This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Molecular Phylogenetics and Evolution 65 (2012) 102–115
Contents lists available at SciVerse ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Molecular phylogeny and dating of Asteliaceae (Asparagales): Astelia s.l.
evolution provides insight into the Oligocene history of New Zealand
Joanne L. Birch a,⇑, Sterling C. Keeley a, Clifford W. Morden a,b
a
b
University of Hawai‘i at Mānoa, Department of Botany, 3190 Maile Way, Honolulu, HI 96822, United States
University of Hawai‘i at Mānoa, Pacific Co-Operative Studies Unit, 3190 Maile Way, Honolulu, HI 96822, United States
a r t i c l e
i n f o
Article history:
Received 4 January 2012
Revised 21 May 2012
Accepted 25 May 2012
Available online 1 June 2012
Keywords:
Astelia
Asteliaceae
Collospermum
Gondwana
Long-distance dispersal
Molecular dating
a b s t r a c t
Asteliaceae (4 genera, 36 species) are found on both continents and island archipelagos in the southern
hemisphere and across the Pacific. The circumscription of Asteliaceae and intrageneric relationships are
poorly understood. We generated a phylogeny including all genera and 99% of the species using DNA
sequence data from chloroplast (trnL, psbA–trnH, rps16, and petL–psbE) and nuclear (NIA-i3) regions.
Relaxed clock methods were applied to infer the age of the family and the timing of cladogenic events.
Generic delimitations change as a result of this study. Collospermum is nested within Astelia and is recognized here only at the subgeneric level. Further, Astelia subgenera Astelia, Asteliopsis, and Tricella are paraphyletic and to achieve monophyly their recircumscriptions are proposed. Despite the presence of
Asteliaceae taxa on multiple Gondwanan landmasses and proposed Cretaceous origins for the family,
radiation of genera was during the Tertiary. The largest and oldest genus, Astelia s.l. (including Collospermum), radiated around the Eocene/Oligocene boundary (ca. 34.2 million years ago (Ma)). Astelia s.l. subgenera diverged from the Oligocene/Miocene boundary onwards (<24.0 Ma). These dates suggest that
current distributions are most likely to be the result of long-distance dispersal. Alpine taxa in New Zealand and Australia radiated during the Late Miocene/Pliocene. These results are congruent with Astelia
micro- and macro-fossil data and suggest that Astelia s.l. either persisted in New Zealand during the proposed Oligocene marine transgression or dispersed from Australia after the subsequent expansion of terrestrial habitat.
Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction
Asteliaceae currently contains 4 genera and 36 species that are
perennial, rhizomatous herbs with linear, alternate leaves. Leaves,
inflorescences, and flowers typically bear distinctive silvery-white
branched hairs that are otherwise rare in Asparagales. Despite its
relatively small size, the family is remarkable for its morphological
variation. Milligania has dry capsular fruit, in contrast to the fleshy
fruit of Astelia, Collospermum, and Neoastelia. Collospermum, Milligania, and some Astelia species have trilocular ovaries, other Astelia
species have unilocular ovaries, and Neoastelia has 3–7 locular ovaries. Skottsberg (including 1934b) suggested that, for Asteliaceae,
bisexual flowers, capsular fruit, and simple hairs, as observed in
Milligania, are plesiomorphic. Breeding systems also vary; Milligania and Neoastelia species are all hermaphroditic, producing
bisexual flowers, Astelia and Collospermum species are typically
⇑ Corresponding author. Present address: Royal Botanic Gardens Melbourne,
Private Bag 2000, Birdwood Avenue, South Yarra, VIC 3141, Australia. Fax: +61
392522413.
E-mail address: Joanne.Birch@rbg.vic.gov.au (J.L. Birch).
1055-7903/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ympev.2012.05.031
dioecious, producing unisexual flowers, and one Astelia species is
gynodioecious.
Distributions of Asteliaceae taxa span Austral and Pacific regions. The center of generic diversity is in Australia with three
of the four genera found there, while species diversity is greatest in New Zealand. Astelia (26 species) is the largest and most
widely distributed of the four genera with species in Australia,
New Zealand, and South America, the Mascarene Islands in the
Indian Ocean, and seven archipelagos in the Pacific. Astelia
species occupy a wide variety of habitats including coastal and
lowland forests, low elevation swamps, alpine fellfields, and high
elevation bogs. Collospermum (four species) is found as an epiphyte in New Zealand lowland forests and in Fiji, Vanuatu,
and Samoa in lowland and tropical montane cloud forests. Milligania (five species) is endemic to Tasmania, where it occupies
habitats ranging from lowland riparian valleys to alpine fellfields. The monotypic genus Neoastelia is a terrestrial herb that
is restricted to temperate rainforests in northeastern New South
Wales.
Generic circumscription within Asteliaceae remains poorly
understood. The most controversial aspect is the segregation of
Author's personal copy
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
Collospermum from Astelia by Skottsberg (1934b). He recognized
the former as distinct based on its simple lateral racemes,
dimorphism of staminate and pistillate plants, basifixed anthers,
long style papillae, and mucilaginous seed hairs, which differ from
the predominantly paniculate lateral racemes, versatile anthers,
and poorly developed mucilaginous seed hairs of Astelia. Within
Astelia, Skottsberg (1934b) recognized three subgenera (Astelia,
Asteliopsis, and Tricella) based on degree of tepal fusion, ovary division, and seed shape. Within these subgenera, he recognized seven
sections (Astelia, Desmoneuron, Isoneuron, Micrastelia, Palaeastelia,
Periastelia, and Tricella) based on leaf venation, pistillode size, seed
surface features, and extent of funicle development. Moore (1980)
considered that a re-evaluation of the circumscription of Astelia
and Collospermum was warranted due to the production of viable
progeny from intergeneric crosses. Further, the inclusion of Milligania in the family has been challenged as this placement was
poorly supported in a cladistic analysis of morphological and
molecular data (Rudall et al., 1998). It has been considered as ‘‘a
divergent outlier’’ in the family. Conversely, Bayer et al. (1998)
questioned the separation of Astelia, Collospermum, and Neoastelia
into distinct genera. Maciunas et al. (2011) identified a sister relationship for Neoastelia/Milligania and Collospermum/Astelia clades
in a cladistic study of morphological data. However, intrageneric
relationships remain largely unresolved as only the sister relationship of Astelia and Collospermum received bootstrap support
greater than 50%. In short almost nothing is clear about the relationships among genera within the family or of their individual
boundaries.
Asteliaceae (Asparagales) has been considered an ‘‘Austral’’ floristic element, as defined by its presence on the eastern Gondwanan
landmasses, Australia, New Zealand, and South America (Fleming,
1975; Fosberg, 1948). This hypothesis is supported by age estimates
for the family’s stem and crown lineages at 102 and 92 million years
(Ma), respectively (Janssen and Bremer, 2004) and by the estimate
for the divergence of Milligania from remaining astelioid genera at
79 Ma (Wikström et al., 2001). Neither study sampled Asteliaceae
comprehensively. Therefore, these estimates can only be taken as
a first approximation. The estimated age of the family dictates that
the most likely explanation for its current distribution is vicariance,
following the path of the separating continents. Later movements
across the Pacific no doubt involved long-distance overwater dispersal, but little is known about the pathways of such dispersal
events.
Fossil records document the presence of Asteliaceae since the
mid-Tertiary. The earliest fossil record is that of Astelia pollen from
Southland, New Zealand in deposits dated to the mid-Oligocene
(Couper, 1960). Macrofossils, including leaf cuticle bearing trichomes, also document the presence of Astelia in Central Otago,
New Zealand in the late Miocene (Maciunas et al., 2011; Mildenhall
and Pocknall, 1989; Pole, 2007). Astelia is not recorded in the Australian fossil record until either the Late Pliocene (MacPhail et al.,
1993) or the Quaternary (MacPhail et al., 1994). A single fossil pollen grain, which may belong to Asteliaceae, is recorded from deposits dated as Oligocene to Pliocene from West Point Island in the
Falkland Islands. This fossil pollen datum, considered either as ‘‘Astelia-type’’ (Birnie and Roberts, 1986) or ‘‘Monosulcites sp. A.’’
(MacPhail and Cantrill, 2006), may indicate a Tertiary presence of
Astelia in South America.
This study reconstructed the Asteliaceae molecular phylogeny
with comprehensive taxonomic and geographic sampling based
on chloroplast and nuclear DNA sequence data. The phylogeny
was used to establish monophyletic groups, referable to genera,
to date the divergence times of major Asteliaceae clades, and to
compare these divergence dates with the timing of major events
in the geologic history of the southern hemisphere.
103
2. Methods
2.1. Taxon sampling
Forty-six Asteliaceae taxa, including all Astelia taxa (26 species,
4 subspecies, and 7 morphotypes), all Collospermum species (4), the
single Neoastelia species, and 4 of the 5 Milligania species were included in this study. Taxa sampled and their distributions are provided in Table 1. Voucher specimen data are provided in Appendix
1. Multiple representatives of Astelia menziesiana, A. nervosa, and A.
graminea were included as there is considerable morphological
diversity within these species that has resulted in morphotypes
being informally recognized (Courtney pers. commo., Moore,
1966; Skottsberg, 1934a, 1934b; Wagner et al., 1999). Outgroup
taxa were selected from the Amaryllidaceae, Hypoxidaceae, and
Xanthorrhoeaceae in the Asparagales.
2.2. DNA extraction, amplification, and sequencing
Total genomic DNA was extracted from fresh and silica preserved samples using Qiagen DNeasy Plant Mini Kits and from herbarium specimens using QIAamp DNA stool minikit (Qiagen,
Valencia, California, USA) according to the manufacturer’s protocols scaled down for the smaller volume of leaf material. Sequence
data were generated from four chloroplast regions (trnL intron,
petL–psbE, psbA–trnH, and rps16–trnK intergenic spacers) (Shaw
et al., 2005, 2007) and one nuclear marker (the third intron in
the nitrate reductase gene, NIA-i3) (Howarth and Baum, 2002).
The primers and protocols used for amplification are provided in
Table 2. Astelia-specific primers were constructed for amplification
of the petL–psbE (psbE2) and rps16–trnK (rps16.2) regions based on
preliminary data obtained using standard primers for each region
(Shaw et al., 2005, 2007). Cloned product was amplified using
the manufacturer’s M13 primers (Promega, Madison, Wisconsin).
PCR amplification of 25 ll reaction volumes included: 2.5 ll of
10 Bioline (NH4) reaction buffer, 2.50 mM of each dNTP,
1.5 mM of Bioline MgCl2, 1.0 mM of each primer, 3.0–5.0 units/
1.0 ll of Taq DNA polymerase, and 10–100 ng of template DNA.
Biolase Red Taq DNA polymerase (Bioline USA Inc, Boston, Massachusetts, USA) was used for amplification of the trnL, psbA–trnH,
and NIA-i3 regions and Biolase Accuzyme Taq polymerase was
used for amplification of the petL–psbE and rps16–trnK regions.
PCR amplifications were carried out on an MJ Research DNA Engine
Dyad PTC 220 Thermocycler (MJ Research Inc., Waltham, Massachusetts, USA).
Reaction products were cleaned with 2 units of Exonuclease I
and 0.4 units of Shrimp Alkaline Phosphatase per 1.0 ll of DNA
(USB, Santa Clara, California, USA) prior to sequencing. Bidirectional sequence data were generated for chloroplast markers directly from amplified product using a 10 dilution of the original
amplification primers. All regions were sequenced for all taxa with
the exception of the petL–psbE region for Hypoxidia sp. (Hypoxidaceae), for which sequence could not be obtained. NIA-i3 sequence
data were obtained for a total of 33 Asteliaceae taxa including Neoastelia, all Milligania, and all taxa in Astelia subg. Asteliopsis and Astelia sect. Periastelia (Figs. 1 and 2). These data were obtained
indirectly from cloned product for 12 taxa (A. australiana, A. graminea, A. graminea ‘‘Dun,’’ A. grandis, A. menziesiana ‘‘fallax,’’ A. menziesiana ‘‘veratroides,’’ A. nervosa ‘‘bronze,’’ A. nervosa ‘‘silver’’, A.
nervosa ‘‘north,’’ A. nivicola var. moriceae, Astelia petriei, and A. skottsbergii) for which high quality data could not be obtained through
direct sequencing of amplified products. Cloning was conducted
with the pGEM-T Easy Vector System as per the manufacturer’s
protocols (Promega, Madison, Wisconsin, USA). At least five and
Author's personal copy
104
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
Table 1
Summary of current Astelia and Collospermum taxonomy (Skottsberg, 1934a, 1934b; Moore, 1966; Wagner et al., 1999) and the geographic distributions of taxa.
