Climate Change And Plant Disease Risk

Water Freedom System

Survive Global Water Shortages

Get Instant Access

Karen A. Garrett, Ph.D.9 Kansas State University

Plant Disease and Ecosystem Services

One of the most important effects of plant disease is its impact on crop plant productivity. Oerke et al. (1994) estimated that damage by disease and insect pests resulted in a 42 percent loss in the eight most important food and cash crops. Pimentel et al. (2000) estimated that 65 percent of U.S. crop losses, $137 billion, were due to introduced pathogens. The effects of plant disease can also be considered within the broader context of ecosystem services, defined as the benefits provided to humans by ecosystems, including services provided by plants and their pathogens (Daily, 1997). Ecosystem services include the following: (1) provisioning services, such as the more obvious provisioning of food, fiber, fuel, and also the provisioning of genetic resources; (2) supporting services, such as

'Associate Professor, Department of Plant Pathology.


soil formation, nutrient cycling, and primary productivity by plants, all of which have great economic value but tend to be appreciated only when there are breakdowns, such as the loss of soil during the U.S. dust bowl; (3) regulating services, such as regulation of climate, disease, and insect herbivory, and water purification; and (4) cultural services, such as opportunities for education, recreation, tourism, and inspiration. The Millennium Ecosystem Assessment10 provides an example of system evaluation based on ecosystem services. Cheatham et al. (in revision) have synthesized perspectives on plant disease and its management in the context of ecosystem services. In addition to the direct effects of disease on crop production, disease and its management by increased tillage, pesticide use, and other methods may reduce services provided by plants such as soil formation and climate and water regulation. Disease may also remove plants that provide important cultural services in addition to the range of other potential services.

Some examples among the many notorious plant diseases illustrate the issues for disease management and the potential impact when diseases cannot be managed effectively. Chestnut blight has had one of the most definitive effects, essentially removing the once common American chestnut from the landscape of eastern North America (Anagnostakis, 2000). Potato late blight is infamous as the proximate cause of the Irish potato famine and continues as a major constraint to potato production, making the use of pesticides a typical part of potato management in many areas (Hijmans et al., 2000). Karnal bunt of wheat offers an example of a disease that does not cause major yield loss, but has an important economic impact on regions where it is present through limits on trade with Europe and other parts of the world where the pathogen has not been detected (Rush et al., 2005). Sudden oak death has changed the structure of some western U.S. forests and threatens to impact forests throughout a much wider area (Rizzo et al., 2005). Soybean rust is a new pathogen to the United States, with the potential to become established throughout much of the U.S. soybean production areas (Pivonia and Yang, 2004). Wheat stem rust was an important pathogen in the United States in the 1900s, motivating the removal of barberry plants that served as an alternate host and supported sexual reproduction of the pathogen. Disease resistance in U.S. wheat has been effective against this pathogen, but now new pathogen types for which this resistance is not useful have arisen in Africa and are likely to arrive in the United States in the near future (Stokstad, 2007).

The effects of climate on plant disease have been a direct object of study for decades. In contrast to many human diseases, the pathogens causing important plant diseases are often present on and around plants, ready to infect when environmental conditions become conducive. This has motivated the development of plant disease forecasting systems based on climatic variables such as

10The Millennium Ecosystem Assessment is an evaluation of the effects of ecosystem change on human well-being assembled from the work of more than 1,360 scientists (see http://www.


temperature and precipitation (De Wolf and Isard, 2007). Although such models are not new, the need to address climate change has placed new demands on these models, the research underpinning them, and policy drawing upon them. This paper begins with a brief introduction to the typical methods for managing disease. The observed and potential effects of climate change on plant disease are then reviewed, with an emphasis on biological thresholds and interactions that may lead to particularly large impacts from climate shifts. The paper concludes with a discussion of research and policy needs for plant disease management in response to climate change.

