Emerging Plant Diseases
and Global Food Security Edite d by
JEAN BEAGLE RISTAINO / ANGELA RECORDS
C H A P T E R
Assessing the Global Impacts of Crop Pests and Diseases
2
Serge Savary ■
Unité Mixte de Recherche, Agroécologie, Innovations et Territoires (AGIR); Institut National de la Recherche Agronomique (INRA); Institut National Polytechnique Toulouse (INPT); Institut National Polytechnique–École d’Ingénieurs de Purpan; Université de Toulouse, Castanet–Tolosan, France
Andrea Ficke ■
Division of Biotechnology and Plant Health, Norwegian Institute of Bioeconomy Research, Ås
Disease management has an important role to play in meeting the growing demand for food quality and quantity (Strange & Scott, 2005). Direct, quantitative yield losses caused by crop “pests” (pathogens, animals, and weeds) are together responsible for losses of 20–40% of global agricultural productivity (Oerke, 2006; Oerke et al., 1994; Teng, 1987; Teng & Krupa, 1980). We also discuss the several facets of crop losses—direct or indirect—with consequences in the short or long term (Zadoks, 1967). A main objective of this chapter is to show that “20–40% losses” is inadequate to reflect the true costs of crop losses to consumers, public health, societies, environments, economic fabrics, and growers. Because of the diversity of agrosystems in the world, the range of crop health problems, the variability of these problems, and the massive changes world agriculture is undergoing as a result of climate change, globalization, and technology shifts (Chakraborty & Newton, 2011; Gregory et al., 2009; Gustafson, 2011), a clear understanding of the required metric for the impacts of plant diseases is needed now more than ever. We hope that this chapter contributes to this objective.
Several reports (e.g., Alexandratos & Bruinsma, 2012) have indicated that the biosphere is comfortably able to feed the world today—that enough food is produced to meet the needs of the world population today. These reports and others (e.g., Dyson, 1999; Gustafson, 2011; Van Ittersum et al., 2016) also point to the uneven distribution of food production, to progress in food production, and to access to food around the world. In this chapter, the importance of plant diseases, and of crop pests in general, is highlighted with respect to the impacts on food production and food security. The impacts of plant diseases on the several components of food security are stressed, including the insufficiently recognized component of food quality and safety. The types of impacts plant diseases may have, whether chronic (which affect social and economic fabrics in the long term) or acute (which cause disruptions in the food-production and food-access processes), are addressed. Assessing these impacts; anticipating their magnitude; and developing policies, research, and management tactics—all these different objectives entail the same metric: the assessment of crop losses. 13
14
S E C T I O N I A Global Context
■ Crop Loss: A Keystone Concept
to Analyze the Impacts of Crop Pests and Diseases
The Nature of Crop Losses and the FAO Definition of Yield Loss Plant diseases affect crop performance, farms, agrosystems, and societies in many different ways. As a result, the notion of crop loss entails a range of dimensions that need to be considered within a structured typology. An early such typology (Zadoks, 1967; Zadoks & Schein, 1979) has the following structure: ■
■
Direct losses (1) Primary losses (a) yield, (b) quality, (c) cost of control, (d) extra cost of harvesting, (e) extra cost of grading, (f) costs of replanting, (g) loss of income by less profitable replacement crop (2) Secondary losses (a) contamination of sowing and planting material, (b) soilborne diseases, (c) weakening of trees and other perennials by premature defoliation, (d) cost of control Indirect losses (a) farm, (b) rural community, (c) exporters, (d) trade: wholesale and retail, (e) consumers, (f) government, (g) environment
This typology distinguishes direct and indirect losses caused by plant diseases and, among the direct losses, primary and secondary losses. It does not, however, make a clear distinction between the material losses caused by plant diseases (for example, yield or quality losses) and the economic consequences generated by such losses. As we progress through this chapter, this typology will be expanded to distinguish these components, yielding the framework shown in Figure 2.1, in which “damage” is used as a synonym for “crop loss” in accordance with more recent literature (Main, 1983; Savary et al., 2006; Zadoks, 1985). The bulk of the information on crop losses concerns the first category of the above list—the direct, primary losses in yield. The concepts of potential (or theoretical) yield, attainable yield (that is, crop not injured by disease), and actual yield (of the actually harvested crop) provide yardsticks to measure yield differences, especially yield losses. These concepts enable the assessment of progress and objectives in disease management (Chiarappa, 1971, 1981; Rabbinge et al., 1989; Zadoks, 1967, 1985). The potential yield (Yp) of a crop is determined by the genetic makeup of cultivated plants, current temperature regimes, and radiation, which are called “yield-defining factors.” In reference to the Yp, all other plant requirements are supposed to be met optimally at any time and at each crop development stage throughout a growing season. Yp is therefore achieved without any limitation of nutrients and