Genus/subgenus
Section
Species/subspecies/morphotype
Geographic distribution
Asteliaceae
Astelia Banks & Soland. ex R. Br.
Astelia
Palaeastelia
A. hemichrysa (Lam.) Kunth
Mascarene Islands: Reunion
Asteliella
A.
A.
A.
A.
A.
A.
A.
A.
A.
Desmoneuron
Astelia Banks & Soland. ex R. Br.
Asteliopsis
Isoneuron
Micrastelia
Periastelia
Astelia Banks & Soland. ex R. Br.
Tricella
alpina R. Br. var. alpina
alpina var. novae hollandiae Skottsb.
linearis Hook. f. var. linearis
linearis var. novae zelandiae Skottsb.
papuana Skottsb.
subulata (Hook. f.) Cheeseman
nadeaudii Drake & F. Br.
solandri A. Cunn.
trinervia Kirk.
A. banksii A. Cunn.
A.
A.
A.
A.
A.
A.
A.
A.
A.
A.
Australia: Tasmania
Australia: NSW, Victoria
New Zealand: South, Stewart, Auckland
New Zealand: North, South
Papua New Guinea
New Zealand: South, Stewart, Auckland, Campbell
Society Islands: Tahiti
New Zealand: North, South
New Zealand: North, South
New Zealand: North
neocaledonica Schltr.
pumila (G. Forst.) Gaudich.
argyrocoma A. Heller ex Skottsb.
menziesiana Sm. ‘‘fallax’’
menziesiana Sm. ‘‘forbesii’’
menziesiana Sm. ‘‘menziesiana’’
menziesiana Sm. ‘‘veratroides’’
waialealae Wawra
rapensis Skottsb.
tovii F. Br.
New Caledonia
Argentina, Chile, Falkland Islands
Hawai‘i: Kaua‘i
Hawai‘i: Maui, Moloka‘i
Hawai‘i: Maui, Moloka‘i, Lāna‘i
Hawai‘i: Hawai‘i
Hawai‘i: O‘ahu, Kaua‘i
Hawai‘i: Kaua‘i
Austral Islands: Rapa
Marquesas Islands: Ua Pou, Nuku Hiva
A. australiana (J. H. Willis) L. B. Moore
Australia: Victoria
A. chathamica (Skottsb.) L. B. Moore
A. fragrans Colenso
A. graminea L. B. Moore
A. graminea ‘‘Dun’’
A. grandis Hook. f. ex Kirk
A. nivicola Cockayne ex Cheeseman var.
nivicola
A. nivicola var. moriceae L. B. Moore
A. nervosa Hook. f.
A. nervosa Hook. f. ‘‘bronze’’
A. nervosa Hook. f. ‘‘north’’
A. nervosa Hook. f. ‘‘Stokes’’
A. petriei Cockayne
A. psychrocharis F. Muell.
A. skottsbergii L. B. Moore
New
New
New
New
New
New
Collospermum Skottsb.
C. hastatum (Colenso) Skottsb.
C. microspermum (Colenso) Skottsb.
C. montanum (Seem.) Skottsb.
New Zealand: North, South
New Zealand: North
Fiji: Viti Levu, Vanua Levu, Kandavu. Vanuatu: Espiritu Santo,
Tanna, Aneityum
Samoa: Upolu, Savaii
Milligania J. D. Hook.
M. densiflora J. D. Hook.
M. johnstonii F. Muell. ex Benth.
M. lindoniana Rodway ex Curtis
M. stylosa (F. Muell. ex J. D. Hook.) F. Muell.
ex Benth.
Australia:
Australia:
Australia:
Australia:
Neoastelia J. Williams
N. spectablis J. Williams
Australia: New South Wales
Hypoxis hirsuta (L.) Coville
Hypoxis hygrometrica Labill. subsp.
hygrometrica
Curculigo capitulata (Lour.) Kuntze
Hypoxidia sp.
Crinum asiaticum L
Bulbinella hookeri (Hook.) Cheeseman
Dianella sandwicensis Hook. & Arnott
South eastern Canada and eastern United States
Australia: New South Wales
Tricella
C. samoense Skottsb.
Outgroups
Hypoxidaceae
Hypoxidaceae
Hypoxidaceae
Amaryllidaceae
Xanthorrhoeaceae
Xanthorrhoeaceae
up to seven replicates per taxon with cloned products were sequenced, with the exception of A. menziesiana ‘‘veratroides,’’ A.
nervosa ‘‘bronze’’ and A. nivicola var. moriceae, for which only three,
two, and two replicates were obtained, respectively. NIA-i3 sequence data could not be obtained for taxa in Astelia subg. Astelia
or Collospermum (Figs. 1 and 2). Sequence data from the NIA-i3 region was extremely variable and, therefore, difficult to align between distantly related clades. Therefore, it was considered
Zealand:
Zealand:
Zealand:
Zealand:
Zealand:
Zealand:
Chatham
North, South, Stewart
South
South
North, South
South
New Zealand: South
New Zealand: North, South, Stewart
New Zealand: North, South
New Zealand: North
New Zealand: South
New Zealand: South
Australia: New South Wales
New Zealand: South
Tasmania
Tasmania
Tasmania
Tasmania
Asia
Seychelles
Southeast Asia
New Zealand
Hawai‘i
inappropriate for resolution of families and sequence data were
not generated for outgroup taxa.
Sequence data were generated using an ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit (PE Biosystems, Foster City, CA) at the Advanced Studies in Genomics, Proteomics, and
Bioinformatics facility at the University of Hawai‘i at Mānoa. Sequence fragments were edited in Sequencher 4.9 (Gene Codes Corporation, Ann Arbor, Michigan, USA) and manually aligned in
Author's personal copy
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
105
Table 2
Genomic regions, primers, and protocols used for amplification and sequencing.
Region
Primer name
Sequence (50 –30 )
Reference
Amplification protocol
Chloroplast
trnL
trnL50 UAAF (TabC)
GGGGATAGAGGGACTTGAAC
Taberlet et al. (1991)
94 °C (2 min), 35 (94 °C (2 min), 54 °C (1 min),
72 °C (2 min + 7 s/cycle)), 72 °C (7 min)
psbA–trnH
3 trnL
psbA
CGAAATCGGTAGACGCTACG
GTTATGCATGAACGTAATGCTC
Sang et al. (1997)
95 °C (7 min), 35 (95 °C (30 s), 57 °C (30 s), 72 °C
(1 min)), 72 °C (10 min)
rps16–trnK
trnH(GUG)F
rps16x2F2
CGCGCATGGTGGATTCACAATCC
CGCTCAACCTACAGGAACT
Tate and Simpson (2003)
Shaw et al. (2007)
petL–psbE
trnK(UUU)x1
rps16.2
petL
TTAAAAGCCGAGTACTCTACC
CGCTCAACCTACAGGAACT
AGTAGAAAACCGAAATAACTAGTTA
Designed for this study
Shaw et al. (2007)
psbE
psbE2F
TATCGAATACTGGTAATAATATCAGC
AAGTGTGAATTAATGGGTTGGG
Designed for this study
NIA-i3F
AARTAYTGGTGYTGGTGYTTYTGGTC
Howarth and Baum (2002)
95 °C (2 min), 35 (95 °C (1 min), 63 °C (1.5 min
0.4 °C/cycle), 72 °C (2 min)), 72 °C (10 min)
NIA-i3R
GAACCARCARTTGTTCATCATDCC
M13F
GTTTTCCCAGTCACGAC
pGEM-T easy vector protocols
(Promega, Madison, Wisconsin)
94 °C (2 min), 35 (94 °C (2 min), 54 °C (1 min),
72 °C (2 min + 7s/cycle)), 72 °C (7 min).
M13R
CAGGAAACAGGTATGAC
0
Nuclear
NIA-i3
UAA
R (TabD)
80 °C (5 min), 35 (95 °C, (1 min), 54 °C (1 min),
72 °C (2 min)), 72 °C (4 min)
95 °C (7 min), 35 (95 °C (30 s), 56 °C (1 min), 68 °C
(2 min)), 72 °C(7 min)
Cloned insert
Sequence Alignment Editor (Rambaut, 1996). For the NIA-i3 region
extensive sequence variation made unambiguous alignment difficult. Alignment of this region was achieved by completing an initial alignment in MUSCLE (Edgar, 2004) of sequence data for taxa
within clades, with assignment within clades identified from a preliminary reference phylogeny based on a maximum parsimony
search of the chloroplast dataset. A second alignment step was
completed to align sequence data for taxa among clades and was
also completed in MUSCLE (Edgar, 2004).
2.3. Phylogenetic analyses
Missing sequence was coded as ambiguous (N) if data for a region were not available or were incomplete. For taxa for which the
NIA-i3 region was sequenced via cloning, the duplicate copies were
combined using Sequencher 4.9 (Gene Codes Corporation, Ann Arbor, Michigan, USA) to form a consensus sequence. The sites that
differed were coded using standard ambiguous base calls (Nomenclature Committee of the International Union of Biochemistry,
1985). Each marker was assessed for its contribution to phylogenetic resolution individually, in combination with other markers
from the same genome (chloroplast), and as a combined dataset
(chloroplast and nuclear). The incongruence length difference
(ILD) test (Farris et al., 1995) implemented as the Partition Homogeneity Test in PAUP 4.0b10 (Swofford, 2002) was used to test
compatibility of data partitions. Constant uninformative characters
were excluded prior to the ILD test to reduce overestimation of
incongruence (Cunningham, 1997; Lee, 2001). Fifty partition
homogeneity replicates were performed based on the heuristic
searches with 20 random taxon-addition sequences, tree bisection–reconstruction (TBR), and the MulTrees option on, swapping
a maximum of 1000 trees to completion.
Maximum parsimony (MP) analyses were performed in PAUP
4.0b10 (Swofford, 2002) on individual, combined chloroplast, and
combined chloroplast and nuclear datasets. Parsimony heuristic
searches were conducted using a two-step process. The first step
included 1000 random taxon-addition sequence replicates using
tree bisection–reconstruction (TBR) holding 10 trees per replicate,
the second step used as the starting trees those obtained from
the first step and completed 1000 random taxon-addition sequence replicates using tree bisection–reconstruction (TBR) with
the MulTrees option on and swapping to completion for a maximum of 100,000 trees. Bootstrap analyses under the MP criterion
were based on heuristic searches with 1000 bootstrap replicates
each with 25 random taxon-addition sequences, tree bisection–
reconstruction (TBR), and swapping to completion on all minimal
length trees.
Tree reconstruction using the maximum likelihood (ML) criterion was performed in RAxML (Zwickl, 2006). The best-fit models
of molecular evolution and model parameters for separate and
combined datasets (excluding gap characters) were determined
using the corrected Akaike Information Criterion implemented in
jModelTest (Guindon and Gascuel, 2003; Posada, 2008). Given that
the bestfit models identified in jModelTest were not available in
RAxML, the closest model available, GTR + C, was used. ML reconstructions were conducted on individual and combined datasets as
documented for MP analyses. All free model parameters were estimated by RAxML. Nonparametric bootstrap values were calculated
using the rapid bootstrap algorithm with 1000 replicates.
Tree reconstruction using Bayesian inference (BI) criteria was
conducted in MrBayes v3.1.2 (Huelsenbeck and Ronquist, 2001).
Again, given that the bestfit models identified in jModelTest were
not available in MrBayes, the closest model, GTR + C, was used.
Six data partitions were specified for Bayesian searches with all
parameters free to vary across each of the data partitions. Bayesian
analyses were performed using Markov Chain Monte Carlo
(MCMC) sampling with two replicates containing four chains
(one hot and three cold chains) each with the heating temperature
set to 0.10. Five million generations were completed with sampling
every 1250 generations. A burn-in period of 1260 generations was
estimated by calculating the average standard deviation of split
frequencies and confirmed by plotting log likelihood values versus
generation time to determine the point at which log likelihood values became stable. All trees generated during the burn-in period
were discarded from each run and the remaining trees were combined to construct a 50% majority-rule consensus tree that was
visualized in MEGA 4.2 (Tamura et al., 2007). Bayesian posterior
probabilities were calculated to estimate internal branch support
of phylogenetic reconstructions.