The Usual Challenges for Managing Plant Disease

Pesticides are a common tool for managing plant disease. For some plant diseases such as potato late blight, crop production without pesticides is currently impractical in many systems. In regions where education about pesticide safety is lacking, some farmers and their families experience chronic pesticide exposure. There are estimated to be between 1 and 5 million cases of pesticide poisoning each year, including many thousands of fatalities (UNEP, 2004). Shifts in pesticide use may thus result in shifts in unmanaged pesticide exposure, so that changes in demand for pesticides due to climate or other factors may have unexpected impacts on human health as well. Other disease management methods may be useful for specific diseases, such as removal of infected plant materials, introduction of biocontrol agents, management for disease-suppressive soils, or use of certified seed to avoid introduction of pathogens.

Deployment of disease resistance genes is often the most attractive option for disease management in agricultural systems. For some diseases, resistance offers completely effective management, whereas for others, effective resistance is not known although partial resistance may still be a useful management component. There is little cost from use of resistance genes to growers or consumers, except that in some cases it may be challenging for plant breeders to combine desired resistance genes with other desirable plant characteristics. Breeding crops for disease resistance also offers challenges in terms of identifying resistance that is durable. The deployment of resistance genes is much more efficient if the genes are useful against pathogen populations for long periods of time even if exposed to large pathogen populations under disease-conducive environmental conditions. Pathogen adaptation to overcome disease resistance is an ongoing problem for the management of many diseases (McDonald and Linde, 2002).

The use of cultivar mixtures is one method of resistance gene deployment that may increase the useful life of resistance genes in some cases. The management of rice blast in China offers a particularly dramatic example of the utility of mixtures for disease management, applied to over a million hectares. Higher-value susceptible rice varieties were grown in strips mixed with strips of lower-value resistant varieties. Both resistant and susceptible varieties experienced a


decrease in disease pressure compared to test plots where single varieties were grown for comparison (Zhu et al., 2000). In this case, it seems that microclimate was an important factor, such that the taller susceptible varieties experience relatively drier conditions when surrounded by the shorter resistant varieties (Zhu et al., 2005).

The fact that agriculturalists have the ability to manipulate crop plant genetics makes plant disease management in agriculture much easier, in some respects, than human disease management. Problems can also arise from this ability, however, as particularly successful crop varieties become widespread. Thus, a common challenge for plant disease management is the general homogeneity of cropping systems in the United States and trends toward greater crop homogeneity in most regions of the world. This homogeneity makes it easier for plant pathogens adapted to the common crop varieties to spread rapidly throughout crop plant populations. Margosian et al. (in revision) have evaluated the connectivity of the four major crop plants in the United States in terms of availability of the crop host species. The connectivity of a landscape for a particular organism, in this case a plant pathogen, is a measure of the ease with which the organism can move through the landscape. Maize and soybean are strongly connected throughout much of their range. Wheat and cotton production are more fragmented, so that pathogen populations cannot move as readily through all production areas. Conversion to biofuel production has the potential to increase crop homogeneity.

Maps of disease risk based on climate can be generated for diseases with reliable and widely applicable forecasting models. For example, Hijmans et al. (2000) mapped the risk of potato late blight based on climate parameters. Using updated forecasting models for potato late blight risk, Villanueva et al. (in preparation) estimated disease risk in the Altiplano region around Lake Titicaca (Figure 2-18). Such models are available for only very well-studied diseases, but Magarey et al. (2007) have developed a general model of infection risk for application in mapping the risk of new pathogens for which detailed models are not yet available. The combination of maps of current and future climatic conditions with models of pathogen risk can be adapted to evaluate changes in global risk in response to climate change. For example, Bergot et al. (2004) predicted the spread of the host-generalist pathogen Phytophthora cinnamomi in Europe.

Implications of Climate Change

Climate change will impact the productivity of agricultural and wildland plant populations through many mechanisms. One method for studying climate change effects on crop productivity is to study the correlation between climate variables and yield to date. Yield is the product of a number of factors, including losses to plant disease; partitioning the effects of these different factors will be necessary to develop a full understanding of the impacts of climate change.

FIGURE 2-18 Estimated potato late blight severity in the Altiplano area of Peru and Bolivia based on weather measures during 2001-2004 used in a late blight forecasting model. The comparable estimates for disease severity in 1995-1998 were for no late blight occurrence in the region. As temperatures increase in the region, the risk of potato late blight may be expected to increase at higher altitudes. Color indicates the level of estimated disease, ranging from green = low to red = high, with a corresponding range in the number of fungicide applications needed for successful potato production. This figure was translated into English from the original (in Spanish) by Mila Gonzalez. SOURCE: Map courtesy of H. Villanueva, R. Raymundo, H. Juarez, W. Perez, and G. Forbes, International Potato Center.