F I G . 2 .1. A relational diagram of concepts associated with crop losses. (Courtesy S. Savary and A. Ficke—© APS)
C H A P T E R 2 Assessing the Global Impacts of Crop Pests and Diseases
water at any development stage and without any injury caused by pathogens, animals, or weeds (as well as by abiotic factors such as frost, flood, or pollution). The attainable yield (Ya) depends on the former yield-defining factors of the Yp, on which an array of yield-limiting factors are overlaid. These yield-limiting factors are inherent in each given production situation considered. These may include a shortage of water and/or nutrients at some development stages as well as excesses of water and mineral compounds, which may cause toxicities. And last, the actual yield (Y) is the yield actually harvested and encompasses all the yield-defining factors (of the Yp), all the yield-limiting factors (associated with a given production situation to explain the Ya), and the yield-reducing effects of injuries caused by harmful organisms (along with other abiotic yieldreducing factors). Specifically, the FAO (Food and Agriculture Organization of the United Nations) definition of yield loss, YL, is the difference between the attainable and actual yield levels: Ya – Y (Chiarappa, 1981). The difference between Y and Ya may be offset by using all disease management options that could prevent disease injuries. Implementing all these options under actual farming conditions would not be economical. However, a fraction of the Ya – Y difference may be offset by using available methods up to an economic optimum, that is, the point at which management costs equal the cost of yield losses. Thus, one may consider an “economic yield” level (Ye) that lies between Y and Ya and coincides with this economical level of management. Ye represents the target of optimized level of disease (or pest) management, strictly from a yield (actual yield or attainable yield) point of view. Therefore, while it is economical to recoup the Ye – Y difference, the remainder of the Ya – Y yield difference, Ya – Ye, corresponds to disease management efforts that would today be uneconomical. From a crop yield perspective, Ya – Ye represents the progress that remains to be made in improving disease management (Chiarappa, 1981; Esker et al., 2012). Such progress may derive from (1) more efficient technology for disease management, allowing higher yield gains per unit management effort (or cost); or (2) cheaper disease management methods, enabling more yield gains for the same management effort. Past and present progress in disease management is based on these levels of yields: Y, Ye, Ya, and Yp (Parlevliet, 1981; Rabbinge et al., 1989; Savary & Zadoks, 1992a; Teng & Savary, 1992; Teng et al., 1993). The definition of what should be an “economic
15
yield” is a critical question that lies beyond the scope of this chapter but represents an area of important multidisciplinary research with important social and environmental dimensions (e.g., United Nations Environment Program, 2007; see also Chapter 9). This perspective is also addressed at the end of this chapter. From the epidemiological and the physiological standpoints, it is useful to note that this framework with three levels of yield (potential, attainable, and actual) and their associated factors constitutes a simplification. For instance, rice brown spot (Cochliobolus miyabeanus) is dependent on the occurrence of drought (Chakra barti, 2001) and yield losses caused by Septoria diseases of wheat (Zymoseptoria tritici and Parastagonospora nodorum) are influenced by cropping practices, especially fertilizer inputs (Leath et al., 1993). In these examples, disease development, and thus the injury caused by an epidemic, depends on the levels of some yield-limiting factors (or their alleviation). The underlying mechanisms of such relationships (Rabbinge et al., 1989; Zadoks & Schein, 1979) may involve the predisposition of plants (Schoeneweiss, 1975) to infection by drought or nitrogen supply (i.e., by yield-limiting factors) or the indirect effects of yield-limiting factors on pathogen cycles (e.g., via microclimatic conditions). Such interactions are reflected by the variability in the yield loss–injury relationships, which we address in the next paragraphs.