The Shimodaira–Hasegawa (SH) test was applied in PAUP
4.0b10 (Swofford, 2002) to test the hypothesis that Astelia pumila
is sister to Astelia subg. Asteliopsis. Constraint trees were generated
Author's personal copy
106
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
(B)
A. menziesiana "fallax"
A. menziesiana "menziesiana"
A. menziesiana "veratroides"
A. rapensis
Subg. Asteliopsis sect. Periastelia
A. argyrocoma
A. waialealae
A. menziesiana "forbesii"
66/-
A. menziesiana "veratroides"
A. tovii
A. argyrocoma
A. menziesiana "fallax"
A. chathamica
63/ 66
A. fragrans
65/76
A. nivicola var. nivicola
A. nervosa “silver”
A. rapensis
A. nervosa "north"
A. grandis
A. tovii
A. graminea
A. australiana
73/92
A. waialealae
A. banksii
Subg. Asteliopsis sect. Isoneuron
A. neocaledonica
A. australiana
- / 100
89/97
93
73/92
A. nadeaudii
A. solandri
A. chathamica
Subg. Astelia sect. Desmoneuron
A. trinervia
A. fragrans
A. petriei
A. skottsbergii
A. graminea "Dun"
A. nivicola var. moriceae
96/97
A. nervosa "Stokes"
65/65
A. graminea
0.97
A. graminea "Dun"
A. grandis
A. psychrocharis
-/89
74/-
Subg. Asteliopsis sect. Periastelia
A. menziesiana "menziesiana"
1.00
A. nervosa "bronze"
A. menziesiana "forbesii"
0.99
1.00
Milligania johnstonii
A. nivicola var. moriceae
Milligania densiflora
Milligania
87/ -
Milligania lindoniana
83/100
90
Neoastelia spectabilis
Subg. Tricella
A. nivicola var. nivicola
Milligania stylosa
Neoastelia
-/0.99
-/89/94
A. nervosa “silver”
A. nervosa "bronze”
A. nervosa "Stokes"
1.00
A. nervosa "north"
1.00
A. petriei
A. psychrocharis
A. skottsbergii
99/86
98/100
1.00
1.00
74/71
Collospermum montanum
0.99
Collospermum samoense
Collospermum
Collospermum hastatum
Collospermum microspermum
98/100
80/92
1.00
100/100
-/-
99/100
0.98
1.00
A. banksii
Subg. Asteliopsis sect. Isoneuron
1.00
A. neocaledonica
64/73
A. nadeaudii
0.98
A. solandri
Subg. Astelia sect. Desmoneuron
A. trinervia
A. pumila
1.00
Subg. Asteliopsis sect. Micrastelia
A. linearis var. linearis
67/71
A. linearis var. novae- zelandiae
0.99
A. subulata
-/66
A. alpina var. novae-hollandiae
0.89
-/68
100/100
1.00
0.93
-/67
100/100
-
1.00
A. alpina var. alpina
A. alpina nov. var.
A. hemichrysa
1
100/100
99/99
0.96
100/100
1.00
1.00
99/100
Milligania densiflora
1.00
Milligania johnstonii
99/100
Milligania lindoniana
1.00
100/100
1.00
Subg. Astelia sect. Palaeastelia
Milligania
Milligania stylosa
Neoastelia spectabilis
100/100
Subg. Astelia sect. Astelia
A. papuana
Neoastelia
Curculigo capitulata
Hypoxidia sp.
Hypoxidaceae
1.00
100/100
1.00
-/79
-
(A)
Hypoxis hirsuta
Hypoxis hygrometrica
Bulbinella hookeri
Xanthorrhoeaceae
Dianella sandwicensis
Crinum asiaticum
Amaryllidaceae
Fig. 1. Bayesian inference 50% consensus topology based on (A) the chloroplast dataset and (B) the nuclear dataset. Numbers above branches are bootstrap values (maximum
parsimony/maximum likelihood) and those below branches are Bayesian posterior probabilities. Bootstrap values >65% and posterior probabilities >0.90 are shown.
Taxonomic treatment is that of Skottsberg (1934a).
Author's personal copy
107
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
A. waialealae
(B)
A. tovii
A. rapensis
A. menziesiana "fallax"
Sect. Periastelia
A. menziesiana "menziesiana"
A. menziesiana "forbesii"
A. menziesiana "veratroides"
A. argyrocoma
A. nervosa "north"
A. nervosa "bronze"
A. nervosa “silver”
A. argyrocoma
Subg. Tricella
A. australiana
A. chathamica
A. fragrans
A. graminea
A. menziesiana "veratroides"
Sect. Tricella
A. graminea "Dun"
A. grandis
A. petriei
A. menziesiana "fallax"
97/99
A. skottsbergii
-/-
A. nivicola var. moriceae
1.00
A. nervosa "Stokes"
A. menziesiana "forbesii"
Sect. Periastelia
1.00
A. psychrocharis
A. menziesiana "menziesiana"
A. nivicola var. nivicola
Collospermum montanum
Collospermum samoense
Subg. Collospermum
Collospermum hastatum
A. rapensis
Collospermum microspermum
A. banksii
A. tovii
Sect. Isoneuron
A. neocaledonica
Subg. Asteliopsis
A. nadeaudii
Sect. Desmoneuron
A. solandri
A. waialealae
A. trinervia
Incertae sedis
Sect. Micrastelia
A. pumila
A. linearis var. linearis
- / 63
A. linearis var. novae-zelandiae
A. nervosa “silver”
A. subulata
Sect. Astelia
A. alpina var. novae-hollandiae
- / 67
Subg. Astelia
A. papuana
0.99
1.00
A. alpina nov. var.
A. alpina var. alpina
A. nervosa "bronze"
A. nervosa "north"
Sect. Palaeastelia
A. hemichrysa
Milligania densiflora
Milligania
johnstonii
Milligania
lindoniana
Subg. Tricella s.l.
A. australiana
Milligania
Milligania stylosa
A. chathamica
97/100
Neoastelia spectabilis
Hypoxidia sp.
Curculigo capitulata
1.00
Hypoxidaceae
Hypoxis hygrometrica
Hypoxis hirsuta
Bulbinella hookeri
Dianella sandwicensis
A. fragrans
Xanthorrhoeaceae
-/-
A. graminea
Amaryllidaceae
0.99
A. graminea "Dun"
Crinum asiaticum
0.2
A. grandis
Sect. Tricella
A. petriei
A. skottsbergii
90/93
1.00
A. nivicola var. moriceae
-/-
A. nervosa "Stokes"
0.90
A. psychrocharis
A. nivicola var. nivicola
78/75
0.99
98/100
100/84
1.00
Collospermum montanum
Collospermum samoense
Subg. Collospermum
1.00
Collospermum hastatum
Collospermum microspermum
99/100
1.00
81/90
98/100
1.00
100/100
-/80
99/99
0.99
1.00
1.00
A. banksii
Sect. Isoneuron
A. neocaledonica
Subg. Asteliopsis s.s.
A. nadeaudii
A. solandri
Sect. Desmoneuron
A. trinervia
A. pumila
1.00
Sect. Micrastelia
Incertae sedis
A. linearis var. linearis
66/ 65
A. linearis var. novae-zelandiae
0.99
A. subulata
-/ 65
A. alpina var. novae-hollandiae
0.91
100/91
-/ -
0.96
0.95
Subg. Astelia s.s.
A. alpina nov. var.
100/100
A. alpina var. alpina
1.00
A. hemichrysa
100/100
100/100
1.00
100/100
1.00
100/100
Milligania densiflora
1.00
Milligania johnstonii
1.00 100/92
1.00
Milligania lindoniana
1.00
1.00
Sect. Palaeastelia
Milligania
Milligania stylosa
Neoaselia spectabilis
100/100
100/100
Sect. Astelia
A. papuana
Neoastelia
Curculigo capitulata
Hypoxidia sp.
Hypoxidaceae
100/100
1.00
Hypoxis hirsuta
Hypoxis hygrometrica
Bulbinella hookeri
Dianella sandwicensis
Crinum asiaticum
(A)
Xanthorrhoeaceae
Amaryllidaceae
Fig. 2. Bayesian inference 50% consensus topology based on the combined (chloroplast and nuclear) dataset. Numbers above branches are bootstrap values (maximum
parsimony/maximum likelihood) and those below branches are Bayesian posterior probabilities. Bootstrap values >65% and posterior probabilities >0.90 are shown. Inset is
the phylogram of the Bayesian inference 50% consensus tree. Taxonomy reflects revisions proposed in the text.
Author's personal copy
108
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
in MacClade 4.07 (Maddison and Maddison, 2003) in which all
relationships were unresolved with the exception of that at the
node being tested. The constraint tree was compared with the optimal ML tree estimated by RAxML. Likelihood scores of the optimal
and the constraint trees were generated with resampling-estimated log-likelihood (RELL) optimization on. Hypotheses were rejected if the log-likelihood score of the alternate tree topologies
were significantly different (P-values <0.05) to that of the optimal
ML tree.
2.4. Divergence time estimation
The hypothesis of molecular clock-like evolution was tested
using the likelihood ratio test (LRT) in PAUP 4.0b10 (Swofford,
2002) based on each individual marker, the chloroplast regions,
and the combined chloroplast and nuclear datasets with a molecular clock enforced and unenforced. LRT resulted in P-values
<0.05 for all tests (excluding that for the trnL region), indicating
that a clock-like rate of evolution is not supported. Therefore, relaxed-clock methods were applied for determination of timing of
cladogenesis. Two relaxed-clock methods were applied to investigate the effect of rate evolution models and associated assumptions on divergence date estimation in Asteliaceae. The first, the
penalized likelihood (PL) method was applied in r8s (Sanderson,
2002). This method uses a parametric model and incorporates an
optimality criterion that introduces a penalty for rapid rate change
between adjacent sites. Divergence time estimates were calculated
using the phylogeny inferred from the ML search of the combined
chloroplast and nuclear data sets. The optimality criterion was estimated using the cross-validation method in r8s (Sanderson, 2004).
Rate variation among sites was modeled using the gamma distribution with a shape parameter value of 0.4070 as estimated by
jModelTest (Posada, 2008). Multiple starts using different initial
starting conditions were used to assess convergence on the global,
rather than local, optimum. Ninety-five percent confidence limits
for divergence times were calculated in the following way: One
hundred bootstrap alignment replicates were generated in Seqgen
1.3.2 (Powell and Vander Kloet, 1997) and branch length was estimated for all replicates based on the optimal ML topology in PAUP
4.0b10 (Swofford, 2002). Node ages were estimated for all replicates in r8s and from these 95% confidence intervals were calculated in Microsoft Excel.
The second method for estimation of divergence time uses
Bayesian Markov chain Monte Carlo analyses applying a relaxed
molecular clock with uncorrelated rates with a log-normal distribution (UCLN). This method was applied in BEAST ver. 1.6.1
(Drummond and Rambaut, 2007). A Yule speciation tree prior,
which assumes a constant rate of speciation per lineage, was specified. As the most appropriate model selected for the combined
chloroplast and nuclear dataset by jModelTest (Posada, 2008) (Table 3) was not available in BEAST the next closest model, GTR + C,
was applied. All parameters were estimated in BEAST. Two independent MCMC analyses of 10 million generations were performed
with sampling every 1000 generations. One thousand trees generated during the burn-in period were discarded from each run and
the remaining trees were combined using LogCombiner ver. 1.6.1.
Parameters were checked using Tracer ver. 1.6.1. An effective sample size of >150 was obtained for all estimated parameters, to ensure parameter space was sampled sufficiently for valid parameter
estimation (Drummond and Rambaut, 2007).