FIGURE 2-18 Estimated potato late blight severity in the Altiplano area of Peru and Bolivia based on weather measures during 2001-2004 used in a late blight forecasting model. The comparable estimates for disease severity in 1995-1998 were for no late blight occurrence in the region. As temperatures increase in the region, the risk of potato late blight may be expected to increase at higher altitudes. Color indicates the level of estimated disease, ranging from green = low to red = high, with a corresponding range in the number of fungicide applications needed for successful potato production. This figure was translated into English from the original (in Spanish) by Mila Gonzalez. SOURCE: Map courtesy of H. Villanueva, R. Raymundo, H. Juarez, W. Perez, and G. Forbes, International Potato Center.


The general trend toward warmer temperatures in recent years in many regions has facilitated study of the correlation between climate variables and yield. For example, rice productivity in the Philippines has shown a negative correlation with night temperature from 1979 to 2003 (Peng et al., 2004). Of course a challenge for such analyses is to account well for all of the other potential causal factors of the system that may vary along with climatic parameters. Lobell et al. (2008) have analyzed which regions of the world are most likely to be confronted with food security issues resulting from climate change, concluding that south Asia and southern Africa are particularly at risk. In another analysis of global agriculture, Cline (2007) points out that potential benefits to agriculture in some areas in the first decades of global temperature change may give the public a false sense of security and make it more difficult to put policies in place to avoid problems from more extreme changes in later decades.

A first step toward understanding wildland plant responses to climate change and the potential for adaptation to new climatic conditions is to address gene expression and underlying genetic diversity in wild plant populations. Travers et al. (2007) studied the effects of simulated precipitation change on big blue-stem, the dominant grass of tallgrass prairie of the U.S. Great Plains. Under the predicted future precipitation patterns with fewer and larger precipitation events leading to longer periods of drought stress, they observed lower expression of a gene associated with the hypersensitive response, a disease resistance reaction. Frank (2007) also studied big bluestem, finding higher infection rates and dampened phytohormonal responses to infection when plants experienced severe drought stress. Studying the diversity of resistance genes in wild plant populations is still challenging because little is known about them and for the moment there are few tools available. Rouse (2007) studied a gene in big bluestem that is related to genes conferring disease resistance in sorghum, finding evidence for historical disease patterns in natural populations that vary in diversity for these genes across a gradient of disease conduciveness.

The effectiveness of disease resistance genes may vary with climatic parameters. For example, Webb et al. (in preparation) found that rice genes conferring resistance to rice blast have different effectiveness depending on temperature. Most resistance genes tested were less effective at higher temperatures, but one of the most effective genes was actually more effective at 35-29°C day-night temperatures than at 29-21°C. These differential responses will influence the selection pressures experienced by pathogen populations as temperatures fluctuate annually and shift over years (Webb et al., in preparation).

Climatic changes and changes in CO2 concentrations can affect plant physiology, growth, and architecture in several ways that influence plant disease risk. On shorter time scales, stomatal closure in response to drought stress makes it more difficult for some pathogens to enter leaves. If plant canopies close earlier in the season due to changed conditions, the increased humidity in canopy microclimates may favor many pathogens. CO2 concentrations are expected to


impact pathogens directly as well, although a model for this impact is unlikely to be simple. For example, in a study of a set of fungal pathogens, Chakraborty et al. (2000) found that some species reproduced more rapidly under increased CO2, while other species reproduced more slowly.

In wildland systems, climate change and increased CO2 concentrations may also have mixed effects. Mitchell et al. (2003) found that the fungal pathogen load in tallgrass prairie increased overall in response to higher ambient CO2. In montaine prairie, Roy et al. (2004) found mixed effects of simulated temperature increases, with some pathogens increasing in abundance and others decreasing. Desprez-Loustau et al. (2007) predicted that the effect of climate change on a set of forest pathogens in Europe will be to increase favorability for the majority of pathogens. In general, rising temperatures may favor soil fungi that cause damping-off in seedlings, sometimes with high rates of mortality, a trend unlikely to be observed in the short term unless studies are designed specifically to look for such effects.

