not the breeding, however, and would be paid just as
much for growing other varieties they hadn't developed themselves. The
alternative seed companies in the United States are often in need of additional
reliable growers. The work isn't usually highly lucrative, but it can provide a
modest return for an enjoyable activity.
 Although the Plant Variety Protection Act is usually of little benefit to amateur
breeders, it's also of little hindrance. You can grow your own seed from
protected varieties as long as you don't sell it. You can sell seed of your own
varieties without protecting them. You can use protected varieties in developing
your own. You are not allowed to use a protected variety to make hybrids to sell
com mercially, however, nor may you merely select from a protected variety
without doing any crosses. But you can cross a protected variety to something
else and then work with the progeny to develop your own variety. And you can
use hybrids in any way you want. They are protected by secrecy, not by law.

In the third world, plant patenting and variety protection isn't recognized.
There are usually no restrictions on using or selling seed.

In Europe, official or national lists profoundly restrict the activities of amateur

breeders and seed savers. To protect patented varieties, there are laws that
prevent the sale of everything that isn't on a special list. These laws are far more
restrictive than those that apply to all other intellectual property rights in Europe
or elsewhere.

It costs time and money to do the evaluations necessary to get a new variety on
the official list. You can afford to do it if you are a seed company with a new
patented commercial variety that is broadly adapted and is expected to sell in
large amounts. But the official-list policy is a barrier to amateurs in getting their
varieties recognized and made available. It doesn't prevent them from offering
their creations through seed saver exchanges, however. (The Seed Savers
Exchange in the United States welcomes foreign members, by the way.)

Once you've developed a new variety, you'll want to get it out where others can
use it. The easiest way is probably to offer your new creation through a regional
or national seed savers exchange.

The Abundant Life Seed Foundation is both a seed company and a nonprofit
foundation dedicated to preserving germplasm of the Pacific Northwest. It
handles native, naturalized, and open-pollinated domesticated varieties. Forest
Shomer, the director, says that the foundation would be happy to carry new
amateur-bred varieties that do well in the Northwest.

The Seed Savers Exchange preserves and distributes open-pollinated varieties

that are not available commercially. That includes the one you've just developed,
director Kent Whealy says.

The Seed Savers Exchange is a marvelous source of breeding material,

contacts, and information. With the exchange's Garden Seed Inventory in one
hand and its most recent annual Winter Yearbook in the other, amateur breeders
have descriptions of all varieties available either commercially or through
members. They can see what has been accomplished in their vegetables of
choice and can figure out whether anyone has ever developed anything like the
variety they're thinking about. In addition, members often list their projects and
interests, exchange information, and trial each other's varieties.

You don't have to live in the United States to join the Seed Savers Exchange.
Since small amounts of seed can be mailed across national borders, you can
participate no matter where in the world you live.

You may want to offer your new variety directly to a seed company. Most of
the smaller, family-owned, or alternative seed companies are happy to introduce
varieties bred by amateurs. In most cases you should

approach just one seed company. Trialing and introducing a new variety is work,
and the introducing company has no legal protection for a public-domain
cultivar. If a new introduction is successful, many companies will soon be
offering it. The introducing company receives no reward for its work - other than
the pride of being the first to offer something special.

By the time you've developed a good new variety, you'll often know which
seed company to offer it to. Look first to the companies from which you buy
seed. If you like what they're doing, they might like what you've done, too.
Territorial Seed Company was the logical choice for Tim Peters's new broccoli.
It is a regional seed company that sells only varieties that do well in the maritime
Northwest, and the company has trialed, introduced, and popularized many new
cultivars of the brassicas. Glenn Drowns and Ewald Eliason both met Jan Blum
through the Seed Savers Exchange. Her company specializes in open-pollinated
and heirloom varieties. Many of the newer or smaller seed companies were
founded specifically to serve home gardeners. Among them, any good new
garden variety is likely to find a home.
A century ago, seed catalogs were full of new introductions that had been bred
by amateurs. Today such contributions are rare. Yet there are not very many
professional vegetable breeders, and their work is focused largely on just a few
of the economically most important vegetables. In addition, professionals
usually must concentrate on varieties intended for commercial producers.
Commercial vegetable production is what feeds most people. We gardeners are
lucky the professionals develop anything for us at all. Fortunately, some
commercial varieties are excellent home garden varieties as well. Also, many
professionals deliberately spin off garden varieties, whenever possible, from
their commercial breeding programs. And sometimes they spend part of their
time working on varieties just for us, whether it is economically justifiable or
not. Most are gardeners themselves, or at least started out that way.

The heyday of the professional breeding of garden vegetables is over. It lasted

from about 1900 to about 1950. Before then, there was no genetics, so there was
no professional plant breeding, in the sense in which we use the phrase today.
During the first half of the century, as the basics of genetics came to be
understood, commercial vegetable production was still largely just a scaled-up
version of garden production. Most vegetables were sold near the farms where
they were grown. Commercial varieties often were similar or identical to home
garden ones. This was a golden age for the development of garden varieties.
Professional breeders, with their newfound knowledge of genetics, created much
of the best we gardeners have today. By 1950, however, most commercial
vegetables were mass-produced and such production was machine oriented. The
needs of commercial growers and home gardeners diverged. Since then the
contributions of professional breeders to gardening have dwindled.

Unfortunately, amateur vegetable breeding also has dwindled. It became

almost a lost art during the decades when the professionals were so active. Now
the professionals have largely left the field, and there is a vacuum. It's time for
amateurs to move back in and fill that gap.

Breeding plants is something every gardener did once. It's something every
gardener should do again. Developing your own vegetable varieties can add a
new dimension to your experience as a gardener and to your satisfaction.

 LENN DROWNS did his first watermelon cross in 1977. It took him
only a few minutes to guess how to cross watermelons and to do it, and that's all
he did that year. In 1978 all he did was to plant a few seeds from his cross
instead of a few more seeds of the standard varieties he was trying. That year
Glenn got his first ripe melon - from one of his own F, hybrid plants. It took
Glenn only a few minutes of work, no extra space, and only one year to produce
breeding material that performed better for him than anything that was
commercially available.

In subsequent years, Glenn inbred his variety. Hand-pollinating the flowers

required several hours per year. The breeding project still didn't require any extra
space, though, since he wanted to grow melons anyway and could both eat the
melons and keep the seed. It was a number of years before `Blacktail Mountain'
was ready for commercial release, but almost from the beginning it was doing
what Glenn wanted most - giving him tasty, fully ripe melons in a climate where
one wasn't supposed to be able to grow watermelons.

It took Ewald Eliason just a couple of years to develop his `Blossom' potato
variety. He made the cross one year and raised plants from the seed the next.
When fall came, there was `Blossom'. All he had to do was keep tubers from it.
Since tubers are actually part of the original plant, they are genetically identical
to it. For a vegetatively propagated plant, as soon as you find an individual you
like, you have an instant new variety.

The space cost of Ewald's breeding could be considered zero, since he grows
and eats a lot of potatoes anyway. He just grows some plants from his own
crosses instead of more plants from an established variety. The experiments,
even when they don't lead to something new and special, are still edible.
 When people say it took them ten years to develop a new variety, they are
usually including several years after the actual breeding was complete, during
which time the new variety was evaluated and increased for commercial release.
Ewald's `Blossom' took just two years to breed, but it was several years more
before it was ready for release. Ewald grew the potato for a number of years
before he felt confident that he had something new and special. Then Seeds
Blum needed to grow and evaluate it, as well as to increase the amount of seed
material for a few years, before the company had enough seed to be able to list
and offer it. Ewald was eating `Blossom' potatoes right from the beginning,
though.

It took ten years for professional plant breeder Calvin Lamborn, of Rogers
Brothers Seed Company, to develop the `Sugar Snap' pea. He actually had snap
peas, though, in the F, generation - in other words, two years after he did the
original cross.

Calvin made that first cross - a cross between a mutant with thick pods and the
snow pea `Mammoth Melting Sugar'- in 1969. He grew the seed from the cross,
the F, seed, in 1970. In 1971 he planted the FZ seed and examined and evaluated
the plants. A number of plants had thick flesh and round pods; the thick flesh
always went together with round pods. He saved the seed from individual round-
podded plants. In 1972 he grew out the seed and found some plants that had
what have come to be known as snap peas.

Calvin increased the seed material he had in the next few years and began
sending samples to seed companies. They weren't interested. No one knew what
to do with a round, edible-podded pea. In 1977 `Sugar Snap' won a gold medal
in the AllAmerican Selection competition. Suddenly everybody was interested.
There was not yet enough seed to introduce `Sugar Snap', so 1978 was spent
increasing the seed - that is, planting most of it and saving all the resulting seed
without additional genetic work.

A seed company will not usually list a new variety until there is enough seed
so that it can fill the orders that are likely to result. This may require 10 pounds
of a new pea or bean for a small company, or hundreds or thousands of pounds
for a larger one. AllAmerican Selection winners are not released for introduction
until there is enough seed to satisfy every company that wants it. Sometimes a
number of generations of seed increase is required before enough seed is
available.

`Sugar Snap' was introduced in 1979. The rest is history. Ten years. Ten years
of Calvin Lamborn's life? No, not at all. Developing `Sugar Snap' was only one
of many breeding projects he had, and, from the viewpoint of an amateur, the
project had largely succeeded by year 3 or 4, because an amateur would be
eating the new pea that year while continuing to refine it and increase the seed.
The amateur would probably then provide a modest seed sample to a seed
company or offer seed through a seed savers exchange. Others would take care
of increasing the seed for commercial release.

Developing your own varieties may not require nearly as many years as is
generally stated, when looked at from the perspective of the amateur. It does,
however, require time. Ewald Eliason, for example, needs time for making his
potato crosses. Making the crosses is easy, but he grows about fifty plants from
the seed of each cross, and it takes more time to grow potato plants from true
seed and transplant than to plant tuber pieces directly in the field. In addition,
while each true seed leads to an instant new variety, most varieties aren't
anything special. It takes time to compile records on each plant, evaluate all the
potatoes, and decide which ones are worth increasing and maintaining. Ewald
doesn't do just one cross either; he does a lot. But, then, he can afford to. He's
retired. He has the time. And he loves his potato hobby.

You don't have to have a whole field of potatoes to breed them, however. You
can choose the scale that you want to operate at. You don't have to make dozens
of crosses. You can make just one. You don't have to grow fifty plants from each
cross. You can grow five or ten - or one. With smaller numbers, you might not be
as likely to find something worth keeping, but you never can tell. Glenn Drowns
made only one watermelon cross and planted just a few seeds from it - and look
what happened!

Some plants are much smaller than potatoes, and you can raise and evaluate
hundreds in a modest garden bed. Others are big and space greedy. Even if your
garden space is modest, you need not limit your breeding projects to just the
smaller plants. There are tricks you can use when you need to grow large
numbers of big plants.

Here is a procedure that Tim Peters recommends for tomatoes: Plant about
fifty seeds of your first generation in a half-gallon pot of soil. When they come
up, select for early emergence and seedling vigor, and eliminate all but about ten.
As the plants grow, remove all suckers and branches, and limit each plant to just
one vine. In addition, you may need to prune the leaves - that is, remove all the
older, bigger leaves to keep the plants from shading each other out. When the
plants start flowering and have set a few fruits, remove the tops. Then continue
to remove all suckers and branches. Pay careful attention to soil fertility. Use
good soil initially and fertilize as necessary.

The plants will be stunted compared to field-grown ones, flowering will be

late, and the fruit will be small. But they ripen. You can determine fruit colors
and shapes. You can tell whether the plants are determinate (shorter bush types)
or indeterminate (continual-bearing vine types). You can tell earliness relative to
other plants grown similarly. And you can select for disease resistance. For the
latter, just shred some infected plant material on top of the soil in the pot.

During the second year of your breeding project, plant fifty seeds from each of
your original ten plants in separate half-gallon pots. Thin, grow, prune, and
evaluate as before. This time, though, save the seed from just one plant per pot,
because you're going to use only ten pots per year. In subsequent years, choose
just one plant per pot for inbreeding. In this way you can inbreed ten different
lines out of a cross with just 5 gallons of soil.

You can use this pot method with peppers also, Tim says, and with some of the
brassicas. It doesn't work with corn, though, or with the vine crops. They need
more space to ripen a fruit at all.

Tim has other tricks for melons, squash, and cucumbers. For example, plant
the seeds in a bed, leaving just 6 inches between them in all directions. Then
proceed as for tomatoes. Limit the plants to one vine, remove all older, bigger
leaves to prevent overshading, and eliminate the growing tip after the first fruits
start to form. You'll be able to select for fruit colors and shapes, earliness,
disease resistance, and many other characteristics - but not, of course, for fruit
size or yield.

How many years a project takes and how much space and time it requires
depends on you and how much you want to do, but it also depends on the plant
you're working with and on the project. Some projects will be impossible no
matter how much space and time you have. Some will be possible but might
require more space (or time) than you're willing to put into them. Others might
be relatively fast and easy.

If you are interested in producing your own F, hybrids, for example, you can
make crosses between different heirloom or openpollinated varieties your first
year and trial the hybrids the second. If any of the plants is better than anything
currently available, your project is a success as of year 2. From then on you just
maintain a few plants of the parent variety so you can continue to produce the
hybrid. Then most of your planting will be of your own hybrid.

In most (but not all) cases, when you cross two different edible cultivars of a
vegetable, all the offspring will produce edible vegetables. The fact that you can
eat your experiments cuts down on the amount of extra space required. But not
all experimental material is edible. If you cross an edible cultivar with a wild
relative, for example, the first few generations of plants may not be edible. In
such cases space costs can mount rapidly. Even when both parents are edible,
most of the offspring may be so inferior that you need to plant a larger space to
get the same amount of food. Developing a useful new variety may be easy, but
there's no guarantee that it will be.

Even when you are working with thoroughly domesticated varieties, you can't
always eat and breed with the same plants. Some of my own breeding work
involves peas, and you cannot, alas, eat a pea and save it too. In addition, peas
must be picked to be kept in production. So when you leave pods on the plant to
let seed develop, the entire plant goes out of production.

Worse yet, I have found that if I try to eat and breed with the same plants, I
always end up eating many of my crosses by accident. The little tags that are
best for marking pea crosses blow around in the wind and get lost in the foliage.
They're obvious enough when you're collecting all the dry pods, but on the live
plant they are easy to miss.

So with peas, at least, breeding definitely requires some additional space above
and beyond what you would plant just for the kitchen. Fortunately, peas are
small plants, and they like to be crowded. I did the first three years of three
different pea breeding projects with about 2 feet of row the first year, about 6
feet the second year, and about 20 feet the third. That was actually for five or six
projects initially, which got trimmed to three by the third year.

Even major projects involving big, spacegreedy plants may not require much
space or time for the first year or so. After the second or third year, you often
have an idea of whether the project is going to work. If it looks good but is going
to take more space or time than you have, you can often borrow land or enlist
help for the critical year or two.

Tim Peters didn't have separate, isolated fields to grow his broccoli plants. But
he could get the same effect by hauling his potted plants around on his bicycle
and recruiting the assistance of his neighbors. Glenn Drowns has often used
fields of abandoned farms for isolation fields; he receives permission to use the
land gratis. Alan Kapuler, whose Peace Seeds is a private plant introduction
station and germplasm collection, evaluates and introduces rare plants, grows
and sells hundreds of heirloom varieties, and conducts dozens of his own
breeding projects. Until 1990 he did it all on borrowed land.

Until recently I had only a few hundred square feet to grow and breed
vegetables. Now I have access to more land than I can use - both garden plots
and entire farm fields - made available to me at no charge by gardeners and
farmers in Corvallis (Ore

gon, where I live) and outlying areas. Developing new varieties of garden
vegetables fascinates most people, and it's a public service. Many people will
help if you explain what you're doing and ask for their help.

Some projects do require huge amounts of space and time. A professional

breeder may need to consider a dozen characteristics of a plant at once and to
grow thousands or even hundreds of thousands of plants to have much chance of
finding even one that meets all the requirements. Often the professional is
working on some major, important commercial crop plant that has been worked
on extensively by hundreds of other professionals who were breeding toward
similar goals. In such cases, to make even a slight improvement in an established
variety might be a major undertaking.

We amateurs will often be working with plants that have received little or no
attention before, so we will often find we can make major improvements quickly
and with just a little work. Even when we work with the common vegetables, we
usually have goals that are different from those of the professionals. Potatoes are
one of the most worked on vegetables. But Ewald Eliason was interested in
potatoes with colored flesh that tasted especially good and did well under
organic gardening conditions.

Commercial cultivars of major crops are difficult to breed in part because it is

usually not profitable to produce the seed unless it can be grown and distributed
in huge amounts. This isn't possible unless the variety has very broad regional
adaptability. That is, a good commercial variety needs to per form well under a
wide range of climates and growing conditions. Amateurs, though, need not be
concerned about broad regional adaptability. They care most about truly
excellent performance in their own specific regions. If a variety also has broad
regional adaptability, that's fine, but it's not essential.

An amateur creation does not need to be of agribusiness significance to be

valuable. If an excellent variety is suitable for only a limited region or for
specialized production methods or uses, it will probably never be carried by the
largest seed companies. But amateurs don't care about that. They can maintain
the seed themselves. They can proudly pass it around to all their friends. They
can offer it through seed saver exchanges. In addition, many alternative, family-
owned, or regional seed companies are eager to introduce unique, regionally
adapted varieties.

A final reason why we amateurs can produce useful new varieties so much
more easily than one might expect, given the heroic efforts sometimes required
of professionals, is that we usually aren't as fussy about what we're after. We
have the luxury of being versatile, of being, even, fickle. We don't have to
produce exactly what we intended to produce. If need be, we can change tracks.

In many cases all we really want is something different, something new and
interesting. And it's easy to get something interesting.

 YOU grow vegetables that have been bred by others, you are
limited by the priorities and the vision of others. When you do the breeding, it is
your vision that matters, and your priorities. 'HEN

The quest for flavor is a big part of what home gardening is all about. Flavor is
often low on professional vegetable breeders' list of priorities, but for the
amateur, flavor is usually somewhere near the top of the list. You can do more
than just breed for good flavor; you can breed for exactly the flavor that you
prefer, flavor that develops under your growing conditions and climate, flavor
relevant to the exact way in which you use the vegetable.

Flavor is a complicated characteristic that

is made all the more complicated because evaluating it is so subjective. My idea

of perfect tomato flavor, for example, is a vineripened `Early Girl Hybrid'
tomato. I like tomatoes that taste both sweet and tart and that have a lot of flavor
and aroma in every bite. Many people prefer mildflavored tomatoes such as the
beefsteak types. When two different people talk about good flavor, they are often
talking about different things.

Many professional breeders work on tomatoes, but they seem to be

concentrating on producing what are called low-acid types. Low-acid tomatoes
may or may not have less acid than normal tomatoes, but they taste as if they do.
They're what I consider mild flavored. When professional breeders do breed for
flavor, they breed for what tastes good to them or to some average consumer. I
am not that average consumer, and what I want is tomatoes that taste good to me.

Flavor is profoundly affected by growing conditions and climate. Part of what

makes `Early Girl Hybrid' so popular in maritime Oregon is that it is able to
develop its full sweet-tart flavor even with cool spring and summer days and
downright cold spring and summer nights. Many tomatoes that are considered
especially flavorful elsewhere either don't ripen at all or ripen but are tasteless.
There are many professional tomato breeders, but only a few operate in areas
where temperatures fall into the forties (°F) at night all the way through the
summer. Whenever you breed for flavor, you are also automatically selecting for
regional adaptation.

Professional breeders also usually breed for tomatoes with a lot of flesh and
little juice. Such tomatoes ship much better than juicy tomatoes. They are also
better for using as slices on a hamburger or in sandwiches; they don't make the
bread soggy.

However, the acids and many other flavor components of tomatoes are mostly
in the juice. I have yet to find any meaty tomato whose flavor I like as well as
that of `Early Girl'. And I don't care that a slice of an `Early Girl' can completely
liquefy a sandwich, because I don't eat many tomatoes in sandwiches. I eat them
like the fruit they are, whole and out of hand. The juicy `Early Girl' often drips
or squirts juice on me, however, and it would be better if the tomato were
smaller so I could get my mouth around it better.

What I want is a `Baby Girl' tomato. It would have the juiciness, tenderness,
and full sweet-tart flavor, as well as the earliness and cold resistance, of `Early
Girl'. But it would be smaller - perhaps even of cherry tomato size.

So far every cherry tomato I've tasted has been meaty (so they travel well) and
either bland or sweet but lacking in tartness. Cherry tomatoes are convenient to
eat, but they just don't taste as good as `Early Girl'. My `Baby Girl' would be
commercially useless, of course, because it wouldn't travel well. But it wouldn't
have to travel any farther than the distance from the vine to my mouth.

Like every gardener who has grown any specific crop for a while, I have my
own individual priorities and preferences - my "wish list" of the characteristics
of the perfect variety. Your ideas and your wish list are no doubt different from
mine. Tim Peters, when he bred from `Early Girl Hybrid', was selecting for
larger, not smaller, size and for meaty tomatoes, not juicy ones.
 There are a lot of professional tomato breeders in the world, but I can't count
on any of them to want exactly what I do in the perfect tomato. In other words, if
I want my `Baby Girl', I'd better breed it myself.

I have saved the seed from `Early Girl Hybrid', and each year I will plant a few
seeds from that material in addition to the standard `Early Girl'. I also have
raided Tim Peters's trash can. I want an `Early Girl' derivative that is almost the
opposite of what he wanted. What did he do with all the material he eliminated
during his breeding project? I asked him.

He had, as I suspected, saved it all, and he was quite happy to send me seed of
lines from his inbreeding that he had discarded because the tomatoes were too
small. Next spring I will grow some plants from that seed in place of my normal
varieties and screen them for flavor. If one of them is suitable, it will be a
number of generations closer to being a stable line than material straight out of
`Early Girl'. I think it's a good idea to save all the seed produced in a breeding
project. Even if it doesn't meet your needs, it might, as in this case, give a head
start to someone else who has different breeding goals.

I will probably also try some crosses of `Early Girl' with varieties that produce
smaller tomatoes. None of this will take any extra space, since I usually raise a
number of varieties anyway and can eat all the tomatoes except the few involved
in the crosses.

Flavor is dependent on many genes and how they interact with each other and
with the environment. It is what we call a quantitative genetic trait - that is, a
trait affected by many genes instead of just one or a few. Quantitative traits are
not especially amenable to any of the new biochemical breeding approaches of
genetic engineering. They are very amenable to traditional plant breeding
methods - the methods that are fully accessible to the average gardener or
farmer.

Other quantitative traits include size, yield, shapes and proportions of various
parts of the plant, tenderness, earliness, lateness, cold hardiness, heat tolerance,
regional adaptation, yield - in fact, most of the essential agriculturally important
characteristics. Specific methods for breeding for quantitative as well as single-
gene traits are covered in later chapters.
Selecting an established vegetable for other than the established use is one of the
major ways new kinds of vegetables are developed. Professional breeders focus
their attention on the most common uses of vegetables. When you do the
breeding, you can focus on your uses.

Consider the lowly cabbage. It started as a plant that was grown for its edible
loose leaves. Somewhere along the line, various people - amateurs all - decided
they cared more about the central buds, the flower bud, the side buds, or the
central stalk, and they bred for plants that excelled at producing those. So we
have head cabbage, cauliflower, broccoli, Brussels sprouts, marrowstem kale,
and kohlrabi, as well as collards and kale.

If you eat a bean as a snap bean, perhaps other breeders will satisfy your
needs. Many professional breeders work on snap beans. But what if you prefer to
use beans as shellies (shelled green seed) instead of eating them earlier as pods?
I doubt that any professional breeder in the world is trying to develop better
shelly beans.

And what if you prefer pole beans to bush types? American bean breeders
have been concentrating on bush beans suitable for harvest by machine. For
machine harvest, the pods should ripen simultaneously; should be held, if
possible, above the leaves; and should be moderately tough - "firm" it's called.
The homegrown bean can be as ten der as possible. The commercial bean needs
to have significant amounts of fiber in it so that it resists damage during machine
harvest and shipping.

Flavor matters to the professionals, too, but it is usually not the top priority. If
a bean bruises and then rots before it gets to the consumer, it doesn't matter how
great it would have tasted had someone eaten it.

You might eat a vegetable raw, whereas others cook it, or vice versa.
Cucumbers are a popular vegetable, and many American professionals work
with them. But how many of them recognize the cuke as an excellent cooked
vegetable and are breeding for better cooking cukes? You may eat a vegetable
very young, whereas others eat it at a different stage. Or perhaps you like the
stalk of the broccoli better than the top and would prefer more stalk and less top.
You might like stuffed squash flowers more than squash and want a variety that
produces a lot of huge flowers and doesn't waste its energy setting many fruit.
 I grew scorzonera (Scorzonera hispanica) for the first time a few years ago.
Scorzonera, or black salsify, is a root crop. It's a perennial that is grown as an
annual. I had only a few plants at the end of the season - not enough to bother
harvesting them and figuring out how to use them - so I just left them in the
garden.

The leaves of the scorzonera plants were long and spearhead shaped. They
lasted past the mild freezes and regrew quickly after the hard ones. That's when I
noticed them and tasted them - and found that they have an excellent, mild
flavor. They taste so much like lettuce that I think if you told someone

they were eating lettuce, they would believe it readily.

When spring came, the scorzonera plants produced copious quantities of large,
mildflavored leaves long before the lettuce could even be planted. I stopped
buying lettuce and filled my salads with "lettucesalsify." Later in the summer the
plants bolted and produced only small leaves, but by then lettuce was available.

Naturally, I am working further with this lovely perennial. I'll start by letting
the original plants get bigger and seeing how they behave. The real
characteristics and virtues of a perennial vegetable often don't appear until the
plants are three years old or so. I'll also obtain and screen additional varieties for
the size, quality, and cold hardiness of their leaves. It will probably be possible
to lift the roots in fall and force them inside as well. And who knows? One of
these days I might even get around to eating the roots.

Scorzonera is a minor crop. Professional plant breeders are doing little or no

breeding work with it. I know of only one group working on it (in Europe), and
they have been working with it as a root crop. In addition, part of what I would
like is a plant that holds its leaves throughout the winter - a characteristic that is
probably possible or advantageous only in mild-winter areas. If I want a better
perennial lettucesalsify, a salsify ideally adapted to my climate and for my
purposes, I'd better get busy and develop it myself.

I may not need to breed a better lettucesalsify, though. Perhaps all I need to do
is explore and better understand the varieties that are already available.
Developing new vegetable varieties includes not just plant breeding but also
plant exploration, plant introduction, variety trials, kitchen trials, and
experimental cooking. They all go hand in hand. They are all fun. The
adventuresome gardener should be ready to do any or all of them - whatever a
specific project requires.

Breeding for better regional adaptation of plants, including extending the

boundaries of the areas in which specific vegetables can be grown, is of great
interest to many amateurs and is very amenable to the breeding methods
available to gardeners and farmers. Glenn Drowns wanted a watermelon that
would grow in the mountains of Idaho, not a watermelon that would do better in
the South, where most of the watermelon breeders are located.

Ken Allan, who lives in Ontario, Canada, is inordinately fond of sweet

potatoes. The sweet potato is an important world vegetable, and many
professionals work with it - but not in Canada. Yet sweet potatoes grow well for
Ken. Some varieties have shorter seasons, but other, later types grow better when
it's cold. So Ken is crossing short-season types with cold-growing types and
plans to select for a short-season type that also grows well in the cold. In other
words, he is breeding for a Canadian-adapted sweet potato. (For a description of
his work, see The Vegetable Gardening Research Exchange, which is listed in
Appendix D.)

Professional breeders for any specific vegetable tend to be concentrated in the

areas where the vegetable is most important as a commercial crop - that is,
where it already

does well. Professionals also are concentrated in temperate regions and work on
vegetables that are important in those regions. Many vegetables that thrive in the
tropics receive little attention. Yet vegetables are perhaps more important in
tropical than in temperate zones. In many areas of the tropics, starchy
vegetables, not grains, are the most important source of calories, the difference
between starving and thriving for millions. There is a lot of room for amateurs to
do important work with plants such as cassavas, sweet potatoes, and yams.

Other crops may be less important on a world basis but crucial in certain
regions. Plantain (starchy) bananas are a major source of calories in some lands.
Achira (Canna edulis), arracacha (Arracacia xanthorrhiza), oca (Oxalis
tuberosa), ulluco (Ullucus tuberosa), and mashua (Tropaeolum tuberosum) are
important root crops in certain areas of South America. (See the discussion of
the book Lost Crops of the Incas later in this chapter.) Amateurs who live where
there are special local crops are in an excellent position to make important
contributions to their regional food base.

You don't have to live in the tropics to make an important contribution to your
regional food base. Professional plant breeders are severely limited in the
amount of time and effort they are able to devote to developing regionally
adapted crops or cultivars. Large seed companies can usually afford to carry only
varieties that sell in large amounts; this usually means cultivars that are widely
adapted. Such cultivars often don't do quite as well anywhere as the varieties that
are fine-tuned to specific areas. Every region is unique and offers unique
opportunities to the imaginative amateur.

Fava beans, for example, thrive in the maritime Northwest. Favas, also known
as faba beans or broad beans, are Viciafaba; common green beans are Phaseolus
vulgaris. Favas love mild, wet winters and cool summers. Only recently have
people started to realize that favas could be a major crop here. The greens, the
green pods, the green shelled beans, and the dried beans are all edible. The dried
beans can be used in soup or turned into tofu or miso, just like soybeans, which
don't grow well here. In fact, fava beans could become the soybean of the
Northwest.

All the work on favas for the maritime Northwest has been done by amateur
plant explorers and breeders and by people associated with the small, alternative
seed companies. Alan Kapuler, of Peace Seeds, started obtaining and trialing
favas about fifteen years ago. He currently offers about a dozen different
varieties that do well in the Northwest.

Steve Solomon, founder of Territorial Seed Company, trialed largeseeded favas

and introduced `Aquadulce Claudia', a largeseeded, good-flavored bean that can
overwinter here. The Territorial line is derived from two plants in a field of
`Aquadulce Claudia' that survived one winter when everything else died. More
recently, Steve has been working with green-manure and smallseeded types. He
has developed a good-flavored small type, he says, which will soon be
introduced under the name `Sweet Lorane'.

Dan Borman (Meristem Plants), who lives in maritime Washington, is

especially inter
ested in favas that will overwinter - that can be planted in fall and used for
greens in the winter and spring as well as for green and dried seed later. He lives
on Orcas Island, Washington, which is actually north of the Canadian border. He
is screening varieties for overwintering ability and for flavorful greens, and he
has worked out methods for making miso from favas as well. I've tried the miso,
and it's excellent. Fava beans also can be roasted to make a rich, full-flavored,
caffeine-free coffeelike beverage.

Farmers in the Skagit Valley of Washington and in southern Oregon have

developed methods for large-scale planting and harvesting of favas, alone or in
combination with grain crops. Seed size is so different that grain and fava bean
seed can be harvested (combined) together and sorted mechanically.

My favorite fava is `Aprovecho Select', which was bred by lanto Evans, an

architect by training. He grew up in Wales, where favas are a common vegetable.
When lanto arrived in Oregon, he looked around and asked, "Where are the fava
beans?" The climate in western Oregon is quite similar to what he was used to.
Fava beans ought to do well here, he thought.

He started collecting favas - from Britain and from wherever his various
travels took him. He has collected many varieties from Latin America that were
brought there by Europeans and grown in relative isolation. They may no longer
exist in their original homelands.

lanto bred `Aprovecho Select' by selecting from a largeseeded European type

he bought from a local feed store. He selected for large bean size, four or more
seeds per pod, and good yield in the Northwest. He did not particularly select for
flavor. The superb, fullbodied, rich flavor of the green shelled `Aprovecho
Select' is just one of those happy accidents.

lanto put together an organization of people, the Fava Project, which publishes
a newsletter on favas and fava research and runs an annual workshop for people
interested in growing, researching, developing, processing, commercializing, or
just eating favas. He also grows and sells about forty varieties. He and other
members of the Fava Project continue to collect new germplasm, develop new
varieties, and spread beans and information around to others. Those interested in
the newsletter or workshop or in buying favas can contact lanto at the address
given in Appendix D.
 Peace Seeds also carries an extensive line of favas, including `Aprovecho
Large', which Alan Kapuler developed from lanto's variety by further selection.

You may grow a vegetable that is established in your area, but perhaps you grow
it in a different way than it is normally grown. If so, you may want to select for a
variety that does well under your growing methods. Or you may end up selecting
such a variety accidentally.

I started breeding pole peas about four years ago. Pole peas are indeterminate
varieties; bush peas are determinate, or, at least, more determinate. Pole varieties
such as `Sugar Snap' grow to about 6 feet high in most areas of the country. Bush
types such as `Oregon Sugar Pod II' grow to 2 to 3 feet. Here in maritime
Oregon, the peas all seem to grow about twice as tall as they do elsewhere.
`Sugar Snap' never quits at 6 feet; it reaches 10 feet readily if the trellis is tall
enough. `Oregon Sugar Pod II' is a handy 4 feet or so.

All I wanted, initially, were disease-resistant pole snap peas. `Sugar Snap', the
classic snap pea, is a wonderful cultivar. It's a tall, vigorous pole type with two
pods per node, and the peas are large and of excellent flavor. But `Sugar Snap' is
susceptible to both pea enation mosaic virus and powdery mildew. Powdery
mildew is common in most areas of the country. Pea enation mosaic virus is
most common in western Oregon and Washington and is of little relevance
elsewhere. Both diseases are so common here that one or the other always cuts
production short in early summer. A variety resistant to both diseases might be
able to produce all the way through our cool summers and, possibly, could be
planted in summer for fall production. In other words, disease-resistant varieties
might double or triple the length of our snap pea season.

One of the major accomplishments of modern professional pea breeders has

been to incorporate genetic resistance to most of the major diseases into many of
the most important commercial varieties. The commercial varieties, however, are
bush varieties. I prefer pole types. They yield so much more per planted area.
They produce bigger pods, and more of them, week after week after week - not a
handful of dinky little pods all at once. In addition, the pods of pole types are
clean and intact and can be used without washing. Bush types produce pods that,
under my conditions, are muddy and slugeaten and that often require
considerable culling, washing, and trimming.
 Professional breeders do little work with pole peas. Jim Baggett, at Oregon
State University, is one of the foremost pea breeders in the world, and he has
developed shelling and snow peas resistant to enation and powdery mildew. He
is also working with snap peas. But he isn't doing anything with pole types. If I
want to be able to grow the pea of my dreams, I'll have to develop it myself. And
so I am.

I started by crossing `Sugar Snap' with `Oregon Sugar Pod II'. `Oregon Sugar
Pod II' is a dwarf snow pea bred by Jim Baggett that is resistant to verticillium
wilt, enation, and powdery mildew. It bears two large pods per node, has
excellent flavor, and can be planted so as to be harvested throughout the summer
and into the fall.

Things went relatively smoothly. I dug back into the professional literature to
find out how to cross peas. The project took only a few inches of row the first
year and a foot or two of row the second year (to grow the F1 plants).

The plants of the first generation after the cross (F,) were all tall plants and had
snowtype peas. This was what I had expected. The genes responsible for height
and basic pod type are well known, so I was better able to make predictions for
peas than when working with many other vegetables. (The genetics of pea height
and pod type are covered in the section on peas in appendix A.) In the second
generation after the cross (F2), some plants were short but most were tall; some
plants had snap peas but most had snow peas. This also was as expected. Some
plants had only one pod per node, however, instead of two. And some plants
were earlier than others. All the peas, snow or snap, had excellent flavor.

Seven of the plants were the kind I was looking for - tall, vigorous pole-type
plants with snap-type pods. I saved the seed from each of them so as to start
seven lines. I am in the process of increasing the seed from each of the lines and
evaluating and selecting for disease resistance.

The project as described so far went as expected. But an extra dimension

developed because of the way I raise my peas.

The normal way to raise peas here is to plant them in early spring. For some
years, though, I have been experimenting with fall planting, trialing various
varieties and establishing planting dates and methods. I regularly plant `Oregon
Sugar Pod II' and `Sugar Snap' in the fall. Both varieties overwinter well here.

To overwinter garden peas, I plant them in midNovember, the same time

`Austrian' field peas are planted here. It is important to plant them late enough so
that by the time the peas emerge, the first early frosts have killed all the aphids.
Aphids are sucking insects that don't seem to hurt the plants much themselves
but that transmit enation and many other viral diseases. The plants freeze solid,
with or without snow cover, off and on during the winter. Whenever it is above
freezing, the plants grow. The overwintered plants start flowering and yielding
peas much earlier in the spring than springplanted peas.

Naturally, I did my breeding work with fallplanted, overwintered material. I

had started the cross with parents that were especially winter hardy, and as I
selected, the less winter-hardy peas were automatically eliminated. By the third
year of the project, I already had material that overwintered better than either of
the parents - peas that grow happily whenever it is above freezing, that can hold
up 6 inches of snow for two weeks without the stems breaking, that can even
withstand freezing rain, and that can be completely covered with ice without
being damaged.

Freeze tolerance and cold tolerance are complex characteristics whose genetics
are not understood. I did know, however, that varieties of vegetables that are
especially cold tolerant are often especially heat tolerant as well, and vice versa.
Often, apparently, it isn't cold tolerance from the plant's point of view; it's
tolerance toward temperature extremes, perhaps even general tolerance toward
environmental stress. So I hoped that my extra-cold-tolerant peas might be
extraheat-tolerant as well - which would make them all that much more able to
last into summer and fall.

I made a midsummer planting of my pea lines in a field in Independence,

Oregon, to see how they would do and to select specifically for summer
performance. Some peas in my lines grew vigorously and yielded peas all
summer. I ate a few of the snap peas, just to have done it, to have eaten Oregon-
grown snap peas in the summer. I saved the seed from the best plants in late fall
and then forgot them for a while. I expected them to get knocked out by the first
freeze, which was imminent.

But they weren't. On November 9, after several freezes had occurred, I

rediscovered the peas. They were still there, and many of the plants were still
flowering and bearing peas. The peas were a little discolored, but they were still
delicious.

These plants are apparently able to take freezing weather at all stages of the
life cycle. That is something I did not, in my wildest dreams, ever imagine or
expect. All I was trying to do was to breed a disease-resistant pole snap pea - a
relatively straightforward genetic exercise. But I was selecting under my
growing conditions, which are unique, and doing so produced a unique result.

The winter of 1990-91 was the worst winter for plants that we've had in the
twelve years I've lived in Oregon. We had long periods of below-freezing
weather, with temperatures frequently down into single digits (°F). Worse yet,
there was at best only an inch or so of snow cover, and often none at all. Many
plants that I never lost before were decimated. Most of my chickpeas, however,
did fine - and they were growing in a wind swept farm field, not in a pampered
garden. They are, as far as I know, the first chickpeas to be overwintered in the
Pacific Northwest.

Chickpea is the English word for the bean and plant of Cicer arietinum. Many
Americans recognize them more readily by the Spanish name garbanzo or
garbanzo bean. In India and Pakistan and throughout the Middle East, chickpeas
are a major grain legume. They come in all sizes, shapes, and colors. In the
United States, only the large creamcolored ones are used, and mostly only in
salads. It's not obvious why.

Although the major uses are of the dry grain, the young green seeds and pods
also can be used as a vegetable, as can the growing tips of the branches.
Chickpeas also produce an acidic exudate that coats the plants and stems. It can
be collected on cloths under the plants or washed off of them and used to make a
drink similar to lemonade. The plants are a foot or two high, 1 to 3 feet across,
and of various shapes depending on the variety. They have lacy foliage that is
quite attractive.

Chickpeas fit especially well into sustainable, low-input agricultural systems.

They produce a tasty, high-protein seed. The seed doesn't shatter; it stays in the
pods until threshing. Threshing is easily achieved on a large scale with standard
combines or on a small scale with the methods used for dry beans.
 Chickpeas, being nitrogen-fixing legumes, don't require high-nitrogen
fertilizers. In fact, they will not yield as well if the soil is too high in nitrogen. In
addition, chickpeas are unusually well endowed with natural defenses against
pests. The leaves and stems are fuzzy and covered with the sticky, acidic
exudate. Insects and slugs seem to find the situation unappetizing. Friends tell
me that deer and rabbits don't like the plants either.

I got interested in chickpeas by accident. It all started when I read Lost Crops
of the Incas: Little-Known Plants of the Andes with Promise for Worldwide
Cultivation, a book published in 1989 by the National Research Council of the
National Academy of Sciences. One of the crops described was nunas, which are
a popping bean. They are the same species as ordinary beans (Phaseolus
vulgaris), but they are varieties that pop. Unfortunately, they are also short-day
plants - that is, they don't flower until day length drops below a certain level.
This means that in the United States, they flower, if at all, only in late fall. In
addition, they also need cool, moderate weather to flower. The only place in the
United States that one would expect to be able to grow them and get them to
flower outdoors would be the maritime Northwest.

The nunas would probably flower here, I thought, but not until about
November. Sometimes we don't have freezing weather until December. The
nunas might make it in a mild year, at least well enough to raise a few plants and
do some crosses. Perhaps I could transfer the genes for a temperateadapted
flowering pattern into the nunas (using recurrent backcrossing, a method I'll
discuss later).

So I called the Western Regional Plant Introduction Station at Pullman,

Washington, which handles the U.S. Department of Agriculture (USDA)
Phaseolus vulgaris germplasm collection, and found myself talking to Rich
Hannan, the USDA horticulturist who is the curator for the beans. He also, as it
turned out, handles the chickpeas.

Rich was willing to send me some nunas, but he had had a lot of requests for
them since that book came out; what he wanted to talk about was chickpeas.
There are popping varieties of chickpeas, too, he said. A friend of his who had
lived in Iran told him that in the Middle East, some types are popped in hot sand,
just like nunas. "Why is everybody interested in popping nunas, which don't
grow here," Rich asked, "and nobody is interested in popping chickpeas, which
do?"

"Because I didn't know that," I responded cheerfully. "They weren't in the

book. Which varieties pop?"

He didn't know. He had about four thousand accessions. An accession

represents a collection event, an acquiring. If a plant explorer buys a pound of
seed from a single basket in a market in Iran, for example, that seed and all its
progeny are an accession. An accession may be the same as a variety, or it may
not; it depends on whether the farmer had one variety in his field or a mix. (In
the USDA-ARS National Plant Germplasm System, accessions have numbers
that start with PI, for plant introduction. I discuss how to obtain germplasm from
the USDA system later.)

Rich had a wealth of germplasm and information about where it was collected
but no information at all on what it was good for. He had tried to get people
interested in popping chickpeas, but to no avail. He talked about them with
everyone who called him about the nunas. He had himself tried to pop some
chickpeas in hot sand, but he had chipped a tooth on them. Undaunted, he tried
two more types and broke a second tooth. He decided he needed someone else to
try the rest.

I didn't think I could handle four thousand accessions of chickpeas. We

compromised. Rich sent me the nunas I had asked for along with twenty
accessions of chickpeas so I could test them and see if any of them popped.

Rich and I made some guesses about what kind of chickpeas might pop. That
first batch were all # 2s. (The smallest size is #I - those that weigh less than 10
grams per 100 seeds. The #2s weigh 10 to 19 grams per 100 seeds. They're about
the same size as popcorn. The largest size is #5 - greater than 40 grams per 100
seeds.) Most were black, but some were brown or cream.

I didn't happen to have any hot sand lying around. I also wasn't too enthusiastic
about the idea of eating sandy chickpeas. So I did my popping the same way I
pop popcorn - in a stainless steel skillet (which can be heated much hotter than
cast iron without smoking) and a little safflower oil (which has a high smoking
point). Peanut oil, sunflower oil, and canola oil also are fine.
 All the chickpeas popped. I popped just five seeds of each variety. I heated a
small amount of oil in the corner of the frying pan until it was just short of the
smoking point. Then I added the chickpeas and stirred them with a fork until
they all popped, usually within about five seconds. I put a little salt on them and
ate them. They didn't expand very much, but they cooked completely and were
tasty. They were delicately crunchy, like a potato chip, not crunchy and hard like
corn nuts. There were obviously differences in flavor among the varieties, but I
couldn't sort that out with such small samples.

I also tried to pop three larger commercial types of chickpeas I'd bought at my
local food co-op. They popped, but they didn't cook. They were mostly hard and
inedible and were a threat to my teeth and dental work. I did my evaluations with
care.

I went back into my legume collection and started popping samples of

everything. I found that just about any bean or pea will pop, but they don't cook,
or don't cook completely. Adzuki beans are about the size of the #2 chickpeas,
and they also pop, but they aren't made edible by popping. The small chickpeas
really are different. As it turns out, there are large chickpeas that pop, too -
chickpeas as big as nunas, chickpeas big enough to make a fine candidate for a
new, delicious, high-protein snack food. But I discovered them later.

James A. Duke's book Handbook of Legumes of World Economic Importance

lists frying as one method of cooking chickpeas. I had initially assumed it was
frying of cooked chickpeas, but now I guessed that it meant exactly what it said -
stir-frying of little whole, dried chickpeas. A woman from India has since
confirmed this for me. Dry, unsoaked small chickpeas are often stir-fried briefly
to pop them, then cooked into various dishes. Of the small types I tested, some
popped so quickly and uniformly that I suspect they have been selected
specifically for stir-frying.

By this time, I had planted my twenty accessions of chickpeas. I knew they

could be planted in early spring (February) here, even though we often get some
of our hardest freezes in February. I figured that if chickpeas can be planted in
February, it also might be possible to plant them in fall and overwinter them. If
so, they would be of much more importance for the maritime Northwest.

Western Oregon gets all its water in the winter, so food plants that can be
planted in the fall to grow during the winter are especially valuable. For a grain
or seed crop, fall planting is especially valuable, because fallplanted plants can
mature their seed in summer, which is the dry season. Springplanted plants
normally mature a crop in late sum mer at the earliest and are threatened by the
fall rainy season, which starts as soon as August some years. Most fallplanted
crops can mature without irrigation. If chickpeas will overwinter here, they
might, like fava beans, be a major addition to the agricultural base of our area.

Chickpeas are traditionally planted in spring, though. Gardeners throughout

the United States and much of Canada should be able to grow appropriate
varieties.

I planted my first chickpeas in midNovember - just a short row (about eight

seeds) of each of the twenty accessions. That was probably later than optimal.
Sixteen accessions each had at least some plants that survived what was the
worst winter in my experience. I think it should be possible to develop
overwintering varieties quite easily by simple selection within available material.

Chickpeas were assumed to be unable to overwinter here. That they are

drought resistant is well known, but they are also thought to be intolerant of wet
conditions. If so, I think the intolerance must be associated with the mature
stages. They don't seem to mind continual rain when young. Most of my
accessions germinated well, even though they were planted as untreated seed in
cold, wet mud, which actually froze solid at least once before they emerged. In
areas with more severe winters, the frost-hardy types presumably could be
planted as early in the spring as the soil could be worked.

After I found out that chickpeas would pop and that they could overwinter
here, I became smitten with them. I began thinking about other aspects of
chickpeas and how they might fit into the agriculture and diet of Oregon and
America.

One way I find interesting projects is by bucking the trends. That is, I take note
of what the professionals are doing and do the opposite. I work on pole peas
when all the professional breeders work on bush types. If professionals are
interested in bush-type squashes, I work with vine types. If everyone else is
selecting under a spring planting regime, I select under fall planting. For
thousands of years people have concentrated on developing annual vegetables. A
mistake, I say. My major focus is on discovering, rediscovering, breeding,
growing, and using perennial vegetables.

Most work with chickpeas in the United States has focused on large types, so I
asked myself if I could think of any reason why little chickpeas might be more
interesting than big ones. And, yes, I could. Little things often cook much faster
than big things. Big chickpeas - the only kinds that are commercially available in
the United States - must be soaked overnight, then cooked for two to three hours.
That doesn't fit very well into the food preparation patterns of most Americans.

I knew that normal-size beans such as pintos also require soaking and cooking
for hours. But little beans like adzukis can be cooked in about forty-five minutes
without soaking. How small did chickpeas come? How fast would they cook
using standard boiling? And what about a larger popping chickpea? I suspected
there was one because Duke's book mentions that popping and eating out of
hand is one of the ways chickpeas are used. The little chickpeas I had been
working with popped, but they were too small to eat out of hand very easily.

I called Rich Hannan and reported on the stir-frying and overwintering, which
was fun. He got rather excited about it. I asked for more chickpeas, and he
promptly sent me about 120 more accessions - all the # Is and a good selection
of every other size and shape, in various shades of cream, tan, brown, orange,
black, green, and mottled and from India, Pakistan, Iran, Iraq, Israel, Turkey,
Saudi Arabia, the former Soviet Union, Spain, Morocco, Ethiopia, and Mexico. I
have been testing and growing them ever since.

All the smallest chickpeas I've tested - the # is and #2s - pop readily and could
probably be used for stir-fries. Many #3s or larger also pop, but not so quickly
and reliably. I have found two large chickpeas, however, that pop very quickly
and completely - within a second or two of hitting the hot oil. The first is a black
and brown mottled #4. The other is an orangish brown pea type with a smooth
skin. [They are, respectively, PI374085 and P1374090.1

Both popbeans have a wonderful flavor, and the flavors of the two are totally
different from each other and from anything else, including boiled chickpeas. I
mean really different - as different as asparagus and ice cream, for example. The
black and brown mottled popbean has a wild flavor; wild rice or buckwheat
groats come to mind, although the flavor is not like either of those. The orange
popbean has a flavor like a roasted nut, though not any known one. Both are
excellent.

I now have more than five hundred acces sions of chickpeas, and I'm still
evaluating them. I suspect there will be many additional popbeans among them.

My method of popping chickpeas, now that I have more practice, is to put a

little oil (I'm using sunflower oil now) in a stainless steel frying pan and heat it
to just below the smoking point. Then I dump in the beans, stir them with a
spatula while they're cooking, and dump them out. They're actually just being
stir-fried.

It's important to stir the popbeans during cooking and not to overcook them.
Small, standard chickpeas may take several seconds (up to about five) to finish
popping. Popping types, even large ones, pop and split open almost as soon as
they hit the oil. If you cook them for more than two or three seconds, you'll
probably burn them. Burned popped chickpeas taste awful. (Creamcolored types,
for example, should still be cream after popping, not brown or black.)

Popping chickpeas expand just a little when they pop; they don't blow up like
popcorn. (The same is true of nunas.) The beans do split open, though, and they
hop around a little in the pan while popping. Small ones occasionally leave and
explore other parts of the room; big ones generally restrict themselves to
jumping from one side of the pan to the other or hopping around in place.

I also did boiling trials. What I have found is that, so far, all the #1 and #2
chickpeas I have tried can be cooked without soaking. The #ls cook in about
fifteen minutes without soaking. The #2s take about forty-five minutes. I think
anything bigger than a #2 is best handled by soaking before cooking. There are
differences in both cooking and flavor within size classes, but defining those
differences will require more work.

Meanwhile, I was also growing chickpeas. I springplanted patches of about

twenty more varieties in the field where my overwintering chickpeas were
growing. They germinated well, but then disaster struck. The recession hit the
owner of the land my chickpeas were on, and he leased it to a neighbor. He
assured me that the neighbor had no intention of using the field my chickpeas
were in, but when I came back to harvest, there was nothing there but a field of
corn. Forty varieties of chickpeas, the seed increase on all my own pea breeding
work, and experimental plantings of thirty or so of my beloved perennial
vegetables all had vanished. (I did have backup seed on everything. The
germplasm, at least, was secure.)

I searched the field and found one small patch of nine chickpea plants in an
area that the plow had missed. They were all nicely dried up and covered with
pods, and they showed no signs of rot or damage, even though the patch had
been irrigated with a sprinkler system. I grabbed the plants and took them home.

The variety that had been spared was a small creamcolored #1, PI359515. I
threshed out the seed from each plant by putting the plant in a cloth bag and
stomping on it. Since this type of seed was almost round, I separated it from the
debris by dumping everything in a box and tilting it a few times in various
directions so the seed rolled away from the junk.

The most productive plant had 834 seeds, which weighed out to 128 grams
(about 41/2 ounces). The average for the nine plants was 55 grams (about 2
ounces). My plants were spaced somewhat erratically. They were originally
planted about 4 inches apart in rows 2 feet apart. The rows can be closer; I used
2 feet for ease in evaluating the plants, not because that much space is necessary.
Optimal spacing will vary with both conditions and variety. Big, bushy plants
can grow to about 3 feet across, whereas other erect plants are only a foot or so
across. Even the sprawling plants hold their pods well up off the ground.

My nine plants differed in size and shape, even though the seed all looked the
same. The most productive plant was large and bushy. The second most
productive plant was very erect and columnar. It might lend itself better to
denser plantings and/or machine harvest. I saved the seed from each of those
plants separately - that is, I started two lines in addition to the original unselected
material based on single seed descent.

The seed I harvested was bigger than what I had planted. The accession was
classified as a #1, and the seed in the package from the USDA weighed out as a
# 1. What I harvested was visibly bigger and weighed out as a #2. I am guessing
that this has to do with the growing conditions.

For the first time, I had enough seed to boil up a real batch of chickpeas. I
pooled the seed from the least productive plants and used part of that. The
unsoaked seed took forty-five minutes to cook, typical for seed of that size. The
flavor was just like that of large chickpeas but slightly more powerful - deli

cious. The chickpeas also popped readily and had a rich, nutty flavor.

I have been experimenting with the large, commercial creamcolored chickpeas,

and I've found that because of their mild flavor, I can eat them as a staple, like
rice or potatoes. They combine well with everything, even sweet things. I put
them in soups, use them as the basis for curries and spaghetti sauce, combine
them into rice dishes, grill cheddar cheese over them for a main course, add them
to oatmeal in the morning, or mash them, drench them with maple syrup, and
call them dessert.

As an experiment, I ate chickpeas every day (often more than once a day) for a
month and found I didn't tire of them. But I definitely did tire of the presoaking
and long cooking. I need the quick-chick types. I'm sure I'll make them a regular
part of my diet once I have enough seed that I can eat it instead of saving it.

It's already obvious to me that differentcolored chickpeas have different

flavors. The creamcolored ones I have tried all have virtually the same flavor. I
also have tried the three colored chickpeas (black, brown, and green) sold by
Abundant Life. They taste very different from the cream ones - more like a wild
grain and less like a nut. I boiled them after soaking, but I didn't especially like
them just boiled. One friend tells me that he's grown the Abundant Life brown,
and it makes especially good hummus. Another friend prefers the black chickpea
`Black Kabouli' (available from Peace Seeds and J. L. Hudson) to commercial
cream ones. He says `Black Kabouli' has a delicious "wild" flavor and that the
creams are "tasteless." I tried `Black Kabouli' just soaked and boiled, and I didn't
especially like it. ('Black Kabouli' will pop, but not well enough or reliably
enough to be considered a popbean.)

My Indian acquaintance told me that the green and brown chickpeas from
Abundant Life look similar in size and color to the ones she is familiar with, but
she isn't familiar with the black or cream types. The way they fix their brown or
green chickpeas is to soak them, boil them, and then cook them into various
spicy dishes. Maybe the problem is that I don't know those spicy dishes. I also
don't know yet whether all the varieties of one color will taste the same or
whether there will be differences.

Chickpeas can probably also be made into all the same kinds of fermented
products for which soybeans are used. My local food co-op already carries a
chickpea miso that I like far better than any soybean miso I have tried. Some
varieties may be better than others for making these products. Chickpeas also
can be used as livestock or poultry feed. Perhaps that's what some varieties were
developed for. Chickpeas can be sprouted and used as a vegetable or in salads.
Are there special varieties for sprouting? And what about the possibility of using
the plants as a green manure crop? It's clear that we still have a lot to learn.

Often we manage to preserve or import the germplasm, the accessions or

varieties of particular plants, but not the information about how to grow, harvest,
prepare, or use them. Often we don't need to breed the plant so much as to
discover or rediscover its uses. In many cases we need to do both - in this case,
for example, to evaluate accessions for cooking characteristics and uses, to
identify accessions that grow well in specific areas of the country and under
specific planting regimes and methods, and to select for high yield among the
material that looks most promising.

One of the major schools of agriculture in the country is at Oregon State

University, but no one there is working on overwintering peas, chickpeas, or
favas. There are very few professional plant breeders in the world, and they can
pursue only a minuscule number of all the projects that might be interesting. We
can't expect them to do everything. If I want chickpeas for overwintering in the
maritime Northwest, I'd better breed them myself. If you want chickpeas for
your growing methods and area, you, too, should get involved.

When we look at the recent accomplishments of professional plant breeders,

the tally is impressive. But it's, nothing compared to what could be achieved if
those few professionals were joined by an army of amateurs.

Rich Hannan, at the USDA Western Regional Plant Introduction Station, has
done a seed increase on the rediscovered popbeans and on # 1 chickpeas so that
he can send small (thirty-to fifty-seed samples) to any researcher who wants
them. A researcher, as far as Rich is concerned, is anybody who wants to do
research with the seed and is willing to tell him what he or she finds out. This
includes amateurs and backyard researchers. (See Appendix C for the ad dress.)
In addition, all seed companies, no matter how small, are welcome to request
material.

While many small chickpeas will pop readily, the two accessions already
mentioned, P1374085 and P1374090 remain the only large ones I've found that
pop readily and become completely cooked just by popping. I think both have
the potential for being developed into fine new snack foods.

The orange chickpea (PI374090) doesn't grow well here in western Oregon
under organic farming conditions. I'm working further with the black-and-
brown-mottled one (P1374085), developing lines for popping and other
purposes. You don't have to wait until mine are ready, however. Both original
accessions are available from the USDA.

My pea and chickpea projects are both interesting in part because I overwinter
my plants - that is, I use other than the ordinary growing method. Any unusual
growing method can produce special conditions for which specially bred plants
may be most suitable. When you do the breeding, you select for plants that
perform well using your growing methods.

My growing methods differ from those of most commercial growers in another

way. I garden organically, and as a result, I probably now have the world's most
slug-resistant pea lines.

There are a lot of slugs here. We call them garden slugs or gray slugs to
distinguish them from the banana slugs found in the hills and coastal mountains
nearby. Banana slugs grow to about 8 inches long. Ours, fortunately, usually
don't exceed about 4 inches.

They're common, though, and they're active year-round, especially during the
winter rainy season. And slugs are very fond of peas.

Many varieties of peas fail to overwinter here, not because they can't take the
cold, but because slugs eat them. Dwarf types generally don't overwinter in my
garden even in a mild year. They use up the energy from the seed in producing a
bushy little plant whose apical tip is within reach of the slugs for a long time.
Once the apical tip has been eaten, the seedling branches, producing more tips
that are still within the slugs' reach and that, because they share the plant's
energy, take even longer to grow out of the slugs' reach. The plant gets chewed
down again and again, until its strength and ability to regrow are gone.

Peas from my lines shoot up first and concentrate on expanding their leaves
and making a plant only after they already have considerable height. Slugs eat
some of them anyway, but the peas are much more vulnerable when they're
smaller. My peas are small for a very short time, so most of them survive.

Most commercial growers or professional breeders would use slug bait as a

matter of course if they were doing pea breeding under such conditions.
Therefore, they would not have selected for slug-resistant plants. Had I
considered breeding for slug resistance, I might well have thought it impossible.
But I didn't think about it. I didn't need to. Since I garden organically, I
automatically selected for varieties that excel when grown organically.

More and more people are gardening organically. The desire to have wholesome,
unadulterated food is a major motive for many of today's gardeners. In addition,
many small farms, and, increasingly, even some large ones, are converting to
more organic, more sustainable, or lower-input methods. Organic gardeners and
farmers often have to grow commercial varieties, however. Such varieties often
do poorly without pesticides, herbicides, and large doses of chemical fertilizers.

Every pattern of agriculture and every agricultural method is associated with

certain crops'and crop varieties that make it work and that work because of it.
The varieties and the methods go together. To try to garden or farm organically
with crops and varieties that were bred to excel under entirely different
conditions is to assume a grossly unfair handicap. To use old-fashioned varieties
that were developed originally under organic growing conditions is sometimes
preferable, but older varieties often have been lost, lack resistance to modern
diseases, or don't yield well enough to be competitive.

We need new varieties resistant to all the modern diseases but adapted to
organic methods, new varieties to support a more diversified, more ecologically
viable agriculture. We need entirely new crops, as well as new varieties of the
old ones, to form the basis for a modern and competitive but fully sustainable
agriculture. Those who envision and strive toward a different agriculture - a
more diversified, more regionally based, or more sustainable agriculture - must
be ready, willing, and able to do some of the breeding work.
You can breed plants because you have a vision of an agricultural future that you
want to help bring about. Or you can breed plants because you just want to play
around. Many gardeners will do some of both. I am strongly committed to
increasing the biological diversity as well as the sustainability of our agriculture.
I work with many new species, especially ones that lend themselves well to
growing sustainably, and I encourage others to do likewise. But I also enjoy just
playing around.

Can I get a leek to cross with a perennial bunching onion? That's what I
wondered while looking at an especially large leek in the supermarket the other
day. I bought the leek and stuck it into the ground in the middle of my bunching
onion patch. Alliums are generally outbreeders - that is, the flowers are usually
fertilized by pollen from other flowers. (Such information on various vegetables
is presented in Appendix A and Table 1.) With just one leek surrounded
completely by onions, perhaps some leek flowers would be fertilized by onion
pollen and would give me a wide cross, a cross between two different species.
And if they did, what in the world would the progeny be like?

If the cross is hard to do, if it requires special methods to overcome

incompatibility barriers, I might not bother pursuing the project. But if the cross
will go by itself - and, after all, many intergeneric crosses will - I'll have a lot of
fun for just the amount of work it took to poke one leek into the ground.

I've already had a good bit of fun just thinking about it and, especially, just
standing in the yard and looking at the leek. It's huge compared to the onions, its
leaves are cut supermarket fashion, and much of the white stem is above the
ground. It looks as if someone just stuck a supermarket leek into the ground in
the middle of an onion patch. It's hard to walk by without noticing the leek, and
the onion patch is adjacent to a much-used public sidewalk. The whole situation
appeals to my sense of the absurd.

You can think as big or as small as you want. You can be serious or
outrageous. You can merely try to transfer resistance to a specific disease into
your favorite kind of pea. Or you can try to develop an entirely new species. You
can even start from scratch - from an edible wild plant -- and try to domesticate
it. There are thought to be about twenty thousand edible wild plants. Only a
couple of thousand perhaps have ever received any breeding attention at all from
any human being.
 And there is no law limiting you to the strictly practical. Ewald Eliason is
dreaming of a potato that can be grown in the flower

garden but will still give potatoes. Ken Allan wants, in addition to better
Canadian sweet potatoes, a pole shelling pea with lots of peas in the pod, good
flavor, and pretty purple flowers.

Amateurs often breed for unusual colors, shapes, or sizes - for novelty for its
own sake, and for beauty. And why not? Having beautiful, unusual, or novel
plants is part of the fun of gardening.

Glenn Drowns wants purple-podded snap peas, and so do I. I want purple-

podded pole snap peas, of course, and they would have to be resistant to enation
and powdery mildew. The purple pods would look beautiful on the vines. They
would look beautiful sliced into salads. The project would take very little space
and time. More important, I like snap peas. I like to grow the plants. I love to
stand in the garden and look at them. And I love to eat snap peas, so I wouldn't
mind eating the excess. In other words, I would enjoy the project. That is the
best of all possible reasons for doing anything.

 N THIS CHAPTER and the next I discuss finding and evaluating
germplasm and experimental material. Not every project starts with a search for
and evaluation of germplasm. In some cases we have two tried-and-true varieties
that are already our favorites and we simply want to combine the best
characteristics of both; we start the project by crossing the two. In other cases a
genetic accident of some sort turns up in one of our plantings and we decide to
explore its possibilities. Or we may have a beloved heirloom variety that shows
variability for a characteristic we care about, so we begin selecting for what we
want. In such cases we already have the germplasm and are familiar with it. But
many projects do start with a search for and evaluation of germplasm.

In some cases the search for and/or evaluation of germplasm is the project. For
example, one or more varieties that would satisfy our requirements might
already be available. We just don't happen to have them or know which ones
they are. My initial work with chickpeas has been of this sort. Thousands of
accessions are available; what is missing is information about what they are
good for. It makes sense for me to start by finding out more about what there
already is. Only then will it be clear what more is needed or where to start
developing it.

However a project starts, it always requires evaluation during its course to

choose what material to use. Evaluating germplasm, like maintaining germplasm
and selection, is common to all plant breeding projects, as well as to all serious
gardening.

Beginning gardeners are usually limited to the crops and varieties that are
commonly available. As their interest grows, they begin trying different varieties
and methods, which is what is known as doing trials. But they often don't know
how to find the best material for their purposes; their trials are not designed,
conducted, or interpreted properly; and the results are often incorrect or invalid.
To a large extent, learning to find germplasm and evaluate it properly marks the
difference between the beginning gardener and the sophisticated one.

Finding germplasm includes obtaining both the physical material and the
information about it. Sometimes the germplasm is more readily available than
the information about it. In this and the next chapter I cover everything from
seed company catalogs and books to obtaining access to the collections of the
USDA. I also consider the problems associated with the vague relationship
between germplasm and variety names and its implications for the plant breeder.

My starting point for varieties and information about them is mail-order seed
catalogs. I go to other, more specialized sources, too, but mail-order seed
catalogs form my basic frame of reference. I have all the catalogs from my
region and many from elsewhere. Some are free; some cost a few dollars.

A complete list of seed companies in the United States and Canada and their
addresses can be found in Kent Whealy's The Garden Seed Inventory (see
bibliography). For an excellent list of seed companies and other sources, both
commercial and professional, throughout the world, see Stephen Facciola's
Cornucopia: A Source Book of Edible Plants (see bibliography).

Companies that charge for their catalogs usually do so because the company is
small or it handles largely rare or publicdomain varieties. Often such companies
have collections that include unique germplasm, and they produce an
information-rich catalog that tells you more about the varieties and what they are
good for than any other source. This information often is based on their own
(unpaid) original plant exploration and variety trials. The catalogs that cost
money are usually well worth it.

Some of the free catalogs also are excellent sources of information. I use the
catalogs from Nichols Garden Nursery, Johnny's Selected Seeds, and Stokes
Seeds (see Appendix D) to obtain information about disease resistance when I
am designing breeding projects. When I lived in Minnesota, I learned how to
garden from the Johnny's catalog. When I moved to the maritime Northwest, I
learned how to garden from the Territorial Seed Company catalog.

I order regularly from about twenty companies. Most are in the Northwest or
the Northeast, which shares some but not all climatic characteristics of the
Northwest. You will develop your own favorites depending on your location and
interests.

The second place I look for germplasm and information about it is the Seed
Savers Exchange. Its annual Winter Yearbook lists thousands of varieties offered
by members that are not available anywhere commercially. My favorite class is
"Miscellaneous." That's where I find many rare perennial vegetables. The Seed
Savers Exchange is open to anyone in the world. The Heritage Seed Program is
an equivalent program in Canada.

I use the Seed Savers Exchange as a source of information as well as

germplasm. Its book The Garden Seed Inventory, edited by Kent Whealy, lists all
the commercial varieties available in the United States and Canada and includes
a compilation of all the various catalog descriptions about them. I frequently find
wonderful information there. In addition, since every variety listing includes all
the seed companies that carry it, I also discover seed companies that are doing
things I'm interested in. The Garden Seed Inventory is in its third edition; plans
are to revise it about every two years.

The Seed Savers Exchange yearbooks list thousands of varieties that are not
available commercially. There is some information with each listing - whatever
the listing member volunteered. In some cases that's not much, but in others it's
quite a lot. When something is listed by five or ten people, the total amount of
information is often very useful, as are the different perspectives.

When I want additional information about a variety listed in the Seed Savers
Exchange yearbook, I sometimes write or call the listing member and ask about
it. I approached Dan Borman, who lives in Washington, over sea kale (Crambe
maritima), for example. It turns out that he, too, is very interested in perennial
vegetables. (He is, in fact, currently organizing a seed company, Meristem
Plants, which will have a heavy emphasis on edible perennials.) Seed Savers
Exchange members frequently list interests, projects, or questions in the
yearbook to find others with specific information or similar interests.

The best single source of information about varieties (including rare species
and varieties) is the book Cornucopia: A Source Book of Edible Plants. This
amazing book, comprising 700 pages of fine print in a large format, lists edible
plants and plant varieties and where to find them. It includes rare as well as
common plants, cultivated and wild edibles, and is worldwide in scope. I have
used it to find rare vegetables that have a single source in some other country, for
example. I also have found information about how rare varieties are used. Yet I
have only begun to explore the possibilities the book represents.

Cornucopia has made a major difference in my ability to find germplasm and

information. I think all serious plant explorers should have a copy of it. For
information about it, see the bibliography (under Facciola).

Organic Gardening and National Gardening magazines both have seed search
or swap services that allow readers to exchange seed. Some seed companies also
have seed search or trial programs. When they do, they describe them and issue
invitations to participate in their catalogs.

When seed catalogs, books, and the Seed Savers Exchange don't provide me
with the germplasm or information I want, I turn back to the mail-order seed
companies, this time to their owners, founders, or managers. Most of them have
much more than what they list in their catalogs. As Alan Kapuler (Peace Seeds)
puts it, "I only list 900 varieties. I have thousands. If you don't find it in my
catalog, ask. I probably have it."

Even when a seed company doesn't have what you want, it might be able to tell
you where you can find it. The seed companies that specialize in preserving
germplasm have extensive interactions with private collectors and knowledge
about material that exists outside any official or commercial channels. Many
such companies' catalogs actually include the offer to help you find something
you haven't been able to get - in other words, they offer plant search services.

The right time to approach these seedspeople is after you are already
thoroughly knowledgeable about the crop you're trying to grow and have tried all
the standard varieties. You can go to them for varieties or information. ("Do you
know of a muskmelon resistant to thus and so?" for example.) Realize, though,
that many of these companies are largely one-person operations and that the
person who knows the plants is usually very busy. In most cases you should
write and be prepared to wait for an answer (until the heavy seed selling season
is over or the harvest is in).

The basic operating rule is to give something back to those who go out of their
way to help you. In some cases what is most valuable is information. In other
cases a company may not list the species because it had only about three
packages of seed. When a company gives you something it has very little of, it's
nice if you can do a seed increase, even a small one, and give the company back
more than you received. In some cases companies will bring that up as part of
the deal.

Exercise some care in what you do with unique information or germplasm with
which you've been entrusted. Don't undermine the commercial position of the
company from which you got it. (If you are unsure about the extent to which you
are free to pass the information or material along and to whom, ask.) These
seedspeople are passionate about plants, and that passion usually matters much
more than money. But there is always a bottom line; even dedicated, passionate,
altruistic people have to survive.

The Agricultural Extension Service has a mission both to do research and to

disseminate it to the general public - which includes you. The extension service
often breeds and does trials of vegetables for your area. It frequently publishes
the results in brochures written for the layperson that it will send on request. You
often obtain the benefit of this work automatically because your regional seed
companies use it in making their choices about listings. In my area, Nichols
Garden Nursery and Territorial Seed Company make extensive use of Oregon
State University and Agricultural Extension Service research and refer to it in
their catalogs.

Various units of the Agricultural Extension Service also often provide a variety
of special services. Mine, for example, has master gardening classes, a phone
number you can call with questions, a plant disease clinic where you can take a
sick vegetable and find out what is wrong with it, and people who can help
identify wild plants, weeds, or garden pests. Some extension service people
conceive of their service as intended primarily for commercial producers. Most,
however, include gardeners in their concept.

University-based plant breeders and horticulturists, personnel at botanical

gardens, and other professionals are important sources of information and
experimental material for amateurs. You should realize, however, that their
mission does not include disseminating material or information to the general
public. They often will help, though, if you approach them properly with a
reasonable request.

Once you are involved with a particular crop, you may want to approach the
main breeders for that crop in the country or the world. Most of them are
surprisingly open to requests from the knowledgeable amateur. Appendix A and
Table 1 include references to the professional plant breeding literature. Use them
to find the world's experts in any given specialization. As with seedspeople, the
right time to approach the experts is after you have already grown all the
commonly available varieties, are familiar with what is generally available, and
have developed a clear idea of what you want to do.

Many university-based plant breeders make their experimental varieties

available for prerelease trial to serious gardeners in the area. In Corvallis
(Oregon State University), breeder Jim Baggett's new vegetables are announced
in the local newspaper; anyone who wants to participate in gardener trials of
them can drop by and pick up seed and report forms. At this stage the varieties
have numbers, not names. They are named only when the final decision to
release them is made.

University researchers in your area also may have such a program. If not, don't
let that stop you from participating in what your region's professional vegetable
breeders are doing. Many researchers are glad to have serious gardeners and
farmers try their material - if they are going to be informed of the results. Call
the breeder who is doing work you're interested in and ask.

Regional seed companies usually credit regional vegetable breeders, so if you

patronize your regional seed companies, you will automatically know the
professionals who are doing work you're interested in. If you have catalogs from
seed companies in many regions throughout the country, you will automatically
know the names of all the professional breeders in the country who are doing
work you're interested in.

In most of these relationships, feedback is the key. It is important to report

back to the researcher on how the variety did. I suggest five basic rules for
dealing with professionals or other amateurs.

1. Sometimes writing is preferable; sometimes writing doesn't get you

anywhere, and calling is better. When you write, you may not include all the
relevant information. When that happens, you usually don't get an answer. If you
call, the expert can ask for the missing information. If you write, type the letter if
at all possible unless it's very short. Always include a phone number if you have
one. That gives the person from whom you are asking help the option of
responding either in writing or by phone. Many people have strong preferences
and will respond graciously given the option they prefer, but not at all otherwise.

2. Do your homework. Don't take up a professional's time asking for varieties

that are available commercially or for information that you can find in seed
catalogs or books that are generally available. These professionals are not
running seed companies or private beginning gardening lessons and won't
appreciate your trying to use them that way. Exhaust the normal channels first. If
the crop is a common one, get some experience growing the available varieties
before you request rare material. Don't ask someone to go to special trouble
sending you a rare broccoli, for example, before you've ever grown broccoli at
all.

3. When you request help, start by explaining what you're trying to do and
where you are with it. ("I've grown this and this and that, but they had thus and
so problems. What I think I need is thus and so, or material I could use to
develop it.") Plant scientists are nothing if not curious about plants. When you
tell them about your project, they tend to get intellectually hooked and are much
more interested in helping. Often they will have useful suggestions or criticisms
or will be able to put you in contact with someone who is doing or has done
what you're thinking about.

4. Keep requests modest initially. I limit my own requests for seed to a variety
or two, for example. Even if I feel greedier than that and would like more, I
exercise self-restraint. There will be other years to request and grow additional
varieties.

5. Once the season is over, call or write your source and tell him or her your
results.

After that the relationship is a two-way one. In subsequent years you can be
more extravagant with requests for seed or information. You will have proven
yourself to be someone who will give something back in return for what you get.
You may find the professional calling you and offering material he or she would
like to see trialed or asking for information. Most professionals have networks of
people with whom they exchange material and information, and most of these
networks include at least some serious amateurs.

There are many things professional plant breeders or other amateurs will help
you with. But you should be aware that it isn't reasonable for you to ask a plant
breeder (professional or amateur) for his or her midproject working material. It's
roughly equivalent to asking a writer for the first draft of her book for you to
revise and publish under your own name; she would not be enthusiastic.

Breeders are much more likely to be willing to send you raw material or
breeding material developed early in a project (such as an F2) or finished
varieties (even before formal release) than to send you material at later stages in
the project. Once they have begun selecting and finalizing a variety, they usually
want to perfect it and release it themselves. They don't want inferior or alternate
versions of their work around. And they, like anyone else, want credit for what
they do.

I often provide access to material that is at mid-stages in various breeding

projects, but this is usually because I'm giving away or sharing the entire project,
not just the material. I find it useful to start many more proj ects than I can
complete, to start projects on a larger scale than I can handle, and to share or
give most of them away as I find people who are interested. In this way I can
help bring about far more work than I can do myself. I have some projects that
are all mine, though. I don't make my midproject working material available for
those.

Three seed companies cater specifically to amateur plant breeders. All offer a
wide range of breeding material as well as established varieties. By "breeding
material" I mean material that has already been genetically manipulated so as to
create a pool from which interesting new varieties can be selected.

In the past much of the potential of breeding material developed by either

professionals or amateurs was lost. In a typical project a breeder might start by
crossing the two most cold-hardy varieties of something and then raise the F, and
F2. Then he or she starts selecting in the F2 for a super-cold-hardy type among a
given choice of sizes, shapes, and flavors.
 The result of this process is a single variety. That variety represents the
breeder's single choice of plant type, flavor, and purposes. In addition, it has
been automatically selected to perform well under the specific soil, climate, and
growing conditions of that one individual breeder.

That same F2 could be used by other people with other preferences to select
other super-cold-hardy varieties adapted to their growing conditions and with
their choices of colors, shapes, and flavors. That F2 breeding material has the
potential to be developed into dozens of new and different varieties.

When companies sell breeding material as well as finished varieties, breeders

who provide the material can watch their initial ideas and efforts being
magnified by those of others. Those who receive the material will be able to save
themselves a year or two or three on an interesting project. They'll begin with
their own ideas and efforts and develop new varieties very rapidly.

Alan Kapuler, of Peace Seeds, offers a variety of breeding materials. His own
breeding work has focused on dehybridizing hybrids. He has developed a
number of publicdomain open-pollinated lines from popular commercial
hybrids. His interests include plant germplasm preservation, heirloom varieties,
and vegetable varieties with especially high nutritional value. For information on
access to his breeding material, see the listing for "Kapuler" in Appendix D.

Ecologist Dan Borman's company, Meristem Plants and Plant Research

Supplies for Cool Climates, also offers breeding material. He is especially
interested in perennial edibles (including vegetables). He carries a wide range of
equipment for amateur plant breeders as well.

Tim Peters' company, Peters Seeds and Research, sells breeding material as
well as finished varieties he has developed. He sells inbred lines and many
varieties that are otherwise unavailable. He has an extensive list of breeding
material of all kinds developed by himself and others. He has an unusual and
extensive collection of breeding material for perennial grains.

Those who want to obtain or offer breed ing material of fava beans should
contact The Fava Project (see Appendix D). Those interested in perennial
vegetables should consult the entry for "Perennial Vegetables" in Appendix D.
 Finally, as of February 2000, I registered the domain name plantbreeding.net.
I'm planning to use this website for one-time releases of my own finished
varieties as well as breeding materials directly to farmers and gardeners.
Through plantbreeding.net, I'll be able to sell breeder's-grade seed and
foundation-grade seed (which is not ordinarily available commercially) as well
as provide much more information about the material than is possible in a seed
company catalog.

In addition, people with valuable collections of germplasm often approach me

to find people who might be interested in it. I'll use plantbreeding.net to help
people with plants to find those who want them. (See "plantbreeding.net" in
Appendix D.)

Importing small amounts of seed from other countries - from foreign seed savers
or from seed companies - is usually legal and relatively simple. (There are
exceptions for certain crops.) Importing live plant material is a totally different
matter. Because of the potential for importing new pests and diseases, live plant
material needs to go through USDA inspection stations, and in some cases

it must be quarantined. In addition, you usually need an import permit. You don't
need any kind of professional status to obtain such a permit.

An excellent article on importing plants appears in the April 1992 issue of

Horticulture. Refer to it for information about USDA rules, obtaining permits,
preparing and shipping plants, and inspection and quarantine.

As a source of both germplasm and information, the USDAARS National Plant

Germplasm System deserves special attention. (USDAARS stands for U.S.
Department of Agriculture-Agricultural Research Service.) The role of the
National Plant Germplasm System is to collect, maintain, and distribute
germplasm and information about it to plant researchers. Some curators feel that
backyard plant breeders and researchers are included in the mission; others don't.
But even those who don't are still very likely to help you if your request is
reasonable.

I know many amateurs who have obtained material from the working
collections of the USDA. In fact, I don't know of any with a reasonable request
who were turned away by a working collection unit. Do note, however, that one
unit, the National Seed Storage Laboratory (NSSL), has as its mission long-term
storage, not seed increase and distribution. Amateurs frequently approach the
NSSL because of its name and then get discouraged when they are turned away.
Usually, however, what they need isn't in the NSSL; it's in the working units.

There are many units of the National Plant Germplasm System in different
areas of the country, and each is responsible for different crop species. In Table
1, 1 list the location of each vegetable in the system. Chickpeas, for example, are
Cicer arietinum. (If you don't know the scientific name, you can find it by
looking up the common name in the index.) In the listing for Cicer arietinurn is
the code "USA-W-6." This means that the species is handled by the W-6 branch
of the National Plant Germplasm System. You should then look in Appendix C
to find that W-6 is the Western Regional Plant Introduction Station in Pullman,
Washington. The address, the name of the principal curator, and the phone
number are given.

You can either write or call the unit. I usually call and ask the person who
answers the phone who handles the crop I'm interested in. Then I ask to talk with
that person. I don't ask for the director of the unit or the principal curator unless
that person happens to be the one who takes care of increase and distribution for
the crop I'm interested in. I always try to talk with the person who is most
directly concerned with the specific plant, the one most likely to care about it.

The USDA system doesn't cover everything. I can find more different unusual
perennial vegetables in the seed catalogs of Peace Seeds, J. L. Hudson, and
Meristem Plants than in the USDA system. And the USDA collections usually
concentrate on wild material and on land races instead of, for example, heirloom
varieties. So you can often find things in the catalogs of alternative seed
companies that specialize in heirlooms that you can't find in the USDA system.

In addition, the USDA folks don't necessarily send out everything they have.
They

usually do seed increases on material before they release it. If they haven't done
a seed increase yet, they usually won't distribute the material.

If you need just an accession or two or five, you will almost certainly be able
to get them if they are available, even if you don't know the accession numbers
and need advice about what you need. On the popbeans, for example, I suggest
you ask Rich Hannan what is available instead of just requesting the ones I've
mentioned. I'll be discovering additional ones, I'm sure, and will be reporting
back to Rich about it. That information then goes into the computer data base
and becomes part of what you have access to.

If you want a lot of something - say, one hundred accessions of something

instead of ten - I suggest that you start more modestly and establish a working
relationship first. Directness works well. You can ask how much you should
reasonably ask for, and you'll get a response that reflects where things are at that
point.

If you get heavily involved in trialing a particular crop, you will probably want
the USDA inventory for the species, if one is available. I have a copy of the
Cicer Inventory, for example. It's an 81-page listing of the Cicer accessions and
information about them.

Each order you receive from the USDAARS National Plant Germplasm
System is accompanied by a packing sheet containing the lines, "This germplasm
is being freely distributed. Continued reciprocal exchange will be appreciated."
Each little seed pack is marked prominently with the words, "Report
performance to the Coordinator of your Region." You are sent the material free,
even of postage, in exchange for a promise to give back that information.

In the 1990s the entire list of holdings of the USDA collections and plant
introduction centers and all available information about the accessions was
cataloged and put onto the Internet as GRIN (Germplasm Resources Information
Network, see Appendix C.)

The search for and evaluation of germplasm is exacerbated by the fact that
variety names don't necessarily correlate in a one-to-one fashion with germplasm
sold under or associated with the name. It's common for different varieties to be
sold under one name. There are dozens of different versions of `Kentucky
Wonder' pole beans, for example. Some have brown seeds; others have white
seeds. Some are early; others are late. Some have strings; others do not.

In some cases, probably, mutations or crosses occurred and led to different

varieties that were not given different names. In other cases different names
might have been given that related to the original name -'Early Kentucky
Wonder', for example. But extra adjectives have a tendency to get lost, so pretty
soon `Early Kentucky Wonder' is being called `Kentucky Wonder', with nothing
to distinguish it from the original. Sometimes someone develops a new variety
that resembles an established one, so he or she appropriates the established name
as a way of describing or selling more of the new variety. Or the developer may
call the new variety `Improved Kentucky Wonder', but the "improved" gets
dropped. `Improved Kentucky Wonder' may be derived from 'Ken tucky
Wonder', and it may not. Anytime I see "improved" as part of a variety name, I
realize that this is a whole new variety that may be similar to the old one in the
ways that matter to me or may not, that may be derived in part from the old one
or may not.

Not only do many varieties get sold under the same name, but the same variety
often gets sold under many names. `Kentucky Wonder' pole beans, for example,
according to the third edition of the Seed Savers Exchange Garden Seed
inventory, is also known as `Old Homestead', `Texas Pole', `Egg Harbour',
`Kentucky Wonder Green Pod', `Improved Kentucky Wonder', `Kentucky
Wonder 191', `White Kentucky 191', 'Kentucky Wonder Brown Seeded', `Old
Homestead Brown Seeded', `Kentucky Wonder Rust Resistant', `Kentucky
Wonder White Seeded', `No. 191', and `White Kentucky Wonder 191'. Some of
these are clearly not synonymous with others, but if you order just plain old
`Kentucky Wonder' from somewhere, you could be getting any of them.

Anytime someone is maintaining a variety whose original name is unknown or

lost, that person has to call it something, so alternate names proliferate. In
addition, seed companies may give new names to old varieties so as to conceal
their sources from other companies.

The situation is even worse when it comes to species vegetables. Many sources
describe Good King Henry (Chenopodium bonus-henricus) as a delicious
perennial substitute for spinach. My Good King Henry, though, is quite bitter.
I've read a number of descriptions and accounts relating to Good King Henry,
and no one mentions that it tastes horrible. Is all Good King Henry as
unappetizing as my line? There aren't any variety names at all, so I can't tell
whether the dozens of seed companies that sell Good King Henry are all selling
the same material or not.
 When a vegetable or herb is sold as a species plant, you have no idea where
the original material was collected. Some lines might derive from wild material
from southern Italy, and others might come from Siberia. One may be cold hardy
enough to overwinter in your area, and the other may not be. Some lines may be
primarily ornamental and of little use as a vegetable, whereas others might be, or
at least might once have been, primarily an edible. There is absolutely no
information in the name that lets you know. All you can do, if you care about it
enough, is to round up as many samples as you can and try them out.

Seeds Blum sells a sweet purple kale called `Ragged Jack'. Peace Seeds,
Nichols Garden Nursery, and others sell a kale called `Red Russian'.
Seedspeople consider the two the same thing. They are willing to buy `Red
Russian', for example, and sell it as `Ragged Jack', or vice versa.

I know of at least three growers for 'Ragged Jack'/`Red Russian', and the way
they grow the variety and maintain it is different. Each grower lives in a different
climatic region, grows the kale differently, and is automatically selecting for a
somewhat different line. The Peace Seeds `Red Russian', for example, is planted
in the fall and overwintered, so it is automatically being selected for
overwintering ability under maritime conditions. It is undoubtedly different from
lines that are maintained by growers who plant in the spring.

Some retail seed companies buy the kale from two different growers, so even
if you buy two packets of seed of the same name from the same company in one
year, the seed in the packets may not represent exactly the same line. And
anytime a retailer changes to a different grower for a variety, it probably also is
automatically changing to an at least slightly different variety - although there
will be nothing in the seed catalog to so indicate. Such changes can be good or
bad, depending on your needs.

Whenever I receive a package of seed from anyone or anywhere, I label it with

the source and the date, because it may not be the same as other packages from
the same or other sources in the same or other years. When I do trials, I often
obtain seed of (supposedly) the same variety from a number of companies and
trial it as if it were all different. I have Good King Henry from half a dozen
sources, for example, and plan on trialing them all in the hope of finding one that
isn't bitter.
 Any plant explorer or adventurer needs to be aware of the poor relationship
between the names of varieties and the varieties. When something matters
enough to you, be prepared to be stubborn. Don't get discouraged if one sample
from one source doesn't give you what you want. Instead, look further.

In most cases you need to grow only a few plants of each sample from each
source, so little gardening space or time is needed for comparisons of varieties
from different sources. The major cost is the time involved in locating and
requesting or ordering the seed.

Finally, just because someone else has tried something and says it isn't any
good or that it doesn't work doesn't mean you shouldn't try it, too. In the first
place, most people don't know how to trial things properly. But even if they
have, they may have tried something that had the same name but was actually
different material. In addition, your conditions are different, and you will
probably use different methods and make different choices about some of the
procedures. And you may be more stubborn or luckier, or you may have more of
a knack for whatever is involved.

 O DO plant breeding, we need to be able to evaluate varieties,
germplasm, and experimental material. Evaluation tells us what germplasm to
work with and which material to go on with. In this chapter I speak in terms of
evaluating varieties, but any lot of germplasm - such as material derived from
our own crosses - is evaluated similarly.

There are two basic kinds of garden trials. Sometimes gardeners compare
various varieties under constant conditions. Other times we're interested in how
best to grow a specific variety or a crop in general, in which case we trial one
variety under various conditions. Many projects combine both objectives,
simultaneously trialing various varieties and various methods. Garden trials of
vari eties or methods are central not just to plant breeding but to all good
gardening.

Garden trials are scientific research. Good trials require good experimental
design, execution, and analysis. But gardening trials also are just playing around,
just trying things. Your trials are your curiosity incarnate. You conduct trials
primarily because it's fun. In this chapter I explain how to get the most
information from trials with the least space and labor - and without losing sight
of the fun.

I love gardening trials. My own trials are almost never just trials, though. I eat
them. Properly designed, a trial plot usually can produce nearly as much food as
the same amount of space planted to a single variety. It yields information in
addition to, not instead of, food.

A trial plot doesn't have to be devoted entirely to a single trial. Many different
trials can be going on at once. Or a plot can be partly trials and partly for straight
food production. A wide bed can have pole bean or pea trials planted down the
middle and other crops planted along the edges, for example. I use this pattern
all the time; it makes excellent use of a small amount of space and produces
large amounts of both food and information.

Trial plots also can be ornamental. I sometimes interplant different-colored

pole beans partly because I want to try them and partly to create beauty on a
trellis alongside a much-traveled sidewalk. If I am trialing only green varieties, I
plant an occasional `Heavenly Blue' morning glory among them. The individual
large blue flowers are truly glorious against the background of the green vines
and beans.

Trials should be a routine part of gardening. Every planting can be a trial. All it
takes is a few plants or a row of the trial variety among those of your standard
planting. Twovariety comparisons where most of the planting is of your standard
usually yield good information for very little work, and they are easy to design,
interpret, and record.

You won't master doing great trials all at once, any more than you can learn to
drive the first time you get in a car. It comes partly with experience. The more
gardening research you do, the easier it gets.

There are many considerations when we design a trial, and they are all
interrelated. The following section is an outline of points to consider when
designing a garden trial. The points are illustrated, explained, and tied together
in the last part of the chapter with examples of some of my own recent trials.

Checklist for Gardening Trial Design

Recording

1. Record things in some fashion that allows you to find all the information
about one trial easily either this year or in future years. I have a separate file for
each kind of crop - mustards, for example. (It used to be a file folder; now it's a
computer file. The principle is the same.) Anything I do with mustards is in the
mustard file in chronological order. All the information for a trial is on a few
pages, not scattered among records for everything I did that year. I can find
anything I have ever done with mustards easily, even if I don't remember which
year.

2. Record only what's important, not everything. If you try to record

everything, your trials become too much work, and it's harder to find the
information that matters.

3. Dates often turn out to matter for unanticipated reasons. Always record the
date you plant. (If the new corn variety doesn't mature, it will be very significant
that you planted much later than you could have.) Flowering and harvest dates
also are often useful. Approximations are usually acceptable (for example, "mid
June," "first week in July," "for the last month or so," "April 10 or earlier").

4. Record the sources of all your seed or material. (As explained in Chapter 4,
varieties by the same name from different sources may be entirely different.) My
records are full of abbreviations such as "AL89" or "J92" (Abundant Life 1989
and Johnny's 1992, respectively).

5. Record information about the weather in some general place where you can
refer to it readily from year to year. It affects all your trials in any given year.

6. Develop regular patterns that help minimize mistakes (such as mixing up

different varieties, skipping something, or planting two things in the same
space).

I do everything left to right and front to back (as I face it). I arrange the seed
on a tray left to right. I plant it left to right. I move each package of seed leftward
on the tray as I put it back so there is a gap between it and the unplanted seed. I
label my field marker (if there is one) and place it before I reach for the next
variety. Even if I get interrupted, I always know where I am. (In complex trials, I
label markers before I go out and order them left to right on the tray, too.)

I always place markers to the left of or directly in front of the varieties they go
with (and record which). In my notebook I always list things with first to last
representing left to right. In crosses I always list the female first.
 What matters is that you have regular patterns ensuring that when your records
say something is a given variety, it really is. Make sure that anyone who works
for you or with you uses your conventions.

7. Record what you plant. There are two hinds of records: those in the garden
and those on paper or computer disk elsewhere. Markers in the garden are never
enough by themselves. Kids and birds pull them up, they fade even

though that kind of ink never did that before, or they vanish mysteriously,
especially if you don't have records elsewhere. You always need something on
paper or disk to help you figure out what is in the garden.

Markers in the garden may not be necessary, but they often add to the
convenience and fun. I like to be able to tell what is happening in my trial as I
watch the plants day by day.

8. If everything you are planting is distinguishable and you already know what
each thing looks like, you may not need any records about what was planted
where. If not, you will. The more things you plant in one trial that you can't tell
apart, the more complicated your experimental design and your labeling and
records have to be.

9. Record anything that goes wrong as it happens. Record anything that might
limit the generality or validity of your results as it happens. ("Soil at left end has
better tilth." "Ground froze right after planting." "Too much lime?") It's hard to
remember later.

10. Do elaborate prior planning if you want or need to; otherwise, don't. I
sometimes design a trial on paper and go out with a list of everything to be
planted and the seed packets all arranged in the predetermined order. But
sometimes I suddenly decide to start a trial right now, while the sun is shining
and that cool breeze is blowing. So I look quickly through my seed, grab
possible varieties, and toss them onto a tray with a piece of paper, a pen, some
wooden markers, and a Magic Marker. Then I go out to the garden and sit on the
ground in front of the space I'm going to plant and shuffle the seed packets on
the tray as I think about them and what I'd like to find out, eliminating some as I
glance occasionally at the available space. Then I plant, labeling markers and
placing them as I go. Finally, I record what I did and what's on the markers.
Elaborate planning has its place, but so does spontaneity.

Designing and Conducting a Trial

11. Figure out exactly what question or questions you are trying to answer.
What related questions don't you expect to be able to answer in this trial? Does
your trial design address the question you're most interested in or only a related
question? (See the discussion of corn trials later in this chapter.)

12. What are the answers likely to be? Are you after qualitative answers or
quantitative ones? (That is, all-or-nothing kinds of answers or degrees of
difference?) How large a difference matters? (You don't need many plants of
each variety to get all-or-nothing answers. You need a lot more to see shades of
difference.)

13. Where should you plant? Near the house or farther out where you tend not
to notice things? Where irrigation is available or not? At home or in the space
you're using at your friend's house? (How frequently will you need to tend the
plot, harvest, or evaluate?)

What about diseases from former crops? What about volunteers? Is there seed
already in the patch that you can confuse with one or more of the varieties you're
planting? You may need to do the trial somewhere where you haven't grown that
crop recently.

14. How much space will you need for the trial, including space required to
separate different varieties? (You may be able to design the trial so as to not need
extra separating space.) How long will you need the space for? When will the
trial end?

15. How many varieties do you want to test? How many are you likely to be
able to handle? Weeding or watering may take more time than planting; harvest
may take more time than weeding and watering; evaluation often takes more
time than anything else.

Are you going to be able to evaluate the flavor of a hundred different

tomatoes? You may want to know about a hundred, but planting a hundred may
very well mean that you don't get around to evaluating any of them. If you do
start more than you can finish, though, don't get discouraged and abandon
everything. Instead decide what is most important and finish that part.

16. The major difference between meaningful trials and useless ones is the
inclusion of controls. If you plant a new variety and it doesn't do well, that
means nothing without controls. There might have been a freeze right after
planting that you didn't notice. You might have used too much fertilizer (or
walnut leaves in your compost) and stunted all plant growth. You might have so
improved the soil that you now have too much nitrogen, causing excess
vegetative growth but poor flowering and fruit yield or lower freeze tolerance.

You need something else besides the new variety in the same patch, something
that you're already familiar with - a standard. The simplest meaningful trial
includes two varieties - the trial variety and the standard.

17. It's often useful to include more than one standard. In elaborate trials, I
usually include the popular standards as well as my own per sonal favorite(s) as
controls. Doing so makes my results more useful to me and to others, and it
makes it more obvious how they fit in with the work of others (who probably
used at least one of the same controls).

18. Whenever you grow out an F1, an F2, or other material derived from a
cross, also grow out a little of each of the two original parents for comparison.
The material derived from crosses always includes plants with unexpected
characteristics. You can best figure out what to do with the unexpected if you
can infer the inheritance patterns involved. For that you need to be able to go
back and compare with the parents.

19. Whenever you evaluate the disease resistance of experimental material, try
to include at least one variety known to be resistant to the disease and one
variety known to be sensitive to it as controls.

20. When trialing different methods, use both positive and negative controls -
controls that bracket the trial methods. For example, if you are testing how much
mulch to use, try to choose your least amount so that it will turn out to be too
little and your largest amount so that it will turn out to be too much. If you are
trying different planting densities, choose a closest spacing that you expect to
stunt the plants and a widest one that is too wide (and will give no benefits over
the second widest one).

21. Consider plants of a different crop for some of the controls. In trials of
overwintering garden peas, I often include the `Austrian' field pea. Garden peas
aren't normally overwintered here, but that field pea is; it's planted in the fall and
grown commercially. If conditions are severe enough to kill 'Austrian', that is
useful information. In the winter of 1990, my chickpeas overwintered in a field
where `Green Wave' mustard died. That mustard has never been winter-killed for
me before; that adds perspective.

22. How many replicas and/or how many plants do you need of each variety
and control? You don't have to have the same number for everything. The
controls are the most important because without them, nothing else is
meaningful. Where space is limited, it is often useful to plant more of the
controls than of the trials.

More plants and repeats are not necessarily better. In some situations one plant
and no repeats will do the job. In other cases a large number of plants or repeats
may be needed. (See the discussion of my tomato trials later in this chapter.)

23. How uniform is the environment? Is one end of the trial patch shaded, or is
the soil different in one end? If the growing area is nonuniform, you probably
need one or more replicas of each variety and control planted in different areas
of the plot so that you can avoid being misled by the environmental
nonuniformities.

24. Practice selective sloppiness. Do things well enough to get the answers you
need in whatever way minimizes the important costs to you (which is sometimes
land, sometimes labor, and for me usually both). Don't do ten times as much as
you need to do under the mistaken belief that the results will be ten times better.
Don't do it ten times more carefully than you need to unless it doesn't cost any
extra time. If you design the trial prop erly, ten times as much work will
probably make the results only slightly better or not any better at all.

For example, don't waste huge amounts of time measuring everything so as to

space it very accurately if small differences in performance aren't going to be
meaningful. (They usually aren't in small trials.) Approximate spacing is good
enough. My hand span is about 9 inches. Four inches is about half a hand. Your
records should reflect how you measured. If you approximate, record the
measurement as an approximation: "about 4 to 5 inches apart in the rows."

25. Should you interplant, or should you plant the different varieties and
controls in separate rows, beds, or patches? Interplanting is better because it
helps correct for the environmental erraticisms of small home plots - more shade
at this end than that, slightly different soil, and so on. However, interplanting
varieties that are not easily distinguishable requires more labeling and record
keeping, so it can end up being a lot more work.

Interplanting is often the way to get good information with the least number of
plants, and it may be essential where the conditions within the planting area
aren't uniform. I often choose varieties for a specific trial that I can easily tell
apart and interplant without elaborate labeling.

26. Are you going to want to save seed from any of the material? If so, you
may need to arrange things with that in mind. For example, the special material
might be at one end. Adjacent to it will be material that flowers at a very
different time or that will be gone by when the special material flowers. Farthest
from it will be anything that would flower simultaneously and might
contaminate it.

27. What part of the experiment are you going to eat, and when? How can you
tell it from parts you need not eat? I've found that if I can't tell what to harvest
and what to save very easily when I'm in the field, I usually don't harvest
anything; the edibles get wasted. I go to some trouble to arrange my plot so that
it is obvious what I can eat. (All the crosses are at one end, for example, or are
elaborately marked not just with little white tags but with Christmas tinsel
draped over them.)

28. Does every seed or plant count, or can you afford to waste seed?

29. If every seed or plant counts, how can you minimize accidents such as
neighbors or visitors walking through the plot, kids or dogs running through, cats
digging in it, deer or rabbits eating some or all of your plants, or birds pulling
things up?