In PL analyses the divergence of Asteliaceae from the Hypoxidaceae was fixed at 79 Ma (stem lineage) and in Bayesian analyses
Asteliaceae (crown lineage) was given a normal prior constrained
with a mean of 79.0 Ma (SD ± 6 Ma) based on divergence time estimates of Wikström et al. (2001). The fossil record, including microfossils (Couper, 1960) and macrofossils (Mildenhall and Pocknall,
1989; Pole, 2007), documents the presence of Astelia pollen in
New Zealand during the Oligocene. Based on these fossil records
a date of 26 million years was used as a minimum constraint for
the clade containing all Astelia and Collospermum taxa. An alternative placement at the node of Astelia subg. Astelia sect. Astelia,
based on the affiliation of the recently described macrofossil Astelia
antiquua Maciunas et al. (2011) was considered, however, support
for the affiliation of this macrofossil is undermined by lack of resolution in the phylogeny. Therefore a more conservative placement
at the node representing the most recent common ancestor for Astelia and Collospermum was used. In PL analyses a minimum age
constraint of 26.0 Ma was applied to the crown node of the clade
containing all Astelia and Collospermum taxa. In Bayesian (UCLN)
analyses a prior of 26.0 Ma was applied to the crown node of the
clade containing all Astelia and Collospermum taxa with a log-normal distribution (mean = 1, standard deviation = 1.5, offset = 25.0)
representing a hard lower bound of 25.0 Ma and a soft upper
bound (95% HPD: 25.1–36.8 Ma).
3. Results
3.1. Sequence characteristics and tree topologies
The number of aligned characters, percent missing data, number of gap characters for each of the five regions and for all datasets, and tree statistics for parsimony and best-fit models of
molecular evolution are presented in Table 3. Thirty-three unambiguously aligned indels were also identified in the chloroplast sequences and were treated as separate characters for phylogeny
reconstruction when the datasets were combined. Of these, nine
(27.3%) were informative for Hypoxidaceae, 20 (60.6%) were informative for Asteliaceae, and 4 (12.1%) were shared across multiple
Table 3
Summary of individual and combined matrix partition statistics, maximum parsimony tree statistics, and models of sequence evolution for individual and combined datasets as
selected using the corrected AIC criterion in jModeltest (Posada, 2008).
a
b
Marker
trnL
psbA–trnH
rps16–trnK
petL–psbE
Chloroplast
NIA-i3
Combined
Number of terminals
Number of aligned characters
Missing data (%)
Number of trees (MPa)
Tree length (MPa)
Number of PICsb
CI
RI
RCI
Number of gap characters
Model of evolution
53
634
0.40
7470
165
53
0.89
0.90
0.80
5
HKY + C
53
711
0.10
9470
161
52
0.80
0.81
0.64
2
TVM + I + C
53
885
2.20
9980
556
181
0.84
0.88
0.74
9
TVM + C
52
1182
3.00
2517
491
142
0.90
0.92
0.83
17
TVM + C
53
3412
1.60
9990
1275
457
0.85
0.89
0.75
33
TPM1uf + C
33
786
14.50
10,000
489
221
0.91
0.94
0.85
n/a
TrN + C
53
4198
10.60
10,000
1726
649
0.86
0.90
0.78
33
TIM1 + I + C
Maximum parsimony searches.
Parsimony informative characters.
Author's personal copy
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
non-monophyletic clades or were autapomorphic. Of the twenty
Asteliaceae indels, four (12.1%) were present in all Asteliaceae,
two (6.1%) were present in all Milligania taxa, one (3.0%) was
shared by M. densifolia and M. johnstonii, one (3.0%) was present
in the clade containing Milligania and Neoastelia, three (9.1%) were
present in all Astelia s.l. taxa, five (15.2%) and three (9.1%) were
informative at the subgeneric and sectional ranks, respectively,
and one (12.5%) was present in all A. nervosa taxa.
ILD tests, comparing length differences between paired regions,
indicated that all chloroplast DNA regions were congruent
(P > 0.05) with the exception of psbA–trnH when paired with other
regions (P = 0.01–0.03). The psbA–trnH region contained an 8-bp
microinversion that appeared to contribute to this incongruence.
No incongruence (P > 0.05) was detected in paired ILD tests for
psbA–trnH with any other region when this motif was excluded.
Therefore, it was excluded from all datasets and congruent regions
were concatenated for subsequent analyses.
Tree topologies inferred from parsimony, maximum likelihood,
and Bayesian searches of the combined dataset were strongly congruent identifying nine well-supported major clades. The Bayesian
inference 50% consensus topology based on the chloroplast dataset
(Fig. 1A), the nuclear dataset (Fig. 1B), and the combined chloroplast
and nuclear datasets (Fig. 2A and B) are presented. Individual chloroplast markers showed limited resolution and are not shown, but
are mentioned in the text if supported clades differed from those obtained in searches based on the chloroplast and combined datasets.
3.2. Phylogenetic relationships within Asteliaceae
Astelia is paraphyletic in all searches based on all datasets as
Collospermum is always nested within it (Figs. 1 and 2). The clade
containing all Astelia and Collospermum taxa, referred to here as
Astelia s.l., is monophyletic. Milligania was monophyletic and Neoastelia was sister to Milligania in all searches based on the chloroplast and combined datasets (Figs. 1 and 2). None of the Astelia
s.l. subgenera Astelia, Asteliopsis, or Tricella (Skottsberg, 1934b)
were recovered as monophyletic in searches based on the chloroplast and combined datasets (Fig. 1). Within Astelia s.l. the four
clades that correspond to recircumscribed subgenera will be discussed (and are referred to by their revised subgeneric names).
The first clade, subg. Astelia s.s. (Fig. 2), includes all sections previously placed in subg. Astelia (sections Astelia and Palaeastelia), with
the exception of section Desmoneuron. The second major clade, Asteliopsis s.s. includes section Isoneuron (previously included in subg.
Asteliopsis), and subgenus Astelia sect. Desmoneuron. The sister relationship of sect. Micrastelia and subg. Asteliopsis s.s. received strong
support in BI searches. However, this relationship received weak
support in MP and ML searches of the chloroplast and combined
chloroplast and nuclear datasets (Figs. 1 and 2). In ML searches
sect. Micrastelia was frequently recovered as sister to the clade
containing subg. Tricella, subg. Asteliopsis sect. Periastelia, and Collospermum (topology not shown). Log-likelihood support for the
latter placement was significantly different (P < 0.01) to that of
the former placement in a Shimodaira–Hasegawa test. All species
previously placed in Collospermum form a clade, recognized here
as subg. Collospermum. That clade is sister to a clade referred to
as subg. Tricella s.l., a well supported clade containing all taxa from
subg. Tricella as well as taxa in subg. Asteliopsis sect. Periastelia
(Fig. 2). Relationships within subg. Tricella s.l. are poorly resolved
and differ among MP, ML, and BI searches.
3.3. Dating estimates
The divergence time estimates calculated using penalized likelihood (PL) and Bayesian (BI) methods are presented in Table 4.
The chronogram derived from the BI analysis is presented in
109
Fig. 3. Divergence date estimates obtained from PL and BI analyses
were congruent, with the PL estimates within the BI 95% highest
posterior densities (HPD) for all clades with the exception of subg.
Tricella s.l., which diverged 0.1 Ma earlier than that interval. In both
PL and BI analyses the most recent common ancestor (MRCA) of
Astelia s.l. diverged at the Eocene/Oligocene boundary (Table 4;
33.4–34.2 Ma). The divergence of the clade containing Astelia s.l.
subg. Asteliopsis s.s., sect. Micrastelia, subg. Tricella s.l., and subg.
Collospermum and the subg. Astelia s.s. clade occurred during the
Early Oligocene (Table 4; 29.1–30.4 Ma). The MRCA of the clade
containing Astelia s.l. subg. Asteliopsis s.s. and sect. Micrastelia also
diverged around the Oligocene/Miocene boundary (Table 4; 21.6–
24.0Ma). The MRCA of Neoastelia and Milligania, and Astelia s.l.
subg. Astelia s.s., diverged during the Miocene (7.6–21.3 Ma). Milligania and Astelia subgenera Collospermum and Tricella s.l. radiated
during the Late Miocene/Early Pliocene (3.9–9.6 Ma).
4. Discussion
4.1. Phylogenetic relationships within Asteliaceae
Asteliaceae is supported as monophyletic in these analyses.
However, the limited taxonomic sampling of closely related Astelioid families precludes definitive determination of the monophyly
of Asteliaceae. Within Asteliaceae, Astelia s.l. is sister to a Milligania/Neoastelia clade. Milligania and Neoastelia are strongly supported as sister (Figs. 1 and 2). Milligania was considered to be
divergent from other Asteliaceae genera due to its semi-inferior
ovary and dry fruit (Rudall, 2003), which differ from the superior
ovaries and fleshy fruit of all other genera. However, ovary position
has been recognized as homoplasious for Asparagales (Rudall,
2003) and lability in fruit type is also documented across the monocotyledons (Givnish et al., 2006). While these characters are synapomorphic for Asteliaceae genera (data not shown) these
characters are not informative for determining generic relationships. Milligania is monophyletic in all reconstructions. A sister
relationship of Milligania and Neoastelia was also identified by
Maciunas et al. (2011) in a phylogenetic study based on morphological data.
A broadly circumscribed Astelia s.l., including Collospermum, is
monophyletic. Astelia s.l. taxa are dioecious or gynodioecious with
flowers that have a pistillode or pistil with a single short or poorly
defined style, a 3-lobed stigma, and fleshy unilocular or trilocular
fruit. The inclusion of Collospermum within Astelia s.l. is consistent
with the close evolutionary relationship recognized for these genera in previous studies. Successful crosses were achieved between
A. nervosa C. hastatum, A. nervosa C. microspermum, and A.
banksii C. hastatum (Moore, 1980) and as a result Moore (1980,
p. 37) stated that the segregation of Collospermum should be reconsidered. Multiple New Zealand Astelia s.s. and Collospermum species
share a chromosome number of n = 35 (Wheeler, 1966).
There are differences, as previously noted, and here we recognize Collospermum at the subgeneric rank within Astelia s.l.. Synapomorphies recognized for Collospermum (Skottsberg, 1934b)
remain valid for subg. Collospermum. These include peltate
branched hairs (in contrast to the hairs of Astelia that are attached
to the leaf surface by a basal stalk), basifixed anthers (in contrast to
the versatile anthers of Astelia), and mucilaginous funicular hairs
that adhere closely to the seeds (in contrast to the mucilaginous
funicular hairs that surround but do not adhere to Astelia seeds).
The two Pacific Collospermum taxa, C. montanum from Fiji and
Vanuatu and C. samoense from Samoa, are sister taxa. The relationships of the New Zealand taxa, C. hastatum and C. microspermum, to
each other and to the Pacific Collospermum taxa are unresolved in
these analyses.
Author's personal copy
110
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
Table 4
Divergence date estimates for Astelia s.l. clades based on analyses of the combined dataset. PL analyses conducted in r8s with the astelioid ingroup fixed at 79 Ma (SD = 6.0 Ma)
(Wikström et al., 2001) and a minimum age of 26 Ma for Astelia s.l. Bayesian UCLN analyses conducted in BEAST with the Asteliaceae stem lineage assigned a prior with a normal
distribution (l = 79.0 Ma, SD = 6.0 Ma, 95% HPD 69.1–88.9 Ma) (Wikström et al., 2001) and the Astelia s.l. stem lineage assigned a prior with a log-normal distribution (l = 1.0,
SD = 1.5, offset = 25.0, 25.0 Ma hard lower bound, soft lower bound at 1, 95% HPD 25.1–36.8 Ma).
Taxon/clade name
Asteliaceae
Neoastelia
Milligania
Astelia s.l.
Subg. Astelia s.s.
Subg. Asteliopsis s.s./sect. Micrastelia
Subg. Asteliopsis s.s./sect. Micrastelia/subg. Collospermum/subg. Tricella s.l.
Subg. Collospermum/subg. Tricella s.l.
Subg. Collospermum
Subg. Tricella s.l.
Sect. Periastelia
a
b
Date estimatesa (million years (Ma))
Penalized likelihood (PL)
Bayesian (UCLN)
69.4b (65.2–71.2)
16.5 (14.2–17.4)
3.9 (3.1–4.6)
33.4 (28.9–34.4)
7.6 (5.8–8.1)
21.6 (18.4–22.4)
29.1 (24.9–30.0)
12.1 (9.8–12.9)
4.0 (2.6–4.7)
4.3 (3.3–4.3)
2.5 (1.8–2.7)
55.4 (36.0–76.0)
21.3 (8.2–37.7)
8.9 (2.6–17.5)
34.2 (26.0–49.1)
12.6 (3.8–24.1)
24.0 (14.0–36.5)
30.4 (19.2–44.6)
16.9 (7.6–27.5)
7.6 (1.8–14.7)
9.6 (4.4–16.1)
3.9 (1.5–6.9)
Asteliaceae (stem lineage) constrained to 79 Ma (as per Wikström et al., 2001).