Range shifts in pathogens are frequently observed. As others have discussed at this workshop, such range shifts can be difficult to interpret. For example, needle blight is moving northward in North America as temperature and precipitation patterns shift (Woods et al., 2005). It is reasonable to think that such range shifts may be driven by changing climatic conditions, but the correlative nature of the data makes it impossible to determine this conclusively. Ultimately these relationships will have to be addressed in projects that combine the full range of factors in field studies as well as more limited and controlled experiments that allow clear conclusions about the effects of factors to partition effects.

The potential importance of extreme weather events is illustrated by the introduction of soybean rust to the United States. It is likely that spores of soybean rust entered the United States via Hurricane Ivan (Isard et al., 2005).11 If such extreme weather events become more common, global movement of pathogens will be accelerated. Soybean rust also offers an interesting example of the potential interactions between two invasive species. The widely introduced and problematic kudzu vine is another host of this pathogen and has the potential to play an important role as a pathogen reservoir during seasons when soybeans are not available for infection. Until now, however, movement of soybean rust has been slower than expected based on some predictions, probably due to environmental conditions that have not been conducive to disease. If the public becomes too complacent about the slower-than-predicted progress of soybean rust across the United States, this may result in more substantial problems if there is not support for needed research and if soybean growers do not prepare adequately.

The ultimate impact of changes in plant disease pressure, in either agricultural systems or wildland systems, will be determined in part by what plant

"Of course, spores of this pathogen may well have entered the United States previously but been unsuccessful in establishing infection. Entry of large numbers of spores may be necessary for an invasive pathogen to "beat the odds."


genotypes or species replace those that have experienced more damage by disease. Eviner and Likens (2008) summarize factors important for predicting the effects of disease on ecosystems, where one of the most important factors may be the functional similarities of infected host individuals versus the species that replace them. Through a broad ecosystem science lens, plant species may be generalized as composing "a single giant photosynthesizing leaf." From this standpoint, damage to one or a few plant species may not be important if other species can play the same role. In eastern U.S. forests, while other tree species increased in abundance to photosynthesize in the place of American chestnuts, they did not provide other important ecosystem services such as production of chestnuts as food for humans and wildlife. Likewise, most agricultural systems are not diverse enough to readily accommodate removal of an important species such as soybeans, if soybean production were to become uneconomical due to a new disease such as soybean rust.

Potential Interactions, Thresholds, and Positive Feedback Loops

If a small change in average temperature or precipitation patterns results in a small change in plant disease risk, this may be relatively easy to accommodate in agricultural disease management and may have little impact on wildland systems. Climate change is a greater concern when interactions serve to amplify the effects on biological systems or when systems are currently near thresholds such that small changes in abiotic drivers may push them beyond the threshold and thus have important effects. Effects may also be exacerbated if positive feedback loops are in place so that increased disease pressure further increases disease risk.

Abiotic environmental conditions are understood to be critically important in plant disease epidemiology, as commonly represented in the "plant disease triangle" (Figure 2-19). The three components of this triangle are a susceptible host, a virulent pathogen (and effective vector, as needed), and a conducive abiotic environment. For example, many fungal and oomycete pathogens benefit from higher levels of humidity. Surprising new disease problems may occur if the susceptible host and virulent pathogen have been present all along and the environment shifts to become more conducive. For example, potato late blight became an extreme problem for Irish food security during the potato famine when wetter years supported rapid disease development. The further interaction between high losses to disease and widespread reliance on potatoes as a primary food led to a disastrous situation.

Allee effects represent one type of threshold. An Allee effect occurs when a species experiences greater limitations on per capita reproduction for small population sizes. Quorum sensing provides an interesting potential mechanism for this type of phenomenon, where bacterial populations may become pathogenic only when intraspecific signaling indicates that a sufficiently large population is present for infection. Smaller population sizes may also make it less likely that


Was this article helpful?

0 0

Post a comment