30. Should you go to special trouble to start the seed or minimize damage from
ordinary pests (such as insects and slugs)? Or is ability to perform under
ordinary conditions part of what you're testing?

31. If every seed or plant counts, don't do anything new. Of course, the more
important the material, the more you'll be tempted to treat it specially. That's fine
as long as the treatment is something you're already familiar with. Don't presoak
seed for the first time when each seed matters, for example. There are almost
always unexpected components of any new method. Learn on material that
doesn't matter.

32. If there is genetic heterogeneity in the planting material, which matters

more, the fastgerminating seed or the slower one? The fastgrowing plants or the
slower ones? Often the fastest-sprouting seed and most vigorous plants are a
result of an accidental cross. They are hybrids. If you are trying to preserve or
"clean up" an open-pollinated variety, you may need to be careful to keep all the
plants, not just the fastest, most vigorous ones. You may have to cull out all
those fast ones to preserve the variety. If you instead cull out the slow plants (or
don't plant the slowest-germinating presoaked seed), you may, by just that one
decision in that one season, lose the entire variety. If you are trying to preserve
germplasm, such as a land race of chickpeas, you want to preserve all the plants
- those that do poorly under your conditions as well as those that do well.

33. You don't have to find out everything in one year. Often you can design a
really good trial only after you know approximately what the answers are likely
to be. It's often most efficient to do a small, crude trial the first year, then do a
much better one the following year(s).

Whenever you are planning to evaluate many varieties and invest major effort,
a small trial that is a subset of the major one can be used the first year to help
shake down your experimental design and eliminate problems with both the
principle and the practice.

34. It's often better to do a number of small trials in different years than just
one massive trial in one year. Different years have different weather, and in
different years you'll probably be using different plots of land and somewhat
different methods. A variety may perform better than a standard one year but still
be inferior to the standard, which performs reasonably well almost every year
(which may be why it's the standard). Only if your work is spread over more
than one year can you begin to see these kinds of differences.

In addition, the differences you see one year can be because of a flaw in your
trial (that section got more fertilizer) or just because of chance. Many small trials
in many years usually give you a more accurate view than one big trial done all
at once.

35. Can you use your normal plant spacing and layout for the crop, or will the
trial require extra spacing or a different layout so that you can tell one plant from
another, get in and examine them, or make crosses?

36. Blind trials are trials in which you arrange things so that you don't know
which plants are which until after the trial is completed and analyzed. They are
most important when you're worried that your opinions or biases might influence
the results or when what is being analyzed is very subjective. Blind trials usually
aren't necessary in gardening work and are not always possible (as when
varieties look different, for example). In addition, they spoil the fun of being
able to watch and understand the trial as it is happening.

One case where I frequently use blind trials is for evaluation of flavor. I don't
bother growing the plants blind, but arrange things so that I don't know which
plant the material to be tasted came from until the tasting is done and the results
recorded.

37. What kind of labeling do you need? Will you need to be able to distinguish
each individual plant or only groups of plants? How can you lay out the garden
so that you can have a simple map that allows you to reconstruct what
everything is after the markers mysteriously vanish?

38. How can you best arrange the varieties within the trial plot so as to
facilitate your being able to identify and evaluate the material? If you are testing
three pole beans that are to be grown on a single trellis, put the yellow one
between the two greens. If the two greens have similar pods, you'll not be able to
tell where one planting begins and the next ends. It takes about a running foot of
empty trellis space to separate the two greens if you plant them side by side - a
foot more space than is necessary.

39. Where appropriate, use plants instead of empty space to separate varieties
you're testing. A few purple-podded pole beans can be used to separate a number
of green varieties that you're testing. The space isn't wasted, since you eat the
beans. I've separated green pole beans with `Heavenly Blue' morning glory or
climbing edible nasturtium plants. Fava beans can make good separators for
earlyplanted material. A single fava bean can be planted between different pea
varieties, for example.

40. Don't cancel a trial prematurely. When one treatment does better than
another, don't rush in there and try to improve things for the other. The
differences you see early in a trial may not hold up later. Further, seeing those
differences was the point, not creating optimal conditions for everything. Finally,
those differences, if real, are not the whole story. (See the discussion of my
potato trial later in this chapter.)

Evaluating a Trial

41. In evaluating trials, take the environment into account. The plant that is
southernmost on a north-south row gets more sun than all the others; it may be
biggest through no virtue of its own. The same is true if a plant is bigger or
earlier than its neighbors but has more space because those next to it didn't come
up. Don't plant in such a way that entire experimental groups shade others.

42. The plants at the edges of patches are usually larger than those in the
middle; you may need to take edge effects into account. Suppose you have a
number of 10-foot-square patches of something and are evaluating yield in
preparation for planting whole fields. You should probably discard all the plants
along the edges of your patches and harvest and evaluate only the plants in the
middle. The yield of just the middle of the patch will give you the best
approximation of what you could expect in a large field (where the contribution
of plants along the edges is insignificant).

43. Notice and record any evidence of genetic heterogeneity. Do the plants that
are supposed to be all the same variety look like it? Or do they appear to be a
mixture of distinctly different types? It may be OK for the variety to have such
heterogeneity - it may even be desirable - or it may not be.

44. A general, subjective impression of performance is frequently adequate.

General impressions can be misleading, though. Impressions about yield can be
especially mislead ing. For this reason, it's often necessary to take at least some
measurements.

45. Chance affects the results of a trial. If two varieties are identical, in any
given trial, sometimes one will look better and sometimes the other. Scientists
use statistical methods to help evaluate the effects of chance. Most good
gardeners develop an excellent feel for what kind of results are meaningful just
as a function of experience.

46. The fewer the plants in a trial, the larger the effects of chance. A 10 to 20
percent difference in yield between two varieties based on, say, ten plants of
each usually isn't reproducible. Even if the varieties are identical, you will often
see differences that large just by chance. But a 10 to 20 percent difference
between two varieties based on thousands of plants of each usually is
reproducible and significant.

47. If you want to be able to establish small differences, you need more plants
than you need to establish larger ones. To ascertain a 10 percent difference in
yield between two varieties, you usually need hundreds or thousands of plants,
not just a few. A 50 percent difference in yield between two varieties is usually
obvious with just a few plants.

48. When you see an apparent difference between two varieties and you are
wondering whether it is meaningful, one trick is to look at the individual plants.
Is all the apparent difference associated with just one or two plants? If the
difference would disappear if a single plant were not there, the difference
probably isn't meaningful. Another trick is to divide your trial mentally down the
middle (which turns it into two smaller trials) and ask whether both trials show
the difference.

49. Notice the range of both varieties. Is nearly every plant of one variety
better than nearly every plant of the other? If so, the difference between the
varieties is probably meaningful, even if it is small. Are all the biggest plants of
just one of the varieties? If so, that is probably meaningful, even if the averages
for the two varieties are about the same.

50. When I need quantitative data, I often measure and record certain plants
instead of all of them. In many cases I record just the best plant ("heights up to 7
inches"). The performance of the best plant is especially important because it
represents what the variety is capable of under my circumstances. The worst
plants aren't as meaningful because in many cases they were just unlucky (for
example, that's where a cat dug).

51. In small plantings the performance of the median plant is often a better
indication of the true average than the average you would calculate by measuring
or weighing all the plants. The median plant is the one that is better than half of
them and worse than half. To get the median, you have to choose and measure or
weigh only one plant.

52. When I don't need quantitative information, I usually don't collect it. I
simply record my general impressions. ("About the same as the standard." "A
little less than the standard, but I'm not sure." "About 3/4 as much as the
standard, but still good enough to try again.") When I do need quantitative data, I
often record the best plant, the median plant, and the worst plant - that is, the
median and the range. I keep more elaborate records only if there is some reason
for doing so.

53. Gardening is both subjective and scien tfic. Acknowledge, appreciate, and
value both the subjective and the scientific. There is no virtue in filling your
notes with numbers and spending a lot of time getting them if the real result -
that you do or do not want to try the variety again - is obvious and is all you
really need. That conclusion may be subjective, but that is the result that matters.

The Bigger Picture

54. When things go wrong, try to evaluate what is left without getting
discouraged. If you've used good design, you almost always get at least some
information if you look at the situation and think about things before you plow
up the plot in disgust. You usually get some of the information you want. You
may end up getting some other information as well. Instead of finding out how
five varieties performed under ordinary conditions, you may find out that when
you go out of town and the friend who promised to water the plot doesn't, only a
particular variety survives at all. That isn't what you wanted to know, but it isn't
totally useless.

55. The worst experiences seem to make the best stories. You may not enjoy
learning that you have done a trial to find out what happens when nothing gets
watered, but your friends will undoubtedly love to hear about it. Try to keep
your sense of humor. There will be other years.

56. As you record and evaluate your trials, ask yourself whether the answers
you got are to the questions you thought you were asking. Sometimes they aren't,
and it isn't necessarily obvious. Don't assume that they are without going back
and reviewing. Thinking about a trial afterward is often the most important part.

57. All experiments are flawed. However you design and carry out your trials,
there will be some flaws. Every trial also has limitations to how general and
applicable it is. Some of these are inherent in the design, and others result from
what happened along the way. Record any flaws or limitations that might have
affected the results or that might have made the trial turn out differently than it
might in some other year, on some other land, or under some other conditions.

What matters isn't to design and conduct an experiment with no flaws or

limitations - which isn't possible - but to recognize and record what the flaws and
limitations are. Examples of limitations I often list are that I planted later than I
would optimally or normally, that part of the planting was more shaded, or that
the summer was unusually cold or hot.

58. Did anything unexpected turn up along the way? Did you find out
something interesting that didn't happen to be what you asked? If so, record it.

59. Once you have your results, you need to decide what to do with them. That
isn't science; that's judgment. And it's subjective. Your results don't dictate your
judgment; they merely inform it. They establish some of the facts that are
relevant. Usually your judgment is to grow something again or not. Or to grow
more of something or not. Or to grow it a different way. You put your trial
conclusions together with everything else you know and care about, and then
you choose.

The choices may not be permanent. You may think of ways to get around the
problem your trial pointed to, for example. You only tried a specific case under
specific conditions. You may decide that that didn't work but something similar
might. So you do additional trials.
 60. In making your final judgments, consider the bigger picture. If an open-
pollinated variety that hardly anyone grows does almost (but not quite) as well
as the fancy hybrid that everybody grows, maybe you should choose the open-
pollinated variety. You can save its seed if you need or want to. In addition,
when you buy the seed, you are probably supporting a small grower and a family
seed company; when you buy the hybrid, you are probably supporting a
multinational giant.

If a rare variety does as well or even almost as well as the common one, maybe
you should grow it just because of its rareness. By growing it, you help to
preserve it and the information about it. And by growing something different
from your neighbors, you might not get their pest or disease when it sweeps over
the whole area; your variety might be resistant. In addition, the pest or disease
your variety might be most susceptible to will be less likely to sweep over your
area, since your neighbors are growing something else. Whenever you grow
something different from your neighbors, you are contributing to the agricultural
biodiversity of your country, your region, and your neighborhood. This is part of
the bigger picture.

Putting It All Together: Some Examples

Mustard Trials, 1991

I am inordinately fond of mustard greens. I plant in March for harvest a few

weeks later, in July for fall and winter use, or in August for use in early spring.
But so far I haven't found any way to have fresh mustard greens during the
summer. To get summer mustard, I would need to plant in late spring; mustard
planted then bolts without yielding much. One variety, `Pak Choy', will not bolt
when planted during that window. But `Pak Choy' doesn't yield for me either
when planted then. It's a mild type that is destroyed by insects and slugs that are
especially efficient in late spring. I also don't much like the flavor.

Prior to 1991, various seed catalogs had been showing me pictures of beautiful
new mustards, enticing me with suggestions about better or different gourmet
flavors, and listing some as "for growing anytime." I wanted to see if "anytime"
included my late spring; I wanted to try some of those new flavors. And I wanted
a rough idea about vigor. How well would other varieties compare to `Green
Wave', my standard, which outgrows most weeds? If mild flavored, would they
grow fast enough to keep ahead of the pests?

The space I had available for the trial was - well, none. Every square inch was
already full of plants. I did have a compost heap that was more or less through
cooking. It was located on a former hard-packed dirt driveway north of the
garage and under a tree. I quickly shoveled one end of the pile into a flat section
about 2 feet wide, 4 feet long, and 5 inches deep. That would be the mustard trial
bed. It would support growth of the plants for a month or six weeks - long
enough for me to taste everything and see what bolted and what didn't. It
wouldn't support growth of full-size plants. But I wasn't expecting full-size
plants; I was expecting premature bolting. (This is an example of how asking
what answers you expect helps you design. In this case it suggested that I didn't
need the land for very long - or even deep enough "land" for "real" plants.)

I would plant `Green Wave' and `Pak Choy' as my controls, along with some
new types. I was expecting the `Green Wave' to bolt quickly before producing
much of anything, because that's what it always does when planted in late spring;
that was the problem. The `Pak Choy' would not yield because it would get eaten
into the ground. Then there would be other mustards, new mustards. Since I was
expecting all-ornothing answers to one question (bolts or doesn't), nearly all-or-
nothing answers to another question (gets decimated by pests or doesn't), only a
rough idea about the third (vigor), and just a taste for a first impression of flavor,
I wouldn't need very many plants per variety I tested.

The next problem in designing this trial was to note the extreme erraticism and
nonuniformity of my growing conditions. The compost was pretty poor stuff.
Some was loose; some was in not totally cooked clumps. Very large populations
of sow bugs were in much of it. The entire patch was in the shade, and one end
was almost in full shade - much more shaded than the other end. The whole bed
was only 2 feet by 4 feet. What this meant to me was that I should interplant. I
needed to have my controls and my experimental varieties spread out over the
entire patch. Separate rows for each variety obviously wouldn't work. I would
need to plant at least two rows of every variety to correct for the variable
conditions in the patch. Three or four would really be better. But there wasn't
enough space for even three or four rows of each of the controls. Even with two
rows of each variety, I would be limited to one or two new varieties.

If I mixed a number of varieties together with the controls and broadcast them
throughout the patch, I would have all varieties and controls in all possible
locations, and the nonuniformity of my patch wouldn't matter. If I interplanted in
this way, I could test several varieties - if I could tell them all apart. In fact, since
I didn't need many plants of each and I was expecting the trial to end with them
still all quite small, if I interplanted, my patch would be a luxuriously adequate
amount of land. I could easily try half a dozen varieties. The big problem would
be telling them apart.

There were a lot of mustards I wanted to try, and I couldn't try them all at once,
so when I went about ordering varieties, I deliberately picked ones that I thought
I would be able to tell apart. (Color photos in the bigger catalogs help a lot.) My
trial would include `Pak Choy', `Joi Choi Hybrid', `Mei Qing Choi Hybrid',
`Tatsoi', `Green Wave', and `Red Giant'. The first four are Brassica campestris
types. `Joi Choi' is a hybrid pak (or bok) choy with green leaves and heavy white
stems. `Mei Qing Choi' is a hybrid Shanghaitype pak choy with light green,
flattened stems. `Tatsoi' is a pak choy with small, deeply intense green leaves
and white stalks.

`Green Wave' is a Brassicajuncea type. Its flavor raw is fiery, much too hot to
eat in more than tiny amounts. I use small amounts raw, shredded fine in salads.
But the real virtue of `Green Wave' is as a cooking green. The chemical
responsible for the hotness is destroyed instantly by cooking. What remains is a
rich, uniquely flavorful green, and the broth it creates is also richly flavorful.
`Green Wave' is my favorite cooking green and soup green.

`Green Wave' also grows faster than anything else I grow. Slugs, insects, and
other pests generally cannot eat it. When properly grown - in cool weather and in
good soil so that it grows rapidly - the entire aboveground part of the plant is
edible, including the succulent stalk and leaf stems. (With older or less optimally
grown plants, the stems and stalks are stringy and bitter.) It also freezes well. I
slice the whole plants into inch-long pieces and blanch them in unsalted water
for a couple of minutes. Then I freeze the greens with the broth.

Somewhere along the line, we took a wrong turn with our breeding of cooking
greens. Growing and thinking about `Green Wave' made me realize this. We
breed cooking greens that are the same as our raw-eating greens. A mistake, I
say. Salad greens and cooking greens should be different plants.
 Salad greens must be mild flavored when raw because we eat them raw. So do
all the bugs and slugs. This makes them hard to grow. We end up fighting the
pests or dumping pesticides on the plants so that there will be something left for
us. But cooking greens don't have to be edible raw. We could raise them much
more easily if they weren't. When planted at the right time of year, `Green Wave'
grows faster than the weeds.

I think `Green Wave' represents a superior agroecological concept for a

cooking green: it can be eaten only by animals that know how to cook. `Green
Wave' mustard fits superbly into low-input, sustainable agricultural patterns. I
think we need more and better cooking greens than we have, and in my own
breeding work `Green Wave' is one of my models.

`Red Giant' also is a Brassicajuncea type. The pictures in the seed catalogs
made it look wonderful, but I had never grown it. I wondered whether it was just
a different-colored version of `Green Wave'. Would it have the same flavor or a
different one? If different, was it a hot type or a mild type? Would it grow
anywhere nearly as well as `Green Wave'?

I mixed seeds of the different varieties together and broadcast it. I didn't bother
labeling anything in the field. I did, however, record the date, varieties planted
and sources, a few brief remarks about the compost, and a curse or two about the
sow bugs - about three lines total.

The `Green Wave', as usual, grew much faster than everything else. I thinned
to cut back the `Green Wave' and create a uniform stand of all the varieties
throughout the patch. There weren't any weeds. I watered the bed every few days
when I tended my ducks, since it was in the same area.

The `Green Wave' bolted quickly before providing any significant amount of
food, as I had expected. Nothing ate it. The `Pak Choy' did not bolt but was
eaten into the ground by pests before providing any significant amount of food,
as I had expected. The `Red Giant' was hot, but not nearly as hot as `Green
Wave'. And its flavor was entirely different. It was not simply a different-colored
`Green Wave'. It both bolted and was eaten into the ground by pests. Everything
else also bolted and was eaten into the ground - except for one plant, which
didn't look like any of the varieties I'd planted and which I'll return to later.
 Only the `Pak Choy' and the hybrid version thereof were bolt resistant under
these conditions. That didn't do me any good because of their pest susceptibility.
I didn't get to taste more than a tiny leaf or part of a leaf of each, but it was
enough to tell me that as far as I was concerned, there was nothing new or
special with respect to flavor that would make me want to grow the new
varieties.

The `Red Giant', though, was distinctly different in flavor. The plants were
beautiful, too - intense purple-red tops of leaves and light green underneath. It
was clearly not quite as vigorous as `Green Wave', but it was pretty good.

None of my six varieties was successful at growing in the mustardless window.

None of the B. campestris types had any remarkable new flavor. But I did find a
new variety that I wanted to work with further -'Red Giant'.

Of the flaws and limitations of the trial, the major one that might affect my
conclusions was the abundance of sow bugs. Sow bugs aren't my normal pest, so
the mustards that were decimated in this trial might behave differently in another
trial with a different major pest. Shade was a concern initially. Had things failed
to bolt, it might have been that they would have bolted with more

sun. That is, had the results been in the opposite direction, I would have listed
shade as a limitation on the generality of the result. As it was, with everything
bolting, shade seemed not to matter.

Also, my impressions about flavor have to do only with warm-grown material.

It's quite possible that the flavors would be quite different under cool-grown
conditions. I expect that they would be different once the plants were touched
with frost.

Nothing suggested that the nonuniform, weird, ridiculous conditions of my

trial bed were any problem. I had a good spread of plants of all kinds all over the
bed. All the plants of the bolting varieties bolted, and none of the plants of the
nonbolting ones bolted, wherever they were in the patch.

It didn't take me any time at all to figure out my trial design. I use this design
all the time. I knew I would use this design when I chose what mustards to order
from the seed companies. There are dozens I wanted to try, but I couldn't do
them all, so I chose ones that I could tell apart readily. Once I've tried all those
that work in this kind of trial, I'll have to use a different design to deal with those
I can't tell apart.

Choosing four varieties that I could distinguish from `Green Wave', `Pak
Choy', and each other took about an hour with the seed catalogs in the winter.
Time required for building the bed and planting: about half an hour. Evaluation
time: about half an hour.

The trial provided good basic information on flavor, bolting, and pest
resistance, and rough information on vigor for four new mustards trialed against
`Green Wave' and `Pak Choy' as controls. It required no real land and only two
hours of work, one of which was lolling around in an easy chair indoors drooling
over seed catalogs.

Interplanting is a powerful technique for getting the most information out of a

small amount of land with erratic conditions. I commonly use a couple of other
interplanting patterns. For example, if I want to trial a mustard that might look
like `Green Wave', I sow one end and the middle of a food production bed with
`Green Wave' and the other end .and the middle with the test variety. That way, if
I can tell the varieties apart, I have the advantages of interplanting in the
midsection and can get a rigorous comparison. If I can't tell the two varieties
apart, I find that out and at least get some information by comparing the plants at
the two ends.

I would usually plant the bulk of the bed in `Green Wave' and only one end
with a little section of overlap in the new variety. That way I get almost as much
food from the bed as I would without the trial, even if the new variety is
worthless.

If I want to trial a number of varieties that I may not be able to tell apart but
that I can tell from `Green Wave', I plant the new varieties in separate rows and
then broadcast `Green Wave' over the entire patch. The varieties are separate
from each other but interplanted with the standard. This design gives a good
comparison of the growth and vigor of each variety as compared to `Green
Wave'. That is more important to me than how they do relative to each other.

Where different kinds of seed are the same size and shape, I usually mix them
and broadcast all of them at once. If they differ substantially, I broadcast each
separately. Afterward, I either rake or water in the seed. I usually try to sow
carefully so that I don't have to thin too much. When I thin, I favor whichever
plants I need to get an even stand with the appropriate proportions of all the
varieties. When the varieties are in separate sections or rows, I use short little
rake strokes within the section or row, not broad strokes that drag dirt and seeds
between different sections.

I mentioned that there was one unusual plant in my mustard trial patch. I noticed
the plant quite early because it was doing so well. It was growing as well as, or
perhaps even a bit better than, the `Green Wave'. After the `Green Wave' began
to bolt, the plant towered over everything else in the patch. It was not any more
bug-or slugeaten than the `Green Wave', even though it appeared to be a mild-
mustard type. And it wasn't bolting.

When I first noticed the plant, I gazed at it quite a while without touching it.
Then I reached out a reverent hand to pluck a single leaf for tasting. And it bit
me!

I withdrew my hand quickly. Then I got down on my hands and knees and
really looked at the plant. It had bright green leaves and very light greenish
stems. I reached out gingerly and touched a leaf. Hairs. It had hairs. Delicate
little hairs all over both upper and lower surfaces. They weren't especially
hurtful, if you were expecting them. Pests, though, apparently did not enjoy
them.

I picked a leaf with its stalk, took it inside, sliced it, and dropped it in boiling
water for a second or so. Then I withdrew the leaf and tasted it. All signs of the
hairs vanished in even an instant of cooking, and the leaf had a mild flavor,
which seemed to be something special. It seemed much more delicious than
cooked greens of other mild mustards. But I might have been imagining the
difference. I couldn't compare the plant's flavor with anything else, since nothing
else produced enough edible greens to cook.

Aha, I thought happily, here, potentially, is another cooking green of the kind I
argue is ideal for organic or sustainable, low-input gardening or farming - a
cooking green that has strong natural defenses against any animal that doesn't
know how to cook. And being both nonbolting and pest resistant, it could be
planted in that mustardless window and give me mustard in midsummer.

I didn't know what the plant was or where it came from. It could have been a
mutant in any one of the varieties I'd planted. It could have been a seed that
resulted from an accidental cross in the field of the growers of any one of those
varieties - a cross with a wild brassica of some kind or maybe even with a turnip.
(Turnips are somewhat hairy, and the root on this thing looked somewhat like a
turnip.) It also might have been a contaminant, something caught in the seed-
processing equipment from whatever was processed before one of my varieties.
But I don't know of any hairy `Pak Choy'-type mustards. Since some of the
varieties I planted were hybrids,

this plant also could have represented an accidental self-pollination of a parent

variety, or it might have been from a seed from my own compost.

The unusual plant overwintered and grew to a huge size - 2 feet high, 4 feet
across, and covered with flowers. I bud-pollinated some flowers in case it had an
incompatibility system I needed to get around, but it wasn't necessary. The plant
produced thousands of seeds, and the seed was nearly 100 percent viable. I have
passed most of this seed on to Miriam Ebert, a friend and fellow gardener.
Miriam and I will work together to select a new, insect-proof, hairy mustard
variety. Should we develop something useful, we will offer third and further
generations of breeding material to others, so that others can derive their own,
locally adapted varieties.

Perhaps that original hairy plant will form the foundation for a new mustard
that satisfies my needs during the hitherto mustardless summer. It may even be
the foundation for a number of varieties derived by others to meet their needs
and preferences. But whatever else, this plant has taught me something. It has
focused my attention on hairiness.

I had not, until this plant suggested the way, recognized the full implications
and possibilities of hairiness. Many wild mustards are hairy. It shouldn't be
difficult to create hairy brassicas, having once decided that there is a reason to do
so. The trick, of course, is the right sort of hairs. They have to disappear when
cooked.

My trial wasn't aimed at finding out whether hairiness could produce pest-
proof but still edible brassicas, but that's part of what I found out. This is an
example of what I mean by a trial producing answers to questions other than the
ones you asked. The single hairy plant could turn out to be more important than
all the rest of the trial, either by providing germplasm for future food or by
providing food for future thought.

Corn Trials; Variable Spacing

I frequently try out new corn varieties. Sometimes I try sweet corns,
sometimes field corns. Right now I am especially interested in old-fashioned
roasting-ear dents.

When I trial corn varieties, I usually plant a 20-foot row of each. At one end of
each row, I space the plants about 6 inches apart; that's close enough so that any
variety will be stunted. Then I plant at wider and wider spacings until, at the
other end of the row, the plants are 3 feet apart. That's far enough apart so that
any variety will have more space than it needs. That way I see how each variety
does when spaced optimally, and I find out how various varieties should be
spaced under my conditions.

Let's suppose instead that I planted a small early variety and a large late one at
a constant spacing of 8 inches apart in the row. The small variety would yield
almost optimally; the large one would probably be so stunted that it wouldn't
mature any ears at all. I wouldn't be asking how the varieties compared to each
other. I would be asking how they compared when spaced 8 inches apart. The
two are very different questions.

At a spacing of 2 feet, the large variety would probably yield more than the
small. If

I wanted to know potential yield per foot of row, I would need to evaluate at
more than just one spacing.

With a 20-foot row with variable spacing, I can get a good idea of how each
variety performs optimally, how each variety should be spaced so as to perform
optimally, and how each variety performs with specific spacings. When I
compare corn varieties grown in beds, I plant them 6 inches apart in all
directions at one end, varying to 2 feet apart at the other end.
 A practicality when you use variable spacing is that you don't want holes in the
pattern. One way to generate a holeless pattern is to plant two or three seeds in a
little cluster in every position where you are going to want a plant. Then thin to
one plant.

When you are interested in how the varieties do, you need to ask yourself
exactly what you mean by that. Is it total yield you care about? Yield in some
part of the season? Yield of material only above a certain size or quality? Yield
early in the season? And if it is yield you're after, is it yield per plant or yield per
unit space?

In my trialing of different corn varieties with different spacings, I really care

about yield per unit of garden space, not yield per plant. But it isn't total yield
that matters to me; it is yield of ears of a certain size and quality. Usually there is
some spacing that stunts the plants a bit and gives two or three smaller ears but
gives the highest total yield per foot of row. And there is some other, greater
spacing that gives less yield per foot but produces larger ears. I generally prefer
larger ears. Little ones cost more labor in preparation per amount of food. I don't
like shucking little ears and getting just a few bites of kernels. So the question I
am most interested in isn't "Which spacing gives the highest yield?" but instead
"Which spacing does the best job of giving me a reasonable number of large
ears?"

To help yourself accurately formulate the question you really want to ask, it
often helps to ask what you do and do not care about or expect to learn from the
trial. In corn trials of the sort I have just described, I can get a good idea about
earliness, size of plants, size and shape of ears and grain, and optimal spacing. I
can get only a general idea about yield. And I can't get more than a rough idea
about flavor. There are only a few plants at optimal spacing for each variety, and
each variety has different characteristics that tell you when to pick the ear. I can't
easily learn the tricks for telling when the ears of a variety are prime unless I
have more plants of it.

If I am impressed by the performance of a variety in my initial trial, I plant an

entire block of it, or two or three blocks, at the spacing suggested by the first
trial. In the second year, I learn to harvest the variety in its prime and can
compare it with other varieties harvested in their prime.
 Tomato Trials, Spring 1992

Miriam Ebert and I are cooperating in some tomato trials. Altogether, we are
trying about twenty cherry-type varieties and about twenty standard large ones.
Miriam is a tomato fanatic. She normally plants twenty or more varieties and
loves to try new ones. My major motive is related to my interest in creating
`Baby Girl' (see Chapter 3). `Baby Girl' would have the flavor of `Early Girl
Hybrid' and would be as early, but it would be open-pollinated and somewhat
smaller. I may also want it to be a high-vitamin orange instead of a low-vitamin
red (see Appendix B). I have a number of breeding projects aimed at creating
`Baby Girl'. But what about the flavor of the cherry and smaller tomatoes that
already exist? Perhaps someone has already bred my `Baby Girl' or has
developed something that I can use as part of the germplasm for developing it.
So I am interested in tasting as many cherry and smaller tomatoes as possible.

We are growing several plants of `Early Girl', since that is the most important
control. They are spread around the patch, interplanted with the other varieties.
We're also growing a number of plants of certain other varieties we care a lot
about. But we have only a plant or two of many varieties. There are reasons that
even one plant of each would probably be all right.

One major reason has to do with the fact that tomatoes are such strong
inbreeders. This means that it is likely that the seeds within each variety are
close to genetically identical. Open-pollinated varieties of outbreeders typically
show much more variability from plant to plant. (I explain why in Chapter 9.)
With a strong inbreeder, an individual plant is likely to be typical; with an
outbreeder, you usually need a number of plants to get an idea of what is typical.

A second reason that we can get away with just one or two plants for some
varieties is that it wouldn't be a tragedy if we lost those plants. We would just try
them again next year. It would be a tragedy if we lost the `Early Girl', though;
doing so would destroy the entire trial. So we need greater numbers of `Early
Girl' just to be on the safe side.

The major question in the trial is that of flavor. We'll be looking for plants that
produce tomatoes that taste at least roughly as good as `Early Girl'. We expect
only an approximation. We will trial any variety that seems anywhere near as
good as `Early Girl' more thoroughly in future years. We also should be able to
get a rough idea about earliness and tomato size, shape, and color.

We don't expect to be able to tell much about yields, though, since each plant
will have different neighbors, which will cause variable degrees of crowding and
shading. And each will be in a different section of the patch.

I often use an initial trial just to find out what a variety tastes like. The trial
will be quick and dirty because I won't be doing it well enough (with enough
different plants in uniform enough conditions) to evaluate yield. I don't care how
much a variety produces of something that I don't want to eat. If I like the flavor,
I'll trial the variety in additional relevant ways in future years.

Potato Trials; The Effects of Mulching

I was involved in a trial in which one bed of potatoes was mulched to see how
that would affect performance compared with unmulched controls. Most of the
planting was done normally. Just a single bed was mulched with an ample layer
of straw. The mulched potatoes came up much more slowly than the others. That
was no surprise, since mulched ground warms up later in spring than bare
ground.

The other participants in the trial remarked on a number of occasions about

how miserable the mulched potatoes looked. They asked me if we should take
off the mulch. I said no without seriously considering it - or explaining. There
was good reason for not even considering removing the mulch, but in retrospect,
I should have explained.

The next time through the field, I wasn't there, and my coworkers unmulched
the potatoes. It was already obvious that they weren't doing as well as the
unmulched ones. They were only about half as high and were meager, spindly
things compared to the bushes the controls had become. But removing the mulch
at that point destroyed the trial. Here's why: The trial was of mulched versus
unmulched. When the mulch was removed, it became a trial of temporarily
mulched versus unmulched. That isn't a useful trial, since we would not normally
do things that way. We destroyed a useful trial and substituted a useless one.

Second, the slower growth of the mulched potatoes wasn't the whole story. We
weren't asking the question "Will mulching slow or not slow initial production of
aboveground plant mass?" We don't eat potato plants. What matters with
potatoes is the potatoes. Slower initial growth of the plant mass may not be
disadvantageous. If the season is long enough, there may be just as high a yield
of potatoes.

Later maturation may avoid important pests or diseases. Later maturation

might be better for storage. Mulching may create better moisture conditions that
more than make up for cooler soil and slower initial growth. Or it might make
the patch more droughtproof. The mulched potatoes may yield more even if they
start slower.

But even if mulched potatoes yield less, mulching might be preferable to us.
The potatoes may be bigger, cleaner, easier to harvest, or more disease free. All
these factors need to be considered.

The initial question was "How will it work if they're mulched?" That is a
subjective question, but that's OK. It was the right question, the relevant
question, the question we really cared about. But we couldn't get the answer to
that question without going all the way and producing the potatoes.

In general, you shouldn't cancel a trial just because an experimental treatment

shows a difference from the controls in a given direction at a given time. Nor
should you try to rearrange the experimental design in midcourse. There may be
exceptions, though. If most of the labor is still to come, for example, you may
prefer canceling a trial that has already developed serious problems. It may be
more efficient use of your time to restart a new and improved version and invest
the evaluation time in it instead of in a badly flawed initial attempt.

The Pea Patch, Spring 1992

My southernmost raised bed lies between the sidewalk and a busy street. It is
about 4 feet wide and 18 feet long. All my pea breeding work and much else has
been done in it. (I did plant peas on a farm in Independence, Oregon, one year,
but that material was tilled under and lost.) The peas occupy a wide row down
the center; the sides of the patch are full of other plants.

The patch is irregular in shape. One side is straight; one has a decided dip. One
end is square; one end isn't. The supports for the trellis are at irregular intervals.
They are not especially straight. Two-by-two-inch boards run across the bed
every few feet, and they are at irregular intervals, too. Some go directly across
the bed, and others go more diagonally. I like asymmetry.

At the moment the bed with my pea breeding also contains trials of ten
different daylily varieties for flavor of the flowers; several `Ragged Jack'/`Red
Russian' kale plants for my salads; a few brassicas (a `Green Wave' mustard and
a 'Red Giant' mustard, a couple of `Marrowstem' kales); some new brassica,
whose identity I don't know, that came from a seed saver in Denmark but that is
flavorful and very vigorous (perhaps I should start thinking about finding out
what it is); three different wild mustards I might want to make crosses with; two
plants of a mutant onion that showed up in my onion patch last year; one
chickpea; one fava bean; and a garlic chive and two huge French sorrel plants at
the end.

The fava bean is one representative of a planting I have at a friend's house. The
chickpea is the most cold-hardy orange popbean from last year's overwintering
trials. When I have plantings elsewhere, I often transplant one representative to
my home garden, where I can better watch it and notice things. But that may be
an excuse. One chickpea growing in my home garden provides me with more
pleasure than a thousand growing elsewhere. I'll be moving the garlic chive and
sorrel plants this year so I can expand the pea planting area. They're holdovers
from a larger perennial planting.

At the north end of the patch is a pile of leaves I dumped there last winter.
Underneath the leaves, undoubtedly, are a lot of large night crawlers (8 inches or
more long) busy softening the soil. When I expand the patch later this year, they
will have done most of the work.

I planted the daylilies last spring and have been redigging and planting around
them since. This year they should all flower, and I'll taste them all. Then I'll lift
the winners, move them to another place, and discard the rest. This bed is not a
good one for perennials because it is between two trees whose roots move up
into the patch if it is not completely redug every year or two. The redigging
usually doesn't happen with the whole bed at once, though.

Last fall I removed the pea vines and planted a combination of brassicas over
most of the bed (around the daylilies). The bulk of it was of `Green Wave'
mustard and `Ragged Jack'PRed Russian' kale, because the primary purpose was
food production and those are my favorites. The `Red Russian' I use mostly in
salads. It's a kale of extraordinary sweetness and flavor - unlike any other; it has
no cabbagey or bitter flavor. Its blue-green leaves and purple stems also are
beautiful. The rest of the brassica planting included trials of various unfamiliar
brassicas, along with small plantings of a few other familiars. I ate most of the
brassicas in the winter and early spring, leaving just a few red kales for salads
and a few other individual plants I might want for various crosses.

The daylilies I selected are almost all intensely yellow or intensely red,
because these are the colors I most want to add to my salads. Three of the
varieties flowered last year. Of them, one small red was much more beautiful to
me than my standard tawny but had pretty much the same flavor. A large yellow
was more beautiful than the tawny but had an inferior flavor. A huge, beautiful,
red tetraploid had a fine, lovely, sweet flavor far better than that of the tawny,
and its petals are thick, with a crisp texture. It is something special.

I don't expect to be able to evaluate the productivity, or even the earliness, of

the daylilies under my conditions. Some are under the tree, and others aren't.
They are all being invaded by tree roots to various extents, and I disturb their
roots to various extents as I dig around them to plant other things. My trials are
directed only at two questions: flavor of petals and beauty. Once I have a number
of winners in flavor and beauty, I'll evaluate productivity and other
characteristics under other circumstances. I want to know about size, flavor, and
production of tubers, too, for example. I also want to know which varieties can
naturalize here - that is, which grow and produce vigorously in an unwatered,
untended bed.

As I mentioned earlier, the peas are planted in a wide row down the center of the
bed. I redug that area and trimmed back the daylily roots before planting. In
most sections the rows are five peas across, spaced about 11/2 inches apart in all
directions. In some sections the row contracts to fewer peas because of the
daylilies (more lovely asymmetry). Down the center of the bed near the ground
run five wires, which evolved from one wire.

I used to use one wire down the middle as the bottom support for the trellis.
But our neighborhood has a lot of cats, which relish digging in fresh gardens.
For planting unique material, excluding cats is essential. I began adding extra
wires to exclude cats. Five wires running the entire length of the pea row at a
11/2-inch spacing is effective in discouraging them. (The wires are about an inch
above the surface of the ground.) Initially I was stringing the additional cat-
suppressing wires after planting, but I soon discovered that it made more sense
to string them before planting and use them as guides. And I no longer use any
of them as support for the trellis. They are now strictly cat suppressors and
planting guides.

With the five wires down the center of the patch and wooden stakes at intervals
along it, I have many short rows laid out on a grid. This grid allows me to
identify every single pea seed by location. That is very helpful in a situation
where every plant is likely to be different and where I will be evaluating and
making records about each. In this case it requires little extra work, since
something has to be done about cats anyway. The five wires work better than
covering the whole area with a strip of chicken wire because they are easier to
string and you can just leave them there as the peas come up. In addition, they
don't interfere as much with weeding.

In my neighborhood, it's useful to have a trellis down the center of beds like
mine even if you don't have anything that needs one. Without it, grown-ups walk
and kids and dogs run through. Some obvious barrier is necessary. For this
reason I put the trellis up as soon as possible after planting, to protect the bed
from gratuitous footprints and rearranging.

There is a modification to the bed to take account of my need to get at the peas
for evaluation and breeding - the 2-by-2-inch boards running across the patch.
They rest on the edge supports for the bed, not on the ground; they run under and
are not attached to the trellis. They're new for me this year. They are a footrest, a
place where I can put one foot up and into the patch so as to reach things better.

I presoak my pea breeding material indoors, so that it emerges very quickly

and is less vulnerable to slugs. In addition, I can avoid taking up planting space
with dead seeds.

The bulk of this year's (1992) pea planting is aimed at evaluating the disease
resistance of the material segregating out of different crosses. In addition to the
segregating material, the planting includes both parent varieties as controls. One
parent, `Sugar Snap', is sensitive to both enation and powdery mildew; the other,
`Oregon Sugar Pod IF, is resistant to both. I have some other known sensitive
and resistant varieties in the patch as well (a few seeds here and there). I don't
expect to be able to evaluate the earliness or yield of my peas from this planting,
given the variable amounts of shade and competition with tree and daylily roots.

I have other pea projects in the patch. There is an F, of a cross between a

purple shelling pea and an edible-podded variety. That's the beginning of an
attempt at getting purple snow and snap peas. Then there is a yellow-podded
snow pea, `Golden Sweet'. It has good-quality pods that are on the small side,
'and it is not disease resistant. The plants are yellowish green, the pods yellow,
the flowers two-toned maroon. The entire plant is beautiful, especially when
distributed among standard types with green plants and pods and white flowers.
As far as I know, `Golden Sweet' is available commercially only from Peace
Seeds. Alan Kapuler says it is one of the varieties that Mendel used.

I love the beauty of `Golden Sweet', but I would prefer better disease
resistance and bigger pods, so I'm planning some crosses with it. I also used
`Golden Sweet' as a separator between some of the other plantings. It makes an
especially good separator because it is easy to distinguish the plants as well as
the flowers and pods from all others. I added a `Golden Sweet' seed here and
there throughout the rest of my planting, too, wherever I felt the extra touch of
color would be appropriate.

Then there is a seed increase and evaluation on another purple-podded variety.

And there are some plants that represent the start of some crosses to obtain tall
forms of certain shelling pea varieties.

The plantings of the controls is material I get to eat and need to be able to
distinguish especially easily, so they are set off clearly. In

some cases tall varieties alternate with short ones. In other cases snow peas
alternate with snap peas. In yet other cases a row of `Golden Sweet' serves as a
divider. The purple variety for seed increase is at one extreme end of the patch.

I have about eight pea breeding projects going on in this one patch, a little bit
of pea food production, the daylily trials, various plants for other projects, and a
good bit of red kale, mustard, and sorrel for my salads. And then there are the
aesthetic values. The bed runs along a busy sidewalk. Small trials I've done with
yellow and purple peas have aroused intense interest from passersby. And I've
responded to that. This pea patch, I am determined, will be both my best and my
most beautiful.

To develop disease-resistant snap peas, I have done crosses followed by

inbreeding and selection. To combine yellow or purple pod color with snow and
snap peas with large pods and disease-resistant plants, I'm doing an initial cross
followed by a very powerful technique called recurrent backcrossing. The wild
mustards will be used in wide crosses with domesticates to try to develop whole
new crop species. The mutant onions are an example of noticing and using
happy accidents. These projects and much more are all going on in one tiny 4-
by-18foot bed between a sidewalk and a busy street.

 HY SHOULD we bother learning any genetics when the
amateurs of the past accomplished so much without it? Early amateur plant
breeders had no idea of the genetic basis for what they were doing. But there
were thousands of them operating over hundreds or thousands of years.
Generations of those amateurs developed new crop varieties. Most individual
breeders probably achieved nothing. In addition, when significant advances did
occur, they were undoubtedly more a result of luck than of skill. Nevertheless,
crops were developed and improved, if only because there were so many
breeders and so much time.

This book is not about how thousands of amateur plant breeders can develop
new varieties given hundreds or thousands of years. It is about how individual
amateur plant breeders can develop new varieties in one year or decade. To be
able to do that, we need to know more about what we're doing than our ancestors
did. We need to know something about the genetic basis of plant breeding.

What we need, however, differs from what is usually presented in beginning

biology or genetics courses in high school or college. Those courses present the
facts and concepts that are most important to our overall understanding of
biology, not those that are most relevant to the practical business of breeding
plants. In addition, they present nothing in enough depth so that we can actually
go out and breed plants much more efficiently.

This book does the opposite. It extracts out of modern genetics only those
concepts that are the basis for plant breeding. It covers those few essentials well
enough so that we can go right out into the garden and use them. With this
background the individual modern amateur should be able to accomplish more in
a few years than thousands of amateurs of yesterday put together could achieve
in their entire lifetimes.

Only ten methods have accounted for more than 90 percent of the
accomplishments of plant breeding, both ancient and modern. They are finding
and evaluating germplasm, maintaining germplasm (seed saving and clonal
propagation), selection, making F, hybrids, using Fzs, inbreeding, outcrossing,
backcrossing and recurrent backcrossing, using new mutations, and doing wide
crosses (to generate entirely new species). You can do them all. This book covers
them all. Of these methods, maintaining germplasm and selection are the
foundation for everything else. They, in turn, are based on understanding the
essentials of plant reproduction. This chapter introduces genetics, plant
reproduction, and selection.

The Basic Observation

The reason that those thousands of amateurs of yesterday managed to

accomplish anything at all is because they were doing something right. They had
noticed something important, even though they didn't understand it. I call it the
basic observation.

This basic observation is that offspring tend to resemble their parents. For
thousands and probably tens of thousands of years, people both noticed and
applied it.

They kept puppies from their favorite dogs, saved seed from their best plants,
and were suspicious if children were not at least vaguely similar to their parents.
Long before people knew anything about the physical basis for inheritance, they
noticed that "like breeds like," and they used that observation to improve their
domesticated plants and animals.

When we save seed from the best plants so as to maintain or improve an old
variety or to develop a new one, we call it selection. We are selecting the best
possible parents for the next generation of seed. We do it because we believe in
the basic observation; we believe that offspring from the earliest parents are
more likely to be early, those from the highest-yielding parents are more likely to
be high yielding, and so on.

Until recently, crude selection - selection based on nothing other than the basic
observation - was the only deliberate method of plant breeding. Combined with
accidental crosses, occasional random new mutations, and a certain amount of
luck, however, it created our domesticated food crops, nearly all of which have
existed in essential form for hundreds or thousands of years. Most vegetable
crops have been improved significantly in the past two centuries, but nearly all
were originally developed hundreds or thousands of years ago by amateur plant
breeders who relied exclusively on "like breeds like" as their only genetic fact
and framework.

The basic observation is only statistically true, however. Offspring do tend to

resemble their parents, but many times they do not. Like does breed like -
sometimes. Selection, alas, works only in certain situations; it is often useless. In
addition, other plant breeding methods are superior to selection in certain
circumstances, or they must be used along with it to achieve maximum
effectiveness.

The basic rule of selection - to keep seed from plants that have the most
desirable parents - is still at the core of plant breeding. But saving seed from the
best parents is sometimes more complicated than it seems, because with plants it
isn't necessarily obvious who the parents are.

Plant Parenthood

A profound advance in plant breeding came about with the realization that
most plants, like most people, have fathers. With plants, however, as with
people, it may not be entirely obvious who the father is. Flowers are the
reproductive organs of the plant. The plant that produces a seed is clearly the
mother of the seed. You can watch the seedcontaining fruit develop from the
flower on its mother. That the seed has a father isn't necessarily obvious.

Human understanding of plant reproduction was undoubtedly complicated by

the fact that plants are much more sexually diverse and versatile than people.
Humans come in two and only two morphologically distinct sexual types - male
and female - and they have only one pattern of reproduction. A male and a
female can mate to produce offspring. Neither can reproduce by himself or
herself.

Spinach plants and humans have a lot in

common. Like people, spinach plants come in two sexes that are
morphologically distinct and have different reproductive roles. Human females
produce eggs, which remain inside their bodies and are fertilized by sperm.
Human males produce sperm, which they deposit inside the reproductive tract of
the female. The sperm swims its way to the egg and unites with it, thus
combining the genetic contribution of the mother with that of the father to
produce a fertilized egg (the zygote). The zygote develops into a baby human
inside the mother's body.

Spinach females produce only female flowers - that is, egg-bearing flowers.
They don't have male flowers - that is, pollen-producing flowers. The eggs of a
spinach female remain in the ovary of the flower and are fertilized by pollen
from spinach males. Fertilized spinach eggs develop into spinach embryos in the
female plant.

Spinach eggs are specialized for staying where they are and producing an
embryo after fertilization. Male spinach plants produce flowers that look very
different from those on spinach females. The male flowers don't produce eggs;
they produce pollen, which, in spinach plants, is dustlike grains specialized for
being blown around by the wind. Would-be spinach fathers produce a lot of
pollen, because only chance determines whether any individual pollen grain will
land on your car or on the ground as opposed to on a stigma of a female spinach
flower. When a spinach pollen grain lands on a receptive stigma, it grows into
and through it, down the style and into the ovary, where it fertilizes the egg.

Spinach eggs and human eggs are essentially equivalent. Both are very large
cells that contain the genetic material of the female along with all the machinery
and raw materials necessary to translate genetic material into an embryo. Sperm
and pollen also are similar. Both are very small cells that contain the genetic
contribution of the male. They contain some of the information for making an
embryo but none of the machinery and raw materials required. Both specialize in
mobility. The sperm has a tail and mechanisms for swimming. Pollen has
specializations that permit it to be transported by wind, insects, or other means.

Plant seeds, however, are not really equivalent to human babies. A seed is
more than just a plant embryo, a baby plant. It is an embryo carefully packaged
with a sophisticated food supply adequate to allow development up until the
embryo is able to survive on its own. It often contains mechanisms that hold the
embryo in a state of suspended animation until circumstances are right for its
development.
 Plants such as spinach that have morphologically distinct sexes are known as
dioecious plants. Some other domesticated dioecious plants are asparagus, yam,
date palm, papaya, pili nut, carob, persimmon, fig, hop, mulberry, pistachio,
jojoba, hemp/marijuana, and muscadine grape. The dioecious pattern of
reproduction is not very common among crop plants, however familiar it seems
to us. Most plants are hermaphrodites. They have both female and male sexual
parts, produce both eggs and pollen, and can function as both mothers and
fathers.

Some hermaphroditic plants, such as bean, pea, tomato, lettuce, and cabbage,
have hermaphroditic (perfect) flowers - flowers that have both male and female
parts. Other plants, such as squash, muskmelon, watermelon, and corn (maize),
are hermaphroditic as plants but have separate, unisexual male and female
flowers. We refer to such plants and such a flowering pattern as monoecious.
Glenn Drowns's watermelons, for example, had separate male and female
flowers, but each plant had both. Corn has tassels (the pollen-producing male
flower) at the top of the plant and female, silk-producing flowers that become
corn ears lower down.

Dioecious plants are the extreme case of outbreeding plants. Like humans, the
only way they can produce offspring is by a mating of two different individuals,
a male and a female. Pea plants, with their perfect flowers, are a good example
of the opposite breeding pattern. A pea flower has both male and female parts
and produces both pollen and eggs. In peas, however, the stigma of the pistil
usually becomes receptive and pollen is usually released and lands on it before
the flower opens. So pea flowers nearly always pollinate themselves. A pea still
has both a father and a mother, but usually the same plant is both. Pea is thus
known as an inbreeding plant. Pea, bean, tomato, and lettuce are among the
vegetables that are primarily inbreeders - that is, primarily selfpollinating.

Some hermaphroditic plants are primarily outbreeders, even though they can
function both as male and female. In cabbage, kale, and many other brassicas,
for example, an incompatibility system usually prevents selfpollination. Any
individual flower functions both as a male and a female and produces both eggs
and pollen. But any given seed usually has two parents, not one. Onion also is a
hermaphrodite, but it is protandrous - that is, the male parts of the flower
develop first and release pollen long before the female parts finish developing
and become receptive. Onion is primarily an outbreeder, but it is self-compatible.
Even though a flower is normally pollinated by some other flower, it can be
another flower on the same plant.

Squash, cucumber, and melons, with their separate male and female flowers,
are primarily outbreeders. But as with onions, nothing prevents a female squash
flower from being pollinated by a male flower on the same plant.

A final major mode of reproduction is important to many plants and alien to

humans. Many plants can clone themselves. They can reproduce a new plant
vegetatively by sending up a shoot from a root fragment or cutting, or by
sending out runners (like strawberries do). Such clones are genetically identical
to the plant from which they were derived. In a sense, they have neither mothers
nor fathers, since they do not arise by sexual reproduction. Any plant normally
grown by planting tubers (or tuber pieces) or cuttings or started from runners is
being reproduced vegetatively instead of sexually - that is, it is being cloned.
Examples include potato, sweet potato, yam, strawberry, and most fruit trees.

In summary, many plants can clone them

selves as well as reproduce sexually, most plants can function as either mothers
or fathers, and the seeds of many plants may have either one or two parents.

Basic Selection

Only for the past two centuries or so have people understood what pollen is
and what it does. Before then, there was no such thing as deliberately crossing
two different plants to combine their desirable characteristics. Any crossing that
happened was a function of accident. There was, however, isolation of different
varieties to keep them pure. The necessity for it must have been established
empirically. The American Indians, for example, grew different varieties of corn
in separate fields. They had come to recognize the need for isolation in
preserving different corn varieties, even though they did not understand what the
corn pollen blowing all over the place had to do with it.

Selection practiced without any understanding of plant reproduction often isn't

very powerful, because if you don't understand the role of pollen, when you try
to save seed from the best parents, you are actually only choosing good mothers.
Each seed has a father, too. In the species that are largely inbreeding, when you
select good mothers, you are automatically selecting good fathers. But for
species that are largely outbreeding, when you select good mothers, all you have
is good mothers. You don't know who the fathers are; they might be undesirable.

Suppose, for example, that you have a patch of melons of a variety whose seed
you are saving. The variety is one of your favorites, but you would prefer a
somewhat larger fruit size. So you try to select for large fruit by saving seed
from the largest melon in the patch. When you grow the seed out the next year,
the melons aren't particularly big. In fact, they don't even look like the same
variety. Something has gone wrong.

Actually, several things have gone wrong. You were trying to use the basic rule
of selection by saving seed from the best, but you made four important but
common mistakes. First, you assumed that the melon is the offspring. It isn't. It
doesn't result from fertilization with the pollen; it is just part of the mother.
Second, you assumed that the seed would carry genes that would generate
melons like the one they started out in. But they don't necessarily, and they
didn't. Third, you didn't choose the best father; you didn't know who the father of
your seed was. Finally, you didn't necessarily even choose the best possible
mother plant.

The flower of a plant is part of the mother plant. It develops into a fruit that is
exclusively maternal tissue except for the seed inside it. The fruit - the melon,
tomato, or whatever - is part of the mother plant, just like the leaves and roots
are. The appearance of the fruit is determined only by the genes and environment
of the mother. It is the same, whatever the characteristics of the plant whose
pollen fertilized the flower from which it developed. Pollination affects only the
seed inside the fruit, not the fruit.

Actually, parts of the seed also are maternal tissue. It is the embryo itself
(within the seed) that is the new individual. So if you pollinate a melon flower
with pollen from a male of a different variety, the melon will look the same as it
would have if the flower had been selfpollinated. In most cases the seed also
looks the same. You can't tell anything by looking at the melon or the seed.
When you grow the seed up, the plants that result are derived genetically from
both the mother and the father. If there has been a cross, the plants will probably
look different from the mother and produce melons that are different from the
ones they came from.
 In some cases in some crop species, the seed may look different if there has
been a cross to a genetically different parent. In peas and beans, for example,
crosses sometimes show up as differences in the first generation of seed. This is
because the huge cotyledon leaves of legumes are the main food storage organ of
the seed, and the cotyledons are part of the embryo. Not all crosses show up in
the seed even in legumes, however.

In the example I gave earlier, you saved seed from your largest melon, but the
next generation produced fruit that looked totally different from the mother. The
flower that gave rise to that fruit was probably pollinated by a bee that had just
visited a flower from a completely different variety - one growing in your
neighbor's garden, for example.

Whenever we try to save seed or do selection, we need to choose both the

female and the male parents. This means that we need to control pollination. In
strongly inbreeding crops such as pea, we don't actually have to do anything,
because left to her own devices, the mother plant also contributes the pollen. So
if we choose a good mother plant, she is also the father plant. Selection is very
easy with inbreeding plants; we just save seed from the best plants.

With outbreeding plants, however, whenever we maintain a variety, practice

selection, or do any kind of plant breeding, we have to deliver the desirable
pollen to the plant and exclude unwanted pollen. This means that before we can
maintain any crop variety or develop any new ones, we need to know whether
the plant we're working with is an inbreeder or an outbreeder. That information
is included in Appendix A and Table 1 for all the common vegetables. In chapter
9 1 explain how to figure it out for yourself in cases where no information is
available.

The final mistake you made in trying to select for large melons by saving seed
from the largest melon was that you didn't necessarily even choose the best
female parent. The size of a melon is affected very much by whether it is the first
melon on the plant or a later one, for example, and by how many melons there
are all together. Suppose your plant had difficulty setting fruit under your
climatic conditions and only managed to set one fruit. That fruit would be much
larger than fruits from sister plants that set several. But would you want a variety
that produced so few fruits, whose ability to set fruit at all was so marginal?
 When we practice selection, we need to consider the entire plant, not just the
fruit on the plant. We would choose a plant whose overall performance came
closest to what we wanted and then either self-pollinate it or cross it to another,
superior individual of the variety. (Which we do depends on considerations I'll
cover later.)

When we act positively and save seed from only certain plants and eliminate
the rest, we call it selection. But sometimes we do the opposite. We save all the
seed except that from the worst plants. This is selection also, but it is usually
called culling or roguing and refers to eliminating the rogues, or bad plants.
Normally there are inferior plants in every generation. When we maintain a
variety, we try to eliminate seed from the inferior plants. In an outbreeding crop,
we do the roguing before the plants flower, if at all possible, so that the inferior
plants don't contribute to the next generation as either maternal or paternal
parents.

In a home garden, we usually don't discard the rogues; we usually eat them. In
a corn patch, for example, we don't have to cull out whole plants and lose their
production. Instead, we can remove the tassels from the inferior plants so that
they cannot produce pollen; then we eat their ears instead of saving them. In
such cases, we always need to consider contamination from other corn patches.
Sweet corn pollen, for example, travels long distances. For this reason most
home gardeners who save their own corn seed must hand-pollinate it.

There are other aspects of selection. Selection doesn't always work. It depends
on the existence of genetic variability for the characteristic in which we are
interested. If there is no genetic variability for fruit size in our melon variety, we
can select year after year and achieve absolutely nothing. We would need to use
other methods.

Selection is both the simplest and the most sophisticated of genetic methods.
In this chapter I've outlined the essentials. I'll take up selection again later. At
that point you will learn when to use selection and how to use it most effectively.
You also will be able to do power selection instead of just basic selection. In
addition, selection is just

one of the ten major plant breeding methods. Most of the other methods,
however, depend on having a better understanding of genes and how they behave
- which is the subject of the next chapter.

 EOPLE HAD the idea that there was such a thing as heredity long before
they knew anything about the physical basis for it. They knew that offspring
tended to resemble their parents, and they supposed that there must be some
hereditary material that offspring got from their parents that caused the
resemblance. (The terms heredity and genetics are equivalent.) Early ideas about
the nature of the hereditary material seem to be that it was some kind of fluid.

It's easy to see how the fluid concept of heredity arose and why it still seems so
intuitively reasonable to the nonspecialist. People probably noticed that when a
person with dark skin married one with light skin, their children's skin color was
generally somewhere in between that of the two parents -

a result that seems analogous to what happens if you mix two different shades of
inky water, for example. Likewise, the offspring of a tall person and a short
person are most often somewhere in between the two in height. And when two
dogs from breeds with different conformations are mated, the pups usually have
a conformation that is quite obviously an average of that of the two parents.
Early theories about the mechanism of inheritance were based on a model that
involved the blending of the genetic fluids from the two parents. Words such as
halfbreed, half-blood, and pure-blood are based on that ancient assumption that
the hereditary material was some kind of liquid or blood, perhaps even
associated with the blood.

Once scientists understood the role of pollen in plant reproduction, they began
to try to decipher the mechanisms of inheritance by doing deliberate crosses of
different varieties of plants. Sometimes the results seemed to fit in with the
concept of blending inheritance. The offspring sometimes did seem as if they
were some kind of blend of the two parents. When a tall variety was crossed
with a short variety, for example, the hybrid often was in between in height.
When an early variety was crossed with a late one, the hybrid often was
intermediate in timing. When white-flowered varieties were crossed with red-
flowered varieties of the same species, in some cases the offspring were pink.
These particular cases fit in with the concept that the genetic material was a fluid
of some kind and that a hybrid between two pure breeding varieties would blend
the genetic fluids and therefore be intermediate between the parent varieties.

But hybrids weren't always intermediate between the two parents. For every
case where a cross between two varieties gave an intermediate hybrid, there
were other cases where the hybrid resembled just one of the parents. And where
there were a number of different characteristics involved in a cross, the hybrid
might be like one parent for some characteristics, like the other for others, and
intermediate for still others. In some cases the hybrid was even more extreme
than or entirely different from either parent. Sometimes a cross of two varieties
with white flowers would result in offspring that had purple flowers, for
example.

In addition, in later generations after a cross, characteristics of the parents that

had not appeared in the hybrid could suddenly reappear in some of the progeny
in the later

generations. These progeny were called throwbacks. The word is meaningless

now, but it reflects the fact that, according to a blending theory of inheritance,
there was no way to explain the sudden appearance of progeny with
characteristics that were not present in the immediate parents.

The scientists of the day simply could not explain or understand what was
going on when they crossed different plant varieties given the conceptual
framework they had for the nature of the hereditary material and the mechanisms
of inheritance. This did not mean that they were eager to hear alternative
explanations from amateur-scientist monks, however.

Gregor Mendel was an Austrian monk. He is considered to be the founder and

father of genetics, although, as far as I can tell, he had exactly zero impact on the
field. He was just too far ahead of his time. The theories he proposed were
correct, but there was no basis for understanding why they were correct. The
framework of facts and concepts that would have made his assumptions
acceptable and his theories logical didn't exist yet. Not enough was understood
about mitosis and meiosis (asexual and sexual cell division).

Forty years later, after people had better microscopes, could see chromosomes,
and had come to understand cell division, there was a physical mechanism to
explain and justify Mendel's assumptions. At that point, in the early 1900s, three
other scientists did work that was essentially equivalent to Mendel's. None knew
anything about Mendel and learned about his work only during the literature
searches they did as part of publishing their own work. So Mendel, the father of
genetics, had no influence on the course of any science, including the genetics he
is considered to be the father of. The three independent discoverers of Mendel's
laws had no trouble getting their work accepted, since the framework for
understanding it existed by the time they came along.

Mendel did a lot of breeding work with peas. It was a fortunate choice because
peas are almost totally selfpollinating. He had a number of different pure
breeding varieties - tall ones, short ones, ones with yellow pods, ones with green
pods, and so on. He made a number of crosses between varieties that differed by
just one characteristic. In each case he collected and planted the F1 seed and
examined and recorded the appearance of the F1 plants. F1 stands for "first
filial" generation. We use the term F1 or F1 hybrid to refer to the first-generation
hybrid made by crossing two different pure breeding varieties.

Mendel took things one step further than many others. He also considered the
F2 generation. That is, he planted seed from the F1 hybrid plants, the Fz seed.
And he examined and recorded the appearance of the F2 plants. The F2 is the
"second filial" generation, or the second generation after a cross between two
pure breeding varieties. It is obtained by crossing two F1 plants. In the case of a
selfpollinator such as peas, you get an F2 generation just by letting the F1 plants
produce seed.

Mendel had two pure breeding varieties that differed in height, for example.
The tall variety always produced tall plants, and so

did their offspring. The short variety always produced short plants. Mendel did
crosses between plants of the two varieties and saved and planted the seed (the
F1 seed). He recorded the appearance of the F1 plants and allowed them to set
seed (the Fz seed). Then he planted the F2 seed and recorded the appearance of
the FZ plants.

Mendel found that the F, hybrids between the tall and short variety were all
identical to each other. Furthermore, they were all tall - every bit as tall as plants
from the tall variety. It was as if the short parents made no hereditary
contribution to the F1 hybrid at all. (According to a blending concept of heredity,
the F, plants should have been intermediate in height.)

When Mendel planted the F2 seed and examined the plants, he found further
unexpected results. The F2 plants were of two distinct types. Some were tall, and
some were short. There was nothing in between. There were 787 tall plants and
277 short ones. That is, about three-quarters of the F2 plants were tall; about
onequarter were short. The tall plants were just as tall as plants from the pure tall
variety, and the short plants were just as short as those from the pure short
variety.

Sometimes Mendel did the original cross using a tall female and a short male.
Other times he did the reciprocal cross (short female x tall male). The results
were exactly the same for all crosses, in whichever direction they were done.

Mendel proposed that inheritance was, basically, an either/or sort of thing.

Tallness and shortness stayed separate, segregated, from each other. They didn't
seem to blend or influence each other at all. (This phenomenon is often referred
to as the Law of Segregation, or Mendel's First Law.) The hereditary material
can't be fluid at all, Mendel said. It must come in units, or factors - lumps, so to
speak - and these factors could exist in two different forms (alleles) that were
inherited as alternatives to each other. Today the factors are usually called genes.
The word factor is still used in certain contexts, though.

Mendel proposed that his two original varieties differed by one gene that
influenced height. That is, the two varieties had different alleles for a gene
involved in determining height.

The F, hybrid plants looked exactly like the plants of the tall variety, but it was
clear to Mendel that they were not genetically identical. They couldn't be, since
what they could pass on to their offspring was different. The tall variety was
pure breeding for tallness. All the offspring were tall, and all the offspring of
those offspring were tall. But the tall F1 wasn't pure breeding for tall. Some of
the offspring of tall F1 plants were short. So the tall F1 must be carrying genetic
material that could determine shortness, even though it wasn't expressed in the
F1.

The numbers of the two different classes of the F2 also were very interesting.
About three-quarters of the F2 plants were tall, and about onequarter were short.

Mendel did six other crosses that produced similar results. When he crossed a
green-podded variety with a yellowpodded variety, for example, the F1 all had
green pods. Then the Fz fell into two distinct classes that resembled the two
parent vari

eties. There were 428 green-podded plants and 152 yellowpodded ones. Again,
the F, resembled just one of the parent varieties, but there were two discreet
classes in the F2 that resembled the two parents. And the numbers approximated
three-quarters of the total for one class and onequarter for the other.
Furthermore, whatever the F1 looked like was the majority class in the F2. The
same kind of results, with the same disappearance of one of the parent classes in
the F1 and same reappearance of it as onequarter of the F2, occurred in all
Mendel's other crosses as well.

Mendel apparently knew enough math to realize that there is something special
about the number 1/4; it's what you get when you multiply two 1/2's. And 1/2 is
really special. It's the probability associated with a random choice between two
alternatives that are equally probable. (Random, in this context, means "by
chance.")

Example: If you flip a balanced coin, it is just as likely to land heads as tails.
That is, there are two possible results, and the choice between them is random,
so the probability of getting a head is 1 in 2, or 1/2.

Suppose you flip two coins. What's the probability of both being heads? Well,
the first one will be heads half of the time, and of that half of the time, the
second coin also will be heads half of the time. So the probability of getting two
heads at once is (1/2 X 1/2) or 1/4. The generalization is called the Product Law
of Probability, and it comes up over and over again in genetics. It states that if
you know the probability of two events occurring separately, and they are
independent events, then the probability of their occurring simultaneously is the
product of the probabilities of their occurring separately. Example: If you toss a
coin and one die, what's the probability that the coin will be heads and the die
will be a three? Answer: The probability of a heads is 1/2. The die is equally
likely to land on any of six sides, so the probability of getting a three is 1/6. The
probability of getting a heads and a three simultaneously is one-half times one-
sixth, or (1/2)(1/6), that is, 1/12.

We also use the Product Law of Probability to describe all possibilities and
their probabilities. When you toss one coin one time, all the outcomes and their
probabilities can be represented as (12H + 12T), where H stands for heads and T
stands for tails. The expression (12H + 12T) shows that there are only two
possible results from one toss of one coin and they are equally probable.

If we toss a second coin, it, too, can be expected to land as (12H + 12T). If we
want to know all the possible outcomes of the toss of two coins, we can simply
multiply the two expressions. That is, all the possible outcomes of the toss of
two coins and their probabilities will result when we multiply ('2H + 12T) by
(12H + 12T). That multiplies out to (114HH + '4HT + 14TH + 1/4TT). In most
cases the two middle terms look identical, because we don't care which coin
landed heads or tails, only that we got one of each. So the expression becomes
(114HH + 1/2HT + 'TT). That is, onequarter of the time we get two heads,
onequarter of the time we get two tails, and one-half of the time we get one of
each.

In case your math is rusty, expressions such as ('4H + 14T) or (a + b) that have
more than one term are called polynomial expressions. You multiply two
polynomials just by multiplying each term in one by each term in the other and
adding all the resulting products. So (a + b)(c + d) works out to (ac + ad + be +
bd). Some other examples: The product of (a) and (b + c) would be (ab + ac).
The product of (d + e) and (f + g) would be (df + dg + of + eg). The product of
(h + i) and (j + h + 1) would be (hj + hk + hl + ij + ik + il). And (m + n + o)(p +
q + r) would multiply out to (mp + mq + mr + np + nq + nr + op + oq + or).

I don't use much math in this book, but you often will want to manipulate
polynomials to figure out the probability of getting progeny of some specific
desired type from a given cross or series of crosses. This tells you how many
offspring (seeds) you need to plant to get what you want. In other cases you can
use such estimates to evaluate whether a specific plant breeding approach is
likely to work at all. The two are often related. If you calculate that you need to
grow ten thousand of something to get one of what you want and you only have
room for twenty, you know that you will need to use some other approach.

To go back to coin tossing, if we toss two coins, what's the probability of

getting at least one heads? We can see from the (1/4HH + 1/2HT + 1 TT)
distribution that it is 3/4. That is, the distribution can be collapsed to (3/4H_ +
1/4TT): three-quarters of one or more heads plus onequarter of no heads, or
three-quarters of one outcome and onequarter of the other. These are exactly the
same numbers that were showing up in Mendel's pea crosses.

Mendel supposed that there was some unit of heredity, some factor, some gene,
associated with the difference in height between the two varieties, and it had two
alternate forms that were different in the two varieties. He further supposed that
every pea plant had two alleles for the height gene in question, and they could be
the same or different. He yet further supposed that during formation of an egg, a
plant passed on just one of its two alleles, and it was chance as to which one it
passed on. Likewise, during formation of pollen, the plant would pass on only
one of the two alleles it had, and it would be random as to which.

When an egg was fertilized, the resulting zygote would get one allele from the
egg and one from the pollen, and would have two alleles. Fertilization was
equivalent to the simultaneous occurrence of two random and independent
events. Which of two different alleles went into an egg would be like whether
one coin landed heads or tails. Which of two different alleles went into the
pollen grain would be like whether another coin landed heads or tails. And
which combination of egg and pollen came together in a fertilization would be
like considering two coins tossed simultaneously.

Mendel used uppercase and lowercase letters to represent different alleles of

genes involved in his crosses. Using this convention, we can name the gene
involved in his tall x short crosses the T factor and the two different alternative
alleles involved in his cross T and t. The tall variety would have the genotype TT
and the short variety genotype tt. The original varieties were both pure breeding.
Another way of envisioning this is that they could pass on only one kind of gene
associated with height. We refer to the condition of the original varieties as
homozygous - that is, each had two identical alleles. The F, hybrid would get a T
from the tall parent and a t from the short parent, so it would have the genotype
Tt. It would be heterozygous - that is, it would have two different alleles.

The F, was tall, however. It looked the same as the original tall variety, but it
was genetically different. (It had to be genetically different from the parent
variety because it could pass on shortness to some offspring, but the tall variety
couldn't.) Mendel used the word genotype to describe the actual alleles present
and the word phenotype to describe what the plants looked like. Apparently, said
Mendel, the genotype Tt looks just like the genotype TT. And he coined some
terms. The T allele was dominant to the t allele; the t allele was recessive to the
T allele. The F, was phenotypically identical to the tall variety, but it was
genotypically different. (The terms don't explain the whys of it; they just
describe the situation.)

Given the rules Mendel set up, what kinds of genotypes would we expect in
the F2, and with what frequencies? Well, the F2 seed is the progeny of a
selfpollination of the F, pea plant, which is of genotype Tt. The F1 plant has two
different alleles, so it can pass on either T or t to its eggs. Mendel assumed that
chance determined which allele the egg got, so half the eggs would get T and the
other half t. Likewise, the F1 pea plant could pass on either allele to its pollen.
Then either kind of egg could combine with either kind of pollen. So one could
get four different kinds of fertilization, and they would be equally probable.

To calculate all possible genotypes that can result from the cross and their
probabilities, we multiply the egg distribution by the pollen distribution. All
possible genotypes of eggs and their frequencies are represented by the
expression (1/2 T + 1/2 t). All possible genotypes of pollen grains and their
frequencies are represented by that same expression. So the expression for all
possible genotypes of zygotes and their frequencies is (1/2T + 1/2 t)('2T + 12t),
which gives (1/4TT + 1/4 Tt + 1/4tT + 1/4tt), which reduces to (1/4 TT + 1/2Tt +
114tt). In terms of phenotypes, this is 3/4 tall + 1/4 short.

According to Mendel's theory, the short plants should all be pure breeding. He
confirmed this by selfpollinating a number of them to obtain an F3. His theory
also specified that the tall offspring weren't all identical. Some should be
homozygous and pure breeding. Others should be heterozygous and should be
able to pass on either tallness or shortness. By inbreeding a number of tall plants,
Mendel showed this, too, was true. In this fashion, he showed that the (3/4 tall +
1/4 short) he obtained in the F2 was actually (1/4 pure breeding tall + 1/2 tails
that carried short + '/4 pure breeding short). He did the same work with his
crosses involving other characters and obtained the same results.

Mendel also did another kind of cross, which we now call a backcross. He
crossed tall F, plants back to the short parent. This cross can be represented as
(Tt x tt). The genotypes Mendel's theory predicts and their

frequencies are given by the expression (1/2 T + 1/2 t)(1 t). That would be (1/2
Tt + 1/2 tt), that is, one-half tall and one-half short. This is what Mendel
obtained. Furthermore, Mendel's theory implied that all the tall offspring of the
backcross should be heterozygous, and he showed that this, too, was true.

Mendel did another experiment. He took the tall offspring from a backcross
and backcrossed again. This cross is again (Tt x tt). It again gave one-half tall
and one-half short. According to a blending concept of heredity, these tails were
only quarter bloods, so to speak, with respect to the tall variety. But they were
every bit as tall as the original tall variety. And when backcrossed yet again, the
offspring were still one half tall and one half short. These talls were still as tall as
the original variety, even though they were now eighth bloods; and the shorts
were still as short as the short variety. There is no way to explain the results of
Mendel's serial backcrosses with a blending concept of inheritance.

Mendel also did various crosses involving two different traits at once, and he
found simple, predictable results in these cases also. When he crossed a tall
variety with green pods to a short variety with yellow pods, for example, the F,
plants were all tall and green podded. The F2 plants were (9/16 tall green + 3/16
short green + 3/16 tall yellow + Vic short yellow).

Mendel realized that his results were what you would expect if the inheritance
of the two traits were independent. In other words, a two-factor cross was just
two one-factor crosses occurring simultaneously. If so, to obtain the classes and
frequencies of both traits simultaneously, you just multiply what you would
expect of each alone. The FZ with respect to height is (3/4 tall + 1/4 short). The
F2 with respect to pod color is (3/4 green + 1/4 yellow). So the Fz considering
both simultaneously is (3/4 tall + 1/4 short)(3/4 green + 1/4 yellow). That
multiplies out to (9/16 tall green + 3/16 tall yellow + 3/16 short green + 1/16
short yellow).
 To portray these results in terms of genotypes, let's assign the letter G to
represent the allele associated with green pods. So the green-podded variety is
GG, and the yellowpodded one is gg. The cross of the tall green variety and the
short yellow one can then be represented (TTGG x ttgg). The results expressed
in terms of genotypes are (3/4 T_ + 1/4tt) (3/4G_ + 1/4gg), or (9/16 T_G_ +
3/16T_gg + 3/16ttG_ + 1/16ttgg).

Mendel performed all possible two-factor crosses with his seven traits, and in
all cases he found the same pattern. In the F2, 9/16 of the plants displayed the
two dominant traits and resembled the F1; 3/16 had one dominant trait and one
recessive; 3/16 had the other combination of dominant and recessive; and V16
had the two recessives.

In each of the two-factor crosses, four classes of Fz progeny were obtained.

Two classes resembled the two original parent varieties. But there were two new
classes, two classes in which the traits of the parent occurred in different
combinations. The cross (tall green x short yellow) produced an FZ that included
both the parental classes, tall green and short yellow, and also the recombinant
classes, tall yellow and short green.

Mendel's explanation of how two different traits are inherited in two-factor

crosses is referred to as independent assortment or recombination. (The Law of
Independent Assortment is Mendel's Second Law.)

The phenomenon of independent assortment or recombination of phenotypic

traits in the F2 forms the basis for much of plant breeding. Independent
assortment allows the breeder to cross two varieties that differ in a number of
ways and to derive new varieties with different combinations of the
characteristics of the parents, or even completely new phenotypes.

 HE GENETIC material of plants and other higher organisms is made
out of DNA, a long, chainlike, linear molecule composed of four different kinds
of subunits. DNA has two profound functions that are essential to its role as the
hereditary material. First, it self-replicates. Given the machinery in the cell, a
DNA molecule can synthesize a duplicate of itself, using itself as the template.
No other biological molecule can do so. Second, DNA encodes information. The
four kinds of subunits can occur in all possible orders along the DNA chain; the
order of the subunits encodes the information required to make proteins.

Proteins are one of the most important classes of substances in living things.
Structural proteins form much of the basic struc

ture of cells. Other proteins, called enzymes, are responsible for catalyzing
nearly all the chemical reactions that take place in cells - all the reactions that
allow cells to grow, divide, develop, and do all the things that cells do. In other
words, DNA, by encoding the information to make proteins, specifies everything
living things need to be and do to be living things.

Today quite a lot is known about the structure and function of specific genes.
But the genes that have been most studied and that are best understood are
mostly those in bacteria, viruses, yeast, and other laboratory organisms such as
mice, as well as those of medical importance in humans. The structure and
function of most genes in most plants is unknown. Peas, tomatoes, and corn are
the only vegetables that have been studied much genetically. Most other
vegetables, even common ones such as squash, are nearly complete genetic
unknowns. Even with peas, tomatoes, and corn, most of the agronomically
important genes are still unidentified.

In most practical breeding of vegetables, we manipulate and recombine genes

without knowing specifically what they do. We usually don't even know what
genes we're working with. It doesn't matter, though. The manipulation and
rearrangement of genes into new varieties using classic plant breeding methods
does not in any way depend on knowing the genetic or biochemical identity of
the genes involved. All it depends on is an understanding of the general
principles.

The other major profound characteristic of DNA, besides its ability to encode
information, is its ability to replicate. If DNA is to be the genetic material, it
must be able to replicate so that each time a cell divides and becomes two cells,
both can have a copy of all the genetic information they need.

In this book I don't go into the biochemistry of how genes code for information
or how they replicate. More about that can be found in any beginning biology or
genetics textbook. One aspect of replication does matter to plant breeders,
however - the fact that it isn't perfect.
 Mutation

DNA replication is usually very exact. If it were totally exact, we would not
need this book. There would be no plant breeding. Plant breeding involves
selecting for genetic

differences, and genetic differences are created by new mutations, or changes -

mistakes in the replication of DNA. We also wouldn't be here to read the book.
Evolution requires selection based on the existence of genetic differences.
Fortunately, DNA replication is very accurate, but not too accurate.

Occasionally when a gene replicates, a mistake is made. The frequency of

spontaneous mistakes - that is, spontaneous mutations - varies with the gene and
the kind of mutation. Mutations of specific types in specific genes commonly
occur between 1 per 10,000 replications and 1 per 100,000. Mutations are rare
when we think in terms of the replication of specific genes. But they are
common when we consider how many genes a complex organism such as a plant
or animal has and how many different ways they can mutate. Humans, for
example, are thought to have 50,000 to 100,000 genes. Most humans probably
contain at least one or more brand-new mutations - altered genes that neither of
their parents carried. And so do most plants.

Once a mutation occurs, the mutant gene self-replicates just as the original
gene did, so the change is passed along to offspring as an altered gene. In many
cases new mutations are lethal. They cause dramatic enough alterations of
function so that the organism cannot survive. In other cases mutations are viable,
but the enzyme involved is altered. There may be much less of it, or it may be
much less active than the original form of the enzyme. In many cases mutations
cause an alteration in one or more characteristics of an enzyme for which they
code. Two enzymes that differ from each other by a sin gle amino acid might
have somewhat different temperature optima, for example. That could translate
into a mutant plant that thrives best at a temperature higher or lower than the
original strain.

Mutation is a random process. If we want peas that grow better in the heat and so
grow an entire field of peas in the heat, we are no more likely to obtain a new
mutation that improves heat tolerance than with a whole field of peas grown
under moderate conditions. In the hot field, however, we would be better able to
identify any mutants with improved heat tolerance, because they would look
different (better than) the heat-sensitive plants. Under moderate conditions the
heatresistant plants would not necessarily be distinct. Growing the plants in a hot
field doesn't cause mutations to heat resistance, but it does allow us to recognize
any that might already be present in the seed we planted. The occurrence of the
new mutation that confers greater heat tolerance is a chemical accident to the
DNA and is in no way caused by or related to whether the pea plants are growing
under hot conditions.

Scientists using simple organisms such as bacteria and viruses have shown that
there is no such thing as directed or purposeful mutation. If I have a tall pea
variety and want a bush form, I might chop off a pea plant so that it bushes out
and then collect the seed. But my efforts to obtain a bush variety in that way
would fail. When I chop off the top of the pea plant, I don't make any change in
the DNA contained in all the cells of the plant. I have turned the individual plant
into a bush, but that bushiness isn't inheritable. The seed of the bush would still
be genetically pure breeding for tall, even though the parent was phenotypically
bushy.

If I want taller pea plants, there is no way I can treat or grow the plants to
create a mutation that will make the plants taller. That specific mutation isn't
likely to occur unless I have hundreds of thousands or even millions of plants.
But it is often the case that the mutation we need occurred at some time in the
past and exists within some variety of our crop somewhere. We learn how to find
such genes, get them out of the variety where they occur, and get them into new
varieties of our own.
 Genes and Alleles

We don't know the exact identity of the gene that was associated with the tall
versus short growth pattern in Mendel's peas. But this pattern in crop plants is, in
a number of cases, associated with differences in the synthesis of the plant
hormone gibberellin. The "tall" version of the gene is usually the form that is
found in the wild. The "short" version, in many cases, has a less active form of
one of the enzymes involved in synthesis of the hormone, so .the plants are
shorter. We refer to two genes as alleles of each other when they are inherited as
alternatives to each other. In molecular terms, alleles are different forms of the
same gene. There can be more than two alleles of a gene in a population of
organisms. But any given individual has only two alleles at the most.

Shorter plants usually cannot compete with the taller forms in the wild; a short
mutant in a patch of tall plants would be shaded out. That problem isn't relevant
when a human plants a patch or field with nothing but short plants. And short
plants may be earlier than tall ones, or less subject to lodging (falling over) in
rain or wind. They also may have a much higher proportion of grain compared to
the rest of the plant. So shorter plants can be advantageous as cultivated crops.
Specific mutations or alleles are not good or bad in and of themselves, but only
within a certain context. An allele that promotes better growth in hot weather
may promote inferior growth in cold weather, for example.

Many systems are involved in determining the size and shape of a plant.
Probably dozens of different genes in peas are capable of mutating so as to
create a tall versus short difference, and hundreds of genes probably influence
plant height to some extent. When we say that tall versus short was a one-factor
difference in Mendel's crosses, that doesn't mean that only one gene is involved
in determining height in peas. All it means is that the two varieties that Mendel
crossed differed by only one gene involving height. The other dozens or even
hundreds of genes involved in determining height must have been the same in
the two varieties he crossed.

Genes, Chromosomes, and Cell Division

Most plants and animals are diploid organisms. Diploid means that they have
two copies of every gene in each cell of their bodies. Genes in higher organisms
such as plants and animals are arranged in structures

called chromosomes. Each chromosome contains the DNA for many genes, in
most cases hundreds or thousands of genes. Each cell has two chromosomes of
each kind. Gametes (eggs and pollen) are an exception; they have only one
chromosome of each kind.

The pea genome, for example, is organized into seven chromosome pairs.
(Genome refers to all the DNA there is in an organism.) Each cell in a pea plant,
except for the gametes, has two chromosomes of each kind for a total of
fourteen. Each pea gamete has only one chromosome of each kind for a total of
seven. The letter n is used to indicate the number of different chromosomes a
species or variety has. The expression "2n = whatever" is used to indicate the
total number of chromosomes, or the diploid chromosome number. Peas are 2n =
14; tomatoes are 2n = 24; lettuce is 2n = 18.

The two chromosomes of a pair are said to be homologous and are referred to
as homologs. What makes two chromosomes homologous is that they have the
same basic genes in the same order. Some of the genes are exactly identical to
the corresponding genes on the homologous chromosome. Other gene pairs
aren't exactly the same; they are slightly different - that is, they are different
alleles of the given gene.

There doesn't seem to be much relationship, at least on a gross level, between

which chromosome a gene is on and its function. Three major genes influencing
height frequently show up in crosses of garden peas, for example. One is on
chromosome 2, one is on chromosome 3, and one is on chromosome 4. (The
numbers are just arbitrary names for the specific chromosomes.) Like wise,
genes involved in flower color and pod color are scattered among the seven
chromosomes. There are two major genes involved in the difference between the
pod characteristics of shelling peas and those of edible-podded snow-type peas.
One is on chromosome 4 and the other on chromosome 6.

The word chromosome refers to the fact that the structures absorb certain
stains that made them easily seen once people began to use microscopes. The
behavior of chromosomes during cell division is dramatic and corresponds
exactly to what one would expect for structures that make up the genetic
material. By the time Mendel's work was rediscovered, people knew
chromosomes behaved in a way that was exactly analogous to the patterns
Mendel outlined. The behavior of chromosomes in cell division explained
Mendel's laws completely.

Two kinds of cell division occur in diploid organisms - mitosis and meiosis.
Mitosis, also called somatic cell division, is the kind that occurs as the body of a
plant or animal grows. Somatic cells are ordinary cells, as opposed to gametes,
which are the sex cells. All vegetative growth - growth of shoots, leaves, stems,
roots, and tubers - is somatic growth and uses mitosis as the mechanism of cell
division. The other kind of cell division, meiosis, is restricted to sexual
reproduction. In fact, it is restricted to just certain cells within the reproductive
organs - those that divide to give rise to the gametes.

The purpose of mitosis is to produce two identical cells where before there was
just one. A mitotically dividing cell gives rise to two daughters that are identical
to it and to

each other. Each daughter cell has the standard quota of two of each kind of
chromosome (and, therefore, two of each kind of gene), just like the cell that
divided to give rise to it.

The purpose of meiosis is to eliminate one chromosome out of every pair so

that the resulting gametes have just one chromosome of each kind (and,
therefore, one gene of each kind). This is done by a special chromosome-pairing
process that is unique to meiotic division. During this process, each chromosome
physically pairs with its homologous chromosome. Then one chromosome of
each pair migrates to one pole of the cell, the other chromosome migrates to the
opposite pole, and a membrane forms down the middle. The result is two
daughter cells each having one chromosome of each pair. Meiosis is sometimes
referred to as reduction division because it reduces the total amount of genetic
material.

Any beginning biology textbook presents sequential pictures of the behavior of

chromosomes during mitosis and meiosis - usually in such detail that the
important differences get obscured. For our purposes we don't need the
mechanical details. All that matters is the overall result.

Somatic cells are diploid; gametes are haploid. When two somatic cells unite
as a result of fertilization, the resulting zygote is diploid again. Since somatic
cell division generates genetically identical cells, when we take cuttings from a
plant or grow a plant from a tuber, the new plants are genetically identical to the
mother plant and to each other. They are clones.

Gametes, however, are not identical to the mother cell from which they arose.
They have only one gene of each pair instead of two. In addition, in most cases
gametes that arise from a single mother cell aren't identical to each other either.
To understand why, let's go back to Mendel's laws. When we follow what
happens when we do a cross, it's easy to see exactly why all Mendel's laws are
the way they are and what is actually happening.

Mendel crossed, for example, a tall pea variety with a short one. We
represented the tall variety as TT and the short variety as tt. We represented the
F, as Tt, and we said that the genotypes and ratios expected in the F2 correspond
to (1 TT + 'zTt + 14tt).

Let's suppose for purposes of illustration that the T gene is on chromosome 2.

The tall parent has two chromosome 2s, and each has a copy of the T gene. As
the tall parent grows, mitosis occurs in the somatic cells, giving rise to more
cells with genotype TT. When the tall parent produces eggs and pollen, they will
have just one copy of chromosome 2, and their genotype will be T. Since both
chromosome 2s had identical alleles for the T gene, all the gametes are identical
with respect to the T gene.

The short parent, likewise, has two copies of chromosome 2, each carrying t,
which is allelic to (a slightly different form of) the T gene. The short parent's
gametes also will have one chromosome 2, but their genotype will be t.

When we pollinate a plant from the tall variety with pollen from the short
variety, we end up with a zygote that has two chromosome 2s, one carrying T
and the other carrying t. In other words, the F1 is genotype Tt.

Now we grow the F1 plants. As they grow from small plants into big plants,
the kind of cell division involved is mitosis, or somatic cell division. Through all
the hundreds of cycles of somatic cell division, the genotype Tt is replicated
exactly. All the cells in the Fl pea plants are identical to one another and have
genotype Tt. But now the pea plant starts making flowers and forming eggs and
pollen - conducting meiosis. In meiosis, the two chromosome 2s pair, and only
one or the other winds up in each gamete. Half the cells get the chromosome 2
with the T and the other half get the chromosome 2 with the t. In other words, it
is complete chance as to which of the chromosome 2s any specific gamete
inherits. The result is that each gamete gets either T or t, and either is equally
likely. Half the gametes carry T and half carry t, and none carry both or none.

To get the F2, the F, plant's eggs are fertilized with its own pollen. Half its eggs
carry T. If all the pollen is equally viable and functional, then an egg carrying T
will be fertilized by pollen carrying T half the time and pollen carrying t half the
time. That is, all the zygotes and their probabilities are (1/2T + '2 t) ('2T + 'zt),
which produces the familiar (14TT + 1/2Tt + 1/4tt) zygotes. It is the pairing of
homologous chromosomes in meiosis and the inheritance of only one
chromosome of each pair that is the physical basis for Mendel's Law of
Segregation.

Mendel's Law of Independent Assortment also is easy to understand. Let's

consider Mendel's cross of a tall, green-podded variety with a short, yellow-
podded variety. As mentioned, three genes that commonly appear in pea crosses
can be responsible for tallness, and we don't know which Mendel was working
with. The three are on chromosomes 2, 3, and 4, respectively. For purposes of
illustration, let's suppose that Mendel's tall gene was the one on chromosome 2.
Mendel's gene associated with yellow pod color was on chromosome 5.

We can represent the cross of the tall green variety and the short yellow variety
as TTGG X ttgg. The F1 is tall and green and is represented as TtGg. Now let's
consider what happens during meiosis of the F1. As we've seen, the two
chromosome 2s pair; half the gametes inherit the one carrying the T, and the
other half inherit the one carrying the t. But the two chromosome 5s also pair,
and half the gametes inherit the one with the G and the other half the one with
the g. The chromosome 2s pair separately and independently from the
chromosome 5s, so the inheritance of T and G are independent. Of the gametes
that have T, half have G and half have g. Of the gametes that have t, half have G
and half have g. Mendel's Law of Independent Assortment is based on the fact
that genes that are on separate chromosomes physically assort independently.

But what if two different genes are on the same chromosome?

 Linkage

Suppose, for example, that genes T and G are both on chromosome 2. Our
cross of the tall green variety with the short yellow would still be represented as
TTGG x ttgg, and the F1 would still be represented as TtGg. But now, when we
consider what happens during meiosis in the F1, only one chro

mosome pair, pair 2, is involved. One of the 2s carries both T and G; the other
carries both t and g. Any given gamete gets only one chromosome 2 or the other,
so half the gametes would be TG and half tg. The F2 would be (1/2TG + 1/2tg)
(1/2TG + 1/2tg), which works out to (3/4T-G- + 1/4ttgg). That's very different
from the (9/A6T_G_ + 3/16T_gg + 3/16ttG_ + 1/A6ttgg) we get when G and T
are on different chromosomes. We recover only the parental classes (those that
resemble the original parents); there are no recombinant classes. There is no
independent assortment of T and G, for the simple reason that they aren't
independent if they are physically on the same chromosome.

Linkage affects our ability to do breeding work. Let's suppose we cross a tall
green with a short yellow, with the objective of getting a tall yellow, for
example. If tall and yellow are determined by unlinked genes, we can recover
tall yellows very easily; threesixteenths of the FZ plants would be tall yellows.
But if T and G are tightly linked together on the same chromosome, we won't get
the desired class at all. With only seven chromosomes, and with crosses that
involve dozens of genes that matter agronomically, linkage of some of the genes
we care about is common.

Fortunately, the effects of linkage on the recovery of recombinant classes is

usually not as extreme as I have portrayed, because a phenomenon called
crossing over occurs. When homologous chromosomes pair during meiosis,
breakages and reunions can occur, so genes that started out linked can become
separated or recombine even though they are on the same chromosome. How
likely this is depends on how far apart the two genes are on the chromosome.

If T and G are very close together,. a crossover between them will be rare.
With no crossovers between the genes, the gametes will all be parental types -
that is, TG or tg. We will get (3/4T-G- + 1/4ttgg) and no T_gg (tall yellow). If T
and G are far enough apart to allow some crossing over, we will get some tall
yellows, though not as many as we would expect from independent assortment.

If T and G are located far apart on a chromosome, crossovers between them

are likely to'occur. They will stay together whenever no crossovers or an even
number of crossovers occur; they will recombine whenever an odd number of
crossovers occur. Half the time, the number of crossovers will be even and half
the time odd. So the gametes expected will be (1/4TG + 114tg + 1/4Tg + 1/4tG)
- that is, exactly the same as we get when the two genes are unlinked.

Thus genes on different chromosomes assort independently because they are

actually physically independent. In addition, most pairs of linked genes act
genetically as if they were independent even though physically they are not.
Only closely linked pairs of genes show their nonindependence by giving us
smaller frequencies of the recombinant classes than we would expect.

When we do crosses to develop new varieties, linkage can either aid or

complicate our efforts. Mendel worked with only seven genes, and none of them
displayed linkage. It is now thought that three of his genes were on separate
chromosomes, another two were far apart on another chromosome, and the other
two also were far apart on yet another

chromosome. In addition, we believe that the other two chromosomes in peas

don't have any of Mendel's genes.

One other basic genetic phenomenon is very common, but Mendel failed to
run into it with his small sample of genes. This phenomenon is codominance.
 Codominance

Mendel worked with only seven different pairs of alleles, and each displayed
complete dominance. That is, in a cross of a tall and a short variety, the
heterozygotes (heterozygous individuals) resembled one of the parents. The
cross was TT X Tt. The F1, which is Tt, was just as tall as the tall parent, even
though it had only one copy of the T gene instead of two. We use the term
dominance to describe such situations. Not all allele pairs display simple
dominance, however. In some cases the appearance of the heterozygote is in
between that of the two parents. The classic example involves a cross between
two pure breeding varieties of four-o'clocks, one with red flowers and one with
white flowers. When the red and white varieties were crossed, the flowers of the
F1 plants were pink. The F2 was 1/4 red-flowered + 1/2 pinkflowered + 1/4
white-flowered.

In this example, we can portray the red variety as RR, the white variety as rr,
and the F1 between them as Rr it is clear from the appearance of the F1 that Rr is
pink. So it is no surprise that the F2 plants with genotype Rr also are pink.
Where alleles display codominance, the F1 is distinguishably different from both
parents, and there are three phenotypic classes in the F2, with the largest class
having the same phenotype as the F1.

In most randomly chosen crosses, most genes - at least 75 percent, I would say
- display simple dominance instead of codominance. But codominance is
common enough that you'll be likely to run into it fairly often.

Purple color in plant leaves and stems frequently shows codominance, with the
hybrid being purple, but not as purple as the parent. A smooth as opposed to a
deeply indented leaf margin also is frequently codominant. Hybrids usually show
an intermediate amount of indentation. Many gene combinations associated with
flower color show codominance.

One important aspect of codominance is that the phenotypes that show up as a

result of codominance are heterozygous and are inherently not pure breeding.
With the fouro'clocks, for example, you could not obtain a pure breeding pink
strain from the cross of the red and white. You could inbreed pinks from this
cross for the rest of your life, but you would still never get a pure breeding pink.
Understanding about the existence of codominance allows you to recognize the
cases where that work would be futile. If you wanted a pure breeding pink, you
would use some other cross to get it.

If we do a cross involving two different genes, and one displays codominance,

this affects the number and kinds of classes we get in the F2. To use an abstract
example, let the cross be AABB x aabb. If both genes display dominance, the
F2, as we have seen, is (3/4A_ + 114aa)(3/4B_ + 1/4 bb), which works out to
9/16A-B- + 3/i6A_bb + 3/i6aaB_ +'I 6aabb. But if the A locus dis plays
dominance and the B locus codominance, then the phenotypes of the Fz would be
calculated from the expression (34A_ + 114 aa)(1/4BB + + 114bb), which works
out to '16A-BB + 38A_Bb + 3/16A-bb + 1/16A-BB + 1/saaBb + VI6aabb.

Exceptions and Apparent Exceptions to Mendel's Laws

There are exceptions to Mendel's laws - some only apparent, some genuine.
Exceptions are common enough that you will run into them fairly often in
practical breeding work, even if you operate on a small scale.

Lethals, Detrimentals, and Steriles

Some genes are recessive lethals. In other words, the plants that are
homozygous for them don't survive, but the plants that are heterozygous for
them do. For instance, the plant with genotype Aa is perfectly OK, but the plant
with genotype as doesn't survive. The seed may abort or fail to germinate, or the
seed germinates but the seedling dies when it runs out of the stored food in the
seed. A detrimental is a gene or genotype that is sometimes lethal or causes a
loss in fitness or vigor.

Where the effect of a genotype is to cause the individuals that carry it to

vanish, you don't see the expected Mendelian ratios. In a cross of Aa X Aa, for
example, we expect + '2Aa + 114aa). But if as is lethal, we get only two classes.
The modified ratio is ('3AA + 2/3Aa). (There are twice as many Aa's as AA's.)

Alternatively, the genotype as may be severely detrimental or lethal only in

some circumstances, so the as class appears, but with a lower frequency than
expected. In addition, if you are following an ordinary gene that is closely linked
to a lethal or detrimental gene, this can skew the ratios of the ordinary gene.
 Some genes are lethal in the gamete stage. Pollen lethals are especially
common. In this case the lethal allele displays normal segregation in formation
of eggs but isn't transmitted via pollen. In other cases the gene may be associated
with poor gamete performance. For example, a gamete carrying allele s may
grow down the style of the female flower more slowly than a gamete carrying S
and fail to be recovered with the expected frequencies in certain crosses.

In addition, some genotypes may be viable but be associated with the sterility
of the plant as a male or a female or both. Some genes are viable and have a
dominant phenotypic appearance as heterozygotes but are recessive lethals as
well. For example, YY plants may have normal green foliage, Yy plants yellow-
green foliage, and yy plants no chlorophyll at all. Thus, the yy plants die as
small, yellow seedlings.

Genes also can be lethal or detrimental only within specific contexts. The
genotype as may fail to survive in cold weather, for example, but have no visible
effect in hot weather. Or it might even outperform A-in hot weather; in that case
it would be advantageous in hot weather but lethal in cold weather.

Lethals, detrimentals, and steriles of various kinds are especially likely to

show up when you save seed from a plant that is nor

mally propagated vegetatively (where such genes are not selected against in each
generation), They also are likely to appear when you forcibly inbreed a plant that
is normally strongly outbreeding - by, for example, doing special tricks to bypass
its incompatibility mechanisms. Simple, commonplace acts such as saving seed
from a potato or bud-pollinating a broccoli can produce progeny that include a
number of weird types.
 Aberrant Segregation

The crosses we have been discussing are crosses involving relatively close
relatives. In these crosses both parent varieties contribute chromosomes with the
same arrangement and order of genes on the chromosomes. If we cross two quite
distant relatives, their genomes may have different chromosomal arrangements.
They have evolved away from each other a bit. The block of genes that is at the
left end of chromosome 2 in one parent, for example, may be on the right end of
chromosome 5 in the other, and vice versa.

In such cases we can still cross the two varieties and obtain the F, without
difficulty. Each gamete contributes one and only one copy of every gene in the
genome, so the F, plants have two copies of each gene, as they should, and they
grow and divide normally. Chromosome pairing doesn't occur during mitotic
division, so the fact that the chromosomes aren't homologous doesn't matter.

However, problems occur during meiosis in the flowers of the F, plants.