Node age fixed.
Our results show that taxonomic recircumscription is also required to achieve monophyly of infrageneric taxa within Astelia
s.l.. Subgenus Astelia s.l. includes sections Palaeastelia and Asteliella,
but excludes sect. Desmoneuron. Skottsberg (1934b) considered a
unilocular ovary to be synapomorphic for subg. Astelia, therefore
he included section Desmoneuron in it. However, these results indicate an alternate placement of section Desmoneuron in subgenus
Asteliopsis. Subgenus Astelia is supported by two synapomorphies:
yellow staminate flowers and seeds possessing a long, curved funicle. Ovary division is homoplasious in Astelia s.l., as it is across
Asparagales (Rudall, 2003), with multiple transitions from the plesiomorphic trilocular to a unilocular state.
A monophyletic subgenus Asteliopsis s.s. is recognized that includes sections Isoneuron, and subg. Astelia sect. Desmoneuron,
but excludes sect. Periastelia. As recognized, synapomorphies for
subg. Asteliopsis s.s. include inflorescences with long, lateral subpanicles (5.7–13.3 cm), staminate and pistillate flowers with a
short perianth tube (0.2–1.1 mm), fruits containing numerous (on
average 12–14) seeds, each with a thickened, ribbed funicle. The
relationships of section Micrastelia are poorly resolved. Therefore
the subgeneric placement of section Micrastelia is not possible
and the section remains unplaced (incertae sedis). While sect. Micrastelia is well-supported as sister to subgenus Asteliopsis s.s. in MP
and BA analyses, an alternate placement, sister to the clade containing Collospermum and subg. Tricella (including sect. Periastelia)
is frequently recovered in ML analyses. Section Micrastelia contains
a single taxon, A. pumila, which is a dominant component of Astelia
moorland in Patagonia and the only Asteliaceae taxon found in
South America. It is morphologically divergent from taxa in subg.
Asteliopsis s.s. and is the only cushion forming taxon in a clade that
otherwise contains epiphytic and terrestrial taxa found primarily
in the understory in lowland to montane forests. The alternate
placement indicated by ML analyses is only slightly more meaningful in light of morphology as A. pumila is morphologically quite different from taxa in that clade. However, it does share seed
characters, including an oblong seed, smooth seed surface, and a
short, truncate funicle with taxa in subg. Tricella s.l. Further data
are required to confidently determine the placement of sect. Micrastelia. For Astelia s.l., small, mat or cushion forming species are
found in multiple, widely spaced clades. Therefore, it is clear that
multiple transitions from the plesiomorphic large, open-clustered
habit to a small, mat or cushion forming state have occurred.
The relationship of Collospermum and subgenus Tricella s.l. as
sister taxa is supported by morphological data. These taxa share
long tepal fusion in staminate and pistillate flowers and staminate
flowers with an obpyriform pistillode. Growth on a terrestrial substrate is plesiomorphic in Astelia s.l. and a transition to ephiphytic
growth has occurred multiple times during Astelia s.l. evolution
including a single transition in the ancestral Collospermum lineage.
Evolutionary relationships indicate that a monophyletic subgenus
Tricella s.l. includes subg. Asteliopsis sect. Periastelia. Subgenus Tricella s.l. is morphologically well defined including taxa that have
orange fruit and large fusiform seeds (>2.0 mm) with a short funicle. Within subgenus Tricella s.l. section Periastelia, which includes
taxa from the Austral, Hawaiian, and Marquesas Islands, is monophyletic. All section Periastelia taxa have small (<0.5 mm) flowers
with a short perianth tube (0.3–0.7 mm), short anthers (0.4–
1.1 mm), and large, elliptic–fusiform seeds (2.0–3.0 mm). Relationships among the remaining taxa in subg. Tricella s.l. remain poorly
understood. However, all taxa in section Tricella have flowers of
intermediate (0.4–7.0 mm) size with a long perianth tube (0.8–
4.0 mm), and large, ovoid–oblong seeds (>3.0 mm).
4.2. Molecular dating and historical biogeography
Results indicate that Asteliaceae crown group radiated during
the late Palaeocene (BA) (Fig. 3) or early Cretaceous (PL). The mean
date for divergence of Asteliaceae at ca. 55.4 Ma as estimated using
Bayesian inference is more recent than estimates for the family
from previous studies (Janssen and Bremer, 2004; Wikström
et al., 2001). This reflects, in part, topological differences in those
studies, in which Asteliaceae diverge early in the astelioid clade,
precluding direct comparison of these divergence dates. The
Asteliaceae phylogeny presented here achieves comprehensive
taxonomic sampling and relationships are generally well supported, providing a robust phylogeny for estimation of divergence
dates for the family. Correct assignment of fossils for calibration of
chronograms is essential for accurate date estimates (Near and
Sanderson, 2004). The fossil data used in this study are the earliest
unequivocally determined Astelia s.l. micro and macrofossils records and are from well characterized fossil deposits (Bannister
et al., 2012; Couper, 1960; Maciunas et al., 2011; Mildenhall and
Pocknall, 1989; Pole, 2007). The placement of the fossil at the Astelia s.l. crown node reflects the presence of Astelia synapomorphies
including stomata with elongate-triangular subsidiary cells,
unsculptured epidermal cells (Maciunas et al., 2011), and spinulose
pollen grains (Cranwell, 1952). This represents a conservative
placement that reflects the affiliation of these fossil data based
on morphological characters as identified by Maciunas et al.
(2011), rather than reflecting the more tentative relationships
Author's personal copy
111
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
A. menziesiana "fallax"
A. menziesiana "menziesiana"
A. menziesiana "forbesii"
A. tovii
NPAC/
EPAC
A. waialealae
A. rapensis
A. menziesiana "veratroides"
A. argyrocoma
A. fragrans
A. chathamica
A. nervosa “silver”
NZ
Subg. Tricella s.l.
A. nervosa "bronze"
A. nervosa "north"
A. australiana
AUS
A. petriei
A. graminea
A. skottsbergii
A. grandis
NZ
A. nivicoa var. nivicola
A. nervosa "Stokes"
A. nivicola var. moriceae
A. psychrocharis
A. graminea "Dun"
Collospermum montanum
AUS
NZ
WPAC
Collospermum samoense
Collospermum hastatum
Collospermum microspermum
A. neocaledonica
A. banksii
A. nadeaudii
c
A. solandri
A. trinervia
A. pumila
Subg. Collospermum
NZ
WPAC
NZ
NCAL
Subg. Asteliopsis s.s.
NZ
SAM
Incertae sedis
A. linearis var. novae-zelandiae
A. subulata
NZ
A. linearis var. linearis
b
A. alpina var. novae-hollandiae
A. papuana
A. alpina var. alpina
AUS/
PNG
Subg. Astelia s.s.
A. alpina nov. var.
A. hemichrysa
a
MASC
Milligania johnstonii
Milligania densiflora
Milligania stylosa
AUS
Milligania
Milligania lindoniana
Neoastelia spectabilis
Neoastelia
Curculigo capitulata
Hypoxidia sp.
Hypoxis hirsuta
Hypoxidaceae
Hypoxis hygrometrica
Creta
Paleo
65 .0
54.8
Eo
33.7
Oligo
23.8
Mio
Pl P
5.3 1.8
Fig. 3. Asteliaceae chronogram with divergence dates estimated for Asteliaceae crown node constrained at 92 Ma (±10 Ma) using Bayesian inference conducted in BEAST.
Constraints applied at points: (a) the divergence of Asteliaceae (stem lineage), (b) the MRCA of Asteliaceae (crown group), and (c) the MRCA of Astelia s.l. (crown group). Bars
indicate 95% height probability densities, dashed lines delimit the Oligocene. Geographic distributions are shown in the bar on the right: AUS = Australia, NCAL = New
Caledonia, NZ = New Zealand, SAM = South America, WPAC = Fijian, Samoan, and Vanuatuan Islands, EPAC = Austral, Marquesas, and Society Islands, NPAC = Hawaiian Islands.
Geological time scale is shown at the bottom: Cret = Cretaceous, Paleo = Palaeocene, Eo = Eocene, Oligo = Oligocene, Mio = Miocene, Pl = Pliocene, and P = Pleistocene.
Geologic times are as per Berggren et al. (1995).
Author's personal copy
112
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
determined from their incompletely resolved phylogeny. In this
study, the availability and incorporation of this internal fossil calibration allows correction for rate variation across the tree, potentially increasing accuracy of date estimates over previous studies
that were calibrated at a single node. For divergence estimates
based on Bayesian methods, calibration dates were drawn from
probabilistic distributions and a hard (minimum) bound and a soft
(maximum) bound was assigned to the Astelia s.l. fossil calibration,
which allowed uncertainty to be incorporated into fossil date
assignment (Ho, 2007; Yang and Rannala, 2006). The recent diverence dates for Asteliaceae taxa estimated in this study are based on
a well resolved phylogeny constrained with well characterized Astelia s.l. fossil data.
This study provides a temporal framework for the evolution of
Asteliaceae genera for the first time. Astelia s.l. is the oldest genus
in the family, estimated to have radiated close to the Eocene/Oligocene boundary. Within Astelia s.l. the clade containing Astelia s.l.
subg. Asteliopsis s.s., Collospermum, and Tricella s.l. is also dated to
the Eocene (Fig. 3 and Table 4). A Late Eocene/Early Oligocene origin for Astelia s.l. is consistent with the single report of a New Zealand fossil from the Eocene (MacPhail et al., 1993, Mildenhall pers.
comm.) the documentation for which has been lost (Mildenhall
pers. comm.). In contrast, radiation of the clade including the Australian endemic genera Neoastelia and Milligania occurred more recently during the Miocene with the divergence of Neoastelia at ca.
21.3 Ma and the radiation of the Milligania crown group at ca.
8.9 Ma (Table 4).
Relationships predicted by a Gondwanan vicariant scenario reflect the sequence and timing of fragmentation of the Gondwanan
continent. Therefore, a vicariant scenario predicts that extant Australian taxa will be sister to extant South American taxa in a clade
sister to a clade containing extant New Zealand taxa. Additionally,
a variance scenario requires that the divergence of New Zealand
taxa/clades must predate the divergence of New Zealand from
the Gondwanan landmass, which occurred around 80 Ma and that
the divergence of Australian and South American clades/taxa must
predate the separation of Australia from Antarctica, which was
completed approximately 30 Ma (Coleman, 1980). The divergence
of Astelia s.l. (a clade dominated in numbers by extant New Zealand
taxa) from the Neoastelia/Milligania clade, (containing exclusively
Australian extant taxa) is dated between the late Cretaceous and
the late Palaeocene. This node, accompanied by very wide 95%
HPD intervals in the Bayesian analysis (36.0–76.0 Ma), spans the
divergence of New Zealand from the Gondwanan landmass. Therefore, a Gondwanan history for diversification of Asteliaceae genera
cannot be ruled out.
However, within Astelia s.l., neither branching sequence, nor
date estimates are consistent with the predictions of a Gondwanan
vicariant scenario. The radiation of Astelia s.l. from ca. 34.2 Ma,
places all subsequent cladogenesis events well after the separation
of New Zealand from Gondwana. Additionally, neither extant taxa
from Australia nor from New Zealand are present in monophyletic
clades. While it is recognized that divergence date estimates that
are calibrated with fossil data, such as these, provide minimum
date estimates, based on current data, the mean dates for the
divergence of New Zealand, Australian, and South American
clades/taxa would have to fall outside the 95% HPD values to be
congruent with predictions based on vicariance scenarios. We consider that evolution within Astelia s.l. is more likely to be the result
of long-distance dispersal.
Date estimates for divergence of the basal Astelia s.l. nodes are
relevant to the ongoing debate regarding the persistence of faunal
and floral taxa in New Zealand during the Oligocene. New Zealand
was a low lying land mass during this time and land area was significantly reduced as a result of extensive marine transgression between 23 and 27 Ma (Cooper and Cooper, 1995; Fleming, 1975).