The Injury–Damage Relationship The assessment of impacts of plant diseases is based on quantitative measurements. An important point in this chapter is therefore to emphasize the quantitative nature of the information used by plant scientists to assess these impacts. Crop loss assessment, that is, the measurement of damage caused by diseases (or pests), involves protocols, experimental procedures (both in the field and under controlled conditions), and a body of statistical and modeling approaches. The methodology for crop loss assessment has been addressed in a series of texts, which we cannot review here, but Table 2.1 provides an overview of important advances with some key references. Crop loss analysis and modeling can be addressed by considering a simplified crop disease system as follows. One may consider a crop harvest, with its quantitative yield and with its many qualitative or other quantitative characteristics (including nutritional value, appearance, food-technological characteristics,
16
S E C T I O N I A Global Context
TA B L E 2 .1. Some important steps in the development of approaches and concepts for the assessment of crop lossesa Approaches and concepts
Domain
References
Plant disease can be measured with scales.
Concepts and methods
Horsfall & Barrat, 1945
Standardized disease measurement serves many purposes, including crop loss assessment, and leads to uniform, shareable information.
Concepts and methods
Large, 1966
Global and crop-wise loss estimates are produced.
Data
Cramer, 1967
Damage mechanisms leading to crop yield reduction can be experimentally quantified.
Experimental measurement
Doling & Doodson, 1968
Scales for different crops and organs are developed to measure disease.
Concepts and methods
James, 1969
Statistical models and mechanistic simulation models can be linked to experimental (field) work to quantify yield losses to diseases and their variation.
Modeling
James & Teng, 1979
International field procedures (such as crop and disease scales or experimental plot sizes) are established.
Concepts and methods
Chiarappa, 1971
Physiological interactions on yield loss between multiple pathogen infections and abiotic factors are experimentally quantified.
Experimental measurement
Van der Wal & Cowan, 1974
Physiological effects (transpiration and CO2 exchange) of infection are experimentally quantified.
Experimental measurement
Ayres, 1976
Statistical models for injury–damage relationships are compared. Concept of yield loss response surface: Crop losses are related to both the level of injuries and the crop development (crop physiological) stage.
Modeling
Teng & Gaunt, 1981
Difference in nature and philosophy between statistical and processbased models established. The scientific meaning of model validation in these two approaches is discussed.
Concepts and methods
Teng & Gaunt, 1981
The concept of damage mechanism is established.
Concepts and methods
Boote et al., 1983; Rabbinge & Rijsdijk, 1981
Disease damage mechanisms are incorporated into an agrophysiological model, and crop losses are simulated.
Modeling
Rabbinge & Vereijken 1980; Rabbinge et al., 1985
Multiple injuries by different pests are artificially manipulated to analyze yield losses.
Experimental measurement
Johnson et al., 1986, 1987
Production situations influence yield losses: Yield losses to multiple diseases and pests depend on levels of injuries in interaction with the attainable yield.
Concepts and methods
Savary & Zadoks, 1992a, 1992b
Multiple disease and pest injuries are incorporated into a process-based agrophysiological model to simulate yield losses.
Modeling
Johnson, 1992
New global, regional, and crop-wise yield loss estimates are produced.
Data
Oerke et al., 1994
Yield loss data corresponding to artificially manipulated levels of injuries caused by 11 yield-reducers (bacteria, viruses, fungi, insects, and weeds) of rice at a range of attainable yields are reported and analyzed.
Data
Savary et al., 2000b
a Courtesy
S. Savary and A. Ficke—© APS.
(continued )
C H A P T E R 2 Assessing the Global Impacts of Crop Pests and Diseases
17
TA B L E 2 .1. Continued Some important steps in the development of approaches and concepts for the assessment a of crop losses Approaches and concepts
Domain
References
The impacts of climate change on crop health, plant disease, and crop losses are considered.
Concepts and methods
Chakraborty et al., 2000
Statistical meta-analysis is introduced into plant pathology, becomes mainstream, and is applied to qualitative (mycotoxin) losses in cereal Fusarium head blight.
Concepts and methods
Madden & Paul, 2011; Paul et al., 2006; Rosenberg et al., 2004
Open-source codes are published for generic, process-based, multiple disease and pest models for rice and wheat. These models reflect effects of production situations on processes and losses.