Homologous chromosomes can't pair. For chromosomes 2 and 5, there are no
proper homologous chro mosomes. Each is homologous partly with chromosome
2 and partly with chromosome 5, and the appropriate parts attempt to pair with
the parts of the chromosome with which they are homologous. When cell
division occurs, many chromosomes get torn apart, tangled around each other, or
left out altogether. Many gametes are inviable. In addition, many that do
function contain major deletions or excessive numbers of genes or sections of
chromosomes that give rise to aborted or inviable seed or abnormal progeny. In
short, the meiotic process that is responsible for causing each gamete to have one
and only one copy of each and every gene - and which produces Mendel's laws -
depends on chromosomal homology. Varieties that are distant enough to have
evolved away from each other somewhat in genome arrangement don't have that
homology.

Crosses between two varieties that are only distant relatives are some of the
most interesting crosses to do. But often we just don't see "normal" Mendelian
segregation in such crosses. We may fail to recover all the expected classes or to
recover them with expected frequencies. The crosses are what we geneticists call
"messy."
 The practical approach is to go ahead and do whatever crosses we want and
look for the class or classes we are interested in. If we can find the appropriate
individuals at all, that is good enough to continue. At later stages of the project,
chromosomal homology recurs as various nonhomologies are eliminated through
pollen and plant lethality. So in later generations of the project, we begin to see
normal Mendelian segregation for characteristics that were messy initially.
 Incomplete Penetrance and Variable Expressivity

So far in this discussion, I have been presuming that the relationship between a
genotype and a phenotype is consistent. This is not always true. Many genes are
variable in their expression. They may be affected in important ways by even
minor differences in the environment, or there may be an underlying stochastic
(chance-associated) mechanism involved in the degree to which they express or
whether they express at all.

The individuals of a pure breeding variety that carries a gene for resistance to a
disease often show considerable variability in their resistance, for example. We
may be able (using methods described later) to establish that the variety is,
indeed, pure breeding for the genes involved in disease resistance, but some
plants are very resistant and others less resistant. And some plants of the variety
may even be fully as sensitive as plants from sensitive varieties.

We use the term variable expressivity for the concept that the degree to which
the character is expressed can vary even when the genotype is constant and the
environment is as constant as we can make it. When the expression of a gene is
so variable that sometimes the gene seems to have no effect on phenotype at all,
we say the gene is not fully penetrant.

If a gene doesn't always express itself, this affects the ratios we obtain in
crosses. If expected classes are (3/4A- + 1/4aa) but as can be distinguished from
A_ only part of the time, then the ratios we observe are skewed more toward A_
and less toward as than we would expect.

Genes with variable expressivity are undoubtedly much more common than we
tend to think, because geneticists almost never choose them to work with unless
they are deliberately studying variable expressivity (and hardly anybody does).
Genes with large and consistent effects on phenotypes are the ones that are
easiest to work with. However, genes with small and/or variable effects will be
present by the dozens in many of the varieties we work with and the crosses we
do.

Variable expression of genes matters a lot in plant breeding. In the first place,
we may need to select for a gene that has variable expression. Many genes of
great agricultural importance are variable in expression or not fully penetrant. A
gene with variable expression that is associated with disease resistance, for
example, may be adequate to allow you to grow the crop; it may not matter that
some plants succumb to the disease as long as most do not. The gene may be
messy to work with but very practical to have and use anyway.

In the second place, if we have a variety that is pure breeding at a given locus
but the gene is variable in expression, the variety will display variability that is
not genetic. When we look at it, our first inclination will be to think that the
variety is not pure breeding. But it may be. If we try to select for the phenotype
associated with the expressed gene, we will be wasting our time; we can select
successfully only if genetic differences are present. We could spend a lot of

time trying to improve the variety via selection and make absolutely no progress.
So it's useful to be able to distinguish when phenotypic variability is caused by
genetic variability and when it isn't. I'll explain how to do that when I come back
to selection.
 Maternal Inheritance

The pericarp, the outermost layer of the seed, is maternal, not embryonic
tissue. So its phenotype reflects the genotype of the mother, not of the embryo
the seed contains. Some other genes whose products operate at very early stages
may also show maternal inheritance - in other words, the phenotype for the
grown plant is established by the genotype of the mother. Maternal inheritance
of this sort is Mendelian, but the expected ratios appear one generation late.
 The Cytoplasmic in Inheritance

More than 99 percent of the genes in plants are arranged on chromosomes

located in the nuclei of cells, and they are inherited according to Mendel's laws.
However, some small fraction of genes are located on organelles in the
cytoplasm, and they aren't inherited in this way. Even though there aren't very
many extrachromosomal or cytoplasmic genes, they have a disproportionate
importance in plant breeding.

The inheritance of genes in cytoplasmic organelles is completely non-

Mendelian. It is almost always exclusively maternal. The egg is a big cell with a
lot of cytoplasm. The pollen is a very small cell with little cytoplasm, which is
not normally transferred to the egg during fertilization. Nuclear DNA comes
equally from both parents, but cytoplasm comes only from the maternal parent.

Once in a while you can do a cross where one of the characteristics you are
interested in is associated with a gene that is cytoplasmic instead of nuclear.
Suppose you do a cross between a variety that has a mottled coloration pattern
on the leaves and one that does not. If you cross a mottled female to a normal
male, some or all of the F, plants are likely to be mottled. But if you cross a
mottled male with a normal female, none of the F, plants will be mottled, nor
will the offspring in any future generations.

The cytoplasm is relevant to us in additional ways. The cytoplasm of a variety

and its nuclear genome have evolved to be compatible. When distant relatives
are crossed, sometimes the cytoplasm and the nuclear genome of the FI or of
some of the FZ progeny aren't compatible. In extreme cases lethality results. In
less extreme cases the plant may be sterile or pollen-sterile. Malesterile
cytoplasms have, in fact, been useful in the production of hybrids. Such
cytoplasms aren't absolutely male-sterile; they are sterile only in the presence of
one or more nuclear genes.

Anytime we are crossing two fairly distant relatives, we need to realize that the
reciprocal crosses might be genuinely different. Both will give rise to the same
nuclear genotype, but they will set it down in the middle of different cytoplasms.
This could have a dramatic effect on the phenotype of the F, and subsequent
generations.
 Many modern crops trace their nuclear genomes to only a dozen or so sources.
But it is common for vast fractions of a modern crop to have only one
cytoplasmic genome. The Southern corn blight in the United States in the early
1970s was caused by the blight sensitivity of a particular cytoplasm - a male-
sterile cytoplasm that was in virtually the entire commercial crop, whatever the
nuclear genome involved.

Whenever we do a cross using a particular variety as the maternal parent, we

are preserving its cytoplasm and discarding that of the male. I think we should
pay more attention to preserving and increasing the diversity of the cytoplasms
in our food crops. All we have to do to preserve a particular cytoplasm is to
make sure that we use it as the female parent when we do crosses.
 Gene Interaction

Often more than one gene is involved in determining a trait. Consider an

abstract example in which A and B are genes that code for different enzymes
involved in the biosynthesis of the plant pigment anthocyanin. Let's also assume
that A is dominant to a and B is dominant to b, and that the A locus and the B
locus are not linked. Let's also imagine that the a allele makes a completely
inactive form of its enzyme, and so does the b allele.

Now let's consider four different pure breeding varieties representing all
possible genotypes with respect to the two loci. Variety 1 is AABB and has
purple flowers. Variety 2 is AAbb and has white flowers. Variety 3 is aaBB and
also has white flowers. And variety 4 is aabb and also has white flowers. Variety
2 would have white flowers because it has no good enzyme for one step in the
synthesis of the purple pigment. Variety 3 has good enzyme for that step, but has
no good enzyme for some other step that is also necessary to synthesize pigment.
And variety 4 lacks any good enzyme for either of two of the essential steps in
pigment biosynthesis.

If we cross variety 1 (purple) with variety 2 (white), the cross is: AABB X
AAbb, and the F, would be AABb (purple). And the Fz would be 3/4AAB_ +
1/4AAbb; that is, 3/4 purple + 1/4 white. This is just a one-factor cross. That is,
there is only one factor - the B gene - that differs between the two strains.
Information about the other gene, the A locus, is only included for later
reference.

If we cross variety 1 with variety 3 we obtain similar results. The cross would
be AABB (purple) x aaBB (white) and the F, would be AaBB (purple). The F2
would be 3/4 A_BB + 114aaBB, that is, 3/4 purple + 1/4 white respectively. This
is also a one-factor cross, but the factor involved is different from the one
involved in the cross of variety 1 and variety 2.

Now let's consider a cross of variety 2 with variety 3. The cross is AAbb x
aaBB, and is a cross of two plants to each other. The F, would be AaBb, and it
would be purple. Each locus displays dominance, that is, one "good" copy of
each gene makes enough enzyme to perform its step. The F, plants have one
good copy of each gene, and thus make enough good enzyme for both steps in
biosynthesis of pigment.

These sorts of situations - where a breeder crosses varieties of identical color

or phenotypes similar in some other regard and obtains an F, of a different color
or pheno type - are quite common. In some cases they have caused great anguish
because the new phenotype was undesirable or inedible compared to the parents.
In other cases the new color or phenotype was useful itself.

But let's carry the cross one step further. The F2 genotypes would be + 3/16
aaB_ + 3/16A-bb + Vi6aabb. And all the classes except the first one would be
white. So the phenotypic classes would be 9A6 purple + 7/io white.

The final possible cross, that of the first variety to the fourth, would be AABB
(purple) X aabb (white) and would give a purple F,. But the F, would be exactly
the same as in the cross of varieties 2 and 3 and would reduce to phenotypic
classes of 9A6 purple + 7/io white.

In a cross of a pure breeding soybean with black pods to one with tan pods, the
F, was black. The F2 was 12/16 black + 3/16 brown + 1/16 tan. Where did the
brown come from? Well, we generate two new genotypes in the F2 of a two-
factor cross, so one or two new phenotypes shouldn't be surprising. In this case,
the cross could be symbolized as EEFF (black) x eeff (tan), and the F, would be
EeFf, which was also black. The F2 genotypes would be 9/16E-F_ + 3/16E_ff +
3/16eeF_ + 11A6eeff. The E_ff and the eeF_ are new genotypes. One of them
must look black and get lumped with the E F group. The other must look brown.

Two different genes that affect the same trait can interact in all kinds of ways.
The basic ratios in the F2 of 9:3:3:1 can become 15:1, 9:7, 9:6:1, or 9:4:3
depending on what phenotypes are generated by the various genotypes. New
phenotypes can appear in the F1 or F2 depending on the specific cross and the
interactions between the genes. And that is all with just two genes, both of which
show dominance.

In a cross with just one factor and codominance, one new phenotype appears in
the F, and F2. In a two-factor cross where both loci show codominance, each
factor generates three genotypes, and the F2 has nine different genotypes. That
means up to seven new phenotypes could appear.
 There are two major genes, p and v, involved in the difference between snow
(edible-podded) peas and shelling peas. Both are responsible for reducing the
amount of fiber in the wall of the pod; they are apparently involved in different
steps or aspects of production of the fiber. Both genes are simple recessives. And
the two loci are on different chromosomes; p is on chromosome 6, and v is on
chromosome 4.

Shelling peas are generally of genotype PPW. Most modern varieties of snow
peas are genotype ppvv, and the pods are edible even when they are quite large.
Some older snow-pea varieties are ppVV or PPvv, and are not as fiber-free as
varieties that are ppvv, or are fiber-free only when the pods are small.

If you crossed a semisnow of genotype ppVV to another semisnow from a

different variety that happened to be PPvv, the F1 would have tough, shelling-
type, inedible pods. The genotypes in the F2 would be 9/16P_V_ + 3/i6P_vv +
3/16ppV_ + 1A6 ppvv. The phenotypic classes we would obtain would be 9/16
shelling + 6/i6 semisnow + 1/,6 snow. (We would not be able to distinguish the
two different classes of semisnows, but we could distinguish them from the true
snows and the shelling types.)

So you could cross two semisnow type peas and, in some cases, recover a
much better snow pea than either parent as 116 of the F2. This is probably how
our modern ppvv varieties were developed.

Note, however, that not all crosses of two semisnows would have the potential
for generating a better snow. Suppose we had only semisnow type snow peas, for
example, and you wanted a better snow pea - one that stayed tender even when it
was a bigger pod. You wouldn't know anything about the p or v genes, and you
wouldn't know that some semisnow varieties are PPvv and others are ppVV. But
as a general principle, you would cross two varieties that were among the best
available with respect to the characteristic that you cared about - pod type, in this
case.

If you happened to cross two varieties that were both ppVV or two varieties
that were both PPvv, then the F1s would still be semisnows, as would all of the
Fes. In other words, you would have just one phenotype in the F2, and it would
be the same as that of the parental varieties and the F1. There would be no
variability for pod type. Even if you grew only twenty or so FZ plants, which
might not be enough to reliably recover a member of a class expected only 1/16
of the time, you would still expect to see at least two classes. So if you didn't,
you probably wouldn't bother growing more seed from that cross. Instead you
would try different crosses.

If,.on the other hand, you crossed two semisnow varieties and obtained an F,
that was a shelling type, you would know right away that you had generated
genetic variability for pod type with that cross. If you then grew up only twenty
FZ plants and found two classes, shelling and semisnow, but failed to find any
snow types, you would probably try more seed from the same cross before
giving up on it and going to a different cross. If you plant only twenty seeds,
much of the time you will miss a class you only expect 1/i6 of the time.

The concept that different varieties may have different genes contributing to
similar phenotypes is a major part of practical plant breeding. Let's suppose we
want a variety that is more cold hardy than anything available. We would make
one or more crosses between the most cold-hardy varieties we could find. We
would be hoping that two varieties would be cold hardy for different reasons and
that from them we could produce a new variety that was more cold hardy than
either parent.

It's also possible to have more than two genes involved in the inheritance of a
single trait in a cross. With flower color, for example, the cross between two
varieties may involve three, four, five, or even more factors. By the time we have
four factors involved, even if all loci show simple dominance, we will have up to
sixteen phenotypic classes in the F2. The cross may produce all kinds of colors
we never imagined, with some genes affecting the kinds of pigments produced,
others affecting the amounts, and

still others affecting the distribution of the pigments in the parts of the flower.

Once a number of factors that affect a trait are involved in a cross, people
sometimes use the terms major genes and modifiers. Two major genes may
account for most of the genetic variation in the F2, for example, so there will be
the expected discreet classes in the F2. But there may be additional variation
within each class caused by the segregation of other genes that, in the context of
the particular cross, have smaller effects on the phenotype. If the cross involves
color, for example, there may be four basic colors in the F2, but various
individuals may have different shades of them. We refer to the genes that
segregate to generate the main classes as the major genes and those that are
involved in creating the shades as minor genes, modifier genes, or modifiers.
The terms are merely subjective, though, as one of those minor genes could be a
major one in another cross.
 Quantitative Inheritance

In many crosses many different genes affect the characteristics we care about.
General plant vigor, for example, is often influenced by many genes, each
having effects that are too small to sort out individually. Sometimes there are
signs of dominance at some loci; often there is codominance. Where many genes
contribute to a characteristic, we don't see discreet classes when we do crosses.
If we cross a high-yielding variety with a low-yielding one, for example, it is
likely that the FZ will include plants that cover the entire range from low to high,
with no discreet classes obvious. The effects of the individual genes appear to be
quantitative instead of discreet. Each gene or allele can be thought of as adding
an increment to the phenotype.

Size, growth rate, vigor, and yield show quantitative inheritance in most
crosses. They also can be affected by single genes that have large enough effects
to show discreet classes. When I make crosses involving tall and short varieties
of peas, for example, I obtain discreet classes of tall and short. But within each
class there is considerable variability for height. In other words, in most of these
crosses one major gene affects height, but a lot of other genes have less dramatic
effects.

Codominance is common with quantitatively inherited characters. A common

model is that several genes are involved whose effects are too small to tally
individually, and there is often codominance. If we cross AAbbCCdd and
aaBBccDD, for example, where the large-letter alleles all contribute to higher
yield and are codominant, then the F1, AaBbCcDd, will have about the same
yield as the parent varieties. Each parent has four of the large-letter alleles, but
among the Fz plants will be some with genotypes having more of the large
letters than either parent. These will average much higher yields than either
parent. There also will be individuals that have fewer large-letter alleles than the
parent varieties and have lower yields.

To consider another example, imagine a low-yielding variety of genotype

aabbccdd and a high-yielding one with genotype AABBCCDD. If we cross the
two, the F, will be intermediate between the two parents, and the Fz will include
individuals of every
yield from as low as the low parent to as high as the high parent. Because the
effects of each allele are small and the environment also influences each plant,
there will be no discreet classes. It will seem exactly as if inheritance were based
on the mixing of some kind of fluid.

In fact, we now know that even in these cases, inheritance is caused by discreet
genes. But when there are many genes and their effects are small, we cannot
decipher what is going on without investigating it specifically, without
understanding Mendelian genes based on how they behave in simpler crosses,
and without sophisticated statistical methods.

Whenever we cross two varieties that are distant relatives, many genes differ,
so many characteristics show quantitative inheritance. These are the kinds of
crosses that most preMendel scientists were trying to learn about heredity. It
wasn't possible to understand such crosses, however, until Mendelian patterns
were understood and there were better statistical methods.
 Predictions and Actuality

It's useful to be able to figure out what classes and frequencies we expect when
we do crosses, because this helps us decide what to do. That is, the predictions
help us evaluate whether a project is possible with a specific approach given the
time and space we have. If we calculate that the class we want is expected to
occur with a frequency of 1/1.024 in the F2, we know we will probably not get
any at all if we raise only twenty F2 plants. If the class we want has a frequency
of 1/16, we have a good chance of getting it with just twenty plants. But how
good?

How many plants would we need to be reasonably sure of getting at least one
of the type we want? If we expect a coin to land heads half of the time and we
toss it twice several times, we won't get exactly one heads every time we do it;
chance matters. Likewise, if we expect a 9:3:3:1 ratio and we raise exactly
sixteen plants, we won't ordinarily get exactly 9, 3, 3, and 1 plants of the
respective classes. Sometimes, for example, we will get none of the class with
the expected frequency of 1/id, and sometimes we will get two.

What we need to know is how many of something we need to grow to be

reasonably sure of obtaining at least one of the class we want. This is something
that we can calculate statistically. The table in Appendix H represents the results
of these calculations for the probabilities that come up most frequently in
genetics. To use the table, look up the probability of the class you want; the table
tells you how many plants you need to be either 95 percent or 99 percent sure of
obtaining at least one of the desired class.

If the desired class is expected 1/4 of the time, we can look up "A" in the table
and see that we need eleven plants to be 95 percent sure of obtaining at least one
of the desired class. If we want to be 99 percent sure, we will need sixteen
plants.

If the class we need has a probability of 1/ib, we should grow forty-six plants
if we want to be 95 percent sure of obtaining at least one of the class. This means
that if we raise forty-six plants every time we need to obtain at least one of a
class with a probability of 1/1o, 95 percent of the time we will succeed, and 5
percent of the time we will not. If we want to be 99 percent sure, we will have to
raise seventy-one plants.

But what if we don't have enough space, time, or seed for forty-six plants, let
alone seventy-one? We may then choose a different project, a different approach
to this project, or a different kind of cross (backcrossing, for example). Or we
may just go ahead with the project and the cross, raise however many plants we
can, and figure on obtaining what we want in more than one step. These options
are aspects of breeding project design, which I discuss in chapter 10 in the
context of specific examples.

 UPPOSE YOUR grandmother had a special bean variety. You remember
it and would love to grow it, but as far as you know, the seed was lost. Then you
find a single seed of the variety caught in a crack in a chest of drawers that
belonged to your grandmother. You plant it, and it germinates. Can you restart
and maintain your grandmother's variety from just the single seed? The answer is
probably yes.

Now let's suppose that you go back and examine all the cracks and crevices in
grandmother's chest of drawers, and you also find a broccoli seed. And you
remember that Grandmother had her own broccoli variety, too. Can you restart
and maintain her broccoli from just a single seed? The answer is probably no.

Beans are an inbreeding plant. Chickpeas,

peas, tomatoes, and lettuce are other common inbreeders. Broccoli is naturally
outbreeding. Corn, spinach, most of the brassicas, and most of the cucurbits are
among the common outbreeders. Varieties of inbreeders and varieties of
outbreeders are essentially different - so much so that it is misleading that the
word variety is used for both.

The genetic structure of a population of plants of an inbreeding variety is

usually very different from that of a population of an outbreeding variety. In
addition, many outbreeders display inbreeding depression. Some also have
incompatibility systems. These factors affect every aspect of how inbreeding and
outbreeding varieties are both created and maintained.

To understand how inbreeders and out breeders are different, and to make the
characteristics of each work for us, we have to understand a bit more about what
inbreeding actually does.

Inbreeding and the Genetic Nature of Inbreeding Crop Varieties

 Imagine that we make a cross of two different varieties of an inbreeder that
generates heterozygosity at the A locus, so that the F, is Aa. Then let's allow the
plant to inbreed one generation to obtain the F2 generation. We know the results:
(1/4AA + 1/2Aa + 1/4aa). What's interesting to us at this point is that one
generation of inbreeding eliminates half the heterozygosity. The F, population
was 100 percent heterozygous at the A locus. One generation of inbreeding gives
us an F2 population in which half are heterozygous (Aa) and half are
homozygous (either AA or aa).

Now let's allow all the plants of the Fz population to inbreed another
generation. And let's assume that each produces equal amounts of seed (or that
we save equal amounts from each deliberately). What will the F3 population be
like?

Well, one-quarter of the F2 population is AA, and the one-quarter of the F3

population these plants give rise to will all be AA. Similarly, one-quarter of the
FZ population is aa, and the one-quarter of the F3 population these plants give
rise to will be aa. The half of the F2 population that is Aa will give rise to
(1/4AA + 1/2Aa + 1/4aa), and since these plants contribute half of the F3
population, their contribution to F3 is (1/2)(1/4AA + 112Aa + 114aa), or ('sAA +
14Aa + 1/4aa). When we add their contribution to that of the other genotypes,
we get an F3 population that is (3/8AA + 114Aa + 3/saa).

The percentage of the population that is heterozygous went from 100 percent
to 50 percent when we inbred one generation. Then it went from 50 percent to 25
percent when we inbred another generation. If we inbreed further, the percentage
of the population that is heterozygous at the A locus will continue to drop by 50
percent with each round of inbreeding.

The reduction of heterozygosity with inbreeding happens at all loci. This

means that we lose, on average, half of the total heterozygosity in the population
with each generation of inbreeding. This has profound implications for the
genetic nature of inbreeding crop varieties.

In an inbreeding crop variety, whenever an event creates genetic heterogeneity

- an occasional cross or a new mutation, for example - sequential rounds of
inbreeding promptly eliminate the heterogeneity and sort it out into homozygous
genotypes. For this reason, when there is variability in an inbreeding crop
variety, it is because it is a mixture of slightly different lines, each of largely
homozygous individuals. In other words, if a population of inbreeders contains
variability at the A locus, it is usually because it contains individuals with
genotype AA and individuals with genotype aa. Little of the variability is in the
form of individuals with genotype Aa.

In other words, individuals of inbreeding varieties tend to be highly

homozygous.

There are three important practical implications of this genetic structure of

inbreeding populations. First, since any given seed or plant is generally very
homozygous, offspring generally resemble their parents very closely. They are,
in fact, very close to genetically identical to their parents unless some mutation
or accidental cross has occurred quite recently in their pedigree. The common
expression "as like as two peas in a pod" is based on an important component of
the nature of reproduction in peas and other inbreeding plants.

The second important practical aspect is that when we choose a superior pea
plant, it is usually the case that the plant is already homozygous for the genes
that make it superior. If I find a chickpea that yields twice as well as the others of
the variety, there is a good chance that I can derive a new variety in one step just
by saving its seed separately from the rest. If the plant is genuinely superior to
the rest genetically, it is likely that all its offspring will be also. (The exception is
if a new mutation or accidental cross occurred recently in the superior plant's
pedigree.)

When I pick one superior plant out of a land race of chickpeas and save its
seed separately, however, I discard most of the variability that was present in the
land race. That could include natural resistance to diseases that might appear in
future years. But it also includes, say, variability for yield, which I'd prefer not to
have. I would like uniformly high yield.

Variability is not necessarily either good or bad in itself. Variability for yield
tends to make the average yield low, so the crop isn't worth growing at all. But
variability for plant color, for example, won't matter at all if the plants all yield
well. Variability for plant shape often is acceptable for hand-harvested
chickpeas, but erect plants are better for machine harvest. Simultaneous
maturation is essential for machine harvest of garden vegetables. For the home
gardener who eats the vegetable fresh, variable maturation over a period of time
often is preferable.

The third practical implication of the genetic nature of inbreeders is how the
preservation of variability relates to the number of plants from which we save
seed. With chickpeas, farmers grow a lot of plants and save seed from them all,
since the dry seed is the crop. Saving seed from many plants maximizes the
probability that all the various related sublines will be preserved. With garden
plants such as peas, however, you have to sacrifice the food production of some
plants in order to save seed. In addition, you don't grow as many plants as a
farmer with a field of chickpeas.

The farmer who saves seed from thousands of chickpea plants is probably
saving seed from every genotype and subline that existed in the variety the
generation before. But if I save peas from ten plants in each generation, I can
save only a few of the sublines that might have been present originally. This
often is adequate, though, since the sublines I save are likely to be those most
suitable for my conditions. In addition, I am not usually the only gardener with a
variety. Different seed-saving gardeners are usually preserving slightly different
lines. Even if each person is maintaining only a single subline, a good amount of
total heterogeneity is preserved.

If I start a new chickpea variety by saving seed from one superior plant, my
new variety has almost no genetic variability, so little will be gained by saving
seed from many plants in each generation. If, however, I want to preserve the
original land race, it's better to save seed from many plants.

You can do a pretty good job of maintaining most of the genetic variability in a
land race or other heterogeneous material by saving seed from a hundred plants
each generation. You can do an excellent job with seed from a thousand plants
per generation. But you can maintain much of the genetic variability in
heterogeneous material with just twenty plants per generation. In the latter case,
you tend to preserve the more common sublines and lose the rarer ones.

You can do a good job of maintaining a relatively uniform inbreeding crop

variety (a subline) by saving seed from five or so plants in each generation. You
can do an excellent job and give yourself plenty of margin for error with ten or
twenty plants.
 The only reason you need to save seed from more than one plant to preserve
your subline of an inbreeder is that new mutations happen all the time, and so do
accidental crosses. If you save seed from just one plant and it happens to be
carrying a deleterious mutation or be the product of a contaminating cross, you
will lose the entire variety. In other words, you can maintain a subline of an
inbreeding crop variety by saving seed from just one good plant per generation.
You save seed from somewhat more than that as a margin for error.

Remember that when you choose one, five, twenty, or however many good
plants, you need to look at the entire plant, not just the fruit. You would not save
seed from the plant with the largest pea pod, for example, if that was the only
pod on the plant.

Note that if you obtain seed from someone who generated it from a single
seed, most of the variability is gone. Even if he or she sends you a hundred
seeds, that doesn't mean you can maintain much variability by saving seed from
a hundred plants in each generation - not when you started with just one subline.
The one-seed generation represents a genetic bottleneck that eliminates virtually
all the variability.

Grandmother's bean variety might have been maintained by saving seed from
just a few plants each generation. If so, it probably had very little variability. By
finding and planting that single viable seed, you will, indeed, be regenerating
Grandmother's variety.

If Grandmother saved seed from dozens of plants in each generation and

created no recent genetic bottlenecks in her maintenance of the variety, her
variety might have contained considerable genetic variability. The single seed
you found would represent just one typical subline; the others would be lost. You
could regenerate one subline of her variety, but not all of them. You could then
preserve the basic line, but not the broader line with all its variability.

If you save seed from a couple of dozen or more plants each year, mutations
will accumulate, and more variability will be created. You'll have your
grandmother's basic line, but with a new repertoire of variability. This is quite
possibly what Grandmother did in the first place. Heirloom garden beans are
often passed on by the ones and twos and half dozens, not by the pound.
Varieties are often passed through bottlenecks, as someone finds or is given just
a single precious seed. The basic types of inbreeding crop plants can usually be
reestablished from just a single seed.
 Heirlooms

Various people have various definitions of an heirloom. Loosely speaking, an

heirloom garden cultivar or variety is just like an heirloom anything else. It's
something that has been passed down from generation to generation, preserved,
and cherished. And like other heirlooms, heirloom garden varieties often have a
flavor of romance.

Some people consider a cultivar an heirloom only if it was, at least at some

point, maintained and distributed informally by gardeners or farmers. Others
don't care how the variety was maintained but feel that there should be a cutoff
point with respect to age - say, at least forty years old. Whatever else, heirlooms
are always openpollinated - that is, nonhybrid - varieties. Many people, myself
included, use the word fairly loosely to mean any openpollinated variety that has
been around for a while.

The word heirloom also suggests excellence and value. Heirloom garden
varieties have survived for decades in the absence of patents or anything else
that makes distributing them especially profitable. They have done so very often
because they are genuinely excellent. They are the varieties our ancestors'
friends or relatives were most likely to give them personally, because they were
the best these people had. They are, of all the varieties that have been developed
and grown, the ones that gardeners refused to do without. Some heirlooms have
broad regional adaptation. Others are fine-tuned to a specific region or climate
and don't do well elsewhere.

Many heirloom varieties - beans, for example - contain a lot of genetic

variability, but many others do not. It depends on how the seed was saved. With
garden varieties of inbreeding crops, you can maintain vigorous varieties by
saving seed from just a few plants each year, as long as they are good plants that
are typical of the variety. When a variety is so maintained, it usually has little
genetic variability. But this usually doesn't matter. If you need more variability,
you can get it by obtaining seed from a number of other people who maintain it
and then pooling the seed. Usually a single, excellent subline of an heirloom
variety of an inbreeding crop is all you need for your home garden.
 Outbreeding and the Genetic Nature of Outbreeding Crop Varieties

Not all heirlooms and other openpollinated varieties are naturally inbreeding.
Many are outbreeding. To understand the genetic nature of outbreeding crop
varieties, we need to understand what happens when plants of an outbreeding
variety sit in a field openpollinating (that is, crossing however they want or
however the bees fly).

Let's imagine that we make a cross between two varieties of an outbreeding

crop and start with an F, population that is 100 percent heterozygous at the A
locus, or Aa. Let's imagine an extreme outbreeder, one with a tight
incompatibility system that prevents self-pollination. If we plant the F, plants in
a patch and leave them alone, none of them will self-pollinate, but they will
cross with each other. That is, they will do the cross Aa X Aa to get an F2
generation. That generation will be (1/4AA + '/2Aa + 114aa), just as was the case
for the inbreeding population we considered at the beginning of this chapter.

What about the F3 generation? That gets a bit complicated. With inbreeders
only three kinds of crosses (self-pollinations for each of the three genotypes) can
occur in the Fz population. With outbreeders there are three genotypes, but there
also are many more than three possible kinds of crosses. This is because each
genotype can cross with plants having any of the other genotypes. So nine kinds
of crosses are possible.

In the outbreeding F2 population, the Aa X Aa cross causes a loss of

heterozygosity, but the AA x as cross creates heterozygosity. For every AA X as
cross, the heterozygosity of the parents is 0 percent, but that of their offspring is
100 percent. In an outbreeding population, the frequencies of the various
genotypes shift with additional generations until the amount of heterozygosity
lost by the Aa X Aa crosses is exactly balanced by the heterozygosity that is
gener

ated by the AA X as crosses. An equilibrium then exists, and the same

frequencies of genotypes occur in future generations. In an outbreeding crop
variety allowed to pollinate spontaneously, heterozygosity is not lost - it is
maintained.

In a population of outbreeder that contains genetic variability, the variability is

usually both within and between individuals. One important practical implication
is that the typical individual of an outbreeding crop plant contains a lot of
heterozygosity. This is why we say "as like as two peas in a pod" but don't say
"as like as two broccoli seeds in a pod." The seeds from an individual broccoli
plant of a variety with standard levels of uniformity are often fairly different
from each other and from the mother. This means that if I find a superior
individual in a field of broccoli, it's fairly likely that it is not homozygous for
whatever genes make it superior. Thus only some of its offspring will likely
inherit its superior characteristics. I can still use it to begin deriving a new
variety, but doing so will probably take longer and be more complex than with
an inbreeding plant.

Most plants are not totally inbreeders or totally outbreeders; they are somewhere
in between. The genetic structure of their populations also falls somewhere in
between, depending on just how much inbreeding and outbreeding they do under
the given circumstances.

Many outbreeding or partially outbreeding plants are characterized by two

other phenomena that complicate our efforts to create or maintain them:
inbreeding depression and self-incompatibility.
 Inbreeding Depression

Many varieties of outbreeders tend to deteriorate in size, vigor, and yield if

inbred too much. Deterioration as a function of inbreeding is called inbreeding
depression. The converse is hybrid vigor - the tendency of many hybrids to be
larger, more vigorous, and better yielding than either parent.

Apparently, genetic heterogeneity makes an important contribution to vigor,

yield, and many other traits in these plants. The classic example is corn. If you
cross two openpollinated varieties of corn, the F, hybrid is usually dramatically
more vigorous and high yielding than either parent, and the effects can be great.
Certain combinations of parents will produce F1 hybrids that yield more than
twice as much as either parent. This hybrid vigor is the basis for the modern
hybrid seed industry.

The other side of the coin is inbreeding depression. If you try to develop a new
variety of corn by saving seed from the best plant in each generation, it won't
work. Even if you are selecting the highest-yielding plant and self-pollinating it,
you will generally find that your yield will deteriorate with each generation. You
would be following the basic rule of saving seed from the best, but you would be
destroying the variety instead of improving it.

We do not entirely understand inbreeding depression or hybrid vigor.

Generally, though, they seem to be associated with the

degree of genetic heterogeneity - that is, the proportion of loci that are
heterozygous instead of homozygous. In an openpollinated line, many loci are
homozygous; there are a lot of AA's, bb's, and CC's (instead of Aa's, Bb's, and
Cc's) within the genotype. When we cross two openpollinated varieties, the F,
hybrid will have many more heterozygous loci than either of the parents. The
more distantly related the two varieties are, the more loci that will be
heterozygous when we make the cross. Generally, crosses between distant
relatives show the most hybrid vigor.

Inbreeding depression is serious for many, but not all, plants. The naturally
inbreeding plants such as peas, beans, and lettuce show little or no inbreeding
depression. It is easy to obtain vigorous varieties that are highly inbred - that, in
fact, can be traced to a single foundation plant. The naturally outbreeding plants,
however, often must be dealt with in such a way that a good deal of genetic
heterogeneity at many loci is maintained. You usually would not want to try to
start a variety of outbreeders from a single foundation plant.

Not all outbreeding plants display serious inbreeding depression. Most of the
cucurbits, for example, show some hybrid vigor, but they don't necessarily suffer
serious inbreeding depression. It is usually possible to derive good varieties of
squash, muskmelons, cucumbers, and watermelons by inbreeding from single
foundation plants. Appendix A and Table 1 include information about inbreeding
depression for various vegetables.

How we go about developing crop vari eties is affected profoundly by whether

the crop displays inbreeding depression. With such crops we have to take the
more complex approach of achieving uniformity for the most important
characteristics, while simultaneously creating or maintaining general genetic
heterogeneity and avoiding genetic bottlenecks. The approaches used with plants
that don't show inbreeding depression often deliberately create genetic
bottlenecks, because the easiest and fastest way to get a uniform pure breeding
variety is to eliminate the genetic variability.

Outcrossing and Self-incompatibility

Many plants have mechanisms that help promote outbreeding and discourage
inbreeding. The extreme case is dioecious plants such as spinach, in which the
male flowers and female flowers are on separate plants. Monoecious plants, such
as most squash and melons, have separate male and female flowers on each
plant. This prevents the self-pollination of a flower, but not the pollination of a
flower by pollen from another flower on the same plant. Many cucurbits have an
additional trick. They tend to open only one new flower per vine per day, thus
additionally discouraging pollination of a flower by pollen from another flower
on the same plant. Flowers can, however, be pollinated by pollen from another
flower on a different vine of the same plant.

Some plants with hermaphroditic flowers also are outbreeders. Onions, for
example, are protandrous - that is, "early male." The stamens develop and
release pollen before the style finishes developing and the stigma becomes
receptive. However, the flower can be readily pollinated by pollen from other
flowers on the same plant. Plants with early female parts also are common.

In yet other plants there is nothing anatomical to indicate that the plant is an
outbreeder. Instead, there is a genetic incompatibility mechanism. The
mechanism prevents self-pollination. It also prevents pollination by plants with
an identical incompatibility genotype (such as plants developed by cloning). It
also prevents crosses of some plants with entirely different plants if they happen
to have the same incompatibility genotype.

Self-incompatibility can be very strong (there is nearly no selfing) or weak

(there is some spontaneous selfing). It often varies from variety to variety within
a crop. In addition, self-compatibility may be affected by the weather. A variety
may be self-incompatible in cool weather and self-compatible, or more so, in
warm weather, or vice versa. Some plants are more self-incompatible at one or
the other end of the flowering cycle. They may be very self-incompatible early
during their flowering but more self-compatible toward the end, for example.

Self-incompatibility can be an inconvenience to the breeder. Whenever we

make an F, hybrid and intend to go on to the Fz with an inbreeder, for example,
we can grow just a few F, plants, because we can self-pollinate them. But if
we're working with an outbreeder, we often grow a larger number of F, plants so
that we are likely to have more than one incompatibility genotype. Then we
make various crosses between them to get the F2, or plant them in a patch away
from others of their kind and let them cross-pollinate.

If an interesting plant turns up in an inbreeder, we can start to explore and

work with it just by keeping its seed, which is inbreeding it. But if something
interesting turns up in an outbreeder, the plant may resist our efforts to self-
pollinate it. Various tricks for overcoming incompatibility barriers are listed in
Appendixes A and B. In some cases, though, we may need to let other plants
pollinate the special plant, then allow those progeny to cross with each other, to
try to obtain what we want in a number of plants in a subsequent generation.
 Saving Seed of Outbreeders

When we maintain seed of established varieties of outbreeders, we usually try

to save seed from at least twenty or more plants each generation. And more is
often better. Exactly how many plants we need to keep enough genetic
heterogeneity to avoid inbreeding depression depends upon the specific crop. If
inbreeding depression is a major problem, more is better. But, as with the
cucurbits, where inbreeding depression isn't a major problem, we may be able to
handle the plants virtually as if they were inbreeders, except we have to control
pollination. Information on the extent of inbreeding depression is listed in
Appendix A and Table 1.

Corn is one of the plants that is most severely and consistently affected by
inbreeding depression. For this reason, it's best to maintain openpollinated corn
varieties by saving seed from at least a hundred good plants in each generation.
With a hundred plants per generation, you can maintain most of the variability.
With a thousand you can maintain nearly all of it.

Plants of outbreeding varieties are often perfectly willing to cross with plants
of different varieties belonging to your neighbor down the street, or to wild
relatives growing as weeds in or near your garden. To maintain outbreeding
varieties, we often must isolate them from others of their kind in some fashion.
Isolation distances considered suitable for each crop are given in Appendix A
and Table 1. Technical aspects of maintaining outbreeding varieties, including
how to cheat at the isolation distances, are covered in Part II.
 Inbreeder or Outbreeder

As previously discussed, how we breed and maintain a variety depends on its

basic breeding system - whether it's an inbreeder or an outbreeder, is self-
incompatible or not, and so on. There are three ways to establish the basic
breeding system: look it up; figure it out experimentally; guess. I use all three
methods, depending on the situation.

Appendix A and Table 1 give basic breeding information for more than 800
plants. If I'm working with a rare or unusual plant, the information I need usually
isn't known. Then I guess. In some cases there are anatomical differences that
make the guess an excellent one. Most of the time, if I'm wrong, I just don't get
the cross I want and have to try a new approach another year. Most of the
obvious guesses work, though.

If the pollen is dustlike and blows around freely, the odds are that the plant
outcrosses to at least some extent. Similarly, if bees or insects visit the flowers,
there is usually some outcrossing - enough that isolation is required for strict
pure production. But insects can facilitate either inbreeding or outbreeding,
depending on the position, arrangement, and anatomy of the male and female
parts of the flower.

Tomato flowers usually hang downward. The style and stigma are within a
cone formed by the fused anthers, which release their pollen inward. Released
pollen drops down over the stigma and fertilizes it. The opening into the fused
anther cone is tiny, too tiny for wind access, as well as for all but the smallest
insects. And it's rare to see an insect in or around a tomato flower. Tomato
flowers also aren't especially noticeable, nor do they have a pronounced odor.
All this makes it no surprise that tomatoes are strongly self-pollinating.

Large, bright flowers or flowers with odors suggest that the plant is trying to
attract the attention of someone for some purpose. That someone is often a
pollinator. In some varieties of tomatoes (including some of the heirlooms), the
style sticks out beyond the rest of the flower. Insects can walk over it, and the
wind can blow on it. These varieties exhibit a greater degree of crossing. The
extent of outcrossing in tomatoes and many other crops often varies with the
specific variety. For example, in some tropical areas, a specific insect visits
tomato flowers; where the insect is present, the tomato is a partially outbreeding
plant.

Even when there is information available

on a plant's breeding system, keep your eyes open. You may have different
weather and different insects than the person who established the information.
Just because something is in a book doesn't mean that your plants and your
insects will act accordingly.

You usually can't tell whether a plant is self-incompatible other than

experimentally. With brassicas, whenever I want to self-pollinate an unfamiliar
plant, I do at least some bud pollination so as to bypass the incompatibility
mechanism if there is one. A good indication is what happens when a plant is
grown by itself. If it sets seed, it is obviously self-fertile to at least some extent.
If it doesn't, the evidence is less conclusive. Pollinators often don't find
individual plants. You may be able to establish whether there is an
incompatibility mechanism simply by doing controlled self-pollination and
pollination with other pollen.

You can establish the extent of spontaneous inbreeding versus outbreeding by

using different varieties whose hybrid is recognizable. Make the hybrid
deliberately so that you can recognize it. Then plant a single plant of one variety
in a patch of the other at your ordinary planting density. Note that the planting
density often affects the extent of cross-pollination by determining whether most
insect trips are between flowers on different plants or between flowers on the
same plant.
 Making and Breaking Hybrids

The commercial hybrid seed industry is a mixture of a biological and a

proprietary sit uation. Hybrids are genuinely superior to openpollinated varieties
for some crops in some cases for some purposes. But commercial hybrid seed
breeders have reason to develop and sell hybrids whether they are better for the
customers or not. Often considerable effort is put into developing and promoting
hybrids instead of openpollinated varieties simply because it is more profitable
for the seed company.

When a seed company develops and sells an openpollinated variety, customers

can save the seed if they want to, instead of returning and buying it from the
seed company in future years. But when a company develops a hybrid, it keeps
the parent lines secret so that only it can produce the seed. This means that
customers have to buy the seed from the company every year or not grow the
hybrid.

The problem is that many people garden partly for the independence of it and
prefer not to be dependent on others for something so basic and essential as seed.
If we give up the ability to produce our own seed, we are giving up something
integral to the gardening experience. We are giving up our role as stewards - and
breeders - of the germplasm we use. And we are giving up the pleasure of
harvesting and cleaning the seed, holding it in our hands, and running our fingers
through it - the essence of next year's crop.

In addition, where one or a few closely related hybrids replace all the heirloom
and openpollinated varieties in an area, the effect is a huge loss in overall
biodiversity. The region comes close to being a monoculture of a single
genotype. Epidemics of pests

and diseases are far more frequent and severe in such situations than they are
when many different varieties and crops are grown. Furthermore, when one of
those more frequent epidemics occurs, if you are growing the standard hybrid,
your crop is fully sensitive to it.

Fortunately, there is an option other than using a commercial hybrid. We can

make our own hybrids, thereby retaining our independence and our control over
our seed and deliberately maintaining and expanding agricultural biodiversity.

When I refer to a hybrid, I mean an F1 of a cross between two different pure

breeding varieties. This is the classic genetic meaning. In the seed industry the
word is often used considerably more loosely. Sometimes Fz material or even
subsequent generations are referred to as hybrids. Sometimes even pure breeding
openpollinated varieties that were derived originally from a cross between two
varieties are called hybrids. This latter use is incorrect or deceptive.

There are two major reasons for using a commercial hybrid instead of an
openpollinated variety - vigor and uniformity. Since the classic hybrid crop is
corn, I'll use it to illustrate. First, with respect to vigor, the best commercial
hybrid corn for your area will usually outperform every available openpollinated
corn you can find, and normally by a substantial margin. The highest
commercial productivity for corn is almost always associated with the use of
hybrids. This isn't necessarily true for other crops.

The crops for which hybrids show the greatest advantage over openpollinated
varieties are those that show the largest inbreeding depres lion. Corn and many
of the brassicas display extreme inbreeding depression, so hybrids for them are
especially advantageous.

For crops that are normally inbreeding or that are outbreeding but don't show
much inbreeding depression, there is usually only a small advantage to hybrids,
if any. The major reason we have so many tomato, squash, and melon hybrids is
proprietary - it is simply better for the seed company (if it can fool us into using
them). In most cases, however, it isn't necessarily better for us.

Second, with respect to uniformity, a good commercial corn hybrid will always
be much more uniform than even the best openpollinated variety. For machine
tending and harvesting, this can be a huge advantage. Each hybrid plant is
almost genetically identical to the others, they all grow to the same size and
shape, and they mature simultaneously.

For machine harvest, any plants that are too different in size or shape don't get
harvested properly. Any that are much earlier than the harvest date usually have
deteriorated. Any that are later than the harvest date never mature at all. The
more the plants are identical and mature identically, the larger the proportion of
them that contribute to the harvest. Synchronous maturity may matter even for
hand-harvested vegetables. I harvest my chickpeas by hand, but I would rather
harvest all the plants in a patch at once instead of having to watch individuals in
the patch and harvest them one at a time as they're ready.

The uniformity of hybrids is sometimes not desirable for home gardeners. A

gardener who likes fresh broccoli might prefer her thirty broccoli plants to
mature variably so that she has a head every few days, not thirty at once.

Territorial Seed Company has an innovative solution to the problem that the
vigor of hybrids may be desirable for home gardeners in cases where the
uniformity is not. For a number of crops, they mix seed of different hybrids that
have different maturity dates. A customer can buy the pure varieties or the mix.
The mix gives you the vigor of hybrids without the uniformity in genotype or
maturity date. You can, of course, always buy a variety of hybrids with different
maturity dates and plant them in separate patches, or mix the seed yourself.

The superior performance of commercial hybrids and openpollinated varieties

is virtually always associated with optimal (commercial) growing conditions -
large amounts of chemical fertilizers and good weather. Under less than optimal
conditions, good regionally adapted openpollinated varieties may perform as
well as or even better than the best hybrid. In good years those who plant hybrids
usually have a higher yield. In bad years those who plant an openpollinated
variety may be the only ones with at least some yield. Normally the introduction
of hybrids goes along with the use of high-nitrogen fertilizers, herbicides, and
pesticides so as to bring out the full potential of the hybrid.

I don't know whether the tendency of modern hybrids to perform well under
only optimal conditions is inherent in the nature of hybrids. I suspect it is not. I
think we have that kind of hybrid simply because that is what the developers of
the commercial hybrids have chosen. Their hybrids reflect their beliefs and
values. If you want hybrids that perform optimally under low-input, sustainable,
more organic conditions, you should be able to develop them. Just start with
parental varieties that do well under those conditions and test your hybrids under
those conditions. When you're done, your hybrids will reflect your beliefs and
values.

An additional criticism of hybrids is that they are nutritionally inferior to

openpollinated varieties. I think this is often true, but I also think it has less to do
with the inherent nature of hybrids than with the specific kinds of hybrids that
are being produced. Little attention is given to nutritional value in the breeding
of food crops. Until recently it was more or less invisible. Superior nutritional
value was probably involved indirectly in the preservation of heirloom varieties,
because people would keep what tasted good (which may or may not correlate
with nutritional value) or because the livestock preferred a variety and
performed better on it. I believe that it should be possible to develop hybrids of
high nutritional value, too, if we select for it directly or indirectly.

Alan Kapuler, who is a geneticist, a biochemist, and a vegetarian, has a special

interest in the nutritional value of vegetables. He's been testing heirloom and
other varieties and building a collection of varieties that are especially nutritious.
His research on the nutritional value of vegetable varieties and the varieties
themselves are available through Peace Seeds (see Appendix D). His catalog is
probably the best starting point for those interested in breeding more nutritious
hybrids or openpollinated varieties.

Peas, beans, and many other largely inbreeding crops don't lend themselves to
the production of hybrids. In addition, there isn't much to be gained, because
they don't display inbreeding depression. There are, however, many hybrid
tomato varieties, and tomatoes also are inbreeders. A hybrid tomato industry is
possible only because most tomatoes produce several hundred seeds per hand-
pollination.

There are many openpollinated lines of tomatoes that are as good as even the
best of the hybrids. I believe the major reason that there are hybrid tomatoes is
the proprietary one - seed companies would rather have us grow something only
they can produce. The hybrids are generally not superior to all available
openpollinated varieties, and in many cases they may not be hybrids at all. They
are openpollinated lines that the seed companies would prefer to have us believe
are hybrids.

In a number of cases, especially with tomatoes and various cucurbits, if you

inbreed a so-called hybrid variety, you see no segregation in the next and all
subsequent generations. That is, the variety behaves exactly as if it were a pure
breeding variety. It's possible that the hybrid came from a cross between two
varieties that were identical for all genes involving plant size, shape, and form;
leaf size, shape, and form; and so on, but were different for something invisible,
such as disease resistance. But I suspect that in most cases where no obvious
segregation occurs when we inbreed a supposed F1 hybrid, it is because it is not
an F1 hybrid; it is a pure breeding variety. It's easy to see why false claims of
hybrid status might be advantageous and profitable.

I began to suspect the status of hybrids a number of years ago when I inbred a
couple of corn varieties and saw no segregation for anything. A few years later I
remarked to Alan Kapuler that I thought many "hybrids" might not be hybrids at
all. His response? "Aha! So you've noticed that too!"

I suspect that seed companies sometimes list an openpollinated variety as a

hybrid in order to deter seed savers or competitors. Why not call the bluff? If
you raise twenty or so putative Fz progeny and see no segregation, you may
already have your openpollinated variety.

Alternatively, the commercial variety might be a true hybrid, but might

represent a cross between two parent lines that were very similar in all the
obvious visible characteristics but different in other critical ways - such as
disease resistance. If so, your Fz would actually be segregating for disease
resistances. Just grow the plants for a few generations under your conditions,
saving seed from the best of each generation. If the material contains genetic
variability for disease resistance, you will automatically select for resistance to
the diseases of your region and growing conditions.

When you derive an openpollinated variety from a commercial hybrid, give

yours a distinctive name. Many openpollinated varieties can be derived from a
single hybrid, and they need names to distinguish them from the parent as well
as from each other.

In the next chapter, a number of examples involve creating openpollinated

varieties from hybrids, that is "dehybridizing the hybrids." In the last part of this
chapter we'll

cover why, when, and how to make our own hybrids.

There are six major reasons for making and using a hybrid instead of an
openpollinated variety. First, it may be a crop, such as corn, for which hybrids
are usually much more vigorous and productive than inbred varieties. Second,
you may have kids; making and testing hybrids is a good breeding project for
them. Third, you may need the potential uniformity of hybrids. If so, you need to
cross inbred lines of openpollinated varieties. Fourth, it is easier and faster to
make and use hybrids in certain situations than to develop openpollinated
varieties. Fifth, you may want proprietary control yourself. Finally, there may be
no other option. Sometimes we can make a hybrid between two varieties that has
all the characteristics we need, but it isn't possible to develop an equally good
pure breeding variety from it.

If you cross almost any two openpollinated corn varieties, you'll see a
substantial increase in plant vigor and yield. If you have kids, making corn
hybrids is an ideal project. It's an excellent illustration of reproductive biology,
the hybrids so outperform the parent varieties, and the entire project only takes a
year.

With kids, an enjoyable approach is to interplant two varieties that have

different colors of kernels. When you interplant a black corn and a white corn,
for example, the white ears will have some black kernels on them. Kids enjoy
picking the black kernels out of the ears for planting. If you want to make corn
hybrids more efficiently, you can use hand-pollination, or plant alternate rows of
the two varieties and detassel one of them (see appendix A).

If you want the most productive hybrids possible, you'll make crosses between
pairs of all your favorite varieties and test them under your conditions. You
should be able to find hybrid combinations that yield as well as or even better
than the commercial varieties, because you are testing in the context of your
exact growing conditions and requirements, and because your hybrids do not
need to have broad regional adaptability.

The hybrids you make by crossing two openpollinated varieties will usually
not be as uniform as commercial hybrids. If you are a home gardener, this may
be preferable. If you are a market gardener or farmer, you may need the
uniformity typical of commercial hybrids. To get it, you must cross two inbred
lines. I suggest that you first test your openpollinated varieties without
inbreeding to identify which combination of two varieties produces the best
hybrid. Then develop some inbred lines from each of those varieties.
 To develop an inbred line, just self-pollinate a few plants for two to four
generations. Then maintain the lines normally. One or more of the lines will
probably be genetically uniform enough to generate uniform hybrids, but still
heterogeneous enough to be reasonably vigorous. Inbred lines of corn do not
necessarily look very uniform, incidentally. Their lack of vigor seems to cause
them to be highly affected by minor differences in the environment. Your inbred
lines will probably look inferior and variable and yield poorly compared to the
lines from which they were derived.

With tomatoes, squash, and melons, we will sometimes make and use a hybrid
simply because it is so fast. When Glenn Drowns crossed two different varieties
of watermelon, the F, hybrid between them was early enough to ripen in his area,
though neither of the parents or other varieties were. He could have used the
hybrid for his crop instead of developing an openpollinated variety. In some
cases, the number of genetic markers involved, or their linkage, or sterility of the
hybrid makes it difficult or impossible to develop an openpollinated variety. In
these cases, using the hybrid is our only option.

In some cases gardeners, especially market gardeners, will want to explore and
develop their own hybrids and use the proprietary characteristics themselves. For
the innovative vegetable gardener who earns a living by finding and introducing
new crops and varieties, a major problem is that other growers immediately copy
them whenever they find a winner, without spending any time, labor, and money
on research. Part of the solution is to keep finding new crops so as to stay ahead
of these copycats. Those who sell produce might want to consider paying some
attention to developing their own exclusive, farm-brand hybrids that only they
can grow. People who have read this book can, of course, use your hybrids to
develop their own equivalent hybrids or openpollinated varieties, but not without
their own research and labor.

N PREVIOUS chapters, I've provided all the theoretical background you need to
breed plants. In this chapter, I show how to put it all together and use it.

The basic genetic techniques are selection, inbreeding, making crosses of various
kinds (including backcrossing), and creating and maintaining genetic
heterogeneity. Selection is a powerful technique when there is genetic variability
for the characteristic we're interested in. It's useless when there isn't. In this
chapter you'll learn how to distinguish between genetic and environmental
variability, to tell when to use selection, and to turn the basic technique of
selection into power selection. You'll also learn how to use back crossing and
recurrent backcrossing, and how to turn basic inbreeding into power inbreeding.

Crosses are the way to introduce genetic variability when your material doesn't
have it already. Crosses can be used to combine characteristics from different
varieties into one variety or to create entirely new phenotypes. In addition, a
very powerful kind of crossing, recurrent backcrossing, can be used to transfer
one or a few genes into some other variety. Many breeding projects start with a
cross. The cross creates the needed genetic variability; the other methods
manipulate the variability in various ways.

Inbreeding is used to reduce the genetic variability in the genetic material so

that it becomes predictable and reliable for the characteristics we care about. It's
a very fast, powerful way of creating new, stable varieties for crop plants that
don't suffer from serious inbreeding depression.

Corn, most of the brassicas, and many other crop plants are outbreeders that
suffer from inbreeding depression if they are made to be too genetically
homogeneous. The obvious breeding method of just saving seed from the best
plants can destroy the vigor and productivity of these kinds of varieties.
Inbreeding depression affects every aspect of creating or maintaining these
varieties. I dealt with this problem theoretically in the last chapter. In this chapter
I explain how to take it into consideration practically and how to use methods
that deliberately maintain adequate amounts of genetic heterogeneity.

Selection is the only method commonly used alone in a project. Most projects
involve a number of methods used at various stages of the project. In this chapter
I illustrate how to mix and match these methods by telling stories about my own
breeding projects and those of my friends.

How we go about doing any specific breeding project often has a lot to do with
the breeding system of the crop involved. Information about the breeding
systems of various plants is given in Appendix A and Table 1. How to figure out
the breeding system where no information is available is explained in chapter 9.
Mechanical details about hand-pollination, making crosses, and tricks for
overcoming incompatibility barriers are given in general in Appendix B and for
specific vegetables in Appendix A.
 Perennial Vegetable Buckwheat

(Selection.)

I recently obtained about a dozen seeds of a Himalayan perennial vegetable

buckwheat, Fagopyrum cymosum, from J. L. Hudson. Alan Kapuler took a few
of the seeds and planted them at his house. I planted a few at mine. I ended up
with two plants, both with attractive heart-shaped leaves. Alan ended up with a
single plant that was much bigger than mine, but he had planted earlier and in
better soil.

The first of my plants to flower did not set any seed until the second began to
flower. It appeared to be self-incompatible. Alan's plant set seed prolifically; it
was apparently selffertile. Of my two plants, one had much more intense purple
stems and leaf veins. These kinds of differences between individual plants are
common when you're working with rare vegetables or with relatively new
material.

I didn't have any a priori ideas about which characteristics were most
desirable, except that I am interested in perennial vegetables that are hardy in my
climate. So I just waited to see whether the plants would survive the winter.

All three plants died down with the first frosts but then sprouted vigorously in
the spring. The plants are very vigorous and produce copious young shoots,
which are the edible parts. All are a mild-flavored cooking green and taste about
the same.

Alan's plant and one of mine are very invasive. My invasive plant had shoots
coming up from spreading roots all across the garden bed, up to about 3 feet
away. That's pretty impressive for a plant started from seed just one season
before - impressive and frightening. My other plant, in another bed, sent shoots
up only right around the plant. This plant is the one with the most intense purple
coloration and is quite attractive.

I have been ripping up shoots from the invasive plant, eating them, and feeding
them to my guinea pigs. When I want more plants, I will make divisions of the
betterbehaved purple. Alan will probably eliminate his plant and clone from
mine, too, if it remains well behaved.

What I've just described is selection at the simplest and most basic level. In
many cases this is all you have to do. We grew only three plants, but each was
different. All were very vigorous once established. My favorite plant has
produced copious amounts of food. I chopped it off at a couple of feet high and
have chopped off shoots frequently enough to keep it contained as a good-size
bush. Now I need to learn more about how to use the plant. I'll also want to keep
an eye on its invasiveness as its root system becomes larger and better
established.

Many rare plants or collections of wild material are genetically very

heterogeneous. All you have to do is look. It's important to look with an open
mind - that is, grow some plants, watch them, live with them a bit, and then
decide what you want. Consider both the virtues and disadvantages of every
characteristic. A self-incompatible perennial buckwheat, if propagated
vegetatively by itself away from others, cannot be propagated by seed, which
can be considered

a disadvantage, for example. But it also will not be shedding seed all over the
place to become weeds, which greatly facilitates caring for the perennial
vegetable patch.
 Orange Popbean

(Selection, determining whether variability is genetic or environmental,

power selection.)

Some varieties of chickpeas are able to overwinter readily here in maritime

Oregon; others are not. Most of the large cream-colored types that resemble the
commercial chickpeas common in the United States don't overwinter here, but
many of the colored types, as well as some of the tiny cream ones, do.

Last year I planted ten seeds of the orange popbean (P1374085) I described in
Chapter 3 as part of my overwintering trials of 150 accessions. Only one plant
survived, so I am saving seed from that plant. Perhaps the plant survived because
it was genetically more cold resistant than the others. But maybe it survived only
because it was lucky - it happened to be in a little more favorable spot and got
better established before the bad weather hit, for example.

If the plant survived because it was genetically superior, I might want to use it
as the foundation for its own variety. So how can I tell whether the variability I
saw was genetic or environmental? Well, if the plant is genetically superior, it
should be able to pass on that superiority to its offspring.

I've saved seed from that plant separate from other seed of the accession (that I
planted later). I will treat that seed as if it were, indeed, a new variety. But next
year I will test that seed against seed produced by the unselected plants. If some
or all of the selected material performs much better than any of the plants in the
unselected accession, I will know that the original plant was genetically superior.
Likewise, if the average performance of the seed of the selected plant is superior
to that of the unselected material, I will know that genetic differences are
involved.

If, on the other hand, the average performance and the range of performance of
the selected material aren't any different from those of the unselected material, I
will conclude that the original plant wasn't genetically superior; it was just lucky.
I won't try to develop a more hardy variety based on that plant as the foundation.

What I am doing physically involves only keeping seed from the best plant
separate for at least one generation. But in actuality, it is quite sophisticated.
When I compare offspring of the best plants with offspring of the average, I
evaluate whether the differences I see in the overall population of plants are
genetic or environmental. This evaluation allows me to tell whether the selection
I did accomplished anything and whether further selection along the same lines
will accomplish anything.

Let's suppose that I find that a number of plants from the selected plant's seed
are much more vigorous and productive than those in the rest of the accession. I
will then use the seed from that original plant as the basis for a new more cold-
hardy variety.

If my new variety shows variability for cold hardiness, I will continue to select
by

saving seed from the best plants each year. I won't throw away the seed from the
rest of the plants, though. I will plant some as a comparison. At some point I will
have eliminated all or nearly all of the genetic variability for cold hardiness from
my variety. At that point the seed from the best plants and that from the worst
plants should perform about the same. When that happens, it means that the poor
plants were poor for environmental reasons, that they were just as good at
passing on cold hardiness to their offspring as the best plants were.

Whenever you see variability in a variety or a population of plants, you can

test whether that variability is genetic or environmental by saving the seed from
the best and the worst separately, then comparing the two the following year.
This method works whether the characteristic is based on one gene or several.
Comparing the best with the average is not quite as powerful, but usually it's less
work and is good enough.

Whenever a variety or population contains genetic variability for a

characteristic you're interested in, you can improve the variety just by simple
selection. If the variability is not genetic, then no amount of selection will be
effective in improving the variety. In order to test the population for the nature of
its variability, you just proceed as if you were selecting, but you evaluate
whether selection is actually happening. Selection plus evaluation is power
selection. Evaluation also tells you when you're done - when you've eliminated
virtually all unwanted variability and can simply maintain the variety.
 Maintaining varieties also requires selection, however. It's usually referred to
as cull ing or roguing. Whether we call it selection or culling/roguing depends on
just how much we're eliminating. When we're creating a new variety, we save
seed from only the best plants in each generation. Most plants aren't allowed to
contribute to the next generation. We call that selection. Once we have our new
variety, we keep seed from nearly all the plants. But there will still be occasional
new mutations or undesirable combinations segregating out. So before the
variety begins flowering, we examine the population and eliminate the
occasional rogue so that it doesn't contribute to the next generation.

High-Yielding Chickpeas

(Selection.)

One of my plants of small, fast-cooking cream-colored chickpeas had more

than eight hundred seeds on it. I kept the seed from that plant separate from the
rest of the accession. If the yield of that lot is better than that of the rest of the
accession, it will represent a first step in breeding for high yield. It is a common
method for use with inbreeding crops and is called single plant descent. All it
really amounts to is keeping your eyes open for really good plants and saving the
seed from them separately. But remember to compare the plants grown from that
seed with the rest of the population to see whether the seed really is different.

With inbreeding crop plants such as chickpeas, when you start with
heterogeneous material such as a land race, you can do a large amount of
breeding by simply observing the plants and noticing any that are different or
superior. As I explained in the last chapter, given the genetic nature of
inbreeding plants, it is likely that in most cases the plants I pick out as superior
are homozygous for whatever genes are involved in their superiority. When this
is true, their offspring are a new variety. I may have a new, pure breeding variety
of chickpeas simply by saving seed from a superior plant.

The Quest for `Golden Snap' Peas

(Doing a cross and going to the F2, followed by selection and inbreeding.)

Alan Kapuler wants a yellow-podded snap pea. There is an heirloom yellow-

podded snow pea available -'Golden Sweet' - that is one of the varieties Mendel
worked with, Alan says. But a yellow snap pea would be a nice addition to the
snap pea repertoire.

There aren't any yellow snap peas. Furthermore, there is no apparent variation
in color within snap pea lines. If, when you grew a snap pea variety, you saw
some plants with pods that were more yellowish, you might try to develop a
yellow snap by simple selection. But you don't see that, so you can't start the
project just by selecting. You first have to introduce the variability you need.

Alan started off by crossing `Golden Sweet' with `Sugar Snap', a green-podded
snap pea. He wanted a new variety that would have the pod color of `Golden
Sweet' and the pod type of `Sugar Snap'. To combine characteristics from two
different varieties, you start by crossing them. The most common way of
proceeding is to go to an F2 and start selection with the FZ or a later generation.

I've mentioned before that snow (ediblepodded) peas differ from shelling peas
by two recessive genes. Both are required for a snow pea that is edible when the
pods are large and the peas have started to form. An additional gene, the N gene,
is associated with pod thickness. Snap peas are ppvvnn, snow peas are ppvvN_,
and shelling peas are usually P_V_N_. So a cross of a modern snow pea variety
and a snap variety is likely to be ppvvNN X ppvvnn. In other words, it's a one-
factor cross. Likewise, the difference between yellow and green pods is
generally a one-factor cross.

I learned about the genetics of pod type and color by reading the chapter on
peas in Mark Bassett's Breeding Vegetable Crops (see bibliography). It includes
lists of the common genes involved in agriculturally important characteristics
and their linkage groups. I give information on some of the most common genes
in the most common vegetables in appendix A. Referrences to the professional
literature on hundreds of vegetables are included in Table 1. You should be able
to track back into the plant breeding literature and find out what is known about
every common vegetable, as well as hundreds of less common ones.

It is not necessary to do that, however. I usually don't, even though I am a

geneticist by training, have a full-scale agriculture library a few blocks from my
home, and am already familiar with the specialized plant breeding literature. In
many cases there is little or no information available on the vegetable I care
about. Often, even if information is available, finding it would take more work
than doing the cross and raising a few Fes. Sometimes there is plenty of
information, but not necessarily on the traits I am inter ested in (such as
overwintering ability or slug resistance). Or there is some information, but it is
contradictory.

In some cases, especially with the more common vegetables, you can find out
enough relevant information to make better guesses about what results you are
going to get when you do a cross. But you will usually do exactly the same thing
whether you have this information or not: you do the cross and go to the F2.
When you get the F2, you learn the essential genetics, whether you knew
anything about it ahead of time or not. And what you find out tells you how to
proceed.

Even if you read a lot of technical sources and make guesses about the
genotypes of your varieties and the genetics involved, you often will have to
revise your assessment based on your own results. When someone says that
some trait involves one factor, that is only with respect to that person's two
parental varieties. That tells you that the genetics of the trait in your cross may
be simple; it doesn't guarantee it.

When Alan Kapuler started his `Golden Snap' pea project, he did not know
anything about the genetics of pod types. The snow versus snap characteristic
could be a one-factor difference in his cross, or it could involve two factors or
more. He suspected that the yellow versus green pod color was a one-factor
difference, since this was one of the traits Mendel worked with, and he'd read
about it. But just because yellow versus green was a one-factor difference in
Mendel's crosses, there was no guarantee that it would be in Alan's, unless he
was doing exactly the same cross. And he wasn't.

So Alan didn't know what specific genes he was working with or what their
patterns of inheritance would be. The simplest possibility would be that pod type
would be associated with a single gene, that color also would be associated with
a single gene, and that the two genes would be independent. When he crossed
the green snap and the yellow snow pea, he got a green snow pea (F,). If it was a
two-factor cross, it was the double-recessive class he was trying to recover; the
expected frequency of the class would be

Alan raised just a few F, plants. As we have seen, the F, plants from a cross
between two pure breeding varieties are all the same, so we don't need many.
Alan grew about twenty F2 plants. More would have been better. As Appendix
H shows, he should have had forty-six plants to be 95 percent sure of getting at
least one of a class that has a probability of 1/e. But Alan had room for only
twenty plants.

Twenty is a useful number of F,s in many cases. It may not be enough to give
you what you want immediately, but it will generally give you a rough idea of
the genetics involved so that you can figure out what to do next. Often it
provides the material you need for the next stage. I often plant only twenty of the
FZ of a cross, even when I know more than two factors are involved.

It's usually easiest to decipher what is going on in a cross if you first consider
each trait separately and then consider how the traits are assorting with respect to
each other. Let's consider the two traits in Alan's cross - pod color and pod type -
separately.

With respect to pod color, the cross was green x yellow and produced a green
F1. Of the twenty or so F2 plants Alan raised, all had pods that were either green
or yellow. There were no new colors or shades of colors. Most of the F2 pods
were green, but a few were yellow. This pattern suggested that pod color in the
cross was associated with a one-factor difference between the two varieties and
that there was simple dominance. More complicated explanations are possible,
but in most cases where you see that pattern, one gene with simple dominance is
what you're dealing with.

With respect to pod color, Alan's cross was GG (green) X gg (yellow), and the
F, was Gg (green). I chose the letters G and g to use as the symbols; you can use
anything, as long as the two symbols you choose have some sort of logical
relationship that makes it easy to remember that they represent alleles of the
same gene. Conventionally, capital letters are assigned to the dominant allele.

With respect to pod type, Alan also got results that were consistent with a
singlefactor cross. The cross was snow x snap. The F, was snow, and the F2 had
two discreet classes, snow and snap, with most of the plants being snow. We can
portray the cross with respect to pod type as SS (snow) X ss (snap); the F, would
be Ss (snow). (We wouldn't normally know what symbols the professionals have
assigned to the genes, so we would choose working symbols of our own.)
 By assigning a different set of symbols to represent pod type, we're assuming
that differences in pod type and pod color are asso ciated with different genes.
This is a good guess, but it doesn't necessarily have to be true. Some genes are
pleiotropic - that is, they have several different apparent phenotypic effects. For
example, the genotype A_ is necessary for formation of the purple pigment,
anthocyanin, in peas. It can affect the color of the pods, the stems, the flowers,
and other parts of the plant, depending on what other genes are present. So
flower color and pod color can be associated with a single gene in some crosses.
In other crosses flower color and pod color may involve two genes, with one
determining whether there is any purple pigment in the plant and the other
determining whether it shows up in the pods. It's common for a gene to have a
phenotypic effect and an effect on growth rate, general vigor, or fertility as well.

Pod color and pod type segregated from each other in the F, plants of Alan's
cross, so they must be associated with different genes. Alan's results were typical
for a two-factor cross, with one factor associated with each trait and with simple
dominance for both loci.

Having looked at each trait separately, we then look at how they behave
together - that is, are they assorting independently, or are they showing evidence
of linkage? In most cases we will find the two genes to be independent, because
they will be on different chromosomes or will be far enough apart on the same
chromosome so that crossing over recombines them freely.

The theoretical expectation in a two-factor cross is a 9:3:3:1 ratio if the two

genes are assorting independently. If they are linked we expect a preponderance
of the parental types. But there isn't much use talking about the exact ratio when
you only have twenty plants. What's more relevant is whether all the expected
classes showed up, and how frequent they are, roughly speaking, relative to each
other.

What an expectation of 9/16GS_ + 3/i6G_ss + 3/16ggS_ + 1/16ggss translates

to if you only have twenty plants is that you expect more of the class with the
two dominants than of anything else, and least of the class with the two
recessives. With twenty plants, you would often not get any of the class with the
two recessives, in fact. But twenty plants is enough to tell you that it's most
likely that two factors are involved in the cross, one associated with each trait,
and that they are inherited independently. That analysis forms the basis for your
decision as to how to proceed.

Knowing that, you would realize that your cross was a completely reasonable
approach to getting what you wanted, and should do the trick, but you just need
a few more F2 plants. And it might have been that what you needed was one of
the more common classes, so twenty plants might have been enough. In addition,
twenty plants is frequently enough to give you better material for the next step
than the original cross.

As it turned out, Alan got all four classes in the F2. He had more green snows
than anything else, a number of green snaps and yellow snows, and only one
yellow snap. The appearance of the F, and the classes in the Fz indicated that the
yellow snap must be genotype ggss - that is, it would be pure breed ing for
yellow snap peas. Alan could save the seed from it and start increasing it, with
the expectation that he already had, in only two years, a pure breeding yellow
snap pea variety.

But there was a catch. The yellow snap pea was setting only one pod per node.
`Sugar Snap' and `Golden Sweet' both usually set double pods. The F, had
probably set doubles. There was variability for pod number in the Fz, so the
parental varieties must have had different genes for double podding. At least two
genes must be involved.

In other words, when Alan did the cross to generate material that had the
required variability for pod type and color, he introduced unwanted variability
for pod number. This sort of thing happens all the time. Anytime you do a cross
to introduce the variability you want, you usually also introduce variability you
don't want and then have to eliminate it.

It wasn't possible to tell much about the inheritance of pod number in the
cross. However, it seemed possible that the single podding might be associated
with one or more recessives. That might mean that the yellow snap with single
pods might give rise only to single-podded offspring, which could be increased
only into a single-podded variety. No doublepodded peas would segregate out.
For this reason Alan did not want to depend on the offspring of the yellow snap
for a new variety.

If single podding was associated with dominant genes, the single-podded

yellow snap's progeny might include some doublepodded types. In that case,
Alan could derive a doublepodded variety from the progeny of

the single-podded pea. So he didn't want to reject the single-podded yellow snap
either.

He saved seed from the single-podded yellow snap. But he also saved seed
from green snaps (G_ss). He knew that some of the green snaps would be
heterozygous for yellow, that is, Ggss. In fact, twothirds of them would be. The
reason for the number twothirds: The F2 is (1/4 homozygous green + 1/2
heterozygous green + 1/4 yellow). There are twice as many heterozygous greens
as homozygous greens. That is, of the greens, twothirds are heterozygous for
yellow. Their genotype is Ggss. And they will produce the desired yellow snaps
as one-fourth of the next generation.

Alan also kept seed from the doublepodded yellow snows (ggS_). Two thirds
of them would be ggSs, and would also produce the desired yellow snap class as
one-fourth of the next generation.

When Alan raised only twenty Fzs from his cross, that turned out to be enough
to get one of the double-recessive class - but it wasn't enough because of other
things segregating in the cross. The twenty Fzs were enough, however, to give
him essential information about the genetics involved, and he could use the
information to make good choices about how to continue the project and get
what he wanted with an additional step.

Normally, even if the desired class is obtained in the F2 generation, you raise
seed from a number of the plants and select further. Anytime you do a cross,
other things will be segregating besides the genes you're primarily interested in.
Some plants will be more vigorous than others, some will be earlier, and so on.

The basic pattern, though, is that you do the cross, raise a few F,s, and then
raise twenty or more Fes. If you obtain the desired class in the F2, you self-
pollinate the plants (if it is an inbreeding crop). If you don't get what you want in
the F2, you use what you've learned about the genetics from the results of your
cross to decide how best to proceed - whether to try a different cross, stick with
the one you have and raise more Fes, or use some of the Fes to derive what you
want in one or more additional steps. After you have some of the desired class,
you increase your interim variety by inbreeding further and selecting the seed
from the best plants in each generation.

If many genes are involved in the cross instead of a few, the process takes
longer, but the same general approach applies. If some of the genes are linked
instead of independent, that can either speed things up or slow them down, but in
most (but not all) cases the basic approach is still effective.

You may or may not know anything about the specific genes involved in your
cross, but in most cases that doesn't change what you actually do. You'll use your
knowledge of the general principles as applied to the actual results from your
own crosses to decipher the basic genetics involved. Then you'll be able to figure
out how best to continue.
 PurplePodded Peas

(Crossing, selection, power inbreeding, backcrossing, recurrent

backcrossing.)

I want a purplepodded snap pea. I don't know of any. The snap peas we have
don't show any variability for purple color, so if I

want a purple snap, the best way to start would be to cross a purple snow pea and
a green snap pea. But there aren't any purple snow peas either. Every purple pea
I've seen has been a shelling pea.

The purple shellers are beautiful. The plants are intensely green, the stems are
greenish purple, the flowers are two-tone maroon, and the purple pods are lovely
against the background of the foliage. Even varieties described as being
ediblepodded have always turned out not to be when I checked them out.

I obviously need to start by crossing a purple nonsnap to a nonpurple snap and

using genetic recombination as the basis for developing a variety that is both
purple and a snap. But I also want the enation resistance and powdery mildew
resistance that matter in my area. So I started by crossing a purplepodded
shelling pea with `Oregon Sugar Pod II' to develop a diseaseresistant purple
snow pea.

There are two main diseases involved, and the purple shelling pea isn't
resistant to them. The enation probably involves one gene and the powdery
mildew two. In addition, there are two genes involved in the difference between
shelling and edible pods; the edible pods are the combination of the two
recessives. That's five genes all together, and for at least four of them, I need the
recessive combinations.

What I've read about pod color suggests that it takes three dominant genes to
produce purple pods. One is involved in anthocyanin production in the entire
plant; the others extend it to the pods. To use arbitrary letters, let's say A B_C_ is
the genotype required for the purple pods, ddeeff is the genotype required for the
resistance to the two diseases, and ppvv is the genotype associated with the
edibility of the pods.
 The cross would be AABBCCDDEEFFPPVV (purple, disease sensitive,
shelling pea) x aabbccddeeffppvv (green, disease resistant, snow pea). I'm
guessing about parts of the genotypes, assuming the worst. The purple variety
might carry one of the diseaseresistance-conferring genes, for example. And the
green variety might have one of the dominants that helps produce purple pods.
But I should assume the worst, since any cross can introduce additional factors
that I don't know about. (Such as the single-pod-pernode characteristic I've
already mentioned.)

This is an eight-factor cross. I want a new variety of genotype

AABBCCddeeffppvv. I could derive it very easily if I could obtain an FZ plant
of the A_B_C_ddeeffppvv class. But how likely is that? Is it at all possible for a
person whose total pea breeding patch is 18 feet of wide row?

I can calculate it, assuming all the genes are unlinked. With this many genes,
that assumption is questionable, but it gives me a first guess. The probability of
getting the desired phenotypic class is 3/4 for each of the dominants and 1/4 for
each of the recessives. So the probability of getting an individual of phenotype
A_B_C_ddeeffppvv in the F2 is (3/4)3(1/4)5, which is 27/65,536, or 0.0004.
Four peas in ten thousand. That's totally outside the range of numbers I usually
deal with. Even if I had the space for ten thousand peas, I wouldn't have the time
to evaluate them to find the phenotype I want.

But there are other approaches and tricks.

First, I don't have to get the class I want all in one step in the F2. I can do it in
more than one step. Second, I can derive subvarieties that have some of the
characteristics I need, then make a cross between two of those to obtain the final
variety. Third, I can use a trick I refer to as power inbreeding. Fourth, I can use
an especially powerful different kind of technique called recurrent backcrossing.

Let's say I am willing to devote 20 feet of wide row to the project, with the row
five peas wide and thirty peas per foot. That's three hundred peas total. In all
cases I would start the same way. I would cross the purple shelling pea to the
green snow pea. Then I would grow a few F2 plants; that would require just a
few inches of space. Then I would grow three hundred F2 plants.

For approaches 1 or 2 - using more than one step to obtain the desired class or
creating subvarieties - I want the class A_B_C_ ddeeffppvv, but I won't try to get
it all at once. I realize that the purple part of the phenotype is associated with
dominant genes. (The phenotype of the F,s being purple tells me this.) I might,
for example, save purplepodded snow peas and eliminate the others. That is, I
would pay no attention to disease resistance. I would probably be planting very
early so as to get the most and best-quality seed anyway (or overwintering the
plants), and under these conditions, the diseases normally don't show up.

Purplepodded (A_B_C_) peas are expected at (3/4)3 of the F2, or 27/64. Snow
peas (ppvv) are expected at 1/i6. So purple snow peas should be (27/64)(1/I6) of
the total, or about 0.026 of 300, or about 8. (This assumes none of the genes are
linked.)

I would then have a purple snow pea, and I would probably just start using it
as such. The working purple snow line would be pure breeding for ediblepodded
pea type, but not necessarily for purple color. I would cull out any greens that
appeared in subsequent generations. I could start growing and using the pea and
select for resistance to the various diseases in whatever year they happened to
occur in.

But what if none of the peas were resistant to all the diseases? Whenever a
disease appeared, I would take advantage of its presence to isolate sublines of
my variety that were resistant to it. I might not have a line resistant to both, but I
would be very likely to have lines resistant to one or the other. I could then cross
a line resistant to one disease with the line resistant to the other and recover a
variety resistant to both in the F2 or beyond.

Approach 3 is what I call power inbreeding. Start out as before by making and
raising an Ft and an F2. Grow three hundred F2 plants, as before. Save one seed
from each of the three hundred plants and plant those seeds the following year.
Then save one seed from each of those plants, and plant them the next year. And
so on. In the purest form of this approach, you don't do any selection at all. In
this case I would modify the design slightly and keep seed from just the purple
plants each generation.

This approach makes use of the fact that inbreeding reduces the heterozygosity
by one-half each generation at all loci except those involved in traits that we are
selecting for (see Chapter 9). That is, in the F2 the probability of each locus's
being heterozygous is 1/2. In the F3 it is 1/4 (assuming we aren't selecting
deliberately for or against anything). In the F4 only 1/s of the loci are
heterozygous. And in the F5 only 1/16 are. This means that in an unselected F5
generation, the genotypes in the population are (15/32AA + VI 6Aa + 15/32aa).
In other words, the probability of obtaining genotype AABBCCddeeffppvv is
about (1/2)8, or 1/256. And this is if we did no selection at all.

In other words, if we merely inbred each generation without any evaluation

whatsoever or any selection at all, raising three hundred seeds each generation
for five generations, we would have a 1 in 256 probability of obtaining the exact
genotype we want, AABBCCddeeffppvv. That's within the realm of possibility.

The slight modification of keeping only the purplepodded plants would

eliminate most of the genes associated with green pods by the F5 so that nearly
all the F5 plants would be AABBCC. The probability of obtaining the rest of the
desired genotype in the F5 would be about (1/2)5, or '/32. With three hundred
plants, we should get about nine with this genotype.

Practically speaking, we wouldn't be able to identify the diseaseresistant part

of the phenotype (assuming growing the peas very early so as to avoid disease
and get the best seed). So we would just keep all the purplepodded snow peas
and start growing them as a variety. Disease resistance could be selected later
with material planted later in years when the various diseases materialized.

With naturally inbreeding crop plants, just inbreeding for several generations
followed by evaluation at the F4i F5, or F6 stage is one of the most powerful
methods. With peas, where the crosses are difficult to do and a successful one
produces only a few seeds, it is often necessary to go to an F3 or an F4 just to
obtain enough seed so that rare classes are likely to be found. Inbreeding has the
added advantage of allowing more of the classes to present themselves. Initially
after the cross, the classes associated with the dominant genes predominate;
those associated with the recessives are rare.

Whenever we do a cross, we never begin selection in the F,. The earliest

generation in which we begin selection is the F,. But now you can see that there
are many advantages to delaying selection to even later generations in many
cases.
 Delaying selection until late in the project also is often desirable for technical
reasons. Some characteristics cannot be evaluated or selected for every year or
under all conditions. They may depend on the weather or on what diseases
matter that year. Other characteristics are simply difficult to evaluate - flavor, for
example. I would much rather evaluate the flavor of peas in an F5 generation
where each pea is fairly likely to be pure breeding for nearly everything than in
an early generation where the flavor might not yet be established.

Whenever you design a breeding project, figure out not just what class or
classes you want but also how you are going to identify them and how difficult
that is likely to be. It's much easier to evaluate green versus pur

ple pods than whether pods are edible types or shelling types, for example. And
flavor is always difficult to evaluate because it is subjective and it usually varies
with stage of maturity and time of day.

The final approach is backcrossing. To use this approach, I make and grow the
F1, as before. That is, I cross the purple sheller (AABBCCddeeffPPVV) with the
diseaseresistant snow pea (aabbccddeeffppvv) and obtain the F,, which is a
purple sheller (AaBbCcDdEeFfPpVv). Then I cross the F, plants to the snow pea
parent. This is called a backcross - that is, a cross back to one of the parents.

Let's look at what happens at some specific locus, say D, when we do the
given backcross. The F, is Dd. The backcross is Dd x dd. The Dd parent
generates (112D + 1/2d) gametes. The dd parent contributes only d gametes, so
the progeny of the cross will be (1/2Dd + 1/2dd). Ratios involving one-halves
are also typical of all the other genes in the cross, whether dominant or recessive.

The class we want is AaBbCcddeeffppvv. The probability of getting each part

of the genotype in the progeny of the backcross is 'z. The probability of getting
the whole thing all in one step is (12)$, or 1/256. In other words, if we grow
three hundred plants from a backcross, we have a decent chance of getting the
desired class in one step.

As before, though, we wouldn't be able to identify the disease resistance part

of the genotype, so we would just take all the purple snow types and go from
there. The purple part of the genotype is all heterozygous from this kind of cross
- that is, AaBbCc - so we would have to select against greens in subsequent
generations.

Let's look at another aspect of backcrossing. Let's suppose there is one

dominant gene, L, that I would like in one of my pea varieties. I like my variety
exactly the way it is; I just want L in it. I don't want other characteristics from
the variety that carries L, just the one gene. First I would cross the two varieties.
Then I would backcross to my standard variety. Half the offspring would show
L. I would choose one with L and backcross again. Again, half would show L. I
would choose one with L and backcross yet again.

Let's look at what is happening in the genome other than at the L locus.
Consider some other gene that was different in the two varieties, say M. Suppose
the L variety is mm, and my standard variety is MM, where M is dominant. The
original cross is LLmm x 11MM. The F, is LIMm. The progeny from the
backcross are (V 2LI + '211)(12MM + '2Mm). We choose one of the Ll's. Of
those, half are MM and half are Mm. We backcross again. If the plant we chose
is MM, all the progeny are MM; they are fixed for the genotype of our standard
variety. If the plant we chose is Mm, the probability is 1 in 2 that the genotype
will become MM among the progeny. Considering both possibilities, the
probability is now 34 that the genotype is that of our standard variety. With
another generation of backcrossing, the probability will be 7/8 that the genotype
is that of our standard. The same thing is happening at all the loci.

By backcrossing to a given variety, the recurrent variety, we can eliminate

nearly all the genes of the other variety except the specific one we are selecting
for in each generation. Recurrent backcrossing is useful only with dominant
genes. With dominant genes, however, it allows us to transfer an individual gene
from one variety into a desired variety. This is very important in plant breeding,
because often we have a variety that already includes large numbers of genes
that are necessary for the production of what we want in our area.

Wild relatives of crops often have dominant genes that confer resistance to
insects or diseases. These are usually transferred into crop varieties by recurrent
backcrossing. A cross with the wild relative often introduces so much
undesirable genetic variability that good crop types cannot be recovered in an
F2, even if thousands are raised. Recurrent backcrossing allows the breeder to
transfer just the desired gene (or as close as possible to just the desired gene)
into a desired genetic background. Whenever you want just one characteristic
from a given variety and want the rest of the characteristics of the new variety to
be identical to those of an established variety, consider backcrossing.

Where the characteristic in question is associated with recessive genes, you

can still use some backcrossing as part of your approach. You can alternate
backcrossing and inbreeding, for example. The inbreeding is necessary to
identify the desired gene.

You cannot always transfer just one desired gene. The genes that are tightly
linked to it tend to come along. So after recurrent backcrossing starting with a
wild plant, it may not always be possible to eliminate all the undesirable wild
characteristics.

Let's go back to the purplepodded pea. If I use recurrent backcrossing, I don't

have to evaluate for pod type or disease resistance. It allows me to fix a
diseaseresistant genotype, even if the diseases never appear at the appropriate
times during the project. It also doesn't require that I grow three hundred plants.
All I need to grow is enough plants to obtain just one of genotype AaBbCc for
the next backcross. The probability of that is 's. By referring to Appendix H, we
see that to be 95 percent sure of getting at least one plant of a type that has a
probability of '8, we should grow twenty-two plants. That will take me less than
1 foot of row. The entire project can be done in 1 foot of row or less per year
until the backcrossing is over and it's time to inbreed. Even then, only a few feet
of row will be required. With so little space required and no evaluation needed,
the recurrent backcrossing approach can be done in a greenhouse so as to allow
two or three generations per year instead of just one.

It's obvious that backcrossing has some powerful advantages. With peas that
are naturally inbreeding, backcrossing is much more work than inbreeding,
though. To obtain twenty-two seeds, I probably should have at least eight
successful backcrosses. To get those, I would probably need to do about twenty.
That's laborious. To get twenty-two Fzs, I would just have to let one plant set
seed all by itself.

An added consideration is that inbreeding allows new phenotypes to show up

that derive from genes from both parents - phe

notypes you weren't thinking about or planning on. Backcrossing is much more
controlled. You are much more likely to get just what you want and not to see all
kinds of other possibilities.

With all these approaches possible, how am I going after purplepodded peas?
Actually, I'm doing a little of everything. Last year I

crossed a purplepodded sheller to one of my snap pea lines and obtained an F1.
This year I raised three F1 plants and used some of their flowers for backcrosses
to the snap pea line. I let the rest of the flowers produce FZ seed. Interestingly,
although one parent had ordinary green seed and the other had tan seed, the Fz
seed shows segregation for seed colors of various shades and with speckling of
various shades and amounts. It's really beautiful. I have been spending more time
than I will admit just pouring the seed back and forth in my hands and looking at
all the lovely patterns.

I will raise most of the Fz seed to get an F3. And I will backcross the first
backcross generation if any are purplepodded. (I have few enough that I might
not have any.)

This year I also made a number of crosses with a different purple shelling pea
and my snap pea material as well as also crossing it to a diseaseresistant sugar
pea. I'll use some power inbreeding and some backcrossing. You don't have to do
just one or the other. You can, for example, backcross the F, once, then inbreed
the progeny. This gives you a genome that is 3/4 that of the variety you
backcrossed to, on average, instead of '/z. Where most of what you want is
associated with one variety, just a single round of backcrossing before you start
inbreeding can improve the odds dramatically and be very useful.
 Fixing Dominant and Recessive Genes

(Using progeny to identify genotypes.)

Fixing a gene or trait in a variety means to make the variety pure breeding for
it. With respect to pod type in the purplepod project, once I select a plant that has
snow-type pods, it is genotype ppvv, and the pod type is fixed automatically with
just the one act of selection. So fixing a recessive gene that is fully penetrant
(that expresses itself every time) is very easy. I just choose plants that have the
phenotype associated with the recessive gene.

With disease resistance, it may take more than one generation of selection,
because not all sensitive plants necessarily show up. If no aphid that carries the
disease bites the plant early enough to matter, the plant may be unaffected, even
though it is genetically sensitive. Professional breeders frequently inoculate
plants with virus or fungal spores to ascertain their genotypes. Most amateurs
usually depend on just selecting in years when the disease is prevalent and not
worrying about it otherwise.

Some genes express themselves only in combination with other genes. These
genes also usually require more than one round of selection for fixing.

There are two basic approaches to fixing dominant genes. One is to eliminate
the recessives by culling them. To take a specific example, in the purplepod
project we usually design the work to obtain individuals that are A_B_C_. But
we want a variety to be pure breeding for purple pods. That is, we want
AABBCC. The first approach is to massselect - to save seed from all the
purplepodded plants. In the next generation we save seed from purples again.
Each generation the recessive genes associated with green color will become less
and less frequent in the population. This is an easy method, but it doesn't totally
eliminate all the undesired recessive genes.

The other approach is to find a plant that is AABBCC and save seed from only
that one. But all three genes are dominant, so there is only one way to identify
which purples are pure breeding for purple: save their seed and see which ones
give rise only to purple offspring. In other words, we do exactly as we would
otherwise, but we plant the seed from each A_B_C_ plant separately. Some
families will be all purple podded. Others will have both purplepodded plants
and green-podded ones. We eliminate all the families that produce green-podded
plants and save seed only from those that breed true for purple.

If a plant is an outbreeder with inbreeding depression, we normally would not

want to derive a new variety by any means that causes a genetic bottleneck
eliminating most of the needed genetic heterogeneity. For an outbreeder, we
would not try to eliminate all the recessives by identifying a single homozygous
plant and deriving the variety from it. We would probably use mass selection
instead. This would allow us to maintain general heterogeneity; it would nct
allow us to eliminate all the unwanted recessives. We would probably feel that
we could more readily deal with having to cull out occasional recessive
phenotypes each generation than risk ending up with a weak, low-yielding, but
totally pure breeding variety.

If we absolutely could not afford an occasional unwanted recessive phenotype

in the variety, we might generate a number of families that were identified as
homozygous by inbreeding. Then we might combine them and let them cross
with each other to restore heterogeneity and vigor.

A More Winter-Hardy `Green Wave' Mustard

(Crossing and mass selection.)

In the very severe winter of 1990, all my `Green Wave' mustard died. I had
planted several beds of it on the farm in independence, Oregon, but not a single
plant of thousands survived. I had never had `Green Wave' die out completely
like that, although there have been other winters in which the plants were
damaged. I don't like either event. I want a more winter-hardy `Green Wave'.

`Green in Snow' is generally considered to be the most winter-hardy mustard. I

haven't grown it because it is listed in various seed catalogs as a mild type with
no special virtues other than its cold resistance. This might be because we
Americans don't know how to use it. I understand it is a Chinese variety and is
one of their most popular types. It's salted there, and the salted vegetable is stir-
fried. Both `Green Wave' and `Green in Snow' are Brassicajuncea types. I have
begun to accumulate these types to trial them and find out more about how to
use them.
 In the meantime, I have crossed `Green Wave' and `Green in Snow', with the
idea of deriving a new variety that has the vigor, insect resistance, and flavor of
`Green Wave' but the cold hardiness of `Green in Snow'. My hope is that I'll be
able to develop a version of `Green Wave' that will survive any winter here, that
will be vigorous enough to be a good winter green-manure crop, and that will
have the spectacular flavor of `Green Wave'.

I made the cross in both directions - that is, I used some `Green Wave' plants as
the female as well as some `Green in Snow'. Pods developed, and I obtained a
few dozen good seeds from the crosses. I will plant most of the seed in the fall
and let it overwinter to obtain big plants that produce a lot of seed. (I'll hold back
some seed in case we have another severe winter; the F, hybrid might not be
winter hardy.) Then I'll collect Fz seed and plant a substantial plot of that
somewhere to get F3 seed.

Brassica juncea types are said to be mostly inbreeding, but they are insect-
pollinated, so there will undoubtedly be both inbreeding and outbreeding going
on in my Fz and F3 patches. I don't mind. I just want to create a population that
contains plenty of cold-hardy plants. I won't start selecting until the F3 and
subsequent generations. If we have no severe winters before then, I'll do my
initial selections based on plant vigor, form, flavor, and suitability as a winter
green-manure plant for our area.

I also will release F3 seed as breeding material through seed companies so that
other gardeners can get involved. The F3 of the cross between `Green Wave' and
`Green in Snow' should contain all the variability needed to select mustards of
many different types suitable for many different areas.

After the initial crosses to obtain the F,, I won't do any controlled crosses at all.
I'll just grow a mustard patch and allow the plants to cross-or self-pollinate as
they wish. Brassicas frequently display inbreeding depression. I'll generate
plenty of genetic heterogeneity from my initial cross. After that, I'll have
hundreds of plants or more in each generation. In this way I should be able to
maintain nearly all the heterogeneity created by the initial cross, except for genes
specifically selected against in later stages.

Lettuce-Salsify
 (Uncontrolled crosses to generate a genetically heterogeneous population,
selection.)

Dan Borman (Meristem Plants) became interested in scorzonera a number of

years ago. (Scorzonera hispanica is referred to as scorzonera, black salsify, or
oyster plant.) Like me, Dan was interested in the possibility of using the leaves
as a perennial lettuce instead of eating the roots.

I don't know of any information about the breeding system of scorzonera, but it
has large, attractive yellow flowers that are visited by bees. A good guess is that
scorzonera is at least mostly cross-pollinated. It may well be subject to
inbreeding depression. If so, mass selection, not selection of individual

foundation plants, would probably be the best choice of methods. Dan started by
doing uncontrolled mass crosses. (With inbreeders, if you want a cross, you have
to make one. With outbreeders you have the option of letting them do it for you.)

Dan obtained all the varieties of scorzonera he could find from as many
different sources as possible and interplanted them all in one patch. He would let
them cross in all possible combinations so as to generate a population with as
much genetic variability as possible. He planted the seed that resulted from his
first patch of plants in another patch. Then he saved the best plants and culled
out the rest, saved seed again, and planted in a third patch. By this time he was
finding many plants with interesting new phenotypes. Some had leaves 4 inches
across (instead of 1 or 2 inches), leaves with petioles (instead of sessile), and
leaves that were much more tender than any of the material he'd started with.

With each generation Dan makes sure that he keeps at least a dozen plants,
which he allows to cross-pollinate freely. In this way he hopes to maintain a
good level of general genetic heterogeneity.

There has been one hitch in Dan's breeding program. Deer prefer his third-
generation lettuce-salsify. They will walk right past the patch of his original
material in order to eat his selected plants. Greater pest problems are often an
inevitable problem where the variety is to be used as a raw vegetable.

The other side of the coin is that when you deliberately select for pest
resistance, you need to check and make sure the plant is still edible. Many plants
are pest-resistant because they are poisonous or unpalatable. A professional
breeder recently developed a more pest-resistant potato - that turned out to make
people sick. It had higher levels of the poisons that make wild potatoes resistant
to pests - and less desirable to people. I sometimes identify milder, tastier wild
plant individuals by checking those with the most insect damage. The most
delicious daylily flower of the kinds I'm working with is the only one that shows
marked pest damage. I am not the only one who thinks it is a superior raw
vegetable, apparently.

Dan started his project by mixing a number of varieties and allowing them to
cross. The purpose of the crosses was to create genetic heterogeneity. But there
was no way of knowing which cross might generate the heterogeneity needed to
select better leaf types. So Dan, in effect, did all the possible crosses.

A somewhat more rigorous way to generate heterogeneity would be to actually

do all the possible crosses, pool the F1 seed, grow the mixed seed without
selection and allow it to cross however it wanted, and then start selecting the
following year. This design does a better job of making sure that you get all
combinations and possibilities, but it also takes much more work. Dan's
approach generated many combinations, though probably not all, and it made
optimal use of his time.

When you're working with cross-pollinating or partially cross-pollinating

varieties, doing crosses by just mixing and interplanting the varieties is often the
most practical method simply because it is so effortless. In many cases hybrids
between the various varieties are so much more vigorous than the original
material that they can be readily identified in the next generation.

The more breeding projects you have that take care of themselves, the more
breeding projects you can have going at once. And with a lot of breeding
projects, nearly every week presents at least some minor triumph.

Tomato, Corn, Squash, Melons (Dehybridizing the hybrids.)

Many of the accomplishments of modern professional plant breeders are

presented in the form of hybrids. If you know how to dehybridize the hybrids
you can transform these accomplishments into seed that you control. Alan
Kapuler is passionately committed to finding, preserving, and distributing
publicdomain openpollinated varieties. When there is a particularly excellent
hybrid, though, he doesn't do without it. Instead, he uses it to create an
equivalent publicdomain, openpollinated variety. He dehybridizes the hybrids.

In a number of cases, Alan has run into what I referred to as pseudohybrids.

That is, when he attempted to develop an openpollinated version of the hybrid,
the very first step he took - saving seed of the putative hybrid and growing out
the FZ - gave evidence that the "hybrid" probably wasn't a hybrid at all. It was
apparently already an openpollinated variety. In a number of other cases, the
hybrids showed the expected segregation in the F2, and Alan developed his
openpollinated version by selecting through subsequent generations.

Alan had a hybrid sweet tomato, for example, that was very sweet and very
produc tive. He wanted an openpollinated version. So he simply saved the seed
and planted it. Tomatoes are highly self-pollinating. If the "hybrid" were really
an F, hybrid, the plants resulting from its seed would be the F2 generation after a
cross. They would be expected to show segregation for all the characteristics that
differed between the two parents that went into the original cross.

But Alan saw absolutely no segregation. The supposed Fz plants all looked
exactly the same. In other words, the "hybrid" variety didn't act like a hybrid
genetically. It's offspring were all as similar as could be to each other and to the
parent. That is, by definition, what we mean by a "purebreeding" variety.