Astelia s.l. radiated from ca. 34.2 Ma onwards. Subsequently, the
clade containing subgenera Collospermum and Tricella s.l. diverged
from its sister clade containing subg. Asteliopsis and sect. Micrastelia, at ca. 30 Ma. Astelia s.l. subgenera Asteliopsis s.s. and Tricella
s.l. are both dominated by extant New Zealand taxa. If cladogenesis
at the basal Astelia s.l. nodes occurred in New Zealand, as is suggested by the distributions of extant taxa, these divergence dates
imply that Astelia s.l. was present in New Zealand during the
Oligocene.
The presence of Astelia s.l. in New Zealand as early as the Oligocene is also supported by Astelia micro- and macrofossil records,
which have been identified in Early Miocene to the Late Oligocene
deposits from Southland, New Zealand (Maciunas et al., 2011; Mildenhall and Pocknall, 1989; Pole, 2007; Raine et al., 2008). The fossil record suggests that Astelia was found in sub-tropical to warmtemperate forest that was differentiated into canopy and understory strata (Maciunas et al., 2011; Mildenhall and Pocknall,
1989) in New Zealand during this time. In combination, fossil
and molecular dating data suggest that Astelia s.l. was present in
New Zealand during the period of the proposed Oligocene drowning making it one of only a few lineages for which data are consistent with such a scenario (Allwood et al., 2010; Knapp et al., 2007;
Trewick and Morgan-Richards, 2005). An alternate scenario, that
Astelia s.l. was present in Australia during the Early Oligocene
and dispersed from there to New Zealand during the Late Oligocene/Miocene boundary, is also possible. While this scenario cannot be ruled out, it is considered less likely based on both these
dating and fossil data.
The divergence of the South American taxon, A. pumila, from an
otherwise predominantly New Zealand clade is dated to ca.
24.0 Ma. This estimate indicates that Astelia s.l. may have been
present in South America from as early as the Oligocene. This is
congruent with a South American Liliales pollen datum in strata
dated between the Oligocene to the Early Pliocene (in maximum
age limits) and between the middle Miocene to the Early Pliocene
(in best-fit estimates) (MacPhail and Cantrill, 2006). However,
there is disagreement as to whether this record is of Asteliaceae.
Birnie and Roberts (1986) determined it to be Astelia-type pollen,
whereas MacPhail and Cantrill (2006) determined it only as
‘‘Monosulcites sp. A’’. Therefore, congruence of these results with
palynological data awaits accurate determination of this pollen
grain. The presence of Astelia pumila in South America is considered
to be the result of long-distance dispersal from a New Zealand origin. A stepping-stone long-distance dispersal pathway from New
Zealand to South American via Antarctica would have been available for long-distance dispersal during the Oligocene as coastal forests were present in Antarctica until the Miocene when cooling and
extensive ice development eliminated coastal vegetation (Truswell, 1990).
Multiple dispersals between New Zealand and Australia are inferred for Astelia s.l. as Australian taxa are widely separated in Astelia subgenera Astelia s.s. and Tricella s.l. In subgenus Astelia s.s.
Australian taxa are basal and New Zealand taxa are in a nested
clade suggesting a long-distance dispersal event from Australia to
New Zealand. In contrast, in subgenus Tricella s.l. Australian taxa
are nested in subg. Tricella sect. Tricella, which otherwise contains
only New Zealand taxa, suggesting at least one long-distance dispersal from New Zealand to Australia. Both subgenera, Astelia s.s.
and Tricella s.l., radiated during the Miocene (Fig. 3 and Table 4)
and both are dominated by taxa from subalpine/alpine habitats.
Despite its relatively recent origin, subg. Tricella s.l. contains more
than half of the New Zealand Astelia s.l. taxa. The radiation of these
clades coincides with the increasing availability of high elevation
habitat in Australia and New Zealand as a result of climatic cooling
and tectonic activity during the Miocene (Galloway and Kemp,
1981; Smith, 1982). High mountains were largely absent from
Author's personal copy
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
southeastern Australia until the Middle to Upper Miocene (Galloway and Kemp, 1981; Smith, 1982) (although Ollier (1986) suggests the Kosciusko uplift may have occurred earlier). Tectonic
activity commenced in New Zealand with the Kaikoura Orogeny
in the Miocene (from 15 Ma) and reached a peak during the Pliocene (Ollier, 1986; Winkworth et al., 2005). The recent radiation
of these taxa in New Zealand may have been facilitated by the
availability of cool-temperate habitats as a result of the Pliocene
mountain building in New Zealand. A recent origin and rapid radiation has also been proposed for other New Zealand alpine plant
taxa (Lockhart et al., 2001; Meudt and Simpson, 2006; Wagstaff
and Garnock-Jones, 1998; Winkworth et al., 2005).
Milligania and Neoastelia are endemic to Tasmania and New
South Wales, respectively. These genera are associated with wetter
climates and habitats that typically provide good conditions for
fossil preservation. However, neither genus is documented in the
pollen record. The Australasian fossil record is sparse in many
areas, particularly for monocotyledons (Herendeen and Crane,
1995). Additionally, neither genus is widespread in contemporary
habitats. Rather, they typically have restricted distributions within
which they are locally common. In light of the divergence dates
estimated for these taxa, their absence in the pollen record may
indicate similarly restricted distributions in the past. If contemporary habitats of extant taxa are indicative of historical habitats of
ancestral taxa, suitable habitat would have been available during
the time periods estimated for these radiations. Neoastelia spectabilis, is currently found in cool-temperate Nothofagus rainforest in
northern New South Wales (Williams, 1987). Moist, closed rainforest vegetation dominated by Nothofagus persisted in southeastern
Australia until the Middle Miocene (Nott and Owen, 1992). Extant
Milligania species are now found on alluvial soils adjacent to lakes,
on riparian cliffs in temperate rainforest, in cushion plant-dominated moorlands, and as part of alpine bog vegetation in southern
and western Tasmania. In Tasmania lowland rainforests were present as early as the Early Oligocene (Hill, 2004) and a well-developed subalpine and alpine vegetation was present from the early
Miocene (Hill and Gibson, 1989).
5. Conclusions
This study, providing date estimates based on the Asteliaceae
phylogeny with near complete taxonomic representation, is important for understanding southern hemisphere biogeography. We
provide the first phylogeny resolving the relationships of Asteliaceae taxa and use this phylogeny to investigate the historical biogeography of the family. Although it is found across Australia, New
Zealand, and South America, the genus Astelia s.l. is not Gondwanan. Its crown radiation diversified at the Eocene/Oligocene
boundary and subgenera radiated during the Oligocene onwards.
The timing of Astelia s.l. evolution presented here, alongside Astelia
micro- and macro-fossil data, implies that Astelia s.l. was present in
New Zealand during the Oligocene. Astelia s.l. is one of only a few
taxa for which such a scenario is inferred. These results suggest
that a portion of the New Zealand flora may be the result of an earlier phase of long-distance dispersal than has not previously been
recognized. Molecular dating of additional putatively early Gondwanan lineages is necessary to identify other lineages that have
undergone long-distance dispersal during the Eocene or Oligocene
to more comprehensively understand the evolutionary history of
the New Zealand flora.
Acknowledgments
The authors thank the following individuals and organizations
for their support of this research. Access and collection permits
113
and field assistance were provided by the Arthur Rylah Institute
for Environmental Research (Victoria, Australia), Auckland Regional Council (New Zealand), Department of Conservation (New Zealand), Department of Environment and Climate Change (New South
Wales, Australia), Department of Land and Natural Resources (Hawai‘i, United States of America), Department of Primary Industries
and Water (Tasmania, Australia), Department of Sustainability and
Environment (Victoria, Australia), Hawai‘i Volcanoes National Park
National Park (Ha(Hawai‘i, United States of America), Haleakala
wai‘i, United States of America), O‘ahu Army Natural Resources
(Hawai‘i, United States of America), O‘ahu Plant Extinction Program (Hawai‘i, United States of America), Maui Land and Pineapple
(Hawai‘i, United States of America), The Nature Conservancy (Hawai‘i, United States of America). Silica-preserved material was
kindly provided for inclusion of taxa in the phylogeny by Australian National Botanical Garden, Missouri Botanical Garden, Otari
Native Botanic Garden, and individuals J.-Y. Meyer, T. Motley, D.
Strasberg, and A. Whistler. Herbarium specimens were provided
to the Joseph F. Rock Herbarium by the following herbaria: Auckland War Memorial Museum (AK), Herbarium Pacificum (BISH), Allan Herbarium (CHR), Harvard University (GH), Kew Royal Botanic
Gardens (K), National Herbarium of Victoria (MEL), Missouri
Botanical Garden (MO), Herbier National de Paris (P), United States
National Herbarium (US), and Museum of New Zealand Te Papa
Tongarewa (WELT). Funding for field work and sequencing was
provided by the American Society of Plant Taxonomists, Explorers
Club, National Science Foundation DDIG (DEB-0910402), Sigma-Xi,
University of Hawai‘i Foundation and multiple affiliates of the University of Hawai‘i at Mānoa (UHM) including the Botany Department, Ecology, Evolution, and Conservation Biology Program,
Graduate Division, Graduate Student Organization, and the Office
of Community and Alumni Relations. This manuscript benefitted
from comments of A. Sherwood, D. Lorence, and R. Cowie and
two anonymous reviewers.
Appendix A. Voucher specimen information (collector,
collection number, collection location, collection date,
herbarium) for taxa and GenBank accession numbers for DNA
sequence data (trnL, psbA-trnH, 3’rps16-trnK, petL-psbE, NIA-i3)
used in phylogenetic analysis. Missing data for a given region is
listed as ‘‘-’’.
A. alpina R. Br. var. alpina – Birch (JLB373), Australia, Tasmania,
Mt. Field National Park, 8 Jan 2008 (BISH), JX182500, JX182553,
JX182606, JX182659,-. A. alpina var. novae-hollandiae Skottsb. –
Birch (JLB351), Australia, Victoria, Alpine National Park, 20 Dec
2008 (BISH), JX182501, JX182554, JX182607, JX182660,-. A. alpina
nov. var. – Birch and Buchannan (JLB387), Australia, Tasmania,
Southwest National Park, Mt. Eliza Plateau, 13 Jan 2009 (BISH),
JX182502, JX182555, JX182608, JX182661,-. Astelia argyrocoma
A. Heller ex Skottsb. – Birch, Lorence, and Wood (JLB137), Kauai,
Awaawapuhi Trail, 12 Jun 2007 (BISH), JX182503, JX182556,
JX182609, JX182662, JX182711. A. australiana (J. H. Willis) L. B.
Moore – Birch and Downe, (JLB342), Australia, Victoria, Yarra State
Park, 16 Dec 2008 (BISH), JX182504, JX182557, JX182610,
JX182663, JX182712. A. banksii A. Cunn. – Birch (JLB228), New
Zealand, Auckland, Waitakere Ranges, Piha Beach, 16 Dec 2007
(BISH), JX182505, JX182558, JX182611, JX182664, JX182713. A.
chathamica (Skottsb.) L. B. Moore – Birch (JLB399) (cultivated),
New Zealand, Waikanae Beach, 9 Jul 2011 (BISH), JX182506,
JX182559, JX182612, JX182665, JX182714. A. fragrans Colenso –
Birch and Brown (JLB243), New Zealand, Kahurangi National Park,
Heaphy Track, 2 Jan 2008 (BISH), JX182507, JX182560, JX182613,
JX182666, JX182715. A. graminea L. B. Moore – Birch, Courtney,
and Gaskill (JLB257), New Zealand, Kahurangi National Park, Mt.