Modeling
Willocquet et al., 2004, 2008
Disease impacts on ecosystem services, beyond crop losses, are considered.
Concepts and methods
Cheatham et al., 2009
A plant disease epidemic model and a crop loss model are coupled in a geographic information system to model changes in rice yield losses caused by bacterial leaf blight disease.
Modeling
Duku et al., 2016
Yield loss concepts are used to identify and test disease-tolerant cultivars in field experiments.
Experimental measurement
Castro & Simón, 2016
Indirect (polyetic) losses caused by disease are considered in a perennial crop.
Concepts and methods
Allinne et al., 2016; Cerda et al., 2017
or shelf life). These characteristics are reflections of a number of interrelated genetic, ecological, and technical factors, which have all influenced crop growth and development, leading to the considered crop harvest. During the course of the growing season, a disease may have developed that caused injuries to the crop stand and thus may have also affected the crop harvest. As a result, plant disease causes injuries, which in turn may cause damage (crop losses) (Esker et al., 2012; Zadoks, 1985; Zadoks & Schein, 1979). This system is dynamic and nonlinear for three main reasons. First, the dynamics of the disease epidemic, and thus the levels of injury, depend in part on crop growth and development. Second, the epidemic has a nonlinear dynamic, because it also reflects many environmental factors, including those caused by humans. And third, the level of damage varies with the level of injury and does so in a nonlinear fashion because of the complexity of the physiological damage mechanisms affecting the growing crop. Damage (crop loss) may be qualitative (a reduction in the quality of harvested product) or quantitative
(a reduction in the quantity harvested). The bulk of progress that has been made in the field of crop loss analysis and modeling focuses on the quantitative damage, or YL, caused by plant diseases. A core element of the suite of concepts in Table 2.1 is the measure of yield loss, expressed, according to the FAO definition given above, as the difference between the Ya and the Y of a crop, that is, the yield difference between the uninjured crop and the injured crop. YL is thus a quantity that can be experimentally measured and that has a number of key properties: 1. Because YL can be measured within a clear methodological framework, results from different experiments, addressing, for instance, different diseases, different production contexts, or different disease management strategies, can be compared. YL, expressed as a biomass per unit area or as a fraction of attainable yield, thus becomes an objective metric to measure disease risk as well as to assess progress in disease management.
18
S E C T I O N I A Global Context
2. YL can be partitioned into two segments, as indicated above: the amount of yield loss that can be prevented through disease management and the amount that will be lost irrespective of the implementation of such management. If the level of disease management is based on economic reasoning, then one can distinguish the amounts of losses that can “economically” be avoided (to which Ye, the economic yield, corresponds) from those that cannot. YL thus becomes the basis for economic reasoning in disease management with the notion of threshold (Zadoks, 1985). 3. Among the many factors that may influence YL, the nature of the considered disease may come first. Crop loss methodology enables the comparison of plant pathogens (and crop yield reducers in general) in terms of the harm they cause. This enables, for instance, comparative risk analyses, assessment of crop cultivars for their vulnerability (tolerance) to diseases, or the setting of priorities in research in the plant sciences.
Injuries, Crop Loss, Economic Loss, and Uncertainty A consequence of an epidemic, which represents an injury to a crop stand, that is, an alteration of its functioning, is a damage (or crop loss); and an economic consequence of a damage is a loss (Esker et al., 2012; Main, 1983; Zadoks, 1985; Zadoks & Schein, 1979). This is a double relationship: injury–damage–loss is not automatic because not all epidemics cause actual damage (crop loss) and not all damage results in (economic) losses. Instead, epidemics may lead to damage, which may lead to (economic) loss (Rabbinge et al., 1989; Zadoks, 1985). Further, the injury–damage–loss relationships are not linear (Campbell & Madden, 1990; Esker et al., 2012; James, 1974; Large, 1966; Madden, 1983; Madden et al., 2000, 2007; Savary et al., 2006; Teng, 1987). The first relationship, injury–damage, is the damage function, and the second relationship, damage–loss, is the loss function (Teng, 1987; Zadoks, 1985). Damage functions are dependent primarily on the nature of the damage mechanisms triggered by diseases (or, more generally, by harmful agents), within the host crop.