Alan selected a line of openpollinated material that he offers through Peace

Seeds as `Peacevine Cherry'. He suspects that the original hybrid was actually an
openpollinated variety to start with, however, and that `Peacevine Cherry' is
identical or very similar to the original commercial variety.

I think seed companies may sometimes claim a purebreeding variety as a

hybrid in order to deter people from saving its seeds. I have inbred a number of
commercial hybrids and obtained no segregation. This means that either the
"hybrids" were not hybrids, or that they represent crosses between parent lines
that were different only in characteristics that are invisible (such as disease
resistance, for example, when you don't have that disease in your field).

Most varieties listed as hybrids genuinely are hybrids, though. When you save
the seed from them, they show segregation for various characteristics in the next
generation.

Alan Kapuler was impressed with the virtues of a full-season hybrid white
sweet corn, for example, that had nice big ears, good yield, excellent rich flavor,
and attractive, vigorous pinkish-purple plants. He wanted an openpollinated
version. So he just grew the hybrid in a patch by itself and saved the seed. When
he planted it the next year, he obtained vigorous plants with big, uniformly good-
flavored white ears. But there was segregation for plant color. There were
purples and pinks and greens. Alan liked the deep purple plants with their
beautiful purple-husked ears.

So he detasseled the green plants and ate the ears, and saved seed from about
two hundred to four hundred F2 purple plants. He raised about a thousand F3
plants. Only about 5 percent or so were green. He culled the greens again,
always detasseling them before they flowered so that they could contribute
neither as male nor female parents. His openpollinated line began with the F4
material, when he began offering it through Peace Seeds as `True Platinum'.

In order to develop `True Platinum', Alan practiced mass selection starting

with the F2 after a cross. The cross happened to be one someone else made, but
that doesn't change anything. Notice the numbers of corn plants involved. Two to
four hundred F2 plants were the foundation of the line. Not one or two or twenty.
These are the kinds of numbers you should be working with if you want to create
or maintain a vigorous openpollinated corn variety.

Alan developed his openpollinated variety through mass selection instead of

familial selection. This allowed him to keep the numbers high but the work low.
Mass selection has the disadvantage that you never completely eliminate
undesired recessive genes. This means that a few percent of green plants will
always be characteristic of `True Platinum'. With familial selection you could
identify the genotypes of families and pool those that were homozygous for
purple. But the work involved would be too much to do with more than small
numbers. So the variety would end up being founded on a smaller number of
plants.

With corn, it will usually be far more important to work with large enough
numbers to maintain vigor than to eliminate every single unwanted recessive
gene. Most vigorous openpollinated varieties of outbreeding crop plants are
developed via mass selection and contain a small frequency of various unwanted
recessive genes. So they produce a low percentage of off-types every generation.
This is part of the reason why roguing every generation is especially important
with these crops.

Alan also liked a golden patty-pan-type summer squash that was a commercial
hybrid. He wanted an openpollinated line that was similar. So he hand-pollinated
and selfed the variety, then saved the seeds. When he planted the seeds the next
year, he got all kinds of things segregating out - squashes of green and gold, and
of all shapes. It took him five years of inbreeding from plants of the desired type
to select his openpollinated version, which he offers as `Summer Sun'.

Alan also dehybridized a popular hybrid mustard. He simply planted the mustard
in a patch by itself and saved the seed. The Fz showed lots of segregation for
plant type and color. Alan massselected over a number of generations to develop
his openpollinated variety `Prime Choi'.

`Rainbow Inca' Sweet Corn

(Mass Crosses and Mass Selection.)

The final story in this chapter is about Alan Kapuler's creation of `Rainbow
Inca' sweet corn. It was his first breeding project and took place before Peace
Seeds was founded, a time when Alan was living on a commune in southern
Oregon and growing vegetables. He grins when asked about it and says, "That's
my hippie corn."

`Rainbow Inca' didn't start as a breeding project. It began as a spiritual act, a

ceremony. Alan had grown a number of different varieties of corn the previous
year. The ears were of all kinds and colors - flour corns, native Indian corns,
heirloom sweet corns, and others. Some of the ears were especially beautiful.

Alan chose the twelve most beautiful ears and took them out to the field. Then
he shelled each ear and planted it. He shelled the kernels off the ear in order, row
by row, seed by seed, as he walked down the rows in the field - planting as he
went - transferring the pattern of kernels in the ear onto the land. "It was like
unrolling those ears of corn on my field," Alan said.
 When he finished planting the kernels from one ear, wherever he was in the
row, he started the next ear. So the corn from the different ears was all in one
patch, in somewhat intermingled blocks.

Then came the moles. They were espe cially active that year, and they ate a lot
of the corn. Alan had collected a number of heirloom sweet corns. He planted
sweet corn into every spot in his field that was missing a corn plant, without
paying any attention to varieties. This meant that all kinds of crosses could
happen, even between very early and late types. In addition, it meant that he
crossed various sweet corns with all the corns he had originally planted. He
wasn't thinking about any of this at the time; it's just the way it happened.

One of the corns Alan planted was an Incan flour corn with huge, flat white
seeds and plants about 12 feet high. When Alan harvested his corn, the ears on
the Incan corn were especially beautiful, and there were one or two colored
seeds on each ear that represented pollination by a colored variety. There were
yellows, reds, purples, and blues; solid colors, stripes, blazes, and spots; clear
colors and iridescent ones. There were 100 to 150 colored kernels altogether.
Alan picked the colored kernels off the ears and saved them.

He planted about a hundred of the colored kernels the next year. When he
harvested the patch, the ears showed an occasional crinkled seed, representing
sweet types. The genes associated with sweet corn are recessive, so no crinkled
kernels appeared in the original crosses involving the flour-type Incan mother
plants. There were about forty crinkled kernels of all colors.

Alan kept and planted the crinkled kernels the following year. His harvest was
about 5 pounds of kernels, all sweet and of all colors. He selected for large, flat
kernels and planted a couple of ounces of seed. In subsequent years he continued
selecting for large flat kernels of all colors. He also selected for a plant height of
about 8 feet instead of the 10 to 12 feet typical of the original Incan corn. (That's
late enough to be marginal here. Eight-foot plants were enough earlier to be
dependable.) He also selected for ears that were lower on the plant - "So I could
reach them," he says. Lower ears also are larger, so selection for lower ears
automatically selects for bigger ears and higher yield.

`Rainbow Inca' sweet corn preserves the cytoplasm of the original Incan flour
corn and a large amount of genetic variability derived from many sources. The
kernels are of all colors and patterns, huge compared to any other sweet corn,
and broad and flat. The plants are about 8 feet tall. It's of late but reliable
maturity here (meaning it would probably be considered midseason in most
areas). It's undoubtedly been automatically selected for productivity in cool
weather. The flavor is excellent.

Alan didn't realize that he had developed something special until he offered
`Rainbow Inca' through the Seed Savers Exchange and heard the reactions of
those who grew it. It's excellent, unusual, unlike anything else, they said.
`Rainbow Inca' is currently offered through Peace Seeds and Seeds of Change
and remains one of the most unusual corns available.

 P TO this point I've been discussing the breeding of diploid plants,
plants that have two copies of each kind of chromosome, and therefore two
copies of each gene. Some important domesticated plants, however, are triploid,
tetraploid, or hexaploid, and some have even higher degrees of ploidy. That is,
they have three, four, six, or more copies of every chromosome. A tetraploid, for
example, has four entire haploid genomes instead of the two a diploid has. Any
plant with more than two complete haploid genomes is referred to as a polyploid.

Having higher than diploid levels of ploidy often causes a plant to have bigger
flowers and fruit, more intensely colored flowers, bigger or thicker leaves, and
bigger tubers. Potatoes and sweet potatoes are both

polyploids, as are many fruits and ornamental plants.

A number of rare vegetables, as well as many wild edibles, could probably be

turned into instant major vegetables if their edible parts could be made bigger.
One approach to that is through standard breeding. An alternative is through
spontaneous or deliberate chromosome doubling to create a polyploid.

Autopolyploids

In autopolyploids the haploid genome sets are all the same. The potato is an
autotetraploid; the sweet potato is an autohexaploid. In allopolyploids, the
haploid genomes are different. The standard commercial wheat is an
allohexaploid. It has three complete genomes originally derived from three
different species.

Autopolyploids presumably arise initially because cell division isn't perfect.

Once in a while during mitosis or meiosis, chromosomes fail to segregate. If this
occurs in a somatic cell, a tetraploid daughter cell results. This cell has four
chromosomes of each kind instead of two. If properly positioned, this cell could
divide and give rise to a tetraploid branch or section of plant.

Some species seem quite tolerant of changes in ploidy. In others any alteration
in ploidy seems to be severely detrimental or lethal. We don't know why.
Tetraploids of some species may have larger or thicker leaves and flowers, and
the leaves or flowers may be more deeply colored. They also may grow more
rapidly than diploids. (Higher levels of ploidy may or may not show such effects
depending on the species.) In some species autopolyploids may be completely
lethal. In others they may be obtainable, but they may grow more slowly than
diploids, be smaller, or flower later. For most species there seems to be a ploidy
level associated with the maximum size of flowers and other characteristics.
Higher or lower levels are associated with smaller size.

Polyploids are not always vegetatively stable. Chromosomes may be lost

during cell division so that the plant is a mosaic of tissues that differ genetically.
In some cases, a triploid or other-ploid plant experiences chromosome loss
accompanied by selection at the cellular level for the most vigorously growing
combination, and ends up with an apex in which diploidy has been reestablished.
When vegetative instability is occur

ring, you may see branches that are clearly different from neighboring branches,
or sectors - patches of the plant that differ from the rest in color or morphology
or flower characteristics or flower fertility. In these cases, two tubers or cuttings
from the same plant may give rise to plants that are genetically different from
each other and from the parent.

Many polyploids are vegetatively stable, however. Generally, the triploid,

tetraploid, or other-ploid cultivars we use are vegetatively stable. They either are
of species that are naturally stable when polyploid, or they have been allowed to
grow and divide and "sort themselves out" until they have become vegetatively
stable.

Autopolyploids behave differently than diploids when it comes to sexual

reproduction, because they have more than two copies of every gene instead of
two. All of Mendel's laws, predictions, and ratios depend on there being exactly
two copies of each chromosome and each gene. We usually can obtain the
classes we want, but the frequencies of various classes are unpredictable. In
addition there may also be sterility problems.

Triploids

Triploids can be generated when an accidental unreduced diploid gamete pairs

with a normal haploid one. They also may result from the cross of a diploid and
a tetraploid. They are usually sterile. The chromosomes attempt to pair in
meiosis, but they apparently can associate only two by two. All three
chromosomes often associate, since chromo somes begin their pairing at
numerous points along their lengths, so any given chromosome is often paired
with one homolog in one region and the other in another region. When division
occurs, there is no way for three chromosomes to sort themselves into two cells
without creating massive genetic imbalances. The result is often complete pollen
sterility and complete or nearly complete sterility as a female parent.

This doesn't mean that triploid plants are useless. It just means that they
usually cannot be maintained through seed propagation. Often sterility is
advantageous or even essential. The banana of commerce, for example, is a
triploid. Diploid bananas produce hard seeds. The edible banana is a banana with
aborted seeds; the same is true of the plantain. Many apples and pears are
triploid, as is the seedless watermelon. The latter is produced by crossing a
tetraploid variety with a diploid one.

Any perennial or plant that can be propagated from tubers, divisions, or root
cuttings or by other vegetative means can be maintained and used even if it is a
triploid. Triploid varieties of ornamentals may be especially useful. Flowers may
be bigger and richer in color. And since they don't set seed, the individual
flowers may last longer on the plant. The plant may have a prolonged blooming
period - one that might be achievable with a diploid only by pinching off the
flowers to prevent seed set.

Triploids and other plants that have odd numbers of haploid genomes can
normally reproduce only vegetatively. But plants with even numbers of haploid
genomes - tetra ploids, hexaploids, and octaploids - may be fertile.

Doubling Chromosome Numbers

 Usually, the doubling of the chromosome number to produce an autotetraploid
can be accomplished deliberately. It frequently occurs spontaneously, especially
in regenerating tissue. Tetraploid suckers on tomato plants are common, for
example. In some cases we can obtain polyploids simply by chopping off the top
of the plant and watching for lateral shoots that look different. Chromosome
numbers also can be doubled deliberately by using the poison colchicine.

Colchicine inhibits the formation of the spindle apparatus in cells by binding to

the protein subunits from which the spindle is made. The spindle apparatus is
part of the mechanical structure involved in moving chromosomes around during
cell division. If colchicine is applied at the right concentration for the right time,
some dividing cells will fail to have normal segregation of their chromosomes,
but then will go on with normal growth and division. Some such cells will be
tetraploid. The result, with somatic tissue, is a mosaic of diploid and tetraploid
tissue. With a little luck, some new shoots arise from some of the tetraploid
tissue and look different enough so you can identify and propagate from them.

Colchicine is usually applied to vegetative tissue as a cream. It must be applied

to growing areas such as an apical tip or the shoots of a tuber. When used with
seeds, seeds are usually presoaked in water containing colchicine before
planting.

Colchicine is a dangerous poison and is not readily available to amateurs. If

you want to use it, and you haven't formal scientific background or affiliations,
you may need to collaborate with a professional. Often, however, colchicine is
not necessary to obtain what we want. Spontaneous chromosome doublings are
common. If we have a triploid plant that is producing hundreds of sterile
flowers, for example, often one branch or one flower has seeds. That branch or
flower represents a spontaneous chromosome doubling, and the seeds are likely
to be viable.

Allopolyploids

Allopolyploidy involves different genomes. Wheat and many grains as well as

many brassicas are allopolyploids. Allopolyploidy arises naturally when a
spontaneous wide cross between two different species is followed by
spontaneous chromosome doubling. It is a common mechanism for evolution of
new plant species. We can create allopolyploids deliberately by first doing a
wide cross between two different species and then doubling the chromosome
number. After chromosome doubling, the plant is functionally diploid, since it
has two of each chromosome of each of the species from which it was derived.
Such allopolyploids are called amphidiploids.

In the next chapter I explore the deliberate creation of allopolyploid vegetables

further under the subject of wide crosses. To create an allopolyploid vegetable
deliberately, we start by doing a wide cross.

 HEN MENDEL explored the genetics of diploids, the crosses he
did were of varieties that were all close relatives. Often we want to make crosses
that are of distant relatives or even entirely different species. These are called
wide crosses. There are two kinds of wide crosses: crosses between plants that
are distant relatives but of the same species and crosses between plants that are
of entirely different species. Wide crosses between different species are rare in
animals but common in plants. They are, in fact, a major mechanism of plant
evolution.

Wide Crosses Within Species

When a wide cross is between two plants that are members of the same
species, obtaining the cross is usually easy, as is growing the F, plants. But when
the plants are distant relatives, there is often chromosome nonhomology - that is,
the two plants may have different arrangements of their genes into
chromosomes. Such differences arise because breakage and reunion of
chromosomes can occur, and chromosomes sometimes reunite with
arrangements different from those they started with. Such chromosomal
mutations can be preserved, just as single-gene mutations can, and are a part of
the evolution of genomes of plants.

When distant relatives are crossed, a block of genes that is on one chromosome
in one parent may be on a different one in the other. When the hybrid between
them reproduces sexually, there may be aberrations in chromosome pairing and
segregation. Some gametes can receive two copies of some genes and none of
others. Some chromosomes get torn in two as the cell divides, having two of the
region that attaches chromosomes to the division mechanism. Others lack that
region entirely and get lost.

The wide-cross hybrid may produce many inviable gametes. Many zygotes
also may die and lead to aborted seed or weak, aberrant plants. In short every
nonhomologous chromosomal region in the hybrid causes problems during
meiosis. In a hybrid between distant relatives, there can be dozens of
nonhomologies, each creating all these problems with respect to segregation.

In such hybrids, low viability of pollen, eggs, or seed can be a problem. In

addition, if we are following genes that are within or linked to a region of
nonhomology, we may be unable to obtain the recombinants we want. We may
even be unable to recover the genes that went into the cross because of linkage
to or association with areas of the chromosome that are involved in the
nonhomologies.

If the genes of interest to us aren't involved in the nonhomologies or linked to

them, we usually can get what we want. But often we see all kinds of "junk"
segregating out of the cross, there are problems with sterility, and normal
Mendelian ratios don't apply. If we can recover the class we want at all, though,
we can go ahead, even with the other problems. In future generations we tend to
eliminate the nonhomologies and start seeing Mendelian genetic behavior. Many
ordinary crosses have some degree of nonhomology and show "messy" ratios in
the first generations after the cross.

Sometimes the problems with low pollen viability or general sterility are so
great that it is difficult to do any breeding work with the hybrid. Often
backcrossing is a useful way to deal with such hybrids. If we make the hybrid
and then try to obtain an F2, we're dealing with a situation in which the
probability of obtaining both good eggs and good pollen is low, so the
probability of a good egg's being fertilized by good pollen is even lower. If the
probability of a good gamete of either kind were 1 in 1,000, for example, the
probability of a good zygote would be 1 in 1 million. We might never get one.

But if we make the hybrid and then backcross it to one of the parents, one of
the kinds of gametes is normal. In the preceding example, we would expect 1 in
1,000 good seeds instead of 1 in 1 million. Many plants produce several hundred
or even several thousand seeds, so the odds of 1 in 1,000 can be quite workable.
(With orchids, even odds of 1 in 1 million are workable, since orchids commonly
produce millions of seeds per flower.)

Pollen seems more sensitive than eggs to chromosomal aberrations. It is more

likely to work if we use good, normal pollen on questionable eggs than the
reverse. In backcrossing we're most likely to succeed if we use the hybrid as
female and fertilize it with pollen from one of the parents.

Wide Crosses Between Species

Many of the most interesting wide crosses are between entirely different
species. These have the potential for giving rise to entirely new species of plants.

When plants from different species are crossed, the chromosomes are usually
different enough that they don't pair at all. If a diploid plant with twelve
chromosomes is crossed with a diploid plant from another species with sixteen
chromosomes, for example, a gamete with a haploid set of six combines with a
gamete with a haploid set of eight. The resulting hybrid has fourteen
chromosomes, but none of the chromosomes in the haploid set of six is
homologous with any in the haploid set of eight.

The result is a plant that, practically speaking, has fourteen chromosomes but
is haploid. Such hybrids can usually grow vegetatively, but they can't reproduce
sexually. As previously mentioned, sexual reproduction involves diploidy.

If the chromosomes of the wide-cross hybrid are doubled, however, it will

have twenty-eight chromosomes in fourteen proper pairs. It will be a diploid, but
one with more chromosomes or kinds of chromosomes than either of its parents.
In some cases the plant remains sterile. But in many cases fertility is restored
completely. The result is an entirely new species.

Where a variety propagates vegetatively from tubers or divisions, the

chromosome doubling to recreate diploidy may not be necessary.

The results of wide crosses can be quite unpredictable. We don't know what
the phenotype is going to be when we combine such different haploid genomes
into one plant. In some cases there are partial chromosomal homologies instead
of a complete lack of homology, so segregation may occur in subsequent
generations, or it may not.

Somatic growth of hybrids may or may not be stable. If it isn't, selection for
more rapidly growing and stable material happens as the plant grows, so that the
genome from one part of the plant to another isn't the same. You may see
obvious sectoring. Sometimes after enough growth and sorting out, when
vegetative stability is restored, the plant may have the genome of just one of the
parents again. Or it may have mostly the genome of one parent with a new
chromosome or two from the other. Or it may have a genome that combines
some chromosomes from one parent and some from the other, but not the
complete genomes of both. In many cases, however, the hybrid does retain both
entire parental genomes.

In most cases there is some difficulty in making wide crosses between plants of
different species. This is virtually by definition. Different species are different
because they don't ordinarily cross under normal conditions. Sometimes the
difficulty is minor or technical. The plants normally grow on opposite sides of
the world, for example. Or they flower at different times. In some such cases
there may be no physical barriers to the cross when we hand-pollinate. In many
cases, however, there are major incompatibility barriers that must be overcome
in order to cross two different species. Various methods for overcoming
incompatibility barriers in ordinary as well as wide crosses are given in the last
half of Appendix B.

In general it is common to be able to obtain crosses of different species within

the same genus, and occasionally it is possible to obtain crosses between plants
in different genera in the same family. Sometimes a cross between two species
can be made to work with a given individual or variety but not with others. It is
very common for wide crosses to be possible in one direction only, so it is
especially important to try crosses in both directions. Basically, we just have to
try things.

Don't be discouraged from trying a wide cross because the experts say it's not
possible. They are using different plants of different varieties and a different set
of special tricks for overcoming incompatibility barriers, and they have different
weather conditions. Failing to obtain a wide cross isn't significant; obtaining it is.
And you need to obtain it only once.

I think the use of wide crosses, followed by chromosome doubling where

necessary, has tremendous unexplored potential for creating new vegetables. The
unpredictability of the results discourages much interest in the approach by
professionals, but I think the very unpredictability makes it all the more fun for
the amateur.

With much of standard breeding, we want a particular thing and we go about

figuring out how to get it. But with wide crosses, I think the most productive
approach is to do every cross that we think might be interesting. In cases where
we are successful, we can then spend some serious time figuring out what the
new plant is good for.

 HERE ARE two basic approaches to plant breeding. First, there is the
deliberate approach. We decide to try to get a specific something that is bigger,
better, tastier, or more beautiful, and we design and conduct a project aimed at
achieving our predetermined goal. But there is another approach. Something
interesting turns up in our garden. We didn't plan it or ask for it, but there it is.
Now that we have it, we decide we would like more of it.

Many important agricultural advances undoubtedly trace their origin to some

genetic accident that happened in the garden of someone who was observant and
curious. Knowing how to take advantage of the accidents can add a lot to our
plant breeding repertoire. We can't plan to have an accident, but we can be ready
for it when it occurs.

Genetic accidents are commonplace. I had five last year alone. Two were in
my farm field vegetable patch that was about a third of an acre in extent. The
other three were in my small home garden that is only about 200 square feet in
size. The farm field probably had dozens of interesting accidents in it, but I
wasn't there looking at every individual plant regularly. What's critical about
genetic accidents isn't so much their occurrence as their being observed. A
much-loved, muchwatched home garden is especially fertile ground for the
discovery and exploitation of happy accidents.

Three major classes of genetic accidents can provide interesting opportunities:

somatic mutations, meiotic mutations, and accidental crosses. My accidents of
last year were one obvious somatic mutation, one probable meiotic mutation,
one accidental cross, and two events of less obvious origin. I discussed one of
these - the appearance of a hairy, pest-proof, nonbolting mustard - in Chapter 5.
In this chapter I illustrate how you can use happy accidents when they turn up in
your garden.

The Seed Savers Exchange Winter Yearbook provides space where members can
ask questions and invite interaction with others who are interested in similar
things or have relevant information. In the 1991 edition, member Davis
Huckabee asked about something that had turned up in his garden: "Do tomatoes
produce bud sports as some tree fruits do? Last year one branch of a `Porter'
tomato produced fruit much larger...."

The answer is yes. Tomato plants, like fruit trees and every other living thing
on the planet, do experience mutation. If the mutation occurs in a somatic cell
that gives rise to a branch, one entire branch will be genetically different from
the rest of the plant.

Sport is one of those old, traditional words that dates back to the time when
people didn't have any concept of genes or mutations. When one plant in the
field was obviously different from its fellows, it was called a sport. When one
branch on a tree was clearly different from the others, it was called a bud sport.

Somatic mutation has been especially important in developing new varieties of

perennial plants such as fruit trees. Many traditional varieties were discovered
because someone noticed a branch whose fruit was different from the other fruit
on the tree. The person then made cuttings for rooting or grafting and established
new trees with the new genotype. Vegetables also can have somatic mutations
that, if properly positioned, can give rise to branches or parts of the plant that are
genetically different from the rest of the plant.

Davis should be able to exploit his tomato sport quite easily. Since tomatoes
are automatic inbreeders, saving the seed actually amounts to self-pollinating the
flowers on the mutant branch - that is, inbreeding from them. That's an excellent
approach. It's usually the best approach unless there is some reason to do
otherwise.

When it comes to the next step, the important thing is to realize that the new
tomato mutation is both heterozygous and dominant to its corresponding allele.
Here's why: Tomatoes, being diploid, have two copies of every gene. It isn't
likely that both copies will simultaneously mutate in the same plant. Instead, one
gene presumably mutates, and its allele is still of the original form. That is, the
mutant branch is heterozygous for the mutant allele. If the mutant allele were
recessive, the branch would look the same as all the rest (by definition). But it
doesn't. So the mutant allele must be dominant to (or codominant with) the
original allele. In other words, if the genotype at some locus in the original
variety was, say, bb, the unusual branch has a genotype of Bb.

Once we realize that the mutant tomato branch is heterozygous for the new
mutation, we also realize that the seed from tomatoes on the mutant branch is
going to show segregation for the new mutation. That is, the seed will not be
pure breeding for the new mutation initially. So we grow a number of plants
from the seed, because we know that only some of them are going to be of the
mutant type. The expected genotypes in the next generation will be (1/4BB +
Y2Bb + 114bb). The genotypes Bb and bb correspond to largefruited and
normal-fruited phenotypes. Genotype BB is new; we can't predict what it will
look like.

If there is simple dominance, BB will look the same as Bb. It might instead
have even larger fruit than Bb. Or homozygosity for B might have detrimental
effects. It's possible that BB plants will be lethal or that the fruit will be
deformed and undesirable.

The most probable result is that BB and Bb will look the same, that there is
simple dominance. In this case, we can derive a pure breeding variety with large
fruits quite easily. We just inbreed and save seed from largefruited types for a
number of generations, as I've described previously.

Inbreeding and selecting large plants for a number of generations doesn't cost
any work at all. Instead of growing just `Porter', we grow some plants that are
'Porter' and more that are a larger `Porter'. In addition, all we need is a few plants
in each generation - no more than we would have been growing anyway.

There is a faster way to get the new pure breeding variety. It depends on
deliberately identifying one or more plants that are homozygous for the
mutation instead of on the general nature of inbreeding to generate
homozygosity.

The first year we grow several (more than eight) plants from seed of the
tomatoes on the mutant branch. Then we save seed from at least eight of the
largefruited plants. The

second year we grow at least eleven plants from the seed of each of those eight,
keeping track of which plants had which mothers. Most of the families have both
largefruited and normal-fruited types; their mothers were heterozygous for the
mutation. But some families are all largefruited; their mothers were
homozygous. If we choose one of the plants from one of the latter families, we
will have, as of year 2, a new pure breeding variety.

Why did we choose the numbers eight and eleven in the previous paragraph?
Because we expect offspring of the mutant branch to be three-quarters
largefruited and one-quarter normal, and we realize that if we select largefruited
offspring, 2/3 of the largefruited offspring should be heterozygous (Bb) and 1/3
homozygous (BB). Then we refer to Appendix H to figure out how many large
plants we need to be at least 95 percent sure of having one homozygote. If
something has a probability of 1/3 we need eight plants to be 95 percent sure of
getting one. Next we need to figure out which plant(s) is (are) homozygous.

We identify homozygous plants by growing a small batch of offspring from the

seed of each. The ones capable of segregating out the recessive gene clearly
aren't homozygous. We have to grow enough offspring of each to see the
homozygous segregant if it is possible, but we don't want to grow more plants
than we need to for that purpose. How many offspring do we need to grow to be
95 percent sure of finding at least one homozygote segregating from a
heterozygote if that is what's going on? We look at our table again. If a mother
plant is heterozygous, the proba bility of a homozygous offspring is 1/4. How
many plants do we need to be 95 percent sure of getting at least one of
something that has a probability of 1/4? The table tells us eleven.

So if we are happy with being 95 percent sure about things and we want to do
the minimum possible work to generate a pure breeding variety in the least
possible time, we grow eight plants from seed of tomatoes from the mutant
branch, and we screen them by growing eleven plants from the seed of each.
That's eighty-eight plants total in the second year. But then we're through; we
have a pure breeding variety.

The family-line breeding program gives you a pure breeding variety faster than
just saving seed from a single largefruited plant or two each generation. But it is
more work.

Davis Huckabee saved seed from his mutant tomato branch and planted it.
Most of the plants had large tomatoes, but some had the standard `Porter' size.
He is now cheerfully growing and inbreeding his new larger `Porter' to get a pure
breeding variety.

You can often take cuttings from vegetables. It isn't normally done with annuals,
because it is usually easier to start them from seed. My friend Miriam Ebert
starts late plantings of tomatoes by simply cutting suckers off her early plants
and sticking them in the soil. Brassicas can be propagated by shoot or root
cuttings (see Appendix A). Little Brussels sprouts heads can be rooted and will
flower and set seed as if they were baby cabbages.

If you have a small mutant branch of something, you may be able to increase it
by such vegetative means. Or you might cut off the entire top of the plant above
the mutant

branch so that it becomes the main leader. You can also increase material by
grafting, or use grafting in emergencies.

My friend Dan Borman went out into his garden one day and found that a very
special sunflower seedling had been nipped off by a slug. The top of the seedling
was lying on the ground. He quickly sliced the base of the stem into a wedge.
Then he cut the top off a seedling of a common variety and made a Vshaped slit
into the region between the two seedling leaves. He slid the special seedling's top
into the V. The graft took. The special plant was rescued. Neither he nor I have
seen or read anything about either grafting sunflowers or grafting seedlings.
What matters is that he tried it.

Not every different-looking branch represents a new mutation. Sometimes pests

or diseases are responsible. In corn, for example, some fungal diseases produce
hormones that alter the sexual expression of the plant so that there may be
kernels on the tassels. In such cases the funny branch or tassel may not be
genetically different at all.

My experience with somatic mutation was even simpler and more

straightforward than the story of the `Porter' tomato, because mine occurred in a
vegetatively propagated perennial - the `Egyptian' onion, also called the
`Walking' onion - a top-setting form of Allium cepa. `Egyptian' onions make
bulbs at the top of their flower scapes in the position where flowers would
normally be. That is, they use part of the mechanism of sexual reproduction for
vegetative reproduction and don't reproduce sexually at all.

The scapes bend over and plant the bulbs - thus the reference to "walking." Or
you can plant the bulbs yourself, using them as sets, which is what I do. Some
people use the top sets as little onions, but I have always found them too small to
be worthwhile preparing that way. `Egyptian' onions also divide at the base to
form bunches, though not as prolifically as true bunching onions. I have about a
dozen clumps.

Last year one section of one of my clumps produced scapes with two to four
much larger top sets per scape instead of a lot of little ones. I think I might prefer
having a few large top sets, as the bigger top sets make better planting stock.
And these are so big that I might be able to plant them very deep and have an
onion with a long, blanched stalk. That's possible with big top sets but not small
ones.

Normally, to get blanched green-onion stalks, people grow seedlings, then

transplant them into trenches and hill them up as they grow. But that's too much
work for me. I've found that the biggest top sets of `Egyptian' onions can be
planted several inches deep and will come up and make a nice onion with a
blanched stem. But only the biggest top sets have enough stored energy to do
that, and my standard variety produces few big top sets. In addition, bigger top
sets might be more useful to me in the kitchen.

I planted two top sets from the large-topset-producing scapes in a second

patch. One produced a plant with unusually large, thick scapes and large top sets
- some up to an inch across. The leaves may or may not be bigger or thicker than
usual; I can't tell. The

scapes are so thick and sturdy that they have remained erect, and it is now time
to harvest the onions. In other words, the mutant walking onion doesn't walk
anymore. If that's consistent, it could be handy.

I would much rather have scapes that stay upright until it's time to harvest the
bulbs than ones that bend over and break and dump the bulbs in the mud before I
get around to harvesting them. I would rather harvest the bulbs and use them or
plant them where I want them than have them trying to plant themselves
randomly in an already crowded garden. The plant produced twelve large top
sets on four scapes. The largest was 11/4 inches across. Many of the top sets also
developed scapes with batches of smaller top sets on them; there are thirty of
these smaller, secondary top sets.

The plant that developed from the other large top set is a completely ordinary-
looking plant with normal-size bulbs and scapes that bend over and dump the
bulbs in the mud. Apparently, not all the tissue in the scape with the large bulbs
was mutant. The original plant I noticed must have been a mosaic.

There are two obvious possible explanations for the new mutation. One is that
it is a single-gene dominant mutation of some sort, as in the tomato example
described previously. The other is that the mutation is a chromosomal one, most
likely a spontaneous chromosome doubling, so that the mutant plant is
tetraploid.

From a practical point of view, it doesn't make any difference whether the new
mutation is of a single gene or a change in ploidy, since I'll be increasing and
maintaining the material vegetatively. I'll just plant all the bulbs from the giant
plant and keep any that also produce giants. All the scapes on the giant plant
produced big bulbs, so I think this plant may be of purely mutant tissue, or very
nearly so. If it is, all the bulbs should produce giant plants. If not, some of them
may produce normal plants. I'll cull out any normals, of course.

As it stands, it looks as if I already have a new vegetable variety - and for no

more work than just noticing something and picking two bulbs and shoving them
into the ground. I wasn't especially thinking about how useful big top sets would
be before the mutation happened. I wasn't thinking much about the `Egyptian'
onions at all. I was only eating them. But once the mutant appeared, I began
thinking about its potential. With happy accidents, the question usually isn't
"How can I get thus and so?" Instead it is "Here is something; what might it be
good for?"

In fall of 1990 I planted about 800 seeds of `Aprovecho Select' fava bean in a
field in Independence. It's a large-seeded type with spectacular flavor as a green-
shelled bean. It overwinters here in an ordinary year. But the winter of 1990-91
was unusually severe. All the fava bean plants died - except one. That one wasn't
even damaged.
 To me the situation represented not a setback, but a wondrous opportunity.
There are no large-seeded favas that are fully as cold hardy as I would like.
Perhaps the surviving plant could be used to breed one.

My tactic would be to inbreed the plant. But favas can outcross to a significant
extent, so I transplanted the plant to my home garden, where I could watch and
pamper it, and where it would be isolated from other fava beans.

By mid July, my special plant had produced nice large pods with big seeds. All
the seeds had a pale, opalescent green skin and white hilums. `Aprovecho Select'
has tan seeds with black hilums. I don't know of any large-seeded fava with
white hilums. And I had never seen that skin color before, either.

I showed the seed to lanto Evans, who was the original source of the
`Aprovecho Select' seed I had planted. He said the green skin is the same color
as that of certain Japanese varieties they have grown in past years - back before
they realized how much isolation the favas require. The Japanese varieties don't
have white hilums, however. A few of the `Aprovecho Select' seeds I planted had
white hilums, though. I think it likely that my special plant is a descendent of an
accidental cross between a white-hilum `Aprovecho Select' and a green-skinned
Japanese variety.

I planted thirty seeds from the special plant in fall of 1991 and increased the
seed stock to a few hundred. I've now passed most of the material over to Dan
Borman. His ordinary winters are similar to my most severe ones, and he is
passionate about fava beans. He'll do further seed increase and select for winter
hardiness under his conditions.

By now you should see an obvious pattern to how to approach most happy
accidents. Start by inbreeding, unless there is some reason to do otherwise - such
as using a species with an incompatibility system or strong inbreeding
depression. If the plant is a natural inbreeder, just save the seed. If it is enough of
an outbreeder that isolation is usually required to maintain pure varieties, either
hand-pollinate or isolate.

My happy accidents all involved plants that were inbreeders or that are
vegetatively propagated. Happy accidents in outbreeders that don't suffer greatly
from inbreeding depression are usually handled similarly. However, if the plant
has an incompatibility system, you may not be able to self-pollinate it.
Furthermore, if it is of a species that displays inbreeding depression, you cannot
usually derive a vigorous variety on the genetic base of just a single plant. In
these cases, you may need to cross the unusual plant to one or more others (or
allow it to cross) to produce a population of plants from which you can obtain
more of the unusual type in future generations.

If you find a single unusual corn plant, for example, you would realize that you
cannot develop a vigorous variety through inbreeding from the one plant.
Instead, you would use a careful combination of inbreeding and outbreeding. For
example, you might cross the plant with a hundred ordinary plants, and save the
seed. Then you plant that seed in a patch by itself. None of the plants might
resemble the unusual one. But you wouldn't be discouraged. You would allow
them to cross-pollinate each other. In the next generation, you will probably have
many of the unusual plants, enough to form a solid genetic base for a vigorous
variety.

When happy accidents appear in inbreeders, you usually develop their

potential by inbreeding. When happy accidents appear in outbreeders, you often
must do some outbreeding to generate an adequate population base. Then you
inbreed and mass-select within the population.

 VERY CULTIVATED plant was once a wild plant somewhere. Every
edible wild plant represents a possible new vegetable. Every niche in your yard
that doesn't have edible things in it represents an area where an edible wild plant
could be growing if you could figure out which one and find it. Our modern style
of living in houses, for example, creates patches of partial or solid shade. So do
the trees we like to live among. Most people try to grow sunloving grass under
their trees and north of their buildings. But the woods are full of edible plants,
often perennials, that thrive on partial to full shade.

Generally, it seems to have been the sunloving plants that were brought into
cultivation. Now, with so many people living in homes with only small and
usually partially shaded yards, might be a good time to look for edible plants to
fill the shady niches.

There are said to be twenty thousand edible plant species. Of these, only a few
hundred have received any breeding work from any member of humanity. Did
our ancestors tally every edible species and choose exactly those with the most
potential? Did they find and domesticate all the best possibilities? I'm sure that
they did not. First, only some peoples were involved in domesticating plants;
vast areas of the world were populated by people who weren't agriculturalists.
None of the native edible plants of the northwestern United States were
domesticated, for example, because the natives were migratory hunter-gatherers,
not farmers. Our region is full of delicious edible wild plants that are already
perfectly adapted to growing here.

Elsewhere in North America native agriculturalists concentrated on growing

just a few staples. They collected most vegetables from the wild instead of
domesticating them. All native North American edible wild vegetables represent
essentially unexplored terrain for domestication.

Even if our ancestors had evaluated all possible wild edibles for their potential
as domesticates, it wouldn't mean much. Even today no one can evaluate the
potential of a plant before the breeding work with any degree of certainty. Even
in areas of the world long occupied by domesticators and agriculturalists, there
are undoubtedly many more wild plants that would make good domesticates.

Finally, what makes a good potential domesticated vegetable depends not just
on the plant but on the cultural patterns of the potential users. There are new life-
styles; new ways of handling homes, gardens, and land; new knowledge about
and concern over nutrition; new food preparation methods. All create possible
niches for new vegetables.

What about vegetables bred for microwaving, for example? Microwavable

popbeans? Corn on the cob that tastes good when microwaved as whole ears
without opening them up and removing the silks first? I recently read that waxy,
moist potatoes such as `Yellow Finn' are best for microwaving; that's the first
reference I've seen to specific varieties and microwaves.

We have very efficient food dehydrators now, too. Some edible wild plants are
not usually used because they have toxic substances that dissipate during drying.
Such plants may now have potential for the casual gardener.

Until recently, gardens in America were planted in the spring and harvested in
the fall, and then tilled. Perennial vegetables just didn't fit into that pattern. Now
more and more people are using permanent beds for part of their vegetable
gardening. Perennials fit into a permanent bed style of gardening extraordinarily
well. Every alteration in our cultural patterns creates opportunities for creating
new vegetables - new varieties of established vegetables as well as entirely new
crop species.

Professional plant breeders are doing little work on domesticating new crop
species. Professionals need to get grants for their work, and it's easier to get
money for work on something that is already important. When Rich Hannan
received phone calls from professional plant breeders who were interested in
Phaseolus vulgaris popbeans (nunas) and he told them about Cicer arietinum
popbeans, they responded, "But what I have funding for is Phaseolus." I had
funding for neither; I was free to be open-minded.

Very few professional plant breeders are working on domesticating new

vegetable species, and those that are do so as only a small part of their efforts. I
would guess that more than 99 percent of the professional vegetable breeding
work is devoted to fewer than one hundred species.

Corn has been profoundly improved by modern professionals, but it was

created by amateurs. The same is true of almost every food crop we have. The
work of the modern professional plant breeder is directed almost exclusively at
improving the crops that were created by the amateurs of yesterday. Who will
create tomorrow's new crops and new crop species? Who else but today's
amateurs.

How long does it take to domesticate a plant? It depends on what you mean by a
domesticated plant. That's something that geneticists and anthropologists argue
over. There is a tendency for them to claim that domesticating any plant takes
hundreds of years but also to define domestication so that something isn't
considered domesticated unless it has been cultivated for hundreds of years. That
argument is somewhat circular. It's clear, though, that in some cases a single
genetic mutation is enough to turn a wild plant into a much more useful
domesticate. Such mutations often turn up in gardens by accident and are often
present in wild populations. We can find them just by looking and collecting.

For my own purposes I think of domestication in the broadest and loosest way.
Anytime we collect a wild, naturalized, or escaped plant and deliberately grow it
near our domicile, we are engaged in domestication. Sometimes we have
changed the plant genetically and sometimes not. Note, though, that when we
choose a particular plant or population of plants in the wild to start our own
population, we have done breeding in the sense that our population differs
genetically from the average for the wild population.

The best place to start in domesticating new vegetables is with the edible wild
vegetables that are native to or have naturalized to your region. You can consult
books on wild edibles or try to find out what plants the native people of your
region used and how they used them. Then you go collecting, or refer to others
who do or have. Note that many of the alternative and regional seed companies
carry seed of native wild plants. In my own region, Abundant Life Seed
Foundation, Meristem Plants, Peace Seeds, and Seeds Blum all carry seed of
native edible species.
 My own preference is for plants that are more agroecologically adapted to my
conditions. I already have plenty of vegetables, inherited from others elsewhere,
that will grow here when watered out of season and pampered. I want new
vegetables that produce tasty, different food, and for less work.

I also want vegetables with new, delicious flavors, not ones that are merely
palatable. And I am not willing to do a lot of work to get a small edible part. Or
to spend hours preparing anything. Or to boil anything in two changes of water
and throw away most of the vitamins and minerals. Or to dry something, pound
it into flour, and mix a small amount of that flour with wheat flour to make bread
that would have tasted just as good without the addition. You will have your own
preferences.

When you go collecting edible wild plants, keep the conditions in your yard or
garden firmly in mind. Don't bother collecting from marshes unless you have a
marsh. Nearly everyone has some areas of partial as well as deep shade where
woodland plants might thrive. And most gardeners have at least some areas of
full sun or nearly full sun that could be home to prairie or meadow wildlings.

The characteristics of the plant in the field usually aren't a very good indication
of what it will be like under garden conditions, with richer soil and optimal
spacing. It's a good idea to collect from a number of different areas whether the
plants seem particularly impressive or not. They'll almost always be much
bigger, more succulent, and more productive once they get their roots into good
garden soil.

Individual wild plants and populations may vary a lot in ways that matter to
you. They can be milder or stronger in flavor, be bigger or smaller, have leaves
of various shapes, and so on. You'll probably start by collecting something
haphazardly and growing it. If it's a winner in your garden, you will often go
back to the wild and make additional collections looking for specific
characteristics.

Whenever you collect, you will have to decide whether to lift plants from the
wild or collect seed. If the plant is rare, or is rare in the area where you found it,
you shouldn't collect whole plants. You should collect seeds. (Some areas even
have laws against digging up wild plants.) An alternative might be to dig down
and remove a section of the root for a root cutting or to remove a shoot or two to
root. If the plant has a number of crowns, you might be able to remove a small
piece for yourself and leave the bulk of the plant where it is. The object is to
obtain some germplasm while leaving the wild population unharmed.

Seed represents greater genetic diversity than a single plant or clones from
one, and it can usually be transported and stored more easily. However, seed of
wild plants frequently does have one or more special dormancy mechanisms, so
you may have some work to do before you can figure out how to get your wild
plant's seed to germinate. (The J. L. Hudson catalog is one of the best sources of
information on how to germinate seed from rare or wild plant material.)

Some seed has a tough seed coat that must be penetrated in some fashion
(nicked with a knife or file, for example, or abraded a bit with sandpaper) before
it will germinate. Other seed needs a cold treatment to mimic winter. Seed that
ripens in late summer or fall and germinates in spring is often most easily
handled by planting in fall and waiting for spring.

Generally, edible wild vegetables are usable as soon as you get the plants
established, which is usually a few months for most annuals, two or three years
for perennials started from seed, and one or two years for perennials started from
plants or cuttings.

Dan Borman (Meristem Plants) collected plants of Lomatium nudicaule, an

edible in the family Apiaceae, from the wild and moved the plants to his garden.
The greens have become one of his major vegetables and staples.

Alan Kapuler (Peace Seeds) is in the process of building a collection of all the
wild Apiaceae of the Northwest. Many are edible; many were used by the
Indians. Some are potherbs. Some produce large, succulent roots. Alan also has
two different species of camas (Camassia leichtlinii and Camassia quamash).
Camas was one of the Indians' staples. The plant was so important that
harvesting rights to various camas fields were a major issue.

A number of professional plant explorers and breeders in Louisiana have been

working extensively with the American groundnut (Apios americana), which
was collected wild and used extensively by the Indians. (See the article by
Blackmon and Reynolds listed in the bibliography and Janick and Simon's
Advances in New Crops.) They have made extensive collections from the wild
and are evaluating them. Among the characteristics they are selecting for are
fused (bigger) tubers, plants that have tubers closely spaced underneath them
instead of dispersed thinly over several square feet of ground, and plants that
don't twine and are able to grow on the ground instead of needing support.

My major attempt at domestication so far involves the dandelion (Taraxacum

officinale). Dandelion leaves can make a delicious dish of greens in the early
spring, and the plants are perennial.

There are plenty of dandelions growing in fields around here. But the ones
here in town are in lots frequented by dogs or in yards where people mow them
down or spray them with herbicides. I can walk a few blocks to a park area and
collect dandies, but that isn't very convenient. When I want to eat something, I
want to be able to go to the garden and have it there. I need my dandelions
handier.

I started by ordering seed of the variety `French Broadleaf' and planting it. It
grew slowly. That's not surprising for a young perennial. The plants looked and
tasted exactly like the wild dandelions that were growing all over the place,
though. There was nothing especially broad about the leaves. In the summer the
plants were covered with powdery mildew and stopped growing entirely. They
were still quite small; now they were both small and miserable looking. The wild
dandelions growing in the lawn a few feet away looked happy and vigorous.

I was going about things backward, I decided. Why should I be trying to grow
an imported dandelion when my region was full of naturalized dandelions -
dandelions that have been selected over hundreds of generations to grow here
without any special pampering? If the import tasted better or was more
productive than the natives, that might be reason enough. But that didn't seem to
be the case.

So I forgot about other people's dandelions and started paying attention to the
native ones. They come in various sizes and shapes. Some are bitter even early
in the spring, and some are mild flavored until quite late. Every time I went
anywhere and saw a particularly large, luxurious dandelion in flower, I took
away some of its seed. Dandelions along roadways, dandelions in clear-cuts,
dande lions in the gardens of my friends - all were scrutinized.
 One winter day when I was out walking my dog, I noticed a dandy that had
produced large leaves in late winter before most plants did much. I carefully
marked the plant. Later in the spring, after the dandelions began flowering, I
went back and collected seed from the early one. Then, as I came back to the
house with my thoughts still on dandelions, I happened to notice an especially
large dandelion with very big leaves that was growing right in my driveway. I
tasted it. It was mild - as mild or milder than any dandy I had ever tasted. And
the leaves were huge. I had walked right by it on the way out to collect
dandelions elsewhere. I named it `Driveway Surprise'.

`Driveway Surprise' was in danger of getting mowed down. So I lifted it,

separated it into a major plant and a number of root cuttings, and moved it to my
garden.

But the clones of `Driveway Surprise' produced plants with small, dissected
leaves instead of big, broad leaves. And they weren't especially mild. They
tasted pretty much like the rest of the dandies in my yard. In the summer they
got powdery mildew, too, though not nearly as badly as the imported dandelion
had. They never seemed to quit growing completely or to look quite as miserable
as the import. It was time, however, to rethink my approach.

Maybe little plants produced small, dissected leaves. Many wild plants seemed
to have both kinds of leaves, dissected ones and broad ones. Or maybe it had to
do with the sun. As I thought about it, it seemed to me that the tall, luxurious,
broad-leafed dande lions also were the mildest ones at any given time of year,
and they were always growing in fairly heavy shade.

I was still going about things wrong, I decided. When I transplanted `Driveway
Surprise' to my garden, I had arranged for it to have adequate sun. Maybe that
was counterproductive. It had been growing in a gravel driveway area north of
the garage and under a tree. It must have had very close to full shade. The other
thing I did when I transplanted the clones to my garden was that I put them in a
year-round water situation. Wild dandies around here normally don't get any rain
in the summer. That section of driveway was never watered except by natural
rain, and the plant was big enough that it must have survived and thrived through
at least a couple of years.

I left the dandy clones where they were in my garden, but I planted other
plants south of them. This spring the dandelions were heavily shaded, and I got
my first taste of large, mild, tender `Driveway Surprise' dandelion greens.
Interestingly enough, the flowers also are quite large - up to about 2 inches
across and more a solid ball than flat - and the large, unopened flower buds make
a nice vegetable. The flowers on the original `Driveway Surprise' as it was
growing in the driveway were nothing special. But the plant was growing in
gravel and hardpacked clay, not good garden soil.

My `Driveway Surprise' leaves stayed mild far into the summer, long after the
wild dandelions - even those growing in full shade - had turned bitter.
Throughout the spring and early summer, I could boil the leaves in water, drain
them, and serve them with salt, pepper, and vinegar, just like any delicious
green. (I didn't have to boil them in two changes of water, as is often done with
wild dandelion greens.) I had only enough leaves to experiment with, not enough
to satisfy. During midsummer, the leaves became inedible because they got
powdery mildew, and ultimately they became bitter. The plants continued
growing, however. I'll remove the old growth in the fall and let the plants start
over making fresh, clean leaves.

I think I'm still doing things wrong, though. When I can get to it, I'll build a
patch north of the garage in practically full shade for `Driveway Surprise'. I
won't water the patch at all. I'll subdivide and further clone the plants I have in
late fall after the rains start. They will have all the rain they need to get
established. They probably won't give me anything to eat the first spring, but the
following year, in late winter and early spring, I should have plenty of greens. I
should never have to water them. The plants will, presumably, go dormant in late
summer when there is no water available. And with no watering, they probably
won't get powdery mildew. Once established, the dandy patch should produce
good spring greens year after year with no work at all. And the huge golden
flowers will be attractive as well.

I don't yet know whether `Driveway Surprise' is actually milder than the wild
dandies around here. Part of the mildness seems to come from growing the plant
in the shade. Another part of it, as with many plants, is keeping the leaves picked
so that I am always harvesting young, fresh leaves. Leaves that have been
around for a while are always bitter. Growing in good garden soil, the plant
grows leaves much faster than the dandies in my lawn, so the fact that the leaves
are milder doesn't necessarily mean there is a genetic difference.
 It's also possible that other wild dandies would have flowers as big as
`Driveway Surprise' if they were equally pampered. I haven't done any side-by-
side comparisons of `Driveway Surprise' and other wild types. If I were going to
offer it for sale, I would need to do that so I would know what special features
my plant had. But I'm not. I need only to grow and eat dandelions. I don't need to
sell them.

I also haven't grown out seed from 'Driveway Surprise'. I've just cloned it. My
first priority is to establish a big enough patch of `Driveway Surprise' so that I
will have all the dandelion greens I want in the spring.

What about the seed of my dandy that might float all over and irritate my
neighbors? Well, I usually eat the buds before they open. When I miss them, I
leave the beautiful flowers but pick the seed heads before they open. I don't want
a garden full of dandelions any more than my neighbors want them in their
yards. And I want my dandies producing more leaves and buds, not seed.

When I first started playing with wild-collected dandelions, I treated them

exactly as if they were a typical domesticated crop. Our normal garden crops are
imported from elsewhere, and we usually grow them by altering our conditions
to make them as much as possible like those where the vegetables originated. I
did not realize the agroecological implications of working with a plant that was
already perfectly adapted to growing in my area. I now view the agroecological
implications as one of the most important reasons for collecting and cultivating
wild edibles.

Dandelions growing wild in my region have been selected to be able to survive

summer drought and to do their growing in the wet winter and spring. That
means that they could make a nice permanent bed of edible and ornamental that
would require no water and no care. Dandelions that grow wild in your area will
probably have different characteristics. If you live in an area with severe winters,
for example, your wild dandelions might make especially big roots and then die
back and go dormant in winter. Such a dandelion might put on a heavy burst of
leaves in early spring. Or it might have potential as a forcing crop - a vegetable
that produces a root that is harvested and then put in a warm place to develop a
crop of greens during the winter.
 My dandelion might not be able to go dormant in winter and may not even
survive in a harsher winter. Then again, the ability to go dormant whenever there
is a reason to may be one of the basic characteristics of all dandelions. This
seems to be the case with sea kale (Crambe maritima). It can go dormant in
either winter or, because you forgot to water it, summer. It doesn't go dormant if
you water it in summer. Allium vineale, a local naturalized wild onion or garlic,
represents the other extreme. It goes dormant and dies back in summer whether
it is watered or not.

Working with native edibles broadens your perspective on gardening. It invites

you to reexamine all your gardening methods and how they fit into your life-
style, your cultural

patterns, your cooking and eating habits, and the ecology of your region. It helps
you identify the underlying assumptions and imagine other possibilities.

We need native vegetables. But we need more than just the plants. We also
need entirely different ways of growing and using them, ways that capitalize on
their unique adaptation to our climate and their special properties.

When it comes to growing and using an unfamiliar plant, some precautions are
advisable. First, make sure you've properly identified the plant. (I frequently
consult the weed specialists at my local agricultural extension service; to them,
this stuff is "weeds.") Second, read what is available about how your plant is
processed or used as an edible. Some plants have edible parts and poisonous
parts, for example. Some are edible only at certain times of the year or only with
special processing.

Don't believe everything you see in print, though. Many authors copy from
each other's writing without ever trying the plants themselves. Some of the
plants the books call edible taste awful when prepared by the methods they
describe. Don't believe anything until you've tried it yourself. And try things
cautiously and carefully.

Realize that you may be unable to eat a specific plant even when others can.
There are individual differences in ability to digest specific compounds; human
biological variability frequently shows up when you experiment with an
uncommon food plant. I know at least three people who suffer from severe
digestive upset and abdominal pain if they try to eat sunroot (Jerusalem
artichoke, Helianthus tuberosum). A good guess is that they can't digest inulin,
the unusual starch characteristic of sunroots.

You also may be allergic to specific plants, even if they are a fine food for
others. I am allergic to all plant pollen. Many people can use the flowers and
buds of a number of species as edibles, but I have an instant allergic reaction to
them. If I remove the stamens of daylilies before they shed pollen, I can use the
flowers as a vegetable; otherwise I cannot.

Whenever you are trying a new food plant for the first time, whether it's a wild
plant or not, always start with a single bite and then wait a day. The next time try
a small portion and again wait a day. On the third try you can have a full serving.
Once you've established that the plant is a good food for you, don't assume that it
is a good food for a guest; he or she might be unable to digest it or be allergic to
it.

The other major precaution with respect to experimenting with unusual or wild
plants involves their potential for becoming pests. Is your plant invasive? Will it
take over your entire yard and that of your neighbors? Will it shed floating seeds
that spread all over? Neighbors tend to prefer that your plants stay in your yard.
That nice edibleshoot bamboo can send out 40-foot underground runners and
turn your entire neighborhood into one huge bamboo stand unless you confine it
rigorously.

Some counties have laws against growing certain plants or allowing them to
grow on your land. In some counties, for example, it's

illegal to grow dandelions. Some edible wild plants are not allowed to be
brought into certain states or grown there because of their known potential as
agriculturally important weeds.

Many wild edibles are potentially major weeds even though they aren't illegal.
You need to take care that your plant adventures don't create problems for
yourself and others. Sometimes edible plants aren't cultivated for very good
reasons. You should not limit your care to the requirements of the law. Follow
the principle that your neighbors have a right to decide what to grow and
shouldn't be growing something of yours unless they planted it themselves.
 I grow unfamiliar species in small raised beds surrounded by regularly mowed
grass. If they prove to be moderately invasive, I can keep them contained. If they
seem dangerously invasive, I can cover the entire bed with plastic and destroy
them.

Plant breeding starts as soon as you choose what germplasm to perpetuate, as

soon as you choose to grow this plant or seed from this population and not that
one - that is, as soon as you bring a plant home and start growing it. If you
become enamored of a specific wildling, you may make many collections. And
you may do crosses of this plant with that one to try to combine desirable
characteristics of both. Everything you learned about how to breed the more
common plants applies to the rare vegetable or wildling. And, of course, you will
be practicing selection with every generation.

The fact is there are 20,000 plants that have useful edible parts. I exclude species
like pine trees, whose underbark is edible, and include only species that have
sizeable seeds or roots or leaves or stems or flowers or pods that are eaten
somewhere in the world. Out of those 20,000 only about a hundred have been
brought to an advanced agronomic level. Indeed, it's been said that a mere 22
crops feed the world. Wheat, rice and maize by themselves are supposed to
provide 50% of the weight of food for the world. This is a tiny, tiny larder from
which to feed a planet. It is a dangerous vulnerability.

- Noel Vietmeyer, "The New Crop Era," in Advances in New Crops

HE IRISH potato famine occurred because Irish peasants depended

almost exclusively on just one staple crop - potatoes - and on only a few closely
related varieties. Growing just one staple worked well for years. That's why they
did it. An individual farmer could support his family better with potatoes than
anything else. But then suddenly, without warning, a year came when growing
potatoes as the only staple didn't work; it didn't work at all.

Irish potatoes were sensitive to late blight. Because everyone grew potatoes
and was growing blight-sensitive potatoes, once the blight broke out, it spread
like wildfire. The entire crop was lost. Those who could emigrated. Thousands
starved to death.

In the past few decades, agricultural patterns have changed so that this
situation - fields planted in monoculture, entire regions dominated by single
varieties of single crops - has ceased to be the exception and has become the
rule. Such conditions are ideal breeding grounds for major epidemics of pests
and diseases. From the Southern corn blight in the United States in the 1970s to
the outbreak of whiteflies in California in the early 1990s, we are obtaining
increasing evidence that our current agricultural patterns are seriously flawed.
 We simply have too few food-crop species. And we are growing too few
varieties of those we have. Combined with unsustainable and ecologically
destructive agricultural practices, it is a recipe for disaster. We need to make
some major changes in our food production patterns. And we need new crops to
facilitate new patterns.

We must expand the food base. We need to be growing and using thousands of
crop species, not a few dozen. We need hundreds of varieties of every major
crop, not just a few. In addition, we need new crops and crop varieties that are
fine-tuned to specific regions, crops that require minimal inputs and labor, and
crops that help to preserve the land instead of destroying it. We also need to
integrate the preservation of general biodiversity into agriculture. In short, we
need to be practicing not agriculture but conservation agriculture. We need to
expand the horizons of agriculture on this planet.

Many professional plant breeders, agronomists, horticulturists, and

conservation biologists recognize the need for increased agricultural biodiversity
as well as new, sustainable, environment-friendly methods. They are doing what
they can. But there are too few professionals, and there is too much to do.

I believe that expanding the horizons of agriculture will depend predominantly

on the contributions and involvement of amateurs. This book is an endeavor to
enlist, empower, and encourage all gardeners and farmers everywhere to
participate in this effort.

I began breeding plants just for fun. I thought I would just be applying what I
already knew about genetics and gardening. But being involved in breeding my
own varieties changed everything. I think about plants and gardening differently,
see more possibilities, and am quicker to recognize and examine underlying
assumptions. Breeding my own vegetable varieties has expanded my horizons.

Developing your own varieties is fun, and it adds new dimensions to your
gardening. The value of breeding new, useful plants transcends the pleasure,
though. A new food plant is a unique contribution to society. To be able to make
that contribution is deeply rewarding.

When you start breeding your own plants, your contribution starts long before
you develop anything, though. The minute you do a cross, bring home a wild
plant, grow something unusual, or even just maintain and grow a rare variety,
you achieve a small increase in agricultural biodiversity. The means are part of
the end; you succeed in part just by beginning. As soon as you begin breeding
your own vegetable varieties, you begin expanding horizons.

Here is information about more than eight hundred interesting plants: scientific
names, common names, families, and life-styles; chromosome numbers, basic
breeding systems, flowering patterns, flower types and modifications, average
crosspollination frequency, major pollen vectors, and incompatibility system
information; recommended isolation distances, seed yields, location in the
USDA-ARS National Plant Germplasm System, and references.

I've placed the greatest emphasis on cultivated vegetables, edible wild plants
that can be used as vegetables, and relatives of both that might be used for
breeding. A secondary emphasis is on cultivated and wild fruits, including tree
fruits, because there is no clear distinction between fruits and vegetables. Most
plants that are normally consid

ered fruit crops can also be used as vegetables, and many are obviously both,
depending on the variety, the part of the plant used, or what you happen to be
doing with it. Many grains and nuts also are included.

All species are listed in alphabetical order by scientific name. If you don't
know the scientific name, you can find it by looking up the common name in the
index. The first species listed for each genus includes the family the genus is in.

Sometimes information about a single species is scattered among a number of

entries because different sources use different scientific names. I cross-
referenced where I could, but I usually did not try to correct the taxonomy; thus
there are contradictions in this table because there are contradictions in the
sources. In addition, information is sometimes listed under "spp." instead of
under the exact species. Where contradictory information exists, I sometimes list
a species twice to keep each set of information and its references intact. When
you look up something, it's a good idea to scan through all the entries for the
genus to check for contradictions or information listed under synonyms or "spp."

This table is, in part, a compilation of compilations. You can, however, find the
original sources by using the references in each entry.
 Plants are a mutable and variable lot. When I include a specific piece of
information, all it means is that someone, somewhere, when looking at one
particular cultivar or even one particular plant, thought he or she saw it that way.
There could be differences from cultivar to cultivar or population to population.
And people sometimes misidentify the plants they are describing.

No information on uses of the plants is included, as doing so would have made

the table larger than the rest of the book. Almost all the entries are included in
Stephen Facciola's book Cornucopia, which compiles all the available
information about the uses of three thousand cultivated and edible wild species.
Cornucopia gives references for everything and also includes sources for the
plants. See the bibliography for all the references listed here.

Categories of Information and Key to Abbreviations

Listed below are the basic categories of information and the most commonly
used

terms and abbreviations, given in the order in which they appear in the entries.

1-Life-style. Whether annual, biennial, or perennial, and whether an herb, a

shrub, a tree, a vine, grass, or a liana (woody vine). Examples: "l-a,v" refers to
an annual vine; "I-p/a,h,s" is a perennial that is raised as an annual and may be
an herb or a shrub (depending on cultivar, climate, growing methods, or growing
conditions).

Much of the information about life-styles and the system of abbreviations

describing it were taken from Duke, Hurst, and Terrell, Economic Plants and
Their Ecological Distribution.

2n = Diploid chromosome number. Information is usually from Duke, Hurst, and

Terrell, Economic Plants; Darlington and Wylie, Chromosome Atlas of
Flowering Plants; or Zevan and Zhukovsky, Dictionary of Cultivated Plants and
Their Centres of Diversity. These are compilations, but they give the original
references.

gf-Genomic formula. Helps trace genetic and evolutionary relationship between

species. See explanation in section on brassicas in Appendix B.
b-Basic breeding system. "b-o" and "b-i" refer to outbreeding and inbreeding,
respectively. Note that, practically speaking, most plants can both inbreed and
outbreed; what matters is the degree. Always look further in the listing for
information on crosspollination frequency and recommended isolation distances.

fFlower types, flowering patterns, flower specializations and modifications. f-rn,

f-d, and f-h refer to species that are monoecious, dioecious, and hermaphroditic
(perfect flowers), respectively. Flower modifications and specializations are
given after the basic flowering-pattern description. Examples: "f-rn plant
protandrous" means the species is monoecious -that is, plants have separate male
and female flowers, but the male flowers tend to develop before the female ones;
"f-h flower protogynous" means the species has perfect flowers, but the female
part of the flower develops and is receptive before the male part develops and
releases pollen.

x-Average percent crosspollination (contamination) frequency in inbreeding

plants and major vectors. Example: "x-3% by insects." Note that "x" varies
depending on region, weather, insects present, and other factors, so "x" figures
are intended to give you just a rough idea. When the frequency matters, you can
determine it yourself, as has been described in chapter 9.

s-Self-compatibility information. "s-i" and "s-c" are self-incompatible and self-

compatible, respectively. A listing of "s-i" means that at least some cultivars are
partially or totally self-incompatible. Other cultivars in the same species may be
self-compatible. Likewise, where incompatibility exists, it may be slight, causing
only a minor reduction in seed vigor or seed set, or it may be so major that no
seed forms at all without crosspollination.

v-Major pollen vector (in outbreeding plants). Example: "v-bees, flies."

d-Recommended isolation distances as given by various sources. Distances are

given in meters unless otherwise specified. Note that there are many ways to get
around or minimize the need for isolation, so don't get discouraged if you see a
huge isolation distance. (See part II.) Since different sources have different
levels of authoritativeness and usefulness (depending, for example, on your
climate), recommendations of the various sources have been kept distinct by
using a number following the "d." The d-figure is usually least authoritative, as it
represents older information that has been superseded in many cases.
d-Recommended isolation distances as given by Frankel and Galun, Pollination
Mechanisms, Reproduction, and Plant Breeding.

d2-Recommended isolation distances as given by George, Vegetable Seed

Production. These distances are considered appropriate for producing
foundation seed in Great Britain.

d3-Isolation distances recommended by the curator in charge of the species for

the USDA-ARS National Germplasm System collection or by some other
appropriate professional breeder or curator of the group. Names and positions
of sources are given.

d4-Recommended isolation distances as given by Fehr and Hadley,

Hybridization of Crop Plants.

y-and y2-Yield of seed when grown for seed:

y-Information referring to large commercial plantings; mostly from George,

Vegetable Seed Production. Usually in kilograms per hectare (kg/ha).
Sometimes from Duke, Handbook of Legumes of World Economic
Importance.

y2-Information from Jeavons, How to Grow More Vegetables, 2nd ed. (There is
a third edition now available.) Refers to small-scale, intensive planting in
raised beds using organic/biodynamic methods. Given in lb/100 sq.ft.max.,
which is pounds per hundred square feet of raised bed and represents the
maximum attained by Jeavons under excellent gardening conditions in
California. These figures are usually more relevant to small-scale backyard
growing.

USA-Location in USDA-ARS National Plant Germplasm System. Abbreviations

following "USA-" are those used by the system. See Appendix C for meanings
of abbreviations and addresses, phone numbers, and curators in charge of the
various collections. Information is from an unpublished list. Example: "USA-
CR-COR" means the species is in the clonal repository in Corvallis, Oregon.
"CR-COR" is an abbreviation used by the National Plant Germplasm System.
You can find it by looking in Appendix C.
r-References - sources of information for the entry. The major sources are
indicated by the abbreviations that follow. Occasional sources are written out.
Complete informa references is given in the bibliography. Sample reference: "r-
d,f,g,Smith 1982" would mean the entry is a compilation of information from
Duke, Hurst, and Terrell; Frankel and Galun; George; and Smith's book or article
of 1982. Abbreviations used are as follows:

r-b Bassett, Breeding Vegetable Crops. r-d Duke, Hurst, and Terrell, Economic
Plants and Their Ecological Distribution. A list of one thousand important
plants. This is where to go to find information about where things will grow.

r-d2 Duke, Handbook of Legumes of World Economic Importance. Best

compilation of breeding, growing, and use information on all the legumes.

r-d3 Darlington and Wylie, Chromosome Atlas of Flowering Plants.

r-f Frankel and Galun, Pollination Mechanisms, Reproduction, and Plant

Breeding.

r-f2 Fehr and Hadley, Hybridization of Crop Plants.

r-f3 Facciola, Cornucopia. Compilation of all available information on uses of

three thousand species, with references. Also gives sources for all the plants.

r-g George, Vegetable Seed Production.

r-j Jeavons, How to Grow More Vegetables.

r-1 Unpublished list. An internal document from the USDA-ARS National Plant
Germplasm System that is not readily available.

r-pc "Personal communication." Who told me is listed in parentheses after the

reference. Example: "r-pc(Dan Borman)."

r-po "Personal observation," which means I've worked with the species myself
and noticed or established the fact.

r-r Ryder, Leafy Salad Vegetables.

r-s Shinohara, Vegetable Seed Production Technology of Japan, vol. 1.

r-s2 Simmonds, Evolution of Crop Plants.

r-y Yamaguchi, World Vegetables.

r-z Zevan and Zhukovsky, Dictionary of Cultivated Plants and Their Centres of
Diversity. There is a more recent second edition that I didn't know about when
I compiled this table.
 General Abbreviations Used

cv., cvs. cultivar(s)

kg/ha kilograms per hectare

lb/100 sq.ft.max. pounds per hundred square feet in raised bed; represents
maximum yield under good conditions

m meters ("miles" is spelled out)

max. maximum

sp., spp. species (sing.), species (pl.)

ssp. subspecies

var. variety
 List of Plants

Abelmoschus esculentus (= Hibiscus esculentus) Malvaceae okra 1-a,h 2n =

72,144 b-i f-h x-5-20% by insects, hummingbirds d-400m d2-500m y-1,500
kg/ha; 500 in the tropics y2-9.3 lb/100 sq.ft.max. USA-S9 r-d,f,f3,g,j,l

Abelmoschus moschatus (= Hibiscus moschatus) ambrette, musk mallow 1-

a,b,p,h,s 2n = 72 USA-S-9 r-d,f3,l,z

Actinidia chinensis Actinidiaceae chinese gooseberry 1-p,s,v 2n = about

116+,160 USA-CR-DAV r-d,f3,1

Aegle marmelos Rutaceae bael, bengal quince I-p,t 2n = 28,36 r-d,f3

Agave sisalana Agavaceae sisal 1-p,h 2n = 138-150 b-o f-h protandrous v-

probably wasps, bees, insects; possibly also bats, gravity r-d,f3,s2

Albiziajulibrissin Mimosaceae silk tree 1-p,t 2n = 26,52 r-d3,f3

Allium All alliums that flower are b-o and f-h and protandrous, many extremely
protandrous. Usually s-c. Some form bulbs in flower heads and don't flower.

Allium allegheniense Liliaceae 2n = 14 r-d3

Allium ammophilum 2n = 16, 32 r-d3

Allium ampeloprasum (= A. porrum = A. babingtonii) leek, levant garlic,

perennial sweet leek 1-b/a,h 2n = 16,24,32,40,48(r-b) 2n = 32 (r-d3) b-o f-h s-
c v-bees d21,000m y-500 kg/ha; 600 max. y2-9.8 lb/100 sq.ft.max. USA-W-6
r-b,d,d3,f,f3,g,j,l,s2,z

Allium ampeloprasum wild leek 2n = 32 r-d3

Allium ampeloprasum var. elephant garlic 1-a,b/ a,h r-f3

Allium amplectens 2n = (14),21,28 b-apomictic r-d3

Allium angulosum mouse garlic 2n = 16-100! USA-W-6 r-d3,1

Allium ardelii 2n = 14 r-d3

Allium artemisietorum 2n = 16

Allium aschersonianum 2n = 16 r-d3

Allium babingtonii see A. ampeloprasum

Allium bakeri 2n = 32 r-d3

Allium bidwelliae 2n = 28 r-d3

Allium canadense tree onion 2n = 16 r-d3,f3

Allium carinatum 2n = 16,24,25,26 r-d3

Allium carmeli 2n = 14 r-d3

Allium cepa onion 1-b/a,h 2n = 16(r-b,s2) 2n = 16,32(r-d) 2n = 16,32(r-d3) b-o

f-h protandrous Male sterility exists and is used to make hybrids. x-93% s-c
Demonstrates some inbreeding depression. v-bees, flies, insects d-800m d2-
1,000m y-5001,000 kg/ha; 2,000 max. y2-(regular or tor pedo)10.3 lb/100
sq.ft.max. USA-NE-9 r-b,d,f,f3,g,j,l,s2

Allium cepa var. ascalonicum (= A. ascalonicum) shallot 2n = 16 r-b,d3,f2,g,j

Allium cepa Aggregatum group potato onion 2n = 16 r-b,f3

Allium cepa var. viviparum USA-NE-9 r-1

Allium cepa x Allium fistulosum Egyptian, tree, top-setting, or walking onion I-

p,h Propagated vegetatively from top sets or bulb divisions. Some varieties
flower before making top sets, others make top sets without flowering. r-
b,f3,po

Allium cernuum wild onion, nodding onion, lady's leek I-p,h 2n = 14,(16!) USA-
W-6 r-d3,f3,1

Allium chinense rakkyo 1-p/a,h 2n = 24,32(r-d) 2n = 16,32(r-z) r-d,f3,z

Allium chloranthum 2n = 16 r-d3

Allium ciliare 2n = 32 r-d3

Allium condensatum 2n = 17 USA-W-6 r-d3,1

Allium coppoleri 2n = 16 r-d3

Allium cyaneum 2n = 32 r-d3

Allium deseglisei 2n = 32 r-d3

Allium desertorum 2n = 16 r-d3

Allium dumetorum 2n = 16 r-d3

Album fistulosum Welsh onion, Japanese bunching onion 1-p/a,p/b,h 2n = 16 b-

o f-h s-c y2-39.6 lb/100 sq.ft.max. Yield listed for "bunching onion," which I
am assuming is A. fistulosum, but it might not be. USA-NE-9 r-
b,d,d3,f3,j,l,s2,z

Allium fuscoviolaceum 2n = 16 USA-W-6 r-d3,1

Allium fuscum 2n = 16 r-d3

Allium hirsutum 2n = 14 r-d3

Allium harataviense 2n = 18 USA-W-6 r-d3,1

Allium hunthianum 2n = 16 r-d3

Allium ledebourianum 2n = 16 USA-W-6 r-f3,l,z

Allium lusitanicum perennial Welsh onion 2n = 16 r-d3

Allium macranthum 2n = 14,28 r-d3

Allium margaritaceum 2n = 16,32 r-d3

Allium materculae 2n = 16 r-d3

Allium meteoricum 2n = 16 r-d3

Allium modestum 2n = 16 r-d3

Allium moly lily leek 2n = 14 b-apomictic r-d3

Allium moschatum 2n = 16 r-d3

Allium neapolitanum daffodil onion 2n = 14,21,35,28 r-d3

Allium nipponicum 2n = 16,32 r-d3,z

Allium nutans 2n = 16-108! r-d3

Allium odorum fragrant-flowered garlic, Chinese leek flower 1-p,h 2n = 16,32 r-

d3,f3

Allium oleraceum field garlic 2n = 32,40 USA-W-6 r-d3,f3,1

Allium pallasii (= A. pallyssium) 2n = 16

Allium pendulinum 2n = 14,(18!) r-d3

Allium porrum see A. ampeloprasum

Allium proliferum 2n = 16 r-d3

Allium pseudoflavum 2n = 18 r-d3

Allium pyrenaicum 2n = 16 USA-W-6 r-d3,1

Allium roseum 2n = 32 r-d3

Allium roseum var. bulbiferum 2n = 48 r-d3

Allium rotundum 2n = 32 r-d3

Album sativum garlic I-p/a,h 2n = 16(r-b) 2n = 16,48(r-d) 2n = 16(r-z) 2n = 16(r-

d3) USA-W-6 r-b,d,d3,f3,j,l,s2,z
Allium sativum var. ophioscordon (see also A. scorodoprasum) rocambole garlic
USA-W-6 r-f3

Allium schoenoprasum chives 1-p,h 2n = 26,24,32(r-d) 2n = 16(24,32)(r-z) 2n =

16,24,32(r-d3) b-o f-h s-i v-bees USA-W-6 r-d,d3,f,f3,l,z

Allium scorodoprasum (see also A. sativum var. ophioscorodon) rocambole 2n =

16,24 USA-W-6 r-d3,1

Allium scorzoneraefolium 2n = 14 r-d3

Allium senescens 2n = 32,48 USA-W-6 r-d3, f3,1

Allium sihhimense 2n = 32 r-d3

Allium splendens 2n = 32 1-h USA-W-6 r-d3,f3,1

Allium spp. For uses and sources of these and additional wild and cultivated
species, see r-f3.

Album spp. See pp. 228-230.

Allium stellatum prairie onion 2n = 14 r-d3,f3

Allium tartaricum tartar onion 2n = 28-32 r-d3

Allium tel-avivense 2n = 16 r-d3

Allium thunbergii 2n = 16 r-d3,f3

Allium triquetrum 2n = 18 r-d3,f3

Allium tuberosum Chinese chives, leek, or leek flower; gynmigit 1-p,h 2n =

16,24,32(r-d) 2n = 32(r-d3) 2n = 16,32(r-z) b-o USA-W-6 r-d,d3,f3,1,z

Allium ursinum ramsons 2n = 14 r-d3,f3

Allium validum Pacific onion, swamp onion I-p,h 2n = 28 r-d3,f3

Allium victorialis 1-h 2n = 16,32 USA-W-6 r-d3,f3,1

Allium vineale crow garlic 1-p,h 2n = 32,40 r-d3,f3

Allium viviparum 2n = 16 r-d3

Allium yunnanense (= A. mairei) 2n = 16,32 r-d3

Allium zebdanense 2n = 18 r-d3

Alocasia indica (= A. macrorrhiza?) Araceae 2n = 28 r-s2

Alocasia macrorrhiza giant taro 1-p/a,h 2n = 26,28 r-d,l,s2

Althaea officinalis Malvaceae marsh mallow 1-p,h 2n = 42(about 40-44) USA-

NC-7 r-d,f3,l,z

Amaranthus angustifolius Amaranthaceae 1-a,h 2n = 32(34) r-z

Amaranthus caudatus Inca wheat, quihuicha 1-a,h 2n = 32,34 USA-NC-7 r-d,d2,

d3,f3,1,z

Amaranthus caudatus var. love-lies-bleeding 1-a,h y-0.5 lb/sq.yd. r-J. L. Hudson

catalog

Amaranthus cruentus African, Chinese, Spanish,

or bush spinach or greens, amaranth 1-a,h 2n = 32,34 b-o v-wind d2-500m y-up
to 2,000 kg/ha USA-NC-7 r-d,f3,g,l,z

Amaranthus dubius African, Chinese, Spanish, or bush spinach or greens,

amaranth 1-a,h 2n = 64 (Might be a tetraploid form of A. cruentus.) b-o v-
wind d2-500m y-up to 2,000 kg/ha USA-NC-7 r-f3,g,l,z

Amaranthus gangeticus See A. tricolor

Amaranthus hybridus slim amaranth 2n = 32 USA-NC-7 r-f3,z

Amaranthus hypochondriacus (= A. leucocarpus) grain amaranth, huazontle,

princess feather 1-a,h 2n = 32,34(r-d) 2n = 32(r-z) USA-NC-7 r-d,f3,1,z

Amaranthus lividus livid amaranth 1-a,h 2n = 34 USA-NC-7 r-d,f3,l,z

Amaranthus mangostanus See A. tricolor

Amaranthus mantegazzianus (= A. edulis = A. caudata ssp. mantegazzianus) 2n

= 32 r-d3,E3,z

Amaranthus paniculatus (= A. cruentus?) 2n = 32 r-f3,z

Amaranthus spinosus thorny pigweed 2n = 32(r-z) 2n = 34(r-d3) USA-NC-7 r-

d3,f3,1,z

Amaranthus spp. Most amaranths are b-i,o and v-wind, perhaps also bees with
species with colorful flowers. Grain species are usually b-i,o f-m s-c.
Arrangement of flowers and flowering pattern usually produces a mix of self-
and crosspollination. "Each of the many cymes of the inflorescence is initiated
by a single staminate flower followed by an indefinite number of pistillate
flowers, often over a hundred. Stigmas of the earliest pistillate flowers are
receptive before the staminate flower opens; most of the later pistillate flowers
develop after the staminate flower has abscessed. However, cymes of different
ages are present on each indeterminate inflorescence and pollen transfer
among them probably makes selfing more common than crossing" (r-s2).

Amaranthus spp. grain types y2-16 lb/100 sq.ft.max. r-j

Amaranthus tricolor (= A. gangeticus = A. melancholicus = A. tricolor var.

mangostanus) African, Chinese, Spanish, or bush spinach or greens, amaranth
1-a,h 2n = 32,34 2n = 32(r-z) 2n = 34(r-d3) b-o v-wind d2-500m y-up to 2,000
kg/ha USA-NC-7 r-d,d3,f3,g,l,z

Amaranthus viridis green amaranth 2n = 34 r-f3,d3

Amphicarpaea monoica (= Falcata comosa) Fabaceae hog pea, ground peanut 2n

= 20 r-f3,z

Anacardium occidentale Anacardiaceae cashew 1-p,t 2n = 40,42 USA-CR-MAY

r-d,f3,l

Ananas comosus Bromeliaceae pineapple 1-p,h 2n = 50,75 b-o f-h s-i v-

hummingbirds USA-CR-MIA r-d,f,f3,1
Anchusa officinalis Boraginaceae true bugloss 1-b,p,h 2n = 16 USA-NC-7 r-
d3,f3,1

Anethurn graveolens Apiaceae dill I-a,b,h 2n = 22 USA-NC-7 r-d,f3,1

Angelica archangelica Apiaceae angelica 1-b,p,h 2n = 22 r-d,f3,1

Annona cherimola Annonaceae cherimoya 1-p,t 2n = 14,16 b-o f-h flowers

strongly protogynous v-beetles USA-CRMIA r-d,f,f3,l

Annona muricata soursop I-p,s,t 2n = 14,16 r-d,f3,l

Annona reticulata custard apple 1-p,s,t 2n = 14,16 r-d,f3,1

Annona squamosa sugar apple 1-p,t 2n = 14,16,28 USA-CR-MIA r-d,f3,1

Anthriscus cerefolium Apiaceae chervil 1-b,p/ a,h 2n = 18 r-d,d3,f3,1

Anthriscus sylvestris woodland chervil, cow parsley 1-p,h 2n = 16,18 r-d3,f3

Apios americana (= A. tuberosa = Glycine apios) Fabaceae American groundnut

1-p,v 2n = 22 b-o s-i v-insects r-f3,z, Blackmon & Reynolds 1986, Reynolds,
et al. in Janick & Simon, eds., Advances in New Crops 1990
Apium carvi see Carum carvi

Apium graveolens var. dulce Apiaceae celery I-p/b?,a,b,h 2n = 22 b-o f-h flower
protandrous x-30% by insects s-c v-insects d2-500m or more y-500 kg/ ha y2-
9.9 lb/100 sq.ft.max. USA-NE-9 r-d,f,f3,g,j,l,r

Apium graveolens var. rapaceum celeriac 1-a,b,h d2-500m or more USA-NE-9 r-

d,f3,g

Apium graveolens var. secalinum leaf celery r-f3, r

Arachis hypogaea Fabaceae groundnut, peanut 1-a,h 2n = 4x = 40 b-i f-h x-

occasionally by bees d-0 y-1 tonnes/ha y2-24 lb/ 100 sq.ft.max. USA-S-9 r-
d,d2,f,f3,g,j,1

Aralia cordata Araliaceae udo 1-p,h 2n = 28 USA-S-9 r-d,f3,l,z

Arctium lappa Asteraceae great burdock, edible burdock, gobo 1-p/a,b/a,h 2n =

32,36 r-d,f3,l

Armoracia rusticana (= Cochlearia armoracia) Brassicaceae horseradish 1-p,h 2n

= 32(r-s2) 2n = 28,32(r-d,z) Propagated vegetatively. Highly seed sterile. r-
d,f3,j,l,s2

Artemisia abrotanum Asteraceae southernwood 1-p,s 2n = 18 r-d,l

Artemisia dracunculus tarragon 1-p,h 2n = 18,36,54 r-d,f3,l

Artemisia vulgaris mugwort I-p,b,h 2n = 18,16,36 USA-W-6 r-d,f3,1

Artocarpus altilis Moraceae breadfruit 1-p,t 2n = 56?,54?,81 r-d,l

Artocarpus heterophyllus jackfruit 1-p,t 2n = 56? USA-CR-MIA r-d,l

Asclepias cornutii see A. syriaca

Asclepias curassavica Asclepiadaceae bloodflower 1-p 2n = 22(r-d3) r-d3,1

Asclepias incarnata milkweed 2n = 22 r-d3,f3

Asclepias spp. 2n = 22 for all eight species listed in d-3 USA-S-9 r-d3,1

Asclepias speciosa milkweed, showy milkweed 1-p,h 2n = 22 r-d3,f3

Asclepias syriaca (= A. cornutii) common milkweed, silkweed 1-p,h 2n =

(20),22,(24) (r-z) r-d3,f3,1,z

Asclepias tuberosa pleurisy root, butterfly weed 1-p,h 2n = 22(r-d3) USA-W-6 r-

d3,f3,1 Asimina triloba Annonaceae pawpaw 1-p,t 2n = 18,16,2x? USA-NC-7
r-d,l

Asparagus acutifolius Liliaceae wild Mediterranean asparagus 2n = ? USA-NC-7

r-b,f3,1

Asparagus densiflorus sprenger asparagus I-p,h 2n = 60 USA-NC-7 r-d,l

Asparagus maritimus 2n = ? USA-NC-7 r-b,l

Asparagus officinalis asparagus 1-p,h 2n = 20,40 b-o f-d However, male flowers
have vestigal stigmas that sometimes set an occasional seed; female flowers
have nonfunctional vestigal stamens. Also, andromonoecious plants (which
have perfect and male flowers on the same plant) exist; their anatomy suggests
that they are largely self-pollinating (r-b). v-insects Fresh pollen can be stored
in vials in a home freezer for six months; longer with lower temperature and/or
control of relative humidity (r-b). USA-NC-7 r-b,d,f,f3,l,z

Asparagus springeri 2n = ? Resistant to Fusarium but doesn't cross with A.

officinalis. r-b

Asperula odorata Rubiaceae sweet woodruff I-p,h 2n = 44 r-d,f3,l

Atriplex canescens Chenopodiaceae fourwing saltbush 1-p,s USA-W-6 r-d,f3,1

Atriplex hortensis orach, mountain or German spinach, butter leaves, sea

purslane 1-a,h 2n = 18 USA-W-6 r-d,f3,l

Atriplex spp. saltbush b-o f-d v-wind USA-W-6 r-f,l

Avena spp. Poaceae oat b-i f-h Largely cleistogamous; upper flower usually
 imperfect. x-up to 10% by wind d-0 y2-13.2 + lb/100 sq.ft.max. USA-NSGC
r-f,f3,j,l

Averrhoa bilimbi Oxalidaceae bilimbi 1-p,t 2n = 22,24 r-d,f3,1

Averrhoa carambola carambola 1-p,t 2n = 22,24 USA-CR-MIA r-d,f3,l Bactris

gasipaes Arecaceae peach palm 1-p,t 2n = 30 r-d,f3,1

Balanites aegyptiaca Zygophyllaceae desert date 1-p,t 2n = 16,18 r-d,f3,1

Barbarea verna (= B. praecox) Brassicaceae creasy greens, upland or winter

cress 1-b,p,h 2n = 16 r-f3,z

Barbarea vulgaris yellow rocket, common winter cress, upland cress 1-b,p/a,h 2n
= 14,16 2n = 16(r-z) r-d,f3,l,z

Basella alba Bassellaceae Malabar spinach 1-p/b,a,h 2n = 48,44 USA-S-9 r-

d,f3,1 Bellis perennis Asteraceae daisy b-i f-h d-f,f3,1

Benincasa hispida Cucurbitaceae waxgourd, winter melon 1-a,h,v 2n = 24 b-o f-

m USA-NE-9 r-d,f3,l,s2(p306)

Beta vulgaris ssp. cycla Chenopodiaceae Swiss chard, chard, spinach beet 1-
a,b/a,h b-o fh flower protandrous s-i v-wind d-2km; 5km from ssp. esculenta
y-1.1 tonnes/ha; 2 max. y2-29 lb/100 sq.ft.max. USA-NC-7 r-d,f,f3,g,j,l

Beta vulgaris ssp. esculenta sugar, fodder, or red beet; beetroot; mangold;
mangel I-a,b/a,h 2n = 18,27,36 b-o f-h flower protandrous s-c,i v-wind d2-
500-1,000m; 5km from ssp. cycla y-1,000 kg/ha; 2,000 max. USA-NC-7 r-
d,f,f3,g,j,l

Bombax ceiba Bombacaceae red silk cotton 1-p,t 2n = 46 r-d,f3,1

Borago officinalis Boraginaceae borage 1-a,h 2n = 16 USA-W-6 r-d,f3,l

Borassus flabellifer Arecaceae palmyra palm I-p,t 2n = 36 r-d,f3,1

Brassica campestris ssp. chinensis (= B. chinensis = B. rapa ssp. chinensis)

Brassicaceae Chinese mustard, pak choy, pe-tsai, Chinese leaf cabbage 1-b/a,h
 2n = 20 gf-AA USA-NE-9 r-f2,1,r

Brassica campestris ssp. dichotoma (= B. napus ssp. dichotoma) toria 2n = 20 gf-

AA USA-.NC-7 r-f2,1

Brassica campestris ssp. eu-campestris bird rape, wild turnip rape 1-a,h 2n = 20
gf-AA r-d,f2,l

Brassica campestris ssp. nipposinica (= B. nipposinica) curled mustard 2n = 20

gf-AA

Brassica campestris ssp. oleifera (= B. rapa ssp. oleifera) turnip rape 2n = 20 gf-
AA USA-NC-7 r-d,f2,l

Brassica campestris ssp. pekinensis (= B. pehinensis) Chinese cabbage, Chinese

head cabbage, celery cabbage 1-b/a,h 2n = 20 gf-AA b-o f-h s-i v-bees y2-6.1
lb/100 sq.ft.max. USA-NE-9 r-d,f,f2,j,l

Brassica campestris ssp. rapifera (= B. rapa) turnip 1-b,h 2n = 20 gf-AA b-o f-h
s-i v-bees, insects, wind d2-1,000m y1,500 kg/ha; 1,500 max. y2-14.7 lb/100
sq.ft.max. USA-NC-7 r-d,f,f2,g,j,l

Brassica campestris ssp. ruvo? (= B. rapa ssp. ruvo? = B. ruvo?) broccoli raab
(sprouting turnip) The Seed Savers Exchange lists broccoli raab as B. rapa,
Ruvo group, and refers to it as a sprouting turnip. However, the taxonomy I
am following here calls turnips B. campestris. Most sources don't list broccoli
raab at all.

Brassica campestris ssp. sarson sarson 2n = 20 gf-AA r-f2

Brassica carinata Abyssinian mustard 2n = 34 gf-BBCC USA-NC-7 r-f2,1

Brassica hirta (= Sinapsis alba) white mustard, yellow mustard 1-a,h 2n = 24 gf-
DD r-d,f2,l

Brassicajuncea mustard, brown mustard, Oriental mustard, leaf mustard, Indian

mustard 1-p/a,b,h 2n = 36 gf-AABB b-i f-h s-c x-about 20% (r-f2) v-bees y2-
5.7 lb/100 sq.ft.max. USA-NC-7 r-d,f,f2,j,l,r
Brassica haber (= Sinapsis arvensis) charlock, wild mustard 2n = 18 gf-SS r-
f,f2,1

Brassica napus ssp. oleifera rape I-a,b,h 2n = 38 gf-AACC b-o f-h x-more than
10% s-c v-bees d-40m y2-5.4 lb/100 sq.ft.max. USA-NC-7 r-d,f,f2,j,1

Brassica napus ssp. pabularis rape-kale

2n = 38 gf-AACC USA-NC-7 r-f2,1 Brassica napus ssp. rapifera (= B.

napobrassica = B. napus ssp. napobrassica) rutabaga, swede 1-b,a,h 2n = 38
gf-AACC b-i f-h x-about 20% (r-f2) s-c v-bees, insects, wind d2-1,000m y-
1,500 kg/ha; 2,500 max. y2-5.4 lb/100 sq.ft.max. USA-NC-7 r-d,f,f2,g,j,l

Brassica nigra black mustard 1-a,h 2n = 16 gf-BB USA-NC-7 r-d,f2,1

Brassica oleraceae ssp. acephala kale, collards 1-b/a,a,h 2n = 18 gf-CC b-o f-h s-
i v-bees, flies d2-1,000-1,500m from all B. oleraceae varieties y2-(kale)3.8
lb/100 sq.ft.max. USA-NC-7 r-b,d,f,f2,g,j,l

Brassica oleraceae ssp. botrytis cauliflower, broccoli 1-b,h 2n = 18 gf-CC b-o f-h
s-i However, summer cauliflowers self-pollinate readily compared with other
B. oleraceae subspecies. v-bees, flies d2-1,000-1,500m from all B. oleraceae
subspecies y-400 kg/ha y2(broccoli)-5.5 lb/100 sq.ft.max. y2(cauliflower)-1
lb/100 sq.ft.max. USANE-9 r-d,f,f2,g,j,l

Brassica oleraceae ssp. capitata cabbage 1-b/a,h 2n = 18 gf-CC b-o f-h s-i v-
bees, flies d2-1,000-1,500m from all B. oleraceae subspecies y-700 kg/ha y2-
3.6 lb/100 sq.ft.max. USA-NE-9 r-b,d,f,f2,g,j,l

Brassica oleraceae var. gemmifera Brussels sprouts 1-b/a,h 2n = 18 b-o f-h s-i v-
bees, flies d2-1,000-1,500m from all B. oleraceae subspecies y-600 kg/ha y2-
2.8 lb/100 sq.ft.max. USA-NE-9 r-d,f,g,j,l

Brassica oleraceae ssp. gongylodes kohlrabi 1-b,h 2n = 18 gf-CC b-o f-h s-i v-
bees, flies d2-1,000-1,500m from all B. oleraceae subspecies y-700 kg/ha y2-
20.1 lb/100 sq.ft.max. USA-NE-9 r-d,f,f2,g,j,l

Brassica oleraceae ssp. italica sprouting broccoli, asparagus broccoli 1-p,h 2n =

18 gf-CC v-bees, flies d2-1,000-1,500m from all B. oleraceae subspecies
 USA-NC-7 r-d,f2,g,l

Brassica oleraceae ssp. sylvestris wild cabbage 2n = 18 r-f2

Brassica rapa see B. campestris

Brassica spp. See pages 230-234

Brassica spp. See r-f3 for uses of brassicas.

Brassica tournefortii wild turnip 2n = 20 gf-AA r-f2

Brosimum alicastrum Moraceae ramon I-p,t 2n = 26 r-d

Bunias orientalis Apiaceae Turkish rocket I-p,h 2n = 14,44? r-d3,f3

Bunium bulbocastanum Apiaceae earth chestnut 1-p,h 2n = ? r-f3

Cajanus cajan (= C. indicus) Fabaceae pigeon pea 1-a,p,p/a,s 2n = 22,44,66 b-o

f-h s-c v-honeybees x-13-65% y-1.2 tonnes/ ha y-600 kg/ha ave.; 2,500 max.
(r-d2) y2-24 lb/100 sq.ft.max. USA-S-9 r-d,d2, f,f3,g,j,l

Calendula officinalis Asteraceae pot marigold 1-a,h 2n = 32,28 USA-NC-7 r-

d,f3,l

Caltha palustris Ranunculaceae marsh marigold, American cowslip 1-p,h 2n =

28-60 USA-NC-7 r-d,f3,l

Camassia cusichii Liliaceae 2n = 30 r-d3,f3

Camassia leichtlinii (= C. esculenta) camas, quamash 1-p,h 2n = 30 r-d3,f3,z

Camellia sinensis Theaceae tea 1-p,t 2n = 30,45,60 b-o f-h s-i v-insects r-d,f,f3

Campanula rapunculus Campanulaceae rampion, ramps 1-b,p/a,h 2n = 68,102(r-

d) 2n = 20,102(r-z) r-d,f3,z

Canarium indicum Burseraceae java almond 1-p,t r-d,f3

Canarium ovatum pili nut 1-p,t b-o f-d v-insects r-f,f3

Canavalia ensiformis Fabaceae jackbean Ia,p,p/a,h 2n = 22 b-i x-20% by bees s-
c y2-700-5,400 kg/ha USA-S-9 r-d,d2,f,f3,1

Canavalia gladiata swordbean 1-a,p,p/a,h,l 2n = 22,44 x-more than 20% s-c v-

bees y-1.3 tonnes/ha y2-700-1,500 kg/ha USA-S-9 r-d,d2,f3,g,1

Canavalia plagiosperma oblique jackbean 1-a,h 2n = 22 r-d,d2

Canna edulis Cannaceae edible canna, Australian arrowroot, achira 1-p,h 2n =

18,27 r-d,f3

Capparis spinosa Capparaceae caper 1-p,s 2n = 24,38 r-d,f3

Capsicum annum Solanaceae pepper 1-a,h b-i f-h Protogynous; anthers dehisce
two to three days after flowers open. x-5-10% by bees, thrips (r-D; up to 68%
in India (r-g) Some will cross with C. frutescens. d-30m y-100-200 kg/ha
USA-S-9 r-d,f,f3,g,l

Capsicum baccatum var. pendulum pimentchien I-a,h 2n = 24 USA-S-9 r-d,f3,l

Capsicum chinense aji 1-a,h 2n = 24 USA-S9 r-d,f3,1

Capsicum frutescens pepper, tabasco pepper 1p/a,h,s 2n = 24 b-i f-h

Protogynous; anthers dehisce two to three days after flowers open. x-7-36% by
bees, thrips (r-U; up to 68% in India (r-g) Some will cross with C. annum.
USA-S-9 r-d,f,f3,g,l

Capsicum pubescens rocoto I-p,h,s 2n = 24 USA-S-9 r-d,f3,1

Capsicum sp. cayenne peppers y2-0.1 lb/100 sq.ft.max. r-j

Capsicum sp. sweet/green peppers y2-0.3 lb/ 100 sq.ft.max. r-j

Carduus marianum see Silybum marianum

Carica papaya Caricaceae papaya 1-p,s,t 2n = 18,36 b-o f-d v-wind, insects
USA-CR-HIL r-d,f,f3,l

Carica pubescens mountain papaya l-p,t 2n = 18 r-d,f3

Carissa carandas Apocynaceae karanda 1-p,s,t 2n = 22 r-d,f3,1

Carissa edulis Egyptian carissa 1-p,l,s 2n = 22 r-d,f3

Carissa macrocarpa natal plum 1-p,s 2n = 22 r-d,f3

Carthamus tinctorius Asteraceae safflower 1-a,h 2n = 24,32 b-o f-h s-c v-bees d-
5-90m y2-17.4 lb/100 sq.ft.max. USA-W-6 r-d,f,f3,j,l

Carum carvi (= Apium carvi = Seseli carvi) Apiaceae caraway 1-p/a,b,h 2n =

20,22 USA-NC-7 r-d,f3,l

Carya illinoensis Juglandaceae pecan l-p,t 2n = 32 b-o f-m Variable as to

whether stigma becomes receptive before or after pollen is shed. s-c v-wind
USA-PlO r-d,f,f3,l

Caryota urens Arecaceae fishtail palm 1-p,t 2n = 32 r-d,f3

Casimiroa edulis Rutaceae Mexican apple 1-p,t 2n = 36 r-d,f3

Castanea pumila Fagaceae Allegheny chinquapin 1-p,s,t USA-CR-BRW r-d,f3,1

Castanea spp. American, Japanese, and Chinese chestnuts 1-p,s,t 2n = 22,24 b-o
f-m s-i v-insects USA-CR-BRW r-d,f3,1

Ceiba pentandra Bombacaceae kapok 1-p,t 2n = 80,88,72 USA-CR-MIA r-d,f3,l

Celastrus scandens Alastraceae American bittersweet b-o f-d r-f,f3,1

Celosia argentea Amaranthaceae cock's comb, green soko, white soko v-insects
y-up to 600 kg/ha USA-NC-7 r-f3,g,l
Centranthus angustifolius Valerianaceae 2n = 14 r-d3

Centranthus gilloti 2n = 14 r-d3

Centranthus ruber (= Kentranthus ruber) Jupiter's beard, red valerian 1-p,h 2n =

14 USA-NC-7 r-d3,f3,1

Ceratonia siliqua Fabaceae carob 1-p,t 2n = 24 b-o f-d v-wind, insects USA-S-9
r-d,d2,f,f3,l

Ceratotheca sesamoides Pedaliaceae bungu l-a,h 2n = 32 USA-S-9 r-d,f3,l

Chaerophyllum bulbosum Apiaceae turniprooted chervil 1-b,p/a,h 2n = 22 r-

d,f3,1

Chenopodium album Chenopodiaceae lamb'squarters, mutton tops, fat hen 1-a,h

2n = 18,36,54 USA-NC-7 r-d,f3,j,l,z

Chenopodium ambrosioides epazote, Mexican tea, American wormweed, Indian

wormweed 1-a,p,h 2n = 16,32,36,64 r-d,f3,z

Chenopodium bonus-henricus (= C. esculentus) Good King Henry, allgood,

mercury, fat hen 1-p,h 2n = 36 r-d3,f3,z

Chenopodium bonus-henricus b-o f-h Strongly protogynous. Pistils usually dry

up before anthers emerge and dehisce. On smaller flowers on a scape, anthers
may not emerge at all. A scape that has just started to flower has emerged
pistils on dozens of the earliest flowers but no emerged stamens. v-wind r-po

Chenopodium botrys ambrosia, Jerusalem oak, feather geranium 2n = 18 r-f3,z

Chenopodium capitatum (= C. foliosum) strawberry blite, beetberry 1-a,h 2n =

16,18 USA-NC-7 r-f3,l,z

Chenopodium esculentus see C. bonus-henricus

Chenopodiumfoliosum see C. capitatum

Chenopodium hybridum 2n = 18,36 r-d3

Chenopodium nuttalliae (= C. berlandieri?) huazontle 2n = 36 Seems to be just a
latematuring C. quinoa with edible flower/seed heads (a chenopodial broccoli).
Breeding system presumably like that of quinoa. r-f3,s2,z

Chenopodium pallidicaule (= C. canihua) kaniwa, 1-a,h 2n = 18(r-s2 and others)

2n = 36(r-z) b-i f-h cleistogamous xrare; about 0.1% in greenhouse (r-s2) y-up
to 2,400 kg/ha; 5,000 in experimental plots USA-NC-7 r-f3,1,p,s2,z,Lost
Crops of the Incas 1989(pp 134-136)

Chenopodium petiolare 1-p,h r-Lost Crops of the Incas 1989(footnote, p137) (If
anyone has any of this, please send me some. I want to try it as another source
for introducing perennialism into chenopodial grains and vegetables.)