Author's personal copy
114
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
Arthur, 15 Jan 2008 (BISH), JX182508, JX182561, JX182614,
JX182667, JX182716. A. graminea L. B. Moore ‘‘Dun’’ – Birch
(JLB271), New Zealand, Mt. Richmond Forest Park, Dun Saddle,
Maungatapu Track, 18 Jan 2008 (BISH), JX182509, JX182562,
JX182615, JX182668, JX182717. A. grandis Hook. f. ex Kirk – Birch
(JLB223), New Zealand, Hamilton, Puketaha Gully, 14 Dec 2007
(BISH), JX182510, JX182563, JX182616, JX182669, JX182718. A.
hemichrysa (Lam.) Kunth – D. Strasberg (LR488), Mascarene Islands, Reunion Island, Col Bellevue, 15 Aug 2007 (REU),
JX182511, JX182564, JX182617, JX182670,-. A. linearis Hook. f.
var. linearis – Birch and Michel (JLB276), New Zealand, Southland,
Longwood Range, Bald Hill, 22 Jan 2008 (BISH), JX182512,
JX182565, JX182618, JX182671,-. A. linearis var. novae–zelandiae
Skottsb. – Birch (JLB273), New Zealand, Paparoa Range, Croesus
track, 19 Jan 2008 (BISH), JX182513, JX182566, JX182619,
JX182672,-. A. menziesiana Sm. ‘‘fallax’’ – Birch and Bartlett
(JLB157), Maui, Puu Kukui Watershed Preserve, 31 Aug 2007
(BISH), JX182514, JX182567, JX182620, JX182673, JX182719. A.
menziesiana Sm. ‘‘forbesii’’ – Welton s.n., Maui, Haleakala National
Park, 18 Mar 2009 (BISH), JX182515, JX182568, JX182621,
JX182674, JX182720. A. menziesiana Sm. ‘‘menziesiana’’ – Birch
and Gulizia (JLB107), Hawaii, Hilo Forest Reserve, Huumulu Trail,
10 Sept 2006 (BISH), JX182516, JX182569, JX182622, JX182675,
JX182721. A. menziesiana Sm. ‘‘veratroides’’ – Birch (JLB058), Oahu,
Mt Kaala, 21 Jun 2006 (BISH), JX182518, JX182571, JX182624,
JX182677, JX182723. A. nadeaudii Drake & F. Br. – Yves-Meyer
s.n., Society Islands, Tahiti, Mt. Aorai, 14 Dec 2006 (BISH),
JX182519, JX182572, JX182625, JX182678, JX182724. A. neocaledonica Schltr. – Munzinger 595 and McPherson, New Caledonia,
Province du Nord, 31 Mar 2001 (MO), JX182520, JX182573,
JX182626, JX182679, JX182725. A. nervosa Hook. f. ‘‘silver’’ – Birch
(JLB260), New Zealand, Kahurangi National Park, Mt. Arthur, 16 Jan
2008 (BISH), JX182522, JX182575, JX182628, JX182681, JX182727.
A. nervosa Hook. f. ‘‘bronze’’ – Birch, Courtney, and Gaskill
(JLB263), New Zealand, Kahurangi National Park, Mt. Arthur, 15
Jan 2008 (BISH), JX182522, JX182575, JX182628, JX182681,
JX182727. A. nervosa Hook. f. ‘‘north’’ – Birch and Birch (JLB245),
New Zealand, Tararua Forest Park, Mt. Holdsworth, 12 Jan 2008
(BISH), JX182524, JX182577, JX182630, JX182683, JX182729. A.
nervosa Hook. f. ‘‘Stokes’’ – Birch and Little (JLB251), New Zealand,
Marlborough Land District, Marlborough Sounds, Mt. Stokes, 14 Jan
2007 (BISH), JX182523, JX182576, JX182629, JX182682, JX182728.
A. nivicola Cockayne ex Cheeseman var. nivicola – Birch, Courtney,
and Gaskill (JLB256), New Zealand, Kahurangi National Park, Mt.
Arthur, 15 Jan 2008 (BISH), JX182521, JX182574, JX182627,
JX182680, JX182726. A. nivicola var. moriceae L. B. Moore – Birch
and Brown (JLB233), New Zealand, Kahurangi National Park, Perry
Saddle, 30 Dec 2007 (BISH), JX182517, JX182570, JX182623,
JX182676, JX182722. A. papuana Skottsb. – Rudall 71 Papua
New Guinea, 20 Mar 1999 (K), JX182526, JX182579, JX182632,
JX182685,-. A. petriei Cockayne – Birch, Courtney, and Gaskill
(JLB254), New Zealand, Kahurangi National Park, Mt. Arthur, 15
Jan 2008 (BISH), JX182527, JX182580, JX182633, JX182686,
JX182731. A. psychrocharis F. Muell. – Birch and Beehag
(JLB362), Australia, New South Wales, Kosciuszko National Park,
Kosciuszko summit trail, 30 Dec 2008 (BISH), JX182528,
JX182581, JX182634, JX182687, JX182732. A. pumila (G. Forst.)
Gaudich. – Goodall (RNG2248), Argentina, Tierra del Fuego, 14
Nov 1969 (US), JX182529, JX182582, JX182635, JX182688,-. A. rapensis Skottsb. – Motley and Fenstemacher 2722, Austral Islands,
Rapa Iti (NY), JX182530, JX182583, JX182636, JX182689,
JX182733. A. skottsbergii L. B. Moore – Birch (JLB261), New Zealand, Kahurangi National Park, Mt. Arthur, 16 Jan 2008 (BISH),
JX182531, JX182584, JX182637, JX182690, JX182734. A. solandri
A. Cunn. – Birch and Birch (JLB250), New Zealand, Tararua Forest
Park, Mt. Holdsworth, 12 Jan 2008 (BISH), JX182532, JX182585,
JX182638, JX182691, JX182735. A. subulata (Hook. f.) Cheeseman
– Birch and Blythal (JLB301), New Zealand, Rakiura National Park,
Mt. Anglem, 26 Jan 2008 (BISH), JX182533, JX182586, JX182639,
JX182692,-. A. tovii F. Br. – Wood (KRW10803), Marquesas Islands,
Ua Pou, 2 Jul 2004 (PTBG), JX182534, JX182587, JX182640,
JX182693, JX182736. A. trinervia Kirk – Birch and Brown
(JLB242), New Zealand, Kahurangi National Park, Heaphy Track,
31 Dec 2007 (BISH), JX182535, JX182588, JX182641, JX182694,
JX182737. A. waialealae Wawra – Perlman (SP20093), Hawaii, Kawai, Alakai Swamp, 27 Jul 2006 (PTBG), JX182536, JX182589,
JX182642, JX182695, JX182738. Collospermum hastatum (Colenso) Skottsb. – JLB s.n. (cultivated), New Zealand, Waikanae Beach,
22 Jul 2007 (BISH), JX182538, JX182591, JX182644, JX182697,-.
C. microspermum (Colenso) Skottsb. – Birch and Birch (JLB244),
New Zealand, Tararua Forest Park, Mt. Holdsworth, 12 Jan 2008
(BISH), JX182539, JX182592, JX182645, JX182698,-. C. montanum
(Seem.) Skottsb. – Cahation 2816, Vanuatu, Espiritu Santo, 17
Aug 1985 (BISH), JX182540, JX182593, JX182646, JX182699,-. C.
samoense Skottsb. – Whistler (AW 12035), Samoan Islands, Upolu,
15 Oct 2008 (stored at HAW), JX182541, JX182594, JX182647,
JX182700,-. Milligania densiflora J. D. Hook. – Birch and Jordan
(JLB374), Australia, Tasmania, Mt. Hartz National Park, 8 Jan 2009
(BISH), JX182548, JX182601, JX182654, JX182706, JX182739. M.
johnstonii F. Muell. ex Benth. – Croft 10119 and Richardson (cultivated), Australia, Tasmania, Southwest National Park, 17 Feb 1989
(CBG), JX182559, JX182602, JX182655, JX182707, JX182740. M.
lindoniana Rodway ex Curtis – Birch (JLB371), Australia, Tasmania,
Mt. Field National Park, 8 Jan 2009 (BISH), JX182550, JX182603,
JX182656, JX182708, JX182741. M. stylosa (F. Muell. ex J. D. Hook.)
F. Muell. ex Benth. – Birch (JLB 384), Australia, Tasmania, Mt. Field
National Park, 12 Jan 2009 (BISH), JX182551, JX182604, JX182657,
JX182709, JX182742. Neoastelia spectabilis J. Williams – Birch and
Dwyer (JLB367), Australia, New South Wales, New England National Park, 4 Jan 2009 (BISH), JX182552, JX182605, JX182658,
JX182710, JX182743. Bulbinella hookeri (Hook.) Cheeseman –
Birch (JLB253), New Zealand, Marlborough Land District, Marlborough Sounds, Mt. Stokes, 14 Jan 2007 (BISH), JX182537, JX182590,
JX182643, JX182696,-. Crinum asiaticum L. – Birch (JLB398) (cultivated), Hawaii, University of Hawaii at Manoa, East–West Road, 27
Apr 2011 (BISH), JX182542, JX182595, JX182648, JX182701,-. Curculigo Gaertn. capitulata (Lour.) Kuntze – Steele 1081 (cultivated),
Missouri, University of Missouri Columbia, 29 Sep 2009 (UMO),
JX182543, JX182596, JX182649, JX182702,-. Dianella sandwicensis Hook. & Arnott – Birch (JLB173), Hawaii, Kauai, Na Pali Kona
Forest Reserve, 9 Sept 2007 (BISH), JX182544, JX182597,
JX182650, JX182703,-. Hypoxidia F. Friedmann sp. – Chase (MW,
17,387), (K), Sri Lanka, Kitulgala, 14 Jan 2003 (BISH), JX182547,
JX182600, JX182653,-,-. Hypoxis hirsuta (L.) Colville – Gibson
s.n., Texas, Smithville, Stengl (Lost Pines) Biology Station, 13 Apr
2009 (BISH), JX182545, JX182598, JX182651, JX182704,-. H.
hygrometrica Labill subsp. hygrometrica – Birch (JLB364), Australia, New South Wales, Tumut-Cooma Road, 1 Jan 2009 (BISH),
JX182546, JX182599, JX182652, JX182705,-.
References
Allwood, J., Gleeson, D., Mayer, G., Daniels, S., Beggs, J.R., Buckley, T.R., 2010. Support
for vicariant origins of the New Zealand Onychophora. J. Biogeogr. 37, 669–681.
Bannister, J.M., Conran, J.G., Lee, D.E., 2012. Lauraceae from rainforest surrounding
an early Miocene maar lake, Otago, southern New Zealand. Rev. Palaeobot.
Palynol. 178, 13–34.
Bayer, C., Appel, O., Rudall, P.J., 1998. Asteliaceae. In: Kubitzki, K., Huber, H. (Eds.),
Flowering Plants, Monocotyledons: Lilianae (Except Orchidaceae). SpringerVerlag, Berlin, Germany, pp. 141–145.
Berggren, W.A., Kent, D.V., Swisher, C.C.I., Aubry, M.-P., 1995. A revised Cenozoic
geochronology and chronostratigraphy. In: Berggren, W.A., Kent, D.V., Aubry,
M.-P., Hardenbol, J. (Eds.), Geochronology Time Scales and Global Stratigraphic
Correlation, SEPM Special Publication, pp. 129–212.
Author's personal copy
J.L. Birch et al. / Molecular Phylogenetics and Evolution 65 (2012) 102–115
Birnie, J.F., Roberts, D.E., 1986. Evidence of tertiary forest in the Falkland Islands
(Islas Malvinas). Palaeogeogr. Palaeoclimatol. Palaeoecol. 55, 45–53.
Coleman, P.J., 1980. Plate tectonics background to biogeographic development in
the southwest Pacific over the last 100 million years. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 31, 105–121.
Cooper, A., Cooper, R.A., 1995. The Oligocene bottleneck and New Zealand biota:
genetic record of a past environmental crisis. Proc. Roy. Soc. Lond. B. Biol. Sci.
261, 293–302.
Couper, R.A., 1960. New Zealand Mesozoic and Cainozoic plant microfossils. N.Z.
G.S. Pal. Bull. 49, 1–87.
Cranwell, L.M., 1952. New Zealand Pollen Studies: the Monocotyledons A
Comparative Account. Bulletin of the Auckland Institute and Museum 3.
Cunningham, C.W., 1997. Can three incongruence tests predict when data should be
combined? Mol. Biol. Evol. 14, 733–740.
Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evol. Biol. 7, 214.
Edgar, R.C., 2004. Muscle: multiple sequence alignment with high accuracy and
high throughput. Nucl. Acids Res. 32, 1792–1797.
Farris, J.S., Kallersjo, M., Kluge, A.G., Bult, C., 1995. Constructing a significance test
for incongruence. Syst. Bot. 44, 570–572.