The nonlinearity of injury–damage relationships was, for instance, analyzed by Madden et al. (2000) in the case of plant virus diseases, with the compounding complexity elements of heterogeneous injury distribution in a crop stand and variable timing of epidemic onset. The damage–loss, i.e., the (economic) loss functions (Zadoks, 1985), on the other hand, are dependent primarily on production situations (Savary & Zadoks, 1992b; Savary et al., 2006), including the attainable crop yield, the objectives of agricultural production, market variations, and, more generally, the socioeconomic context in which production is taking place (Rabbinge, 1982). The nonlinearity of these injury–damage and damage–loss relationships is at the origin of difficulties in decision-making in disease management. It is primarily because of this nonlinearity, for instance, that producers are faced with a “gray area” in which uncertainty lies (Zadoks, 1989). The very purpose of sustainable disease management (and of plant protection in general)—from the producer’s standpoint—is to reduce the size of this gray area. In their discussion on the development of disease management strategies, McRoberts et al. (2011) emphasized again the need of jointly considering the probability of disease occurring (and causing a given level of injury) and the probability of such a level of injury causing crop losses. The diagram in Figure 2.1 is presented as a guide to crop loss concepts. This is not a list of possible losses (Zadoks, 1967; Zadoks & Schein, 1979), as discussed earlier in this chapter, but instead a series of concepts grouped within domains and connected according to their relationships. This organization of concepts may enable more flexibility and genericity. First is the “production situation” domain, with its associated crop profile (synthesizing crop management) and its attainable crop performances, including its attainable yield. Next, and associated to the production situation domain, is the “crop health” domain, with elements such as the injury profile, individual injuries, and individual diseases (or pests). Third is the “damage” domain, with its primary and secondary damages, both connected to the production situation and crop health domains and linked to various forms of primary and secondary losses. And fourth is the “economic loss” domain, with its direct economic losses and its indirect economic losses—at the farm, the rural community, the environment, the consumers, and the exporters and trade levels.
C H A P T E R 2 Assessing the Global Impacts of Crop Pests and Diseases
■ Three Facets of Crop Loss The concepts and methodological foundations summarized above and in Table 2.1 provide a basis to further analyze crop losses in different directions. First, crop losses occur in the widely varying production situations of the agricultural world, where a range of crop diseases and pests occur; consideration of production situations and injury profiles is necessary in crop loss analysis. Second, the variability over time and over space is a key feature of crop losses caused by diseases and pests that must be addressed. Third, crop losses do not affect only food provisioning, but actually all the components of food security to varying degrees, depending on the considered pathosystem, crop, and socioeconomic contexts, and this needs to be addressed as well. These three aspects, which lead to facets that expand beyond and around the frame of Figure 2.1, are discussed in this section.
Global Crop Health: Production Situations and Injury Profiles Crop losses occur within the fabric of interconnected components of agriculture: environmental, technological, economic, cultural, and social. These components both explain the occurrence of epidemics and crop losses and determine their consequences. These components can be embedded in the concept of “production situation,” which was defined as the biophysical and socioeconomic context in which agricultural production takes place, is consumed, or traded (Breman & De Wit, 1983; Penning de Vries & Van Laar,
19
1982; Rabbinge & De Wit, 1989; Savary et al, 2017b). A production situation may be simplified into a “crop profile” (Fig. 2.2) consisting of several components accounting for the physical, biological, economic, social, and technological environment of agriculture. In Figure 2.2, for instance, a typical production situation may be represented by a crop profile involving labor (L), water resource (W), fertilizer inputs (F), pesticide use (P), varieties and germplasm (V), and a level of technology (T). The Ya generated in the considered production situation depends on the various components of the crop profile. But crop health, represented by a given injury profile (IP), also depends on the crop profile. As a result, the level of yield losses, YL = Ya – Y, generated in a given production situation by a particular injury profile, depends on both the Ya made possible by the crop profile components and the levels of injuries associated with the injury profile. This approach thus brings to the forefront the Ya as an important marker of production situations as well as a determinant of YL. The nature of agroecosystems and their diversity in attributes and in characteristics are critical if crop health and crop losses are to be addressed globally (Savary et al., 2006; Zadoks & Schein, 1979). The flowchart in Figure 2.2 can be adapted to address different systems such as rice in tropical Asia, wheat in western Europe, coffee in Central America, or peanut in West Africa (Table 2.2) (Savary et al., 2017b). In each system, emphasis on the composition of the crop profile varies from massive labor inputs in tropical rice, high technology and cultivars in European wheat, weed management in African peanut, and combined inputs and cultivars in Mesoamerican coffee.