Chenopodium quinoa quinoa, quinua, petty rice 1-a,h 2n = 36 b-i Plants are
gynomonoecious - that is, they have a mixture of hermaphrodite and female
(pistillate) flowers. x-2-9% Preceeding information from s2. Simmonds has
studied the breeding system of chenopodiums. Contradictory breeding
information exists. r-f lists C. quinoa as b-o f-h v-wind. J. Rea, in Lost Crops
of the Incas 1989 (p159) is quoted in a one-line footnote as saying it is b-i + f-
h usually, but f-m in "a few cases." s2 notes that there is a report of one
population carrying a cytoplasmic/genetic male sterile, so outbreeding.must be
common in at least that population. USA-NC-7 r-d,f,f3,1,s2,z,Lost Crops of
the Incas 1989 (p159)

Chrysanthemum coronarium shungiku, garland chrysanthemum 1-a,h 2n = 18,36

b-i f-h (except ray florets, which are pistillate)

x-about 10% y-130 kg/10 a; 200 max. (r-s) USA-NC-7 r-d,f3,l,s,z

Chrysophyllum cainito Sapotaceae star apple 1-p,t 2n = 24,26 r-d,f3

Cicer arietinum Fabaceae chickpea, garbanzo 1-a,h 2n = 16 b-i f-h

Cleistogamous; self-pollination occurs one to two days before anthesis.
"Flowers visited by bees, but crosspollination is rare." x-small amount by bees
d3-3 feet or whatever is needed to prevent physical mixture of different
varieties; isolation distance needs established at Pullman, Washington; under
those conditions, chickpea is one of the most highly inbreeding species (r-pc,
Rich Hannan, curator of USDA Cicer collection, USA-W-6). Ashworth, in
 Seed to Seed (1991), says that "several researchers" say that crosspollination
occurs commonly and chickpeas cross readily under organic growing
conditions (r-pO). One-half mile is reasonable. y = 400-1,600 kg/ha ave.; can
exceed 2,000 kg; has reached 5,200 on experimental plots (r-d2) y2-24 lb/100
sq.ft.max. USA-W-6 r-d,d2,f,f2,f3,j,1,pc(Rich Hannan),s2

Cichorium endivia Asteraceae endive 1-a,b,h 2n = 18,36 b-i f-h x-15% by flies
d10m Can cross with C. intybus (r-r). y-500 kg/ha; 1,000 max. USA-NC-7 r-
d,f,f3,g,l,r

Cichoriurn intybus chicory l-p,p/b,h 2n = 18 b-o f-h s-i d2-1,000m Can cross
with C. endivia (r-r). y-1 tonne/ha USA-NC-7 r-d,f,f3,g,l,r

Citrullus colocynthis Cucurbitaceae colocynth 1-a,p,h,v 2n = 22,24 USA-S-9 r-

d,f3,1

Citrullus lanatus (= C. vulgaris) watermelon I-a,h,v 2n = 22 b-o Most cvs. f-m;

some, especially older cvs., andromonoecious - that is, have a combination of
hermaphrodite and male flowers on each plant. s-c v-bees d800m y-400 kg/ha
for cvs, with normal seed yield; 250 kg/ha for `Charleston Grey' y2-2.6 lb/100
sq.ft.max. USA-S-9 r-b,d,f,f3,g,j,l,s2

Citrus spp. Rutaceae citrus b-i f-h Many cvs. apomictic. USA-CR-RIV or USA-
PIO r-f,f3,l

Clausena dentata Rutaceae clausena I-p,s 2n = 18 r-d

Clausena lansium wampi 1-p,t 2n = 18 r-d,f3

Clitoria laurifolia Fabaceae laurel-leaved clitoria, butterfly pea 1-p,s 2n = 24

USA-S-9 r-d,d2,f3

Clitoria ternatea butterfly pea I-p,v 2n = 16 b-i s-c USA-S-9 r-d,d2,f3,1

Coccoloba uvifera Polygonaceae seagrape 1-p,s,t 2n = 80,132 r-d,f3

Cochlearia armoracia see Armoracia rusticana

Cocos nucifera Arecaceae coconut 1-p,t 2n = 32 b-o f-m plant protandrous x-

depends on cv. v-wind, insects r-d,f,f3

Coleus amboinicus Lamiaceae Indian borage 1-p,h 2n = 68,32 r-d

Coleus parvilorus ratala 1-a,h 2n = 56,64 r-d,f3

Colocasia esculenta Arecaceae taro, cocoyam, 1-p/a,h 2n = 28 3n = 42 USA-CR-

MAY r-d,f3,l,s2,y

Corchorus capsularis Tiliaceae white jute 1-a,h 2n = 14,28 USA-S-9 r-d,f3,l

Corchorus olitorius tussa jute 1-a,h 2n = 14,28 USA-S-9 r-d,f3,l

Cordyline terminalis Agavaceae ti palm 1-p,s 2n = 152 + r-d,f3

Coriandrum sativum Apiaceae coriander I-a,h 2n = 22 USA-NC-7 r-d,f3,1

Corydalis solida (= C. bulbosa) Fumariaceae 2n = 16,32,24 r-d3

Corylus spp. Corylaceae hazelnut, filbert 1p,s,t 2n = 28,22 b-o f-m s-i v-wind
USA-CR-COR r-d,f,f3,1

Crambe abyssinica Brassicaceae crambe, Abyssinian kale 1-a,h 2n = 90 b-i

mostly, but some outcrossing occurs f-h USA-NC-7 r-d,d3,f2,1

Crambe cordifolia (= C. tataria) colewort, Tatarian sea kale 1-p,h 2n = about 120
USA-NC-7 r-d3,f3(p50),l,z

Crambe filiformis 2n = 30 USA-NC-7 r-d3,1 Crambefruticosa 2n = 30,60 r-d3,1

Crambe grandii fora 2n = about 120 USA-NC7 r-d3,1

Crambe hispanica Spanish colewort 1-a,h 2n = 60 r-d3,z

Crambe hispanica var. glabrata 2n = 30 USA-NC-7 r-f2,1

Crambe hispanica var. hispanica 2n = 60 USA-NC-7 r-f2,1

Crambe hralihii 2n = 30,60,90 USA-NC-7 r-f2,1

Crambe maritima sea kale 1-p,h 2n = 30,60(r-d) 2n = 60(r-z) 2n = 30,60(r-d3) b-

o f-h s-i v-bees d-200m USA-NC-7 r-d,f,f2,f3(p50),l,z,P6ron in Janek &
Simon, eds., Advances in New Crops 1990

Crambe orientalis var. juncea 2n = about 120 r-d3

Crambe orientalis var. koktebelica 2n = 30 r-d3,f3(p51)

Crambe tataria see C. cordifolia

Crambe tatarica Tartar bread plant 2n = 60,120 r-d3,f3(p51)

Crithmum maritimum Apiaceae samphire 2n = 20,22 r-d3,f3

Crocus sativus Iridaceae saffron crocus l-p,h 2n = 14,15,16 USA-W-6 r-d,f3,1

Cryptotaeniajaponica Apiaceae mitsuba, mitsube, Japanese honewort, celery, or

parsley I-p,h 2n = (18),20,22 r-f3,y,z

Cucurnis anguria Cucurbitaceae West Indian gherkin I-a,h 2n = 22,24 USA-NC-

7 r-d,f3,l

Cucumis melo cantaloupe, melon, muskmelon 1-a,v 2n = 24(r-s2) 2n =

20,22,24(r-d) b-o f-m or with perfect and male flowers on one plant plant
protandrous x-0-100% s-c v-bees d2-500-1,000m y-300 kg/ha; 600 max. y2-
2.9 lb/100 sq.ft.max. Note that `Armenian Cucumber' is a C. melo, not a true
cucumber (C. sativus). USA-NC-7 r-d,f,f3,g,j,l,s2

Cucumis metuliferus jelly melon, African horned melon 1-a 2n = 24 USA-NC-7

r-d3,f3,1

Cucumis sativus cucumber 1-a,h 2n = 14 b-o f-m,h plant protandrous Some

modern cvs. have only or almost only female flowers under normal growing
conditions. x-70% s-c v-bees Does not cross with muskmelons, watermelons,
squash, pumpkins, or gourds. (That's a common myth.) However, note that the
`Armenian Cucumber' is not a cucumber but a C. melo, so it will cross with
muskmelons and other C. melo subspecies. d2-1,000-1,500m y-400 kg/ha; 700
 max. y2-4.1 lb/100 sq.ft.max. USA-NC-7 r-b,d,f,f3,g,j,l,s2

Cucurbita ficifolia (= C melanosperma) Cucurbitaceae Malabar, buffalo, or

figleaf gourd 1-p/a,v 2n = 40(r-s2) 2n = 40,42(r-d) USA-S-9 r-d,d3,f3,1,s2

Cucurbita maxima squash, marrow, pumpkin I-a,b,h,v 2n = 40(r-s2) 2n =

40,44,48(r-d) b-o f-m s-c v-bees d2-1,000-1,500m y-500 kg/ha; 1,000 max.
y2-(winter squash)5.7 lb/100 sq.ft.max. USA-NE-9 r-b,d,f,f3,g,j,l,s2

Cucurbita melanosperma see C. ficifolia

Cucurbita mixta squash, cushaw, marrow, pumpkin 1-a,v 2n = 40 d2-1,000-

1,500m y-500 kg/ha; 1,000 max. r-b,d,f3,g,l,s2

Cucurbita moschata squash, marrow, pumpkin 1-a,v 2n = 40(r-s2) 2n =

24,40,48(r-d) d2-1,000-1,500m y-500 kg/ha; 1,000 max. USA-S-9 r-
b,d,f3,g,l,s2

Cucurbita pepo squash, marrow, pumpkin

1-a,v,h 2n = 24,28,40 d2-1,000-1,500m y-500 kg/ha; 1,000 max. y2-

(pumpkin)5.1 lb/100 sq.ft.max. y2-(crookneck, patty pan, or zucchini squash)6.1
lb/100 sq.ft.max. USA-NC-7 r-b,d,f3,g,j,l

Cucurbita spp. squash, marrow, pumpkin, gourd b-o f-m s-c v-bees Do not
usually display strong inbreeding depression. r-f, r-Allard 1960

Cucurbita spp. See pp. 234-238

Curcuma domestica Zingiberaceae turmeric 1-p,h 2n = 32,64,63 r-d,f3

Curcuma zedoaria zedoary 1-p,h 2n = 64,63 r-d,f3

Cyamopsis pedata Fabaceae archucha 1-p/a,v 2n = 32 r-d

Cyamopsis tetragonoloba (= C. psoralioides) guar, clusterbean 1-a,h 2n = 14 y-

700 kg/ha; up to 1,750 irrigated (r-d2) USA-S-9 r-d,d2,f3

Cyclanthera pedata Cucurbitaceae accocha, archucha 2n = 32 USA-S-9 r-d,f3,1

Cydonia oblonga Rosaceae quince I-p,t 2n = 34 USA-CR-COR r-d,f3,1

Cymbopogon citratus Poaceae lemongrass 1-p,g 2n = 40,60 USA-S-9 r-d3,f3,1

Cynara cardunculus Asteraceae cardoon 1-p,h 2n = 34 USA-NE-9 r-d,f3,1

Cynara scolymus globe artichoke, artichoke 1-p,p/a,h 2n = 34 USA-NE-9 r-

d,f3,l,z

Cyperus esculentus Cyperaceae yellow nutsedge I-p,h 2n = 18,108 r-d,f3

Cyperus rotundus purple nutsedge 1-p,h 2n = 108 r-d,f3

Cyphomandra betacea Solanaceae tree tomato 1-p,s,t USA-W-6 r-d,f3,1

Cyrtosperma chamissonis (= C. edule = C. merhusii) Araceae giant swamp taro

r-f3,s2

Dahlia rosea Asteraceae dahlia b-o f-h s-i r-f

Daucus carota Apiaceae carrot I-a,b,h 2n = 18,22 b-o f-h flower protandrous
Displays severe inbreeding depression (r-Allard 1960). s-c v-bees, flies,
insects d-8001,600m Crosses readily with the common wild carrot Queen
Anne's lace. y-600 kg/ha; 1,000 max.; European types in the tropics 300 kg/ha;
Asian types in the tropics 250 kg/ha y2-17.8 lb/ 100 sq. ft. max. USA-NC-7 r-
b,d,f,f3,g,j,l

Dendrocalamus spp. Poaceae dendrocalamus I-p,g 2n = 72 r-d,f3

Desmanthus illinoensis Fabaceae Illinois bundleflower 1-p,h The Land Institute

is working on domesticating this species. USA-S-9 r-f3,1,The Land Report
Spring 1991

Dictamnus albus Rutaceae gas plant I-p,h 2n = 30,36 USA-W-6 r-d,f3,1

Dioscorea alata Dioscoreaceae winged yam 1-p,v 2n = 20,30,40 r-d,f3

Dioscorea batatas see D. opposita

Dioscorea bulbifera air potato 1-p,v 2n = 36,40,54 r-d,f3

Dioscorea composita composite yam 1-p,v 2n = 36,54 r-d

Dioscorea convolvulacea convolvulus yam 1-p,v 2n = 36 r-d

Dioscoreafloribunda floribunda yam 1-p,v 2n = 36,54,72 r-d

Dioscorea japonica glutinous yam 2n = 40 r-f3,z

Dioscorea macrostachya cuculmeca 1-p,v r-d

Dioscorea opposita ( = D. batatas) Chinese yam, cinnamon vine 1-p,v 2n = about

140, about 144 r-d,f3,z

Dioscorea rotundata eboe yam 2n = 40 r-d

Dioscorea spp. yam (not related to American sweet potato, which is sometimes
called a yam) b-o f-d Many have been maintained vegetatively, and in many
seed set is rare; may be more likely if plants are allowed to mature beyond
normal harvest time. USA-CR-MIA r-d,l,s2

Dioscoreophyllum cumminsii Menispermaceae serendipity berry 1-p,v,s,t 2n =

24 r-d,f3

Diospyros digyna Ebanaceae black sapote 1-p,t r-d,f3

Diospyros kaki Japanese persimmon 1-p,t 2n = 54,56 r-d,f3

Diospyros spp. persimmon 1-p,t b-o f-d v-insects USA-CR-DAV r-f,l

Diospyros virginiana common persimmon 1-p,t 2n = 60,90 r-d,f3

Dipteryx odorata Polypodiaceae cumaru tonka bean 1-p,t 2n = 32 b-o v-insects r-

d,d2

Distichlis palmeri Poaceae l-p,g Wild halophyte used by Cucapa Indians to make
bread and atole. Efforts by N. and S. B. Yensen in Tucson, Arizona, are under
way to domesticate it into a salt-tolerant perennial grain. See Wagoner 1990
for review. r-f3,Wagoner 1990.
Dolichos biflorus see Macrotyloma uniflorum
Dolichos lablab see Lablab purpureus
Dolichos uniflorus see Macrotyloma uniflorum

Echinochloa crusgalli Poaceae barnyard grass I-p,a,g 2n = 36,54,56 USA-NC-7

r-d,f3,1

Echinochloa crusgalli var. frumentacea billion dollar grass I-a,g 2n = 36,54,56

USANC-7 r-d,f3,1

Echinochloa pyramidalis antelope grass 1-p,g 2n = 36,54,72 USA-NC-7 r-d,l

Elaeis guineensis African oil palm 1-p,t 2n = 32,36 b-o f-m r-d,f

Elaeis oleifera Arecaceae American oil palm I-p,s,l 2n = 32 r-d

Eleocharis dulcis (= E. tuberosa) Cyperaceae waternut, water chestnut 1-p,h 2n =

30? r-d,d3,f3

Elettaria cardamomum Zingiberaceae cardamom, cardamum 1-p,b,h 2n = 48,52

USA-S-9 r-d,f3,1

Eleusine coracana Poaceae finger millet 1-a,g 2n = 36 USA-S-9 r-d,f3,l

Eleusine indica goosegrass 1-a,g 2n = 18,36 USA-S-9 r-d,f3,l

Elymus canadensis Poaceae Canada wild rye 1-p,g 2n = 28,42 USA-W-6 r-

d,f3,l

Elymus cinereus basin wild rye I-p,g 2n = 28,56 r-d

Elymus condensatus giant wild rye 1-p,g 2n = 28,56 r-d

Elymus glaucus blue wild rye 1-p,g 2n = 28 USA-W-6 r-d,l

Elymus junceus Russian wild rye 1-p,g 2n = 14 r-d

Elymus spp. wild rye b-i f-h d-o USA-W6 r-f,l

Eragrostis chloromelas Poaceae Boer lovegrass I-p,g 2n = 40,60,63 USA-W-6 r-

d,l
Eragrostis curvula weeping lovegrass 1-p,g 2n = 20,40,50 USA-W-6 r-d,l

Eragrostis lehmanniana lehmann lovegrass 1-p,a,g 2n = 40,60 USA-W-6 r-d,l

Eragrostis tef teff 1-a,g 2n = 40 USA-W-6 r-d,f3,1

Eragrostis trichodes sand lovegrass 1-p,g b-i f-h d-0 USA-W-6 r-d,f,l

Eriobotryajaponica Rosaceae loquat 1-p,s,t 2n = 32,34 USA-CR-COR r-d,f3,l

Eruca vesicaria ssp. sativa (= E. sativa) Brassicaceae arugula, rocket, garden

rocket, rocket salad, roquette 1-a,h 2n = 22 gf-EE USA-NC-7 r-d,f2,f3,1,z

Eryngiumfoetidun Apiacea false coriander 1-b,h 2n = 16 r-d,f3

Eucalyptus spp. Myrtaceae eucalyptus I-p,t 2n = 22 b-o f-h plant protandrous s-i
v-bees r-d,f,f3

Eugenia uniflora Myrtaceae Surinam cherry 1-p,s,t 2n = 22 r-d,f3

Euryaleferox Euryalaceae Gorgon water lily 2n = 58 r-d3,f3

Eutrema wasabi see Wasabia japonica
Fagopyrum cymosum Polygonaceae perennial

buckwheat, shakuchiri-soba 1-p,h 2n = 16(r-s2) b-unknown, but "it may

reasonably be assumed to share with F esculentum a heteromorphic
incompatibility system" (r-s2) r-f2,f3,s2

Fagopyrum esculenturn buckwheat 1-a,h 2n = 16,32(r-d) 2n = 16(r-s2) b-o f-h s-i

Strong but not absolute incompatibility. Incompatibility is associated with two
anatomically different flower types. Plants have either pin or thrum flowers.
Pin flowers have short stamens and long pistils. Thrum flowers have long
stamens and short pistils. Usually, thrums can be pollinated successfully only
by pins and pins by thrums, so an individual plant doesn't normally self-
pollinate, even though it serves as both male and female. Whenever self-
pollination occurs, severe inbreeding depression usually results. v-insects d-
hasn't been established (r-f2) y2-30 lb/100 sq.ft.max. r-d,d3,f2,f3,j,l,s2,z

Fagopyrum tataricum Tatarian or Medawaska buckwheat; duck or India wheat 1-

a,h 2n = 16 b-? f-h s-c There is just one anatomical type of flower. Some seed
dormancy? USA-NE-9 r-d,f2,f3,1,p,s2,z

Feijoa sellowiana Myrtaceae feijoa 1-p,s 2n = 22 r-d,f3

Feijoa sp. feijoa b-o f-h s-i,c v-bees r-f

Ferula communis (see also Foeniculum vulgare) Apiaceae giant anise fennel 1-
p,h 2n = 22 r-d3,f3

Ficus carica Moraceae fig 1-p,s,t 2n = 26 b-o f-d v-wasps USA-CR-DAV r-

d,f,f3,1

Ficus elastica rubber plant 1-p,t 2n = 26,39 r-d

Ficus vogelii vogel fig 1-p,t 2n = 26 r-d

Flacourtia indica Flacourtiaceae 1-p,s,t 2n = 22,18 r-d,f3

Foeniculum vulgare Apiaceae fennel 1-p,b/a,h 2n = 22 Bronze, wild, bitter,

sweet, Flor ence, and salad fennels are all F. vulgare. I've seen giant anise
fennel listed as both Foeniculum vulgare and Ferula communis. I don't know
 whether these are the same. USA-NC-7 r-d,f3,l

Fortunella spp. Rutaceae kumquat 1-p,s,t 2n = 18,36 USA-PIO r-d,l

Fragaria chiloensis Rosaceae Chilean strawberry 1=p,h 2n = 56 USA-CR-COR

r-d,l

Fragaria spp. strawberry b-o f-h, h + female, female (male sterile) flowers
protogynous s-c v-bees USA-CR-COR r-f,l

Fragaria vesca European strawberry 1-p,h 2n = 14,35 USA-CR-COR r-d,l

Fragaria virginiana Virginia strawberry 1-p,h 2n = 56 USA-CR-COR r-d,l

Fragaria x ananassa garden strawberry 1-p,h 2n = 56 USA-CR-COR r-d,l

Garcinia mangostana Clusiaceae mangosteen 1-p,t 2n = about 76,96 r-d,f3

Garcinia tinctoria gamboge tree 1-p,t 2n = 80 r-d

Gaultheria procumbens Ericaceae wintergreen 1-p,s 2n = 24 USA-CR-COR r-

d,f3,l

Gaylussacia baccata Ephedraceae black huckleberry 1-p,s 2n = 24 USA-CR-

COR r-d,f3,l

Gladiolus spp. Iridaceae gladiolus b-o f-h s-c r-f,f3

Glycine apios see Apios americana

Glycine max Fabaceae soybean 1-a,h 2n = 40 b-i f-h d-0 y-1.3 tonnes/ha y2-
14.4+ lb/100 sq.ft.max USA-SOY-N r-d,d2,f,f3,g,j,l

Glycine wightii (= Notonia wightii) perennial soybean 1-p,v, l 2n = 22,44 most

b-i; tetraploids may be b-o (r-d2) r-d,d2,f3,1

Glycyrrhiza glabra Fabaceae common licorice 1-p,h 2n = 16 USA-W-6 r-

d,d2,f3,1

Glycyrrhiza lepidota American licorice 1-p,h 2n = 16 USA-W-6 r-d,d2,f3,1

Grewia asiatica Tiliaceae phalsa I-p,t 2n=36 r-d,f3,l

Helianthus annuus Asteraceae sunflower 1-a,h 2n = 14,34(r-d) 2n = 34(r-d3) b-o

f-h flower protandrous x-20-75% s-variable,i Does not display strong
inbreeding depression (r-Allard 1960). v-bees d-800m USA-NC7 r-
d,d3,f3,j,l,Allard 1960

Helianthus maximilianii 1-p,h 2n = 34 USANC-7 r-d,d3,f3,1

Helianthus tuberosus Jerusalem artichoke 1-p,p/ a,h 2n = 102 hexaploid b-o f-h
probably s-i Normally propagated vegetatively. Seed may be obtained when
more than one variety is grown if it is not too cold during the period of
flowering and seed development (r-po). USA-NC-7 r-d,d3,l,po,z

Hemerocallis spp. Liliaceae daylily 1-p,h 2n = 22,33,44 b-o USA-NC-7 r-

d3,f3,1

Heracleum lanatum (= H. maximum) Ariaceae American cow parsnip 1-p,h 2n =

? USA-W-6 r-f3,1

Heracleum sphondylium branca-ursina, European cow parsnip 1-p,h 2n = 22 r-

d3,f3

Hesperis matronalis Brassicaceae sweet rocket 1-p 2n = 24 USA-NC-7 r-f3,l,z

Hibiscus cannabinus Malvaceae kenaf 1-a,h 2n = 36,72 b-i f-h x-2-45% by bees
USA-S-9 r-d,f,f3,l
Hibiscus esculentus see Abelmoschus esculentus
Hibiscus moschatus see Abelmoschus moschatus

Hibiscus sabdariffa roselle 1-a,s 2n = 36,72 b-i f-h USA-S-9 r-d,f,f3,1

Hippophae rhamnoides Elaeagnaceae sea buckthorn 1-p,t b-o f-d v-wind USA-
NC-7 r-d,f,f3,1

Hordeum bulbosum Poaceae bulbous barley I-p,g 2n = 14,28 USA-NSGC r-

d,f3,1

Hordeum jubatum foxtail barley 1-g b-i f-h d-O USA-NSGC r-f,f3,1

Hordeum sp. y2-(barley, beardless)24 lb/100 sq.ft.max r-j

Hordeum spp. "Hordeum species are classified into four sections ... Grain-
producing forms are in the section Cerealia. The basic chromosome number is
seven with species in cerealia being diploid, and those in other sections being
either diploid, tetraploid, or hexaploid." Cerealia includes three cultivated
species and two wild species (H. vulgare, H. distichum, and H. irregulare and
H. spontaneum and H. agriocrithon, respectively). Crosses can be made among
all of them with almost complete fertility, so some consider all five as variants
of a single species. r-f2

Hordeum spp. 1-p,g Wild perennial relative(s) exist that can be crossed with
barley (H. vulgare). r-Wagoner 1990

Hordeum vulgare barley l-a,g 2n = 14,28 b-i f-h Many cvs, are cleistogamous,
can depend on environment and weather. x-up to 10% by wind (r-U;
"generally less than 0.2%" (r-f2) "Six-rowed types tend to have more natural
crossing than tworowed types, and natural crossing tends to be higher in
awnless, awnleted, and naked types" (r-f2). d-0 USA-NSGC r-d,f, f2, f3,1

Hordeum vulgare hybrids barley hybrids b-i f-h + f-male-sterile d-200m d-f

Houttuynia cordata Saururaceae tsi 1-p,h 2n = 56,96 r-d,f3

Humulus lupulus Cannabaceae hops 1-p,v 2n = 20 b-o f-d v-wind USA-CR-

COR r-d,f3,1
Hyscyamus niger Solanaceae black henbane 1-a,b,p,h 2n = 34 USA-NC-7 r-d,l

Hyssopus officinalis Lamiaceae hyssop I-p,h 2n = 12,13 r-d,f3,1

Inga edulis Fabaceae ice cream bean 1-p,t 2n = 26 r-d,d2,f3

Inula helenium Asteraceae elecampane 1-b,p,h 2n = 20 r-d3,f3

Ipomoea aquatica Convolvulaceae swamp morning glory, water spinach 1-p,v,h

2n = 30 USA-S-9 r-d,f3,l,z

Ipomoea batatas sweet potato (unrelated to the yam, Dioscorea spp., although the
orange varieties are called yams in the United States) I-p,p/a,v 2n = 6x = 90(r-
b,f2,s2) 2n = 84,90(r-d) b-o f-h flower protogynous s-i v-bees, bumblebees,
insects Varieties differ widely in ability to flower and set seed. Some require a
shock, such as grafting or cutting the vine one-third of the way through and
then propping the cut open with a matchstick. Seed may need scarification to
germinate. USA-S-9 r-b,d,f2,f3,j,l,s2

Iris spp. Iridaceae iris b-o f-h s-c USA-W-6 r-f,f3,1

Jubaea chilensis Arecaceae chile coco palm 1-p,t 2n = 32 r-d,f3

Juglans cinerea Juglandaceae butternut 1-p,t 2n = 32 r-d,l

Juglans nigra black walnut 1-p,t 2n = 32 USA-CR-DAV r-d,l

Juglans regia English walnut 1-p,t 2n = 32,36 b-o f-m plant protandrous s-c v-
wind USA-CR-DAV r-d,f,l

Kaempferia galanga Zingiberaceae kentjoer 1-p,h 2n = 22,54 r-d,f3

Kerstingiella geocarpa see Macrotyloma geocarpum

Kochia scoparia Chenopodiaceae kochia, summer cypress 1-p/a,b,h 2n = 18

USA-W-6 r-d,z

Lablab purpureus (= Dolichos lablab = Lablab niger) Fabaceae hyacinth bean,

lablab 1-a,p,h,v 2n = 22,24 y-450 kg/ha intercropped; 1,460 kg/ha in
monoculture y-0.9 tonnes/ha (r-g) USA-S-9 r-d,d2,f3,g,l

Lactuca denticulata Asteraceae 2n = 10 r-z

Lactuca indica Indian lettuce 2n = 18 USA-W-6 r-f3,l,z

Lactuca perennis perennial lettuce 1-p,h 2n = 18(r-d3) USA-W-6 r-d3,f3,1

Lactuca quercina 2n = 18 r-l,z

Lactuca sativa lettuce 1-a,h 2n = 18(r-r) 2n = 18,36(r-d) b-i f-h partly

cleistogamous x-1-6% by flies, insects d-10m d22m Isolate from L. serriola. y-
0.5 tonnes/ha (r-g) y2-(head)1.2 lb/100 sq.ft.max. y2(leaf) 2 lb/100 sq.ft.max.
USA-W-6 rb,d,f3,g,j,l,r Also see pp. 213-216

Lactuca sativa Angustan group (= var. asparagina) asparagus lettuce, celtuce,

stem lettuce r-f3,g,z

Lactuca serriola wild lettuce, prickly lettuce f-h b-i Crosses freely with L. sativa.

Lactuca spp. Species with 2n = 16: aurea, bracteata, bourgaei, macrantha, and
others. Species with 2n = 18: chondrillaeflora, perennis, raddiana, saligna,
sativa, serriola, squarrosa, tatarica, viminea, virosa, and others. Species with 2n
= 34: campestris, canadensis, ludoviciana, spicata, and others. L. orientalis is
listed with 2n = 36. r-d3

Lactuca virosa bitter lettuce, lettuce opium 1-a,b,h 2n = 16,18(r-d) 2n = 18(r-d3)

USA-W-6 r-d,d3,f3,1

Lagenaria siceraria Cucurbitaceae clabash, edible gourd, white-flowered gourd

1-a,v 2n = 22 USA-S-9 r-f3,1
Lallemantia iberica Lamiaceae dragon's head I-a,h USA-W-6 r-d,f3,1

Lansium domesticum Meliaceae langsat 1-p,t 2n = 72 r-d,f3

Lathyrus cicera Fabaceae garousse 2n = 14 b-i f-h USA-W-6 r-d3,f,l

Lathyrus hirsutus rough pea 1-a,h,v 2n = 14 USA-W-6 r-d,d2,l

Lathyrus sativus grass pea 1-a,h,v 2n = 14 b-i f-h USA-W-6 r-d,d2,f3,1

Lathyrus tingitanus Tangier pea b-i f-h USA-W-6 r-d,f,l

Lathyrus tuberosus earth chestnut 1-p 2n = 14 r-d3,f3

Lavandula angustifolia Lamiaceae lavender 1-p,s 2n = 54,50,36 USA-W-6 r-

d,f3,l

Lavandula latifolia broad-leaved lavender 1-p,s 2n = 54 r-d,f3

Lavandula stoechas French lavender 1-p,s 2n = 30 r-d,f3,1

Lens culinaris Fabaceae lentil 1-a,h 2n = 14 b-i f-h xrarely (r-d2) d3-3 feet or
whatever is needed to prevent physical mixture of different varieties; one of
the most highly inbreeding species; isolation distances tested and established
in Pullman, Washington (r-pc from Rich Hannan, curator of U.S. Lens
collection, USA-W-6). y-450-675 kg/ha dry cultivated; 1,688 irrigated; 3,000
max. (r-d2) y2-8.0+ lb/100 sq.ft.max. USA-W-6 r-d,d2,f,f3,pc(Rich Hannan),j,l

Lepidium latifolium Brassicaceae 1-p,h 2n = 24 r-f3,l,z

Lepidium meyenii maca 2n = ? r-f3,z

Lepidium sativum garden cress, common cress 1-a,b,h 2n = 16,32(r-z) 2n =

16,24,32 (r-d) USA-NC-7 r-d,f3,l,z

Lepidium virginicum Virginia pepperweed 1-a,b,h 2n = 32 r-f3,d

Leucaena leucocephala (= L. latisiliqua = L. glauca) Mimosaceae leucaena l-p,s,t

2n = 104,36 USA-S-9 r-d,d2,f3,1
Levisticum officinale Apiaceae lovage 1-p,h 2n = 22 USA-NC-7 r-d,f3,l

Leymus arenarius Poaceae strand wheat (wild perennial grain cultivated by the
Vikings) 1-p,g r-f3,Wagoner 1990

Leymus racemosus Volga wild rye (Cool-season perennial grass with large,
edible seeds; relative of rye. Efforts are under way at The Land Institute to
develop perennial grains from this species.) 1-p,g USA-W-6 r-f3,l,The Land
Report Spring 1991,Wagoner 1990

Lilium spp. Liliaceae lily I-p,h 2n = 24(r-d3) b-o f-h s-i USA-NC-7 r-d3, f,f3,1

Linum usitatissimum Linaceae flax 1-a,h 2n = 30,32 b-i f-h x-3% by bees d-0
USA-NC-7 r-d,f,f3,1

Litchi chinensis Sapindaceae litchi 1-p,t 2n = 28,30 USA-CR-MIA r-d,f3,1

Lomatium californicum Apiaceae wild celery parsley I-p,h 2n = ? r-f3

Lomatium dissectum fern-leaved biscuit-root 1-p,h 2n = ? r-f3

Lomatium macrocarpum large-fruited biscuit-root r-f3

Lomatium nudicale lomatium, cow parsley, smyrnium 1-p,h 2n = ? r-f3

Lomatium utriculatum desert gold, porno celery r-f3

Luffa acutangula Cucurbitaceae angled luffa, Chinese okra 2n = 26 b-o f-m

USA-S-9 r-d,f3,l,s2(p306)

Luffa aegyptiaca (= L. cylindrica?) smooth luffa, spongegourd 1-a,v 2n = 26

USA-S-9 r-d,f3,l

Luffa cylindrica (= L. aegyptiaca?) 2n = 26 b-o f-m r-f3,1,s2(p306)

Lupinus albus Fabaceae white lupine I-a,h 2n = 26,22 b-i but bee-pollinated f-h
x10% by bees v-bees USA-W-6 r-d,d2,f,f3,1

Lupinus angustifolium blue lupine 1-a,h 2n = 40,48 b-i bees not required USA-
W6 r-d,d2,l
Lupinus luteus yellow lupine 1-a,h 2n = 52,104 b-i mainly bee-pollinated f-h x-
10-25% by bees USA-W-6 r-d,d2,f,f3,1

Lupinus perennis lupine 2n = 48,96 b-i f-h USA-W-6 r-f,f3,l,z

Lycium chinense Solanaceae Chinese boxthorn I-p,s 2n = 24 r-d,f3

Lycopersicon esculentum Solanaceae tomato 1-a,b,p,h 2n = 24 b-i f-h

Protogynous, but flower and anther cone are pendant, thus facilitating self-
pollination. x-less than 2% by solitary bees, thrips; may be much higher in the
tropics or in long-styled cvs. Note: Some hierlooms and modern cvs. are long-
styled, so you should examine the flowers. d-30m d2-30-200m, largely to
prevent physical mixing y-250-500 kg/ha in the U.S.; 10-50 kg/ ha in Africa
y2-5.5 lb/100 sq.ft.max. USANE-9 r-b,d,f,f2,f3,g,j,l Also see pp. 206-213

Lycopersicon pimpinellifolium currant tomato 1-p,a,h 2n = 24,48 b-i, s-c USA-

NE-9 r-b,d,f3,1

Macadamia spp. Proteaceae macadamia nut 1-p,t 2n = 28,56 USA-PIO r-d,f3,1

Macadamia ternifolia Macadamia nut b-o s-partially c USA-PIO d-f,l

Macroptilium atropurpureum (= Phaseolus atropurpureus) Fabaceae siratro 1-

p,h,v 2n = 22 r-d2

Macrotyloma geocarpum (= Kerstingiella geocarpa = Voandezia poissonii)

Fabaceae Kersting's groundnut, groundbean, potato bean 1-a,h 2n = 20,22 r-
d,d2

Macrotyloma unit forum (= Dolichos uniflorus = Dolichos biflorus) horsegram,

Madras gram I-a,h 2n = 20,22,24(r-d2) USA-S-9 r-d2,f3,1

Malpighia punicifolia Malphighiaceae acerola 1-p,s,t 2n = 20 USA-CR-HIL r-

d,f3,1

Malus spp. Rosaceae apple 1-p,t USA-CRGEN r-d,f3,1

Malva alcea Malvaceae 2n = about 84 USA-NC-7 r-d3,1

Malva brasiliensis 2n = about 112 r-d3

Malva crispa curled leaf salad mallow l-a,h 2n = about 112 r-d3,f3

Malva moschata musk mallow I-p,h 2n = 42 b-o f-h v-bees USA-NC-7 r-d3,l,pc
(Heine Refsing),po

Malva neglecta (= M. rotundifolia) 2n = 42 r-d3

Malva nicaeensis 2n = 42 r-d3

Malva parv flora 2n = 42 USA-NC-7 r-d3,f3,1

Malva pusilla (= M. borealis) 2n = 42,76 r-d3,1

Malva sylvestris common mallow 2n = 42 r-d3,f3

Malva sylvestris var. mauritiana 1-p,b,h r-pc (Heine Refsing)

Malva verticillata 2n = about 84 USA-NC-7 r-d3,f3,1

Mangifera indica Anacardiaceae mango 1-p,t 2n = 40 b-o f-m; some apomixis s-

c,i v-flies USA-CR-MIA r-d,f,f3,1,s2

Manihot esculenta Euphorbiaceae cassava, manioc, tapioca 1-p,s,t 2n = 36

allotetraploid b-o I'm plant protogynous x-0-100% v-wasps, bees, insects d4-
30m; 500m for genetic studies Seed tends to be dormant for three to twelve
months after harvest. r-d,f,f2, f3,s2

Manihot glaziovii ceara rubber 1-p,t 2n = 36 USA-CR-MIA r-d,f3,l

Manilkara bidentata Sapotaceae balata 1-p,t r-d,f3

Manilhara zapota sapodilla 1-p,t 2n = 26 USA-CR-MIA r-d,f3,l

Maranta arundinacea Arantaceae arrowroot 1-p,h 2n = 18,48 r-d,f3

Marrubium vulgare Lamiaceae horehound I-p,h 2n = 34,36 USA-W-6 r-d,f3,l

Matricaria chamomilla Asteraceae wild chamomile 1-a,h 2n = 18 r-d,f3,l

Medicago sativa Fabaceae alfalfa, lucerne Ip,h 2n = 16,32,64 b-o f-h Must be
tripped by insects. x-more than 80% s-i,c Displays extreme inbreeding
depression (r-Allard 1960). v-honeybees, bumblebees, insects d-90m y2-1.1
lb/100 sq.ft.max. USA-W-6 r-d,d2,f2,j,l, Allard 1960

Medicago spp. "Essentially all annual species are cleistogamous and are
exclusively self-pollinated. Generally, the perennial species require tripping,
and will set seed from either self-or crosspollination" (r-f2). For breeding and
life style information on other species, see r-d2; for use as edibles see r-f3.

Melicoccus bijugatus Sapindaceae Spanish lime 1-p,t 2n = 32 r-d,f3

Melilotus alba Fabaceae white sweetclover Ib,a,h 2n = 16,24,32 v-honeybees

Insects required for pollination. USA-NC-7 r-d2,1

Melilotus spp. See r-d2,f3.

Melissa officinalis Lamiaceae lemon balm 1-p,h 2n = 32,64 r-d,f3

Mentha piperita Lamiaceae peppermint 1-p,h 2n = 36,48,64 b-o f-h v-flies USA-
CRCOR r-f,l

Mentha spp. and crosses include field mint, red mint, apple mint, spearmint, and
European pennyroyal. Alll-p,h. USA-CR-COR r-d,f, f3,l

Michelia champaca Magnoliaceae champac l-p,t 2n = 38 r-d,f3

Momordica balsamina Cucurbitaceae balsam apple, bitter melon 1-a,v 2n = 22

USA-S-9 r-d,f3,l

Momordica charantia balsam pear, bitter melon, bitter gourd 1-a,v 2n = 22 b-o f-
m USA-S-9 r-d,f3,l,s2(p306)

Monstera deliciosa Araceae ceriman I-p,l,v 2n = 24,56,60 r-d,f3

Montia perfoliata Portulacaceae miner's lettuce 1-a,h 2n = 12,24,36 r-d,f3

Morinda citrifolia Rubiaceae Indian mulberry 1-p,t r-d,f3

Moringa oleifera Moringaceae horseradish tree 1-p,t 2n = 28 r-d,f3

Morus alba Moraceae white mulberry 1-p,s,t 2n = 28 b-o f-d v-wind USA-CR-
DAV r-d,f,f3,l

Morus nigra black mulberry 1-p,t 2n = 89-308 b-o f-d v-wind USA-CR-DAV r-
d,f,f3,l

Morus rubra red mulberry 1-p,t 2n = 28 USA-CR-DAV r-d,f3,1

Mucuna spp. Fabaceae velvetbean 1-a,h,v 2n = 22 b-i x-rare USA-S-9 r-d,d2,f3

Murraya koenigii Rutaceae curryleaf tree 1-p,s,t 2n = 18 r-d,f3

Musa acuminata Musaceae dwarf banana 1-p,h,s 2n = 22,23 r-d

Musa x paradisiaca banana I-p,h,s 2n = 22,32-35 r-d,f3

Myrciaria cauliflora Myrtaceae Brazilian grape tree 1-p,t r-d,f3

Myristica fragrans Myristicaceae nutmeg 1-p,t 2n = 42,44 r-d,f3

Myrrhis odorata Apiaceae myrrh, sweet cicely I-p,h 2n = 22 r-d,f3

Myrtus communis Myrtaceae myrtle 1-p,s 2n = 22 r-d,f3

Nasturtium microphyllum see Rorippa microphylia
Nasturtium officinale see Rorippa nasturtiumaquaticum

Nelumbo lutea Nelumbonaceae American lotus 2n = 16 r-d3,f3

Nelumbo nucifera (= N. speciosum) lotus 1-p,h 2n = 16 r-d,d3,f3

Nephelium lappaceum Sapindaceae rambutan 1-p,t 2n = 22 r-d,f3

Nephelium mutabile pulasan 1-p,t r-d,f3

Nicotiana rustica Solanaceae b-i f-h USATOBAC r-f,l

Nicotiana tabacum tobacco 1-p/a,h 2n = 24,48,72 b-i x-2-3% by hummingbirds,

bees d-50m USATOBAC r-d,f,f3,1

Nigella sativa Ranunculaceae black cumin 1-a,h 2n = 12 r-d,f3

Notonia wightii see Glycine wightii

Nuphar lutea Nymphaeaceae yellow water lily 2n = 34 r-d3,f3

Nymphaea capensis Nymphaeaceae Cape water lily 2n = 28 r-d3

Nymphaeaflava (= N. mexicana) 2n = 56 r-d3

Nymphaea gigantea Australian water lily 2n = 224 r-d3

Nymphaea lotus Egyptian water lily 2n = 56 r-d3,f3

Nymphaea odorata American water lily 2n = 84 r-d3,f3

Nymphaea rubra red water lily 2n = 56 r-d3

Nymphaea stellata 2n = 28 r-d3,f3

Nymphaea tetragona pygmy water lily 2n = 112 r-d3

Nymphaea tuberosa magnolia water lily 2n = 84 r-d3,f3

Ocimum basilicum Lamiaceae sweet basil I-a,h 2n = 48 USA-NC-7 r-d,f3,l

Ocimum hilimandscharicum hoary basil 1-a,h,s 2n = 76 USA-NC-7 r-d,f3,1

Ocimum sanctum holy basil 1-a,p,h 2n = 64 USA-NC-7 r-d,f3,l

Oenanthejavanica Apiaceae water dropwort 1-p,h 2n = 20 r-d,f3

Olea europaea Oleaceae olive 1-p,t 2n = 46 USA-CR-DAV r-d,f3,1

Opuntia ficus-indica Cactaceae prickly pear 1-p,s 2n = 22,88 r-d,f3

Origanum vulgare Lamiaceae wild marjoram 1-p,h 2n = 30,32 USA-NC-7 r-

d,f3,l

Origanum vulgare marjoram b-i f-h d-90m USA-W-6 r-d,f,f3,1

Ornithopus sativus Fabaceae serradella 1-a,h 2n = 14,16 b-i r-d,d2,l

Oryza glaberrima Poaceae African rice 1-a,h 2n = 24 USA-NSGC r-d,l

Oryza longistaminata African wild rice (wild perennial relative that can be
crossed with rice) 1-p,g r-f3,Wagoner 1990

Oryza rufipogon wild perennial relative that can be crossed with rice 1-p,g r-
Wagoner 1990

Oryza sativa rice 1-a,g 2n = 24 b-i f-h Some cvs. are cleistogamous; varies with
climate. d-3m y2-24 lb/100 sq.ft.max. USA-NSGC r-d,f,f3,j,l

Oryzopsis hymenoides Poaceae Indian rice grass, rice grass, bunchgrass I-p,g 2n
= 48 USA-W-6 r-d,f3,1

Oryzopsis miliacea smilograss 1-p,g 2n = 24 USA-W-6 r-d,l

Oxalis tuberosa Oxalidaceae oca 1-p/a,h 2n = 14,60-70 USA-NC-7 r-d,f3,l

Pachyrhizus erosus (= P. tuberosus) Fabaceae jicama, yambean I-p/a,v 2n = 22 y-

0.6 tonne/ha USA-S-9 r-d,d2,f3,g,l

Panicum miliaceum Poaceae proso millet 1-a,g 2n = 36,54,72 b-i f-h x-more
than 10% by wind d-400m y2-(millet, "regular")30 lb/100 sq.ft.max. USA-
NC-7 r-d,f,f3,j,l

Panicum obtusum vine mesquite 1-p,g 2n = 20,36,40 USA-S-9 r-d,f3,l

Panicum virgatum switchgrass 1-p,g 2n = 18-108 b-o f-h may be apomictic v-

wind d-90m USA-NC-7 r-d,f,1

Papaver bracteatum Papaveraceae scarlet poppy 1-p,h 2n = 14 USA-W-6 r-d,l

Papaver somniferum opium or breadseed poppy 1-a,h 2n = 22,20 USA-W-6 r-

d,f3,l

Passf1ora spp. Passifloraceae passionfruit, granadilla, waterlemon, barbadine I-

p,v 2n = 18 USA-CR-MIA r-d,f3,1

Pastinaca sativa Apiaceae parsnip 1-b/a,h 2n = 22 b-o f-h flower protandrous v-

 30% by bees, flies, insects (Listed as b-i with x-30% by r-f.) d-500m Must be
isolated from commercial crops because a small fraction will bolt in the first
year. y-1,000 kg/ha; 2,000 max. USA-NC-7 r-d,f,f3,g,l

Pennisetum americanum (= P. glaucum) Poaceae pearl millet 1-a,g 2n = 14 b-o f-

h s-i,c y2-18.3 lb/100 sq.ft.max USA-S-9 r-d,f,f3,j,l

Pennisetum clandestinum kikuyugrass 1-p,g 2n = 36 USA-S-9 r-d,l

Pennisetum purpureum elephant grass 1-p,g 2n = 28,27,56 USA-S-9 r-

d,1,Wagoner 1990

Peril la frutescens Lamiaceae perilla 1-a,h 2n = 38,40 USA-NC-7 r-d,f3,1

Persea americana Lauraceae avocado 1-p,t 2n = 24 USA-NC-7 r-d,f3,l

Persea spp. avocado b-o f-h plant protogynous s-i v-bees r-f,f3

Petasites japonicus Asteraceae fuki, huki, sweet coltsfoot 1-p,h 2n = about 87(r-
z) 2n = 87(r-d3) r-d3,f3,z

Petroselinum crispum Apiaceae Hortense group = leaf parsley; Radicosum group

= root parsley (Hamburg)

Petroselinum crispum var. Neapolitanum (leaf) parsley 1-b,p,h 2n = 22 b-o f-h v-

honeybees, flies d-500-1,000m y-800 kg/ha USA-NC-7 r-f,f3,g,j,l,r

Phaseolus acutifolius Fabaceae tepary bean 1-a,h 2n = 22 b-i presumably (r-d2)

USA-W-6 r-d,d2,f3,1
Phaseolus atropurpureus see Macroptilium atropurpureum
Phaseolus aureus see Vigna radiata

Phaseolus coccineus (= P. multiflorus) (scarlet) runner bean 1-p/a,h 2n = 22 b-o

(r-f); i (r-g) f-h x-more than 30% (r-U; less than 40% (r-g) s-c v-honeybees,
bumblebees (required for self-or crosspollination) d-100m; more for stock
seed (r-g) y-1,000 kg/ha USA-W-6 r-d,d2,f,f3,g,l

Phaseolus lunatus lima bean I-p,p/a,b/a,h,v 2n = 22 b-i f-h x-0-80% by bees (r-f);
less than 18% (r-d2) d-45m y-1 tonne/ha y2-(bush)17.8 lb/100 sq.ft.max. y2-
(pole)22.3 lb/100 sq.ft.max. USA-W-6 r-d,d2,f,f3,g,l

Phaseolus multilorus see P. coccineus

Phaseolus mungo see Vigna mungo

Phaseolus vulgaris green, snap, string, common, or dry bean 1-a,h (some wild
forms i-p) 2n = 22 b-i f-h protogynous x-0-1% in many locations; 6-10% in
some places; 1520% in Puerto Rico; often high outcrossing in the tropics.
Thrips cause little outcrossing. Carpenter bees are responsible for high
outcrossing in Puerto Rico. d-45m (r-f); 50150m (r-g); see Appendix B. y-
1,500 kg/ha; 2,000 max. y2-(dry beans: kidney, pinto, red, or white)24 lb/100
sq.ft.max. y2-(bush green beans)17 lb/100 sq.ft.max. y2-(pole green beans)
29.7 lb/100 sq.ft.max. USA-W-6 r-b,d,d2,f,f2,f3,j,l,s2 Also see pp. 221-228

Phoenix dactylifera Arecaceae date palm 1-p,t 2n = 28,36 b-o f-d v-wind r-d,f,f3

Phragmites australis Poaceae reed 1-p,g 2n = 48,36,54 USA-S-9 r-d,f3,1

Phyllostachys dulcis Poaceae sweetshoot or vegetable bamboo 1-p,g r-

f3,Northern Groves catalog

Phyllostachys spp. phyllostachys 1-p,g 2n = 48,54 USA-PIO r-d,f3,l

Physalis ixocarpa Solanaceae tomatillo 1-p/a,h 2n = 24 USA-NE-9 r-d,f3,l

Physalis peruviana Peruvian ground cherry, husk tomato 1-p/a,s 2n = 24,48

USA-NE-9 r-d,f3

Pimpinella anisum Apiaceae anise 1-a,h 2n = 18,20 USA-NC-7 r-d,f3,1

Pinus edulis Pinaceae silver pine 1-p,t 2n = 24 r-d,f3

Pinus quadrifolia parry pinyon 1-p,t r-d,f3

Pinus spp. pine b-o f-m v-wind USA-S-9 r-f,l

Pistacia vera Anacardiaceae pistachio 1-p,t 2n = 30 b-o f-d v-wind USA-CR-

DAV r-d,f,f3,1

Pisum arvense see P. sativum

Pisum sativum Fabaceae pea 1-a,h 2n = 14 b-i f-h x-usually less than 1% in the
 U.S., (r-b) but some cvs. up to 25%; depends on what insects are around d-0
d2-20-100m,

but very short in some countries and mainly to prevent physical mixing y-2,000
kg/ha y2-(bush)21.6 lb/100 sq.ft.max. y2-(pole)12.1 lb/ 100 sq. ft. max. USA-
NE-9 r-b,d,f,f2,f3,g,j,l Also see pp. 216-221

Plantago coronopus Plantaginaceae buckhorn plantain 1-p,h 2n = 10,11,12 r-

d3,f3

Plantago major 2n = 12,24 r-d,d3,f3

Plantago maritima sea plantain 2n = 12,(18),24 r-d3,f3

Plantago ovata large round-leaf plantain 1-p,h 2n = 8 USA-W-6 r-d3,f3

Plantago psyllium psyllium seed 2n = 12 r-d3,f3

Plectranthus esculentus Lamiaceae kaffir potato 1-p,h 2n = 24 r-d,f3

Polianthes tuberosa Agavaceae tuberose 1-p,h 2n = 60,50 r-d,f3

Polymnia sonchifolia Asteraceae yacon, Ilacon I-p,h 2n = 60 USA-S-9 r-f3,l,z,

Lost Crops of the Incas 1989

Populus spp. Salicaceae poplar 1-p,t b-o fd v-wind USA-NC-7 r-d,f,f3,l

Portulaca oleracea Portulacaceae purslane 1-a,h 2n = 54 USA-S-9 r-d,f3,j,l

Prunus spp. Rosaceae almond, apricot, cherry, nectarine, peach, plum b-o f-h s-
c,i v-bees USA-CR-DAV r-d,f,f3,l

Psidium guajava Myrtaceae guava I-p,s,t 2n = 22,33,44 b-o f-h x-35% s-c v-bees
USA-CR-MAY r-d,f,f3,l

Psophocarpus tetragonolobus Fabaceae winged bean, asparagus pea 1-p,p/a,h,v

2n = 26 y-1.l tonnes/ha USA-S-9 r-d,d2,f3,g,l

Pueraria lobata Fabaceae kudzu I-p,h 2n = 24 b-o v-bees USA-S-9 r-d,d2,f3,j,l

Pueraria phaseoloides tropical kudzu 1-p,v 2n = 22 r-d,f3

Punica granatum Punicaceae pomegranate 1p,s,t 2n = 16,18,19 USA-CR-DAV r-

d,f3,1

Pyrus spp. Rosaceae pear 1-p,t b-o f-h s-c,i v-bees USA-CR-COR r-d,f,f3,l

Quercus spp. Fagaceae oak 1-p,t b-o f-m protandrous or protogynous v-wind
USAS-9 r-d,f,f3,1

Raphanus sativus Brassicaceae garden radish 1-a,b,h 2n = 18 gf-RR b-o f-h x-

more than 85% s-c,i v-bees, insects d-2001,000m y-1,000 kg/ha; 2,000 max.
y2-20.6 lb/100 sq.ft.max. USA-NE-9 r-f,f2, f3,g,j,l

Rheum spp. Polygonaceae rhubarb 1-p,h 2n = 22,44 USA-W-6 r-d,f3,j,l,z

Ribes hirtellum Grossulariaceae hairy gooseberry 1-p,s 2n = 30 USA-CR-COR

r-d,f3,l

Ribes nigrum black currant 1-p,s 2n = 16 USA-CR-COR r-d,f3,l

Ribes rubrum red currant 1-p,s 2n = 16 USA-CR-COR r-d,l

Ribes sativum common red ribes 1-p,s 2n = 16 r-d,f3,l

Ribes uva-crispa European gooseberry I-p,s 2n = 16 USA-CR-COR r-d,l

Rorippa microphylla (= Naturtium microphyllum) watercress 2n = 64 s-c r-r

Rorippa nasturtium -aqua ticum (= Nasturtium officinale) Brassicaceae

watercress, green or summer watercress 1-p,h 2n = 32 Can cross-or self-
pollinate. s-c v-flies, insects Produces fertile seed. r-d,g,r

Rorippa nasturtiumaquaticum x microphylla brown watercress 2n = 48 Sterile

triploid resulting from a cross of the diploid and tetraploid species. r-r

Rosa spp. Rosaceae rose 1-p,s b-o f-h s-i USA-NC-7 r-d,f,l

Rosmarinus officinalis Lamiaceae rosemary 1-p,s 2n = 24 b-o v-bees r-d,f,f3

Rubus idaeus var. strigosus Rosaceae American raspberry 1-p,s 2n = 14,21 USA-
CR-COR r-d,l

Rubus occidentalis black raspberry 1-p,s 2n = 14 USA-CR-COR r-d,l

Rubus occidentalis x idaeus purple raspberry 1-P's

Rubus spp. blackberry 1-p,h b-o f-m s-i v-bees USA-CR-COR r-d,f,l

Rubus spp. dewberry 1-p,h USA-CR-COR r-d,l

Rubus spp. raspberry 1-p,s b-o f-h s-c v-bees USA-CR-COR r-d,f,l

Rumex acetosa Polygonaceae sorrel, French sorrel 1-p,h 2n = 12 + XX, females;

12 + XY1Y2, males f-d r-d3,f3

Ruta graveolens Rutaceae common rue 1-p,h,s 2n = 72,81 r-d

Saccharum officinarum Poaceae sugarcane 1p,g 2n = 80,60,90 USA-CR-MIA r-

d,f3,1

Salix spp. Salicaceae willow 1-p,t b-o f-d v-insects USA-NC-7 r-d,f,f3,1

Salvia hispanica Lamiaceae chia 1-a,h USA-W-6 r-d,f3,l

Salvia officinalis sage 1-p,h,s 2n = 14,16 USA-W-6 r-d,f3,l

Salvia sclarea clary 1-b,p,h 2n = 22 USA-W-6 r-d,f3,l

Sambucus canadensis Caprifoliaceae American elder, Canadian elderberry 1-p,s

2n = 36,37,38 USA-CR-COR r-d,f3,l

Sambucus glauca blueberry, elderberry 1-p,s,t 2n = 38 r-d,f3

Sanguisorba minor Rosaceae burnet, salad burnet 1-p,h 2n = 28,54,56 USA-W-6

r-d,f3,1

Sansevieria trifasciata Agavaceae snake plant I-p,h 2n = 36,40,42 r-d,f3

Satureja hortensis Euphorbiaceae summer savory 1-a,h 2n = 45-48 USA-W-6 r-

 d,f3,l

Satureja montana winter savory 1-p,s 2n = 12,30 r-d,f3

Schinus molle Anacardiaceae California peppertree 1-p,s,t 2n = 28,30 r-d,f3

Sclerocarya caffra Anacardiaceae marula nut 1-p,t r-d,f3

Scolymus hispanicus Asteraceae golden thistle, Spanish oyster plant 2n = 20 r-

f3,z

Scorzonera hispanica Asteraceae scorzonera, black salsify 1-p/a,h 2n = 14 r-d,

f3,z

Secale cereale Poaceae rye 1-a,g 2n = 14-29 b-o f-h s-i Does not display strong
inbreeding depression (r-Allard 1960). vwind d-200m y2-24 lb/100 sq.ft.max.
USA-NSGC r-d,f,f3,j,l,Allard 1960

Secale montanum wild perennial relative that can be crossed with rye (S.
cereale) 1-p,g r-Wagoner 1990

Sechium edule Cucurbitaceae chayote, vegetable pear 1-p,v 2n = 28(r-s2) 2n =

24(r-d) b-o f-m r-d,f3,s2(p306)

Sesamum alatum Pedaliaceae wing-sesame 2n = 26 USA-S-9 r-d,f3,l

Sesamum indicum sesame I-a,h 2n = 26,52,58 b-i f-h protandrous x-5-65%

depending on cv., by bees d-180360m y2-5.6+ lb/100 sq.ft.max. USA-S-9 r-
d,f3,j,1

Sesamum radiatum wild sesame 1-p,h 2n = 64 USA-S-9 r-d,f3,1

Sesbania bispinosa Fabaceae agati, canicha 1-a,s 2n = 12,24 b-i d-0 r-d,d2,f3,1

Sesbania exaltata hemp, sesbania, Colorado river hemp 1-p,a,h 2n = 12 r-d,d2

Seseli carvi see Carum carvi

Sicana odorifera Cucurbitaceae casabanana 1-p,v USA-S-9 r-d,f3,l

Silene cucubalus Caryophyllaceae bladder campion 1-p,h USA-W-6 r-f3,1

Silybum marianum (= Carduus marianum) Asteraceae St. Mary's thistle, milk

thistle r-f3,1

Simarouba glauca Simmondsiaceae aceituna 1-p,s,t r-d,f,f3

Simmondsia chinensis (= S. californica) Sim

mondsiaceae jojoba 1-p,s,t 2n = 56,about 100 b-o f-d USA-W-6 r-d,f3,l

Sinapsis alba see Brassica hirta
Sinapsis arvensis see Brassica haber

Sium sisarum Apiaceae skirret 1-p,h 2n = 22 USA-NC-7 r-d3,f3,1

Smyrnium olustrum Apiaceae Alexander's salad greens, black lovage 1-b,p,h 2n

= 22 r-f3,1

Solanum aethiopicum Solanaceae mock tomato 1-p,s 2n = 24 USA-S-9 r-d,f3,l

Solanum aviculare Australian nightshade 1-p,s,t 2n = 46,48,92 r-d,f3,l

Solanum burbankii Mrs. B's nonbitter garden huckleberry 1-a,h USA-S-9 r-f3,1

Solanumfendleri Navajo, fendler, or wild potato 2n = 48 USA-IR-1 r-d,f3,l

Solanumferox ram-begun 1-p,h 2n = 24 USA-S-9 r-d,l

Solanum gilo gilo 1-p,s 2n = 24 USA-S-9 r-d,l

Solanum hyporhodium cocona 1-p,s USA-S-9 r-d,l

Solanum incanum Sodom apple 2n = 24 USA-S-9 r-d,f3,1

Solanum indicum Indian nightshade 2n = 24 r-d,f3

Solanum hhasianum 2n = 24 USA-S-9 r-d,l

Solanum laciniatum kangaroo apple I-p/a,s,t 2n = 48,92 USA-S-9 r-d,f3,l

Solanum macrocarpon native eggplant 1-h,s 2n = 36 USA-S-9 r-d,f3,1

Solanum melanocerasum garden huckleberry, wonderberry, sunberry 1-a,h USA-

S-9 r-f3,1

Solanum melongena eggplant 1-p/a,h 2n = 24,36,48 b-i f-h x-7% by bees,

insects; 6-20% in India at distances less than 50m; no crosspollination at
greater distances y-150 kg/ha; 200 max. y2-0.6 lb/100 sq.ft.max. USA-S-9 r-
d,f, f3,g,j,l

Solanum muricatum melon-pear 1-p,s 2n = 24 USA-S-9 r-d,f3,1

Solanum nigrum black nightshade 1-a,h 2n = 24,36,48 USA-S-9 r-d,f3,1

Solanum quitoense naranjilla, lulo I-p,h,s 2n = 24 USA-S-9 r-d,f3,1

Solanum torvum terongan 2n = 24 r-d

Solanum tuberosum potato 1-a,h 2n = 24,36,48 b-i f-h Many cvs. produce
nonfunctional pollen. USA-IR-1 r-d,f,f3,j,l

Solenostemon rotundifolius Lamiaceae Hausa potato I-p,h 2n = about 84 r-d,f3

Sorghum bicolor Poaceae sorghum 1-a,g 2n = 20 USA-S-9 r-d,l

Sorghum halepense johnsongrass 1-p,g 2n = 20,40 USA-S-9 The Land Institute

has a program of crossing and breeding with S. halepense and S. bicolor in an
effort to develop perennial, winter-hardy sorghum. r-d,l,The Land Report
Spring 1991,Wagoner 1990

Sorghum sudanense Sudan grass 1-a,g 2n = 20 r-d,l

Sorghum vulgare sorghum b-i f-h d-300m; 400m from S. sudanense y2-25
lb/100 sq.ft.max. r-f,j

Sorghum vulgare var. sudanensis Sudan grass 1-g b-i f-h d-0 r-d,f

Sorghum x almum almum sorghum 1-p,g 2n = 40 USA-S-9 r-d,l

Sphenostylis stenocarpa Fabaceae yam bean, yam pea l-p,p/a,h 2n = 18 USA-S-9

r-d,d2,f3,l

Spilanthes acmella Asteraceae toothache plant, para cress 1-a,h 2n = 14,24,52 r-

d,f3

Spinacia oleracea Chenopodiaceae spinach I-a,h 2n = 12 b-o f-d Populations

include male plants, female plants, and monoecious hermaphrodite plants. v-
wind d2-500-1,000m y-800 kg/ha; 2,000 max. y2-10.8 lb/ 100 sq. ft. max.
USA-NC-7 r-d,f,f3,g,j,l,r

Spondias mombin Anacardiaceae hog plum Ip,t 2n = 32 r-d,f3,l

Spondias purpurea red mombin l-p,t r-d,f3,1

Sporobolus airoides Poaceae alkali sacaton 1-p,g 2n = 84-126 USA-W-6 r-d,f3,l

Sporobolus cryptandrus sand dropseed 1-p,g 2n = 18,36,72 r-d,f3,1

Stachys affinis (= S. sieboldii) Labiatae Chinese artichoke 1-p,h 2n = ? USA-

NE9 r-d,f3,1,z

Stevia rebaudiana Asteraceae kaa he'e 1-p,s 2n = 22 USA-CR-MIA r-d,f3,l

Syagrus coronata Erecaceae ouricury palm I-p,t 2n = 32 r-d,f3

Symphytum asperum (= S. asperrimum) Boriginaceae prickly comfrey 2n = 40 r-

d,z

Symphytum officinale common comfrey 2n = 26,about 36,40,about 40,48 USA-

NC7 r-d,f3,l,z

Symphytum peregrinum see Symphytum x uplandicum

Symphytum x uplandicum Quaker comfrey 1-p,h,s 2n = 36 USA-NC-7 r-d, f3,l,z

Synsepalum dulcificum Sapotaceae miracle fruit, miraculous berry 1-p,s,t r-

d,f3,1

Tamarindus indica Fabaceae tamarind 1-p,t 2n = 24 r-d,f3,1

Tanacetum vulgare Asteraceae tansy l-p,h 2n = 18 r-d,f3,1

Taraxacum hoh-saghyz Asteraceae Russian dandelion 1-p,h 2n = 16 r-d

Taraxacum officinale dandelion 1-p,h 2n = 8,24,36(r-d) 2n = 16,24,46(r-r) May

be apomictic. r-d,f3,j,l,r

Telfairia occidentalis Cucurbitaceae oyster nut tree I-p,l r-d,f3

Tephrosia candida Fabaceae white tephrosia 1-p,s 2n = 22 USA-S-9 r-d,l

Tephrosia vogelii vogel tephrosia 1-p,s 2n = 22 USA-S-9 r-d,l

Terminalia catappa Combretaceae tropical almond 1-p,t 2n = 24 USA-CR-
MIA r-d,f3,1

Tetragonia tetragonioides Aizoaceae New Zealand spinach 1-a,h 2n = 32 y2-17.2

lb/ 100 sq.ft.max. USA-S-9 r-d,f3,j,l

Tetragonolobus purpureus Fabaceae asparagus pea, winged pea I-a,h 2n = 14

USA-NC7 r-d,d2,f3,1

Thaumatococcus daniellii Marantaceae miracle fruit 1-h 2n = 20 r-d

Theobroma bicolor Sterculiaceae bacao 1-p,s,t 2n = 20 r-d,f3

Theobroma cacao cacao 1-p,t 2n = 16,20,26 b-o f-h x-more than 30% s-i,c v-
midges, ants r-d,f,f3

Theobroma grandiflorum cupuacu 1-p,t r-d,f3

Thymus serphyllum Lamiaceae creeping thyme l-p,s 2n = 20,24 r-d,f3

Thymus sp. USA-NC-7 r-l

Thymus vulgaris common thyme 1-p,h 2n = 30 r-d,f3

Tragopogon porrifolius Asteraceae salsify, oyster plant 1-b/a,h 2n = 12 y2-27.7

lb/100 sq.ft.max. USA-W-6 r-d,f3,j,l

Trapa natans Trapaceae caltrop, water chestnut 1-a,p,h 2n = 36,40,48 r-d,f3,1

Treculia africana Moraceae muzinda 1-p,t r-d,f3

Trichosanthes cucumerina (= T anquina) Cucurbitaceae snake gourd, club gourd

1-a,v 2n = 22 b-o f-m USA-NE-9 r-d,f3,l,s2(p306)

Trifolium hybridum Fabaceae alsike clover 1-p,p/b,h 2n = 16 b-o f-h x-more

than 90% s-mostly i v-bees USA-NE-9 r-d,d2,f3,j,l

Trifolium incarnatum crimson clover 1-a,h 2n = 14,16 v-bees necessary for

selfor crosspollination USA-S-9 r-d,d2, f3,j
Trifolium pratense red clover 1-b,p,h 2n = 14,28 b-o f-h x-more than 90% s-
mostly i v-bees USA-NE-9 r-d, f,f3,j,l

Trifolium repens white clover 1-p,h 2n = 32,48,64 b-o f-h x-more than 90% s-
mostly i v-bees USA-NE-9 r-d,f3,j,l

Trifolium spp. b-i,o f-h About 30% of species are s-i and v-bees, and some are
cleistogamous. About 70% are b-i. Small-flowered species are often b-i; large-
flowered species, b-o and v-bees. r-f2

Trifolium spp. y2-(sweet, hubam clover)2.2 + lb/100 sq.ft.max. y2-(timothy

clover)0.46+ lb/100 sq. ft. max. r-j

Trifolium spp. See r-f3 for uses of additional species. See r-d2 for breeding
systems of additional species.

Trigonella foenum-graecum Fabaceae fenugreek 1-a,h 2n = 16 USA-W-6 r-

d,d2,f3,1

Triphasia trifolia Rutaceae limeberry I-p,s 2n = 18 r-d,f3

Tripsacum dactyloides Poaceae eastern gamagrass (warm-season bunchgrass) 1-

p,g 2n = 36,72 Wild perennial grain with large, edible seeds. Efforts to
domesticate it into a perennial grain crop are under way at The Land Institute.
For review, see Wagoner 1990. USA-NC-7 r-d3,1,The Land Report Spring
1991,Wagoner 1990

Triticosecale spp. (Triticum x secale?) Poaceae triticale 1-a,g 2n = 42,56 r-d

Triticum aestivum Poaceae wheat 1-a,g 2n = 42 USA-NSGC r-d,l

Triticum carthlicum Persian wheat 1-a,g 2n = 28 USA-NSGC r-d,l

Triticum compactum club wheat 1-a,g 2n = 42 USA-NSGC r-d,1

Triticum dicoccon emmer 1-a,g 2n = 28 USA-NSGC r-d,l

Triticum durum durum wheat 1-a,g 2n = 28 USA-NSGC r-d,l

Triticum monococcum einkorn 1-a,g 2n = 14 USA-NSGC r-d,l

Triticum polonicum Polish wheat 1-a,g 2n = 28 USA-NSGC r-d,l

Triticum spelta spelt 1-a,g 2n = 42 USA-NSGC r-d,l

Triticum spp. triticale and wheat b-i f-h largely cleistogamous x-less than 6% by
wind d-0 USA-NSGC d-f,l

Triticum spp. wheat y2-(durum wheat)26 lb/ 100 sq.ft.max. y2-(Early Stone Age
wheat)17+ lb/100 sq.ft.max. y2-(hard red spring, red winter, or white wheat)26
lb/100 sq.ft.max. USA-NSGC r-j,1

Triticum spp. wheat hybrids f-h + female (male sterile) d-200m USA-NSGC d-
f,l

Triticum spp. See r-f3 for additional information.

Triticum timopheevii timopheev wheat 1-a,g 2n = 28 USA-NSGC r-d,l

Triticum turanicum Oriental wheat 1-a,g 2n = 28 USA-NSGC r-d,l

Triticum turgidum poulard wheat 1-a,g 2n = 28,42 USA-NSGC r-d,l

Tropaeolum majus Tropaeolaceae nasturtium 1-a,h 2n = 28 USA-S-9 r-d3,f3,1

Tropaeolum minus climbing or tall nasturtium I-a,p?,h 2n = 28 r-d,d3,f3

Tropaeolum tuberosum anu, mashua, isano 1-p,h 2n = 42 r-d,d3,f3,Lost Crops of

the Incas 1989

Urena lobata Malvaceae ullucu 1-p,h 2n = 24,36 USA-S-9 r-d,f3,l

Urtica dioica Urticaceae stinging nettle I-p,h 2n = 32,48,52 r-d,f3

Vaccinium angustifolium Ericaceae lowbush blueberry 1-p,s 2n = 24,48 USA-

CR-COR r-d,f3,1

Vaccinium ashei rabbiteye blueberry 1-p,s 2n = 72 USA-CR-COR r-d,f3,1

Vaccinium corymbosum highbush blueberry 1-p,s 2n = 48,72 USA-CR-COR r-
d,f3,1

Vaccinium macrocarpon cranberry 1-p,l 2n = 24 USA-CR-COR r-d,f3,1

Vaccinium vitis-idaea mountain cranberry 1p,s,h,v 2n = 24 USA-CR-COR r-

d,f3,1

Valerianella locusta (= V. olitoria) Valerianaceae cornsalad, mache 1-a,h 2n = 14

rd,d3,f3

Vicia faba (= V. Java) Fabaceae fava or faba bean; broad, bell, horse, tick, or
field bean 1-a,h 2n = 12 b-i,o f-h x-25-50% by bees (r-f2) d2-most people say
300m, but more recent information says 1,000m. y-1,500-2,000 kg/ha y2-
("bell bean") 4.5 lb/100 sq.ft.max. USA-W-6 r-d,d2, f,f2,f3,g,l,s2

Vicia monantha bard vetch 1-a,h 2n = 14 USA-S-9 r-d,d2,l

Vicia pannonica Hungarian vetch I-a,h 2n = 12 b-i bees required f-h d-3m USA-
S-9 r-d,d2,f,1

Vicia sativa vetch 1-a,h,v 2n = 10 b-i bees required f-h d-3m USA-W-6 r-d,d2,f,l

Vicia sativa ssp. nigra blackpod vetch 1-a,v 2n = 12 USA-W-6 r-d,l

Vicia spp. For uses see r-d2,f3.

Vicia villosa winter/hairy vetch 1-a,b,p,h,v 2n = 14 v-bees y2-1.1+ lb/100

sq.ft.max. USA-S-9 r-d,d2,j,1

Vigna aconitifolia Fabaceae moth bean 1-a,h 2n = 22 s-c USA-S-9 r-d,d2,f3,1

Vigna angularis adzuki bean 1-a,h 2n = 22 x-common s-c USA-S-9 r-d,d2,f3,1

Vigna mungo (= Phaseolus mungo) urd bean, black gram I-a,h 2n = 22,24 b-i f-h
up to 42% cleistogamous USA-S-9 r-d,d2,f,f3,1

Vigna radiata (= Phaseolus aureus = Phaseolus mungo) mung bean, green or

golden gram 1-a,h 2n = 22 b-i f-h d-0 y-0.5 tonne/ ha y2-24 lb/100 sq.ft.max.
 USA-S-9 r-d,d2,f,f3,j,l

Vigna umbellata rice bean 1-a,h 2n = 22 s-c USA-S-9 r-d,d2,f3,1

Vigna unguiculata cowpea, yard-long bean 1-a,h 2n = 22 y-0.7 tonne/ha y2-24

lb/100 sq.ft.max. USA-S-9 r-d,d2,f3,g,j,l

Vigna unguiculata ssp. cylindrica catjang, cowpea I-a,h 2n = 22 USA-S-9 r-

d,d2,f3,1

Vigna unguiculata ssp. sesquipedalis yard-long bean, asparagus bean 1-a,p,h 2n

= 22 b-i x-10-15% by ants, flies, bees v-heavy insects USA-S-9 r-d,d2,f3,1

Vigna vexillata aka sasage, zombi pea 1-p,h,v 2n = 22 USA-S-9 r-d,d2,f3,1

Vitis labrusca Vitaceae fox grape 1-p,l,v 2n = 38 USA-GD r-d,f3,l

Vitis rotundifolia muscadine grape I-p,l,v 2n = 40 b-o f-d v-insects USA-CR-

DAV r-d,f,f3,1

Vitis vinifera wine grape 1-p,l,v 2n = 38,40,57 b-i f-h s-some cvs. partly or fully
s-i USA-GD r-d,f,f3,l
Voandezia poissonii see Macrotyloma geocarpum

Voandzeia subterranea Fabaceae bambara groundnut 1-a,h 2n = 22 b-i,o v-ants y-

0.6 tonne/ha r-d,d2,f3,g,l

Wasabia japonica (= W. pungens = Eutrema wasabi) Brassicaceae wasabi,

Japanese horseradish I-p,h 2n = 28 b-o f-h vflies? r-f3,1,pc(Dan Borman),z

Xanthosoma sagittifolium Araceae yautia 1-p,h 2n = 24,26(r-d) 2n = 26(r-s2) r-

d,f3,l,s2

Xanthosoma spp. zanier, yautia, cocoyam 2n = 26 r-s2

Yucca elephantipes Agavaceae ozote 1-p,t r-d,f3,l

Zea diploperennis Poaceae diploid perennial teosinte I-p,g 2n = 20 Can be

crossed with Z. mays to produce fertile hybrids. USA-NC-7 r-f3,1,Wagoner
1990

Zea mays Will cross with wild perennials such as Z. perennis, Z. diploperennis,
and Tripsacum spp. r-Wagoner 1990

Zea mays ssp. mays corn, maize 1-a,g 2n = 20 b-o f-m Often somewhat
protandrous. Displays severe inbreeding depression (r-Allard 1960). v-wind d-
200m d21 km y-1,500 kg/ha; 2,500 max. y2-(sweet corn)10.3 lb/ 100 sq. ft.
max. y2-(fodder corn)22.6 lb/100 sq.ft.max. USA-NC-7 r-b,d,f3,g,j,l,Allard
1960

Zea mays ssp. mexicana teosinte I-a,g 2n = 20 USA-NC-7 r-j,l

Zea perennis perennial teosinte 2n = 40 USA-NC-7 r-f3,l,Wagoner 1990

Zingiber mioga Zingiberaceae temperate ginger, Japanese wild ginger I-p,h 2n =

55 r-d3,pc(Alan Kapuler)

Zingiber officinale ginger 1-b/a,p/a,h 2n = 22,24 r-d,f3,l

Zizania aquatica Poaceae Indian, Tuscarora, or wild rice 1-g 2n = 30 r-f3,1

Zizania palustris northern wild rice I-a,g 2n = 30 r-d,l

Ziziphusjujuba Rhamnaceae common jujube, Chinese date I-p,s,t 2n = 48 USA-

CRDAV r-d,f3,l

Ziziphus mauritiana Indian jujube USA-CRMIA r-d,f3,l


 NE WALL of my living room is lined with shelves full of jars of my
own seed. No wall decoration could provide more beauty, comfort, and security.
Nothing else is such a good conversation piece. Here is a touchstone, a shrine to
what is essential. Here are memories of past seasons and accomplishments, and
dreams for those to come. Here is a symbol that this is, indeed, hearth and home
and homestead. Here is hope for the future.

Why Save Seeds?

Saving seeds is fun. Cleaning the seed, holding the clean seed in your hands, is
magical. Gaze at the seed, run your fingers through it, play with it, and you can
feel the connections.

You're like a child with a gallon bucket of marbles, or a squirrel sitting on a

hollow log full of acorns. Unquenchable joy arises. It is so intense it puzzles you
initially. Then you recognize it. It is the joy that comes from being who you are
supposed to be and doing what you are meant to do.

Seed saving is practical. If you know how to save your own seeds you can grow
rare varieties. Many of the most spectacularly flavorful, unique varieties are not
readily available commercially, either as fruits or seed. One of my favorite
winter squash is `Blue Banana', for example. This squash has a flavor that is
superb, intense, and so different from all other squash that it is like an entirely
different vegetable. But the seed is not available commercially. To grow rare
varieties, you often have to get the seed when and where it is available, then
maintain the variety yourself.

Many superb varieties are not readily available commercially because they
have narrow adaptations to particular regions. `Narragansett Indian Flint', for
example, is said to be the corn that was given to the Pilgrims by the Indians, the
corn that made the jonnycakes that were eaten at the first Thanksgiving.
However, `Narragansett Indian Flint' (also known as `Rhode Island White Cap')
has a narrow ecological adaptation. It likes the quasi-Mediterranean climate of
Southern New England. It is too long-season for Northern New England, and too
heatintolerant for the Southeast or Midwest. It is not widely enough adapted nor
popular enough to warrant its production as a seed crop via normal channels.

`Narragansett' makes wonderful jonnycakes - the very best, I'm told. I don't
know if they are the very best possible, but I've tried them and they are truly
delicious. `Narragansett' also makes rich, full-bodied cereal and polenta. But you
can't buy the seed from commercial suppliers. If you garden in the Midwest, of
course, you don't care, because you can't grow the corn anyway. But given our
current system for mass-producing food and seed crops, the fact that you can't
grow the corn in the Midwest now means that you can't grow it in southern New
England, either - unless you can save your own seed.

Some varieties are not available because they have peculiarities with respect to
production of the seed itself. If a watermelon produces few seeds, for example, it
will not

usually be offered commercially. It's simply too expensive to produce the seed. A
home gardener, though, might be happy to save such seed. And a market garden
might be able to easily produce the handful of seeds needed for a single field's
planting.

Being dependent upon seed companies for your seed means being dependent
upon random fads in foods as well as other people's choices and preferences.
Saving your own seed means independence. It lets you make your own choices
and have your own preferences.

When you save your own seed, the seed is always "available." It is common
these days for all the seed of even very popular varieties to be produced by just a
single grower. If that grower experiences a crop failure, the seed isn't available
anywhere.

Sometimes, even if the seed is "available," you can't necessarily find it. There
can be a poor correlation between variety names and the material you actually
receive. Seed companies often change lines or suppliers, so that what they are
selling one year and the next may be different strains, even though they are
called the same thing.

I once grew a squash that the packet identified as `Red Kuri'. It was a scarlet
teardrop shape, delicious and uniquely flavored. The next time I ordered the seed
the squash was orange-and-green speckled and inferior in flavor. Somewhere,
somehow, the variety had been crossed up. And many companies were selling
the crossed-up seed.

I ordered a packet of seed from each of a half-dozen companies, growing a few

plants from a new company each year. Some companies were selling an inferior-
flavored orange thing as `Red Kuri'. I have yet to find that wonderful scarlet
variety again.

Maybe all the material I've tried in recent years is really something else, not
`Red Kuri' at all. Maybe the variety I liked so much in the first place was
something else. All I can say at this point is ... I wish I had saved the seed.

I like to produce my own seed even of varieties that are readily available
commercially. My own seed is usually bigger, fatter, and more vigorous. I can
plant it earlier than commercial seed. I also have much more of it, so I don't have
to skimp. I can sow generously and then thin, instead of sowing thinly, then
having gaps that have to be replanted later and less optimally.

And with my own seed, the price is right.

When you save seed, you become a plant breeder. You are choosing which
germplasm to perpetuate. This means that you are both deliberately as well as
automatically selecting for characteristics that are important to you, for plants
that are fine-tuned to your needs and growing conditions and region. After you
have saved seed of a variety for a few years, you have your own line of the
variety that is slightly different from anyone else's, and it is usually better
adapted to your needs.

Knowing how to save your own seed also means that you can take advantage
of genetic accidents, ideas, and dreams. Last year, for example, I noticed one
squash plant in perhaps a hundred that was resistant to powdery mildew. I saved
the seed from it. Perhaps I can use it to develop new powdery-mildewresistant
varieties. Powdery mildew after the first fall rains is what ends the squash grow
ing season in my region. Resistant varieties could be very useful. Many new
varieties got their start when some gardener or farmer simply noticed something
that was different and special - and saved the seeds.

We gardeners and farmers care about our direct relationship with soil, plants, and
food. To grow plants from seed bought from others is one level of relationship.
To grow plants from our own seed, to save seeds from our own plants, goes to a
deeper level. It is fulfillment and continuity - plants and people maintaining each
other, nurturing each other, evolving together. It completes the circle.

Saving Seed from Hybrids

Hybrids don't breed true to type from seed. Some hybrids are even sterile,
though most will produce seed. This seed can be used to derive a pure-breeding
variety by the methods described in chapters 9 and 10. Such a variety derived
from a hybrid is a new variety and should be given a new name. It is not the
same as the hybrid from which it was derived. In other words, you can save seed
from hybrids as the first step in creating a new open-pollinated variety, but you
cannot reproduce a hybrid by saving its seed.

This section on seed-saving practice, then, refers to pure-breeding, not hybrid,

varieties.

Seed-Saving Overview

Saving seed is easy. Plants want to make seed. They cooperate fully. To save
seed, all you have to do is let the plants produce seed, then grab it quick before
the birds or squirrels or bugs, and before it gets rained on and molds or sprouts
in the pod.

Saving seed of pure varieties is another thing entirely. Plants don't care at all
about pure varieties. The outbreeders would all rather cross with that strange
inedible ornamental variety down the street in the yard of your neighbor. Even
the inbreeders outcross far more often than they are "supposed to," especially
under organic growing conditions.

To save seed of pure varieties, we need to know something about the

outcrossing tendencies of the crop so that we can isolate it sufficiently from
other varieties or wild plants of the species that it could cross with.

Finally, every variety contains genetic variability. Some of this is desirable and
even essential to the vigor and adaptability of the variety. Some of it, though, is
undesirable. So, we need to grow an appropriate number of plants in order to
maintain the amount of genetic variability that we want. At the same time, we
must select and rogue to eliminate the genes associated with specific kinds of
variability that we don't want.

Given the genetic heterogeneity in most varieties and the greater vigor of the
more wild-type forms, the natural tendency of most varieties is to deteriorate
quickly to something that is far less useful to its human associates. To maintain a
variety we must actively breed in order to counter this tendency

There is actually no such thing as "saving" a pure variety. There is only further
breeding, either deliberate or accidental. We either select in order to hold the
variety in its current form and to eliminate undesirable types, or we select in
order to change the variety in some preferred direction. Both processes involve
exactly the same principles.

Roles and Purposes

"What's my role with respect to this variety?" That's the first thing I ask myself
about every seed-saving project. Am I the sole savior or creator of the variety,
the one person without whom it would be lost forever? Or is my line better than
everyone else's, and especially worthy of preserving and distributing?

Am I planning on building up the precious stock, then giving or selling it to

seed companies or others? Will I be distributing it through the Seed Savers
Exchange? Will many or even all future plantings of this variety all over the
country be descendants of these seeds I hold in my hands today? If so, I will
want to be pretty careful and rigorous. I will use serious numbers of plants, and
serious isolation distances.

Often, however, I'm saving seed just for myself, and I know others have the
variety as well. In that case, I can be quite casual about most nearly everything.
Numbers of plants? I grow what I need for the table, and use special tricks (see
Chapter 19) to deal with maintaining heterogeneity.
 Isolation? It's often minimal. I usually plant so as to be able to recognize
hybrids, which is much easier than avoiding them (see Chapter 18). If I can
recognize hybrids I can eliminate them or not as I choose in future generations.
Who knows? The hybrid might be more interesting than the original material.
And if the seed is just for my own use, what's an outcross or two among friends?

 T'S THE MIDDLE of winter. I am sitting indoors in my big easy chair
while it rains outside. I'm looking at seeds. I always visually examine all the seed
I'm planning to plant each year, and winter is the best time for it. I spread the
corn out in a tray in my lap. I remove any weed seeds and cull out any seeds that
are off-type - or make plans to plant them separately.

Yipe! What's that white flour corn kernel doing here in this lot of flint corn?
Obviously, that's what the seed company ran through the seed cleaning
equipment before the flint. Get that out of there! The `Magenta' corn is supposed
to be magenta. Out with any occasional white or black kernels!

Then there's the 25-pound bag of `Scarlet Runner' beans I bought just to eat.
The seed is smaller than I expected, and I suspect they

are a bush type. I'll plant some. I think I would like to cross a pure black seed
into a runner bush. And it would be nice if I could get the lavender color
separate. Different pure colors would probably have different flavors. Let's look
at all the beans, pull out some to plant, see if there is anything unusual.

So I am sorting through the 25 pounds of beans, one tray at a time. Never can
tell. Maybe ... ha! A black! I sort a bit more. More blacks. Some different shapes.
A nearly solid brown. I never even dared hope for that. I thought it should be
possible theoretically, but never saw one, not in the runner bean species. What a
find! Pure brown line, here we come!

I finish the sorting with 24'/4 pounds of beans left to eat and a half pint of
unusual types. I'll plant some of the ordinary beans, of course. But it is the "off'
types that are the most exciting. To my planting plans I add the "Mutant Ninja
Runner Bean Project."
 Preparation and Planning

The most important part of seed saving happens before you turn over the first
square foot of soil or plant the first seed. It's planning. Saving seed should be
part of every garden's yearly harvest. But it does require extra planning. In this
chapter I give general rules and suggestions about growing seed, a checklist to
review as you plan your plantings. Certain topics - isolation distances, numbers
of plants to grow, and selection - are only mentioned here, and are covered in
separate chapters.

If you are new to seed saving, the planning may seem complicated at first. It
introduces new sets of issues. With a bit of experience and familiarity, however,
you will take seed saving into account automatically, just as our ancestors did.
You'll begin to plan for seed saving just "naturally," and the planning is part of
the fun.

I do most of my garden and field planning during the winter and early spring. I
move all the seed I'm going to plant to a single big box, with the varieties of each
species stacked together. At a glance I can tell how many isolations I need, and
can figure out how to arrange them within the context of everything else I'm
going to plant. I can also tell at a glance whether my plans and dreams fit the
amount of land and time I have. (They never do initially; I always have to
remove some stuff from the box.)

Here are some general rules and considerations:

Visually inspect your planting stock. If the corn seed is supposed to be sweet,
cull out that field-type kernel! If it is supposed to be white, get rid of that yellow.

Even tiny seed is worth examining carefully. You can detect and eliminate
contaminating weed seeds.

Don't, however, eliminate variability in seed type arbitrarily. There is no reason

the mustard seed needs to be all one color.

If you're going to go to the trouble to save seed of something, it's very nice to
know what it is. Record what you are going to plant - exactly what you are going
to plant, including the exact source of the seed. If I bought the lot of seed from
Territorial in 1997, for example, "T97" will follow the variety name in my notes
as well as on the field marker. "98;T97" would represent seed from my own
growout in 1998 of seed I obtained originally from Territorial in 1997.

Remember that many very different lines are being sold under the same variety
name, and that seed companies may change suppliers from one year to the next.
You haven't identified anything very well with just a variety name. You have to
write down the source and the year.

Plan, organize, and label things in the field so you can tell which variety is
which. Give plants enough space so that you don't confuse different varieties.
The plants need the space anyway. Crowded plants may yield decent vegetables,
but they don't produce very good seed.

It's better to alternate varieties that are distinguishable. Separate the two green-
podded varieties of pole beans with a purple-or yellow-podded one, for example.
Or alternate species and accomplish both identification and isolation
simultaneously. Alternate sections of common beans (Phaseolus vulgaris) with
runner beans (Phaseolus coccineus), for example.

If you're going to do hand-pollinations with squash, thin to one plant per hill,
and plant hills far enough apart so that you can easily trace a vine to its source.

Rotate seed crops. This is useful for all crops as a means of minimizing pests
and disease, but it is critical for seed crops that are capable of volunteering. If
you plant squash and volunteers come up all over the patch, you'll be hard
pressed to know who anybody is in July (before fruits develop), when you need
to hand-pollinate flowers for breeding or seed saving. Plant the squash where
you did not have any squash last year.

Gophers, moles, and cats often rearrange plantings gratuitously. All the more
reason to try to choose varieties for each row and for adjacent rows that are
distinguishable. I still have squash seed from 1998 with labels like "98 big green
mystery 6-2." Gophers were especially active in the squash patch that year, and
they rearranged several sections of seeds before they germinated.

Then there was the year I entertained myself and the local squirrels by letting
them have all the unneeded squash seed as we ate the squashes during the winter.
Between the squash and the oilseed sunflower seeds I put out for birds, those
squirrels had a wonderful winter. I enjoyed watching them stuff their little
cheeks. It was amazing how much squash and sunflower seed they ran off with.

That spring, when I went down to the garden to plant some ornamental
sunflowers I was planning to continue breeding, I noticed that there were already
things that looked quite a lot like sunflowers coming up in the area I wanted to
plant. In fact - they were sunflowers. Oilseed sunflowers.

That year, oilseed sunflowers and cullsquash-seed squash came up all over my
garden and compost heap as well as those of my next-door neighbor. I had
forgotten something important. Humans are not the only creatures that plant
things. The squirrels planted all the excess seeds I gave them. They planted them
in all the soft, easy places to dig - gardens and compost piles.

All varieties outcross to some extent. The socalled outbreeders simply do more
of it than the socalled inbreeders. Furthermore, there is much more outcrossing if
you garden or farm using organic methods. Living land is full of pollinators of
many species.

Every variety will need isolation from other varieties as well as from wild
plants of the same species if you want to save seed and keep the variety pure.
How much isolation you need depends upon the specific variety, the situation,
and your goals and purposes with respect to the seed. Isolation is covered in
Chapter 18. Organizing your plantings so as to provide appropriate isolation is a
major part of the planning.

How many plants do you need to save seed from? This issue is covered in
Chapter 19.

In some cases you can eat and save seed from the same patch. In other cases
you have to choose one or the other. You can't both eat a green bean and save the
seed, for example. But you can eat and save seed from the same winter squash.

You'll often need to use one part of the patch for seed saving, or mark the
plants you want to save.

Are the needs of the seed crop going to be different from those of the crop if
you were just growing it as a vegetable? If plants of the seed crop must be
allowed to dry out, you may want to put the planting in a separate bed or at one
end of your garden. That way you can stop watering the seed crop when the
plants begin to dry without interfering with the watering of the rest of your
garden.

Seed crops are also good candidates for growing using drip irrigation methods.

Do you want adequate seed or the very best possible seed? In many cases,
adequate seed is all you really need. If you want to save the best possible seed,
you'll probably have to give the plants extra space, extra nutrients, and optimal
watering.

Almost all seed crops need to stay in the ground somewhat beyond eating
stage in order to produce seed. They often need more space too, because they are
going to be much bigger than regular vegetable plants before they're done
making seed. In many other cases, however, you grow the plants the same way
you would for just a vegetable crop, and simply use certain plants or a certain
section for seed saving.

Some plants, the biennials, produce seed only in their second year. So, if you
are growing most types of onions, carrots, or cabbage, for example, you will
probably need to lift and store plants during the winter, then replant them in
spring. See Suzanne Ashworth's book Seed to Seed for more information about
growing various biennials for seed.

Some crops are very easy to save seed from. Self-pollinating annuals such as
beans, peas, and tomatoes can be handled with very little more work than it takes
to grow the vegetable crop.

Unfortunately, you probably can't save seed of everything you might like to
grow. You can grow some things as vegetables that you can't do as a seed crop.
For example, any crop that is late-season for your area is unlikely to produce a
seed crop reliably, since the seed crop requires extra growing time. Even if you
can produce seed some years, the seed is likely to be inferior to what you can
buy commercially.

You can seed-save early and midseason sweet corns easily, for example. The
latest variety you can produce as a sweet corn crop will need to be maintained by
others who have a longer or warmer growing season.

Planting

The following are a few simple rules related to planting.

1. Inspect all seed before you plant it. Eliminate weed seeds. Remove off-types
and either eliminate them or plant them separately.

Remove any seed that has a bumpy skin or atypical blotches. Such seed is
usually from diseased plants.

Smaller or lighter seed is often from imma ture plants or pods, and doesn't
germinate as well as typical seed.

2. Never plant all the seed at once or in the same year if it is material you can't
afford to lose and can't replace.

3. Don't experiment on seed you can't replace. When the seed is rare or unique,
it is tempting to give it extra care and treat it specially. Resist the urge, and use
your normal methods. Try out any new methods on seed you don't care about.
Don't presoak seed for the first time with a rare lot of seed, for example.

4. Plant plenty of seed if you have it available, then thin to the desired number
of plants. That way you will be selecting for material that germinates vigorously
under your conditions. If you plant minimal amounts of seed and keep every
plant, you won't be doing much selecting.

Tending Plants

1. Watch plants for disease. Some diseases are seedborne. Don't save seed from
diseased plants or plants that are stunted, or that produced mostly sterile pods, or
that failed to thrive. If anything looks diseased, and I don't recognize the disease,
I either get a professional diagnosis fast, or I eliminate the plants fast, to be on
the safe side.

Universities with agricultural colleges usually have faculty or staff who can
help you identify an unfamiliar disease. In many cases they can do it from just a
good description. If they can't help, they usually can tell you whom to take or
send specimens to.

2. Practice selection and roguing, as described in Chapter 20.

3. If establishing isolation depends upon detasseling (corn) or removing flower

scapes on some plants, remember to do that during the season.

4. Mark plants or sections of plants that are reserved for seed saving. Mark
them really obviously and prominently. Bright orange surveyor's tape is good - I
use big swaths of it. Tell everyone involved in harvesting what the markers
mean. After all, it's hard to save seed from a fruit someone has already eaten
accidentally.

Harvesting seed is covered in Chapter 21.

 ONSIDER Buffalo Bird Woman, a Hidatsa Indian woman, working in
her garden in North Dakota in the mid-1800s. She doesn't know anything about
the role of pollen in reproduction of plants. She has never heard of isolation
distances. Yet she saves all her own seed. She even raises extra seed corn to
trade to others. A string of her seed-grade ears of corn is so valuable, it can be
traded for a tanned buffalo skin. (See Buffalo Bird Woman's Garden: Agriculture
of the Hidatsa Indians, by Gilbert L. Wilson; Minnesota Historical Society Press,
1987.)

Yet, today, in order to save seed, we need to talk extensively about isolation
distances. How did Buffalo Bird Woman get away with not needing to know
about all this?

Imagine a pioneer family with their fields and gardens. They save seed from
their own

corn, soup beans, pole beans, and other crops. They don't know any more about
plant genetics than the Indians. Yet, they, too, are "getting away with it." How is
this possible?

Traditional Seed Saving and Isolation Distances

I believe the explanation lies in a number of factors, a combination of social

and cultural habits, as well as a few useful myths:

1. Distinct varieties were planted separately. Different corn varieties were each
planted in plots of their own, never mixed together.

2. Sizes of the plantings were large enough so as to provide automatic

isolation.

3. Plantings were often separated from those of neighbors by pastures, woods,

or other kinds of plantings.

4. Often, each family grew just one variety of a species. Grandma had "a
favorite bean" that had been passed down in her family, not several favorite
beans.

5. Where the family grew two or more varieties of a species, the different
varieties often had different growing niches, maturity times, or purposes that
dictated that they would be grown in different areas.

6. If a family grew two or more varieties of a species, the plants and the seeds
looked distinctly different. (A family would not grow two varieties of white
beans that looked the same, for example.)

7. The gardener or farmer was thoroughly familiar with her varieties, so that
offtype plants or seeds were easily recognized.

8. Gardeners and farmers had gardening and farming parents, and very
traditional patterns of planting that had been passed down through the
generations. Seed savers had seed-saving parents, and inherited the patterns as
well as the seeds.

9. Where scientific knowledge did not suffice, useful myths filled in the gaps.

10. Maintaining pure varieties was sometimes not important.

11. Even when "pure varieties" were important, the concept of "pure variety"
was much looser, making people more tolerant of less-than-perfect seed saving.

Buffalo Bird Woman planted five different corn varieties: hard white (flint), hard
yellow (flint), soft white (flour), soft yellow (flour), and sweet. She planted each
in a separate field. The corn varieties had maturity times ranging over the entire
season, with the soft

white being earliest. Differing maturity times helped provide isolation.