Fleming, C.A., 1975. The geological history of New Zealand and its biota. Biogeogr.
Ecol. New Zealand 27, 1–86.
Fosberg, F.R., 1948. Derivation of the flora of the Hawaiian Islands. In: Zimmerman,
E.C. (Ed.), Insects of Hawai‘i: A Manual of the Insects of the Hawaiian Islands,
Including an Enumeration of the Species and Notes on their Origin, Distribution,
Hosts, Parasites, etc. University Press of Hawai‘i, Honolulu, pp. 107–119.
Galloway, R.W., Kemp, E.M., 1981. Late Cainozoic environments in Australia. In:
Keast, A. (Ed.), Ecological Biogeography of Australia. Dr. W. Junk bv Publishers,
The Hague, The Netherlands, pp. 53–80.
Givnish, T.J., Pires, J.C., Graham, S.W., McPherson, M.A., Prince, L.M., Patterson, T.B.,
Rai, H.S., Roalson, E.H., Evans, T.M., Hahn, W.J., Millam, K.C., Meerow, A.W.,
Molvray, M., Kores, P.J., O’Brien, H.E., Hall, J.C., Kress, W.J., Sytsma, K.J., 2006.
Phylogenetic relationships of monocots based on the highly informative plastid
gene ndhF: evidence for widespread concerted convergence. In: Travis, J.T. (Ed.),
Monocots: Comparative Biology and Evolution (Excluding Poales). Rancho Santa
Ana Botanic Garden, Claremont, California, pp. 28–51.
Guindon, S., Gascuel, O., 2003. A simple, fast and accurate method to estimate large
phylogenies by maximum-likelihood. Syst. Bot. 52, 696–704.
Herendeen, P.S., Crane, P.R., 1995. The fossil history of the monocotyledons. In:
Rudall, P.J., Cribb, P.J., Cutler, D.F., Humphries, C.J. (Eds.), Monocotyledons:
Systematics and Evolution. Royal Botanic Gardens, Kew, London, pp. 1–21.
Hill, R.S., 2004. Origins of the southeastern Australian vegetation. Philos. Trans. Roy.
Soc. Lond. B. Biol. Sci. 359, 1537–1549.
Hill, R.S., Gibson, N., 1989. Macrofossil evidence for the evolution of the alpine and
subalpine vegetation of Tasmania. In: Barlow, B.A. (Ed.), Flora and Fauna of
Alpine Australasia: Ages and Origins. Commonwealth Scientific and Industrial
Research Organization, Canberra, pp. 205–217.
Ho, S.Y.W., 2007. Calibrating molecular estimates of substitution rates and
divergence times in birds. J. Avian Biol. 38, 409–414.
Howarth, D.G., Baum, D.A., 2002. Phylogenetic utility of a nuclear intron from
nitrate reductase for the study of closely related plant species. Mol. Phylogenet.
Evol. 23, 525–528.
Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 17, 754–755.
Janssen, T., Bremer, K., 2004. The age of major monocot groups inferred from 800+
rbcL sequences. Bot. J. Linn. Soc. 146, 385–398.
Knapp, M., Mudaliar, R., Havell, D., Wagstaff, S.J., Lockhart, P.J., 2007. The drowning
of New Zealand and the problem of Agathis. Syst. Bot. 56, 862–870.
Lee, M.S.Y., 2001. Uninformative characters and apparent conflict between
molecules and morphology. Mol. Biol. Evol. 18, 676–680.
Lockhart, P.J., McLenachan, P.A., Havell, D., Glenny, D., Huson, D., Jensen, U., 2001.
Phylogeny, radiation, and transoceanic dispersal of New Zealand alpine
buttercups: molecular evidence under split decomposition. Ann. Mol. Bot.
Gard. 88, 458–477.
Maciunas, E., Conran, J.G., Bannister, J.M., Paull, R., Lee, D.E., 2011. Micene Astelia
(Asparagales: Asteliaceae) macrofossils from southern New Zealand. Aust. Syst.
Bot. 24, 19–31.
MacPhail, M.K., Alley, N.F., Truswell, E.M., Sluiter, I.R.K., 1994. Early Tertiary
vegetation: evidence from spores and pollen. In: Hill, R.S. (Ed.), History of the
Australian Vegetation: Cretaceous to Recent. Cambridge University Press,
Cambridge, pp. 182–262.
MacPhail, M.K., Cantrill, D.J., 2006. Age and implications of the Forest Bed, Falkland
Islands, southwest Atlantic Ocean: evidence from fossil pollen and spores.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 240, 602–629.
MacPhail, M.K., Jordan, G.J., Hill, R.S., 1993. Key periods in the evolution of the flora
and vegetation in western Tasmania I. the Early–Middle Pleistocene. Aust. J. Bot.
41, 673–707.
Maddison, D.R., Maddison, W.P., 2003. MacClade 4: Analysis of Phylogeny and
Character Evolution. Sinauer Associates, Sunderland, Massachusetts, USA.
Meudt, H., Simpson, B.B., 2006. The biogeography of the austral, subalpine genus
Ourisia (Plantaginaceae) based on molecular phylogenetic evidence. South
American origin and dispersal to New Zealand and Tasmania. Biol. J. Linn. Soc.
Lond. 87, 479–513.
Mildenhall, D.C., Pocknall, D.T., 1989. New Zealand geological survey. N.Z. G.S. Pal.
Bull. 59, 128.
Moore, L.B., 1966. Australasian asteliads (Liliaceae). N.Z. J. Bot. 4, 201–240.
115
Moore, L.B., 1980. Hybrid asteliads (Liliaceae). N.Z. J. Bot. 18, 37–42.
Near, T.J., Sanderson, M.J., 2004. Assessing the quality of molecular divergence time
estimates by fossil calibrations and fossil-based model selection. Philos. Trans.
Roy. Soc. B 359, 1477–1483.
Nomenclature Committee of the International Union of Biochemistry, 1985.
Nomenclature for incompletely specified bases in nucleic acid sequences. Eur.
J. Biochem. 150, 1–5.
Nott, J.F., Owen, J.A.K., 1992. An Oligocene palynoflora from the middle Shoalhaven
catchment New South Wales and the Tertiary evolution of flora and climate in
the southeast Australian highlands. Palaeogeogr. Palaeoclimatol. Palaeoecol. 95,
135–151.
Ollier, C.D., 1986. The origin of alpine landforms in Australasia. In: Barlow, B.A. (Ed.),
Flora and Fauna of Alpine Australasia. CSIRO, Melbourne, Australia, pp. 3–26.
Pole, M., 2007. Monocot macrofossils from the Miocene of southern New Zealand.
Palaeontol. Electron. 10, 15A.
Posada, D., 2008. JModelTest: phylogenetic model averaging. Mol. Biol. Evol. 25,
1253–1256.
Powell, E.A., Vander Kloet, S.P., 1997. Foliar venation characteristics of Vaccinium
sect. Macropelma and sect. Myrtillus (Ericaceae): a comparative analysis. Can. J.
Bot. 75, 2015–2025.
Raine, J. I., Mildenhall, D.C., Kennedy, E.M., 2008. New Zealand Fossil Spores and
Pollen: An Illustrated Catalogue. Version 1, GNS Science, Lower Hutt.
Rambaut, A.E., 1996. Se–Al: Sequence Alignment Editor. University of Oxford,
Oxford, England. <http://evolve.zoo.ox.ac.uk/software/Se-Al>.
Rudall, P.J., 2003. Unique floral structures and iterative evolutionary themes in
Asparagales; insights from a morphological cladistic analysis. Bot. Rev. 68, 488–
509.
Rudall, P.J., Chase, M.W., Cutler, D.F., Rusby, J., de Bruijn, A.Y., 1998. Anatomical and
molecular systematics of Asteliaceae and Hypoxidaceae. Bot. J. Linn. Soc. 127,
1–42.
Sanderson, M.J., 2002. Estimating absolute rates of molecular evolution and
divergence times: a penalized likelihood approach. Mol. Biol. Evol. 10, 101–109.
Sanderson, M.J., 2004. A nonparametric approach to estimating divergence times in
the absence of rate constancy. Mol. Biol. Evol. 14, 1218–1231.
Sang, T., Drawford, D.J., Stuessy, T.F., 1997. Chloroplast DNA phylogeny, reticulate
evolution, and biogeography of Paeonia (Paeoniaceae). Am. J. Bot. 84, 1120–
1136.
Shaw, J., Lickey, E.B., Beck, J.T., Farmer, S.B., Liu, W., Miller, J., Siripun, K.C., Winder,
C.T., Schilling, E.E., Small, R.L., 2005. The tortoise and hare II: relative utility of
21 noncoding chloroplast DNA sequences. Am. J. Bot. 92, 142–166.
Shaw, J., Lickey, E.B., Schilling, E.E., Small, R.L., 2007. Comparison of whole
chloroplast genome sequences to choose noncoding regions for phylogenetic
studies in angiosperms: the tortoise and the hare III. Am. J. Bot. 94, 275–288.
Skottsberg, C., 1934a. Astelia and Pipturus of Hawai‘i. Bernice P. Bishop Mus. bull.
117, 3–39.
Skottsberg, C., 1934b. Studies in the genus Astelia Banks et Solander. Kungl. Svenska
vetenskapssakademiens handlingar 14.
Smith, J.M.B. (Ed.), 1982. A History of Australasian Vegetation. McGraw-Hill,
Sydney, Australia.
Swofford, D.L., 2002. PAUP: Phylogenetic Analysis Using Parsimony (and Other
Methods).
Taberlet, P., Gielly, L., Pautou, G., Bouvet, J., 1991. Universal primers for
amplification of three non-coding regions of chloroplast DNA. Plant Mol. Biol.
17, 1105–1109.
Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: molecular evolutionary
genetics analysis (MEGA). Mol. Biol. Evol. 24, 1596–1599.
Tate, J.A., Simpson, B.B., 2003. Paraphyly of Tarasa (Malvaceae) and diverse origins
of the polyploid species. Syst. Bot. 28, 723–737.
Trewick, S.A., Morgan-Richards, M., 2005. After the deluge: mitochondrial DNA
indicates Miocene radiation and Pliocene adaptation of tree and giant weta
(Orthoptera: Anostostomatidae). J. Biogeogr. 32, 295–309.
Truswell, E.M., 1990. Cretaceous and Tertiary vegetation in Antarctica: a
palynological perspective. In: Taylor, T.N., Taylor, E.B. (Eds.), Antarctic
Paleobiology: Its Role in the Reconstruction of Gondwana. Springer-Verlag,
New York, NY, pp. 71–88.
Wagner, W.L., Herbst, D.R., Sohmer, S.H., 1999. Manual of the Flowering Plants of
Hawai‘i. University of Hawai‘i Press, Bishop Museum Press, Honolulu, Hawai‘i.
Wagstaff, S.J., Garnock-Jones, P.J., 1998. Evolution and biogeography of the Hebe
complex (Scrophulariaceae) inferred from ITS sequence. N.Z. J. Bot. 36, 425–437.
Wheeler, J.M., 1966. Cytotaxonomy of the large asteliads (Liliaceae) of the North
Island of New Zealand. N.Z. J. Bot. 4, 95–113.
Wikström, N., Savolainen, V., Chase, M.W., 2001. Evolution of the angiosperms:
calibrating the family tree. Proc. Roy. Soc. Lond. B. Biol. Sci. 268, 2211–2220.
Williams, J.B., 1987. Neoastelia. In: George, A.S. (Ed.), Flora of Australia;
Hydatellaceae to Liliaceae. Australia Government Publishing Service,
Canberra, Australia, pp. 493.
Winkworth, R.C., Wagstaff, S.J., Glenny, D., Lockhart, P.J., 2005. Evolution of the New
Zealand mountain flora: origins, diversification and dispersal. Org. Divers. Evol.
5, 237–247.
Yang, Z., Rannala, B., 2006. Bayesian estimation of species divergence times under a
molecular clock using multiple fossil calibrations with soft bounds. Mol. Biol.
Evol. 23, 212–226.
Zwickl, D.J., 2006. Genetic Algorithm Approaches for the Phylogenetic Analysis of
Large Biological Sequence Datasets under the Maximum Likelihood Criterion.
The University of Texas at Austin, Austin, Texas. <http://www.bio.utexas.edu/
faculty/antisense/garli/Garli.html>.