F I G . 2 . 2 . A systems framework for production situations, injury profiles, and yield losses. (Courtesy S. Savary and A. Ficke—© APS)
20
S E C T I O N I A Global Context
TA B L E 2 . 2 . Crop profile components, injury profiles, and examples of dynamics of interactions in four crop systemsa
System Irrigated and rainfed rice, tropical Asia
Main crop profile Main components b components of injury profiles L W F P V T
Bacterial blight, Xanthomonas oryzae pv. oryzae Stem rot, Magnaporthe salvinii Sheath blight, Thanatephorus cucumeris Brown spot, Cochliobolus miyabeanus Leaf blast and neck blast, Pyricularia grisea Rice tungro, Rice tungro bacilliform virus and Rice tungro spherical virus Planthoppers, especially Nilaparvata lugens and Sogatella furcifera Rice whorl maggot, Hydrellia philippina
Dynamics
References
Risks in rice crop health arise from individual diseases and their combinations in syndromes. Four drivers of agricultural change (labor, water, fertilizer, and land availability) translate into a few distinct crop profiles (production situations, PS). Risk factor analysis leads to four conclusions: (1) PSs, as wholes, represent very large risk factors for occurrence of disease syndromes; (2) PSs are strong risk factors for individual diseases; (3) drivers of agricultural change represent strong risk factors of disease syndromes; and (4) drivers of change taken individually represent small, but significant, risk factors for individual diseases.
Savary et al., 2000a, 2011a
Wheat health involves the fivedisease pathosystem: leaf and yellow rust, Fusarium head blight, powdery mildew, and Septoria tritici blotch. Three disease syndromes (two of which are dominated by leaf rust or Septoria tritici blotch) are identified and associated with climatic years, wheat cultivars, and agricultural (chemical) extensification. Climatic years, wheat cultivars, and crop management, in this decreasing order, define disease epidemic risks.
Daamen, 1981; Daamen et al., 1989; Savary et al., 2016a, 2016b
Leaf folder, Cnaphalocrocis medinalis Dead hearts and white heads (stem borers), several species, including Scirpophaga incertulas, S. innotata, Chilo suppressalis, Sesamia inferens Weed infestation, many species Wheat, western Europe (France)
F P V T
Brown rust, Puccinia triticina Yellow (stripe) rust, Puccinia striiformis Fusarium head blight, several species, including Gibberella zeae, Fusarium graminearum, G. avenacea, F. avenaceum, F. culmorum, and Microdochium nivale Powdery mildew, Blumeria graminis Septoria tritici blotch, Zymoseptoria tritici
(continued ) S. Savary and A. Ficke—© APS. = labor; W = water; F = fertilizer; P = pesticides; V = varieties and germplasm; and T = technology level (see text and Figure 2.2).
a Courtesy b L
C H A P T E R 2 Assessing the Global Impacts of Crop Pests and Diseases
21
TA B L E 2 . 2 . Continued Crop profile components, injury profiles, and examples of dynamics of interactions in four crop systemsa
System Coffee, Central America
Main crop profile Main components b components of injury profiles F P V T
Coffee leaf rust, Hemileia vastatrix Brown eye spot, Cercospora coffeicola Leaf miners, Leucoptera coffeella Phoma leaf blight, Phoma costaricensis Root-lesion nematodes, Pratylenchus coffeae Root-knot nematodes, Meloidogyne exigua Dieback, Colletotrichum spp.
Dynamics
References
Four groups of coffee-based agrosystems are identified, ranging from extensive (low-input, perennial polyculture) to intensive (unshaded high-input monoculture). In each group, drivers of coffee production vary as well the levels of injuries. Syndromes of crop health depend on production situations. Each coffee-based agroecosystem group corresponds to varying levels of pest and disease injuries, crop yield, and ecosystem service provision, aside from coffee production.