Buffalo Bird Woman knew that "corn travels." That is, she knew that growing
different types in fields adjacent to each other didn't work, that off-color and
offtype kernels would be found in the ears. She had a cultural myth that the
kernels of one type could "travel" and get into ears of another type if the fields
were too close together. The myth of "traveling" helped her keep varieties pure
by encouraging appropriate isolation.

Sometimes, Buffalo Bird Woman tells us, women with adjacent fields made
arrangements so that the adjacent corns would be the same variety. They also
tried to separate fields of different types. They learned from experience, for
example, that corn traveled more over "soft ground." In other words, the corn
was better isolated by untilled ground, which probably contained trees and brush
that broke up wind patterns.

Finally, Buffalo Bird Woman saved only the best, biggest, most true-to-type
ears for seed. In this way, she would recognize and eliminate many of the crosses
that did occur. She and her neighbors grew only five varieties of corn, each
distinctly different in appearance, and each thoroughly familiar. Recognizing
crosses and eliminating them was relatively easy. It didn't matter very much how
many crosses occurred if she could eliminate them from her seed crop each
season.

Buffalo Bird Woman also had five kinds of beans, which appear to have been
bush drying beans. Each had a different color. Her pattern for planting beans and
corn put each in hills. She alternated rows of corn hills with rows of bean hills.
Nine rows of corn with alternating hills of beans made a block that was
separated from the next block with a row of hills of squash.

Buffalo Bird Woman kept beans of one type together in planting, threshing,
and storing. It's a good guess that different types were planted in different
blocks. These cultural and agricultural traditions gave pretty good isolation
distances. If your mother was a successful seed saver and you learn and always
follow her traditional planting orders and patterns, you don't need "book
learning" about isolation. Your cultural pattern provides it.

Buffalo Bird Woman's squash were yellow or green or mottled, and of various
shapes. She seems to have practiced no special methods to isolate different
varieties, and apparently didn't have distinct varieties. Rather, she had a very
diverse population of squash (Cucurbita maxima), and used all the shapes and
colors as summer squash, either cooked immediately or dried in slices for winter
use.
Now, imagine a pioneer family in a small settlement. Pioneer Man has two
varieties of corn. One is early season and yellow. It will make a crop no matter
what. The other is a full-season white that provides the best possible yield in an
average or good year. In a bad year, though, it doesn't make a crop at all.

The maturity times are so different that the corns are largely isolated
automatically. The early-season corn is yellow. The biggest, plumpest ears are
the first ears on each plant, and they all form before the later, white corn is
flowering. Just saving the biggest, prettiest ears for seed is enough to guarantee
purity of the yellow corn.

The late tassels on the yellow corn can contribute pollen to the first ears on the
late white corn. However, when this happens, pale yellow kernels appear on the
white ears. Merely choosing fully white, true-to-type ears of white corn for seed
is enough to eliminate nearly all crosses with the yellow.

Furthermore, the yellow corn is a small plant, and the white is big. The hybrid
plants are intermediate, and are regularly eliminated.

A small creek with marshy land and alders separates Pioneer Man's farm from
one neighbor. A woodlot separates the corn from that of the other closest
neighbor, who is growing the same two varieties anyway. These two varieties are
fine-tuned for excellence in this particular region. Everyone grows them.

Pioneer Woman has her favorite variety of pole bean for green beans. She
grows just the one variety. She has a bush bean for shelly beans. It's planted
fairly far from the pole bean, because it's short. The pole beans are the tallest
thing in the garden, and are the last row.

There are bush drying beans, but they are a field crop. They're a long way from
the pole beans.

Pioneer Woman has both hot and sweet peppers in her garden. Her mother
taught her that if you grow hot and sweet peppers side by side, the sweet peppers
will be hot. So Pioneer Woman grows hot and sweet peppers in opposite corners
of the garden.

In fact, hot peppers cannot influence the flavor of sweet peppers. But if you
grow the two types adjacent to each other, and save the seed, you'll have crosses.
The crossed, hybrid seed, when grown up the next generation, will generally
produce hot peppers.

The idea that hot peppers can influence the flavor of adjacent sweet peppers is
a myth, but it's a useful myth. If you believe this myth, you isolate your peppers.
You don't need to isolate them just to grow good sweet peppers, but you do need
to if you are saving seed from sweet peppers.

The larger your plantings and the fewer the varieties you grow of any one
species, the easier the seed saving. Seed saving is easiest for farmers and people
with large gardens. With large plantings, you just plant different varieties in
separate blocks and save seed from the middle. That alone gives you pretty good
isolation for seed for home or farm use. Seed saving is also easiest if your
plantings are far from those of neighbors.

All the added detail about isolation for seed savers of today is necessary
because most of us these days don't have plantings that are very large, let alone
far from those of our neighbors. Nor do very many of us have seed-saving
parents whose traditions we could learn directly. Further, many of us want to
grow lots of varieties of each species, not just one. And some of us are growing
seed for distribution or sale, not just for home or homestead use.

We need to know more about isolation distances because we need to know

what we can get away with. And we need to learn various tricks so that we can
get away with as much as possible.

Isolation Distances for Organic Farmers and Gardeners

In order to maintain pure varieties, we have

to isolate them from other varieties of the same species. If they are outbreeders,
we have to isolate them. If they are inbreeders, we still have to isolate them,
especially if we have healthy organic land that is alive with myriad species of
pollinators. If the crop is "highly self-pollinating," it just means that the
percentage of hybrids will be lower, not that there won't be any.

When I wrote the first edition of this book, I accepted conventional wisdom
about isolating inbreeding crop plants. I said: "With the plants that are very
highly inbreeding, only a little isolation is required - often just enough to prevent
physical mixing of the plant material." This is probably valid for the corporate
and university fields where data on isolation distances are most often determined
experimentally. However, I have come to the conclusion that it is not true at all
for those of us who grow garden or farm crops organically

In retrospect, I should have become suspicious just by the "official" listings of

pollinators. If you look in Table I, you'll see that for most of the insectpollinated
crops the listed pollinator is usually honeybees, or bumblebees, or both. Go to
any healthy organic garden or farm, though, and look for the pollinators there.
I'm talking about a long-time organic garden, one that includes or is flanked by
permanent unmowed and unsprayed areas where complete insect life cycles take
place.

If it's a farm-scale operation, there may be a little creek with wild banks. There
are natural, unmowed corners and strips along fence lines. The less-fertile,
rougher ground has been left as a woodlot, or is being allowed to regenerate
whatever is natural for the area. A huge blackberry thicket, more effective than
any fence, separates the property from the next-door neighbor. It's a haven for
wildlife of all kinds, including insects.

When you examine plantings flowering in such healthy circumstances, you

find that they usually have a dozen or more different pollinators working them.
There are honeybees and bumblebees, of course, but also many types of small
solitary bees, and dozens of different species of flies. I tried once to count the
different species pollinating a few square feet of flowering mustard. I tallied at
least a dozen flies and half a dozen bees of various kinds within just a few
minutes of watching.

I didn't realize the implications of all this healthy-land "pollinator pressure,"

however, until I crossed up eight varieties of soup beans and thirty lines of
chickpeas.

Beans are often considered one of the more highly inbreeding species. Backyard
seed savers frequently give them no isolation at all, or only a few feet. There is
controversy in the professional literature as to how inclined beans are to
outcross. Most researchers claim they are highly inbreeding, and that little
isolation is needed. A small minority report frequent outcrossing. I knew that
such things have a lot to do with the climate and the insects of a region, so I took
the precaution of consulting Jim Baggett.

Jim Baggett had been the vegetable breeder at Oregon State University for
decades, and beans were one of his major crops. OSU is right here in Corvallis,
in the Willamette Valley of western Oregon, just a few miles from my home and
various farms where I raise crops. So I figured that whatever isolation distance
Jim used for his beans would be good enough for me.

Jim told me that beans are really very highly self-pollinating here. He had no
trouble with cross-contamination of adjacent plantings. In fact, he said, the rates
of outcrossing are so low that you are more likely to get offtypes from new
mutations than from an outcross to an adjacent planting. All you need for
isolation is enough distance to allow you to keep from physically mixing the
seed when you harvest.

So I planted my soup bean varieties using three feet of isolation in between

different varieties. I planted `Jacob's Cattle', `Hutterite', `Gaucho', `Pinto', `Black
Coco', and others. In between the varieties of common beans I planted beans of
other species such as soybeans or teparies so that the "three feet of isolation"
didn't actually cost me any space. I wanted to see which varieties would do well
in my region, and to get a taste of each. I would continue to grow the ones I liked
best.

All went well that first year. When I harvested the beans they looked exactly as
I had expected; that is, each batch was uniform and typical for the variety. That
year, I got a seed increase on the better producers, but not enough to eat any
batches of beans without wiping out my planting stock.

Some of the beans were so late in maturing that they proved marginal for me.
Others didn't yield well enough to bother with. Several thrived and yielded quite
well though. `Black Coco' was especially impressive. It's a large, black, round
bean that's very early and very high-yielding, and it has upright, vigorous,
purplish bushes. `Gaucho', a small, deep golden bean with a long shape, also did
very well.

The following year, I planted longer rows of the half-dozen varieties that had
looked most promising the first year. This time I harvested enough beans to eat a
good batch or two of everything and have plenty of each pure variety to continue
on with.

But ... pure variety? What were these long beans doing in among the round
`Black Coco' seeds? And why were some of the `Gaucho' beans speckled gold
instead of solid gold?

My beans weren't pure varieties anymore. The `Black Coco' was all black, but
not all the beans were round. Some were big and longshaped, like a black Jacob's
Cattle'. And some were shaped quite a bit like a black `Gaucho'. And there were
intermediate forms as well. The `Gaucho' beans were mostly the gold color they
were supposed to be. But there were a few that were the size and shape of pintos,
and they were speckled like pintos. They looked just like pintos, except they
were gold pintos.

And so it went. Every batch of seed was crossed up. In every batch of beans
was stark, unambiguous evidence of crosses - multiple crosses in fact. Every lot
of beans had crossed with several of the other lots. It was fascinating to just look
at the beans and figure out who had been up to what with whom. It was not,
however, much of a success as far as maintaining pure varieties.

The first year that I had grown beans, everything had crossed with pretty much
everything else in the whole patch. There was no sign of it on the beans I
harvested the first year, however. This result is typical. The shape of a bean is
determined mostly by the shape of the pod, which in turn is determined by the
genes of the mother plant, not those of the seed. Some of the outer layers of the
seed coat are also determined by the mother plant. The result is that you usually
can't tell a cross has happened by looking at the beans that result from the cross.
Instead you find it out one year later when you plant those beans and see what
kind of plants and what kind of seeds they produce.

If Jim Baggett, Plant Breeder Extraordinaire, doesn't need to isolate his bean
varieties, I wondered, how come mine were all crossing? Jim has created new
varieties, increased the seed up to a hundred pounds or more, released it to seed
companies, and so on. There is no possibility of his being wrong about his need
to isolate bean varieties. Why did my need for isolation seem to be so different?
I had a similar experience with chickpeas (garbanzo beans). I was unable to find
any hard data on isolation distances. Rich Hannan, at that time curator for the
USDA chickpea collection, told me that chickpeas are one of the most highly
inbreeding of all species, that they needed only about three feet of isolation - just
enough to prevent physical mixture.

I planted about thirty different accessions from the USDA collections. I gave
them five feet of isolation between varieties. As usual, I didn't "waste" the
isolation space. I planted heirloom soup peas in between the sections of
chickpeas.

Most of the chickpeas didn't perform well. This is not surprising, since
chickpeas are not normally grown in western Oregon. Even where they are
grown (in the more and climate of eastern Washington, for example) the seed is
treated with fungicides. I was not treating anything. I wanted beans that could be
grown organically.

About six accessions did quite nicely. They yielded reasonable quantities of
seed that, as expected, looked just like the seed I had planted. One of the most
productive varieties was, delightfully enough, the black-andbrown-mottled
popbean I talked about in Chapter 3.

The next year I planted the half-dozen varieties that had done best the year
before. This time, however, I noticed something funny about the plants. They
didn't seem to be very uniform.

Pure varieties of chickpeas have either green plants and produce white or light
seeds, or they have pinkish or purplish plants and make darker-colored seeds
(black, brown, tan, or green). The first year, each variety had been uniform for
seed type and plant color. The plants were either all green or they were all
pink/purple. But this year the green-plant varieties weren't all green. About 5 to
10 percent of the plants were pinkish/purplish. And blocks of pink/purple plants
didn't look very uniform either. Some plants were much lighter pink than others.
Uh oh.

Several uh ohs, in fact. Uh ohs in every single one of my varieties. Each

variety had a number of crosses. I harvested plants with atypical color or leaf
size and threshed them separately, and they showed segregation for seed types
and colors. Offtype seed also showed up in the main threshing. That is, the

plants I could recognize as crosses were, indeed, crosses. But I couldn't

recognize all the crosses. Back to the drawing board. I sent off to the USDA for
replacement samples and started over.

In conversations with both Jim Baggett and Rich Hannan, I've found that their
university fields have very little pollinator activity. Commercial agribusiness-
style tilling and spraying is the norm. The fields are, relative to organic
plantings, pretty sterile.

My observations about outcrossing of the inbreeders under organic conditions

may also explain some or even most of the contradictory observations in the
professional literature about pollinators, outcrossing frequencies, and necessary
isolation distances.

I suspect that those growers with healthier, more species-rich, vital lands -
those who are closest to growing organically, in other words - are the ones who
are noticing more pollinators and hence more outcrossing, and who require a
greater isolation distance between varieties than that which is usually specified.

In fact, even those farmers who use sprays and till fencepost to fencepost may
have an organic-growing situation with respect to pollinators if they have
nonagricultural land such as woods or marshes or permanently untilled ground
nearby.

Isolation Basics

In order to maintain pure varieties, we need to prevent their becoming

contaminated by outcrossing with alien varieties. There are two methods
involved: isolation and roguing.

We prevent or minimize accidental crosses by isolating the variety from

foreign pollen. That is, we grow the variety in such a way that it is not exposed
to foreign pollen, or the exposure is limited.

We rogue for a variety of reasons, but one of the main purposes is to eliminate
hybrids created by accidental crosses in the prior generation. In other words, we
rogue partly because we usually do not have perfect isolation. Whenever
possible, we rogue before the plants begin to flower, so that the rogue
contributes not at all to the next generation, either as female or male parent.

Maintaining pure varieties usually requires both isolation and roguing in every
generation.

Accidental crosses between two varieties often produce very vigorous hybrids.
This means that if we cull out all the slowestgrowing material in a planting of a
variety, we may be eliminating all the pure material and keeping all the hybrids.
To maintain pure varieties that are true to type, we have to eliminate the rogues
and keep the plants that are true to type. To do this, we need to know what "true
to type" is for the variety. This is easy if we are familiar with the variety, and
much harder if we aren't.

When I am growing a variety for the first time, I usually plant just a small
amount and let it grow without roguing the first year, and I don't save seed. I've
found that it usually takes me one season of growing a variety before I know
what "true to type" is. I need to see whether the biggest corn plants produced
ears that are true to type, for example. Or whether the vining squash plants or the
bush type ones yield the good tasting fruit. If I try to save seed the first year, I
am too likely to rogue the wrong plants.

Becoming familiar with the variety before trying to save its seed helps in other
ways, too. If I'm not saving seed, I don't need to isolate at all. I can put a known
variety I need to save seed from in one end of the garden or field, for example,
and the dozen others of the species that I am "trying out" or just growing for the
table at the other end.

Furthermore, more than once I've gone to heroic measures to isolate something
that was already thoroughly crossed up before I got it. If the material is already a
mess, and I still want it, I have to "clean it up." In that case, the first couple of
years I grow the material, I don't need to isolate it much. The genetic mishmash
that's already in the variety is larger than the additional mess introduced by a few
more gratuitous crosses. Only later, when the variety is purer, does serious
isolation become more important.

If I am growing seed just for my own use, I often use roguing instead of
isolation to maintain pure varieties. I'll arrange adjacent varieties so that I can
recognize the F, hybrid plants.

When I save seed, I always write on the label card specific information about
which other varieties of the species I grew that year and how large and how close
the plantings were. That way, when I plant such seed, I know which hybrids to
be watching for, and I can make a good guess about how many to expect.

If I'm growing seed for distribution or ultimate distribution to others, then I

practice "serious isolation."

Isolation Overview

Isolation involves protecting our pure variety from foreign pollen of three
kinds.

First, there can be wild plants and/or weeds that are close relatives of our crop
and able to cross with it. In some cases they are the same species. In a few cases
they are technically a different species but are in the same genus, and can still
cross enough to create problems.

For example, the common weed known as Queen Anne's lace and the
cultivated carrot are both members of the species Daucus carota. If there is
Queen Anne's lace flowering in or near the patch with our carrots, we'll get
crosses. To save carrot seed, we need either to eliminate the Queen Anne's lace
or to make sure none is flowering at the same time as our carrots.

Wild mustards often cross with various domestic mustards or other brassicas,
even when they are not the same species. In my garden, all the brassicas seem to
be able to cross with each other. The hybrids, however, are often almost
completely (though not totally) sterile.

Second, we need to protect our variety from pollen of other varieties of the
same species that we might be growing.

Finally, in many cases, we also need to protect our variety from wild or
cultivated material being grown by neighbors and others in the area.

There are three basic ways to establish isolation distances. We can look them
up. (See Table I and/or Suzanne Ashworth's book Seed to Seed.) We can
establish them ourselves. (See Chapter 9, page 120.) Or we can guess.

I use all three approaches. I usually start by looking them up. If no information is
available, I usually guess based upon flower anatomy and pollinators. Even
when there is information, we must still do some checking, watching, and
guessing, because our conditions and pollinators aren't identical to those of
others.

Differences in regions, growing conditions, and pollinators are part of the

reason why different sources sometimes give different information, and why I
give multiple numbers and sources in Table I where they are available. If you
grow using organic methods, the larger outcrossing frequencies and isolation
distances are more likely to be the relevant ones.

I always watch an unfamiliar plant as it flowers and note the flower anatomy
and the pollen vectors. Flower anatomy often gives us clues as to whether the
plant is primarily an inbreeder or primarily an outbreeder. If the pollen is loose
and blows around in the wind, then wind is at least one of the vectors, and some
isolation is undoubtedly required. The plant may be primarily inbreeding or
primarily outbreeding depending on whether pollen from the same flower or
from other flowers usually reaches the stigma, and whether there is an
incompatibility system.

If something unfamiliar is being pollinated by bees, we can make a good guess

about isolation distance by considering what is necessary for other bee-
pollinated species. For the bee-pollinated outbreeders, though, we'll often find
that the distances are in the range of '/4 to 1 full mile. There are ways of isolating
plants other than by distance, however.

Factors that Affect the Need for Isolation

1. For most practical purposes, plants that are generally inbreeding need less
isolation than those that are deliberate outbreeders. The primary inbreeders
include tomatoes, common beans, and peas, for example, but not runner beans or
fava beans, which outcross regularly instead of only occasionally. The term
"inbreeding" is, at best, relative, however.
 2. With insectpollinated crops, what's the pollinator pressure? Under organic
growing conditions, even the "inbreeders" may cross and may need more
isolation than is necessary just to prevent physical mixture of the seed.

However, many people grow some of the more inbreeding crop varieties side
by side for years without experiencing enough crossing to be noticeable.

3. Do you need absolute purity? Can you tolerate no outcrosses whatsoever? If

so, you need very much greater isolation than if you can tolerate, say, a fraction
of a percent of outcrosses.

Most seed obtained from seed companies has a fraction of a percent of

contamination in it already. If you rogue out four or five outcrosses per thousand
seeds that were in the material when you got it and introduce one or two new
ones, that may be good enough.

4. The size of the plantings matters. A single 50-foot row of each of two
varieties of beans separated by 50 feet in a garden, with, say, a corn patch in
between, will be more than well enough isolated for most purposes. But if you
have twenty 50-foot rows of each

variety, with the blocks still separated by 50 feet, you'll get many more crosses.
If your bean rows are 50 feet away from a 200-acre commercial planting, you
may get so many crosses that trying to save pure bean varieties in that garden
becomes impractical.

5. Arrangement matters too. Bees tend to fly up and down rows, for example.
But they often skip sections of rows. They are much more likely to skip 20 or 50
feet down the same row, however, than to fly over 50 feet of corn patch to a
different row.

6. If you have two different varieties separated by 50 feet, it matters a whole

lot what the 50 feet is. Visual and wind barriers are better than bare ground or
low plantings. Flowering plantings that insects can "clean themselves up on"
may be better than nonflowering plantings.

7. How much space do you have? For most of us, isolation is a matter of
compromises between what might be optimal and what is actually possible.
 Isolation Distances, Absolute and Practical

The isolation distances recommended by Suzanne Ashworth in her book Seed

to Seed take all the available professional literature into account and give
numbers that are stringent enough to establish absolute isolation under virtually
all conditions. These distances are quite large. Most of the distances for the
legumes, for example, are in the range of `/4 to 1 full mile. For chickpeas, for
example, Ashworth notes that some researchers say that crossing is common.
She recommends a distance of '12 mile.

Where absolute purity of a variety is required and no crosses at all can be

tolerated, I believe isolation distances for the socalled inbreeders need to be
about the same as those for the deliberate outbreeders. In most cases, distances
ranging from about 1/4 to 1 full mile.

Most of us these days don't have the ability to isolate most plantings by V4 to
1 full mile. Fortunately, however, only rarely do we need to. In very few cases
are we working with absolutely pure varieties.

Most home gardeners will find that they can raise different varieties of most
insectpollinated crops and keep them acceptably pure with isolation distances of
about 20 to 100 feet between separate rows or blocks of different varieties.

The 20-to 100-foot minimum works best when the plantings are not too large,
and when the grower combines the basic isolation distance with other tricks to
maximize isolation effectiveness.

Alan Kapuler, for example, gardens and grows seed crops here in western
Oregon on 2 acres of organic garden that is alive with pollinators. Yet he can
grow dozens of different varieties of common beans and runner beans at once
and save seed, and maintain and do seed increases on pure varieties.

He does this by alternating rows of trellises of runner beans and common

beans. Usually each trellis is separated from the next by a small block of corn as
well. This gives good enough isolation to save and increase seed for sale, even
though there may be only about 40 feet between one variety of beans and the
next variety of the same species. Alan gets occasional crosses, but they are rare.
Generally, the seed he sells is much "cleaner" than what he starts with, and it is
clean compared to other seed that is available commercially.

Growing alternate rows of different species of beans on trellises provides both

a visual barrier and material for the bees to "clean themselves up on" between
different varieties.

Unless you have absolutely pure varieties and need absolute isolation, I suggest
the following Basic Rule for Everyday Seed Saving: Isolate things as well as you
can with the space you have. If you don't have much space, use any of the
various tricks and methods given later in this chapter and make the most
effective use of the space you have.

Even if your isolations are less than optimal, with a little luck, most of the time
you'll get away with it. And, of course, sometimes you won't. Sometimes you'll
get crosses. I crossed up eight varieties of beans and thirty varieties of chickpeas,
however, and the world did not end.

In some cases, absolute purity is desired or required, that is, no outcrossing at all
can be tolerated. Here's an example:

My black-and-brown-mottled popbean, a chickpea, outcrosses. The plants are

of a size, color, and form that represents primarily dominant characteristics, so
crosses are usually not recognizable.

The characteristics I'm most interested in, flavor and popping, seem to be
destroyed totally by outcrossing and are not easily recovered by backcrossing.
That is, an outcross destroys the variety, but invisibly. Further, no signs of
outcrossing show up in most cases until a full generation after the hybrid has
contributed pollen enough to contaminate everything.

In addition, one offtype in a thousand could break one tooth per offtype. At a
rate of one offtype per thousand, almost everyone who ate a few pounds of
popped bean per year would break at least one tooth. With ordinary beans, the
offtype, if eaten, has trivial consequences. At worst, the bean cooks, looks, or
tastes different from the rest. An offtype popbean, however - one that didn't pop
and wasn't soft when you were expecting it to be - can do you serious personal
injury.
 Finally, I am breeding material that I believe (on my more optimistic days) to
have the potential to be the start of a whole new snack food industry. If so,
thousands of pounds of chickpeas might someday be growing from those I'm
breeding and maintaining right now.

I isolate this chickpea from small garden plantings by at least `/4 mile. I simply
don't grow it within miles of any large commercial planting. Given the situation,
in this material, no crosses at all can be tolerated.

Isolation Tricks and Methods

1. Where large isolation distances are required, we can grow just one variety to
seed in any given year. Note that in many cases we can grow numerous varieties.
We just can't have more than one flowering at the same time. (Consider what
else is growing in the region or area, though. The air in summer is full of corn
pollen, for example, whether we plant any or not.)

2. We can isolate in time instead of with distance. We can grow an early, a

midseason, and a late variety, for example, if the flowering periods don't overlap.

3. We can hand-pollinate. Squash and other cucurbits, for example, are easy to
hand-pollinate for small-scale seed saving.

4. Visual barriers help minimize outcrossing performed by insects. A 20-foot

block of corn in between two different bush bean varieties is much more
effective in isolating them than 20 feet of open space or space with other low
plantings.

If I have two kales that flower at the same time, and I want to save seed of
both, I'll often plant them on opposite sides of the house. Just watch the
pollinators and see what they do. In one 4x8-foot patch, individual bees will
work the kale until it is "used up" for the moment (until more flowers open and
anthers dehisce). Then they go to the sage flowers within the same patch, not to
the kale in a patch out of sight on the other side of the house.

5. Barriers such as tall corn or hedges or trees also break up wind patterns.
Wind carries both windborne pollen as well as pollinating insects.
 6. We can grow as many varieties as we want, but remove flower scapes or
flowers from all but one for the relevant period. With brassicas, for example, I
frequently cut the flower stalks off the varieties I don't want seed or
contaminating pollen from. But I keep the plants so that I can continue eating
from them. What matters isn't how many varieties I have in the same small
space, but how many varieties I have flowering simultaneously in that space.

7. We can cage the plants or plantings so as to eliminate or restrict pollinators.

Reemay (or some similar floating row cover material) can be used to exclude
many insects. Screens or mesh are sometimes used. With outbreeders, though,
we may have to introduce clean insects into the cage to get pollination.

8. We can bag individual plants or inflorescences and either hand-pollinate or

introduce appropriate insects.

9. We can eat the edges of the patch and harvest just the center. Rich Hannan
tells me, for example, that 250 yards of isolation is needed for fava beans under
his conditions but only 50 yards if it is planted with material the bees can clean
up on. In other words, he could plant many fava varieties in adjacent 150-yard
plots if he harvested just the central 50 yards of each block for seed.

10. For seed intended for personal use, it is often easier to depend upon being
able to recognize and eliminate hybrids than to prevent them through heroic
isolation measures. I often isolate brassica varieties by just 20 feet or so, if I
know I can recognize the hybrid. The official isolation distances are usually in
the range of '/4 to 1 mile. I always get a few hybrids under such conditions.

Pure pink/purple plants crossed with pure green ones give pinkish plants that
are obviously different from the green parent (but which may or may not be
distinguishable from the pink/purple parent). Varieties that have leaves with
deeply indented margins, when crossed with those having smoothmargined
leaves, usually yield hybrids with leaves of intermediate type. Tall varieties
crossed with short varieties give hybrids that can be intermediate or tall
depending upon the specifics. So we can detect the cross in seed of the short
variety, but not necessarily in seed of the tall one.

With corn you can just detassel hybrids to eliminate their pollen; then you can
keep the plants and eat the ears. I usually eliminate flower scapes instead of
whole plants so that I can eat the plants instead of wasting them.

11. Instead of excluding foreign pollen, we can just swamp it out. There is
always corn pollen in the air during spring and summer, for example. But we
may be able to get away with maintaining a variety by open pollination if we
have a substantial patch or field of a single variety and harvest for seed only
from the middle. There is foreign pollen in the air in the middle of the patch, but
it is outnumbered by thousands or millions of times by pollen from the plants in
the immediate vicinity.

12. We don't need to save seed of everything every year. It's easier to save
enough seed for several years. That way we need fewer good isolations in any
given year.

13. We don't have to plant everything on our own land. Friends and neighbors
will often be willing to let you grow a variety in their gardens when you need the
isolation - especially if you're in the habit of giving them seed.

14. On a somewhat larger scale, we can lease land elsewhere to grow some of
our seed crops, or contract with a grower to grow certain seed crops for us to our
specifications.

Official numbers with respect to isolation distances are largely derived from
open fields. I believe that the distances needed, even for crops such as corn, are
not nearly as large when there are lots of houses, trees, bushes, fences, and
hedges to break up wind patterns.

My yard and the surrounding area have lots of trees, hedges, and flowering
plants of all kinds that provide ample food for pollinators. There is enough for
bees to do so that they don't very often go farther than between adjacent yards on
one trip. Small flies are easily blown great distances by wind, but the wind
patterns are broken up by houses, trees, and hedges.

I routinely examine the yards of my neighbors within two houses of my own in

all directions to see what they are growing that might contaminate seed crops of
mine. With

houses, trees, bushes, and complicated wind patterns, it usually doesn't seem to
matter what more distant neighbors are doing with respect to insectpollinated
crops. Windpollinated crops are another matter, though. (And corn is an extreme
case.)

In short, look up isolation distances to give yourself a basic idea. If you see
isolation distances that are larger than the land you have available, though, don't
get discouraged. Look at your specific situation, notice your wind patterns,
watch your pollinators - and get creative.

 OW MANY plants of a variety do we need to save seed from in any
given generation in order to maintain the variety properly?

There are times when one plant is not only adequate, it's the exactly optimal
number. There are other times when saving seed from just one plant will destroy
the variety in a single generation.

In many cases one plant is adequate for some purposes, but not others, or
adequate, but not really optimal.

There are four reasons why we usually save seed from more than just one plant
in any given growout of a variety

First, we don't want our entire year's seed ruined by an accidental cross or an
undesirable new mutation. If we save seed from just

one plant, and it is involved in such an accident, the entire year's growout might
be affected.

Second, we usually want to maintain some or nearly all of the genetic

heterogeneity that might exist in the variety. The term "genetic heterogeneity"
refers to the fact that the individual plants in the variety have different
repertoires of genes in them. In order to preserve the entire range of genes
present in the variety, we save seeds from an adequate number of individual
plants each generation.

Third, some plants have incompatibility systems. If you actually grow only
one plant, it may be sterile or nearly sterile. If it produces seed, the seed is likely
to be of poor quality or to give rise to inferior, wimpy seedlings. If you grow
only two or three plants, they may happen to have the same incompatibility
genotype, giving the same results as if you were growing just one plant.
 Fourth, many outbreeding plants are subject to inbreeding depression. The
greater the extent to which a species tends to suffer from inbreeding depression,
the more important it is to maintain as much genetic heterogeneity as possible in
the variety - and therefore the more important it is to save seed from many plants
instead of just a few.

So the answer to the "How many plants?" question depends upon whether the
crop is a basic inbreeder or a basic outbreeder, whether it has a self-
incompatibility system, and the extent to which it is subject to inbreeding
depression. You can find this information about the crop you're interested in by
looking it up in Table I and Appendix A. (Table I is organized by scientific
names; if you don't know the scientific name, look up the common name in the
index.)

(Note: In Chapter 9 1 discuss the basic genetic nature of inbreeders and

outbreeders and give the reasons behind the pragmatic suggestions featured in
this chapter.)

Inbreeders and Numbers

Inbreeders give us the most options. Many a bean and pea and tomato variety
has undoubtedly been saved through just a single seed, or just a few. The
inbreeders don't mind inbreeding (even though they may outbreed to various
extents too). They don't suffer from inbreeding depression.

Most seed savers try to save seed from at least twenty good plants per
generation for most inbreeding crop varieties under most circumstances.
"Twenty good plants" means

that you grow out more than twenty so that you have twenty left after selection
and roguing. Twenty plants per generation is enough to preserve most of the
genetic heterogeneity in most varieties. Forty plants per generation will allow
you to maintain nearly all the heterogeneity in most inbreeding varieties.

In a few cases, where the plant material is very heterogeneous, larger numbers
might be warranted. To maintain a very heterogeneous landrace of an inbreeder,
for example, a USDA curator might choose to use a minimum of one hundred
plants per generation.
 With inbreeders, however, maintaining genetic heterogeneity is optional. The
individual plants are vigorous whether we maintain the genetic heterogeneity in
the overall variety or not. The main reason we want to maintain the genetic
heterogeneity is that it is the basis for future selection, adaptation, and evolution
of varieties.

Most seed savers save seed from at least twenty good plants per generation for
crops like beans or peas; the plants are small, and we're growing lots more than
that anyway. However, many seed savers save seed from just a single tomato
plant, for example, or just a few.

When we save seed from just a single tomato plant, what we are preserving is
actually a single subline of the variety, not the whole variety (see Chapter 9).
That subline is often all we need, though. And when many other people are also
saving seed of the same variety, many sublines are being preserved overall.
Anyone who needs more genetic heterogeneity in the material can obtain the
variety from several different sources and simply combine the lines.

Sometimes saving seed from just one plant is optimal. Whenever we notice a
superior individual plant of an inbreeding crop type, we usually save its seed
separately and use that seed as the beginning of a new line or new variety. When
we do this, we are eliminating most of the genetic heterogeneity of the original
line in return for the advantage of the superior characteristic we noticed.

There is another situation in which saving seed from just one or a few
individuals of an inbreeding crop is best: when the material has been crossed up.
Then, we can "clean up" the material by saving seed of several individuals
separately so we can identify one plant that is still pure. We use the seeds of that
plant to "restart" the line.

Outbreeders and Numbers

Corn, kale, and sunflowers are examples of outbreeding crops. Kale and
sunflowers have incompatibility mechanisms to prevent selfpollination. Corn is
capable of selfpollinating, but it is very subject to inbreeding depression.

With most outbreeders, maintaining the genetic heterogeneity of a variety is

not optional. When a variety has lost too much genetic heterogeneity, the
individual plants are wimpy. That is, they germinate poorly, grow poorly, yield
poorly, and die (or yield nothing at all) at the slightest excuse. Even if they yield
something, it is too small and/or too inferior to be worth the trouble. The loss of
vigor caused by loss of heterogeneity in a variety is called "inbreeding
depression."

For most outbreeders, with just a few exceptions, we never try to save seed
from

one or just a few plants. Instead, we save seed from at least forty to two hundred
plants in each generation, depending upon how subject the crop is to inbreeding
depression.

Corn is very subject to inbreeding depression. Most seed savers try to save
seed from at least one hundred good plants every generation. Many prefer two
hundred. (See Appendix A for seed saving information on corn.)

Most professional germplasm curators would want to save seed from at least
two hundred corn plants, and possibly many more - up to one thousand if it is
desirable to maintain nearly all the genetic heterogeneity, or if the variety or
landrace is very heterogeneous.

For outbreeding crop plants that are not as subject to inbreeding depression as
corn, we usually choose a number between forty and one hundred as the
minimum.

Squash is an outbreeder that is an exception to the rules. It doesn't have an

incompatibility system and isn't subject to inbreeding depression. We treat it as
an honorary inbreeder. We're usually very happy to save seed from just twenty
squash plants. Often, we save seed from a lot fewer than twenty because the
plants are so big, and hand-pollination is involved. (See Appendix A for seed
saving information on squash.)

Avoiding Inbreeding Depression

In order to avoid inbreeding depression you need to do two things. First, you
have to save seed from adequate numbers of plants in every generation so that
genetic heterogeneity in the variety is maintained.
 Second, the plants need to be grown so as to facilitate cross-pollination rather
than selfpollination.

If I plant sixty kale plants all together in a patch and space them two feet apart
in all directions, for example, the bees will frequently move from flowers on one
plant to flowers on another. If I space the plants six feet apart, the bees will
mostly move from flower to flower on a single plant. If the variety is capable of
selfpollination at all, it will be mostly selfpollinated under these conditions and
there will be inbreeding depression.

It doesn't suffice to plant sixty plants if you arrange them so that each is
basically all by itself with respect to pollination. Remember that even one
generation of inbreeding causes a loss of half the heterozygosity in each plant of
the next generation (see Chapter 9). Even one generation of forced inbreeding of
a plant subject to inbreeding depression can result in a severe loss of overall
vigor and yield.

If a variety is sufficiently strongly selfincompatible, and spacing is such that

pollinators don't cross-pollinate the plants, the result will be little or no seed.

If you interplant the kale such that they are even farther apart and separated by
other flowering plants, nearly all of the bee-trips will be from flower to flower
on one plant. Interplanting has its virtues, but consider the cross-pollination
aspect.

If you plant all the kale in a single row, plants will usually be pollinated by
their immediate neighbors. In small plantings of outbreeders, patches or blocks
are preferable to single rows.

If inbreeding depression results from failure to allow or provide for proper

cross pollination, the variety may still have enough heterogeneity in it, but the
individual plants don't. They are too highly homozygous and thus not vigorous.
You can regenerate the variety completely in a single generation, though, just by
planting with proper spacing and in a block. That way, the next generation of
seed is mostly from all kinds of crosses instead of mostly from selfings.

If, however, the variety doesn't contain enough heterogeneity (because you
saved seed from too few plants, for example), you can't regenerate it just by
planting it properly for one generation.

Normally, we do not try to save seed of an outbreeding crop such as corn via
seed from just one plant. There is an exception, however.

An F, hybrid plant between two fairly different varieties is heterozygous at

many loci. It has a tremendous amount of heterogeneity within just the single
plant - more than enough for an entire variety. We can save seed from a single
plant of such a hybrid and use it to breed a vigorous new variety.

What If I Don't Have Enough Room?

With most outbreeders, it's important to save seed from adequate numbers of
plants in order to maintain the vigor of the material. But you don't have to grow
them all in one year. You can grow a few each year, for example, and pool the
seeds from different years when you draw seed for planting.

In addition, there is a difference between optimal and practical. You may be

able to get away with just twenty kale plants per year, for example, instead of the
minimum of forty recommended - especially if you grow them in a block instead
of a row so as to maximize cross-pollination of the plants in all possible
combinations.

With inbreeders, if you don't have enough

space for the optimal number - well, life is full of compromises. Just grow as
many plants as you have room for (or need for the table). You might not be able
to preserve all possible genetic heterogeneity in the material, but you don't need
it all to have a good line.

 N SOME WAYS selection is obvious: we save seed from the best plants.
Yet in other ways selection is complex, subtle, and full of surprises. It is
amazingly easy to select for exactly the opposite of what we want.

To use selection most effectively, we need to know how to choose as well as

how to evaluate selection criteria.

Here are some guidelines.

Selection Basics

1. You must practice selection, culling, or roguing as part of every generation

when saving seed of a particular variety. There is really no such thing as variety
"maintenance." New mutations and outcrosses always occur. Varieties evolve.
You have to practice selec tion in order to hold a variety in the form that you
consider desirable. Unselected or improperly selected varieties deteriorate
rapidly.

2. Varieties are often rogued more than once in the season. That is, you go
through and remove inferior or atypical plants. In a patch of brassicas, for
example, I might go through once and eliminate any premature bolters. Then I
might examine the rest of the plants as they come into flower and eliminate those
that are smallest.

3. Roguing of outbreeders should be done before flowering if at all possible, so

that their pollen as well as their seed is eliminated from the next generation. This
isn't always practical or possible, however. Sometimes we can't identify the
rogue until later.

In corn, for example, we may not recognize a rogue until we see the mature,
off-type ears. We can easily eliminate the rogue plant's ears, but by then it's too
late to call back the pollen it contributed to all its neighbors. When we select or
rogue based upon ears, we are selecting using the mother plants only. We
eliminate about half of the undesirable effect of a rogue, but not all of it.

Selection based upon just mother plants is practical and effective, but not as
effective as selection based upon both parents.

4. Consider the possible virtues of a new mutation, accidental cross, or rogue.

You must eliminate it from your original variety. But that doesn't mean you have
to discard it. Perhaps it deserves consideration as the basis of a new variety.

I often put the interesting rogue at one end of the row and save seed from the
other end.

5. You can usually arrange to eat rogues and culls instead of wasting them.
You can chop the flowering scapes off the inferior brassicas, for example, so you
have more time to eat the leaves.

6. Any time you select, rogue, or cull, always look at the whole plant, not just
the individual fruits. You must select the best mother and father plants, not the
best individual fruits. Being best involves more than producing a single
impressive fruit. (See Chapter 6, pages 81-82.)

7. In choosing the best parents, remember edge effects - that is, the fact that
plants at the ends of rows or on the edges of plantings have more sun, water,
nutrients, and space than the rest. If a centrally located plant looks impressive,
that's more meaningful than if an edge plant thrives. Examine the edge plants
and figure out how being on the ends

or the edges affects things so you can mentally correct for this factor.

Did the superiorlooking plant have the same amount of space as the others? Or
did it have more than the others because its nearest neighbors died?

8. In choosing the best parents, remember environmental effects. Was that

superiorlooking plant the only one that wasn't shaded by the house? Or is it the
plant that is closest to the fertile spot created by last year's burn pile? Maybe the
plant really isn't genetically superior. Maybe it is only lucky.

9. Hybrids caused by an accidental cross may germinate quicker and grow

faster than the pure seed. Examine your material for trueness to type before you
do any culling. If you presoak seed, don't plant just the fastestgerminating seed.
Plant it all.

10. Do your roguing only after plants are big enough so that you can tell what
should be rogued.

11. With unfamiliar material, plant sparsely enough so that you don't have to
thin until plants are large enough to evaluate. You might thin out all the little,
slow-growing pure plants and keep all the accidental-cross hybrids.

12. If you plant under conditions that are atypical and save the seed, you may
be selecting for material that is inferior under your normal conditions.

If, for example, I plant `Oregon Giant' pole bean in late spring or early
summer, I would not want to save seed from the material. I would be saving seed
from the plants that did best under warm-weather conditions. Such material
might do poorly under ordinary conditions. When you select, think about what
you're doing. What is the essential purpose of the variety? Why are you growing
it?

The "purpose" of `Oregon Giant' in my garden is to thrive and produce from

earlier plantings than anything else, to yield earliest from early plantings, and to
grow happily in cold mud - not to produce earliest or best from warm-weather
plantings.

13. Select for natural resistance to soilborne fungal pathogens by planting

untreated seed, and plenty of it. (See 14.)

14. Select for a variety's ability to germinate vigorously under your growing
conditions by planting much more seed than you need for the number of plants
desired. Then thin to the best plants. Selection for many important but invisible
characteristics will happen automatically.

15. Select for the ability to grow and thrive under organic growing conditions
by growing plants under those conditions and saving seed. You don't have to
know what diseases you're selecting for resistance to, or what genes or genetic
characteristics matter. Just select as usual based upon the various visible qualities
and performance characteristics that matter to you. The additional selection for
organic-growing capability will happen automatically.

Selection Complexities, Subtleties, and Surprises

16. If you plant too early, so that the earlygerminating material is killed by
frost, you may be selecting for slow germination, thicker seed coats, or seed
dormancy mechanisms.

17. If you eat the early produce and save seed from later material you may be
selecting for lateness.

18. If your variety has an extended harvest because various plants mature at
different times, and you want to maintain that characteristic, you must save seed
from plants with the entire range of maturities.

It's easy to save seed from just the earliest plants, because you notice them, and
the seeds are ready first. If, however, you want to develop or maintain a variety
that will have an extended harvest period, you'll probably need to collect seed
more than once during the season. If you let the early-maturing seed sit around
until the late material is ready, the earlier seed may shatter and get lost, or be
rained upon and ruined. This would mean that you would be accidentally
selecting for lateness.

19. In selection, there are usually tradeoffs. For example, selection for more
fruits per plant usually means that the individual fruits will be smaller. Selection
for two pods per node on peas generally gives pods that are about as large as
those that are alone at the node. But more than two pods per node is usually
associated with smaller pod size.

20. Ears from early varieties of corn are generally smaller than those from
main-season varieties. You can select for larger ears on an early variety. But you
can't expect ears on an early variety to be as large as those of the best later
varieties. Later varieties have longer to develop. Generally, they have bigger
plants and more leaf mass to support bigger ears.

21. If you select the biggest ears of corn from your harvest as seed ears, you
might be selecting for lateness. (The biggest ears are often produced by bigger,
later plants.) Or you may be selecting for plants that have only one bigger ear
instead of two or three. The single bigger ear may be more desirable. Or you
may prefer the greater total productivity usually associated with more ears per
plant.

22. Selecting for high productivity sometimes results in a lower sugar content.
That's because the same energy is going into making sugar for a larger amount of
fruit.

23. When you are creating new varieties, you can't select for everything
desirable simultaneously. If you try to select on the basis of six different
characteristics, for example, you will probably not be able to select very
effectively for any of them. Decide what your priorities are. You can select most
effectively if you concentrate on just one or two characteristics.

24. On the other hand, when you rogue established varieties, you're usually
eliminating just a few plants for any given reason. You can rogue for many
reasons simultaneously, as long as you have enough plants left for a good seed
crop.

25. Saving seed from a volunteer? Consider why it is a volunteer.

For instance, that volunteer fava bean may represent a seed from a plant that
shattered its seed earlier or more easily than the rest. Save its seed and you could
be selecting for shattering, which makes seed saving much more difficult.
Shattering is especially inconvenient in a grain crop.

Or perhaps that volunteer mustard is from a seed you planted last year that
didn't germinate. Instead, it sat dormant all season, overwintered, and sprouted
the next spring. That seed may represent a plant that has mutated back to wild-
type seed dormancy mechanisms. If you save its seed you may be dealing with
erratic, unpredictable germination from there on out.

26. Usually, you should not eliminate variability in an arbitrary manner. Often
certain characteristics go together in ways that aren't apparent. You might
eliminate all the different colors of seed in a variety of mustard, for example, so
you have just one color. But maybe resistance to a disease important in your area
was linked to one of those genes that you just discarded. Now you can't select
effectively for disease resistance.

We need consistency for certain characteristics for the crop to be useful to us.
We usually try to eliminate variability for flavor, for example. But we don't need
the mustard seed to be all one color. With soup beans, though, color is a
component of flavor. So we usually do want a soup bean variety to be one color.

27. Some plants produce more seed than others. If we pool the seed from all
the plants we save seed from, the plants that make more seed will contribute
more to the next generation. We will be automatically selecting for seed
production. This may be good if the crop is a dry bean, for example, where the
seed is also the useful edible portion. It isn't necessarily good if the plant is a
mustard, where a plant that produces a lot of seed may have been one that
produced a small amount of edible leaves.

Sometimes we need to correct for differences in seed production. One method

is to save a constant amount of seed from all the plants we've chosen.

Selection for the Purpose of Germplasm Preservation

Sometimes germplasm preservation is our primary goal in saving seed. In that

case, we try to keep all the genes and genetic heterogeneity in the material,
whether this is useful or optimal for us or not.

When USDA curators maintain varieties, for example, they try to maintain all
the heterogeneity. If they were to select for material that yields best at their
location, they might be discarding genes important to the potential grower who
lives in another area. If they selected a certain color or flavor or plant form, they
would be eliminating the genes needed by someone who wants the opposite
characteristics.

When the primary goal is germplasm preservation, we try to avoid

automatically or deliberating adapting the material to our personal preferences or
specific situation. We use our knowledge about selection to avoid selecting.

We keep seed from all the plants that are true to type, not just the best. (We
still eliminate the rogues that represent outcrosses, however.) We may even keep
equal amounts of seed from all the plants so as to avoid accidentally selecting for
plant size, vigor, yield, or seed production under our conditions.

Evaluating a Selection Program

It's useful to evaluate your selection program and selection criteria. Sometimes
the variability that we see isn't genetic. If the variability for a characteristic isn't
genetic, selecting based on that characteristic doesn't accomplish anything. It
doesn't change the variety.

For example, my black-and-brownmottled popbean produces seed that is either

black-and-brownmottled or black. The difference isn't trivial. The black beans
have a more intense flavor when popped.

So I hand-sorted the beans, then planted the mottled seeds in one row and the
black seeds in another. I found that both rows gave me the same proportion of
mottled beans versus black ones in the next generation. This meant that the seed
color differences in this variety were not heritable.

Here's another way of stating the same thing: Something other than genes is
responsible for the variability in seed color in my popbean variety.

Yet another way of saying it: There is no genetic heterogeneity for seed color
in this material.

If the beans could talk they might describe the situation this way: "These color
differences are not our fault!"

All the above versions of the statement have the identical implication: I could
select from now till doomsday without having any effect on the seed color of my
popbean.

Of course, one then wonders why I am getting two colors of beans. Threshing
out individual plants provided the answer. It's differences in harvesting that
matter. Seeds from pods that dry fully before the plant is cut are always mottled.
Seeds from pods that are somewhat green when the plant is harvested are green
initially, but they dry out black.

If I harvest a plant only after it is completely dry, all its seeds are mottled. If I
cut a plant that has half dry pods and half greenish ones and leave it to dry, when
I thresh I get a combination of types.

Knowing that the color difference isn't genetic, I don't waste time, space, or
seed by selecting on the basis of seed color. Instead, I focus on other
characteristics which are highly influenced by genetic differences in this
material, such as plant form, yield, and disease resistance. At the same time, I
realize that I can always get more black seeds, if I want them, by modifying my
harvesting appropriately.

Many of the differences that we see between various plants of a variety have
more to do with the environment than with genes. Some plants landed on better
soil, weren't nibbled on as much by insects, and so on. Selection works only on
genetic differences, not environmental ones.

There are three simple ways to evaluate a selection criterion. One is the approach
I took with chickpea seed color. Plant two different batches of seed that are
selected differently. Do they give you the same range of types? If so, the
variability is not genetic. It's environmental in some fashion.

A second method involves planting both selected and unselected seed. Instead
of discarding the seed from the inferior plants, you save at least some of it. Then
you plant some of that seed side by side with the selected seed. Then you
compare the two plantings. Is the seed of the rejected plants actually inferior? Or
is it yielding plants that are just as good as those from seed of plants you're
selecting? If both batches of plants are equivalent, your selection criterion isn't
working.

Sometimes the seed of the plants you are rejecting produces the superior batch.
If so, your selection is effective, but it is backfiring. (The next job is to figure out
why.)

Sometimes the selection is working but has negative side effects you will want
to know about and consider. Red color in corn seed, for example, is often
associated with less vigorous plants, lower yield, and smaller ears.

The third method involves planting seed from different years side by side.
Then you compare. Is your selection working? If so, is it having any unexpected
effects as well? The results can prove interesting and instructive.
 Does Selection Always Work?

In a word, no.

Sometimes selection is effective and sometimes it isn't. That's why it is often

worth our while to evaluate whether any given selection criterion is working in a
given situation. Avoiding ineffective selections allows us to focus on effective
ones. I can select much more powerfully for disease-resistant popbeans, for
example, if I don't discard half the beans every generation over differences in
color that are not genetic.

Selection works by genes, not by magic. The effectiveness of selection

depends upon underlying genetic heterogeneity for the characteristic in our
material.

 ANY WONDERFUL heirloom pole bean varieties have been saved
one pod at a time. Whenever the gardener was in her garden, she picked the pods
that had started to dry, and tucked them in her pocket. Back in the kitchen, she
tossed the pods in a basket with others of their kind to finish drying. During
winter, she brings out the baskets full of different types of pods and starts
shelling them out by hand in the evening, while talking and relaxing with the
family.

The kids, even the littlest ones, are allowed to help. And they want to. They
feel honored. They consider it a privilege. "Look at this one!" they exclaim over
each unusual type, every new discovery.

I think that our fascination with seeds is hardwired into our genes as a species.
Watch

young children opening bean pods and finding the seeds. You can see that they
know they are doing something special. You can see that they sense the magic.
 Harvesting

When should you harvest plants for seed? I let the plants proceed with their
natural process of making and dispersing seed for as long as possible. I intervene
only at the last possible minute, so that the seed disperses into my hands instead
of all over the ground, or into the bellies of self-appointed sharers.

My simple all-inclusive rule for fruits is to allow them to become as mature as

possible without actually rotting, falling off, or being eaten by bugs or slugs, or
stolen by birds or squirrels.

Adequate seed of some crops can be obtained from fruit that is at eating stage,
but the best-quality, largest, most vital seed comes from fruits that were
harvested way past prime eating stage. You can save adequate seed from eating-
stage tomatoes, for example. But when the tomato has softened way past
edibility the seed is bigger and fatter and will germinate more enthusiastically,
giving rise to more vigorous seedlings.

You can save good melon seed from the fruits as you eat them. But the best
seed comes from fruits that you have allowed to mature beyond eating stage in
the field. I tie a large chunk of bright surveyor's tape around the stems of fruits
I've chosen for seed so that I don't harvest them for the table by accident.

For plants that produce dry seed, such as beans or mustard or most flowers, I
prefer to let the plants or stalks mature and dry out completely in the field. I
harvest earlier only if I must. Generally, this is because of wet weather or the
presence of competitors - birds or other critters that could beat me to the seed.

Some seed crops are harvested before the plant is completely dry because the
seed shatters, that is, pods release the seed to fall out on the ground. With
buckwheat, for example, the earliest, biggest seed starts to drop long before all
the seed is mature. Getting a good yield of a crop that has shattering seed
requires guessing at the best harvesting time. The goal is to recover most of the
early, big seed and as much additional seed as possible. Usually, the best
harvesting time is just a bit after the earliest seed begins to shatter.

For small amounts of seed for my own use, my basic seed harvesting
equipment

includes plastic freezer containers, paper bags, index cards, and a pen to write
with. Whenever I harvest for the table, I take baskets for the fresh vegetables and
the extra containers for seed.

In my main garden I water regularly, and water can ruin seeds once they have
started to dry. So I snip flower heads and clip pods here and there as they start to
dry, and put them in containers. I record what each seedhead or pod is on an
index card and toss it in the container with the seeds. When I get back to the
house, I tuck the open bags and containers away on a shelf to let the seeds finish
drying.

I often write quite a lot on the index card that accompanies the seed. Seed type,
harvest date, whether anything else of the same species was growing in the
garden, and, if so, what it was and how close - all this goes on the card, as well
as anything unusual that prompted me to save seed from this particular plant.
When I clean the seed, it goes into a new container, and the card goes along.
With the card, I don't have to recopy all that information each time I change
containers or clean seed.

For growing larger amounts of seed, the basic harvesting equipment includes
more paper bags, some 5gallon buckets, and tarps. You will often plant an entire
row or section of a row or bed of the crop specifically for seed production. And
you should try to arrange harvesting so as to optimize the quality of seed and
minimize the labor.

As an example, I grow a lot of brassicas such as kale and mustard. I plant them
in beds away from the vegetables that need water in the summer. I don't water
them at all. In spring they are growing actively, and we have plenty of rain here
in Oregon to cheer them along. In May, when the plants start to want to dry
down, the rains stop. Then I need to water the rest of the garden, but the
brassicas I allow to dry out completely.

Oregon, where we have little or no summer rain, is ideal for such seed crops.
Rains (or overhead watering) damage seed quality once the seed begins to dry.
So if you have summer rains or your plants are exposed to overhead watering,
you will usually let the plants dry out only partially. Then you'll cut the plants
and stack them loosely somewhere under cover to let them finish drying.

When I grow mustard in fields away from home, for example, I leave it until
the plants are completely dry. Then I clip the stalks near their bases and toss
them on a handy tarp. I transport the plants back home as well as store them on
the tarps. I set the tarps in the shade somewhere, under cover if rain threatens,
and fluff the plants up a bit so that air continues to circulate through them. That's
a good description of how it works when everything for this seed crop is
optimal, including the climatic pattern and the weather.

I used to handle mustard in my home garden exactly the same as that in the
field. In my home garden these days, though, I can't leave the seed mustard so
long anymore, because of LBBs (my acronym for Little Brown Birds). As soon
as the mustard starts drying, hordes of LBBs descend upon it and have a
wonderful time stripping the seeds out of the pods.

When the flock descends, each drying mustard bush quivers all over, every
branch alive with dozens of tiny twitchings and rustlings. It would be
enchanting, except for the fact that those birds are capable of eating every single
seed of the crop in just a few days, leaving the bushes with nothing but stripped,
empty pods. On a plant that one day earlier had thousands of pods, each
containing several seeds, you can search for minutes without finding a single
good seed.

The first season I got LBBs, I lost most of the mustard seed from my earliest-
maturing lot. When a later-maturing batch started to dry, however, I was ready. I
watched the drying plants carefully for the first signs of bird damage. As soon as
I saw those first stripped branches, I harvested the plants and stacked them
loosely on a tarp on my back deck to finish drying. The result was a back deck
and tarp full of bird droppings, and still no seeds.

I had overlooked the fact that birds fly. The LBBs flew to the garden, found
their breakfast missing, and required all of about one nanosecond to spy it up
there on the deck. So up they came. Right in front of my very eyes, just five feet
away on the other side of the glass sliding door, they ate all my seeds.

These days, I pile my homegrown mustard plants on a tarp in the driveway on

the other side of the house. The LBBs haven't seemed to notice it there. At least
not yet.

Other seed-saving friends of mine in the area, when told about the birds, said,
"Aha! So you have those too, huh?" Apparently, you can get away with major
seed-saving activity in one location without being molested for a couple of
years, but a couple of years only. Soon the LBBs discover what you're up to.

I have plans for the future that involve whole fields of mustard. I hope to avoid
the LBBs by the ordinary measures of crop rotation. Crop rotation is usually
practiced in order to promote soil fertility and minimize plant diseases. However,
it's also very helpful in avoiding buildups of pests, or in minimizing the extent to
which existing levels of pests focus their attentions on your crop and develop
inconvenient habits.

Bush bean plants and pods are usually allowed to dry as much as possible before
harvest. If you have summer rain or overhead watering, you may need to harvest
partially dry plants and allow them to finish drying under cover. A general rule is
to let half the pods dry, then clip and harvest the plant.

If you don't have to worry about water, you can leave the plants until they dry
fully in the field. In that case, you still need to harvest them before the pods
begin to shatter and dump the seed on the ground.

Pole beans are sometimes grown on the ground for a seed crop. (This is how it is
done commercially) But usually a home gardener will be growing the pole beans
on a trellis or on corn plants in the main garden and eating most of the beans.
Seed saving consists of just choosing one section for seed production, and
picking the pods as they dry. Once a plant matures and begins to dry off the first
few pods, it doesn't produce many more green beans. So it's best to use one
section for seed saving and keep the main planting well picked so that the plants
continue to

produce properly for the table. (The same applies to peas.)

When you harvest dry plants such as mustard or beans, you want as little
shattering to occur as possible. When you thresh plants, however, you want as
much shattering as possible; that is, you want the plants to release nearly all their
seed readily. I often harvest dry plants in the morning, when the air and plants
are moist from dew. At this time, there is little loss of seed from shattering. I
usually thresh in the afternoon or early evening, when the plants are as dry as
can be, and release their seed most readily.

Even a fully dry pile of plants is easiest to thresh in the afternoon of a sunny
day. The plants rehydrate just slightly each night and morning and dry out again
each afternoon. If you thresh in the morning, or on a damp day, you not only get
a significantly smaller portion of the seed released, but you have to work much
harder for what you do get.

Many people harvest dry plants such as brassicas or beans by pulling up whole
plants and tossing them on tarps. I never do it that way. Instead, I always clip the
plants and leave the roots and dirt in the field, where they belong. I find the root
balls contribute rocks and clumps of dirt that are much harder to separate and
clean out of the seed than is the dry plant debris.

Equipment useful in harvesting includes gloves to protect your hands from

sharp, dry plants, containers or 5gallon buckets for fruits, and tarps of various
sizes for holding and carrying dry plants. The 5gallon plastic buckets are a
standard in the food industry. You can get used ones for free from restaurants.
My local food co-op sells the used buckets for $1 (cleaned).

The tarps can be bought for prices ranging from $5 to about $20, depending
upon size. You'll want several of various sizes for both harvesting and
transporting plants as well as threshing seed. I have handy little 4x6-foot tarps,
as well as sizes ranging up to 15x20 feet or so, which hold about as much as I
usually want to carry.

There are such things as one-row combines and small-scale threshing

machines. One of these years, I keep telling myself, I'm going to need to get a
small used stationary thresher. Even the smallest versions of such equipment,
however, are likely to cost more than you'll want to spend unless you are selling
seed - and fairly large amounts of seed.

Generally, people who are growing up to about one hundred pounds of

brassicas or beans, ten pounds of tomato seed, and a few hundred pounds of dry
corn, for example, are harvesting it by hand using the simple tools I have
mentioned: gloves, 5gallon buckets, and tarps.
There are specialized tricks for harvesting certain crops. For example, I harvest
seed from bunching onions (Allium fistulosum) by clipping the flower stalks as
they dry and turning them upside down inside open paper bags. When the seed
heads are completely dry, most of the seed will fall out spontaneously or with
just the gentlest shaking.

The plants don't dry all their scapes simultaneously. If you wait until the last
scapes dry before clipping them, most of them will have

already dropped most of their seed. To prevent this, you need to go through the
patch every few days and pluck the dry scapes.

Furthermore, if you break up the seedheads by rubbing or stomping on them as

we do with many dry-seed crops, the debris from the dried flowers is of such a
size that it is almost impossible to separate from the seed.

I give some of the tricks associated with specific vegetables in Appendix A.

Seed to Seed, by Suzanne Ashworth, treats seed saving species by species, and is
an excellent source of such information. In addition, whatever the crop, with
experience you will develop your own additional repertoire of techniques, tricks,
and methods.
 Threshing and Cleaning

One fall evening I went over to the Kapuler family home for dinner. There was
a big pile of dry bean plants on a tarp in the middle of the living room floor.
Alan and Linda Kapuler, their kids, and the invited guests all shelled beans onto
the tarp as we visited. The task is simple. You pick up a dry plant or chunk of
plant, shell out the pods, and toss the empty plant (empty pods all still attached)
onto a separate pile. It's very companionable.

By dinner time, there was a pile of plant debris and a tarp full of dry beans that
were clean except for a few stray dry leaves.

Once upon a time, people always worked with their hands as they talked. They
repaired tools, mended, spun, knitted. And I doubt not that they handthreshed the
smaller lots of seeds. It's a healthy pattern. I often listen to classical music when
I handthresh seeds. I don't have a TV, but if I did, you can bet I would be
shelling out beans while I watched it.

There are two basic kinds of seed-processing situations. "Dry processing" refers
to seeds associated with dry pods and other dry plant debris. Beans, peas, and
mustard are examples. The goal is to separate the seed from all the rest of the dry
plant material. In most cases this is necessary because the junk takes too much
room to store, and it is full of bugs or eggs of bugs that would eat most of the
seed during the winter.

"Wet processing" refers to seeds that are associated with fruits. The seed has to
be separated from the fruit, washed, and dried. Often, a fermentation process is
involved. Tomatoes, squash, and cucumbers are examples of crops whose seeds
require wet processing. To wash seed you simply put the seed in some water and
rinse it. To fermentationprocess seed you put the seed in some water and let it sit
around for a while. Then you wash it.

When you wash the seed, you remove various germination inhibitors that
prevent the seed from germinating inside the wet fruit. So, after the washing, it's
important to dry the seed quickly to keep it from sprouting. Those are the basics.

I don't always clean small lots of tiny dry seeds if they are just for my own
personal use. It depends upon how much space the uncleaned seed takes and
how hard it is to remove the debris.

For example, I don't clean shungiku (ed ible chrysanthemum) seed. I don't
produce much of it, and it is just for my own use. The seeds are tiny and of such
a shape that they don't separate from the debris very easily. So I don't bother to
clean it.

I free the seed from the dry flower heads by rubbing them between gloved
hands. This breaks up the flower heads and produces a seed/debris mix that is
alive with tiny insects of many sorts. These insects would be happy to eat all the
seed during the winter. So I put the seed/debris mix in a quart jar and put it in the
freezer. This kills the insects and insect eggs. I broadcast the whole thing when I
plant. It isn't elegant, but it saves time and it works.
 Dry Processing

There are two stages in dry processing. First we use mechanical force of some
type to release the seed from the pods or plant material. Then we clean so as to
separate the good seed from the bits of leaf, pod, stem, chaff, and other junk.

If you have just a few pods or seedheads, you can rub them between gloved
hands and separate the seeds. Or you can put them in a tray and run a rolling pin
over them. Just try a few methods until you find one that works. With beans or
peas, for a few ounces of seed you can simply open the dry pods by hand over a
big bowl and strip out the seeds.

A few dry bean plants can be put inside a pillowcase and rubbed around or
stomped on a little to release the seeds. This method is especially useful for
threshing individual plants.

Supposedly, you can beat individual plants or masses of dry plants against the
inside of a barrel. I've never tried it, though. The last barrel I actually ever saw
was in a small country store, and it was full of pickles. I don't think the barrel
method would work for most plants - only with plants that release their seed
extraordinarily easily. I do have a line of big, 6-foot-high, grain-type fava beans,
however, that thresh out so easily that I thresh them simply by hitting them once
or twice against something. The "something," however, is (surprise!) a big tarp
on the driveway with the edges propped up to make a big tray.

For larger amounts of any dry plant and seed material, tarps are very handy. I
pile dry mustard or bean plants on tarps in the field and use the tarps to transport
them. The plants then dry in the shade on the tarps until they are completely dry
and I am ready to thresh. (Direct sun damages seed.) Threshing consists of
turning on some marimba music and dancing on the pile of dry plants until most
of the seeds have been released.

Loose seed can be separated from debris by screening with sieves of some
type. You can buy sieves of various mesh sizes that are constructed specifically
for seed cleaning. (See Appendix E.)

My screens are three pieces of hardware cloth I bought in a hardware store.

These come in rolls 3 feet wide. You can have the store chop off whatever length
you want. I have pieces of each of the three smallest of the mesh dimensions, 1/8
inch, 1/4 inch, and `/2 inch.

My 'a-inch and '4-inch screens measure 2x3 feet and are unmounted. My '/2-
inch screen is 3x6 feet and is mounted on two

bamboo poles wired to the two long edges. With this set of screens, I can
separate just about anything. Sometimes I use the mesh to remove all the large
stems and plant debris and let the seeds and small stuff through. Other times I
use mesh of a size that retains the seed and lets all the finer debris through.

Winnowing involves separating seeds from debris by using wind or air. A

traditional Native American method for winnowing is to toss the seeds in a
basket in the wind. You are supposed to toss the seeds and small debris skillfully
into the air and catch the seeds gracefully, repeating the process until the chaff
has all blown away.

I don't know anyone who can actually do this. It has certainly never worked for
me. Perhaps our winds here in western Oregon are just too variable. We seem to
have either no wind, or variable winds of constantly changing direction. It's hard
enough just to avoid losing the seed. Getting good separation of chaff and seeds
just doesn't happen. In addition, the wind always changes direction and blows
the chaff right back into my face.

I find it more useful to use a window fan during a period of minimal wind. I
don't usually try to pour seed from one container into another in front of the fan,
though, as is often suggested. I lose too much seed, or retain too much debris, or
both. Instead, I prefer to pour my seed in front of the fan onto (of course!) a big
tarp.

When I clean mustard seed, for example, I start by hauling the tarp with the
dry plants to the concrete driveway, which I've swept to remove any rocks and
debris. Then I dance and stomp and shuffle all over the plants on the tarp. The
pods mostly separate from the plant as well as splitting open and releasing the
seed.

The next step is to remove the stems and large plant debris. I put down another
tarp and prop up the lower edge and corners with three big rocks. (My driveway
slopes, and round things such as mustard seeds roll.) For this I use my 1/2-inch
hardware cloth, the big piece that is mounted on bamboo poles.

I put this hardwarecloth frame down on the empty tarp and prop up one end
with a concrete brick. Then I ball up the tarp with the seeds and debris and dump
the mess onto the frame. All the big stems and plant fragments stay on the
hardware cloth. All the seed, chaff, and empty pieces of seedpod pass through
the screen. I discard the big stuff, bundle up the tarp with the seeds and debris,
and dump the seed/debris mix into a 5gallon bucket.

Next comes the winnowing. I set a window fan on a box above one end of a
big tarp (about 10x15 feet, for example) and attach it to a long extension cord.
The bottom of the fan is about 1'/2 inches above the tarp. Much higher and too
many seeds bounce off the tarp; much lower and you don't achieve as good a
separation. I put rocks in a few places under the edges of the tarp to turn it,
temporarily, into a big tray. Then I turn on the fan so that it is blowing
lengthwise down the tarp.

Now, using a smaller container, I scoop up seeds and debris, a quart or so at a

time, and pour them in a steady stream back and forth above the fan. The wind
from the fan hits the debris and pushes it downwind above the tarp. The dust and
lightest debris blows away

entirely. Somewhat heavier debris goes all the way down to the end of the tarp.
Still heavier debris settles nearer the fan. Heavy scraps of stem settle nearest the
fan. There will usually be a swath about 2 or 3 feet wide about halfway down the
tarp that contains nearly all the best seed, and this seed is nearly completely
clean.

This method gives you separation for seed quality as well. Light, small,
inferior seed travels farther than the best seed.

After the winnowing, I bundle the tarp so as to dump the debris at the two ends
onto handy little 4x6-foot tarps. Then I bundle the big tarp again so that an edge
near the good seed is over the edge of a big stainless steel bowl. Into the bowl
goes the clean seed.

From the stainless steel bowl, the seed goes into a quart jar and directly into
the freezer. That's because there are always lots of insects on the debris, and the
seed carries insect eggs. I always freeze-cycle all my brassica seed as the last
step of threshing and cleaning.

I often use a couple of large stainless steel bowls for lots of seed that are too
small to require the fan and tarp treatment. I put the seeds and fine debris in one
of the bowls and shake it a bit. The chaff and junk will migrate to certain points,
leaving much cleaner seed everywhere else. I gather the areas of clean seed and
move them to the second bowl. I discard the regions containing seed with lots of
debris. (Generally, I have lots more seed than I need, but not necessarily lots
more time.) After a few passes, most of the seed is (mostly) clean.

One of my favorite tools for cleaning bean and chickpea seed is a large
cardboard box top of the sort that tables and large appliances come in. Mine are
about 4x8 feet, with edges about 4 inches high. In other words, they are big
cardboard trays.

I put a cardboard tray on the ground, lift one end to tilt it a little, and dump the
seed/debris mix into the high end. The seed bounces over the debris and rolls
down to the other end. Some seeds roll better than others, depending upon shape.
But debris and chaff rolls little if at all, and I can tilt the box to various degrees
and jiggle it as well. The result is almost perfectly clean seed after one or two
passes.

When I clean large amounts of chickpeas, for example, I first thresh them out
by dancing on the plants on a tarp. Then I bundle up the edges of the tarp to form
a funnel-shaped bag and shake it a bit. The seed and pod pieces and smaller
twigs settle to the bottom of the bag. The big plant stems and pieces remain on
top. I gently set the tarp down and open it up, and remove and discard that
surface layer of large junk. Then I pour the mix of seeds, pod pieces, and smaller
debris into a 5gallon bucket.

Next comes the cardboard tray treatment. I sit down crosslegged with one end
of the cardboard tray resting on the pavement and the other in my hand. I use a
quart-sized container to dip up portions of the seed/debris mix with the other
hand. One hand dips and delivers a stream of plant material to the upper end of
the tray. The other hand controls the tilt of the tray and jiggles it appropriately.
 Seed that is roundish and big, like most beans and peas, will bounce right over
the accumulating debris and roll right down the

cardboard tray and accumulate, as almost totally clean seed, at the bottom end.
By dumping the seeds from greater or lesser heights above the tray, I control
how hard they bounce. By the angle of tilt and the jiggling, I control how much
momentum the seeds have as they continue to roll. Once you get the hang of it,
it's an efficient and satisfying method.

Every once in a while I stop to remove the good seed from the bottom end of
the tray, and dump out the debris accumulating at the top end.

I also use my big cardboard trays to thresh out individual chickpea plants,
something I do a lot of because of my chickpea breeding work. The individual
dry plants start off in separate paper bags. That's how I harvest individual plants
in the field, collecting the most productive or interesting-looking plants. In each
bag is an index card giving the plant's identity and information about why I
decided to collect it separately. I leave the paper bags open so that the plants can
continue drying. The paper bag retains any pods that might fall off the plant
during transportation or storage.

Later, when I get around to it, I thresh out the individual plants. I put a whole
dry plant into one end of my big cardboard tray on the driveway, and stomp and
dance on the plant to thresh out the seeds. I remove the three or four biggest
plant stem chunks by hand and toss them onto a nearby 4x6-foot tarp.

Then I raise the end of the tray with the debris and seeds a few inches and
jiggle it a bit, so the seeds roll away from the debris to the other end of the tray. I
lift the clean seed out of the far corner of the tray with the index card label, then
put both seed and label in an open plastic freezer container to finish drying
and/or for later examination (counting, recording seed characteristics, weighing
the yield, etc.).

Corners and crevices of cardboard boxes can retain seeds from one batch and
dump them into the next. I tape inside flaps and corners of cardboard boxes so
that they don't retain seed easily.

Tarps, too, can contribute to crosscontamination between various batches of

seeds, especially with little seeds such as brassicas, which can stick to minor
flaws in the tarps. I use new or newer tarps for seed work, and patch rough
places with tape. I also take precautions to preserve tarps. I sweep the concrete
free of rocks and sticks before I put a tarp with seeds down and dance on them.
Small rocks underneath the tarp can cause tears or blemishes when danced upon.
Dragging tarps over rough ground causes even worse tears and holes. I use older
tarps for such purposes.

To avoid crosscontamination between batches of seeds, I shake tarps and boxes

thoroughly after I use them. Then I inspect them thoroughly before I use them on
the next batch of seeds.

Clothes can also be a source of crosscontamination. Beans can fall into pockets
with one batch of seed and fall out when you bend over the next. Rolled up shirt
sleeves or cuffs are almost sure to transfer seeds from one batch to the next.
Even the seams of long-sleeved shirts can trap brassica seeds. I wear short-
sleeved T-shirts with no pockets when I am cleaning seed. And I shake and brush
myself off thoroughly between batches of different seeds along with all the
cleaning equipment.

For specific information on harvesting, threshing, and cleaning corn, see

Appendix A.
 Wet Processing

Seeds of many fruits and berries can be cleaned just by washing them in water.
Other fruits, such as tomatoes and cucumbers, have seeds that are coated with
gelatinous material. The gel makes cleaning and handling the seeds difficult or
impossible, and it usually also contains germination inhibitors. So we need to
add a fermentation step to the process, to digest the gel and release the seed.

After fermenting or washing, the seeds are wet and their germination inhibitors
have been removed. Therefore, to prevent germination or any loss of viability
the seeds must be dried promptly. Little seeds can be dried quickly in sieves or
on screens, given a wind or a fan to keep air moving over them. Larger wet
seeds, such as fermentationprocessed squash, require auxiliary drying of some
sort. I use a food dehydrator set on 95°F for up to 8 hours. (See the next section.)

Tomato seed, for example, is surrounded by a gelatinous layer that contains

germination inhibitors. In processing seed, this layer is either removed by an
acid treatment or digested by fermentation. For most gardeners, the fermentation
method is easiest, and it also eliminates many seedborne diseases.

Saving tomato seed is especially satisfying because tomato seed is so

expensive. In addition, it's very easy to produce much better quality tomato seed
than anything you can buy commercially.

I process tomato seed in two different ways, depending upon the amount of
fruit and seed involved. I often save seed from just a single fruit or a few fruits
from one plant. With just a few fruits, I squeeze the juice with the seeds out of
the tomatoes and ferment just that. That way I don't have the problem of
separating the seeds from the pulp after the fermentation.

I start by allowing the fruits to ripen way past the edible stage - on the vine if
possible, indoors otherwise. To process, I start by cutting each fruit in half across
the pattern of internal seed cavities. I use my thumbs to dig and squeeze the juice
containing the seeds out into a bowl. I put this fluid in a clear plastic cup of
suitable size.

I like clear cups so I can see the fermentation happening, that is, see the
bubbles forming and rising. I use very tiny cups if I have just a small amount of
juice, because the surface portion will be discarded. I use containers that are
small enough so that the tomato juice is at least an inch deep.

I put the seed-cavity juice with seeds in the cup, put an index-card label under
the cup, and leave the cup and its contents alone for one to five days. I don't add
water; that would just slow things down. I don't stir. How long the fermentation
takes depends largely on temperature. Usually, a mat of mold will form at the
surface of the juice. At some time after that, the seeds will be "done."

After the fermentation has been proceeding for a day or two, as indicated by
the bubbles, or after the surface mat of mold forms, I test the seeds, and I test
them every day until they are done. I dip gently into the container,

remove some fluid with seeds, and put it into a small strainer. I hold the strainer
under the faucet to wash the seeds. Then I feel the seeds. If the gel around the
seeds is all gone, they're done, and I'm ready to end the fermentation. If the test
seeds still have gel on them, I wait a day and test the batch again.

If you ferment the seeds too long, viability of the seed and vigor of the
resulting seedlings is harmed. Grey or brown instead of bright cream-colored
seed indicates overly long fermentation. Most instructions for saving tomato
seed give an arbitrary time for the fermentation. I think it is very worthwhile,
however, to test the fermentation and choose the right endpoint instead of just
guessing.

When the fermentation step is over, I remove the mat of mold at the top of the
tomato juice and discard it. Then I pour the rest of the tomato goop into a
strainer. I wash the seeds by holding the strainer under the faucet.

I dry the seeds in the same strainer. I use the flow of water from the faucet to
spread the seeds evenly out in a layer on the strainer. I then set the strainer
somewhere that has good air circulation, such as in front of a fan.

After only a few hours, the seed will be "semidry." Then I "rustle up" the seeds
in the strainer with my fingers, spreading them around and abrading them a little
against the strainer. The object is to separate the seeds from each other so that
they don't dry together in clumps. (I've found this "rustling up" useful in the
drying of all wet-processed seeds.) Then I spread the seed out in the strainer and
forget about them for a while. (Several days or more.) By then the seeds are dry
enough to store.

With small batches of seeds from a few tomatoes, I do the processing indoors.
If there's more than a few ounces of tomato glop fermenting, I do the entire
process outdoors. Even a few ounces of fermenting tomato glop smells pretty
powerful.

For a big batch of seed, I start by harvesting the whole, overripe tomatoes into
5gallon buckets. Then I stomp them. I have a heavy-duty plastic garden tub that's
big enough to hold a person and a few gallons of tomatoes. I put tomatoes in the
tub in a layer up to about 6 inches deep, put on rubber boots, and hop in. I stomp
the tomatoes only well enough to break each one into a few big chunks. That's
good enough to release most of the seeds. It's tempting to stomp the tomatoes
much more. It obviously releases a larger proportion of the total seed and it's
also really quite a lot of fun. In fact, it's easy to get carried away. If you don't
exercise control, in just a few minutes you'll have stomped the tomatoes to
smithereens.

Unfortunately, it is very hard to separate tomato seeds from tomato-fruit

smithereens. But it's very easy to separate the little seeds from big intact tomato
chunks. So restraint in stompery is warranted.

After the stomping I transfer the tomato material to 5gallon buckets for
fermentation. I usually add a little water in order to facilitate the release of the
seeds from the fruits and to make it easy to stir the mess. (Not a lot of water;
perhaps about 50 percent as much as the volume of tomato material.) I stir the
fermenting material every day, and test some of the seed each day until the seed
is "done," as described previously.

Next comes screening. I usually add much

more water to the tomato glop after fermentation and right before screening in
order to help free the seeds from the tomato chunks and pulp. I use my three
hardwarecloth screens to separate the pulp from the seeds. I put the biggest-mesh
screen over an empty 5gallon bucket and pour the tomato glop through into the
next bucket. Then I grab another empty bucket, and the next-smallermesh
screen. And so on.
 When I'm done, I have one or more 5gallon buckets full of tomatoish-colored
water with tomato seed at the bottom. I pour off the excess water. Then I pour
the final amount of water with the seed in it out onto a propped-up old screen
door. I use a stream of water from the hose to finish rinsing the seed and to
spread it around evenly on the screen. Then I set the screen with the seeds in
front of a fan for drying.

I remember to "rustle up" the seeds after they are dry enough (but not too dry)
so that, after the final drying, I end up with nearly all individual seeds instead of
clumps. That's all there is to it.

One cautionary note: It's very easy to transfer tomato seeds from one batch to
another. They tend to stick to your hands and to the buckets and screens. Wash
your hands and rinse strainers or hose down buckets and screens between
batches of different seeds.

You can process a pound of tomato seed at a time pretty easily using methods
involving nothing more than a few 5gallon buckets, a good solid stomping tub, a
fan, hardwarecloth screens, and a good chunk of window screen. Actual seed
yields per amount of fruit vary wildly depending upon variety, but an average is
about one pound of seed for every 100 pounds of fruit.

I describe wash-processing and fermentationprocessing of squash seeds in

Appendix A. Seed to Seed gives detailed information on processing methods for
just about every conceivable vegetable seed.
 Drying Seed

Seed dried at room temperature usually has a moisture content of 10-20

percent. That is not dry enough to store in airtight containers or to freeze. Such
seed should be stored in paper envelopes or bags or other containers that allow
some air exchange. Seed stored in paper is subject to attack from insects or pests,
however. In addition, many kinds of seed have insects or insect eggs in them as
they come from the field. Such seed will be destroyed inevitably if we don't take
measures to kill the insects and insect eggs. Freezing is a good preventative
measure.

Storing seed in glass jars or, for bigger amounts, in 5gallon buckets is an
effective way of preventing infestations of both insects and rodents.

Only very dry seed can be stored in plastic bags, jars, or other airtight
containers. And only very dry seed can be frozen.

Most little seeds like mustard and lettuce can be dried well enough without
artificial drying. Only if you live in a desert area, though, are you likely to be
able to dry bigger seed such as corn and beans to the "very dry" point without
artificial methods.

The way you tell whether a batch of seed is "very dry" - that is, dry enough to
store

in airtight containers or to freeze - is by testing it.

Test small seeds or thin ones such as squash by bending them. If they are "very
dry" they will snap instead of just bending. With bigger seeds such as corn,
beans, or peas, do the "hammer test." Take a few seeds outside and put them
down on a piece of brick or concrete. Hit each seed with a hammer. If the seed is
"very dry" it will shatter like glass when hit. If seeds are inadequately dry, they
will smash or mush instead of shattering.

"Very dry" corn or bean seed has a moisture content of about 6 to 8 percent.
This level of moisture is optimal for storage. Too much moisture and the seed
remains physiologically active and continues to respire; longevity is affected.
Too little moisture and viability as well as longevity will be affected.

To dry seed to the "very dry" stage for freezing or storing in airtight containers,
you can use either silica gel (see Appendix E for sources) or a food dehydrator
that has a thermostat. You can't use a home oven to dry seed. Its thermostat
controls don't go low enough. How fast seeds dry over silica gel depends on how
much seed you have, how wet it is, and how much gel you're using. Use the
"very dry" test to tell when the seed is dry enough. If you use equal weights of
seed and gel, a few days or less is usually sufficient.

Silica gel is likely to be most useful for small amounts of seed. Bigger batches
of seed take lots of silica gel, which you then need to dry in the oven to reuse. I
think it's easier to dry the seeds in a dehydrator.

Many people use small round food dehydrators to dry seeds. You set the
thermostat to 95°F and dry the seeds for up to about 8 hours. Eight hours is
enough to take big squash seeds that have just been fermentationprocessed all
the way from fully wet to very dry.

I've used a couple of the round dehydrators for years. They have several
disadvantages. The holes in the middle of the circular trays greatly reduce the
potential amount of seed each tray can hold, and the trays are small to begin
with. There is also only a constant air stream, which is divided among however
many trays you use. So if you have eight trays instead of four, the drying will
take twice as long. In addition, you need to rotate the trays during the drying,
because the bottom trays get more air than the top ones.

Nevertheless, these round dehydrators only cost about $50, and they can do a
good job on small lots of seed. If you get one of the round dehydrators, get one
with a thermostat that goes down to the 95°F that you'll need for seed drying. In
addition, it's important that the design be such that the air flow is divided and
moves across the trays and out. Some dehydrators force air through all the trays.
The air goes through the first tray and then through the second, etc. With such a
design, only the material in the first tray is dried very effectively.

I recently bought an Excalibur dryer. I needed it for drying summer squash.

The little round dehydrators just don't do the job when it comes to drying pounds
and pounds of fully wet vegetables. A one-pound summer squash fills a little
round dryer to capacity, and one one-pound summer squash is nothing.

The Excalibur quickly became my pre ferred dryer for drying seed as well. The
large square trays with no holes in the middle are a lot more useful. The air
stream is constant over all the trays instead of variable. And the dryer has both a
thermostat and a 26-hour timer. Furthermore, by omitting alternate trays, one can
even dry big items such as whole ears of corn.
 Protecting Seed from Insects

There are quite a lot of insects whose major purpose in life is to eat all the seed
you have saved. All organic pea seed raised here in western Oregon has weevils
or weevil eggs in it, for example. Unless you do something to kill the insects,
they will develop during storage and destroy nearly every seed. In some areas of
the country, weevils are a serious problem in beans. Nearly everyone has
problems with corn. Grain moths and many other insects love it.

There are three basic approaches to killing insects in seed. Some people
sprinkle diatomaceous earth in with the seeds so as to kill or discourage insects.
I've never used this approach. My seed collection overlaps too much with my
food supply. I have received batches of seed coated with the gritty white stuff,
though, and I can vouch for the fact that it doesn't look or taste very appetizing.

To kill weevils and other insects and insect eggs I put seed in glass canning
jars with airtight lids. Then they go into the freezer for at least several days. (The
seed must first be fully dry).

When I need the seed, I remove the jar from the freezer a day ahead of time. It
isn't a good idea to open a jar of seed right after it has been removed from the
freezer. Moisture will condense all over the seed, and I doubt the sudden shock
does the seed any good.

A third method for killing insects in seed uses dry ice. I'm planning to start
using this method on the larger amounts of seed such as the 5gallon bucket loads
full that I seem to be accumulating. Many university research departments and
medical clinics use dry ice. You may be able to find one that will sell you the
modest amounts you need, or they will be able to tell you the name of a source in
your locale.

Use at least 3 ounces of dry ice for a 5gallon bucket of grain. For larger
amounts, use at least 1 pound of dry ice per 30 gallons of volume. Crush the dry
ice and mix it with grain in the bottom 2 inches of the container. Add the rest of
the grain and cover loosely for 8 hours or more to let the dry ice vaporize into
CO2. The CO, being heavier than air, displaces the air. After the 8 hours, seal the
container. Insects will die from lack of oxygen.
 Once you have killed the insects and insect eggs in seed, you need to store it in
such a way that it cannot become reinfested. Paper packets and bags don't hold
off a serious onslaught of insects for very long. For seed I care about the most, I
use glass jars or 5gallon buckets.
 Protecting Seed from Rodents

Sooner or later, if you have lots of seeds lying around in open containers
drying, or protected only by paper or plastic, you will attract mice or rats.

A few years ago in my house, a mouse suddenly appeared and started

processing my

squash seeds into more mice. I preferred the squash seeds.

Being a believer in biological controls, I did what I considered the logical

thing. I started transferring all my seeds to jars and other rodent-proof
containers. Simultaneously, I got a baby ferret. I figured the ferret could eat the
mice.

I wish to report that a ferret in a seed collection is a whole lot worse than mice.
Mice only remove a few seeds here and there for personal use. They don't love to
play in seed, turn over or open all the containers, scatter all the seed on the floor,
and steal all the empty containers to roll around the room as toys. Mice don't get
excited by noise, and push all the glass jars of seeds off the shelves just to hear
them go "crash."

Also, mice are small. They can't drag a 1pound package of corn seed away and
hide it under the furniture for later investigation.

With a ferret around, if I leave an opened box from a seed company alone for a
few hours, when I come back, half the packets of seeds are missing. They just
vanish, only to be discovered at random later in one of the ferret nests occupying
an empty drawer I hadn't looked in for a while - where the packages of seed are
lovingly intertwined with all those socks I was also missing.

Getting the ferret did solve the mouse problem, though. By the time I had
"ferretproofed" my seed collection, it was also mouse-proof.
 Storing Seed

The basic rule for storing most seed is "cool and dry." Seed stores best in the
coolest, driest place in your house. On the other hand, seeds look best and
provide the most emotional and visual gratification if displayed cheerfully and
gloriously on shelves in the living room. Take your pick.

Most books and articles on seed saving provide you a chart showing the
supposed longevity of different types of seed. Other than learning that onion-
family seed is quite short-lived, however, I haven't found such charts too useful.
They are talking about average commercial squash seed, for example. My
squash seed is from the biggest, best fruit of each plant, which was cured
optimally before cleaning. Commercial seed comes from all the fruit, which
aren't cured at all.

In my experience, prime, handcrafted, homegrown seed keeps much longer

than standard commercial seed. I expect my wellgrown squash or corn or bean
seed to last at least five years at room temperature, for example, with only
modest losses in germinating ability, and to be usable for even longer.

For longer-term storage, you can freeze seed and leave it in the freezer.


 HE FLAVRSAVR tomato was the first genetically engineered food to
be released and sold to the public. FlavrSavr contains an artificial gene designed
to inhibit the function of the gene that makes pectinase. Pectinase is an enzyme
involved in tomato softening. The idea is that, if the tomatoes stayed hard, they
could be ripened a little longer on the vine before commercial harvest. Most
commercial tomatoes are picked hard and ripened artificially. Tomatoes allowed
to ripen on the vine taste better, but are normally too soft for commercial
picking, shipping, handling, and storage.

For years both scientific and political controversy raged over whether
genetically modified food in general - and the FlavrSavr tomato in particular -
should even be allowed to be grown in open fields, let alone

sold to anyone for eating. All through the Sturm and Drang, however, my
thoughts ran on a more simple-minded level. What I was thinking was: "But this
isn't going to work."

A hard tomato isn't going to seem very much like a tomato, even if it tastes like
one. Furthermore, control of fruit ripening is a highly integrated, holistic
process. I doubted that you could change fruit softening without also having a
serious influence on flavor.

"These molecular geneticists don't know about pleiotropy," I muttered to

myself. Pleiotropy refers to the effects of genes on characteristics other than the
"primary" one. I'll discuss the concept more "genetically" later in this chapter. At
this point, let me put it in philosophical terms. Pleiotropy is a genetic version of
the ancient Taoist understanding that you cannot do just one thing.
 "The way the world works is like a bow," the Taoist sage Lao Tzu pointed out
more than two thousand years ago. "When you pull the string, the top comes
down, the bottom comes up, and all the parts move."

One day about 2,300 years ago, another ancient Chinese Taoist, Chuang Tzu,
went walking in the mountains with his friend Hui Tzu, a famous logician. They
sat and rested at the edge of a cliff, their feet dangling over, and they talked
philosophy. Hui Tzu argued that everything should be defined in terms of its
usefulness.

"What about the ground you're sitting on?" Chuang Tzu asked. "You're only
using that little piece under your rump, right?"

"That's right," said Hui Tzu.

"But what would happen if I took a shovel and dug away this little piece
behind that?" asked Chuang Tzu.

"I'd fall," Hui Tzu admitted sheepishly.

"The useful depends upon the useless," Chuang Tzu said cheerfully. "Isn't it
amazing how everything is so connected?"

Molecular geneticists tend to be mechanists. They altered just one gene

involved in softening of the tomato, figuring they would get a tomato just like
the original except that it wouldn't soften. I didn't think so. I thought that if you
change that one gene, you'd probably affect a whole lot of things, but especially
flavor.

I Meet the FlavrSavr Tomato

All doubts aside, however, the idea of the FlavrSavr fascinated me. I awaited
its release eagerly. A firm tomato wouldn't seem very much like a tomato, even if
it tasted the same. But there are plenty of hard vegetables that taste great. And
even if the tomato tasted very different, it might still taste good. It might be a
whole new vegetable. I wanted to tastetest that tomato! And I was prepared to
give the FlavrSavr every possible chance. I wouldn't even require it to conform
to any concept of what tomatoes should be. I wouldn't require it to fit into the
"tomato niche." I'd simply try it, and if it tasted good, if need be I'd think of new
niches.

And of course, I thought, I might be wrong about flavor. Maybe the tomato
would taste just like an ordinary tomato, only be firm. Who knows? Certainly
the molecular geneticists all seemed to think that was what should be expected.

For years, however, the controversy in the media was as close as I got to a
FlavrSavr tomato. Finally, in 1995, FlavrSavr was released and arrived under the
brand name of MacGREGOR'S® tomatoes at my local Safeway supermarket. I
knew when they were coming, and rushed out to the store the very first day they
became available.

There they were, at long last, sitting in a huge bin under a huge sign
proclaiming them as MacGREGOR'S®. They looked beautiful. Big and deep red
and perfect. They were firm, even though fully red. Each tomato was marked
with a little label stuck on it proudly identifying it as a MacGREGOR'S® tomato
and boldly stating, though in very fine print towards the bottom of the label,
"grown from genetically modified seeds." I bought two tomatoes, bagged them
carefully, and scampered home.

Finally came the long-awaited taste test. I cut a small slice from one of the
tomatoes and put it in my mouth. The taste was so bad I almost spewed it out by
reflex. Instead I deliberately spat the offending morsel into the trash, then
examined the tomato carefully. Had something really awful been spilled on it? I
couldn't see anything wrong with the tomato.

I cut a small chunk from the other tomato and sampled it. I spat that out too.

I told myself that it really wasn't a taste test if I didn't swallow. So I took
another very tiny portion. But I just couldn't swallow. The body rebelled. I spat
that portion out too, then rinsed my mouth out with water.

It's easy to describe the flavor of those tomatoes. They tasted like gasoline.

The tomato did not merely taste like a bad tomato. It didn't taste like a tomato
at all. There was absolutely no component in the flavor that suggested "tomato."
Had I done a blind taste test in which I was asked to categorize substances as
food or fuel, I would have classed the FlavrSavr as fuel.

In all I had read about whether the FlavrSavr should be released, no one had
mentioned that the tomato tasted awful. The tomato was subsequently
withdrawn, in order, as it was put, "to develop cultivars of more acceptable
flavor." In other words, the tomato was withdrawn because it tasted awful.

FlavrSavr was, for all practical purposes, withdrawn by my Safeway store

within three days. The tomatoes were still there in the bin, but the sign
identifying them had been removed. Those tomatoes stayed there for weeks. I
doubt if anyone ever bought them more than once except strictly by accident.
Apparently, the one bin more than satisfied the needs of the community for this
tomato for the rest of the decade.

Genetic Engineering and Genetically Modified Food

A little personal preamble to this section: I am a "classical" geneticist turned

molecular geneticist turned "traditional" plant breeder. As a young scientist I got
my start in Drosophila (fruit fly) genetics, working my way through
undergraduate school at the University of Florida in the Drosophila lab of Henry
Wallbrunn (the man to whom this book is dedicated).

While at Florida I migrated into the molecular biology lab of Arthur Koch,
then moved on to graduate work in fungal and molecular genetics with John R.
Raper at Harvard University. Thereafter followed a stint on the faculty in
Genetics and Cell Biology at University of Minnesota.

I started in the field that was almost the definition of classical genetics. My
academic work subsequently was in molecular genetics. During the last twenty
years, however, my concern has been with agroecological issues, and for the last
ten I've been doing standard plant breeding on a variety of crops.

Most "traditional" plant breeders come from backgrounds in agriculture and

have little experience in molecular biology. Most molecular geneticists come
from biochemistry traditions, and don't know much about classical genetics,
about agriculture, or about growing or breeding plants. I feel that my personal
participation in all these fields gives me an excellent perspective from which to
see genetic engineering in its full scientific, historical, and technological context.
 Genetic engineering involves one of two types of modifications. In one case,
the genetic engineers synthesize a gene deliberately designed to perform a
certain function in a certain crop. More commonly, the desired gene occurs in
some other, unrelated, organism, and they start by extracting it.

In both cases they proceed by redesigning the desired gene and linking it up to
various other genes and pieces of DNA. Some are control genes to make the
desired gene function in the right tissues in the target-crop plants. Others are
DNA snips that are good at jumping from organism to organism and inserting
into DNA. Yet others, such as genes for antibiotic resistance, are included to
facilitate identification and manipulation of the cells containing the invented-
gene construct after it has been inserted into plant cells.

Then the gene engineers insert this DNA construct into cells of the crop
growing in artificial media. The antibiotic resistance helps them select for and
identify which cells and cell lines have experienced a successful insertion of the
DNA construct into the plant cell's chromosomes. Cell lines containing the
construct must then be grown into whole plants capable of reproducing with the
new DNA as part of their genetic makeup.

The methodology of genetic engineering is basically biochemical. It involves

manipulating DNA and cells in the laboratory. The traditional and modern plant
breeding methods promulgated in this book are field methods involving whole
plants. They do not require laboratories, do not construct artificial genes, and do
not transfer any genes from any organism to any unrelated one.

These ancient as well as modern field methods transfer and manipulate genes
only

in exactly the same way as the plants themselves do, that is through normal
reproductive processes. Traditional and modern plant breeding exclusive of
genetic engineering is merely a speeding up and guiding of natural reproductive
processes. Genetic engineering though, is based upon a dramatically different
process.

Genetically modified food (or GM food for short) is food from plants that have
been bred using genetic engineering. Such plants contain laboratory-constructed
artificial genes or genes from very unrelated species.
 Some spokespeople from the GM food industry are fond of claiming that
genetic engineering is an old technique that has been going on for thousands of
years. This claim strikes me as dishonest, or at the very least deceptive. Plant
breeding using the normal reproductive processes of plants has indeed been
going on for thousands of years. But genetic engineering, which involves
physically transferring DNA from one organism to another in the laboratory, is
not merely more of the same.

Plant breeding has certain natural limitations not shared by genetic engineering.
Speaking as a plant breeder, these limitations can be frustrating. Speaking as a
consumer, these limitations are reassuring. You cannot, for example, take a gene
from peanuts and put it into corn using standard plant breeding methods. That
limitation is a matter of life or death to a person who is allergic to peanuts.

There is tremendous controversy over the ethics, safety, and social and
environmental consequences of genetic engineering in general, and genetically
modified foods in partic ular. These issues are outside the scope of this book. For
a thorough and readable appraisal and evaluation of all aspects of the subject I
recommend a little book by Luke Anderson, Genetic Engineering, Food, and Our
Environment (Chelsea Green Publishing, 1999).

In the rest of this chapter I'll focus on a practical comparison of the advantages
and limitations of standard plant breeding and genetic engineering.

Standard Plant Breeding versus Genetic Engineering

Genetic engineering involves directly manipulating DNA. You can't do it

without a specialized background and a laboratory filled with several hundred
thousand dollars worth of specialized equipment. Does this mean we gardeners
and farmers are out of the game now that genetic engineering has become a
reality?

I don't think so. In fact, there is more need for our contributions and expertise
than ever before. The fact is, there are many limitations as to what can be
accomplished with genetic engineering. Much of what we care about most as
gardeners and farmers is not very amenable to the genetic engineering approach.
And almost all of what matters in breeding plants for organic or sustainable
agriculture is best approached using standard plant breeding methods.
 An additional aspect is the matter of the scope of the vision. All our food crops
were originally created with standard plant breeding methods. All genetically
engineered varieties represent relatively small changes in thoroughly established
varieties, not whole new crops or whole new crop species.

Genetic engineering could not create corn starting from wild teosinte today.
Nor could it have created any of the other food crops we have today. Genetic
engineering is very powerful at making single changes in existing crop varieties.
But someone needs to breed the crop or the basic variety in the first place.

Genetic engineering excels at manipulating single genes. Standard plant

breeding methods work with the whole genome at once. The whole genome is
the natural level for evolution, including the human-directed evolution that
develops entirely new crop species. I believe that most of the big advances
during the lives of those of us now living - most development of whole new
crops, for example - will continue to come from standard plant breeding.

Finally, the universities and multinational seed companies have blindly jumped
upon a single genetic engineering bandwagon. They're all going in one direction.
The main driving force behind the choice of research problems seems to be
mostly a matter of whether the problem is one that can be approached by genetic
engineering - not whether the goal is important or needed.

Most plant breeders have retired or are retiring and are being replaced with
genetic engineers. Wisdom would have added genetic engineers to a full
complement of field-experienced plant breeders using standard methods. Instead,
the universities and big corporations have virtually abandoned standard plant
breeding. In doing so, they have all but abandoned many of the most interesting
and important problems.

Genetic engineering excels at physically manipulating single genes. It is at its

best in situations where single genes have a known and definitive effect on the
characteristic of interest. Only a few agricultural crops are very well known
genetically. Even in the genetically best-known crops such as corn, peas, beans,
and tomatoes, only a few of the simplest traits are very well understood.

Genetic engineering requires exact information about specific genes and a

thorough understanding of the relationship between various traits and the
underlying genes and biochemistry. In only a very few cases for a very few types
of plants do we yet know those relationships well enough for genetic engineering
to be able to produce desirable new varieties.

One of the great virtues of standard plant breeding is that we don't need to
know anything about the genes that determine the characteristic we care about,
or about what any of those genes do. We can guide a variety towards our goals
without being limited by our understanding of the underlying biochemistry.

An even more important limitation of genetic engineering, however, is that most

characteristics of agricultural importance are quantitative genetic traits. That is,
they are determined by many genes, not just one or a few.

Flavor, yield, size, shape, earliness, lateness, cold hardiness, heat tolerance,
drought resistance, and other components of ecological adaptation are all traits
that involve many genes. Such traits are not especially amenable to the
techniques of genetic engineering. These traits include nearly every
characteristic that is most important to gardeners and small-scale farmers.

The great strength of genetic engineering is its ability to change a single gene in
an established variety while leaving the entire rest of the genome intact. This is
very difficult to do with standard breeding methods. The closest we can come to
transferring single genes into an established variety is via recurrent backcrossing
(see pages 138-140). However, there are always some blocks of genes too tightly
linked to the desired gene to be separated.

A second major limitation of standard plant breeding methods is that sometimes

the gene we need doesn't exist anywhere in the whole plant species. If there is a
useful gene in mustard, and we want it in tomatoes, we're out of luck if we are
using standard plant breeding methods. Genetic engineers have the option of
isolating and modifying the mustard gene and inserting it into the tomato. In
some cases they can even synthesize an artificial gene and use it.

However, once inserted into the tomato, the alien or artificial gene may not
have the anticipated effect, and, even if it does, there may be many other effects
that were unpredicted and are unacceptable to the farmer or consumer. This
brings us back to the phenomenon of pleiotropy.
Pleiotropy refers to "secondary" effects of genes. The classic experiment was
done by Dobzhansky in the early days of Drosophila (fruit fly) genetics. In
Drosophila, there are hundreds of mutant strains available that dif fer from the
basic wild-type stock by a single mutant gene. This is possible because the
mutations arose in that specific wild-type background originally.

Each mutation has a characteristic effect on the phenotype of the flies that
carry it - such as altering the eye color to white, or causing stunted wings, or
increasing the dark pigment in the body. When you actually work with the
mutant fly stocks, though, you notice that each mutation affects other phenotypic
traits as well.

The flies have a characteristic pattern of bristles, for example. I remember

noticing how it seemed that every mutation, whatever its primary effect, seemed
also to alter the bristle patterns. You work with flies under a low-power
dissecting scope. If you're observant, you soon start wondering whether every
gene affects everything else.

Dobzhansky's experiment asked the question, "Does every gene affect every
phenotypic characteristic?" In order to address the issue fairly, he considered
many common genetic mutations, and he examined their influence on an
unrelated part of the phenotype. He chose to look at the effects of all the
mutations on the dimensions of the spermatheca.

The spermatheca is an internal organ in the fruit fly female that allows it to
store sperm between matings. The spermatheca is usually invisible, so
Dobzhansky could be assured that he was not choosing it because of
preconceived ideas.

So Dobzhansky measured the spermatheca in wild-type flies and in ones that

carried w, for example (the mutation associated with white eyes), and in many
other strains that differed from wild by just a single gene. The results were
unambiguous.

Each mutation had originally been identified and characterized by some simple
major effect on the external appearance of the fruit fly. Every gene, however,
also influenced the shape of the spermatheca.
 Everything affects everything. You can't change just one thing. Everything is
connected.

If you change any gene whatsoever in a tomato plant, this will affect the flavor
of the fruit, the yield of the plant, and every other characteristic. The real issue
is, "By how much?" and "Does it matter?" If you change a gene known to be
central to fruit ripening, of course, this will affect fruit flavor, and probably
profoundly.

Pleiotropy means that when a genetic engineer makes a change in a single

gene, the overall result is less predictable than I think most genetic engineers
these days are supposing. Concomitantly, a large number of genetically
engineered plants will probably turn out to be undesirable or unworkable in the
last stage of their development.

All this does not mean that genetic engineering is fatally flawed. It only means
that the most powerful advantage of molecular genetics - that of being able to
make simple single changes in genes - is actually less powerful than one might
suppose, because simple changes in genes will not reliably translate into simple
changes in phenotypes.

Practically, pleiotropy means that plenty of the ideas and projects that are
theoretically possible and that can be produced technically in the lab are likely to
fail at the last stage, when it comes to producing desirable foods from plants
with appropriate yields.

Flavor depends upon levels of dozens, sometimes hundreds, of components.

For that reason alone, I think flavor will always be one of the most difficult
subjects for genetic engineering to address. Yield also depends upon hundreds of
genes. I believe that many genetic engineering projects will fail because the
genetic modifications will have unavoidable effects on yields.

You can do stupid things with standard plant breeding, but you are more likely to
do stupid things with genetic engineering. It took genetic engineering to create
and release a tomato as bad as the FlavrSavr.

Whenever you make a cross to get segregants better than both parents, for
example, you also get segregants that are worse than both parents. Sometimes
much worse. I have produced plenty of weird or unpalatable flavors on
individual plants here and there in my various breeding projects.

However, the plant breeder works with whole plants and their phenotypes at
every stage of the project. Anything bad-tasting or unpalatable is eliminated as
soon as it appears. Standard-bred varieties are tastetested as they are developed.
By the time a standard-bred variety is released, it is already an old friend to its
creator, who has been eating it and feeding it to her friends and family for years.

In contrast, genetic engineers start a project based upon certain assumptions

and ideas, then spend almost the entire course of the project just manipulating
DNA and plant cells. Only at the very end of the project do

they have real plants. Only then do they get to find out whether their
assumptions and ideas were correct and discover any unpredicted side effects.

The standard plant breeder is like a person who builds a go-cart and tests it at
every significant stage of development - at each stage discarding what doesn't
work and refining what does. But the genetic engineer has to build his go-cart by
design and theory alone, without benefit of any testing. He drives it for the first
time in the race. Only if every assumption was correct and everything works
exactly as it is supposed to is the go-cart likely to run well enough to finish the
race.

Imagine further that the genetic engineer is allowed to use titanium in his go-
cart, but not wood, screws, or nails. However, he has access to jet engine parts as
well as go-cart parts....

We can breed food crops that are actually poisonous, not merely bad-tasting,
though these categories overlap. We can do this with either standard plant
breeding or genetic engineering. With standard plant breeding, there are two
contexts in which we are most likely to produce something poisonous.

Context 1: We have started a project by crossing an edible species or variety to

an inedible, poisonous, or wild one. In this case, we're usually trying to get some
useful characteristic out of the poisonous or wild material back into an edible
variety. Given a poisonous parent, poisonous progeny are expected in the project
and are selected against routinely.
 Context 2: We are selecting for insect or herbivore resistance. Most natural
plant species have toxins of various kinds that protect them from being eaten..
Often, domestication of a plant species involves selecting for lower levels of
these toxins. If we select for insect resistance, we may be selecting for higher
levels of the toxins once again. Higher toxin levels may make the plants
unpalatable and/or poisonous to humans as well.

One needs to be cautious whenever selecting for insect resistance or toxin

levels or making crosses involving high toxin levels. I don't know of any plant
breeder who has actually poisoned himself or herself, incidentally. Most plant
toxins are bitter or foultasting. You don't get poisoned because you don't eat the
plant.

A much-quoted recent plant breeding incident involves the breeding of insect-

resistant potatoes. The potatoes had such high levels of solanin in them that they
could make people sick. (Solanin is the toxin found in wild potatoes, and in the
skin of green potatoes.) These toxic potatoes were developed using standard
breeding methods. They were, as expected, never released.

It is obvious that any genetic engineering involving pest resistance or toxin

levels would have to be evaluated carefully, just as would any standard plant
breeding involving these characteristics.

Genetic engineering is engineering. The genetic engineer has some ideas and
creates a design, then attempts to impose it on the plants. Genetic engineering is
limited by knowledge and imagination. Genetic engineers can only create what
they can imagine.

Plant breeders also often start with ideas and designs and attempt to impose
them on

plants, but that stage is quickly transcended. What soon develops is more like a
cooperation between plants and person - a two-way partnership - a conversation.
Plant breeding is limited only by what it is possible for the plants to do. The
plant breeder can and often does breed things that she never could have
imagined.

Everything in This Book Is Illegal with Genetically Engineered Varieties

 There is no inherent reason why genetically engineered varieties would have to
exist in an entirely different legal framework than standard-bred ones, but the
fact is, currently, they do. To make use of the genetically engineered varieties
requires accepting an unprecedented legal situation. This new legal frame of
reference outlaws garden-and farmbased seed saving and plant breeding.

Genetic engineering, as it is currently practiced, produces patented genes.

Plant Variety Protection (PVP) does not patent individual genes. It only gives
limited protection to a whole variety, that is, to a specific arrangement of genes. I
can legally use a PVP variety in my own breeding program to breed something
even better. Breeders may use the material without restriction to create the next
generation of varieties.

in contrast, the new genetically modified genes themselves are patented. You
are legally not allowed to save seed, not even for your home garden. You are
legally not allowed to make any crosses with the varieties or do any breeding
with them. There can be no farmand garden-based plant breeding with these
varieties, no adapting them to specific regions or growing conditions.

With patented genes, everything I'm encouraging you to do in this book is

illegal.

With the invention of "terminator" technology, companies like Monsanto, owner

of the patent on it, can envision making it physically impossible for gardeners
and farmers to save seed, not merely illegal. Seeds with terminator technology
carry inactive genes that kill plants in the embryo stage. The genes are activated
when the seeds are treated with a triggering chemical such as tetracycline. The
treated seed produces plants whose seed is sterile. This allows a company to sell
seeds that are only capable of producing one generation.

Monsanto, in response to considerable bad press and a furious public outcry,

has stated that they will not develop this terminator technology for license to
others. However, I have seen nothing in their statements that promises that they
will not incorporate terminator technology into their own varieties.

I think terminator technology is so attractive to seed companies that unless we

specifically outlaw it, we will soon have it out in our fields. We need laws that
make the use or possession of terminator technology a criminal (not merely a
civil) offense, and one involving serious prison time.

I view terminator technology as the genetic engineering version of poisoning

wells. Those who develop terminator technology may have some legitimate
purposes they think they are trying to serve. So, undoubtedly, did most of those
who poisoned wells.

Genetic Engineering and Sustainable Agriculture

I believe genetic engineering has tremendous potential to provide tools for

breeding for sustainable agriculture. However, I am underwhelmed by its current
tally of actual accomplishments. I would be impressed by a new crop that grew
so well it outgrew all the weeds. I am not thrilled by a new crop that has a minor
change that allows you to dump huge amounts of this herbicide on it instead of
huge amounts of that one.

I believe most breeding for sustainability is better and more easily done by
standard methods than through genetic engineering. There are some exceptions.
One field of genetic engineering that has great potential for sustainable
agriculture is the development of new genes for viral resistance in plants. Viral
resistance can also be developed with standard methods, of course. But, as the
Chinese say, why not walk with both legs?

In addition, there are other things I can also imagine that would be
sustainabilityenhancing, and that could be accomplished only with genetic
engineering.

However, I cannot imagine any circumstance in which I would be willing to

grow a variety whose seed I am not allowed to save. Any advantage to
sustainable agriculture viewed at the level of the specific crop would, in my
eyes, be more than offset by the sustainability-destroying bigger picture.

Analyzing the virtues or liabilities of herbicide resistance and insect resistance

is much more problematic than analyzing viral resistance. You can argue that
herbicide-resistant plants constitute sustainable agriculture if it means you'll now
be dumping huge amounts of Roundup on the plants instead of huge amounts of
something even worse. Similarly, herbicides used with no-till farming methods
can be considered sustainable if you're cultivating a hillside where the soil
washes away when it is tilled.

In my eyes, these arguments are comparable to those of the man who says he's
treating his wife compassionately these days, because he has started beating her
with a horsewhip, whereas he used to use baseball bats and chains.

If a farmer is abusing his land, his neighbors, and the environment badly
enough, and some of the new genes allow him to continue abusing them, but less
badly, he may claim this is a move toward sustainability, relatively speaking. He
might even be right. But he shouldn't expect the rest of us to cheer.

Meanwhile, back in the kitchen, I'm not eating genetically modified foods. I am
both allergic to and intolerant of wheat gluten. Most prepared foods contain at
least small amounts of wheat gluten. Small amounts are usually not listed on
labels. This fact has caused me serious and repeated grief. We need laws that
require full labeling of every ingredient in food, GM or not. As it is, our labeling
laws are inadequate. Unlabeled GM food very much exacerbates an already bad
situation.

Until full labeling is required by law and is on all GM foods, I am voting with
my dollars. I am avoiding GM foods. I'm buying organic. I'm growing my own.
And I am starting my own company to sell fully labeled non-GM foods - starting
with corn meal and corn products made from flour corns and flint corns of
named varieties. (See specificfoods.com in Appendix D.)

 ANDWICHSLICE" is a new class of summer squash. It is unusually
delicious as a raw squash, and it is delicious at a stage that is much bigger than
optimal for flavor for other summer squashes. It is best when 4 to 6 inches long
and 3 to 5 inches across. Slices of the squash are the perfect size to fit into
hamburger buns or between slices of bread for sandwiches.

The squash has such an unusual firm, crisp texture, you don't even need the
bread for a sandwich. This is wonderful for me, because I love sandwiches, but I
can't eat wheat. I can put a slice of cheese between two slices of squash. I can
even make an allvegetable sandwich - a thick slice of 'Brandywine' tomato and a
thin slice of Walla Walla' sweet onion in between two slices of Sandwichslice.
Or I can cut the squash lengthwise and in half, to get a shape that fits perfectly
inside a taco shell.

Small chunks of the squash are wonderful with dips or cheese. They're great in
"cucumber" salads. I even eat a lot of Sandwichslice just plain and out of hand.

The first Sandwichslice squash plant turned up in my squash patch in 1998 and
began producing the first Sandwichslice squashes in late July. This was just a
year after I started the breeding project. Or maybe no years at all. You see, I
never started a breeding project to create Sandwichslice.

It never would have occurred to me that a summer squash could taste so good,
especially at such a large size. Had anyone suggested I try to develop
Sandwichslice, I would have dismissed the idea at once on the grounds that it
was impossible. I didn't even like summer squash. Up until Sandwichslice I
didn't grow them. I certainly didn't like them raw.

In 1998, the Year of the Sandwichslice, I wasn't trying to breed summer

squash. I was trying to breed a bigger and better `Sugar Loaf' winter squash.
`Sugar Loaf' is a winter squash that was bred by Jim Baggett at Oregon State
University. It belongs to the species Cucurbita pepo, which also includes most
Halloween pumpkins, most summer squash, most acorns, `Spaghetti', `Small
Sugar Pie', and the delicatas. `Sugar Loaf' is a delicata type. `Delicata' and
`Sweet Dumpling' are other members of this flavor class.

`Sugar Loaf' is sweeter and more intensely flavored than other delicatas, and
the flavor is wonderful and distinctive. `Sugar Loaf' is also more fine-grained,
and has drier, denser flesh than any other pepo variety I've tried.

Winter squash is a major staple for me. Properly grown winter squash of a
gourmet variety such as `Sugar Loaf' is grand eating. The problem is, `Sugar
Loaf' and the other delicatas are small. The biggest fruits only run about 2
pounds. By the time you clean them there is just a few bites of meat.

Many people who use vegetables in small amounts as tiny side courses like
these small fruits. But the small fruits drive me crazy. I need a fruit big enough
to provide the main dose of carbohydrates and calories for the meal, not a tiny
little side course.

Furthermore, `Sugar Loaf' is harder to clean than most fruits, and the flesh is
only about half an inch thick. I like the flavor of `Sugar Loaf' as well as that of
any squash whatsoever of any species. But I frequently end up eating just the
biggest few and letting the average ones go to waste because they are just such
trouble to clean for the amount of meat in them.

Sometimes some of my `Sugar Loaf' fruits don't even make it out of the field.
It's lots harder to harvest, transport, and store little squash than the same number
of pounds of bigger squash. At the end of the season, amid all the harvesting
pressure and physical fatigue, I tend to leave small squash in the field.

I am often tempted not to grow `Sugar Loaf' at all, but I really can't bring
myself to do that. The flavor is so spectacular. In addition, these squash have a
unique niche.

`Sugar Loaf' and the other pepo "winter" squashes are really "fall" squashes.
The big winter squash of good eating quality, such as `Sweet Meat' and all the
Hubbards, belong to the species Cucurbita maxima. Maximas mature late and
require a full month of curing for optimal quality and storage. Pepo varieties
usually mature earlier, and are prime within a few days of harvest. I need a good
pepo in order to have squash to eat during the fall while my big maximas are
finishing off in the field and then curing indoors.

I really love and need those `Sugar Loaf' fruits. But I need them to be bigger. A
4-to 6-pound version would be ideal for my purposes. That's small enough to fix
in the microwave oven as a meal for one or two serious squash-eaters. Even a 3-
pound fruit would be workable if I could get thicker flesh at the same time.

There are other reasons to do some breeding work with `Sugar Loaf'. It is not
especially early. The seed is small, and it doesn't germinate very well from the
earlier plantings I prefer. The plant is not vigorous compared to some. The vines
grow out to about 3 or 4 feet. Compared to much of what I grow, either bushes
or vines, `Sugar Loaf' is not very impressive.

It would be nice if I could develop a variety with the `Sugar Loaf' flavor, but
with bigger fruits and larger, more vigorous vines. I'd also like it to be able to
germinate in cold mud and to grow better in cold weather. I wouldn't mind
thicker flesh, either.

So how do I go about getting a bigger and better `Sugar Loaf'?

Why Not Just Select?

One classic approach to getting a bigger `Sugar Loaf' would be to just select
within the `Sugar Loaf' variety for bigger fruit size. Simple selection is both the
most basic and one of the most powerful of plant breeding methods under many
circumstances.

A couple of observations, though, suggested that the "just select" approach was
unlikely to be very effective in this particular case.

First, `Sugar Loaf' plants don't show much variation in fruit size from plant to
plant. What size differences there are seem to be obviously related to the
variations in soil fertility or spacing from one plant to the next.

Second, I've saved true-self-pollinations from plants here and there over the
years. By a "true-self-pollination" I mean that the flower was pollinated by
pollen from the exact same plant. When the plant at the end

of the season seemed to be giving bigger fruits than average, I would plant that
seed out the next year along with "straight-run" - that is, unselected seed.

I've never seen any obvious difference. In other words, seed from plants with
the biggest fruits doesn't give me any bigger fruits than straight-run seed.

These two observations suggested that there is not much genetic heterogeneity
for fruit size in the `Sugar Loaf' variety. Selection works by changing gene
frequencies from generation to generation. It isn't effective when the breeding
population is genetically uniform for the characteristic of interest.

I've discussed the basic genetic relationship between selection and genetic
heterogeneity in Chapter 10 and elaborated further in Chapter 20. This time, let
me forego more scientific explanations and use an analogy.

Suppose you have a basket full of balls, some golf balls and some tennis balls.
You pour part of the balls into a second basket through a screen that is of
variable size mesh. The screen bounces away or retains most of the tennis balls,
but only a few golf balls. The second basket will have a higher proportion of golf
balls than the basket you started with.

The first basket is the collection of genes in a variety that has genetic
heterogeneity for size. The screen is your selection method. The second basket is
the collection of genes in the next generation.

Now imagine instead that you start with a basket that contains only tennis
balls, and pour them through the screen into another basket. The second basket
will contain all tennis balls too. It doesn't matter what kind of screen you use;
you still get another basket full of tennis balls. The original basket that has all
tennis balls contains no underlying heterogeneity for ball size or type. Selecting
based upon that basket just doesn't get you anywhere.

It's likely that there is a little genetic heterogeneity for size in `Sugar Loaf'. If
so, if I worked very hard at "just selecting" for a number of years I could
probably increase the average fruit size a little. But I want much bigger fruits. I
didn't think there was enough genetic heterogeneity in `Sugar Loaf' to get a very
much bigger fruit via simple selection.
 Choosing the Right Cross

To get a bigger `Sugar Loaf', then, the obvious method is to start with a cross.
The purpose of the cross is to introduce genetic heterogeneity for fruit size, and
to contribute genes that will enhance fruit size as much as possible.

The simplest scheme is to cross `Sugar Loaf' to something with much bigger
fruits - the bigger the better. If there were something much bigger that had good
fruit quality, it would be a logical candidate. None of the C. pepo varieties with
good flesh quality (as winter squash) are very big, though.

The biggest pepo I'm familiar with that has good fruit quality is `Small Sugar',
which is only about 6 pounds. Like `Sugar Loaf', `Small Sugar' doesn't have
especially vigorous vines, and doesn't germinate or grow well in cool weather.

In choosing the right partner for a cross, everything you know about plant
breeding and gardening and growing plants comes into play. All the varieties you
have trialed in your

life help. They give you ideas, extend your understanding, and become the
primary candidates for involvement in your breeding projects.

I decided to cross `Sugar Loaf' to a big Halloween pumpkin, then recover good
fruit quality by recurrent backcrossing to `Sugar Loaf'. I had a candidate in mind
for crossing to `Sugar Loaf', one that has all kinds of agricultural characteristics
that I would love to have in a new variety. It was `Connecticut Field'.

`Connecticut Field' interested me for myriad reasons beyond just fruit size. The
other factors have to do with my vision of a more sustainable agriculture. I want
to breed crops to facilitate the change to a more sustainable agriculture. I want to
breed varieties that specifically enhance the viability and profitability of the
farms that are practicing sustainability.

In addition, I believe sustainable agriculture requires agriculture, not

agribusiness. I think true agriculture is usually only possible on farms smaller
than those that have become associated with the agribusiness pattern. So I
consider that anything I can breed that creates special opportunities for small
farmers, organic farmers, or family farmers is a useful contribution.

I can best describe some of these ideas by talking about my concept of an

agroecologically "ideal squash" and `Connecticut Field'.

The Agroecological Virtues of a Squash

`Connecticut Field' is a classic orange Halloween pumpkin. At up to about 30

pounds with no special treatment, the variety is one of the biggest-fruited pepo
varieties. There are several orange Halloween pumpkins that are listed as
producing fruits of a hundred pounds or more, but these are C. maxima varieties.
I can't cross them with the C. pepo variety `Sugar Loaf'.

None of the big C. pepo Halloween pumpkins have good flesh quality. This is
no surprise, since they are sold by the pound, and it is cheaper to produce water
than dry weight. The most productive ornamental varieties always have watery
flesh compared to the much smaller pie pumpkins. `Connecticut Field' has fairly
thick flesh that is watery, coarse, and bland.

`Connecticut Field' is very vigorous under my growing conditions. Even

commercial seed usually germinates vigorously from early plantings. The vines
are vigorous. The leaves are huge. The seeds are big. The fruits are early for
such a big fruit - often a full month earlier than anything of even nearcomparable
size.

The variety is an old public-domain heirloom. I can make crosses with it

without legal problems or repercussions.

The plants seem to grow better in cold weather than most squash or pumpkins.
Here in western Oregon, temperatures fall into the forties every night for the first
third of the growing season. Days with highs in the sixties are more common
than those with highs in the seventies until well into the growing season.

I like to plant my squash in April. I like to have the plants well established
before the cucumber beetle population gets too high. In addition, the early
planting allows me to ignore irrigation for the first month or more, because the
ground is still moist from winter rains.
 With later plantings, the small plants are not able to reach down into the moist
layer of soil, which recedes day by day. Later-planted squash need more
watering and pampering. They often fail completely because of cucumber
beetles, flea beetles, or water problems. Later plantings work better with
transplantings than with direct seeding.

Here's my idea of the perfect growing methods and the perfect variety: In early
spring before any of the insects are out in significant numbers, I plant the seeds
of this ideal variety by direct seeding. Then I go away. I come back at the end of
the season and harvest the fruit.

The plants would be so deeprooted that they would be able to tap into the
water table and need no irrigation. The vines would be vigorous and spreading,
and would occupy the ground they covered so thoroughly that weeds would be
shaded out. No weeding would be necessary.

Actually, I would undoubtedly visit the squash often for breeding work and to
handpollinate for seed saving. But I enjoy visiting and observing and breeding
plants lots more than I enjoy hard physical labor.

In reality, I would have to irrigate some, because western Oregon gets no

effective rain in the summer after about the beginning of May. The top several
feet of soil dry out completely thereafter, with moisture receding progressively
day by day. Even if plants can reach deeper water, this may not be adequate for
good growth. It is the process of expand ing into and growing in fertile soil that
provides nutrients to the plant. This fertile soil is usually located in the upper
layers.

I can imagine squash plants so deeprooted and aggressive that they need no
water in the first half of the season, though. And perhaps these squash would
need only a good soaking a few times in the last half of the season - instead of
the twice-weekly watering usually considered minimal here.

I can also imagine such vigorous, spreading vines that one weeding around the
plants early in the season and one tilling between the hills would be all that was
needed. After that the plants would take over and shade out any weeds.

The agroecologically ideal squash variety would be able to perform

magnificently under organic growing conditions, of course. And it would have
such an extensive root system and efficiency at utilizing available nutrients that
it would be able to grow optimally on more modest levels of nutrients than most
squash require.

It's possible that the ideal squash would be better at establishing useful
associations with root mycorrhizae than most squash.

My ideal squash would have big seeds. The squash varieties that germinate well
and grow well enough to establish plants from my early plantings all have quite
big seeds.

There are a number of factors that might be involved. First, those early-planted
squash often don't get much direct sun. Sometimes it rains or is cloudy nearly
every hour of every day for weeks. Under these conditions, a big seed with
plenty of stored food may be needed to give the seedling a good start.

In addition, my squash get only organic fertilizers, which are not very high in
nitrogen. Nitrogen is a particular problem in early spring soils in western Oregon
because, unlike most soil nutrients, the nitrogen is part of compounds that are
water-soluble. The nitrogen is leached out of our soils by our heavy winter rains.

Organically grown crops are dependent upon new nitrogen newly released
from organic matter in the soil. The release of nitrogen may be from old or
freshly added organic matter, but it requires microbial action. In the cold, wet
soil of a western Oregon spring, however, there isn't much microbial activity.
Nitrogen availability for organic crops is a problem.

Perhaps big seeds allow my squash plants to get established based upon
internal reserves of nitrogen, instead of needing much nitrogen from the soil
initially.

`Connecticut Field' has huge seeds, and germinates very vigorously from those
early plantings. `Sugar Loaf' has little seeds, and germinates very poorly from
early plantings.

(My observation about the importance of big seeds applies only to my early
plantings. In later plantings, when the weather is warm and sunny, there is not
any very obvious correlation between seed size of a variety and germinating
vigor.)

My ideal variety would have big leaves. It would grow well in cold, cloudy
weather.

Many varieties that do manage to germinate from my early plantings don't

thrive thereafter. They germinate, but the seedlings quit growing at about the size
that represents their having used up the stored food in the seed. After a while, the
root rots or the plant just disappears. If the plant survives, it becomes a stunted
plant that makes no fruit.

Plants that establish themselves well and grow vigorously during that early
period all have big leaves. I don't know of any smallleaved variety that
establishes itself well in my early plantings. I am guessing that this is because
most days are cloudy, and the light is of the low levels that come through clouds,
not the high levels associated with direct sun. With little actual sunshine, perhaps
big leaves are needed in order for the plant to be able to do enough
photosynthesis to support its growth.

`Connecticut Field' has huge leaves, up to more than a foot across. `Sugar
Loaf' has little leaves about 5 inches across. When an occasional `Sugar Loaf'
seed does manage to germinate from an early planting, the seedling usually dies
or develops into a stunted plant. Later plantings of `Sugar Loaf' germinate and
grow well. If they don't get wiped out by cucumber beetles, that is.

I could speculate that squash plants with deep green leaves ought to do better in
those cloudy, early-spring days than those with paler green leaves. Deep green
leaves presumably have more chlorophyll. It seems reasonable to suppose that
they would be able to conduct more photosynthesis in low light levels.

I've frequently had deep green leaf color and lighter color segregating side by
side in crosses, however, and the deep-green-leaved plants do not appear to have
any advantage.

Often, deep green color can indicate better

nutrient utilization, especially of nitrogen. So, for example, within a breeding

project, I note the greenest corn plants, because these are probably the ones that
are obtaining higher amounts of nitrogen from the soil. They may be better
rooted, or better at absorbing nutrients, or, possibly, better at establishing
mycorrhizal associations that help them acquire nutrients. The pattern doesn't
seem to work for squash, though.

I've also speculated that mottled leaves should impair the plant relative to solid
green ones (with more chlorophyll, presumably). But this doesn't check out in
the field either. When I've had mottled leaves and solid ones segregating from
crosses, there is no correlation between the mottling and vigor or growth of the
plants from any kind of planting.

My ideal squash would have deep, extensive roots. I think root growth patterns
are related to and established by some of the same genes as vine growth patterns.
That is, I think varieties with big vigorous vines also have big vigorous root
systems, and bush varieties have shallower and smaller root systems. If I'm right
about this, the convenience of bush squash in some circumstances comes at the
expense of their ability to tap into deeper moisture and fertility.

I've done some work with germinating various kinds of squash seeds rolled up
in damp paper towels that confirms my ideas about the correlation between shoot
and root growth patterns. I orient all the squash seeds in one direction on a paper
towel, cover them with another paper towel, then add just enough water to create
dampness, not wet ness. Then I roll up the paper towels with seeds, and stack the
roll on its end in a sealed container. The roots emerge and head downward
between the paper-towel layers. I open the container once a day and blow into it
to provide air exchange.

Within a few days, there are seedlings with nice little root systems. If a variety
is a vining type it makes a taproot that is much longer than the shoot. The taproot
has just a few branches, and they are short compared to the main root. That is,
the root displays strong apical dominance.

If a variety is a bush type it makes a root about the same size as the shoot. And
the root is very bushy. There is little apical dominance. That is, there are many
branches, and branches are almost as long as the bigger root they emerge from.

Seedlings from Fz seed that is segregating for bush character show roots
ranging from strong taproots to bush types. In other words, the seedlings show
segregation for root type.

Under my growing conditions, vining varieties are much more likely to survive
and thrive than bush varieties. I think this is because I don't start watering until
summer, and the moisture content in the upper foot of soil is quite variable. The
vine seedlings probably reach below this layer in just a few days.

Bushes cannot be distinguished from vines in the field until the plants are more
than a foot across. Both vines and bushes start out looking like bushes. After a
month or so, the vines start running while the bushes continue to grow as bushes.
Apparently the root systems are different right from the beginning, however.

Vine form correlates with good early estab

lishment of seedlings in my early plantings. If I have vines and bushes

segregating in an F2, for example, and I plant heavily and thin to the biggest 10
percent of the plants, I will have discarded virtually all the bushes before I can
even identify them based upon the top of the plant. Instead of getting'/4 bushes, I
might get none, or say, one in ten.

If I want bushes out of the cross, I need to plant the seed sparsely so I can keep
every plant until the plants are a foot or two across. At this point, the vines start
to run, but the bushes keep acting like bushes. Only then can I identify the
bushes and thin so as to retain the best bushes.

If I want vines, I don't have to fret that only '/4 of the plants will be vines. I just
plant excess seed in each hill, then thin to the biggest plant per hill when the
plants are about 4 inches high or so. They'll almost all turn out to be vines.

I wait till the plants are 4 inches high or more to give them a chance to grow
based upon their own root systems for a while. Early emergence of seedlings for
bushes seems to be just as good as for vines. The seedlings are using stored food,
not their root systems. The extra vigor of the vines only shows up after the plants
have had to provide their own support for a while.

`Connecticut Field' generally sends out two long, vigorous, rambling vines per
plant. If the root conformation of `Connecticut Field' resembles the vines, these
roots should be able to reach far down into the soil for water.

I prefer vigorous spreading vines to vigorous rambling vines, though. A

spreading vine makes more branches and more fully occupies a given amount of
space instead of just sending a couple of big vines shooting through the vicinity.
The rambling vines leave plenty of room for weeds to grow around and between
the vines. This is inconvenient. The wandering vines prevent the use of a tiller
between the rows, but don't shade the ground thoroughly enough to solve the
weeding problem.

`Sugar Loaf' only grows out to about 4 feet or so, but it has more branched and
spreading vines that more thoroughly occupy the space they take than do the
vines of `Connecticut Field'.

I hoped that from a cross between `Connecticut Field' and `Sugar Loaf' I might
develop a variety with the branching pattern of `Sugar Loaf' and the vine and
leaf size of `Connecticut Field'. This would result in a big, vigorous, spreading
vine form of growth.

The vine style of growth facilitates interplanting. Bush squash, the beloveds of
agribusiness, make a dense shade and are high enough that they just don't
interplant well. I grow a lot of Indian-style corns that work well in hills planted
four feet apart or so in all directions, with up to four corn plants per hill. Vining
squash can make good use of the space between those hills. (For an article on
some of my corn work, see the Bibliography)

Interplanting is easy to do in small operations that are planted by hand, but

much harder to do on an agribusiness level. Bush plants are much easier for
agribusiness to deal with than rampant vines.

For someone who plants by hand, big plants can be less labor-intensive than
small ones, as

long as they yield as well per square foot of area. If I can get the same yield from
one big spreading vine that takes up 100 square feet as from ten plants that take
up 10 square feet each, the big plant makes for a whole lot less work. That's just
one plant that has to be planted as well as weeded around early in the season.
The rest of the space can be tilled until the vines spread.
Agroecological adaptation has to be defined in terms of both a specific growing
region as well as a specific planting and growing style. If I plant later, that
changes everything. So do any changes in growing style, such as going from
overhead watering to row drip, or from direct seeding to transplanting.

If I transplanted, for example, bush plants might be optimal. A more restricted,

denser root system might be easier to confine and transplant without damage
than a sparser, more extensive, "wilder" one.

(However, if I were going to start planting by transplanting, I would test this

idea and see if it is actually true.)

Whether vines with their (presumably) extensive but sparse root systems make
better use of soil fertility than bush plants, with their (presumably) shallower but
denser root systems is going to depend upon where the fertility and water is. If
all the fertility is in the top couple of feet of soil, the bushes might do better if
they do a better job of colonizing that soil layer with their roots.

Furthermore, these thoughts about roots are based upon guesses, assumptions,
and looking at some seedlings in paper towels. Perhaps the vines thoroughly
colonize the upper layers of soil after they finish shooting their big taproots
through. It's amazing how little we understand about roots and water and
fertility.

The important thing is to think about and understand your soil, climate, weather,
plants, and growing style. Then figure out how to build your understanding into
your breeding. Let your mind and ideas and speculations flow free. But always
test all the speculations with direct experiments and observation in the garden or
field.

What about fungicide-treated seed? Well, I eschew it. I want germinating vigor
incorporated into the genes of the seed. And I want natural vigor to soilborne
fungi that matter in my region. Treated seed makes it impossible to select for or
maintain good natural resistance or vigor.

Treated seed is like a crutch. If you use it, you will have to continue using it.
You won't learn which varieties are adapted to your region. You won't learn
appropriate planting styles. And you won't be doing seed saving or plant
breeding.

I use treated seed of a variety only when nothing else is available, and only for
the time it takes to save my own untreated seed. If the variety can't grow without
seed treatments from home-saved seed it doesn't belong in my field.

Treated commercial squash seed seems to germinate very poorly in my early

plantings, incidentally - much worse than ordinary untreated commercial seed,
and very much worse than my homegrown seeds. I don't know why. If I had to
depend upon commer cial treated seed, my early-planting style would be
completely impractical.
 The Grand Plan

My plan was to cross `Connecticut Field' and `Sugar Loaf', then do recurrent
backcrossing to `Sugar Loaf' some number of generations. Then I would look for
a plant with `Sugar Loaf' flavor and quality that could be inbred to establish a
new variety.

In breeding for a particular flavor, I usually find it most practical to backcross

to the variety with the desired flavor for at least one or two generations. To
obtain specific flavors, I think you are usually dealing with a dozen or more
genes simultaneously. To recover a specific flavor from an Fz isn't very likely
without growing out thousands of plants. (See Chapter 10 for how to evaluate
various plantbreeding tactics.) `Connecticut Field' and `Sugar Loaf' probably
differ by hundreds of genes that have serious effects on fruit flavor and quality.

An F, plant of `Connecticut Field' and `Sugar Loaf' would get half its genes
from each parent. If that F, is backcrossed to `Sugar Loaf', the offspring will be
3/4 `Sugar Loaf'. If some of those offspring are backcrossed to `Sugar Loaf' yet
again, the resulting offspring will have 78 of their genes identical to `Sugar
Loaf'. (See Chapter 10.)

In a population that is '8 `Sugar Loaf', it should be much easier to find plants
with `Sugar Loaf' flavor. With a little luck and a little gentle selection along the
way, the `8 genome from `Connecticut Field' would contribute some fruit size
and some advantageous agroecological characteristics as well.

This backcrossing scheme, with a minor hiccup at the beginning, is still in

progress. The rest of this story focuses on something else - something
unexpected that came out of the project.
 Choosing the Cytoplasm

My plan was to cross `Connecticut Field' female flowers with pollen from
`Sugar Loaf'. The reason for doing the cross in the direction described - that is,
using `Connecticut Field' female flowers, not `Sugar Loaf' ones - has to do with
the cytoplasm. We determine which cytoplasm we retain when we choose which
plant to use as the female parent.

The genes we usually talk about are nearly all nuclear genes - that is, they are
located on the chromosomes in the cell nucleus. But there are also a few genes
located in certain organelles in the cytoplasm, and these aren't inherited in a
Mendelian fashion at all. They are maternal. They come from the mother only.

In plants as well as animals, some genes are located in the mitochondria, and
are involved in energy metabolism. Plants have chloroplasts, as well, which also
have some of their own genes. Chloroplasts are the central organelle involved in
photosynthesis.

If I do the cross using `Connecticut Field' female flowers, the F, plants will
receive all their chloroplast genes, all their mitochondrial genes, and one dose of
nuclear (Mendelianstyle) genes from `Connecticut Field'. In addition, they will
receive one dose of nuclear genes from `Sugar Loaf', the pollen-parent.

If I use `Sugar Loaf' female flowers, I'll get all the mitochondrial and
chloroplast genes from `Sugar Loaf' instead.

Plant breeders tend to ignore the possibility of serious genetic differences in

the genes located in the cytoplasm. They usually make crosses in whatever
direction is most convenient technically. I think though, that we should be
paying more attention to the role of the cytoplasm.

`Connecticut Field' germinates and grows better in cold weather and under low
light conditions than most varieties. It is quite possible that those characteristics
could involve genes in the cytoplasm. So I wanted the cytoplasm from
`Connecticut Field'.

Disease resistance is sometimes also known to be cytoplasmic. Whenever I'm

doing a cross involving disease resistance, I use female flowers from the
resistant parent, just in case the resistance is associated with the cytoplasm.

Whenever I cross a rare variety to a common one, other factors being equal, I
try to preserve the cytoplasm of the rare variety. I think we need a greater
biodiversity of cytoplasms in our gardens and on our farms as well as a greater
diversity of nuclear genetic combinations.
 The Reality

In 1997, 1 planted both `Connecticut Field' and `Sugar Loaf' so I could do the
beginning cross. The `Connecticut Field' grew happily. The `Sugar Loaf' didn't
come up and had to be replanted, as usual.

The `Sugar Loaf' flowered too late to make the cross. I pulled flowers off the
'Connecticut Field' to keep it flowering. Finally, late in the season, instead of
losing the year and wasting the `Connecticut Field', I crossed it to something else
I had around that was half `Sugar Loaf'. The half-`Sugar Loaf' was an F, between
`Sugar Loaf' and a bush acorn. It was from an entirely different breeding project.

I used the `Connecticut Field' as the female parent, as desired. So my cross

gave me seed that carried `Connecticut Field' cytoplasm and a nuclear genome
that was '2 'Connecticut Field', '4 `Sugar Loaf', and '/4 bush acorn.

In addition, each seed would be expected to be genetically different from every

other. Each represents hundreds of different genes segregating from the `Sugar
Loaf'/acorn F, then combined with a haploid genetic complement from the big
pumpkin. A row of plants grown from such seeds should be as wild and weird as
it gets.

In 1998, 1 swayed back and forth right up until planting time about whether to
plant the "weird stuff," or to forget that mess and start over with `Connecticut
Field' and `Sugar Loaf' and try again for the simpler cross. The simpler cross
would require only a few plants in the next generation, because an F, is fairly
uniform. The cross I had made, however, would be segregating for `Sugar Loaf'
and acorn characters. That meant I would need more plants in order to retain the
`Sugar Loaf' characteristics. And the material was only '/4 `Sugar Loaf'.

The day that I planted the squash, I was still undecided. I finally went ahead
and planted a whole 60-foot row of the "weird" stuff. There were so many
interesting characteristics in that material, I couldn't resist planting it just to see
what it would do.

I halfway regretted "wasting" the row almost immediately. I only had about ten
rows, and squash is one of my main winter carbohydrate and vegetable staples.
A 60-foot row is a lot of space to produce plants I'm not expecting to be able to
eat. A cross of something with inferior fruit to something with superior ones
usually gives you inferior fruit in the first generation. The acorn/`Sugar Loaf' F,
was a good-quality delicata type, but the `Connecticut Field' was not. I expected
only inedible offspring in the first generation.

This was one of my standard early plantings. Wherever I had enough seed, I
planted six seeds per hill, to be thinned to the best plant. I do this for pure
varieties as well as my experimental material. If you plant lots of seed and
eliminate most plants, keeping just the best, you get strong selection for ability
to germinate under your conditions. If you conserve seed and avoid wasting it,
you'll be keeping nearly every plant that manages to come up. That gives you
little selection for germinating ability or early growth.

Then we had three weeks of winter. It rained. And it rained. We had

temperatures down into the forties every night, often not much better than the
fifties during the day. There may have been a few minutes of sun on one day, but
not much more. I would probably need to replant everything.

By then I had changed my mind about the weird material. With the late
planting, I needed more space for food, since I always lose so much from late
plantings. So I headed back out to the field at the first break in the weather,
figuring I'd get that weird row redone and out of the way. I would overplant with
something else, something I could eat this year, not just curiosity and dreams.

So I drove over to the field, draped my big peasant's hoe over my shoulder,
filled my fanny pack with the packets of seeds, and slogged out to the field. If
any plants of the "weird stuff" were up, I had decided, I would hoe them under
and replant anyway.
 The Squash Speak

When I got out to the field, the situation was as bad as I had thought. Row 1,
no plants. Row 2, a few straggling plants here and there, only from one of my
breeding projects, nothing from standard varieties. Only at best one seedling per
hill of the six seeds planted. Most hills with nothing. None of the few existing
seedlings looking very happy. Row 3, nil.

Row4....

"Hey, hey, hey! Hello!" yelled a vigorous row of crowded little plants.

"Well, hello ..." I responded. I had never been accosted by a bunch of squash
seedlings before, so I was a bit taken aback.

There was the most beautiful row of squash seedlings I had ever seen. There
were five or six seedlings in every hill. The single row of a dozen hills had about
seventy vigorous plants. The entire rest of the field, of ten rows, only had
perhaps a dozen plants total. And these plants already had their first real leaves.

"Whatcha up to, Person? Whatcha doin' with that big hoe?" So asked the
nearest seedling, one in a hill about a third of the way down the row.

"Oh, er, this?" I said. "Heh heh. Nothing. Nothing at all. I just like to carry it
around." I quickly set the hoe aside.

I did no replanting that day. Those plants had earned their place in the field.
 Carol Falls in Love

For a long time I walked up and down that row of little plants, squatting here
and there and just gazing, contemplating. Fascination. Awe. Admiration.
Commitment. Bonding. It was a bit like participating in a birthing. Something
important is happening. Something is coming into being. I don't know what it is,
but I know my role. I will nurture, guide, facilitate.

The rains continued, and turned into six extra weeks of winter. There was
essentially no sun. Every day was cold and rainy or overcast. It never actually
got to below freezing at night, but it never got very far above freezing either.

Finally the sun came out, and I headed over to the field to replant. Nearly all
the seedlings that had come up during the initial cold had subsequently died. The
entire squash patch was empty except for the weird material and parts of three
other rows, all involving other breeding projects of mine.

The weird row was baby plants now. They had managed to grow with virtually
no sunshine at all.

"How did you do that?" I asked the squash in the nearest hill. "How did you
grow so big with no sun?"

"Ha ha! There was plenty of sun! There were several minutes of sun back there
in the second week. Enough for anybody!"

"But what about the cold?" I asked.

"Very invigorating," said the squash.

"Squash aren't supposed to like cold."

"They aren't?"

"No, look," I said, pointing around at the empty rows. "Everybody else died."

"Good thing those wimps are out of the way," said the squash. "We're going to
need the space."
I replanted all the empty sections in the other rows. Then I thinned the weird
row. It was painful, but it had to be done.

I was expecting segregation for bush versus vine type. With a heterozygous
bush/vine crossed to a vine, I expected half heterozygous bushes and half vines.
I wanted vines. As mentioned earlier, I can select heavily for vines simply by
thinning to the biggest plants when the plants are about 4 inches or larger.

So my first round of selection was simple. I thinned each hill from five or six
young plants to the biggest one. This gave me a dozen plants in the row, just one
per hill. A month or so later, the result was eleven vines and one bush, which I
eliminated.
 Disaster and Opportunity

The ancient Chinese symbol for "crisis" is a combination of the symbols for
"disaster" and "opportunity"

The 1998 growing season was a double disaster. First, there was the six extra
weeks of winter. Then I replanted. Then, just as my replanted squash were
emerging and most vulnerable, there was a plague of cucumber beetles. Every
seedling had a dozen or more beetles on it. Nearly every one was eaten to the
ground within a day or two of emerging. By then, it was too late to replant again.

So there I was with only about a quarter of a squash patch occupied virtually
entirely by plants that were experimental material - plants whose fruits could not
be expected to be edible at this stage in the project. I love growing squash, but
the basic idea is to get something to eat. I decided it was time to learn more
about summer squash. I had read somewhere that most winter squash make good
summer squash too. I had not checked this out, because I didn't especially like
summer squash. Now, though, I had too few squash plants and probably no good
winter squash. I could take this opportunity to tastetest all my winter-squash-
breeding plants to see if they were good as summer squash.

I also learned how to dry summer squash and evaluated all the plants as to
their flavor as dried squash. The results were interesting, but are a story for
another day.

Meanwhile, the squash were growing, growing, growing. The vines were exactly
what I wanted, all of them. They were very vigorous, very aggressive, spreading
vines that fully occupied the space they took and shaded out the weeds. They all
had huge, magnificent leaves. The vines rooted at the nodes wherever they had
good contact with the soil.

There was quite a lot of variability from plant to plant in leaf shapes and
colors, but not overall plant form. (I had eliminated the bushes.)

When the plants began fruiting, the fruits were of various types. They ranged
from flattened disks to spheres to short loaf shapes like `Sugar Loaf'.
('Connecticut Field' is variable in shape, sometimes a little taller than round and
sometimes a little flatter.) Most plants had fruits that were light green striped and
spotted with darker green, like immature `Sugar Loaf'. Some were solid green of
various shades. Some turned orange in storage and some didn't.

I picked most of the fruits to keep the plants flowering, because I wanted at
least one backcross to `Sugar Loaf' and one selfpollination on each plant. The
plants responded by acting like summer squash - cranking out flowers and
summer squash with abandon. By early August I had backcrosses to `Sugar Loaf'
as well as true-selfpollinations on every plant. The backcross to `Sugar Loaf'
represented the original project I was trying to do. The selfpollination
represented the fact that I had no idea what these plants were good for yet, but
suspected that they were good for something just as they were.

To put it another way, the backcross represented my attempt to move the

material closer toward `Sugar Loaf'. But the selfpollination was the best way to
preserve the genetic combinations of these individual plants who already existed,
and who had so charmed and impressed me.

Meanwhile, I was tasting all my squash as summer squash. The cooked flavor of
the squash from the twelve different plants was edible, and as good, I thought, as
most true summer squash - which is to say, nothing special. They all tasted more
or less the same. As I've said, I don't like summer squash.

Then I started tasting the squash raw, starting with the last plant in row 4, the
weird row. My first reaction was that I had made some mistake, and picked up
something other than a squash. The raw squash was truly delicious. It was sweet,
quite firm, and very flavorful, with a flavor unlike anything else at all. The fruit
was a loaf shape of about a pound at this young stage, and was light green with
darker green stripes.

I sliced the squash into chunks and ate the whole thing with little bits of aged
Cheddar cheese. I, who had always sneered at summer squash, who always
picked them out of salads at other people's houses, if they were so crude as to
include them. I, whose attitude had been that anything that tasted good with
some raw squash in it would taste even better without.

I tried squash from two other sibling plants that I brought home in the same
lot. They were every bit as uninteresting raw as most summer squash.
 I ran back to the squash patch and harvested a 1-pound squash from each of
the other plants and tried them. Nothing. None of them tasted very good. It was
just the single plant that had fruit that tasted good raw.

That single plant seemed to be producing an entirely new vegetable. A unique

flavor. An intense sweetness for an immature squash. Firm flesh in the stage up
to about 6 inches long and 4 or more inches wide. Just a perfect size to slice for
sandwiches. Sandwichslice was born.

By fall, the Sandwichslice foundation plant and her siblings had spread into huge
vines 40 feet across, and had entirely overgrown the adjacent rows. ("We told
you so," said the squash.)

I had kept the squash picked except for the handpollinations. I wanted to see
how they would do as summer squash, and you have to keep summer squash
picked. The plants kept cranking out squash. By October the squash patch was
mostly dead from powdery mildew, which is often what ends my squash-
growing season.

But Sandwichslice and her siblings were still cranking out squash. The old
central portions of the squash were heavily mildewed. But powdery mildew
mostly affects older growth. These vigorous vines had massive amounts of
vigorous new expanding growth and new roots at many nodes. They kept
growing and producing as much as ever.

There was an unusually mild, late fall. We finally had some near-freezing
weather. Almost all the squash plants that hadn't already died from powdery
mildew were killed by the cold. Sandwichslice and her sibs merely quit
flowering temporarily. After it warmed up again, they cheerfully resumed
flowering, and began producing fruit faster than ever on the huge plants.

The last week of November, long after fresh local summer squash had
vanished from gardens and markets, I took a Sandwichslice and a chunk of
cheese with me to a germplasm meeting. There were seed company people,
USDA germplasm people, and lots of organic market gardeners. Most of them
tasted a chunk of the just-plain, raw Sandwichslice. Everyone seemed to like it
and to be amazed.
 "You'll have to teach people how to eat this," said Rich Hannan, head of the
USDA/ARS Western Regional Plant Introduction Station.

"It's very easy," I said. "First, you open your mouth. Then, you put in the
squash...."

What Rich meant, though, was that nobody expected a squash to be good
enough to eat out of hand.

The squash was apparently every bit as good as I had thought it.

I harvested the mature fruit from Sandwichslice and her sibs and evaluated it.
They ranged in size, counting the biggest fruit of each plant only, from 6 pounds
to 14. That was about what I had expected. Mature Sandwichslice fruits were
blocky and loafshaped, and weighed up to 12 pounds. They were cream-colored
and light-green-striped initially, then turned to tan and greenstriped when fully
ripe. `Sugar Loaf' is also tan and greenstriped. The mature Sandwichslice fruit
looks like giant `Sugar Loaf'.

As a mature squash, Sandwichslice was quite interesting. It had an intense

aroma when chopped open - an aroma quite like that of `Sugar Loaf'. The flesh
also was intensely flavored, and of `Sugar Loaf' type. But it was also watery and
coarse. Because of the texture, it was nowhere near being a good winter squash.
The sibling plants all produced mature squash that were watery, coarse, and with
little flavor or aroma.

At this point, my project split into two parts. I would continue to try to create a
better `Sugar Loaf'-type winter squash. For that I would use the Sandwichslice
backcross to `Sugar Loaf'. (That project is underway, and is another story.)

I would also develop a Sandwichslice summer squash variety. For the

continued development of a `Sandwichslice' variety, I would plant out a row of
the true-selfpollination of Sandwichslice.

Sandwichslice was already a good Sandwichslice. But I could not expect the
plant to be pure-breeding or stable. I would need to breed further, select, and
stabilize it.

During the winter, I wondered and worried as to whether Sandwichslice was a

oneshot deal. What fraction of the next generation would be Sandwichslice type
and quality? All? None? One out of a hundred? A thousand? It all depended
upon what genes and what genetic configurations were involved in creating that
flavor and texture of young fruit. All I could do was hope. There is a big
difference between a single good plant and the start of a new variety.
 Sandwichslice

In 1999 I planted my squash as part of a friend's melon operation. This was a

new experience for me. Black plastic and row drip irrigation instead of overhead
watering. I worried about the Sandwichslice material. I gave them 20 feet
between rows and 10 feet between plants. But I was concerned as to whether
their root systems would have full access to the space underground. It all
depended upon whether the row drip irrigated the entire field, or just a 3-foot-
wide band underneath the plastic. My normal squash planting involves overhead
watering, not row drip or black plastic.

I also wondered whether the vines would be able to root at the nodes. Even if
the entire field was adequately moist, the top 6 inches or more of soil could be
expected to be dry. Could vines root at the nodes under these conditions?

I love big vines, but as I thought about it,

the only time I had used row drip irrigation, I had quite poor production on the
big vining types of winter squash. Perhaps with a row drip scheme, big vines
would be counterproductive. It's quite possible that bushes or semi-bushes would
work better in such circumstances. They might, I speculated, have a root system
that took better advantage of restricted growing space, if restricted growing
space was the row-drip reality.

I planted a short row of the Sandwichslice breeding project, a short row of the
backcross to `Sugar Loaf', several other breeding projects, and lots of winter
squash for eating. In addition, I planted about thirty different kinds of summer
squash. If I was going to be breeding summer squash, it might help if I was
somewhat more familiar with them, I figured.

I also wondered about how unique the Sandwichslice flavor was. Is there
anything else with such good flavor raw? Is there any summer squash that is
delicious raw at such a large size?

We had another very late spring. I had to do a lot of replanting. The

Sandwichslice offspring, however, jumped up out of the ground and grew, vastly
outgrowing all standard varieties and all my other breeding project material as
well.

By June, when watering began to be necessary, I had seven vigorous plants, all
of them with spreading vines that had taken over a circle of about 10 feet,
already way beyond the row drip and black plastic.

They kept growing. They made a few fruits. I ate a fruit from the earliest plant.
It was delicious. It was another Sandwichslice. Elation. Joy. Relief.

After the first few fruits, though, subsequent fruits dried up. The plants
continued flowering prolifically, but produced nothing. They didn't look very
happy either. And they didn't root at the nodes in the dry surface soil. When I
walked the rows, the plants grumbled and growled at me. One plant with
beautiful cream-colored blocky fruits (when she made any) cursed at me every
time I went by.

Meanwhile, the bush summer squash plants of thirty different varieties, finally
established in a later planting, had begun producing. Most of them had already
outproduced the Sandwichslice-project plants, in spite of being much smaller
plants with much less space between plants and rows.

After a month of trying to do handpollinations on Sandwichslice plants, only

to have all the fruits dry up, I got the picture. The overall growing method gave
adequate amounts of water to small melon plants and small restrained vines like
`Sugar Loaf', and to small, well-behaved bush summer squash. But it was not a
suitable growing method for my huge Sandwichslice vines. They did not
produce. I was fearful that they were in danger of actually dying. (The bush
summer squash, however, were doing fine.)

So I put an overhead sprinkler in to succor the Sandwichslice. They began

growing even faster, began setting buckets full of fruit, and started rooting at the
nodes.

"This is more like it!" the cream-fruited plant said. Thereafter, she greeted me
cheerfully every time I walked the rows.

Of the seven Sandwichslice plants in that generation, all had the characteristic
flavor of the original Sandwichslice. One, however, made vines that only ran out
to 10 feet, and its fruits were only up to about 4 pounds. It wasn't a good
Sandwichslice type, because the fruits were already past prime at 4 inches
across, being quite near full size.

Another plant made fruits with good flavor, but the texture was coarse instead
of fine-grained. The coarse grain made for inferior eating.

Yet another plant had fruits with an unpleasant off-flavor on top of the
Sandwichslice flavor.

One plant was so late that it did not even begin flowering before the end of the
summer, even though the plant was as big as any. The fruit was excellent quality,
though.

Two plants had round fruits. The rest were loafshaped. One plant had cream-
colored fruits that matured to light-green- and greenstriped. The rest had light-
green- and greenstriped fruits.

Of the seven plants, three gave me fruits that were really delicious, and had the
other good characteristics too: reasonable earliness; huge, vigorous vines with
big leaves; excellent growth in cold weather; and fruit of appropriate size for my
purpose.

Meanwhile, I tasted all the other summer squash I was growing. I discovered a
couple that I liked cooked and a couple I liked dried, but none whatsoever that I
liked raw Furthermore, there was nothing even marginally edible raw at the stage
when it was 4 inches across. Sandwichslice really is something special.

The obvious way to proceed with development of Sandwichslice is by planting

out seed from the true-self-pollinations of the best two or three plants. As is
typical with such projects, the good plants undoubtedly still include much
unwanted variability. I am cheerfully willing to tolerate a great deal of
variability, but not for fruit quality or stage of usefulness.

I can get rid of some of the genetic heterogeneity by selecting against

undesirable characteristics. When one does this one never totally eliminates
undesirable recessives, however. Undesirable recessives affecting fruit flavor
and quality aren't acceptable.
 A more effective way to eliminate undesired variability is by simply
inbreeding the best plant for an additional generation or more. As explained in
Chapter 9, each round of inbreeding causes a loss of half the remaining genetic
heterogeneity.

My approach will be to inbreed a few lines from the best plants each
generation until I have uniformity and stability for fruit flavor, quality, and size.
Then I'll mass-select thereafter, to retain as much unrelated heterogeneity as
possible.

I think disease resistance and overall plant vigor might be as much related to
the general heterozygosity of the plants in many cases as to specific genes. If I'm
right about this, it will be best to inbreed Sandwichslice as little as possible. Yet
some inbreeding will be essential to achieve consistency in fruit quality.

To Market, to Market, to Sell a New Squash .. .

It's Spring of 2000. I'll soon be planting out the Sandwichslice. My guess is
that this year most, but not all, of the plants will give me fruit of prime type and
quality. I should be able to begin to sell the fruit in local mar

kets this year, while I continue to perfect and stabilize the variety.

All I need to do is plant hills 4 feet apart, then thin to 10 feet after tasting one
fruit per hill. I'm guessing I'll get at least 100 pounds of fruit per plant, and
maybe 200 or more.

It is likely that it will take at least another two to four years before I have a
stable variety. The years won't be wasted, though. I can eat all the Sandwichslice
I want and sell it to local stores, restaurants, or in farmers' markets.

In addition, it will take at least two to four years before I understand the
variety well enough to know how to grow it optimally. What kind of yield will it
give? Will it yield as well per square foot as ordinary bush summer squash?
What is the optimal spacing?

How will it behave in intercroppings? Most squash can't tolerate shade. But
Sandwichslice can grow in early spring with nothing but cloudy weather. Maybe
it will be more shade-tolerant than most squash. If so, that would greatly expand
its usefulness in intercropping.

What are Sandwichslice's water needs? Obviously, overhead watering is basic.

But will the plants be able to take less watering than bush types in an overhead
watering system, as I originally imagined? The huge surface area of leaves might
mean the plants will require as much watering as bushes, whether deeper-rooted
or not. There is lots of exploring to do. I can be learning how best to grow
Sandwichslice, however, while I finish breeding it.

I have a workable variety already, though - one that represents a new class of
squash, a variety good enough to eat and even to sell, right now - a mere two
years after beginning.

This section gives detailed information on breeding some of the most popular
vegetables - tomato, lettuce, pea, bean, alliums (onions), brassicas, squash and
pumpkin, and corn. The breeding methods for these few vegetables cover
virtually the full range of methods. By reading through Appendixes A and B and
the general material in Part II and using Table I as a reference, you should have a
good start toward being able to breed any vegetable, whether you have any
specific information about it or not.
 Tomato

The tomato, Lycopersicon esculentum, is in the family Solanaceae. Other

vegetables in the family include L. pimpinellifolium (current tomato), Capsicum
spp. (peppers), Cyphomandra betacea (tamarillo), Physalis spp. (husk tomatoes
and relatives), and various species in the genus Solanum, which includes the
potato, eggplant, naranjilla, and garden huckleberry.

The tomato (L. esculentum) and current tomato (L. pimpinellifolium) can be
crossed easily.. There are many wild species of Lvcopersicon with which the
cultivated tomato can be crossed with more difficulty and which have served as
sources of dominant genes for disease resistance via recurrent backcrossing. (See
chapters by Rick in Hybridization of Crop Plants and Tigchelaar in Breeding
Vegetable Crops.)
 Breeding System

Tomatoes are basically inbreeders. In modern cultivars the pistil never emerges
outside of the fused cone of the stamens. The anthers are on the innermost
surface of the stamen cone and shed pollen inward. Since the flowers are
pendant (hang downward), the pollen drops over the stigma, thus causing self-
fertilization.

Many heirloom varieties or varieties with L. pimpinellifolium or wild species

in their ancestry have pistils that extend beyond the stamen cone and are exposed
to pollinators. These varieties are much more likely to be cross-pollinated.

Tomato stigmas become receptive about a day before the flowers open. Pollen
begins to dehisce somewhat later than the stigma becomes receptive, but still in
the bud stage. The stigma remains receptive and pollen continues to shed during
the entire time the flower is open, which may be one day to a week, depending
on conditions.

Tomatoes grown outdoors shed their pollen when jostled by the wind.
Greenhouse tomatoes need to be jostled to release their pollen and set fruit.
Shaking the plants regularly will suffice.
 Isolation

Ashworth, in Seed to Seed, summarizes the situation: "The extent of cross-

pollination in tomatoes has been a controversy among seed savers for a long
time. Some say that crossing is rampant, while others have never seen crossing
after years of growing different varieties next to one another.

The specific repertoire of pollinators present is undoubtedly a major factor.

The other major factor that affects the isolation distance is whether your
varieties have retracted or protruding stigmas. If all your varieties have retracted
stigmas and you don't see any insects working the flowers, you may be able to
use just a few feet of isolation - the distance involved when you alternate rows of
tomato varieties with rows of other plants, for example. Varieties with protruding
stigmas are more subject to being crossed by insects and will need more
isolation.
Inbreeding Depression

The tomato, a natural inbreeder, does not display inbreeding depression.

 HandPollination and Crossing

Tomato flowers are emasculated and pollinated in the late bud stage. At this
stage the bud has reached full length, the sepals have started to open, and the
petals have started to change color from light to bright yellow, but the petals are
still tightly shut. Once the petals start to open, it is too late; anthers are already
dehiscing. (See illustrations on pages 306-307.)

Once the buds are emasculated, they can be pollinated immediately or in a

separate step the same day or later. Sometimes repeat pollinations are used to
increase the success rate. George describes emasculating and cross-pollinating at
separate times, with timing dependant on the region and the weather: "For
example, in Californian field production, flowers are emasculated early in the
morning (from about 06:30) on the same day as pollination; whereas in North
European greenhouse hybrid seed production, flowers are emasculated up to two
days in advance of pollination."

Note that you can see when tomato (or other plant) pollen is released by using
a hand lens. Use this general information to help you, but the exact timing
depends on the variety and the weather, and the final authority is the plants.
Ultimately, you establish whether dehiscence has occurred at a certain stage
under your conditions not by reading, but by looking at the plants.

Emasculation of tomato buds is performed as follows. Choose buds at the

appropriate late bud stage. One or more buds on a cyme may be at the right stage
at once. You may have to figure out a right time of day to catch the buds in the
right stage. It will usually be sometime in the morning. The highest success is
with bigger buds, the first few on a cyme. Fruits and uncrossed flowers are
usually removed to eliminate competition with the cross.

At the appropriate late bud stage, the corolla - the ring of petals (yellow) - is
folded up around the stamen cone (also yellow), and the two are attached at their
bases. Thus there is a two-layered cone encircling the pistil. In emasculation this
double cone is lifted off as a unit, leaving a bare pistil with a ring of sepals
(green) at its base.

To remove the double cone, insert one tip of a fine forceps between the anther
cone and the pistil and the other between the calyx (sepals) and the corolla
(petals) and gently pull the cone off, taking care not to injure the pistil. You may
need to dissect away one petal and stamen first and/or to remove the remaining
stamens afterward.

You can transfer pollen by using pieces of anther cone or by scraping pollen
from the cone off anthers with a triangular dissecting needle. Just-opened
flowers are the best and easiest source of pollen.

As is true for most Solanaceae family members, tomato pollen is long-lived

and stores well. It will remain viable for weeks at room temperature. If
desiccated and refrigerated, it will last a number of months.

Emasculated flowers do not need to be covered to prevent contamination, as

there is no pollen to entice insects and no bright parts to signal to them. Under
hot, dry, or windy conditions, emasculated flowers may need to be covered with
glassine bags to prevent their drying out. Outdoor crosses are most successful if
done under cool, relatively wind-free conditions.
 Hybrid Seed Industry

There is an extensive hybrid seed industry for tomatoes, even though they
don't display inbreeding depression and thus there is no special biological
advantage to the hybrids. The best open-pollinated varieties of tomatoes are as
vigorous as the best hybrid varieties. The major reason for the industry is the
financial advantage to seed companies in producing and promoting proprietary
varieties that no one else can multiply.

Commercial hybrid seed production depends on hand-emasculating and

handpollinating the plants. Genetic male sterility has started to be used in
making hybrids. This saves the emasculation step, but handpollination is still
required. Hybrid tomato seed is economical to produce because the plants and
produce are high-value ones for which people will tolerate high seed costs,
because each successful pollination produces two hundred or more seeds, and
because hybrid seed is mostly produced in Taiwan with inexpensive but skilled
labor.
 Genetics and Breeding

Much of the recent professional work has been aimed at developing machine-
harvestable determinate varieties for processing, as well as those with good
shipping and holding qualities for the supermarket trade. Another trend is toward
developing specialized varieties associated with every use, growing method, and
growing region. Such specialized cultivars are replacing general ones
commercially.

Tomato varieties for greenhouse producers need to be indeterminate so as to

maximize productivity, for example, and need resistance to some pests and
diseases that aren't very important outdoors. Field crops need to be resistant to
the major diseases prevalent in the area. Dominant genes conferring resistance to
many diseases and pests have been discovered in wild varieties and transferred
to commercial cultivars by recurrent backcrossing.

The tomato is one of the most popular vegetables for genetic experimentation,
and many of the genes of commercial importance have been identified and
mapped. (For chromosome maps of a few hundred of the genes, see Tigchelaar.)
Some of the genes of greatest agricultural significance are given in the following
paragraphs. Symbols that start with small letters are recessive; those that start
with large letters are dominant. (Most of the information is from Tigchelaar; the
rest is from Mike Courtney's review in the 1989 Summer Edition of the Seed
Savers Exchange.)

Genes that affect plant growth habit include sp, br, and d, which are self-
pruning, brachytic, and dwarf, respectively. Self-pruning means determinate.
Nearly all determinate varieties carry it. A few extreme bush types, especially
those for pot culture, carry brachytic or dwarf genes. `Redbush' carries br;
`Epoch' and `Tiny Tim' carry d.

The gene c is associated with the potatoleaf phenotype.

Two different genes, j-1 and j-2, can cause jointless pedicels. In plants with
jointed pedicels, the tomato stem breaks in picking so that the tomato has a
section of stem on it. When tomatoes with jointless pedicels are picked, the
tomato separates from the stem at the fruit.
 Dominant genes for disease resistance have been transferred from wild species
by recurrent backcrossing. These genes could be transferred into heirloom and
home garden favorites by simple recurrent backcrossing.

Four different genes confer resistance to leaf mold - Cf-1, Cf-2, Cf-3, and Cf-
4. Fusarium wilt immunity for race 1 and race 2 is associated with genes I-I and
1-2, respectively. Resistance to verticillium wilt, Septoria, late blight, Alternaria,
and Stemphylium is conferred by genes Ve, Se, Ph-1, Ad, and Sm, respectively.
Tobacco mosaic virus resistance can be conferred by Tm, Tm-2, or Tm-22.
Nematode resistance is conferred by Mi.

There is genetic variability for insect resistance of various kinds, but most
professional breeders and growers depend on insecticides instead. Organic
gardeners and farmers could be pursuing the genetic approach. Tigchelaar
includes a summary of work that has been done and a list of wild species
resistant to various pests from which the needed genes could be obtained. See
Rick and Tigchelaar for information about crosses with wild species.

Male sterility is associated with a number of genes. The ones I've seen listed
are all recessives.

The u locus is associated with the uniformity of ripening. Many older

cultivars have green shoulders when the rest of the fruit is ripe. Modern
cultivars usually ripen uniformly.

The gene nor^ is nonripening and is found in `Longkeeper'. Other genes that
affect ripening include rin (ripening inhibitor), Gr (green ripe), and Nr (never
ripe), among others.

The gene pat-2 causes parthenocarpic fruit - that is, plants carrying it can
produce fruit without setting seed. The cultivar `Severianin', for example, carries
this gene. Varieties with parthenocarpic fruit can set fruit without being
pollinated, so they can set fruit in weather that is too cold for pollen release or
function. Fruit so set is seedless. Other genes also can cause parthenocarpic fruit.

Mike Courtney provides an excellent summary of tomato color genetics in the

1989 Summer Edition of the Seed Savers Exchange, which I've used for this
discussion. Tomato color is determined by a combination of the color of the skin
and the color of the flesh. Skin color is yellow or colorless, the difference being
determined by a single gene: y+_ is yellow; yy is colorless. Flesh color involves
a number of genes but is basically yellow, orange, or red.

The red tomato of commerce has a yellow skin and red flesh. Tomatoes with a
colorless skin and red flesh are pink. Yellow flesh color is determined by the r
locus, with rr being yellow and r+_ being red. Yellow tomatoes have yellow skin
and yellow flesh; white, ivory, and lemon tomatoes have colorless skin and
yellow flesh.

The color of the flesh depends on the content of carotenoids. These include
various forms of carotene, all of which are orange, and lycopene, which is red.
Of these, betacarotene is of nutritional significance because the body can convert
it to vitamin A. Lycopene has no vitamin A value. Various genes control the total
amount of carotenoids; others determine which form of carotenoid predominates.
Red tomatoes have most of their carotenoid in the form of lycopene. Yellow
tomatoes have no lycopene and only trace amounts of betacarotene. Orange
tomatoes have no lycopene and large amounts of carotene.

Crimson tomatoes carry the gene ogc and have somewhat more lycopene and
somewhat less carotene. Crimson tomatoes have a yellow skin. When crimson
flesh is combined with a colorless skin, purple tomatoes result.

Tangerine tomatoes are associated with the gene t. They have no lycopene and
most of their carotenoids in the form of zeta-carotene. Among these are `Golden
Jubilee', `Orange Early', `Orange Tall', `Golden Delight', and `Sunray'. Tangerine
flesh and colorless skin result in a muted gold tomato. Two examples are
`Glecker's Gold Glow' and `Golden Glow'.

Apricot-colored fruits are associated with the gene at. They have reduced
lycopene levels but normal betacarotene levels. The gene Del has increased
delta-carotene. This is one cause of orange fruits. The gene hp (high pigment)
increases the amount of lycopene and carotene, as well as the amount of
chlorophyll in the immature fruit. `Redbush' has this gene. The gene dg (dark
green) also increases the amount of chlorophyll in the immature fruit and has
more lycopene and betacarotene in the mature fruit.

The gene gf (green flesh) is responsible for mature fruit that retains some of its
chlorophyll. In r+ forms the green and red pigments yield red-brown fruit. In r
forms, which have no red pigment, the mature fruit is green. `Evergreen' is an
example of the latter.

The gene gs is associated with green striping. The fruit is either red and gold
striped or green and gold striped depending on the configuration at the r locus.

Orange fruits are associated with two dominant genes, B and MoB. Plants
carrying just B (e.g., `Caro Red') have no lycopene and five times as much
betacarotene. Plants carrying both B and MoB have ten to twelve times as much
betacarotene as normal. `Caro Rich' carries both B and MoB. B is closely linked
to sp (self-pruning; determinate) on chromosome 6.

The tomato is second only to the carrot as a source of betacarotene in the

American diet. Thus it is particularly unfortunate that red has been seized upon
as the color of choice for tomatoes. Orange tomatoes fare well in blind taste
tests; but red tomatoes fare better in tests where participants can see the color.
Commercial breeders and growers are limited by this popular bias, but home
gardeners can breed and grow any color tomato they want.

Given that B and MoB are both dominant, it would be easy to transfer them to
other varieties by crossing them with `Caro Rich', followed by recurrent
backcrossing. Onefourth of each backcross generation would be expected to be
deep orange and to have high

betacarotene content. Perhaps amateur plant breeders should spearhead a drive

toward breeding more nutritious vegetables and lead the way with orange
versions of all their favorite tomato varieties.

Tomato flavor depends on the presence and the amount of dozens of

substances, and so it has complex inheritance patterns. Most of the acid and
many of the other components of tomato flavor are in the juice, however, so
meaty tomatoes are inherently not as full-flavored as juicy tomatoes. Meaty
tomatoes are often preferred by those who like mild-flavored tomatoes, though.

Harvesting, Processing, and Storage

Tomato harvesting, processing, and storage is covered in Chapter 21.

 References

Charles M. Rick, "Tomato," in Hybridization of Crop Plants, pages 669-680;

George, Vegetable Seed Production, pages 208-223; Edward C. Tigchelaar,
"Tomato Breeding," in Bassett, Breeding Vegetable Crops, pages 135-171;
Ashworth, Seed to Seed, pages 151-155; Mike Courtney, "The Genetics of
Tomato Fruit Color," 1989 Summer Edition of the Seed Savers Exchange, pages
33-36.
 Lettuce

Lettuce, Lactuca sativa, is in the family Asteraceae (Compositae). It is thought

to derive from the wild species L. serriola (wild lettuce, prickly lettuce) and
crosses freely with it, so much so that the two might be regarded as within one
species.

Other genera and species in the family Asteraceae that may be found in
Appendix A are Arctium lappa (gobo), Bellis, Calendula, Carthamus,
Chichorium endivia (endive), C. intybus (chicory), Chrysanthemum, Cynara
cardunculus (cardoon), C. scolymus (artichoke), Dahlia, Helianthus annuus
(sunflower), H. tuberosus (sunroot, Jerusalem artichoke), Inula, Matricaria,
Petasites, Polymnia, Scolymus, Scorzonera hispanica (scorzonera, black salsify),
Silybum, Stevia, Tanacetum, Taraxicum officinale (dandelion), and Tragopogon
porrifolius (salsify, oyster plant). (I listed these because of all the interesting
wide-cross possibilities.)

Stem lettuce, also called asparagus lettuce or celtuce, is a form of lettuce that
has been bred for the stems instead of or in addition to the leaves. It is the same
species as ordinary lettuce (L. sativa) and crosses freely with it. Celtuce is
sometimes referred to as L. sativa var. asparagina or as being in the Angustan
group.
 Flowering

Most lettuce varieties are day neutral. Those used for winter production in
greenhouses are usually long-day plants.

In crisphead lettuce the seed stalks need assistance to emerge. One option is to
slice through the head to expose the core. Some breeders thump the head hard
with the palm of their hand to break the leaves off at the base. Alternatively, the
leaves may be removed, or the plants may be deheaded so that axillary shoots
develop.
 Breeding System

Lettuce is a strong inbreeder. The flower is made up of numerous tiny, perfect

florets. All

the florets open on the same day. The flowers open, pollen is released, the style
finishes emerging, and the stigma becomes receptive and is fertilized all within a
few hours. The flowers open just once on one day. Each floret produces only one
seed.

Within the individual floret the anthers are fused into a tube that encircles the
style. The inner surface of this tube sheds pollen onto the style as it emerges. The
style accumulates a coating of self-pollen as it emerges and before it is receptive.
As soon as it becomes receptive, the self-pollen fertilizes the flower.
 Isolation

The literature is somewhat contradictory concerning isolation. According to

George, "Although up to 5% cross-pollination has been observed in lettuce in
some areas most authorities regard it as a self-pollinating crop and only specify a
physical barrier (e.g. adjacent sections of greenhouses) or a minimum of two
meters between different cultivars."

Lettuce will cross with and should be isolated from its wild relative, L. serriola
(prickly lettuce). Celtuce must be isolated from other lettuce varieties.
Inbreeding Depression

Lettuce is a natural inbreeder and does not display inbreeding depression

 HandPollination and Crossing

Lettuce is one of the more difficult vegetables to do crosses with. (See

illustration on page 306.) Professional breeders depend on washing the pollen off
the stigmas at exactly the right time: after it has all been shed but before the
stigma becomes receptive and the pollen has germinated. A fine mist of water is
directed at the entire flower. The right time for washing is after the style has
grown beyond the anther cone but before the stigma lobes have begun to curl
outward. "This method produces about 25%-75% crosses if done properly,"
Ryder says. The chance of success can be increased by using more than one
washing. One still must be able to identify hybrids from selfings, however. That
is, even these "controlled" lettuce crosses are actually fertile X fertile crosses.

The florets do not develop with complete synchrony. When the first ones are
ready to emasculate by washing depends on the variety and the weather. You
have to observe the plants and figure out their timing under your conditions.
Usually it is in the morning - early morning on warm days, later morning on
colder or overcast days.

Individual florets can be emasculated by hand and then pollinated. The anther
cone is removed with forceps early on the morning of flowering. This cone is
difficult to remove; the flowers are tiny and delicate, the process is time-
consuming, and a successful cross produces just one seed. Usually fertile X
fertile crosses, with washing as described, are done instead.
 Genetics and Breeding

Much of the professional breeding of lettuce has been oriented toward disease
resistance and achieving greater uniformity of harvest time. Other important
trends involve shaping cultivars for specific regions and growing methods.
Recently, for example, growing lettuce in greenhouses during the winter has
become an important production

method. Varieties for winter greenhouse production need to be able to grow well
with less light and heat than outdoor varieties.

In spite of the popularity and importance of lettuce, it is not very well known
genetically. Breeding is done largely by using general principles. Somewhat
more than fifty genes have been identified; there are no real chromosome maps.
Many of the genes involve disease resistance; usually the resistance is only to
specific races of specific diseases. For a description of the genes involved and
breeding lettuce for disease resistance, see Ryder, Leafy Salad Vegetables. A few
of the other genes that have been identified are described in the following
paragraphs (from Ryder's chapter in Bassett).

The formation of anthocyanin, which produces red or reddish leaves, is under

complex control. In general, hybrids between red and green varieties are red or
reddish. Two dominant genes, C and G, influence anthocyanin formation. The
distribution of the pigment, however, is controlled by another gene, the R locus.
R, R, Rb", R1, and r are all alleles at the R locus. R is associated with full red
leaves; R' is spotted; Rs" has red-brown spotting; Ris tinged with red at the leaf
margins.

The gene Sc (scallop) describes the leaf margin.

The U locus controls leaf lobing. U is nonlobed, u° is oak-leaved; u is lobed.

The gene t is associated with slow seed stalk formation, or bolting resistance.

Genotype ww has white seeds; W_ has black seeds. The y locus is associated
with yellow seeds. Genotype yyW has yellow seeds. Genotype yyww has white
seeds.
 Given the difficulty with which lettuce crosses are made, it would be nice to
have a number of known marker genes to help identify hybrids. Unfortunately,
there aren't many. The color and leaf-shape genes are most effective because
they can be identified in seedlings. Seed-color genes also are often used, with
identification of the hybrids being delayed until bolting and seed formation.

Harvesting, Processing, and Storage

Lettuce seed ripens irregularly. Ashworth suggests that the way to get the
maximum amount of seed is to shake the plants daily over a container or bag.
Alternatively, entire plants are harvested when much of the seed is dry, then put
upside down in bags. Seed can be freed from the plants by shaking or rubbing
them between the hands. Screens of various sizes are used to separate seed from
debris. Winnowing usually isn't effective, as the debris blows about the same as
the seed.
 References

Ryder, Leafy Salad Vegetables, pages 13-94; Edward J. Ryder, "Lettuce

Breeding," in Bassett, Breeding Vegetable Crops, pages 433-473; Ashworth,
Seed to Seed, pages 92-94; George, Vegetable Seed Production, pages 120-132.
 Pea

Garden and field peas, Pisum sativum, are in the family Fabaceae
(Leguminosae). Field peas used to be referred to as a separate species, P arvense,
but there is no justification for this, as garden and field peas can be crossed
freely.
 Breeding System

Peas are strong inbreeders. They normally self-pollinate before the flower
opens. The flowers are protogynous; the stigma becomes receptive well before
the pollen is shed. This facilitates crossing, allowing the bud to be emasculated
and handpollinated in one step.

In moderate weather pollen can remain viable for several days. The stigma
remains receptive until a day or more after the flower wilts.
 Isolation

According to Gritton's chapter in Bassett, crossing of peas in the United States

is normally less than 1 percent. The pea is, in most areas, one of the most highly
inbreeding crops, and the isolation distances used are often just a few feet - just
enough to prevent physical mixture of plants or seeds.

However, this picture is complicated by the fact that in some areas peas can be
extensively cross-pollinated. Reported cross-pollination frequency for peas
ranges from 0 percent in New York to 60 percent in Peru, where a number of
insects cross-pollinate them.

Official isolation distances in Europe are 100 meters. Ashworth (Seed to Seed)
recommends 50 feet.

I believe most seed savers in temperate regions can use just a few feet of
isolation between varieties - such as is represented by separating different pea
varieties by a row of something else - as long as you inspect the plantings and
plants (buds and flowers) for insects and insect damage that might suggest more
care is needed.
Inbreeding Depression

Pea is a natural inbreeder and does not display inbreeding depression.

 HandPollination and Crossing

Open pea flowers have three distinctive kinds of petals. (See illustration on
page 307.) The outermost, or standard, is the one that opens up wide into two
lobes when the flower opens. The next inward petal is the wing, which surrounds
the center of the flower. Interior to the wing is the keel, the innermost petal that
protects the pistil and stamens. Peas are handpollinated in the bud stage, and in
buds these three petals are all folded up, so you must penetrate three layers of
petals.

The outer two petals, the standard and the wing, are normally opened by hand
and left intact so that they can close afterward and protect the fertilized stigma.
The innermost petal, the keel, must be slit open or removed. Most people remove
it. Pea flowers need not be bagged after handpollination.

Handpollination is done when the buds are as big as possible but before pollen
has been released. In my experience this stage varies widely with both variety
and weather. For many varieties under cool (normal earlyspring) conditions, the
late predehiscent stage is when the petals of the bud have grown out to just
barely beyond the sepals. But plants and buds can grow at temperatures that are
too cold for pollen dehiscence. Under these conditions buds grow without
dehiscence until it warms up. So in cool weather the buds may be quite large and
the petals may extend far beyond the sepals when the anthers dehisce. But in
warm weather the same variety may dehisce at a much earlier stage, when the
buds are much smaller and the petals are still shorter than the sepals. Exact
timing also is influenced by variety.

It's much easier to work with the big, latedehiscing buds associated with cool
weather. Don't try to do crosses when the weather is too cool for them to take,
however. (You learn to tell this by noticing what kind of weather pods normally
set in.)

Buds lower on the plant also are bigger, easier to work with, and more likely to
make successful crosses, and they produce more seed. So whenever possible,
make crosses with material grown early in the year, and do them early in the
flowering season. Midsummer flowers pollinate when the bud is so tiny that I
have not been able to do crosses with them at all. I plan to try again, but with
growing the plants in the shade. I also may try emasculating them at a very early
stage and then pollinating them a couple of days later.

It takes a little practice to manipulate pea flowers and do crosses. Figure on

ruining several flowers before you acquire the skill. I usually start by stripping
off the sepal that is directly in the way of opening up the standard and wing.
Then I open up the standard and wing and hold them open with one hand. With
the thumb and forefinger of my other hand, I grasp the keel about halfway down
and tear it off so as to bring most of the stamens with it but leave the style
untouched and undamaged. When I'm very clever about it, I get all ten stamens.
Usually one to three remain, and I remove them with forceps. Alternatively, you
can remove the keel in pieces with forceps and then remove the stamens.

I use pollen that has been shed the same day I do the crosses - that is, from
buds somewhat bigger than those I'm using as females or from just-opened
flowers, depending on the weather. Fresh pea pollen is bright yellow and moist-
looking. Old pollen is pale and dry-looking.

Pea pollen can be stored for up to six days after dehiscence. (I think this means
in a refrigerator.) Whole buds or flowers that have just dehisced probably could
be wrapped to prevent them from drying out and then refrigerated. Pea pollen is
said to be viable for about a year if vacuum dried and stored at 25°C. I think this
means that you can store it for an extended time by desiccating it and putting it
in the freezer, as described in Appendix B. I intend to try it.

I usually transfer pollen by using the whole male flower with the petals
stripped off, by using anthers, or by removing the keel of the male flower and
using it. The latter has the advantage that it doesn't sacrifice the male flower, just
removes the extra pollen. To use the male keel I pull it off as described for
female flowers; it constitutes a little bag of pollen into which the female stigma
can be dipped.

The stigma is at the very end of the style. That subterminal brush on the style
is not the stigma.

After handpollinating, I close the standard and wing back around the pollinated
pistil. I am told that Baggett's group uses a bit of tape to hold the entire flower
closed for additional protection from drying out. They fold a small piece of
Scotch tape in half, then open it again and use it to clamp the flower closed. I
plan to try this trick next season.

My own experience is that crosses take best in mild weather. About half my
crosses take under optimal conditions; much fewer take late in the season or in
hot weather. Perhaps the tape trick will improve the odds on crosses done under
these conditions.

Pea seed, if mature enough that it has accumulated most of its dry matter, can
be germinated without drying down. Just shell out the fresh, mature, but not dry
peas and soak them in water for a few days until they germinate. (Change the
water once or twice a day to provide plenty of oxygen and discourage bacterial
or mold growth.) This trick can cut about a month off the time required to do a
generation of pea breeding. It also works with all the other legumes I have tried.

Labeling pea crosses is trickier than with many vegetables because the stem
that supports the bud is small and delicate. I use tiny labels used for pricing
jewelry, and the smallest labels in the class. These are about `/2 inch long and '/4
inch wide and are supported with a slender thread.

It's very difficult to breed and eat from the same pea plants. The cross labels
twist around and hide, so it is easy to eat the crossed pods by mistake. The care
needed to examine the plants for labels makes harvesting too much work. And
maturing pods cut the production of the plant short; peas must be kept picked to
stay in production. For these reasons, I usually plant separate patches for
breeding and eating.
 Genetics and Breeding

There are three basic classes of pea varieties, depending on their intended
purpose: dry (soup) peas, shelling peas for use as green shelled peas, and
ediblepodded peas. Some peas also are used for green manure or livestock feed.
Home gardeners usually grow shelling and ediblepodded types. Recent breeding
efforts have concentrated largely on developing bush types with simultaneous
pod ripening for picking by machine and on incorporating disease resistance into
these varieties. Much additional work has been aimed at obtaining peas of every
given maturity for every purpose for each of the major pea-growing areas.

In shellers professional breeders aim for small pea size. This is because
consumers have learned to associate large peas with pastprime peas. The smaller
peas fool them into thinking the peas are prime. That isn't necessarily true. I
think home gardeners should give some attention to breeding for large peas, as
these produce more peas for the same amount of shelling work. The classic
variety `Alderman' (aka `Tall Telephone') has huge peas that are among the
tastiest.

Professional breeders have been concentrating almost entirely on dwarf

varieties. For home gardeners there are many advantages to tall, pole varieties.
We home gardeners will probably have to breed them ourselves. In many cases
all you need to do is transfer a gene for tall into established dwarf varieties by
recurrent backcrossing.

So far professional breeders have released only green-podded shellers or

ediblepodded varieties. This is boring. We home gardeners should do something
about that, too. Many bean varieties have purple or yellow pods. The genes for
colored pods also exist in peas, waiting for us to use them.

In short, in spite of the fact that the pea is one of the most popular vegetables
that receives a relatively large amount of breeding attention, there are still plenty
of opportunities for amateurs to work with peas to fill the needs and niches that
are being neglected.

Genetically speaking, the pea is one of the best studied vegetables. See
Gritton's chapter in Bassett for a list of many of the known genes, chromosome
maps, and references to the current professional literature. Some of the genes
that are most relevant agronomically are described in the following paragraphs
(taken from Gritton).

Seed shape is determined by a number of genes. The two most important loci
are R and Rb. The recessive allele at either of these loci causes the seed to be
wrinkled instead of round. The round seed becomes starchy faster than the
wrinkled seed, but it germinates better in cold weather. Round types are usually
used for early cultivars; otherwise, wrinkled types are used. Wrinkled genotypes
are rrRb_, R_rbrb, and rrrbrb. Round seeds are R_Rb_.

The round versus wrinkled phenotypes are apparent at the seed stage, since
they affect the starches in the cotyledons, which are part of the embryo.
Segregation in the F2 seed can be seen as you shell pods on the F, plants.

Cotyledon color can be green (ii) or yellow (I-). You may have to pierce the
seed coat to see the color of the cotyledons. In addition, color fades with age and
is most obvious in fresh seed. Canning cultivars traditionally have a light green
color that resists leaching in the canning fluid. Freezer cultivars traditionally
have darker green peas. (There are now dual-purpose cultivars, too.) The genes
pa and vim are associated with dark green seed and foliage color.

Plant height is controlled by three genes: cry, la, and le. Alleles associated with
tallness are dominant.

Two independent genes, p and v, are associated with reducing the fibrous
membrane on the inside of the pod. Modern ediblepodded varieties have
genotype ppvv Shelling peas are generally PPVV Crosses of ediblepodded peas
and shellers produce F, plants that have shelling-type peas. Plants carrying Dpo
have tough, leathery pods that dehisce readily when mature.

The gene n increases the thickness of the pod wall and causes the overall shape
of the pod to be round instead of flat in cross section. All snap peas have the
genotype ppvvnn.

Synthesis of anthocyanin (the purple, red, and blue pigment) requires the
presence of the dominant gene A. Plants of genotype as have white flowers;
green, not purple, pods; and no purple in the leaf axils or seed coats. Plants of
genotype A_ have some purple, but other genes control where it is distributed.

Purple pods require the presence of three dominant genes: A, Pu, and Pur. The
purple ring in the leaf axil requires A and the dominant gene D. (Plants that are
dd have green leaf axils even if they can produce purple pigment.) Genes b and
ce (in the presence of A) change the shade of color in the flower to pink and
rose, respectively.

The official name of the gene associated with yellow pod color is gp. Yellow
pods are recessive to green.

Purple, brown, or black color or spotting in the seed coat is controlled by a

number of

genes (see Gritton). All of them require A to be expressed. Dozens more have
been identified that control the number of basal branches, stem fasciation, leaf
morphology, leaflet margin shape, conversion of leaflets to tendrils or vice versa,
and so on. Additional genes have been identified that control the presence,
absence, or amount of wax on the pods and the rest of the plant. Other genes
control the distribution of the wax.

Two genes, fn and fna, are associated with the number of flowers at a node.
This characteristic is very much influenced by the environment.

Some genes that determine the shape of the pod are It (25 percent wider), Bt
(blunt apex), and Con (affects curvature).

Yet more identified genes are associated with dimples on the seeds, texture of
the seed surface, seed shape, and whether seeds stick together in the pod. There
are also genes that change the pollen color or drastically alter flower anatomy.

The genes En, Fnw, Fw, sbm, and er-1 are associated with resistance to pea
enation mosaic virus, fusarium near wilt, fusarium wilt, pea seed-borne mosaic
virus, and powdery mildew, respectively.

I have seen no information on the genetics of pea flavor other than the basic
sweet wrinkled-seeded versus the round-seeded characteristic. I also have seen
no genetic information on size of pods or peas, number of peas in the pod, cold
hardiness, heat resistance, or many of the other characteristics that matter most
agronomically. Yet these characteristics are clearly inherited. It is instructive that
even with peas, where so much is known compared to most vegetables, large
parts of every breeding project still must be guided largely by general principles.

Harvesting, Processing, and Storage

Harvesting, processing, drying, and storage of peas is covered in Chapter 21.

 References

Earl T. Gritton, "Pea Breeding," in Bassett, Breeding Vegetable Crops, pages

283-319; Suzanne Ashworth, Seed to Seed, pages 139-140.

Common Bean (Phaseolus vulgaris)

Phaseolus vulgaris, the common bean, is the major type of dry bean used in
North America and, in its green bean form, is one of the major fresh, canned,
and frozen vegetables. Some home gardeners also grow varieties that are shelled
and eaten as immature seeds. Two other species within the genus that are
common as commercial beans are P coccineus (runner beans) and P. lunatus
(lima bean).

Common beans can be distinguished from runner beans, which have similar
seeds, by the fact that runner bean seedlings have cotyledons that remain in the
ground when the seed germinates. In common beans the cotyledons emerge
aboveground when the bean germinates. Fava beans, too, have cotyledons that
emerge. However, favas have their hilum at the end of the seed, not on the side
as common beans do.

To a large extent the varieties grown for use as dry beans, green beans, and
shellies are all different varieties, although there is

some overlap. Dry beans, optimally, have a fibrous pod and are stringed; these
characteristics make them easy to thresh. In addition, the seeds develop and dry
rapidly once the pods reach full size.

The ideal shelly bean also has pods that are fibrous and stringed, and therefore
easy to open by hand. In addition, the seeds develop rapidly after the pod reaches
full size, but they don't dry down quickly. They stay in the fresh green state for a
while.

The ideal green bean should have little fiber in the pod, because the pod is the
edible crop. Optimally, it is stringless. In addition, its seeds should develop
slowly after full pod size is reached so that it stays in the right stage for green-
pod use as long as possible. (Once seeds start to develop, the pods invariably
toughen and become more fibrous.)

Green beans for canning usually have white seeds so they don't color the
canning liquid. Green beans for use as a fresh commercial crop need to have a
bit more fiber in the pods than those for canning or freezing so that they can
stand up to shipping and handling.

Among home garden varieties, especially of pole beans, there is overlap in the
uses for many cultivars. However, any cultivar that is optimal for one of the
three main uses is inherently suboptimal for the others.
 Breeding System

Beans are basically inbreeders. Like peas they have three kinds of petals:
standard, wing, and keel. Also like peas the style and ten stamens are enclosed
within the keel, and dehiscence and self-fertilization occur in the late bud stage.
In beans, however, the style is twisted into a coil, the stamens are twisted around
the style, and the keel is twisted as well. The anthers are actually pressed against
the stigmatic surface such that they shed their pollen directly on it. According to
Bliss, the stigma is receptive from at least two days before the flower opens to at
least one day after.
 Isolation

Isolation theory and practice for the common bean is discussed extensively in
Chapter 18.
Inbreeding Depression

As natural inbreeders, beans do not suffer from inbreeding depression.

 HandPollination and Crossing

Bean crosses are often done in a greenhouse. This isn't essential, but it appears
that mild conditions, adequate moisture, and preventing the crossed flower from
desiccating are important considerations. Outdoor crosses are usually done in the
morning, when the turgor of the plants and flowers is higher.

It isn't possible to provide a good drawing of the anatomy of the bud of bean
because of the twisting of the parts. You can learn the basics by reviewing the
material on pea.

According to Bliss, buds are at the right stage for emasculation when they are
plump, showing color, and due to open the next day. He notes, "If self-
pollination has already occurred, the stigma will be swollen, light green rather
than white, dried rather than sticky, and with some remnant pollen attached." I
have never done any bean crosses, but I suspect that the situation is similar to
that for peas - that dehiscence occurs in the late bud stage, but exactly when
depends on variety and weather. You have to watch for the right stage for the
variety and check the weather on any given day.

Apparently it is much easier to injure the style or stigma in a bean cross than a
pea cross, and it's harder to get into the bud physically. The standard and wings
are opened and held back as with the pea. Then most or all of the keel is
removed with forceps to expose the style, stigma, and anthers.

Flowers that have just opened are used as the source of pollen, Bliss says.
They may be used immediately or refrigerated for use later. There are two
common pollination methods. In both, the pollen is transferred from the stigma
of the male flower to the stigma of the female. In the rubbing method, the stigma
of the male flower is made to extrude from the keel by pressing on the wings; the
pollinated stigma of the male is then rubbed against the stigma of the
emasculated flower.

In the hooking method, the pollinated male stigma is removed with forceps,
rubbed against the female stigma, and hooked through the style so that it remains
near the stigma in the female flower. One report gives success rates of 30 to 40
percent for rubbing and 70 to 80 percent for hooking.
 The bud is taped shut after pollination to prevent desiccation. A 4-centimeter
strip of Scotch tape is formed into a circle with the sticky side inward. It's placed
over the standard. It should encircle only the petals, not the sepals. A small piece
of wet tissue may be put inside the tape circle to raise the humidity around the
bud.

Because of the difficulty of emasculating beans, fertile X fertile crosses are

often done instead. This type of cross is practical whenever the cross will yield
recognizable hybrids. It is much easier to do because the bean's stigma can be
made to protrude from the keel when the wings of the flower are pressed back.
The exposed stigma can be pollinated without permanently opening the flower
or removing the keel. The stigma can self-pollinate later, of course, but the
applied pollen has a head start. Sometimes the stigma of the male is hooked
behind that of the unemasculated female, clamping it in place and preventing the
female stigma from returning into the bud, where it can be self-pollinated.

A. Unemasculated flowers. The two central buds (arrows) are at the correct stage
for emasculation.

B. Emasculated flowers and parts removed.

 Lettuce Floret Development

In stage 1, the stigma has not yet emerged from the anther sheath surrounding it.
In stage 2 the stigma has started to emerge. It is already covered with pollen, but
because it is not receptive, self-pollination has not yet occurred. Stage 3 is the
best stage for removing pollen by washing. Stage 4 is too late: the stigma has
become receptive, and self-pollination has occurred.
 Pea Flower Development

Bud 2 is at the stage that is often optimal for emasculation and pollination. But
sometimes bud 1 or bud 3 is the correct stage, depending upon the variety and
the weather. Bud 4 is too old; it will have already self-pollinated. It would be a
good source of fresh pollen.
 Pea Flower Anatomy

Here is a flower with the petals removed.

Onion Flower
Brassica Flower
 Genetics and Breeding

Extensive breeding work has been done on developing cultivars for dry use
and for green-pod use. I know of no work on developing better shelly varieties.
It seems arbitrary to me that the major fresh use for peas is as shellies and the
use of whole pods is secondary, whereas for beans the reverse is true. I think it's
a matter of historical accident. I would like to see home gardeners produce
dozens of shelly varieties and turn the shelly bean into as popular a vegetable as
the shelling pea.

Most varieties grown primarily for dry use are bush cultivars. Traditionally,
green-pod varieties were pole types. The two major green bean classics, `Blue
Lake' and `Kentucky Wonder', are pole beans. In this century, in the United
States, commercial green beans have come to be harvested by machine, and bush
green beans predominate. In Japan, where farmland is used very intensively, pole
beans are the major commercial green bean type.

Silbernagel's chapter in Breeding Vegetable Crops does not give a summary of

the genes that have been identified as affecting agriculturally important
characteristics, and I know of no such summary anywhere else. Silbernagel does
give references to the professional literature that provide gene lists and
information about breeding for specific characteristics. You can develop a
general feel for the kinds of situations involved by reading the section on the
genetics of peas in this appendix.

A number of traits are used as marker traits, however. Anthocyanin (purple or

red pigment) inheritance is complex but is often used as a marker to identify
hybrids. Hybrids between purple and green varieties are usually purple. That is,
purple-podded is dominant to green-podded, purple-flowered to white-flowered,
purplish foliage to green foliage, and so on.

Crosses of pole and bush cultivars normally give pole-type Fs.

Colored seed is dominant to white seed. The seed coat is maternal tissue,
however, and as such it expresses the genes of the mother, not those of the seed.

Harvesting, Processing, and Storage

Harvesting, processing, drying, and storage of beans is covered in Chapter 21.
 References

M. J. Silbernagel, "Snap Pea Breeding," in Bassett, Breeding Vegetable Crops,

pages 243-281; George, Vegetable Seed Production, pages 192-199; F A. Bliss,
"Common Bean," in Fehr and Hadley, Hybridization of Crop Plants, pages 273-
284.