Allinne et al, 2016; Avelino et al., 2006, 2007
A path of agricultural intensification, involving (1) better weed control, (2) increased crop density, and (3) some mineral fertilizers, is tested in a network of field experiments. As production situations are improved, attainable yields increase. However, with increasing attainable yields, levels of rust and late leaf spot injuries increase too, leading to increased yield loss. Improvement of production situation beyond a threshold (attainable yield) of approximately 1 ton/ha must involve some disease management (via host plant resistance).
Savary & Zadoks, 1992a, 1992b
Thread blight, Corticium koleroga Ceratocystis canker, Ceratocystis fimbriata Peanut, West Africa
L F T
Groundnut rust, Puccinia arachidis Late leaf spot, Phaeoisariopsis personata Early leaf spot, Cercospora arachidicola Aspergillus damping-off, Aspergillus niger, A. flavus Botryodiplodia collar rot, Lasiodiplodia theobromae Sclerotium stem rot, Sclerotium rolfsii
Plant pathogens and pests are thus integral components—and not superimposed elements—of agroecosystems. As a result, any changes in the ways agroecosystems are managed are likely to influence diseases and pests and the crop losses they may cause. This is a founding principle of integrated pest management (Kogan, 1998; McRoberts et al., 2003; Teng & Savary, 1992). Table 2.2 provides summarized examples of such evolutions (Savary et al., 2017b).
Conversely, the impact of a particular disease may require the consideration of the simultaneous impacts of other diseases and pests in addition to the considered production situation and attainable yield (Esker et al., 2012). One may consider that, when several diseases or pests are at play in a given system, they interact via their damage mechanisms. In many cases, this interaction is a competition between yield-reducing organisms, leading to less-than-additive effects between pests
22
S E C T I O N I A Global Context
and diseases, as has been shown in potato in the United States (Johnson et al., 1986, 1987), in rice in Southeast Asia (Savary et al., 2000b), or in wheat in western Europe (Willocquet et al., 2008), for example.
The Variability of Crop Losses Generated by Emerging, Acute, and Chronic Epidemics A typology of epidemics has been proposed to describe the different patterns of plant disease epidemics over time and space (Savary et al., 2011b): (1) Chronic epidemics are generally mild and regularly occur over large areas; (2) acute epidemics occur infrequently, sometimes at a very high level of intensity but over comparatively restricted areas; and (3) emerging epidemics occur under exceptional conditions and may potentially affect very large areas. Brown spot of rice (Bipolaris oryzae) is a typical example of a disease causing chronic epidemics and losses in South Asia. Many viral diseases, such as rice tungro in Southeast Asia or maize streak in sub-Saharan Africa, are examples of acute epidemics. The wheat stem rust epidemic caused by the UG99 variants of Puccinia graminis f. sp. graminicola is a good example of an emerging epidemic. The typology of chronic, acute, and emerging epidemics is flexible. It allows for transition, that is, for inbetween classification. Potato late blight, for instance, a disease exhibiting acute epidemics in today’s western Europe, may also be seen as emerging when one refers to the epidemic of the nineteenth century in northwestern Europe (Zadoks, 2008) or to the European invasion of the new mating type of the pathogen. Similarly, stripe rust is chronic in northwestern Europe or in parts of North America and China; but the epidemics (e.g., Chen, 2005) in the U.S. Midwest, western Europe, central and southern Asia, China, and Australia may be looked upon as emerging, since they are associated with new forms of the pathogen. One may transfer the typology from epidemics to crop losses and consider (1) epidemics causing chronic crop losses, that is, crop losses occurring regularly over large areas at comparatively low levels; (2) epidemics causing acute crop losses, that is, crop losses occurring infrequently over small or comparatively restricted areas at sometimes very high levels; and (3) emerging epidemics, affecting potentially very large areas and potentially causing heavy crop losses. Great attention is paid to the acute and, even more so, to emerging epidemics (the Irish Potato Famine of the nineteenth century was indeed associated with an emerging
pathogen). This should not, however, distract attention from the importance of chronic plant diseases. The Great Bengal Famine (Chakrabarti, 2001; Padmanabhan, 1973), for instance, was associated with a chronic epidemic of rice brown spot. Another example of chronic plant diseases are the Fusarium diseases (Munkvold, 2003) of maize in subSaharan Africa, with their terrible consequences on human health (Dutton, 2009; Wild & Gong, 2010) because of the mycotoxins produced. Aside from significant quantitative crop (yield) losses, plant disease in this case leads to impacts in an altogether different dimension.
Disease and Pest Impacts on the Components of Food Security Plant diseases have played a role in history (Zadoks, 2008) and have, for instance, been associated with food shortages and famines. Historical causation of famines is an extremely complex field. However, there is little doubt that plant diseases may have contributed, sometimes critically, to the occurrence of famines and sometimes to their exacerbation (Savary et al., 2017a; Zadoks, 2008). Such impacts are necessarily associated with extreme epidemiological and crop loss events, that is, emerging or acute epidemics. Aside from massive losses associated with major economic, social, and human crises, plant diseases commonly also cause regular attrition of crop performance through chronic epidemics. Chronic epidemiological and crop loss patterns affect the overall sustainability of agricultural and social systems from ecological, social, and economic standpoints (Savary et al., 2011b). Crop performance is not measured only in terms of quantitative outputs (i.e., yield); crop production has nutritional, aesthetic, and cultural dimensions. Critical among these dimensions is the fulfillment of food security. Food security has been defined as “[a condition] when all people, at all times, have physical, social and economic access to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active and healthy life” (FAO, 1996). Consideration of the components of food security (Savary et al., 2017a) provides a useful framework to address some of the multiple impacts of plant diseases. These components are (1) availability and primary production; (2) availability and imports and stockpiles; (3) physical access and supply chain; (4) economic access; (5) stability of food availability; and (6) utility, safety, quality, and
C H A P T E R 2 Assessing the Global Impacts of Crop Pests and Diseases
23
TA B L E 2 . 3 . Examples of impacts of plant diseases on the components of food securitya,b Wheat rusts in world’s main Potato wheatlate blight, production Ireland, areas, 19th century today
Fusarium head blight (scab), U.S. Midwest, today
False smut of rice, Asia, today
Brown spot of rice, South Asia, today
Coffee rust, Central America, today
1. Availability. Primary production
++++
+++
++
++
+++
–
2. Availability. Import, stockpiles
+++
++
++
+
++
+
3. Access. Physical and supply chain
+
+
+
–
++
+
4. Access. Economical
++
+
+
+
++
+++
5. Stability of food availability
+
+
+
+
+++
–
6. Utility, safety, quality, nutritive value
–
–
++++
+++
++
–
Component of food security affected
a Adapted b ++++
from Savary et al., 2017a. Table courtesy S. Savary and A. Ficke—© APS. = Very large impact on the considered component; +++ = large impact; ++ = strongly affected; + = affected; and – = not affected.
nutritive value. All six components may be affected by plant diseases (Table 2.3) (Savary et al., 2017a). Plant diseases such as potato late blight, wheat rusts, and brown spot of rice can severely affect the first component (primary production), and they affect the second (imports and stockpiles). But some of these that greatly reduce availability (components 1 and 2; Table 2.3) also greatly affect other components, such as component 6 (food utility, safety, quality, and nutritive value), as in the case of rice brown spot. Other plant diseases are important not because of the losses they cause in production (component 1) but because of the reductions they cause in the quality of harvests (component 6), as exemplified by Fusarium head blight on small cereals, the Fusarium diseases of maize, or rice panicle diseases such as false smut. Plant diseases may also affect food security, even if the affected crop does not directly contribute to food provisioning. Plant diseases also are likely to affect food security in quite different ways, depending on the economic contexts of agriculture and of the consumers. Thus, the components of food security may be linked with the nature of the considered crop (i.e.,
food or non-food crop) and the structure of markets (i.e., weakly organized and local to strongly organized and global) (Savary et al, 2017a). This classification enables researchers to address the impacts of diseases on non-food crops, for instance, the impact of coffee rust on food security in Central America (Avelino et al., 2015).
■ From Local Impacts of Individual
Diseases to the Global Impacts on Crop Health
The baseline information—the individual datum— to analyze and model crop loss is obtained from measurements made on samples at the individual field scale. This information, however, when collected in a standardized manner from fields grown in a range of environments and analyzed with other information, can be used to generate estimates at larger scales. Such estimates can in turn inform strategic decisions for setting priorities in research and for policy making. A few