Alliums (Onions)

Onions, garlic, chives, leeks, and shallots are all members of the genus Allium.
For information about which varieties correspond to which species, see Table 1.
 Flowering

Many onions are biennials or biennials raised as annuals. The common bulb
onion, for example, is an Allium cepa. It is ordinarily grown to produce a bulb
the first year. For seed production the bulbs are inspected, rogued, and stored for
replanting the following year. Many onions need a cold treatment to induce
flowering. For most varieties two weeks in the refrigerator will suffice.

Perennial species such as A. fistulosum normally don't flower the first year but
do flower in subsequent years. Some species and some varieties don't flower at
all, or flower only under special conditions. Instead, they make bulbils in the
flower heads (top sets) or multiply by division at the base.
 Breeding System

Alliums form large clusters (umbels) of tiny flowers. Allium flowers are
perfect, but alliums are outbreeders. Bees and flies of various types and other
insects are the pollinators. Alliums are selfcompatible, but individual flowers
usually don't self-pollinate because of protandry. Pollen shedding in an
individual allium flower occurs and is often completed before the stigma
becomes receptive, thus preventing self-pollination. However, many flowers in a
single umbel are opening, pollinating, and becoming receptive at various times,
so fertilization of flowers by

pollen from other flowers in the same head is possible and common.

According to Pike's chapter in Bassett, pollen begins to be shed shortly after

the flower opens, weather permitting, and is shed over a period of three to four
days. The stigma becomes receptive when the pistil reaches full size. In some
cases this is after pollen release is completed. In other cases the stigma becomes
receptive after, but the same day as, the pollen dehisces (Dr. Kalloo's Vegetable
Breeding vol. 1, page 25). The stigma remains receptive for several days.

Cytoplasmic male sterility exists and is the basis for the hybrid onion seed
industry (see Pike).
 Isolation

For commercial production, isolate alliums by 1,000 meters from all others of
the same species. You also may cage plants in small plots, but if so,
handpollination or introducing insect pollinators is necessary. Fly pupae can be
placed in the cages so that they emerge clean and ready to go to work. To collect
blowflies, place a piece of meat in an open jar and put it outdoors. Wild flies will
lay eggs in it, and the larvae will pupate on the sides of the jar. Caged flies will
live longer if water is available.

Most seed savers will want to avoid all these problems by growing just one
variety of each allium species. Note that varieties that aren't flowering don't
count. You can grow other varieties of A. cepa bulbing onions if you, for
example, are growing just one variety for seed and the others are making bulbs,
not flowering. Usually your neighbors' gardens present little or no problem,
because they normally won't have flowering alliums, only bulb-producing first-
year material.
 Inbreeding Depression

Alliums generally display inbreeding depression, enough so that self-

pollinating for even two generations causes problems with vigor.
 HandPollination and Crossing

Handpollination can be performed by emasculating the flowers and then

transferring the pollen several days later. Most of the flowers on the umbel must
be sacrificed to prevent contamination, however, and individual flowers produce
just a few seeds.

Allium crosses are usually performed as fertile X fertile crosses instead of as

handpollinations. Commercially, small numbers of plants of the two varieties to
be crossed are caged together, and pollinators are introduced. Alternatively, you
could transplant two plants to some area away from other varieties of the same
species and let the bees do it.

My crosses of alliums have all been wide crosses, and I do them as

handpollinations. I start by removing all the opened flowers on an umbel as well
as buds too small to work with. I open and emasculate the buds that are big
enough to work with by removing all their petals, sepals, and stamens. (See
illustration on page 307.) I don't try to cover the umbel. I depend on the lack of
petals, pollen, and nearby flowers to discourage pollinators. I have not checked
this method rigorously, but I've never seen pollinators on any flowers I've
prepped this way.

Under my conditions, most species of alli

ums seem to finish releasing pollen days before the styles finish growing, so
none of my prepped flowers are ready to hand-pollinate. I inspect my prepped
flowers on subsequent days until the most advanced style reaches full size. Then
I hand-pollinate it and all the others on the umbel. (This is probably the right
time for the oldest flower and too early for the others.) I repeat the
handpollinating daily for the next few days. In other words, each flower I prep
receives several doses of pollen, some too early, others probably too late, and,
hopefully, one absolutely optimally. This takes very little time compared to that
required for prepping the flowers in the first place.
 Genetics and Breeding

Most of the breeding work has been with bulbing varieties of A. cepa. There
are three basic kinds of onions raised for bulbs: fresh eating onions, storage
onions, and dehydrating onions. In addition, some A. cepa varieties are used as
scallions. Because onions have a precise photoperiod requirement for bulbing,
much effort has been spent on developing all kinds of onions for each specific
region.

The breeding work seems to be based largely on general principles. There are
very few identified genes in onions. The genotype vv has virescent seedlings; gl
gl has glossy foliage; pr pr is resistant to pinkroot.

The inheritance of bulb color in A. cepa is controlled by three loci: I, C, and R.

I and C control whether there is any pigment. R controls whether it is red or
yellow. R_ is red; rr is yellow I is a dominant pigment inhibitor. Pigmented
plants must be ii. II is white, no matter what the other genes are; Ii is buff. The
other inhibitor is a recessive. C is colored; cc is white. To have colored bulbs, a
plant must carry genotype iiC_. If it does, its color is determined by the R locus.
iiC_R_ is red; iiC_rr is yellow. In summary, red varieties are iiCCRR, yellow
varieties are iiCCrr, and whites may be either II____ or __cc__.

Harvesting, Processing, and Storage

Allium seed heads begin dropping seed from the earliest flowers as, or very
shortly after, the latest flowers dry up. And various seed heads are ready at
various times. So for a seed crop of an allium, it is important to watch the plants
and harvest the various seed heads as they are ready. I usually harvest at least
three times during the season. Once some of the seed on a head has dried up, it
should not be exposed to rain or watering. It should be removed and taken
indoors to finish drying.

Small amounts of seed can be cleaned by rubbing seed heads between the
hands and then winnowing. Once the seed heads have been broken up, though,
the debris is hard to separate from the seed. Often sufficient seed can be obtained
by shaking whole seed heads without breaking them up; much seed drops free,
and the debris isn't generated. When I need a lot of seed, I sometimes break up
the seed heads by rubbing but don't bother separating seed from debris. I sow it
all.

Allium seed is more easily damaged by mechanical processing that most seed.
It must be handled gently. Allium seed also does not store as well as the seed of
brassicas, legumes, and many other common vegetables.
 References

Leonard M. Pike, "Onion Breeding," in Bassett, Breeding Vegetable Crops,

pages 357-393; H. A. Jones and L. K. Mann, Onions and Their Allies; Dr.
Kalloo, Vegetable Breeding vol. 1, page 25.

Brassicas (Cabbages and Relatives)

The brassica family, Brassicaceae, includes numerous important vegetables.

This section covers the genus Brassica. Additional genera and species in the
family are covered in Table 1, including Armoratia rusticana (horseradish);
Barbarea spp. (cress); various Crambe spp., including Crambe maritima (sea
kale); Eruca vesicaria (rocket); Hesperis matronalis (sweet rocket); various
Lepidium spp., which include maca, garden cress, and Virginia pepperweed;
Raphanus sativus (radish); Rorippa spp. (watercress); and Wasabiajaponica
(Japanese horseradish).

Brassica oleraceae includes cabbage, kale, collards, broccoli, sprouting

broccoli, cauliflower, Brussels sprouts, and kohlrabi. B. campestris includes
turnips, Chinese cabbage, pak choy, Chinese mustard, turnip rape, and a few
others. B. napus includes rutabaga, swede, rape, and rape-kale. B. juncea
includes leaf mustards. Condiment mustards and wild mustards account for a
few other species. For details see Table 1.

On Brassica taxonomy I am following Downey, Klassen, and Stringam, as

found in Fehr and Hadley's Hybridization of Crop Plants. The abbreviation gf-in
Table 1 refers to genomic formula. The chromosomes of the various species are
microscopically distinguishable, so it is possible to understand the evolution of
and relationships between the various species by examining the chromosomes.
The genomic formula indicates which genomes are involved.

Example: A is the haploid B. campestris genome, so Table 1 includes the

symbol gfAA for B. campestris, indicating it is a normal diploid. B. nigra (black
mustard) is gf-BB. B. juncea (leaf or India mustard, of which my much-beloved
`Green Wave' is a representative) is gf-AABB. In other words, it is a tetraploid,
an amphidiploid, and was derived by a cross of a campestris and a nigra
followed by chromosome doubling.
 Flowering

Biennial brassicas must be overwintered to achieve flowering. Whether this

can be done outdoors depends on the cold hardiness of the variety and the
region. Most brassicas require some period of cold to flower. Brassicas that form
tight heads, such as cabbage, may require scoring of the outer leaves to allow the
flower stalk to emerge. Cut two slashes in an X into the top of the head.
 Breeding System

Brassicas have perfect flowers. Most brassicas are strong outbreeders.

Varieties of the species B. oleraceae, B. campestris, and most others are usually
self-incompatible. (An exception is summer cauliflowers, which are usually
selfcompatible.) The tetraploid species B. juncea and B. napus, however, are
selfcompatible inbreeders. The strength of the incompatibility varies from
variety to variety and plant to plant, as well as under different conditions. The
normal pollen vec tors are bees, flies, and insects of various kinds.

In brassica flowers the stigma becomes receptive some days before the bud
finishes developing and opens. In some species and varieties the stigma
protrudes slightly and is exposed during the last day before the flower opens.
The anthers dehisce some hours after the flower opens. When the flowers open
and the anthers dehisce depends strongly on the weather.
 Vegetative Propagation

Brassicas are easy to propagate vegetatively by shoot or root cuttings. These

methods can be used to increase or maintain a specific plant for further
evaluation or to overwinter or apply cold treatments to induce flowering.

For shoot cuttings, cut off a vegetative shoot, place it in potting soil, and set it
in a moist place to root. Rooting hormone may be used. For root cuttings, use
root sections about the thickness and length of a pencil. Make a direct cut across
the upper end and a slanting cut across the lower end of each. Place them
vertically in potting soil or in the ground.
 Isolation

For commercial production brassicas are isolated from all others of the same
species by 1,000 to 1,500 meters. Note, for example, that cabbage, kale, collards,
broccoli, cauliflower, and Brussels sprouts are all B. oleraceae and will cross
freely with each other. The other species are equally diverse (see Table 1). Some
mustards are B. juncea. Other plants we call mustards are actually leaf or stem
forms of Chinese cabbage (B. chinensis). Which is which is often not obvious.

In my garden, with its myriad of pollinator species, all the brassicas seem to
cross to some extent whether they are the same species or not. Hybrids between
the species are usually highly, though not totally, sterile. I isolate all brassicas as
if they were one big (somewhat confused) species.

There are a number of wild mustards of various species as well, and it usually
isn't obvious what species they are. I eliminate any wild mustards growing near a
variety I am growing for seed.
 Inbreeding Depression

Most brassicas are subject to inbreeding depression. The exceptions are those
that substantially inbreed naturally, such as summer cauliflower and the B.
juncea varieties (see Table 1).
 HandPollination and Crossing

Handpollination of brassicas is done in the bud stage. If the plant is self-

incompatible, emasculation prior to doing crosses isn't required. According to
instructions given by Dickson and Wallace for cabbage, buds can be used from
one to four days before opening. Self-incompatibility has developed by the day
before flowering, however, so if bud pollination is being used for self-pollinating
an incompatible plant, earlier buds should be used. (See illustration on p. 307.)

Generally the opened flowers are removed, and the six to eight buds on the
inflorescence - representing all those that will open for the next four days - are
opened, emasculated if necessary, and handpollinated. After handpollinating, it is
necessary to exclude pollinators. Dickson and Wallace specify a cheesecloth bag
around the blossoms.

I used to spend much time making brassica crosses by hand. I never do it that
way anymore. I just plant the two different varieties or species side by side. I'll
always get a few percent of crosses from the plants at the edges of the blocks
where the two varieties are adjacent. A few thousand seeds from these plants can
be sown in a few square feet of soil the following year, and the hybrids
recognized as seedlings.
 Hybrid Seed Industry

Because most brassicas are so subject to inbreeding depression and display

such large amounts of hybrid vigor, there is much natural advantage to hybrid
brassicas.

My favorite mustard, `Green Wave', is an open-pollinated type and is more

vigorous than any other mustard I've seen, including the hybrids. But `Green
Wave' and other B. juncea types are in a very real sense hybrids already, because
they are tetraploids containing two entirely different brassica genomes from
different species. This apparently gives them the vigor of hybrids, but they still
have the genetic stability of any other open-pollinated variety.

The hybrid seed industry for brassicas has been based on the incompatibility
system. For information on the incompatibility system and how it's used to
produce hybrid seed, see Dickson and Wallace in Bassett. Cytoplasmic sterility
also exists and is starting to be used to produce hybrid seed.
 Genetics and Breeding

A few of the genes that have been identified in cabbage are given in the
paragraphs that follow (from Dickson and Wallace).

T is a dominant gene causing the plant to be tall. Most cabbage cultivars are tt;
tall cabbage plants tend to fall over. Ax is associated with axillary buds.

The A locus controls anthocyanin. There are various alleles. A recessive

inhibitor, c, suppresses anthocyanin formation in all parts of the plant; cc has no
anthocyanin, whatever the rest of the genotype. B is associated with color
limited to the midrib. G and H are intensifiers associated with deep purple color.
In plants carrying M the color is magenta instead of purple. Other genes also
affect color (see Dickson and Wallace).

The gene En is associated with entire leaves (renamed from E). W is

associated with wide leaves. The gl locus (glossy) has a series of alleles that
determine the amount of wax on the stems. With Hr-1 the plant has hairy first
leaves and hairs on the margin, with "sometimes poor expression in
heterozygote." Smooth leaves are associated with sm along with wr. Pet controls
whether the plant has petioles and leaves instead of sessile leaves. Four genes -
W-1, W-2, W-3, and W4 - are associated with frilling of leaves.

K is a dominant gene associated with heading. Hybrids between heading and

nonheading forms are intermediate. Hybrids between varieties with pointed
heads and those with round ones have round heads, but the inheritance isn't
simple. Crosses between varieties with savoy heads and those with hard heads
produce hard-headed offspring. Other such patterns are as follows, with the type
the hybrid resembles being listed first: annual habit versus biennial; early
maturity versus late; few wrapper leaves versus many; wide core versus narrow.

Core length is associated with two codominant genes with additive effects.
Head splitting is determined by three codominant genes that act additively.
Crosses of red and green plants are usually pink. Savoy leaf texture involves
three or more genes.

For information on genetics of disease resistance, see Dickson and Wallace.

Harvesting, Processing, and Storage

Brassica seed harvesting, processing, and storage is covered in Chapter 21.

 References

Michael H. Dickson and D. H. Wallace, "Cabbage Breeding," in Bassett,

Breeding Vegetable Crops, pages 395-431; R. K. Downey, A. J. Klassen, and G.
R. Stringam, "Rapeseed and Mustard," in Fehr and Hadley, Hybridization of
Crop Plants, pages 495-509.
 Squash and Pumpkins

The words squash, pumpkin, and marrow are applied more or less
indiscriminately to fruits of four species: Cucurbita pepo, C. maxima, C.
moschata, and C. mixta. Generally, most of what are referred to in the United
States as summer squash are C. pepo. Many of the winter squash are C. maxima.
Some are C. moschata or C. mixta. The words pumpkin and marrow tend to be
used for some varieties within each of the species but not others.

There are anatomical differences between the species, but most people know
which variety is which because the seed pack or catalog gives the species or they
look it up. The C. pepo species includes the acorns, many of the orange
pumpkins, the summer crooknecks, the scallops, the zucchinis, `Spaghetti', and
many others. C. maxima includes other pumpkins, the Hubbards, the bananas,
'Arikara', `Buttercup', `Golden Delicious', the Kuris, and many others. C.
moschata includes the butternuts, the cheeses, the melon squashes, and others. C.
mixta includes `Tennessee Sweet Potato', most of the cushaws, and others. For
extensive lists of which cultivars are in which species, see Ashworth's Seed to
Seed, The Garden Seed Inventory, or a Winter Yearbook of the Seed Savers
Exchange.
 Breeding System

Squash and pumpkins are outbreeders, but they seem to have some genetic
characteristics typical of inbreeders. (See the sections on inbreeding depression
and on genetics and breeding.) The flowering pattern is monoecious; there are
separate male and female flowers.

The first few flowers are usually male; then both male and female flowers are
produced. Honeybees and bumblebees are the usual pollinators. Squash and
pumpkins are all selfcompatible.

The female and male flowers are easy to tell apart because the females have a
swollen area at the base, which represents the ovary. It looks like a little fruit and
will develop into one after fertilization. The female flowers produce nectar only;
the male flowers produce both pollen and nectar.
 Isolation

The recommended isolation distance for different varieties within a species of

cucur

bits is 1,000 to 1,500 meters. The four different species don't cross with each
other spontaneously, however, so you can grow one of each without isolating
them from each other. That's only appropriate if you are 1,000 meters or more
from any neighbor growing a variety of the same species.

Most seed savers are not 1,000 meters from the nearest squash-planting
neighbor and/or they want to grow more than one variety of a species. So they
hand-pollinate. The separate male and female flowers, the large flower size, the
high success rate, and the fact that you get hundreds of seeds from each
handpollination make handpollination especially easy.
 Inbreeding Depression

In spite of their basic outbreeding anatomy, squash and pumpkins display little
or no inbreeding depression.
 HandPollination and Crossing

The stigmas of squash don't become receptive until the flower opens, at which
time pollen also dehisces. Therefore, it is necessary to emasculate at one time
and pollinate later.

The normal procedure is to tape male and female blossoms shut with masking
tape applied to the end of the flower bud. This should be done in the late
afternoon or early evening of the day before the bud is due to open for the first
time. These buds are beginning to show color, and the petals may be showing
signs of opening at the ends. If the bud is taped earlier in the day, the tape will
damage the bud as it grows. It's useful to put a flag or marker of some kind in the
ground next to each taped flower; otherwise you might have a hard time finding
the flowers again.

Male flowers as well as females must be taped. Otherwise bees can get to and
remove the pollen before you do, and the little that is left is contaminated.

Handpollinations are done the next morning, as soon as possible after the
pollen is released from the anthers. This is usually about 9 A.M. in my garden.
Release of pollen happens earlier on warm dry days and later on cool or wet
ones. When male flowers are ready to use, the pollen is yellow and fluffylooking
as it clings to the anthers. Pick male flowers to use, as pollen applicators. Untape
them by removing the end of the flower with the tape and tear off the petals. It's
best to use all the pollen from a male flower for each female flower.

After the male flower is in hand and prepared, untape the female flower by
removing the tips of the petals with the tape. Then brush pollen from the male
onto the stigma of the female. Retape the end of the female flower closed with a
new piece of tape.

Some bees will bite into taped flowers before or after the handpollination.
Check for holes in the taped flowers or trapped bees.

Cross information is initially recorded on a tag on the stem of the female

flower. Once the fruit has developed, make sure to write the information on the
fruit before it is harvested. Tags tend to get lost during harvest.
 Hybrid Seed Industry

There is enough hybrid vigor in squash and pumpkins so that hybrid varieties
can be

advantageous. Hybrids may be earlier than parent varieties and may show
increased yield.

The hybrid seed industry is based largely on chemical sprays that inhibit the
development of male flowers. The variety to be used as the female is planted in
rows that are sprayed. For every so many rows of the female variety, there is a
row of the desired male variety, which isn't sprayed. A more laborious method is
to remove the male flowers from the female variety by hand and provide rows of
the male variety as pollinator.
 Genetics and Breeding

Although squash and pumpkins and many other cucurbits clearly are basically
outbreeders, they seem to have adapted secondarily to being inbreeders.

Whatever the cause, squash and pumpkins have the anatomy of outbreeders
but seem to have some of the genetic characteristics of inbreeders. Inbreeding
depression is not usually a problem in working with them. Extensive inbreeding
is used in developing most squash and pumpkin varieties. After an initial cross,
desired offspring are selected and inbred, often for a number of generations. In
other words, the breeding approach is very much as it is with peas and other
natural inbreeders, except that you do the inbreeding by handpollination.

Most breeding of squash and pumpkins is based on the general genetic

principles, not on specific information about the genes involved. Not many
specific genes have been identified. A few of those that have been follow (see
Whitaker and Robinson for more).

Some of the following genes that affect skin color have been identified in C.
pepo: B (bicolored fruit, deep yellow), C (colored fruit, green), 1 (lightens fruit
color), r (recessive white), W (dominant white), Y (yellow), and St (longitudinal
stripes). Two other genes affecting fruit skin color in C. maxima are bl (blue at
maturity) and Rd (red at maturity). In C. pepo, wf is associated with white flesh.

In C. pepo, Di, Hr, and Wt are associated with disk-shaped fruit, hard rind, and
warty skin, respectively. Bi and cu affect the amount of cucurbitacin, and thus
the bitterness of the fruit.

In both C. pepo and C. maxima the gene Bu is associated with bush growth
form.

Harvesting, Processing, and Storage

Squash and pumpkins must be allowed to mature fully if they are to be used
for seed. In addition, there is an after-ripening process. Seed quality and viability
are improved if the harvested fruit is stored a month or more before the seed is
extracted. Most home gardeners remove the seed whenever they eat the squash,
as long as it has had at least a month for after-ripening.
 Small amounts of seeds can be put in water and worked with the hands to
remove flesh and debris. The seeds are then spread in a single layer on a tray and
allowed to dry. (They will stick to paper.) It's not always easy to get the seeds
clean using this method, however, and it's a lot of work for more than a few
seeds.

Alan Kapuler's method for processing cucurbit seeds yields cleaner seeds with
higher germination rates and is suitable for small or larger amounts of hand-
processed seeds. Alan opens the fruit and puts the seeds

and pulp into buckets of water. Both float. He leaves them in the water for one to
three days - until the seeds sink. It is important to look at the seeds every day and
catch them immediately after they take up enough water to sink. At this point the
good seeds with large kernels inside sink, and they are easily separated from
both the debris and the seeds with poor germ.

After separating and rinsing the seeds, Alan immediately puts the seeds on
trays in his food dehydrator and dries them. If you clean seeds using soaking, it
is important to dry them promptly and rapidly.
 References

Thomas W Whitaker and R. W Robinson, "Squash Breeding," in Bassett,

Breeding Vegetable Crops, pages 209-242; Ashworth, Seed to Seed, pages 114-
120; Carol Deppe, "Custom-Crafted Vegetables I & II" (Squash seed saving and
breeding.) National Gardening Magazine, June/July & August, 1998.

Corn (Maize)

Sweet corn and the various field corns - flour, flint, and dent - are all Zea
mays. So are popcorn and the ornamental corns. The kind and arrangement of
starches in the endosperm of the seed determine the basic kind of corn. Flint
corn seed has a lot of flinty endosperm and just a little of the floury kind. Flour
corn has mostly floury endosperm with just a little flint. Dent corn has a good bit
of both types of endosperm. You can tell which class a variety is by cutting open
a few of its seeds with wire cutters (longitudinally) and looking. The flinty
endosperm looks like flint or glass and is either yellow or white. The floury
endosperm is soft, grainy, and white, like chalk.

Sweet corn carries one or more genes that alter starch synthesis, causing the
kernels to be sweeter when green but shriveled when dried. The wrinkled seeds
have a flintyappearing endosperm, but its composition is different from that in
the flint types.
 Breeding System

Corn is an outbreeder. It is monoecious. Male flowers, called tassels, emerge

from the top of the plant. Female flowers, called ear shoots initially, emerge
from some of the leaf axils on the plant. Once fertilized, they develop into the
ears.

Each silk in an ear shoot or ear is actually the very long stigma of an individual
flower. The silk is receptive over its entire surface, so if a pollen grain falls
anywhere on it, it germinates, grows up the silk, fertilizes the egg, and produces
a single kernel. If only a few such fertilizations happen, the corn ear will be
mostly blank and have just a few seeds. Corn is selfcompatible.

The pollen of most varieties begins to dehisce a few days before any of the ear
shoots are ready. The silk is receptive from when it first appears to until it dries
up. The pollen is very sensitive to heat and to drying out. It is mostly shed in the
morning as soon as it warms up and the humidity drops. Corn pollen is short-
lived. It usually lives from minutes to a few hours.
 Isolation

Corn pollen is light and is blown long distances by the wind. In areas where
corn is common, the air is full of pollen from the fields and gardens in the
region. The isolation distance recommended by George is "up to one kilometer."
Various factors may reduce the amount of isolation you need.

You may not need as much isolation for a very early variety if it flowers before
most of the corn in your area. You may not need as much isolation for a white
variety if most of the corn in your area is yellow. Pollen from yellow corn will
produce yellowish kernels if it fertilizes white corn, and you can cull such
kernels out by hand. You also can reduce the isolation distance if you have a
large enough patch so you can save seed from only the plants in the middle.
Most home gardeners have too many corn-growing neighbors for any of these
methods to be reliable, and they save corn seed by handpollinating.
Inbreeding Depression

Corn displays extreme inbreeding depression.

 HandPollination and Crossing

Ashworth gives an excellent description of the process in Seed to Seed, which

also includes photographs of every stage. Basically it involves bagging tassels to
collect pollen; covering ears before the silks appear, to prevent contamination;
and doing the pollination.

Corn ear shoot bags and corn tassel bags may be purchased from the
companies listed in Appendix E. When I first started doing handpollinations, I
substituted envelopes for ear shoot bags and ordinary grocery bags for tassel
bags. But it doesn't usually rain in the summer here. The professional shoot and
tassel bags are water resistant.

To bag the ear shoots, you have to learn to recognize them before the silks
emerge. Where they develop on the plant (at what node) and what they look like
depends on the variety.

To bag an ear shoot, grasp the leaf whose axil it is hiding in and strip it down
and off the plant. Then pull the ear shoot bag down over the shoot and firmly
into the crevice between the shoot and the stalk. You may need to gently slice the
leaf that goes around the shoot and the stalk to be able to pull the bag down far
enough to secure it.

You can bag plants over a few days without pollinating them. If you
accumulate bagged plants over several days, you will need to cut off the silks
occasionally every two or three days so they don't emerge from the bag.

The day before you plan to pollinate, cut off the tip of the ear shoot to expose
the silks or a fresh area of silk. The following day, this area will be a brush of
dense silks of ideal length for pollinating. (If you bag the ear shoots the day
before you plan to use them, you cut off the tips at the same time.)

Obtain pollen by bagging tassels the day before you do your handpollinations.
Tassel bags are much larger than ear shoot bags. Give the tassel a good shake to
get rid of any foreign pollen that may have landed on it. Then position the bag
over the tassel and fold the bottom of the bag up so that the bag is tight around
the tassel. Staple the bag thus, or attach it with paper clips. In the late morning
(or later) of the next day, give the tassel a shake to free as much pollen as
possible before removing the bag.

To hand-pollinate, dust pollen onto the

brush of silks on the ear shoot. Then use a tassel bag to cover the ear.

Cover the ear loosely enough so that it will be able to grow without being
constricted by the bag. This is necessary initially to protect the ear shoot from
contamination. Later, what matters is that you've usually written the cross
information on that bag, so you don't want it wandering away from the ear. Pull
the tassel bag down over the shoot, then fold the lower ends around the stalk and
staple or clip them there.

When I'm doing handpollinations involving crosses, I plant the corn rows at
least 3 feet apart. Otherwise, it's too hard to get into the patch without jostling
the plants, causing them to release clouds of pollen, and increasing the
probability of contamination. I also walk through the patch with care to avoid
bumping plants and making them release pollen.

Corn pollen does not store easily. You can store it for about a week, though, if
you collect it soon after it starts to be shed for the day and refrigerate it under
conditions of high relative humidity (such as with a damp paper towel in the
container with the pollen).
 Hybrid Seed Industry

Corn displays extreme inbreeding depression and extreme hybrid vigor. There
are many advantages to using hybrid varieties.

There are two main methods for producing hybrid corn commercially: using
male-sterile cytoplasm and detasseling. The detasseling method is usable with
any variety, not just specialized ones with special cytoplasm. It also can be used
on any scale. This is the method I describe here (after the description in George's
Vegetable Seed Production).

To produce a hybrid, plant alternating rows of the two parents. Usually you can
plant two to four rows of the variety you're using as the female for each row of
the variety used as the male, depending on how much pollen the male variety
produces and how efficiently it achieves pollination under your conditions.

When tassels of the male variety begin to emerge, remove them before they
open and shed any pollen. This is called detasseling. You must notice and
remove the tassels on tillers as well as those on the main plant. "It may be
necessary to detassel the crop up to seven times, at two day intervals, to ensure
that it is complete," George says. Home gardeners and farmers can easily
produce their own hybrid corn using this method.

If you are less than a kilometer from other corn patches and still want to
produce your own hybrid corn, plant extra rows of the variety you want as the
male parent around the whole patch to saturate the air with the desired pollen
and overwhelm any contaminating pollen in the air. Under these conditions you
normally hand-pollinate to maintain pure varieties. But you won't be saving seed
from the hybrids, so occasional contaminants won't matter.
 Genetics and Breeding

Corn genetics has been studied extensively. Coe and Neuffer's 112-page
chapter in Corn and Corn Improvement describes genes and genetic maps. So
much variability is present in corn and so many genes are known to be involved
in every trait that I am going to limit my discussion to just a few traits of the
seed itself.

The appearance of the kernel depends in part on the chemical composition of

the endosperm. The endosperm is triploid tissue. In the fertilization of corn, the
pollen tube grows down the silk and contributes two sperm nuclei to the egg.
One sperm nucleus unites with the egg to become the diploid zygote, which
develops further to become a baby plant inside the seed (the germ). The other
sperm nucleus unites with two polar nuclei from the egg to become a triploid.
That triploid nucleus divides and gives rise to the endosperm, the storage
material that constitutes the rest of the interior of the seed.

The endosperm is normally yellow or white. The outer layer of the endosperm,
the aleurone, is often colored. At least fifteen genes are known to affect color in
the aleurone alone. The very outer layer of the seed is called the pericarp and is
maternal tissue.

The characteristics of the pericarp depend on the genotype of the mother plant,
not the seed. Characteristics of the aleurone and endosperm depend on the genes
of the seed. But the aleurone and endosperm have two copies of egg genes and
one copy of pollen genes instead of one of each.

The Su locus is the major gene responsible for all the early varieties of sweet
corn. Sweet corns are susu; flint and other field corns are SuSu. Just that single
gene is enough to cause the greater sweetness and wrinkled seed. However,
many other genes are involved as well. Most sweet corns have thin, tender
pericarps, for example, and other characteristics of texture and flavor.

The other major gene that can cause the sweet phenotype is sh2 (shrunken). It
is a different gene, but the recessive condition also causes sweet flavor and
wrinkled kernels. Both the sweetness and the wrinkling are more extreme in sh2
varieties, enough so that there is more of a problem with seed viability and
shattering. The varieties listed as supersweet are usually sh2 types.

Since both sweet genes are recessive, two varieties that are of different types
must be isolated from each other if they are to produce sweet kernels. They
differ by two loci, one variety being susu Sh-2Sh-2 and the other being SuSu sh-
2sh-2. If the former is pollinated by the latter, for example, the endosperm will
be susuSu Sh-2Sh-2sh-2. It will have at least one dominant gene at each of the
two loci and will be of the field type.

If you have blocks of the two different types of sweet corn near each other, the
kernels pollinated by other plants within the variety will be sweet, but those
pollinated by the other variety will develop faster and be hard and tough.
Likewise, any field corn near a patch of either sweet corn variety will give rise to
sweet corn ears that have some hard, nonsweet kernels in them.

Breeders now have other genes and combinations that can be used to generate
the sweet phenotype as well. There are also modifiers of the other genes. One
especially useful one is se, sugary enhancer. This recessive gene, in the presence
of susu, increases the sugar content and delays the formation of starch. It is
especially useful because it gives us supersweet varieties that don't have as many
seed viability problems as those generated using the sh-2 gene.

A number of different genes can cause the

floury phenotype. When flint or sweet corn is pollinated by a flour corn, the
kernels are usually of the flour type.

Many dozens of genes are involved in the color of the seed. I'll mention only
the practical matter that white varieties pollinated by any colored variety
produce colored kernels and that black varieties crossed with any other color
variety produce black or blackish kernels.

Harvesting, Processing, and Storage

Corn to be saved for seed is allowed to dry on the plants for as long as
possible. Ears are then removed and stored in a dry place until there is time to
deal with them. Home gardeners usually shell out seed corn by hand. It's very
common to have insects of various kinds in homegrown corn seed, so common
that treatment of some sort to kill insects and insect eggs is often essential. I dry
my corn seed to "extra-dry" and then freeze it, at least temporarily, as described
in Chapter 21.
 References

George, Vegetable Seed Production, pages 288-293; Ashworth, Seed to Seed,

pages 187-196; Karl Kaukis and David W Davis, "Sweet Corn Breeding," in
Bassett, Breeding Vegetable Crops, pages 75-584; W. A. Russell and A. R.
Hallauer, "Corn," in Fehr and Hadley, Hybridization of Crop Plants, pages 299-
312; E. H. Coe, Jr., and M. G. Neuffer, "The Genetics of Corn," in Sprague, Corn
and Corn Improvement, pages 111-223, Carol Deppe, "Parching Corn," National
Gardening, June 1997.

This appendix covers the technical aspects of making crosses and hand-
pollinations.

Making Crosses

There are two basic aspects to doing a cross. One is physically doing it or
arranging for it to happen. The other is recognizing it when it has happened.
Usually we need to do both. "Physically doing it" means collecting pollen and
pollinating by hand; "arranging for it to happen" means getting the bees (or wind
or other pollen vectors) to do the cross for us. The second aspect, "recognizing
when a cross has happened" means being able to distinguish the hybrids
resulting from the cross from any self-pollinations, or selfings, of the maternal
parent.

Sometimes we carefully transfer pollen from one flower to another after

emasculating the recipient. (Details on how to collect and transfer pollen and
emasculate flowers depend on the species and are covered in Appendix A.)
Sometimes, however, the flowers are too small or otherwise too difficult to
emasculate, so we do a fertile x fertile cross, which means that we pollinate the
female with pollen of our choice without emasculating it first. This usually
means that the resulting seed is a mixture of two kinds that resulting from
fertilization by the pollen that we transferred deliberately and that resulting from
fertilization by pollen from the female flower itself. In other words, the seed may
represent both the desired cross and undesired selfings. Thus we must depend on
our ability to distinguish the hybrids from the selfings.

Even when we do deliberate hand-pollination after emasculation, there are

often some contaminating selfings. For that reason crosses are often done in the
direction that permits distinguishing the hybrids from the selfings. Here direction
refers to which variety is used as the female parent. You can cross a short pea
variety with a tall one in one of two ways: short female x tall male, or tall female
x short male.

In the cross short female x tall male, we emasculate flowers of a short plant
and fertilize them with pollen from a tall plant. Since tall is dominant to short,
the hybrids - that is, the successful crosses - will be tall. Any accidental selfings
will simply be plants of the maternal variety, which will be short. When we grow
up the offspring, we can easily distinguish which are hybrids and which are
selfings.

The reciprocal cross - the cross done in the opposite direction - is tall female x
short male. Again, we expect the hybrids to be tall. And again, we expect
accidental selfings to be of the maternal variety. But in this case the maternal
variety is tall, so the accidental selfings are tall, and both the successful crosses
and the selfings look the same. There is no way to distinguish which is which.

Most geneticists would do a cross of a tall variety and a short variety in the
direction that involved using the short variety as the female unless there was
some strong reason to do otherwise. That way, the accidental selfings could be
distinguished from the crosses and eliminated.

If two varieties look different, the F, hybrid can't resemble both of them. If we
do

the cross in both directions, in at least one direction we should be able to

distinguish the hybrid from the maternal parent. Often we can distinguish it from
both parents. Anytime we grow up an F,, we should grow some of each parent
variety along with it and compare it to both.

The genes or traits that are most useful for distinguishing hybrids from selfings
are called marker genes, marker traits, or markers. Those that show up in the
seed, the seedling, or the young plant are the most useful, because we can
discard the mistakes at an early stage. For example, round-podded peas when
crossed with wrinkle-podded peas produce a round-podded seed. If the wrinkled
variety is used as the maternal parent, the mistakes are obvious. If a grain of
pollen from the mother has fertilized one of the eggs that produced a seed, there
will be a wrinkled seed in the pod among the round ones.

In most cases tall plants, when crossed with short ones, produce tall or
intermediate hybrids. Determinate (bush) plants crossed with indeterminate ones
(vine types) usually produce indeterminate hybrids. Plants with purple pigment
in the leaves or stems crossed with those without any purple pigment usually
produce hybrids that are purple, purplish, or pink. Plants with smooth-edged
leaves when crossed with plants with dissected leaves usually produce hybrids
with leaves that are intermediate in shape. These characteristics are all
distinguishable in young plants or even seedlings.

In some cases the two parents are so similar in appearance that we do not
expect to be able to distinguish the hybrid from either parent. In other cases we
want to do a cross in a given direction because we want to preserve the
cytoplasm of a particular parent. In these cases we emasculate accurately and
make exactly the cross we want if at all possible.

Even when a cross is made exactly and properly, the progeny may derive from
only the mother. This phenomenon is called apomixis.

Apomixis is an aberrant form of meiosis in which seed is produced without any

contribution of genetic material from pollen, which means that this genetic
material is excluded from the zygote, and hence the next generation. What
matters to us is that apomixis interferes with our ability to do crosses, because
the plant doesn't use the pollen we give it.

Some species use apomixis as a regular form of reproduction. In many

dandelions, for example, the seed is from ordinary-looking flowers but isn't
derived via fertilization. In dandelions, though, as in many plants, the extent of
apomixis depends on weather conditions, variety, and/or part of the season.

In most plants apomixis represents the rare accident instead of the ordinary
form of reproduction. In others it happens fairly commonly, but normal sexual
derivation of seed is still the norm. Even where apomixis is rare, we may see it
when we do crosses - especially if we have excluded all ordinary pollen and
provided only pollen that is somewhat incompatible (such as from a different but
related species). Whenever we do a cross, especially a wide cross or a selfing of
a plant with an incompatibility system, we need to realize that the rare apomictic
events may be more common or easier to achieve than the cross.

We always record crosses to distinguish which parent is maternal and always

compare presumptive F, progeny with the two parents. When all the progeny of
an apparent cross resemble only the maternal variety and show no characteristics
of the paternal variety, an accidental selfing or apomixis may be responsible.

When working with outbreeding species, it is often easier to arrange for a cross
to happen and then screen the offspring for the hybrids than to hand-pollinate. I
often just grow or transplant two plants, one of each variety ("partners," I call
them), in a spot where they are mostly by themselves. I plant them very close
together - even bend them so that their flowering branches intertwine. That way,
as insects pollinate, they will often be going from one plant to the other. They do
the cross for me.

Remember that just putting two plants together and letting bees, wind, or
whatever do the crossing works only with outbreeding plants. With strong
inbreeders the only way to get a cross is to transfer the pollen yourself.

To make a cross between two different varieties, we need to have them both
flowering at the same time, or we need to be able to store pollen. If I want to
cross an early corn and a late one, for example, I plant the late one as early as
possible in the spring and the early one later. I often use the variable spacing I
described earlier in Chapter 5. Each variety to be crossed is planted in a row with
spacing ranging from 6 inches at one end to more than is necessary at the other.
The purpose in this case is to spread out the period during which pollen and
plants ready to be pollinated are available. The more crowded the plants are, the
later they mature. With a range of spacings I can have plants pollinating and
being ready for pollination over a very much longer time period than with any
single spacing. This usually gives me appropriate material for crossing, even
when the maturity dates of the varieties are significantly different.

You can speed up corn pollen by several days by cutting the plant down and
standing it in the shade somewhere, or even by cutting the tassel off. Many kinds
of flowers will shed pollen somewhat earlier if you remove them and allow them
to start drying out.

Some kinds of pollen can be stored for long periods, and some can't. Where
information on storing pollen is available, I include it in Appendix A. In most
cases, however, no information is available. When I want to store pollen for just
a few days, I usually put freshly shed pollen or even whole flowers that have just
shed pollen in a vial or plastic bag to prevent them from drying out. Then I put
the container in the refrigerator. Most, though not all, pollen will keep for a few
days to a week or two in this fashion.

When I want to store pollen for longer periods, I usually desiccate and freeze
it. I put the pollen (or anthers with pollen, but without the filaments) in gelatin
capsules (the kind used for drugs and vitamins - many but not all pharmacies sell
empty gela

tin capsules). I put the open gelatin capsules on a layer of silica gel, seal the jar,
and leave it at room temperature for a few hours. Then I open the jar, close the
gelatin capsules, put them in a vial with more silica gel, cap it, and put it in the
freezer.

To do a cross with frozen pollen, I open the vial and remove one capsule. Then
I open the gelatin capsule and leave it lying around at room temperature for an
hour or two so that the pollen or anther can rehydrate before I use it.

Hand-pollination

For many uncommon vegetables and for wild species, you have to figure out
the breeding system and/or how to do crosses yourself, as there isn't any
information available. Even with common vegetables, the available information
is only a first approximation. Details depend on the exact variety and the
weather. You'll have to fine-tune the techniques to your varieties and conditions.
So whether you are working with a rare vegetable or a common one, you need to
be able to figure out breeding patterns and crossing techniques for yourself.

With peas, for example, the line drawings in Appendix A show that the proper
stage for pollination is when the flower-petal part of the bud extends slightly
beyond the sepals. This is long before the bud is full size -very long before it
opens - and just before the pollen is released. Since the stigma is already
receptive then, the flower is opened, emasculated, and pollinated all at once.

In early spring, with large-flowered varieties, I can make pea pollinations very
easily, because the buds become quite large before the pollen is shed; the line
drawings are more or less accurate. But later in the year (or whenever there is a
sudden warm spell) the situation changes. Pollen is shed much earlier, when the
bud is much tinier and at a stage when the petals don't yet extend beyond the
sepals. The dwarf varieties I've worked with have smaller flowers, and many
also shed pollen at an earlier stage.

There are six steps to hand-pollination: (1) obtaining pollen; (2) identifying
flowers or buds at the right stage to be pollinated; (3) emasculation where
appropriate; (4) transferring pollen; (5) protecting the fertilized female flower
(from foreign pollen or drying out) where appropriate by covering or sealing it
or using other methods; (6) eliminating competition from other flowers and
fruits where appropriate.

First of all, it is important to have the appropriate conditions. There is no point

in making hand-pollinations during weather that is too hot or too cold for pollen
germination or fruit set, for example. Watch the species and varieties you are
interested in and become familiar enough with them that you know when the
flowers normally open and when the pollen is normally released (time of day
and weather), as well as what kind of weather conditions are required for
ordinary pod or fruit set. Among the conditions that may matter are stage of the
plant, part of the season, weather during the day of pollination and the next few
days, and time of day.

It's usually easiest to get crosses to succeed

if they are early in the flowering season but not necessarily the very first flowers.
With peas, for example, I wait until the weather warms up enough so that
naturally fertilized pods will set. Then I do the crosses. Early flowers (weather
permitting) are much more likely to take crosses and will produce more seed
than flowers higher on the plant. I don't bother doing any crosses during hot
spells; even natural pollinations don't set pods very well then.

With peas I eliminate any other flowers or pods at the node where I am making
a cross so as to eliminate competition for nutrients. I usually eliminate noncross
pods that are lower than my cross or crosses as well. With some species it's
helpful or even essential to remove all fruits that formed before the one to be
crossed. In other cases eliminating competition isn't necessary.

Most pollen becomes viable just shortly before the anther dehisces (sheds it). In
general, it is normally maximally fertile and viable right after it is shed. Some
pollen continues to be viable for several days. Other pollen is dead within a few
hours. Whenever possible, I use pollen that has been shed that day. Freshly shed
pollen is usually moistlooking and more brightly colored than old pollen. Once
you've looked at several flowers of a variety and dissected a few buds at various
stages, you'll learn to recognize what fresh pollen looks like.

The information in Appendix A tells you when various flowers dehisce (just
after the flower opens, at a particular bud stage, and so on). In reality, though,
dehiscence is partly determined by the plant's maturity and partly by the weather.
Pollen frequently doesn't shed in cold or wet weather, even if the bud or flower
is at the right stage. In such cases the bud or flower usually continues to grow
and the pollen sheds at a later stage - usually very promptly within hours of
when it warms up or dries out.

For similar reasons most flowers don't shed pollen early in the morning; they
shed after it warms up and the fog or dew burns off. As well, most flowers don't
shed on rainy days. For some species crosses "take" much better if done at a
particular time of day. For most it doesn't matter much as long as the time and
weather permit pollen shedding.

You can see when pollen has been shed. Just touch an anther gently and see
whether your finger comes away colored. I also pay attention to bees and other
pollinators. If they are out working my flowers, it's because pollen or nectar is
available.

Bees and other insects remove pollen, so you have to get there first. With
brassicas, for example, I often pick the flowers I want as male parents in the
morning before the pollen is shed and bring them inside. By a few hours later,
the flowers outdoors are releasing pollen, and the bees are out grabbing it anther
by anther almost as fast as it is released. The anthers dehisce individually over a
period of time, so it's not easy to find pollen once the insects become fully
active. In addition, insects, pollinators or not, may have contaminated it.

Books for professional plant breeders usually show a variety of tools for
transferring pol len. A fine-tipped brush and a jar of 95 percent ethyl alcohol for
sterilizing things are common. You can obtain this alcohol in denatured form
from a drugstore, but I don't use it. I sometimes use rubbing alcohol (70 percent
isopropyl alcohol) to clean my hands and forceps when I have contaminated
them with unwanted pollen or am going between varieties. I haven't checked
rigorously to see whether the rubbing alcohol kills pollen; I depend as much on
using ample paper towels to dry my hands afterward as on the alcohol. Most of
the time I don't even use the towels or alcohol. I work with only one variety at a
time and go in and wash my hands and forceps in between.

With big flowers I just pick the flower that I want to use as the male, strip
away anything in the way, and touch the anthers with loose pollen to the stigma
of the female, using the rest of the flower as the handle. Alternatively, I pick
individual anthers with a pair of forceps and use them.

My hand-pollinating tools include a magnifying glass, a scalpel with

disposable blades (the fine-pointed type), laboratory forceps, flat toothpicks,
rubbing alcohol and paper towels (sometimes), a Magic Marker, and little labels.

I use the scalpel and forceps for dissecting tiny flowers when I'm trying to
figure out how to work with them. Flat toothpicks are great for transferring
pollen when the whole male flower or stamen isn't appropriate. I use a different
one for each cross and throw them away. The forceps are my major tool. I use
the kind that have fine-pointed tips and that open when you press them and close
when you release them, not the reverse. You can purchase tiny labels with strings
in an office supply store. For peas, I use very tiny ones, used for pricing jewelry.
Forceps, scalpels, and other hand-pollinating supplies are available from
Meristem Plants (see Appendix E). Ordinary tweezers and razor blades will do if
the material isn't too small.

Identifying female flowers at the right stage to be receptive and then

emasculating them can be difficult or easy, depending on the species.
Emasculation is not always necessary. The cucurbits have separate male and
female flowers; spinach is dioecious - it has male and female flowers on separate
plants. Even with perfect flowers, you may not need to emasculate if the plant
has a relatively strong incompatibility system or if you're getting in there before
self-pollen has shed and you expect to be able to distinguish hybrids from
selfings.

When you do need to emasculate, you have to do it before the pollen is

released and when the anthers are immature enough that pollen is not released by
handling. When possible, try to find a stage at which the stigma is receptive and
the anthers aren't, so that you can emasculate and pollinate at the same time. This
is easy with species such as peas, where the stigma becomes receptive long
before the anthers finish maturing. With species whose male parts mature at the
same time as or before their female parts, you need to emasculate at one stage,
mark the flower so you can find it again, and then pollinate at a later stage.

Pollen shedding is obvious, so you can easily figure out when to emasculate.
Usually

you want to emasculate as late as possible so that the bud or flower is as big as
possible and easy to handle. It isn't as obvious when the stigma is receptive.
With many flowers it seems to become receptive about as the style reaches full
size, but it can be earlier or later. If you are working the methods out yourself,
start by figuring out the latest stage at which you can emasculate, then try
several pollinations of emasculated buds at that stage and see if they work.

After doing a cross, you may need to cover the flower to prevent
contamination by foreign pollen. In some cases it is simply to keep the pollinated
stigma and style from drying out, because the hand-pollination has destroyed all
or part of the flower that usually protects them. Covering the crossed flower isn't
always necessary.

With inbreeders like peas, for example, which are not worked by or attractive
to insects, there is not much danger of contamination. Covering the crossed
flower is not necessary. In other cases, if the petals are removed, there is
nowhere for the normal pollinators to land, or they may not notice the "flower"
without the petals. So removing the petals may be enough to exclude pollinators.
In yet other cases the pollen attracts the pollinators, and if there are no anthers,
they won't land on the flower.

Watching the pollinators at work can give you ideas for how to do crosses with
a given species, as well as for shortcuts that might be suitable for your
conditions and pollinators. Appendix A includes information about covering
flowers for specific vegetables. You may be able to get away with less or have to
do more depending on your weather and your pollinators.

Overcoming Incompatibility Barriers

Up until now I've been talking about ordinary crosses that take well
spontaneously - compatible crosses. But often we want to do crosses that are
incompatible. There are two basic kinds of incompatibility. First, there is
incompatibility caused by a genetic incompatibility system. This kind is
designed to prevent inbreeding. In plants such as many of the brassicas, self-
pollination or pollination by plants with the same incompatibility gene(s) or
genotype doesn't take.

The second kind of incompatibility is normally associated with preventing

crossing between different species. In some cases there are few or no barriers
between making crosses between different species in the same genera or even
different genera in the same family. The biological barrier between the species in
nature may arise from the fact that they flower at different times, have different
pollinators, are located in different ecological zones, or are separated
geographically, not from physical inability to cross. In many cases, though,
crosses between species don't "go" spontaneously, even once a person has done
the task of pollinating.

The same kind of tricks are used to get around incompatibility barriers of both
types. Usually, when someone gets a rare cross to go, he or she reports the trick
used, but in most cases no one knows for sure whether it was the trick that
caused the suc

cess. In other words, doing a difficult cross is usually a matter of just trying
things.

Crosses will often go in one direction but not the other. When doing wide
crosses, it's especially important to try each species as the female and as the
male. Even in cases where both crosses can be obtained, the results will be
entirely different, since they set a hybrid genome down in the middle of two
completely different cytoplasms. Sometimes only one of the combinations is
viable. Where both are, there are often differences in appearance, growth rate,
vigor, or other properties. And there is no way of predicting ahead of time which
combination will be most useful or interesting, even if both are possible. You
just have to try them and see.

Different individual varieties and plants may differ in their ability to

participate in wide crosses. If I try a cross between a wild perennial mustard
plant and `Green Wave', for example, and it doesn't go, that doesn't necessarily
mean I have to start trying the more laborious forcing methods. The cross might
go readily between the wild plant and a different variety of mustard, even of the
same species.

When trying an incompatible cross, first try to do the cross normally, but with
two modifications. First, do the cross as early as possible after the female
becomes receptive, to give slowgrowing pollen as long as possible to grow
before the flower is dropped. Pollinate the flower in the bud stage in cases where
the stigma is receptive then.

Second, make sure to exclude all possible sources of pollen other than what
you provide. In many cases the incompatible pollen does germinate and start to
grow down the style, but its germination and/or growth is much slower than for
compatible pollen. Under normal circumstances, where compatible pollen is
present, the incompatible pollen is beaten out by the faster-growing competition.
If you have excluded all compatible pollen, the cross may go.

;`All compatible pollen" includes that of wild relatives, too. When you are
doing an ordinary cross, you often don't need to worry about contamination with
pollen from the field down the street. But if you are doing a difficult cross in
which you would be happy to get just one successful seed from your work, the
rare pollinations from fartraveling bees may be much more frequent than the
success rate for your cross.

In some cases where slow growth of incompatible pollen occurs, the flower
normally dries up and falls off before the pollen grows all the way down the
style and achieves fertilization. A number of tricks can help circumvent this
problem. The simplest one is to pollinate as early as the bud is receptive. In
many plants the stigma is receptive much earlier than the flower opens. By
pollinating in the bud stage, slowgrowing pollen has much longer to reach the
ovaries.

In addition, in some plants, such as the brassicas, the incompatibility

mechanism that prevents self-pollination develops late in the maturation of the
flower. There is a stage at which the stigma is receptive but not yet capable of
recognizing and excluding incompatible pollen.

Many other tricks depend on just giving the alien pollen more time to do its
job. Use

of hormones that promote fruit set may help, because part of what they do is to
slow down flower drop. Some breeders cut off the stigma and upper part of the
style of a flower and put the pollen directly on the cut stub. This shortens the
distance the pollen tubes have to grow, and in some cases the pollen also may
bypass exclusion mechanisms associated with the stigma. Just covering the
pollinated flower to keep it moist may allow the flower to stay on long enough
for a difficult cross to take.

Many incompatible crosses are much more likely to go at certain periods.

Selfincompatible crosses may be rare among fertilizations early in flowering, for
example, but common among the latest flowers. Incompatibility of both kinds is
sometimes circumvented by weather, either artificial or deliberate. A plant may
be willing to accept incompatible crosses late in its flowering period if no other
pods or fruits have been allowed to remain on it. Heat or cold shocks can
sometimes cause incompatible crosses to take. Selfincompatibility often depends
on climate or weather.

In many cases it appears that something in the sticky juices associated with the
stigmatic surface may be necessary for germination of the correct pollen or may
prevent germination of the wrong stuff. Another idea is that the right pollen
might have substances that are necessary for germination. Various tricks are
based on these ideas.

Sometimes breeders transfer the stigmatic "goo" from a plant that the pollen
would be compatible with to the stigma of the incompatible plant before
pollination. I wonder whether it might not be more effective to pollinate a
compatible plant, then shortly thereafter transfer the goo with the (hopefully)
germinating pollen in it to the incompatible plant. Washing the goo off the
incompatible plant also may be necessary.

What about pollinating a compatible plant with the desired pollen and letting it
germinate, then cutting the stigma and style off the desired female, slicing off the
stigma with the germinating pollen, and putting it directly on the stub? That way
the pollen would be stimulated to germinate and start growth by the compatible
stigma and not be excluded by any mechanisms, either positive or negative,
associated with the stigma of the incompatible plant.

A simpler related method is to mix live pollen from the desired male parent
with dead pollen from a compatible plant and use that to hand-pollinate. The
idea is that enzymes or proteins on the surface of the compatible pollen may be
necessary to set up a reaction in the stigmatic surface or break down the surface.
With a huge amount of compatible pollen, the reaction is general enough that
foreign pollen is no longer excluded.

You might be able to kill the pollen by heat-treating it, but heat also denatures
many proteins. I think the first thing I would try would be to collect the pollen
and le.t it age. Then I'd try some crosses with it to check the killing method.

Henry Wallbrunn, geneticist and orchid breeder, tells me that he often achieves
difficult orchid crosses by mixing live pollen. Suppose he wants a cross of a
variety-A female and a variety-B male, but that cross is difficult (or up till then
impossible) to ob tain. He chooses a third variety, variety C, that crosses with
variety A readily and whose hybrids with variety A he expects to be able to
distinguish from those between A and B. Then he mixes pollen from varieties B
and C and uses it to pollinate variety A.

Another possibility that I think might work in at least some cases is to slit open
the ovary of the living plant with a razor and push in some pollen. A recent
report describes fertilizing excised ovaries with pollen in vitro. Apparently
growth of a pollen tube down the style isn't always necessary. My suggestion is a
low-tech version.

Sometimes there is incompatibility between the developing embryo and the

endosperm of the seed. Incompatible pollen germinates and achieves fertilization
and an embryo results, but the endosperm - the seed's food supply - fails to
develop or isn't compatible with and can't be used by the embryo. The high-tech
solution to this problem is embryo rescue. The embryo is dissected out and
cultured sterilely in a tissue culture medium that provides nutrients and
substitutes for the endosperm. I wonder whether the same thing might be
possible in a lowertech way by transferring the hybrid embryo to a normal seed
whose embryo has been removed. I haven't tried it.

In many cases where embryo rescue is required to achieve a cross, the

reciprocal cross may be possible without it. The reciprocal cross has the same
embryonic genome, but it has a different endosperm.

So far the major "trick" I have used to get wide crosses to go is to look at the
weather forecast and do them during a period of ideal weather.

Collections are listed alphabetically by collection site code followed by site
name, address, and phone number. For information about how to use this list and
how to approach and obtain material from the USDA system, see Chapter 4.

CLOVER. Clover Collection. USDAARS, Department of Agronomy, University

of Kentucky, Lexington, KY 40506; (606) 257-5785.

COTTON. Cotton Collection. USDA-ARSSPA, Route 5, Box 805, College

Station, TX 77840; (409) 260-9209.

CR-BRW. National Clonal Germplasm Repository-Brownwood. USDAARS,

Pecan Research, Route 2, Box 133, Summerville, TX 77879; (915) 646-0593

CR-BYR. National Clonal Germplasm Repository-Byron. USDAARS,

Southeastern Fruit and Tree Nut Research Laboratory, PO. Box 87, Byron, GA
31008; (912) 956-5656.

CR-COR. National Clonal Germplasm Repository-Corvallis. USDAARS, 33447

Peoria Road, Corvallis, OR 97333; (503) 757-4448.

CR-DAV. National Clonal Germoplasm Repository-Davis. USDAARS,

Department of Pomology, University of California, Davis, CA 95616; (916) 752-
6504
CR-GEN. National Clonal Germplasm Repository-Geneva. USDAARS, New
York State Agricultural Experiment Station, Geneva, NY 14456-0462; (315)
787-2390.

CR-HIL. National Clonal Germplasm Repository-Hilo. USDAARS, c/o

Beaumont Agricultural Research Center, 461 W Lanikaula Street, Hilo, HI
96720; (808) 959-5833.

CR-MAY. National Clonal Germplasm Repository-Mayaguez. USDAARS,

Tropical Agricultural Research Station, P.O. Box 70, Mayaguez, PR 00709-
0070; (809) 831-3435.

CR-MIA. National Clonal Germplasm Repository-Miami. USDAARS,

Subtropical Horticultural Research Station, 13601 Old Cutler Road, Miami, FL
33158; (305) 238-9321.

CR-ORL. National Clonal Germplasm Repository-Orlando. USDAARS, Route

2, Box 375, Groveland, FL 32736; (904) 787-5078.

CR-RIV. National Clonal Germplasm Repository-Riverside. USDAARS, 1060

Pennsylvania Avenue, Riverside, CA 92507; (714) 787-4399.

DBMU. Database Management Unit. USDAARS-PSI-GSL-GRIN, Building

001, Room 130, BARCWest, Beltsville, MD 20705; (301) 344-3318.

FLAX. Flax Collection. USDAARS, Walster Hall, Room 206A, North Dakota
State University, Fargo, ND 58105; (701) 237-8155.

GD. National Plant Germplasm Quarantine Laboratory. USDAARS, 11601 Old

Pond Drive, Glenn Dale, MD 20769-9157; (301) 344-3003.

IR-1. Inter-Regional Potato Introduction Station. USDAARS, Peninsula

Experiment Station, Sturgeon Bay, WI 54235; (414) 743-5406.

NA. National Arboretum. USDAARS, 3501 New York Avenue NE, Washington,
DC 20002; (202) 475-4836.

NC-7. North Central Regional Plant Introduction Station. USDAARS, Iowa

State University Ames, IA 50011; (515) 292-6507.
NE-9. Northeastern Regional Plant Introduction Station. USDAARS, P.O. Box
462, New York State Agricultural Experiment Station, Geneva, NY 14456-0462;
(315) 787-2244.

NPMC. National Plant Materials Center. USDA-SCS, Building 509, BARC-

East, Beltsville, MD 20705; (301) 344-2175.

NSGC. National Small Grains Collection. USDAARS, Small Grains Germplasm

Research Facility, P.O. Box 307, Aberdeen, ID 83210; (208) 397-4162.

NSSL. National Seed Storage Laboratory. USDAARS, Colorado State

University, Fort Collins, CO 80523; (303) 484-0402.

PEANUT. Wild Peanut Collection. USDAARS, Plant Science Research

Laboratory, 1301 N. Western, Stillwater, OK 74075; (405) 624-4124.

PIO. Plant Introduction Office. USDA-ARSPSI-GSL, Building 001, Room 322,

BARCWest, Beltsville, MD 20705; (301) 344-3328.

S-9. Southern Regional Plant Introduction Station. USDAARS, 1109 Experiment

Street, Griffin, GA 30223-1797; (404) 228-7255.

SBMNL. Systematic Botany and Mycology Lab. USDAARS, Building 265,

BARC-East, Beltsville, MD 20705; (301) 344-2681.

SOY-N. Soybean Collection-North. USDAARS, University of Illinois, W-321

Turner Hall, 1102 S. Goodwin Avenue, Urbana, IL 61801; (217) 244-4346.

SOY-S. Soybean Collection-South. USDAARS, Delta Branch Experiment

Station, PO. Box 197, Stoneville, MS 38776; (601) 686-9311.

TOBAC. Tobacco Collection. USDAARS, Crops Research Laboratory, PO. Box

1555, Oxford, NC 27565-1555; (919) 693-5151.

W-6. Western Regional Plant Introduction Station. USDAARS, Room 59,

Johnson Hall, Washington State University, Pullman, WA 99164-6402; (509)
335-1502.

The National Plant Germplasm System holdings and data are now all in a
database called GRIN, the Germplasm Resources Information Network. You can
access GRIN through the Internet to obtain information or request material. Here
are the relevant addresses.

home page www.ars-grin.gov/npgs/

holdings www.ars-grin.gov/npgs/holdings.html

holdings by site www. ars-grin. gov/npgs/stats/site . summary. stats

search www.ars-grin.gov/npgs/searchgrin.html

accession queries www.ars-grin.gov/npgs/acc/acc-queries.html

taxonomy index www.ars-grin.gov/npgs/tax/index.html

taxonomy search www.ars-grin.gov/npgs/tax/taxgenform.html

germplasm request www.ars-grin.gov/npgs/orders.html

order form for germplasm www.ars-grin.gov/npgs/order.html

Plant Variety Protection Office home page

www.ams.usda.gov/science/pvp.htm

PVP search www.ars-grin.gov/cgi-bin/npgs/html/pvlist.p

The following is not a comprehensive list of worthy seed companies or
organizations. It is only a list of those referred to in this book. It emphasizes seed
companies in the Northwest, and those with special interests similar to mine. For
the best available list of seedsaving organizations, seed companies, and other
sources throughout the world, see Stephen Facciola's book Cornucopia II.

Abundant Life Seed Foundation. P.O. Box 772, Port Townsend, WA 98368;
(360) 385-5660; Fax (360) 385-7445. abundant@ olypen.com. Openpollinated
seed of garden and native plants of the North Pacific Rim.

Allan, Ken. 61 S. Bartless Street, Kingston, ON K7K 1X3, Canada. allan@adan.

kingston.net. Sweet potatoes, tall peas, organic openpollinated vegetable seed.

Borman, Dan. See Meristem Plants.

Bountiful Gardens. 18001 Shafer Ranch Road, Willits, CA 95490; (707) 459-
6410; Fax (707) 459-6410. Sole U.S. distributor for Chase Seeds of England.
Openpollinated.

Deep Diversity Seed. P.O. Box 15700, Santa Fe, NM 87501. Subsidiary of Seeds
of Change. Deep Diversity comprises the more exotic part of the Kapuler
germplasm collection. (The rest of the Kapuler collection is in the Seeds of
Change catalog.) Organic, openpollinated seed.

Deppe, Carol. See Web sites plantbreeding.net and specificfoods.com.

Drowns, Glenn. See Sand Hill Preservation Center.

Fava Project, The. Anyone interested in exploring, researching, breeding,
growing, or eating fava beans may join. Contact Dan Borman at Meristem
Plants.

Fedco Seeds. P.O. Box 520, Waterville, ME 04903-0520; (207) 873-7333. (No
telephone orders.) Seed for cold climates and short growing seasons.

Garden City Seeds. P.O. Box 204, Thorp, WA 98946; (509) 964-7016; Fax (800)
968-9210. www.gardencityseeds.com. Northernacclimated, openpollinated
seeds; perennial vegetables.

Irish Eyes, Inc. (Includes Ronniger's Potatoes, Kalmia Onions, and Greg
Anthony Seeds.) PO. Box 307, Thorp, WA 98946; (509) 964-7000; Fax (800)
964-9210. www.irish-eyes.com. Over 80 potato varieties; garlic; multiplier
onions and other alliums, seeds. Free catalog.

J. L. Hudson, Seedsman. Star Route 2 Box 337, LaHonda, CA 94020.

Openpollinated; many rare varieties.

Johnny's Selected Seeds. Foss Hill Road, Albion, Maine 04910-9731; (207) 437-
4301; Fax (800) 437-4290. www.johnnyseeds.com. Seeds for northern climates.

Kapuler, Alan. See Peace Seeds, Deep Diversity Seeds, and Seeds of Change.

Meristem Plants and Plant Research Supplies for Cool Climates. Box 254, Deer
Harbor, WA 98243. Openpollinated seeds; perennial vegetables.

Native Seeds/SEARCH. 526 N. 4th Avenue, Tucson, AZ 85705; (520) 622-

5561. Fax (520) 622-5591. Preservation of traditional Native American crops
and wild relatives of U.S. Southwest and northwestern Mexico. Openpollinated.

Nichols Garden Nursery: Herbs & Rare Seeds. 1190 North Pacific Highway,
Albany, OR 97321-4598; (541) 928-9280; Fax (541) 967-8406.
nichols@gardennurserycom. www.gardennursery.com. Seeds and plants. Many
rare varieties.

Northern Groves. Four Wands Farm, 23818 Henderson Road, Corvallis, OR

97333; (541) 929-7152. More than 50 hardy bamboos, including a number of
edible types.
Oregon Exotics Nursery. 1065 Messinger Road, Grants Pass, OR 97527; (541)
846-7578. Incredible collection of rare, recently collected perennial vegetables,
tubers, and trees, including hardy citrus.

Peace Seeds: A Planetary Gene Pool Resource and Service. Alan Kapuler's seed
company for a number of years. Peace Seeds was being acquired by Seeds of
Change just as I was writing the first edition of this book. Kapuler became
research director for Seeds of Change. The Kapuler germplasm collection is
available through two catalogs. The main Seeds of Change catalog covers the
varieties sold in larger amounts. The Deep Diversity catalog covers the rest, with
many offerings being very rare. See entries for Deep Diversity and Seeds of
Change.

Peaceful Valley Farm Supply: Organic Growers Source Book and Catalog. PO.
Box 2209, Grass Valley, CA 95945; (530) 272-4769; Fax (530) 272-4794.
www.groworganic.com Organic seeds. Organic growers supplies and equipment.

Perennial Vegetables. Seed companies that have especially good collections of

perennial vegetables include Deep Diversity Seed, Garden City Seeds, Meristem
Plants, Nichols Garden Nursery, Irish Eyes, Peters Seed & Research, Oregon
Exotics, Perennial Vegetable Seed Co., Seeds of Change, and Wild Garden
Seeds.

Perennial Vegetable Seed Co. PO. Box 608, Belchertown, MA 01007. Perennial
vegetables.

Peters, Tim. Tim Peters was working for Territorial Seed Co. at the time I wrote
the first edition of this book. He subsequently out-migrated and started his own
seed company, Peters Seed and Research.

Peters Seed and Research. 407 Maranatha Lane, Myrtle Creek, OR 97457; (541)
874-2615. The best collection of perennial grains I've ever seen, as well as many
experimental and original varieties, and an excellent collection of Northwest
heirlooms. Openpollinated organic seed.

Pinetree Garden Seeds. Box 300, New Gloucester, ME 04260; (207) 926-3400;
Fax (888) 527-3337. www.superseeds.com Focuses on gourmet quality and
flavor for the home garden.
plantbreeding.net. Web site for farm-and garden-based plant breeding and
seedsaving.

Plant breeding links. Information on availability of my own varieties and

publications. Additional topics of interest for those who have read Breed Your
Own Vegetable Varieties. Online bookstore for plant breeders and seed savers.
www.plantbreeding.net.

Redwood City Seed Company. PO. Box 361, Redwood City, CA 94064; (650)
325-7333. www.ecoseeds.com. Openpollinated. Many rare Native American and
other traditional varieties.

Richters. 357 Highway 47, Goodwood, ON LOC 1AO, Canada; (905) 640-6677;
Fax (905) 640-6641. catalog@richters.com inquery@ richters.com. Many rare
herbs and vegetables.

Ronniger's Potatoes. See Irish Eyes, Inc.

Salt Spring Seeds. PO. Box 444, Ganges, Salt Spring Island, BC V8K 2W1,
Canada; (250) 537-5269. Certified organic, openpollinated seed. Huge selection
of beans and grains.

Sand Hill Preservation Center. Heirloom seeds and poultry. 1878 230th Street,
Calamus, IA 52729; (319) 246-2299. (No calls Sunday or Monday)
Openpollinated vegetable seeds, especially vine crops and tomatoes. Also, many
rare and traditional breeds of poultry.

Seeds Blum. This company, founded by Jan Blum, was one of the earliest
promoters, rediscoverers, and purveyors of heirloom vegetables. Jan Blum has
retired from this phase of the seed business, and sent out her last Seeds Blum
catalog in 1998.

Seeds of Change. P.O. Box 15700, Santa Fe, NM 87506-5700; (888) 762-7333.
www.seedsofchange.com. Certified organic openpollinated vegetable, flower,
and herb seed.

Seeds of Diversity Canada. PO Box 36 Station Q, Toronto, ON M4T 2LT,

Canada; (905) 623-0353. www.seeds.ca/ This is the major seed savers exchange
in Canada. Membership is $20-$35. Includes three publications a year and
provides access to more than a thousand varieties.

Seed Savers Exchange. 3076 North Winn Road, Decorah, IA 52101-7776; (319)
382-5990; Fax 319-382-5872. Membership $35-$45. Includes three huge
publications per year, the Winter Yearbook being the one that lists all the
thousands of varieties offered by members.

Seed Savers Exchange in Canada. See Seeds of Diversity Canada.

Southern Exposure Seed Exchange. PO. Box 460, Mineral, VA 23117; (540)
894-9481; Fax (540) 894-9481. www.southernexposure. corn. Huge collection of
traditional openpollinated varieties adapted for the Southeast.

specificfoods.com. This website will sell completely labeled, 100 percent gluten-
free, non-GM foods made from specific named plant varieties. The initial focus
will be on corn meal and corn products made from traditional varieties of corn.

Territorial Seed Company. PO. Box 157, Cottage Grove, OR 97424-0061; (541)
942-9547; Fax 541-942-9881. www.territorialseed.com. Specializes in varieties
for the Maritime Northwest (west of the Cascades). Distributes one catalog to
the Maritime Northwest and another nationally.

Vegetable Garden Research Exchange, The. This publication, run by Ken Allan,
provides an outlet for exchanging information among home-gardener researchers
and plant explorers and breeders. Receive and exchange information on cultivar
trials, culture methods comparisons, plant breeding, and much else. It's
scientifically rigorous, and is worth much more than the cost of membership. For
information, write Ken Allan, The Vegetable Garden Research Exchange, 61
South Bartlett Street, Kingston, ON K7K 1X3, Canada.

Wild Garden Seed. Shoulder to Shoulder Farm, PO. Box 1509, Philomath, OR
97370; (541) 929-4068. Unusual salad greens, many farm-bred varieties, and
naturalized, native, and semi-domesticated wild edible plants of the Maritime
Northwest. Many segregating mixtures from which farm-adapted new varieties
can be easily selected. Semi-wholesale. Minimum order $50.

Abundant Life Seed Foundation Screens of various sizes for cleaning seed,
cloth bags for seed, a hand-operated seed thresher, and selfsealing seed
envelopes.

Bountiful Gardens

A Seed Saver's Kit, which includes silica gel for drying seed, data forms, and
the book Growing to Seed; silica gel refills.

Meristem Plants and Plant Research Supplies for Cool Climates

One of the main objectives of this new company is to provide for the needs of
amateur plant breeders and researchers. It carries an extensive line of seed
saving, plant breeding, and plant research supplies, forceps of various kinds, ear
shoot and pollen bags, and dissecting and standard microscopes.

Southern Exposure Seed Exchange Silica gel, ear shoot bags, drawstring muslin
bags (for excluding pollinators while allowing normal flower development),
selfseal seed packets, light and heavy Zip-Loc bags, heat-seal pouches and
equipment, parafilm, seed saver vials, corn sheller, various kinds of tapes,
markers, and labels.

See Appendix D for addresses.

This chart tells you how many plants you classes of the given probabilities.
(Adapted need to be 95 percent or 99 percent sure of from C. North, Plant
Breeding and Genetics in obtaining at least one of the desired class, for
Horticulture.)

accession The material in a germplasm collection that results from a single
collecting event - such as a collector buying beans from one basket in a
market. The basket may contain just one variety or a mixture of varieties. A
sample of beans from a basket in the next market would get a different
accession number, even if they looked identical. They may or may not be the
same variety. So an accession may contain more than one variety, and a variety
may be represented in a collection by more than one accession number.

alleles Different variations of one gene at one locus.

allopolyploid See ploidy.

amphidiploid A polyploid that originates through the combination of two

different genomes such that it behaves like a new diploid species.

anther See male parts of the flower.

anthesis The opening of a flower.

apornixis The process of producing seed without any union of gametes. Pollen
may or may not be necessary to stimulate the process, but it doesn't contribute
genetically. The seed produced may or may not be genetically identical to the
mother, depending on the kind of apomixis involved.

asexual Without sex.

autopolyploid See ploidy.

backcross To cross one of the progeny of a cross back to one of the parents. A
backcross is such a cross.

bolt To produce flower stalks.

bud pollination A hand-pollination performed when the flower is in the bud

stage. Often useful in overcoming incompatibility barriers.
bud sport A mutant branch or section of a plant that arises when a mutation has
occurred in the somatic tissue that gave rise to the bud.

calyx The ring of sepals on a flower.

chimera An individual whose cells are of two or more genotypes.

chromosome The structural units containing the genes.

cleistogamy The act of releasing pollen and selffertilizing before the flower
opens.

clone To create genetically identical individuals. Clones are the individuals so

created. Vegetative propagation of plants produces clones.

codominant See dominant, codominant, and recessive.

colchicine A cell-division poison used to double chromosome numbers.

corolla The ring of petals on a flower.

cotyledons The first leaves of the embryonic plant within the seed that are used
as a food supply for the germinating embryo.

cross A fertilization involving two different parents, or to make or cause such a

fertilization. Opposite of self-pollination.

crossing over A breaking and reunion that interchanges corresponding parts of

homologous chromosomes. It allows genes on the same chromosome to
recombine unless they are very close together.

cull To remove inferior or undesirable individuals. Same as to rogue (remove the

rogues).

cultivar A cultivated variety. Professionals use the word cultivar rather than
variety. Common use favors variety. In this book I usually use variety, but I
also use the two words interchangeably.

cytoplasm The rest of the contents of a cell other than the nucleus.
cytoplasmic inheritance Inheritance associated with the cytoplasm instead of
genes on chromosomes (and in the nucleus).

day-neutral See photoperiod.

dehiscence of anthers Pollen release.

determinate and indeterminate growth patterns Indeterminate plants have shoots

that remain vegetative at the apex; they produce flowers only subterminally.
Determinate plants have shoots that grow vegetatively for a while but then end
in flowers. This restricts overall size and often also produces a bushier form.
Bush beans and bush tomatoes are determinate. Pole beans and vine tomatoes
are indeterminate.

detrimental Used to describe a gene, allele, or mutation that causes an

impairment or disadvantage to the individual that carries it. Often used as a
noun. Example: Recessive lethals and detrimentals are common in populations
of plants, especially outbreeders.

dioecious Varieties or species that have male and female flowers on separate
plants.

diploid See ploidy.

DNA Deoxyribonucleic acid, the genetic material.

dominant, codominant, and recessive If plants of genotype Aa look the same as

those of genotype AA, we say the A allele is dominant to the a allele or that the
a allele is recessive to the A one. If, however, plants of genotype Aa look
different from both AA's and aa's, we say that the A and a alleles are
codominant. Strictly speaking, the terms dominant, codominant, and recessive
apply only to relationships between alleles of a gene, not to traits. That is, it's
correct to say that T is dominant to t but not to say that tall is dominant to
short. Tall might not be dominant to short in some other cross where different
genes are involved. The use of the words to describe traits promotes sloppy
thinking but is such a useful shorthand that it's almost impossible to avoid. I
usually try to avoid it, but give in occasionally.

edge effects Plants at the ends of rows or on the edge of a planting are often
 bigger, bushier, earlier, and/or more productive than the rest because they have
more space, get more light, or have less competition for water or nutrients - in
other words, because their environment is better. Edge effects often have to be
considered when you're trying to evaluate the performance of individual plants
or groups of plants.

egg The female gamete that combines with a nucleus from a pollen grain to
make a zygote.

emasculate a flower To remove the male parts of a flower before pollen is shed.

embryo The part of the seed that grows into the new plant after the seed
germinates.

endosperm A storage tissue in the seed that is used up by the embryo in the
process of germination. Endosperm characters are traits associated with the
endosperm of seeds.

expressivity Not all genes or genotypes express themselves consistently. If a

single genotype has a variable degree of expression within a given single
environment, it's said to have variable expressivity.

F1, F2, F3, etc. The progeny of the specified generation after a cross. When two
different varieties are crossed, the seed they produce is the F, generation. When
it is planted, it grows into F1 plants. A cross of two F1 plants produces F2
seed, and so on.

factor A one-factor cross is a cross in which the two parents differ by one pair of
alleles at one locus. (Example: AA X aa.) The number of factors in a cross is
the number of pairs of alternatives. The cross AABB X aabb is a two-factor
cross. The term factor dates from before people knew that the factors are really
genes. Twogene cross would be ambiguous (Would AABB X aabb be a
twogene cross or a four-gene one?), so factor is still used in this context.

female parts of the flower The pistil is the whole female unit of the flower. The
eggs, ovary, style, and stigma are parts of the pistil. The stigma is at the end of
the pistil and is the special surface on which the pollen lands and germinates.
The style is the structure that supports the stigma. The pollen tube grows down
 the style to reach the ovary. The ovary is at the base of the style and contains
the eggs.

fertile X fertile cross A cross made without emasculation of the flower, generally
because the flowers are of a type that makes emasculation too diffi cult. In a
successful fertile X fertile cross, only some of the offspring are the cross; the
rest are the unavoidable self-pollinations.

fix To fix a gene, allele, or trait in a population or variety is to achieve

homozygosity for it in all individuals in the population so that the population is
then pure breeding for it.

frankenfood A derogatory reference to food derived from genetically engineered

plants or animals. It is a contraction of Frankenstein food, recalling the
artificially constructed monster in the novel Frankenstein by Mary W. Shelley.
Frankenstein killed his creator.

gamete The haploid sex cells that result from meiotic (sexual) cell division -
pollen and eggs.

gene The basic unit of inheritance. A gene is the section of DNA that codes for
one protein or subunit of a protein.

genetic engineering The creating or altering of organisms or varieties of

organisms by direct physical transfer of DNA.

genome All the genetic material in a cell or an organism. See also germplasm.

genomic formula See ploidy

genotype and phenotype The genotype is the exact genes in an organism (TT, Tt,
or tt, for example). The phenotype is what the organism looks like (tall or
short). You identify the genotype by knowing the genetics and parentage
involved. You identify the phenotype just by looking at the plant. The
phenotype depends on both the genotype and the environment. If I chop the
top off a plant of genotype TT, its genotype is still TT, but its phenotype
changes from tall to short.

germplasm Like genome, the term germplasm refers to all the genetic material,
 but the context is usually broader and the emotional value is different. We refer
to the genome of individuals or species, but the germplasm of species or
genera, or the germplasm in a country or the world. The germplasm of North
America would be all the genes in all the species, wild and domestic, that are
found or grown in North America. The two terms overlap in formal definition
but rarely in practice. Genome is a technical, unemotional term. Germplasm,
with its associations of germ/core/life and plasm/life/blood, has mystical and
spiritual overtones. We use the word germplasm when we care about it. It is
saving germplasm from extinction that biologists and many others are so
passionately concerned about.

gibberellin A plant hormone involved in controlling plant growth, differentiation,

flowering, fruit set, and other processes.

GM, GMO, GM foods Genetically modified. Genetically modified organism.

Genetically modified foods. All refer to genetically engineered organisms or
products derived from them.

haploid See ploidy

heirloom variety A variety that has been around for a while, that has been
handed down from gardener to gardener or generation to generation.

hermaphrodite A flower that has both male and female parts. Same as perfect
flower.

heterosis Increased vigor of a hybrid as compared with the two parental

varieties. Same as hybrid vigor

heterozygous and homozygous An individual that has two different alleles at the
locus under consideration is heterozygous at that locus. (Example: Aa.) Such
an individual is called a heterozygote. An individual with identical alleles at
the locus (AA or aa) is homozygous and is a homozygote.

hilum The scar or spot on the seed that marks the area where it was attached to
the ovary.

homologous chromosomes In diploids, the two chromosomes that have the same
or similar genes on them and that pair during meiosis.
homozygous See heterozygous and homozygous. hybrid vigor Increased vigor of
a hybrid as compared with the two parental varieties. Same as heterosis.

inbreed To breed two closely related individuals. In the extreme case, to self-
pollinate.

inbreeding depression A loss of vigor in a variety or population associated with

inbreeding. Especially typical of plants that are natural outbreeders. (See
Chapter 9.)

incompatibility mechanism A genetically determined physiological mechanism

that prevents or restricts selffertilization or fertilization by plants or pollen with
related incompatibility genotypes.

incompatible Two plants are incompatible when they are unable to fertilize one
another. A plant is self-incompatible when it is unable to fertilize itself.

indeterminate See determinate and indeterminate growth patterns.

inversion A chromosomal mutation in which one block of genes on the

chromosome has been inverted and therefore carries the genes in reverse order.

isolation distance The distance apart two different varieties of a single species
must be to prevent their cross-pollinating (and thus contaminating) each other.

heel See standard, wings, and keel.

lethal Used to describe a gene, allele, or mutation that kills the individual
carrying it in some specific context. Often used as a noun. Example: Recessive
lethals are quickly eliminated from populations that are strongly inbreeding.

linkage If two different genes are on the same chromosome, the genes are said to
be linked.

locus A specific position on a chromosome or chromosome pair that is occupied

by a specific gene.

long-day plants See photoperiod.

male parts of the flower The stamen is the basic male part of the flower. Its parts
are the anther, stalk/filament, and pollen. The anther is the part at the top that
produces the pollen. The stalk or filament is the rest of the stamen, the part that
supports the anther.

mass selection Selection based on working with a population as a whole as

opposed to working with various families or other groupings separately

meiosis and mitosis Meiosis is sexual cell division. It results in gametes. It

occurs only within the pollen-or egg-producing tissues within flowers. Mitosis
is the ordinary cell division that occurs whenever a plant grows or reproduces
vegetatively. Mitosis results in more ordinary cells.

mitosis See meiosis and mitosis.

modifier A gene that influences or alters the effect of some other gene with a
larger effect.

monoecious Used to describe varieties or species that have separate male and
female flowers on each plant.

mutation and mutant A mutation is an alteration in a gene or chromosome. A

mutant is the changed thing (DNA, protein, phenotype, gene, chromosome, or
individual) that has the mutation in it.

nucleus The organelle in the cell that contains the chromosomes.

off types Individuals of a population or variety whose characteristics are

atypical or undesirable. Also called rogues.

open-pollinated variety A stable variety, a variety that breeds true from seed.
(Normally maintained by allowing natural pollination under field conditions,
but hand-pollination is sometimes used to achieve the equivalent.) Not a
hybrid.

outbreed To cross-pollinate instead of to self-pollinate. Alternatively, to cross to

a distant relative or an unrelated plant instead of self-pollinating or crossing to
a close relative. Opposite of inbreed.
outbreeder A variety or species that naturally cross-pollinates more often than it
self-pollinates.

ovary See female parts of the flower.

penetrance The degree to which a gene or genotype expresses itself. If some

individuals of a genotype express a characteristic and others do not, even
within the same environment, we say that the gene or genotype has incomplete
penetrance.

perfect flower A flower that has both male and female parts. Same as
hermaphrodite.

phenotype See genotype and phenotype.

photoperiod Length of day. A plant that has a photoperiod requirement for

flowering is one whose flowering is at least partly under the control of the day
length. Long-day plants begin flowering only after the day length exceeds a
specific minimum. Short-day plants begin flowering only after the day length
falls below a given level. In northern temperate zones, spring-flowering plants
are often long-day plants and fall-flowering plants are often short-day plants.
Day-neutral plants are plants whose flowering is not dependent on
photoperiod.

pistil See female parts of the flower.

pleiotropy A gene's influence on traits other than the primary one that it is
defined in terms of.

ploidy The number of haploid genomes in a cell, plant, variety, or species.

Gametes of most plants are haploid. This means they have one copy of each
kind of chromosome. Other than gameteproducing tissue, most plants are
diploid - that is, they have two copies of each kind of chromosome. Polyploid
plants, varieties, or species have more than two complete haploid sets of
chromosomes. They may be triploid, tetraploid, hexaploid, etc.

In autopolyploids the chromosomes are all derived from a single species. In

allopolyploids the chromosomes are derived from two or more different
species. Genomic formulas are used to describe the origin of polyploids. If
 three diploid species have genomic formulas of AA, BB, and CC, respectively,
then AABB would be a tetraploid that arose from a cross of the first two
species followed by chromosome doubling. In other words, it would contain
all the chromosomes typical of both species (as identified microscopically).

pollen The male gametes of plants.

polyploid See ploidy

popbean Any legume variety whose seed both pops and becomes fully cooked
and edible during the quick exposure to heat during popping. The popbeans
discussed in this book are all chickpeas, Cicer arietinum.

protandry A situation in which pollen is released before the stigma matures and
becomes receptive.

protogyny A situation in which the stigma matures and becomes receptive before
the pollen is released.

pure breeding variety A variety that is homozygous for everything relevant. This
means that all members of the population or variety have the same genotype
and give rise to offspring with the same genotype. Same as true breeding.

recessive See dominant, codominant, and recessive.

reciprocal crosses Reciprocal crosses involve the same parents, but with different
sexes. If we cross a female from a tall variety with a male from a short variety,
the reciprocal is a cross of a male from the tall variety with a female from the
short variety

recombinant and recombination In a cross of two parent varieties, some classes

of F2 offspring resemble the parents and some do not. Those that resemble the
parents are called parental classes. Those that do not are called recombinant
classes, and the individuals in them are recombinants. They have different
combinations of the characteristics of the parents. In the cross tall green X
short yellow, F2 progeny that are tall green or short yellow are parental types.
Those that are tall yellow or short green are recombinants. The process of
producing these new types is called recombination.
recurrent bachcrossing A breeding approach in which progeny of a cross are
backcrossed to one of the parent varieties, then some of the resulting progeny
are backcrossed again to that same parent variety, and so forth. The parent
variety that is crossed to the offspring in each generation is called the recurrent
parent/variety. Recurrent backcrossing is used to transfer one or a few genes to
a different variety

scape A flower stalk.

selection The process of choosing the appropriate individuals to use in creating

or maintaining a variety. To select is to choose which germplasm to perpetuate
or continue with.

self A cross involving only one parent; a self-pollination. To self is to self-

pollinate. Opposite of cross.

short-day plants See photoperiod.

somatic Having to do with the body. Opposite of sexual. Somatic cells include
all the cells in a plant except those involved in producing gametes.

species A population of organisms capable of interbreeding in nature.

(Interbreeding means producing the normal number of fully fertile offspring.)

sport Old-fashioned word meaning mutant. A mutant is a sport with a modern

education.

stable variety A variety that reliably produces similar plants from seed to seed
and year to year; a pure breeding variety. To stabilize a variety is to do
breeding work with it until it becomes a stable variety

stamen See male parts of the flower.

standard variety or method A well-known variety or method that is included in a

trial or experiment as a standard for comparison, that is, a control.

standard, wings, and keel The three distinctive kinds of petals of a legume
flower. The standard is the outermost petal that spreads out to either side of the
open flower. The wings are the middle layer
of petals between the standard and the keel. The keel is the inner petal that
encloses the pistil and stamens.

stigma See female parts of the flower.

style See female parts of the flower.

subspecies Any subgroup within a species that seems distinctive and meaningful
to whoever is talking about it. This word has no official biological meaning but
is too useful to do without.

terminator technology A combination of genes that may be engineered into a

plant variety for the purpose of physiologically preventing their reproduction
by seed.

tetraploidy See ploidy

translocation A chromosomal mutation in which blocks of genes from two

different chromosomes have exchanged positions.

triploid See ploidy

USDA-ARS U.S. Department of AgricultureAgricultural Research Service.

variety A population of plants that are predictable and defined with respect to
characteristics that matter to the grower. Open-pollinated varieties breed true;
hybrid varieties do not.

vector A pollen vector is the means by which pollen of a given species is

normally dispersed. Examples: wind, water, insects, hummingbirds, gravity

vernalization An exposure to cold required by some plants before they will

flower.

wide cross Ordinary crosses are between individuals of the same species. Wide
crosses are between more distant relatives - such as members of different
species or even different genera.

wings See standard, wings, and keel.

zygote A diploid cell created when the pollen and egg fuse. It divides and grows
to give rise to the embryo, which (after seed germination) develops into a
whole new plant.

Some of the material listed here is for a lay audience. Much of it, however, is
intended for professional plant breeders or plant breeding students. If you want
to delve into the more advanced literature, the trick is not to let yourself be
intimidated. It helps to skim as many different books as possible initially. Some
authors will explain certain words and concepts, and others will explain others.

Many of the more technical books are out of print and/or very expensive, but
in the United States they can be obtained through interlibrary loan.

Allan, Ken. The Vegetable Garden Research Exchange. (See Appendix D.)

Anderson, Luke. Genetic Engineering, Food, and Our Environment. White River
Junction, Vermont: Chelsea Green Publishing Co., 1999. Very readable short
book that provides thorough coverage of all aspects of the genetic engineering
issue.

Ashworth, Suzanne. Seed to Seed: Saving Techniques for the Vegetable

Gardener. Decorah, Iowa: Seed Saver Publications, 1991. Most complete
single source of information on seed saving for all common and many rare
vegetables.

Bassett, Mark J., ed. Breeding Vegetable Crops. Westport, Connecticut: Avi,
1986. Fourteen separate chapters covering fourteen vegetables: asparagus,
beans, cabbage, carrots, corn (sweet), cucumbers, lettuce, onions, peas,
peppers, squash, sweet potatoes, tomatoes, and watermelons. Each chapter
includes a section on the general botany of the vegetable, floral botany and
controlled pollination (doing or avoiding crosses), major breeding
achievements of the past, an overview of current breeding goals, genetics, gene
lists, and an extensive bibliography of the current professional literature. This
book assumes an extensive background in genetics and plant breeding, but
much of the book is useful to amateurs, too.

Bates, David M., Richard W Robinson, and Charles Jeffrey, eds. Biology and
Utilization of the Cucurbitaceae. Ithaca, New York: Cornell University Press,
 1990.

Blackmon, W J., and B. D. Reynolds. "The Crop Potential of Apios americana -

Preliminary Evaluations." HortScience 21, no. 6 (1986): 1334-1336.

Burton, W. G. The Potato. 3d ed. New York: Halsted, 1989. Professional.

History, growing, breeding, diseases, storing.

Darlington, C. D., and A. P. Wylie. Chromosome Atlas of Flowering Plants.

London: Allen & Unwin, 1955.

Deppe, Carol. "Parching Corn." National Gardening Magazine May/June 1997.

Describes my work on rediscovering and using true parching corn.

See my web site wwwplantbreeding.net for a current list of my publications.

Duke, J. A., S. J. Hurst, and E. E. Terrell. Economic Plants and Their Ecological
Distribution. 1975. A 16-page table of one thousand important plants and their
chromosome numbers, life-styles, ecological characteristics, and requirements.

Duke, James A. Handbook of Legumes of World Economic Importance. New

York: Plenum, 1981. Invaluable for anyone interested in working with
established legume crops, in developing new ones, or in preserving legume
biodiversity.

Facciola, Stephen. Cornucopia II: A Source Book of Edible Plants. Vista,

California: Kampong Publications, 1998. An amazing compendium of edible
plants, varieties, and sources. Encyclopedic 700 pages of small print, large
format.

Fehr, Walter R., and Henry H. Hadley, eds. Hybridization of Crop Plants.
Madison Wisconsin: American Society of Agronomy and Crop Science
Society of America, 1980. Revised in 1982. Instruction on exactly how to
make crosses. Many pictures and diagrams. No background in genetics or plant
breeding is required. Crops covered are bean, broad bean, chickpea, cowpea,
crambe, lentil, maize/corn, pea, peanut, potato, rapeseed, mustard, soybean,
sugar beet, sunflower, sweet potato, and thirty or so field and forage crops.
Doesn't cover genetics or general plant breeding.
Frankel, R., and E. Galun. Pollination Mechanisms, Reproduction, and Plant
Breeding. Monographs on Theoretical and Applied Genetics 2. New York:
Springer-Verlag, 1977.

George, Raymond A. T. Vegetable Seed Production. 2nd ed. New York: CABI
Publishing, CAB International, 1999. For the professional seed

producer, but not too technical for amateurs. Researchers frequently need to deal
with small lots of experimental seed, so it includes smallscale as well as large-
scale methods. Covers most major and many minor crops. Includes information
on pollination mechanisms and isolation requirements of the plants and on
harvesting, threshing, and storing the seed.

Hawkes, J. G. The Potato: Evolution, Biodiversity and Genetic Resources.

Washington, D.C.: Smithsonian Institute Press, 1990.

Janick, Jules, and James E. Simon, eds. Advances in New Crops. (Proceedings
of the First National Symposium on New Crops: Research, Development,
Economics, held in Indianapolis, 1988.) Portland, Oregon: Timber, 1990.

Jeavons, John. How to Grow More Vegetables Than You Ever Thought Possible
on Less Land Than You Can Imagine. 4th ed. Berkeley, California: Ten Speed
Press, 1982. Revised frequently. The classic on biodynamic gardening. Raised-
bed, biodynamic growing methods for fifty-nine vegetables, twenty-seven
grains or field crops, thirty cover or fodder crops, and miscellaneous others.
Gives planting distances, plant requirements, vegetable/grain/fodder yields,
and seed yields. The only published source on seed yields for some crops,
especially on seed yields in the context of backyard rather than commercial
methods.

Jones, H. A., and L. K. Mann. Onions and Their Allies. London: Leonard Hill,
1963. For both professionals and growers. Still the classic on the alliums and
one of the few sources of breeding information for species other than Allium
cepa.

Kalloo, Dr. Vegetable Breeding, Vol. 1. Boca Raton, Florida: CRC Press, 1988.
First of a series of three volumes by Dr. Kalloo, professor and head of the
Department of Vegetable Crops, Haryana Agricultural University, Hisar, India.
 This first volume covers the basic techniques of vegetable breeding, including
methods for hybridizing.

Vegetable Breeding, Vol. II. Boca Raton, Florida: CRC Press, 1988. This volume
covers genetics of and breeding for resistance to disease, insects, and
nematodes, and tolerance to environmental stress. Chapter 2, pages 95-131,
"Insect Resistance in Vegetable Crops," should be of special interest to anyone
wanting to develop insectresistant varieties.

Vegetable Breeding, Vol. III. Boca Raton, Florida: CRC Press, 1988. Chapter 2
of this volume (pages 41-59) is "Breeding for Quality and Processing
Attributes in Vegetable Crops."

Kloppenburg, Jack R., Jr. Seeds and Sovereignty: The Use and Control of Plant
Genetic Resources. Durham, North Carolina: Duke University Press in
cooperation with the American Association for the Advancement of Science,
1988. Separate chapters by various authors representing all possible points of
view.

Lovejoy, Ann. "Importing Plants." Horticulture, April 1992, pp. 16-20.

National Research Council. Lost Crops of the Incas: Little-Known Plants of the
Andes with Promise for Worldwide Cultivation. Washington, D.C.: National
Academy Press, 1989.

National Research Council. Lost Crops of Africa: Volume I; Grains. Washington,

D.C.: National Academy Press, 1996.

Nieuwhof, M. Cole Crops: Botany, Cultivation, and Utilization. London:

Macmillan, 1979. An introduction to genetics for beginning horticultural
students.

North, C. Plant Breeding and Genetics in Horticulture. London: Macmillan,

1979. An introduction to genetics for beginning horticulture students.

Robinson, Raoul A. Return to Resistance: Breeding Crops to Reduce Pesticide

Dependence. Davis, California: agAccess, 1996. Documents the efficacy of
breeding for disease resistance via mass selection. Encourages farm-based
plant breeding.
Ryder, Edward J. Leafy Salad Vegetables. Westport, Connecticut: Avi, 1979.

Shapiro, Howard-Yana, and John Harrisson. Gardening for the Future of the
Earth. New York: Bantam Books, 2000. Sustainable agriculture one garden at a
time. A tale told via visits with Deppe, Jeavons, Kapuler, Mollison, and others.
Wonderful color photography.

Shinohara, Suteki. Vegetable Seed Production Technology of Japan Elucidated

with Respective Variety Development Histories, Particulars. Vol. 1. 4-7-7
Nishiooi, Shinagawa-ku, Tokyo: Shinohara's Authorized Agricultural
Consulting Engineering Office, 1984. Translation of a 1978 book by Seibundo
Shinkosha Ltd. Originally intended for Japanese seed growers. The only
source of breeding information for certain vegetables.

Simmonds, N. W, ed. Evolution of Crop Plants. New York: Longman, 1976.

Sprague, G. F, and J. W Dudley, eds. Corn and Corn Improvement. 2d ed.

Madison, Wisconsin: American Society of Agronomy, Crop Science Society of
America, and Soil Science Society of America, 1988.

Tsunoda, S., K. Hinata, and C. Gomez-Campo. Brassica Crops and Wild Allies:
Biology and Breeding. Tokyo: Scientific Societies Press, 1980.

United States Department of Agriculture. 1992 Yearbook of Agriculture: New

Crops, New Uses, New Markets; Industrial and Commercial Products from
U.S. Agriculture. Office of Publishing and Visual Communication, USDA.

Wagoner, Peggy. "Perennial Grain Development: Past Efforts and Potential for
the Future." In Critical Reviews in Plant Sciences, Boca Raton, Florida: CRC
Press, 1990.

Whealy, Kent, ed. The Garden Seed Inventory, 5th ed. Decorah, Iowa: Seed
Saver Publications, 1999. Revised frequently. A compilation of names,
descriptions, and sources for all nonhybrid vegetable varieties available
commercially in the United States and Canada. As much fun as two hundred
seed catalogs, which is what it repre

J J~ Winter Yearbook. For a lay audience. Published annually by the Seed Savers
 Exchange; comes with membership. Not available through bookstores or
libraries. Gives you access to thousands of vegetable varieties maintained by
members that are not available commercially.

Wilson, Gilbert. Buffalo Bird Woman's Garden. St. Paul, Minnesota: Minnesota
Historical Society Press, 1987.

Wittwer, Sylvan, Yu Youtai, Sun Han, and Wang Lianzheng. Feeding a Billion:
Frontiers of Chinese Agriculture. East Lansing: Michigan State University
Press, 1987. Mind expanding. A tremendous source of ideas and perspective.

Yamaguchi, Mas. World Vegetables: Principles, Production and Nutritive Values.

Westport, Connecticut: Avi, 1983.

Zeven, A. C., and P M. Zhukovsky. Dictionary of Cultivated Plants and Their

Centres of Diversity: Excluding Ornamentals, Forest Trees and Lower Plants.
2d ed. Wageningen, Netherlands: Centre for Agricultural Publishing and
Documentation, 1982. Gives scientific names, chromosome numbers, centers
of diversity, and a few words about use for thousands of species.

Sustainable living has many facets. Chelsea Green's celebration of the
sustainable arts has led us to publish trend-setting books about organic
gardening, solar electricity and renewable energy, innovative building
techniques, regenerative forestry, local and bioregional democracy, and whole
foods. The company's published works, while intensely practical, are also
entertaining and inspirational, demonstrating that an ecological approach to life
is consistent with producing beautiful, eloquent, and useful books, videos, and
audio cassettes.

For more information about Chelsea Green, or to request a free catalog, call
toll-free (800) 639-4099, or write to us at P.O. Box 428, White River Junction,
Vermont 05001. Visit our Web site at www.chelseagreen.com.

Chelsea Green's titles include: