Climate change – impact on crop growth and food production, and plant pathogens

Abstract

Climates are changing worldwide at rates not seen previously in geological time. This affects food production itself and the growth and reproduction of plant pathogens which reduce crop yield and quality. There is a need to develop an understanding of the implications and impacts of climate change on natural biodiversity, artificial landscapes as well as production agriculture (defined here as a generalization embracing all of the primary uses of land for agriculture, horticulture and forestry), since these form parts of an integrated continuum. Currently, 20–25% of harvested crops worldwide are lost to pre- and post-harvest diseases and climatic change is expected to increase these losses. Climatic change results in increasing variability and altered scales of temperature, rainfall and wind velocity and periodicity. These changes affect the activities and vigour of aerial and soilborne pathogens. Some pathogens capable of devastating crops and harvested produce have become more active and damaging because their geographical ranges expand as a consequence of climate change. Human populations are increasing rapidly, resulting in greater demands on all natural resources which far outstrip supplies. Natural biodiversity is being damaged, frequently beyond repair and not infrequently with little or no knowledge of the characteristics of the plant genotypes being lost. Only very recently have analyses of ecosystem services begun revealing the intricate and delicate webs of unappreciated natural assets which are vital for human sustainable (used here to describe processes and actions which balance resource-use with resource-availability and conserve economic, environmental and social welfare) survival. The combination of climatic change, expanding human demands for food, resources and space and the increased activities of plant pathogens presents pathologists with immense challenges. Previously, visionary plant pathologists have contributed hugely to solving humanity's problems. Current challenges compounded by climate change offer opportunities for the exploitation of our profession's capacities for working with ‘one foot in the furrow (and) one hand on the bench’ with increasing relevance for society at large.

Résumé

Le climat se modifie à un rythme jamais vu à l'échelle des temps géologiques, et ce, à la grandeur de la planète. Cette situation influence la production de nourriture ainsi que la croissance et la reproduction des phytopathogènes qui altèrent la qualité et le rendement des cultures. Il faut comprendre les conséquences et les répercussions des changements climatiques sur la biodiversité, les paysages artificiels ainsi que sur l'agriculture entendue ici au sens général du mot, englobant tous les principaux usages du sol pour l'agriculture, l'horticulture et la sylviculture, étant donné que ceux-ci font partie d'un ensemble intégré. Actuellement, de 20 à 25 % des récoltes à l'échelle de la planète sont perdues à cause de maladies qui surviennent avant ou après la récolte, et l'on s'attend à ce que les changements climatiques accroissent ces pertes. Ces derniers engendrent l'accroissement de la variabilité des températures, des précipitations ainsi que de la vitesse des vents et de leur périodicité. Ces changements influencent l'activité et la vigueur des agents pathogènes aéroportés et terricoles. Certains agents capables de dévaster des cultures et des récoltes sont devenus plus actifs et plus nuisibles à cause de l'expansion de leurs aires de répartition engendrée par le réchauffement climatique. De plus, les populations humaines, elles aussi, croissent rapidement, ce qui entraîne une plus forte demande sur toutes les ressources naturelles, demande qui excède l'offre de beaucoup. La biodiversité naturelle est compromise, parfois sans possibilité de récupération, souvent sans que nous connaissions les caractéristiques des génotypes de plantes à jamais perdus. Ce n'est que très récemment que des analyses des fonctions écologiques ont commencé à révéler la trame complexe et fragile du capital naturel incompris essentiel à la survie de l'Homme, utilisé ici au sens du mot pour décrire les processus et les actions qui maintiennent en équilibre consommation et disponibilité des ressources, et qui entretiennent le bien-être économique, environnemental et social. La combinaison des changements climatiques, de la demande croissante des humains pour la nourriture, les ressources et l'espace, de même que les activités accrues des phytopathogènes, constitue un défi considérable pour les pathologistes. Auparavant, des phytopathologistes visionnaires ont contribué activement à résoudre les problèmes de l'humanité. Les défis actuels, aggravés par les changements climatiques, offrent à notre profession des occasions de se démarquer, ‘un pied dans le labour, l'autre dans le labo’, et ce, en faisant preuve d'un intérêt croissant pour la population dans son ensemble.

Keywords:

• biodiversity
• climate change
• crop loss
• disease
• food security
• plant pathogens
• population
• resource supply
• sustainability

Mots clés:

• approvisionnement en ressources biodiversité
• changements climatiques
• durabilité
• maladies
• perte de récolte
• phytopathogènes
• population
• sécurité alimentaire

Introduction

The world's climates have changed over geological time ever since this planet became sufficiently stable to form an atmosphere. Current scientific and political concerns centre on the accelerating speed and scope of change compared with what has been experienced in previous geological eras and the implications for mankind's survival and that of all the other species which inhabit this planet with us. This review does not examine causative factors for climate change, since there are emotive overt and covert political debates surrounding the issues of anthropogenic versus alternative causes which might initiate climate change. Such debates only serve to cloud the scientific and social importance of climate change without contributing towards developing the means for its mitigation. Here, scientific biological and physical evidence for the occurrence of climate change is identified and aligned with its potential effects for the growth and reproduction of those microbes which destroy crops in the field, in post-harvest storage and the food distribution chains. Since plant pests and pathogens are currently responsible for upwards of 25% crop and post-harvest losses, any changes to their habitats which exacerbate damage and loss have very serious implications for security of the world's food supplies.

Ensuring a sufficiency of adequate, safe and reliable food supplies in the face of burgeoning demand is a huge practical challenge to the ingenuity and expertise of scientists. Adding further challenges due to escalating losses resulting from the increasing activities of microbes propelled by environments that are more favourable for their growth and reproduction simply adds to the scale of these challenges. Finding and developing solutions for these challenges is the biggest task that science has ever faced. These imperatives have been brought into even sharper focus recently by the scientifically valid correlation between civil conflict and global climate (Hsiang et al., 2011), with the occurrence of catastrophic natural events (Min et al., 2011; Pall et al., 2011) and increased frequency of tropical cyclones over the northern Indian Ocean linked with adverse effects for human health caused by aerosols composed of black carbon and sulphate (Evans et al., 2011).

Climatic change – biological evidence

Natural biological events are some of the most potent indicators of climate change. Phenological studies are now viewed as sources of primary data of great value in tracking the impact, scale and rate of climatic alteration. The UK possesses some of the most extensive records. Springtime events are traced in one of the longest running archives from a single location initiated by the landowner and naturalist Robert Marsham working from his farm at Stratton Strawless, Norfolk from 1736 to 1797 (Fiske, 2008; Spinks & Lines, 2008). Marsham recorded 27 signs of the initiation of spring during this period, from the first snowdrop (Galanthus nivalis) flowering to the first cuckoo (Cuculus canorus) singing, and noted how timing was affected by weather. A cold period delayed springtime events and warm conditions speeded them up. Leafing by the mountain ash (Sorbus intermedia) varied from 5 March to 2 May conditioned by the prevailing weather. In 1740, after a fearful winter, spring was badly delayed and it was March before Marsham heard a song thrush (Turdus philomelos) sing and hawthorn (Crataegus monogyna) bushes did not bloom until June, two months later than usual. Marsham's recording was continued by his descendants until 1958, providing two centuries of data. In Scotland, there is a similarly substantial body of information dating back to the early 1700s, illustrating the effects of climate change (Last et al., 2003). Climate change disturbs the synchrony between temperature and photoperiod, and because insects and pathogens show individual patterns of response to temperature, carbon dioxide and photoperiod, the result is a loss of evolved phasing. This changes the relationships between plants and the environment. The closeness of this relationship was demonstrated for woody perennials such as Rhododendron spp. and cultivars by Dixon & Biggs (1996) for acclimation to temperature change following the cessation of dormancy. More recently, data from Salisbury, Wiltshire containing 64 000 records dating from 1950 identified that the growing season for many fungi has increased as ambient temperatures have risen (Gange et al., 2007). Over this 50-year period, fungi doubled the length of their reproductive season from 33 days to 74 days. Some fungi are now reproducing twice per year instead of once. It is suggested that changes seen in the growing season of fungi are the most substantial for any group of organisms on Earth (Gange et al., 2007).

The first formal and scientifically validated link between observed global changes in physical and biological systems and human-induced climate change predominantly from increasing concentrations of greenhouse gases was demonstrated by Rosenzweig et al. (2008). They surveyed 29 500 data series of which 90% (P << 0.001) demonstrated that changes at the global scale were in the direction that would be expected as responses to global warming. In biological systems, 90% of the datasets showed that plants and animals were responding consistently to temperature change. This is predominantly illustrated by phenological change with earlier blooming, leaf unfolding and spring arrivals. Events on the current scale have not happened on Earth in the past three-quarters of a million years (King, 2005). In addition, previously no one single living species has had full control of the Earth's entire resources and reproduced itself in unprecedented numbers at a very rapid rate. Humans now have these abilities and, as a result, the Earth's resources are in imminent danger of exhaustion and the environment is changing in a manner that exacerbates this process. Dangers threatening the natural world are highlighted by a recent analysis of 76 634 separate French forest surveys made over the period 1965–2008, which warns of the dangers of habitat fragmentation in lowland forests resulting from increasing temperatures (Bertrand et al., 2011) whereas previously it was thought that woodlands at higher altitudes were the most vulnerable.

Climatic change – physical evidence

Scientific interest in climatic change was probably initiated by the Swedish physical chemist Arrhenius (1896), who estimated that doubling the carbon dioxide content of the atmosphere could increase the Earth's temperature by 5 °C. Later, he refined his calculation to a 2.1 °C rise. Longstanding physical meteorological records are contained in The Central England Temperature Archive (Anon., 2011a ) dating back to 1659. This is the world's oldest continuous weather record. It applies to an area within the triangle formed between the towns of Lancaster, Bristol and the outskirts of London and was compiled by Gordon Manley (1902–1980), a Durham University meteorologist. He meticulously worked through thousands of documents left by enthusiastic amateurs such as landed gentry and vicars who made daily recordings between the 17th and 20th centuries. Manley spent 30 years working through their journals, cross-checking and calibrating thermometer readings until he achieved a standardized record. The Meteorological Office at Exeter now keeps these records up to date. They demonstrate how climate has changed over this extended period. Further evidence for change comes from the analysis of long-term (1957–2006) datasets for Antarctic surface temperatures. Overall, that continent is warming at about 0.1 °C per decade with the most obvious warming taking place in winter and spring and over West Antarctica (Steig et al., 2009) and with similar happenings in the Arctic (Anon., 2009a ).

Based on subsequent predictions of the Intergovernmental Panel on Climate Change (IPCC) (Watson et al., 1998, Giorgi et al., 2001), the economist Stern (2008) defined the current situation in the following terms – ‘If no action is taken to reduce emissions, the concentration of greenhouse gases in the atmosphere could reach double its pre-industrial revolution levels as early as 2035, virtually committing us to a global average temperature rise of over 2 °C. In the longer term, there would be more than a 50% chance that temperature rise would exceed 5 °C.’

Agricultural implications

Some 1.5 billion ha of land is used worldwide for crop production and of this 960 million are in developing countries (Fischer et al., 2002). In the last 30 years, the world's cropped area has expanded by approximately 5 million ha annually, with Latin American countries accounting for 35% of this increase through deforestation, with substantial deleterious impacts, for example on the Amazon Basin (Tollefson, 2011). Land is a primary resource that cannot be created. There is, therefore, a finite amount beyond which the cropped area cannot be increased. About 40% of the world's arable land is now degraded to some extent and most of that land is in the poorer nations in densely populated, rain-fed farming areas where overgrazing, deforestation and inappropriate land-use compound other problems. About 3 billion ha (one-fifth of the world's land surface) sustains forest ecosystems. Countries such as Russia, Brazil, Canada, USA, China, Australia, Congo and Indonesia account for 60% of the world's forest land. In the 1990s, 127 million ha of forests were cleared and only 36 million ha replanted. Africa has lost 53 million ha of forest mainly converted into fragile cropped land (Tollefson, 2011).

Two-thirds of the world's population live in areas receiving 25% of the annual world rainfall. About 70% of the world's fresh water is used for agriculture and that figure rises to 90% in nations relying on extensive irrigation. Currently, 30 developing nations face water shortages and by 2050, this could increase to 50 nations mostly in the ‘developing country’ category. Water scarcity and the degradation of arable crop land are the most serious obstacles inhibiting future increases in food production (Dixon, 2009a ).

Under-developed regions

Against this backdrop, Smith & Almaraz (2004) have summarized the dangers of climate change to crop production. Extremes in temperature are dangerous to crop production especially where growth is accelerated due to additional atmospheric carbon dioxide. More northerly zones may become wetter and warmer, which could benefit crop production in the short term, but the tropics and subtropics could become hotter and drier. Calculations based on three out of four of the Climate Change Models show consistent expansion in the areas of arid land in developing countries. Africa is thought to be the region most vulnerable to negative impacts of climate change on crop production (Challinor et al., 2007).

Currently, 1.08 billion ha of land in Africa has a growing period of less than 120 days. With climate change by the 2080s, this can expand by 5–8% (equal to 58–92 million ha). This change would be accompanied by a loss of 31–51 million ha of land in favourable growing zones with growing period lengths of 120 to 270 days per year. About 1 billion people worldwide, and of that 180 million in Africa, live in vulnerable zones currently relying on agriculture for their existence. By the 2080s, land areas with increasingly severe constraints for crop production in the world zones will be Central America and the Caribbean, Oceania and Polynesia, northern Africa and west Asia. In southern Africa, an extra 11% of land could suffer severe constraints to cropping. By the 2080s, decreases in potentially good agricultural land are predicted to occur in northern Europe (with the UK and Ireland particularly affected); southern Europe (especially Spain); northern Africa (especially Algeria, Morocco and Tunisia); southern Africa (especially South Africa); and in east Asia (especially China and Japan). Countries such as Venezuela, New Zealand, Mozambique, Sudan and Uganda are individually nations with good agricultural land that are especially vulnerable (Dixon, 2009a ).

Food insecure regions are identified by Lobell et al. (2008) as South Asia, China, southeast Asia, east Africa, central Africa, southern Africa, west Africa, Central America and Caribbean, Sahel, west Asia, Andean Region and Brazil, since they contain a notable share of the world's currently malnourished populations. Across all these regions, the wheat crop appears to be most vulnerable to damage from climate change. In their study, Lobell et al. (2008) used statistical crop models and climate change projections for 2030 from 20 general circulation models. Two regions, south Asia and southern Africa, are predicted to suffer the worst effects of climate change on their agricultural productivity, with substantial effects on their populations. Currently, however, uncertainties vary widely by crop and hence priorities for adaptation depend on the risk attitudes of the investment providers.

In broad-level analyses of Chinese agriculture, Liu et al. (2004) used country-level cross-sectional data on agricultural net revenue, climate and other economic and geographical data from 1275 agriculturally dominated counties. Under most climate change models, higher temperatures and more precipitation would have an overall positive impact on China's agricultural output. But these impacts vary greatly seasonally and regionally. A realistic study comes from Russia, where it is suggested that a shortage of water for irrigation may override any advantages accrued from temperature increases and the availability of high-grade soils for grain and other crop production. As a consequence, Dronin & Kirilenko (2008) analyzed strategies for food security based on previous agrarian systems used at periods in Russian agricultural history, viz. free-market, big commune-war communism, developed socialism and fortress-market employed to provide interregional food exchange. They deliberately omitted the strategy of compensating for shortfalls in food by substituting with imports. The free-market model outperformed the others but the fortress-market also succeeded as no regions were threatened by grain shortages. Several adaptation measures are identified, such as moving meat production northwards and the exploitation of genetically modified cultivars. The authors note that increased irrigation could mitigate some effects of climate change especially in Southern Russia. But they admit that water supplies will become severely restricted and hence this should not be seen as a route for adaptation.

Developed world

North America is the largest producer of agricultural exports to the rest of the world and therefore climate changes there can have implications for everyone else. Reviewing climate change effects in detail, Changnon & Hollinger (2003) studied the production of corn (maize) in the Midwestern USA. There appears to be a potential for up to 40% increases in rainfall here since there has been steadily increasing rainfall over the past 50 years in Midwest USA. But this translated into little additional yield unless the rainfall coincided with the drought-stressed summer period. The impact of increased soil moisture for good or ill depends on timing and season. Using two climate-change models, the UK Hadley Centre for Climate Prediction and Research model and the Canadian Centre for Climate Modelling and Analysis model for studies of wheat production in the Great Plains region of the USA, Weiss et al. (2003) concluded that yield and percentage kernel nitrogen could not be maintained at even current levels especially in the arid part of Nebraska. This translates into a loss of quality in the flour required for breadmaking (Blumenthal et al., 1996). These authors identified needs for new cultivars to increase nitrogen uptake and translocation, as simply adding extra quantities of nitrogen fertilizer is not an agronomically, economically or environmentally sensible answer.

A broad-scale review of major crops by Chen et al. (2004) identified, as might be expected, that climate-change effects varied for different crops in the USA. For corn (maize, Z. mays) precipitation and temperature have opposing effects on yield levels and variability – increased rainfall raises yield and decreases variance. Temperature increases have the reverse effect. For sorghum (Sorghum vulgaris), higher temperatures reduced yields and yield variability. Increased rainfall raised sorghum yields and its variability. The authors also used the Hadley and Canadian climate change models, and these indicated that future variability decreased for corn (Z. mays) and cotton (Gossypium spp.) but increased for soybean (G. max). Increased variability equates with unreliability in harvest volume which is an unwelcome outcome for all sections of the food chain from field to plate. Drought is predicted to be one of the main consequences of climate change in the USA (Anon, 2009b ). According to the US Global Change Research Program, ‘warming over the mid-latitude land-masses such as continental United States is predicted to be higher than the forecast average global warming. Much of the inland USA faces a rise of between 5 °C and 6 °C on the current emissions path (that is, ‘business as usual’) by the century's end, with a substantial fraction of that warming occurring by mid-century’; severe drought could become a biennial event in the USA with a dust bowl stretching from Kansas to California (Romm, 2011).

Further north, evaluations have been made of the impact of climate change on spring wheat, maize, soybean and potato crops in seven agricultural regions of southern Quebec. These indicated an increase in carbon dioxide and temperature with resultant acceleration in crop maturation encouraged by reduced soil moisture availability. Adaptive moves would be needed to cope with these negative effects caused by climate change (Brassard & Singh, 2008). A similar conclusion comes from studies of wheat production in parts of south Australia which will cease to be economically viable (Luo et al., 2007) based on critical yield thresholds without mitigation. Farmers' adaptive options depend on innovations in agricultural science and technology, including genetic alteration and traditional plant breeding integrated with sound sustainable husbandry.

The potential impact of climate change on European agriculture has received considerable attention. As in the wider world, there are some initial beneficiaries, particularly in more northerly areas. In northern Europe, yields are expected to increase as new crops and cultivars emerge in the short to medium term. For example, analyses indicate that in the short term, German farmers may benefit from climate change, with maximum gains where the temperature increase is +0.6 °C. In the longer term, however, there may be losses (Lang, 2007). This work is based on theoretical modelling which is unable to take all variables into account. Warming decreases the crop growing period which reduces yields, but increased precipitation linked to higher temperatures and carbon dioxide concentration raises yield for crops such as winter wheat and soybean (Alexandrov et al., 2002). Spring barley (Hordeum vulgare) is the most important cereal crop in Central and Western Europe (Trnka et al., 2004) because of its use for animal feed. In the Czech Republic, soil water content increases with climatic change. This is a key factor in determining yield. Yields can increase by 54–101 kg^.ha⁻¹ per 1% increase in available soil water content on sowing day. Doubling of carbon dioxide concentration increased yield by 13–52% and opportunities for earlier sowing further enhanced yield (Trnka et al., 2004). This makes areas of western central Europe attractive for cereal production in the short to medium term.

Adverse effects can be expected in southern Europe, where anticipated water shortages may reduce yield but farmers could adapt their husbandry to prevailing conditions aided by technological progress. Adaptation needs to be quantified and built into simulation models to determine the impact of climate change (Reidsma, 2007). The variability of sugar beet (Beta vulgaris) yield (measured as coefficient of variation) (Jones et al., 2003) can increase by 50% (from 10 to 15%) compared with 1961–1990 figures, with serious implications for commercial planning in the beet sugar industry. Climate change is expected to result in yield increases of around 1 T^.ha⁻¹ of sugar in northern Europe and comparable losses in yield in France, Belgium and west-central Poland over 2021 to 2050. These figures mask significant increases in yield potential due to earlier spring conditions and accelerated growth probably offset by losses due to drought stress during the taproot filling growth stages. The effects of carbon dioxide concentration on biomass production are approximately linear from 360 to 700 ppm CO₂ (Demmers-Derks et al., 1998).

Areas with existing drought problems will suffer from a doubling of losses and there may be serious new problems in north-eastern France and Belgium. Overall, west and central Europe will potentially see losses from drought rise from 7% (1961 to 1990) to 18% (2021–2050). In Spain, High Resolution Climate Models (HRCMs) were used to study potential yields and showed crop failures of winter wheat in the south but yield increases for spring wheat in northern and high altitude areas (Minguez et al., 2007). In Turkey, a study by Umetsu et al. (2007) considered the Lower Seyhan Irrigation Project using an expected value–variance (E–V) model. Under water constraints, farmers chose to grow high-value crops such as watermelon (Citrullus vulgaris), citrus (Citrus spp.), cotton (Gossypium spp.), fruits and vegetables. But because of the increased cost of water, gross revenues fell. In Israel, adaptation by increasing irrigation and nitrogen use were advocated by Haim et al. (2008) as a means of mitigating the adverse implications of climate change by 2070 to 2100 for wheat and cotton production. Since water supplies for the entire Middle Eastern region will be at a considerable premium by the latter part of this century, advocating such strategies for adaptation may in reality not be feasible or equitable.

Wine production is a good example of a worldwide product where clear geographical differences in advantage or disadvantage emerge from climate modelling. Changes in currently cool climate areas, such as the Mosel Valley, Alsace, Champagne and the Rhine Valley, could lead to more consistent vintage quality and potentially enhance the ripening of warmer climate-adapted cultivars (Jones et al., 2005). But in those regions where currently grape (Vitis vinifera) cultivars are grown close to the climatic optimum, for example southern California, southern Portugal, the Barossa Valley and the Hunter Valley, these regions may become too hot for high-quality wine production. Winter temperature changes would also affect viticulture by making regions that experience hard winter frosts (e.g. Mosel Valley, Alsace and Washington) less prone to vine damage, while other regions (e.g. California and Australia) may have such mild winters that latent bud hardening may not be achieved and cold-limited pests and pathogens may increase in both number and severity as a result of warmer winter conditions.

Historical analyses such as that conducted by Therrell et al. (2006) for maize yield over the period 1474 to 2001 demonstrated a close link between food supply and climatic change. The implications of changing climate have been recognized scientifically for well over a century (Johnson & Smith, 1965; Smith, 1965; Froud-Williams et al., 1996; Simons, 2007). Change may be beneficial, at least in the short term, as demonstrated by Magrin et al. (2006) who showed that Argentinian yields of wheat (T. aestivum), maize (Zea mays), sunflower (Helianthus annus) and soybean (Glycine max) have recently benefited from increased precipitation, decreased maximum and increased minimum temperatures presumably due to climatic changes.

Some economists make the assumption that by 2080, consumers will be much richer than today and separated even more from agricultural production processes, earning their income in largely urban non-agricultural occupations. Hence, they postulate that changes in consumption will depend more on food prices and on income differences than on local agricultural production. They suggest further that the share of undernourished in the world total population would fall below 20% when an arbitrary index of 130 is reached, whereby aggregate food supply exceeds aggregate food requirements by 30%. Hunger is completely eliminated where this index reaches 170. Fischer et al. (2002) postulate that ‘the trade system will (only) mitigate local climate-change impacts when consumers can afford to buy food on the international market – (but) food prices rising due to climate change may put an extra burden on those consumers who depend on imports, even without a region experiencing direct local climate-change impacts on production conditions’ (my parentheses). The economic and climate change models give starkly different prospective outcomes for 2080. Either ‘climate change impacts on agriculture will increase the number of people at risk of hunger’ or ‘with rapid economic growth and a transition to stable population levels, poverty, and with it hunger though negatively affected by climate change, would become a much less prevalent phenomenon than it is today’ (my italics).

As a general conclusion, where world climate gets warmer and where temperature rise is extreme, this could be detrimental to both human health and to food supply. The spring, summer and autumn seasons become longer and this effect is more dramatic at higher latitudes. In these areas, the climate becomes drier, for example in parts of Canada, and this could reduce the area available for hard red spring wheat production. In Quebec, fruit trees would move northwards and reduced snow cover would make it difficult for forage legumes to survive in winter. Glaciers have retreated around the world by up to 30% in the 20th century. The result is less water flowing through rivers, hence reduced amounts available for irrigation. Extreme weather events are becoming more common, increasing droughts and tropical storms are making crop production more difficult. The incidence of extreme weather events is now linked with climatic change (Schiermeier, 2011). Higher temperatures accelerate the breakdown of soil organic matter. Less organic matter means lower yields because of a lack of nutrients and water. Lower soil fertility reduces the availability of microbes antagonistic to plant pathogens. As natural forms of biological control diminish, the incidence of pathogens and the diseases that they incite can increase. As a counterweight increased carbon dioxide concentration raises soil organic matter content resulting in greater microbial activity. Increased carbon dioxide means more photosynthate for nitrogen fixation encouraging the growth of C₃ plants. But soil erosion increases as a result of more severe wind events. Rising sea levels also mean that adjacent land becomes more saline.

Effects on crop growth

Blackman's Principle of Limiting Factors (Blackman, 1905) states that ‘when a (biological) process is conditioned as to its rapidity by a number of separate factors, the rate of the process is limited by the pace of the slowest factor’ – this applies equally now as it did a century ago. Blackman's ‘slowest’ factor may now be interpreted as one which causes greatest stress and therefore inhibits plant growth and reproduction. Hence, although for example, rising carbon dioxide concentration in the atmosphere may within certain limits increase the rate of photosynthesis, this is counterbalanced by rising temperature. The basic principles of plant physiology which are likely to govern responses to climate change are understood in considerable detail (Grace & Zhang, 2006; Morison & Moorecroft, 2006; Morecroft & Keith, 2009). Despite some geographical zones of advantage, reduced transpiration and resultant higher temperatures in plant foliage leads to accelerated tissue senescence. Whether these effects are beneficial or not depends on the extent to which temperatures rise and exceed the optimum for efficient photosynthesis. Overall, the data suggest that elevated carbon dioxide may have positive benefits for C₃ plants, including yield stimulation, improved resource-use efficiency, more successful competition with C₄ weeds, less damage from ozone toxicity and in some cases greater pest and pathogen resistance (Fuhrer, 2003). Although warming accelerates plant development, it reduces grain filling, limits nutrient-use efficiency, increases water consumption and favours C₄ weeds as compared with C₃ crop plants. Changes in the water balance and amount of water available in the soil are crucial for crop growth. In grasslands, 90% of the variance in primary production can be accounted for by annual precipitation (Campbell et al., 1997). Calculations using the Penman–Monteith equation predict that potential evaporation increases by about 2–3% for each 1 °C rise in temperature (Lockwood, 1999). While biomass and yield may increase with rising carbon dioxide concentrations, patterns of dry matter allocation to roots, shoots and leaves also changes. Root to shoot ratios increase with elevated carbon dioxide, favouring root and tuber crops. Conversely, rising temperature and reduced transpiration limit biomass and seed production drops. Non-structural carbohydrate levels increase but protein and mineral nutrient contents fall; hence, food quality declines both for herbivores and for humans (Dixon, 2007).

Plant pathology

Possibly one of the best definitions of ‘plant pathology’ is that of Professor R.K.S. Wood (Imperial College, London) (Wood, 1967): ‘a science rooted in the practice of learning about the diseases of economically important plants in order to reduce losses of crops caused by them’. This echoes Professor J. C. Walker's (University of Wisconsin, Madison) (Williams & Marosy, 1985) earlier contention of an academic discipline firmly rooted with ‘one foot in the furrow’. Both recognized the need for a combination of knowledge of the disease causing agents, diseases themselves, an understanding of crop production and of the environments experienced by both the pathogen and the associated crop(s) over time. The immense problems associated with the interaction between climate change, plant disease and food (in)security demands the combined skills of laboratory-based molecular pathologists and field-based practitioners (Andrews, 2010).

Approximately 25% of crop production is lost to the ravages of pests and pathogens between the field and consumer's plate (Main, 1977; James & Teng, 1979). Climate change will alter phasing of lifecycle stages and their rates of development for pests and pathogens and associated antagonistic organisms (Chakraborty, 2011). It may modify mechanisms of host resistance and host–pathogen relationships. The geographical distribution of hosts and pathogens will alter. The level of crop losses will increase while the efficacy of control measures (Coakley et al., 1999) could fall when faced with greater populations of pests and pathogens. Increased fecundity of fungi results from elevated carbon dioxide concentrations and higher temperatures. Increased rainfall events would reduce weather-windows for spray application and allow greater likelihood of contact sprays being washed off from canopies. Raised carbon dioxide could increase the thickness of epicuticular waxes resulting in slower penetration of pesticides. Rising temperatures could increase the range of pathogens as suggested for Phytophthora cinnamomi by Brasier (1996). Similarly, increased spread is likely for rice blast (Magnaporthe grisea), wheat scab (Fusarium spp.), stripe rust (Puccinia striiformis) and powdery mildew (Blumeria graminis). Boag et al. (1991) estimated that each 1 °C rise in temperature would allow soil-borne nematodes to migrate northwards by 160–200 km. Recent epidemics of wheat stripe rust (yellow rust) (Puccinia striiformis f. sp. tritici) appear to result from an increase in prevalence of strains adapted to warmer temperatures and capable of overcoming the long-standing resistance genes Yr8 and Yr9 as determined by physiological race surveys conducted in the western states of the USA (Garrett et al., 2009).

The effects of climate change on pest and pathogen outbreaks are already being seen in the UK and western Europe. For example, insect pests such as diamondback moth (Plutella xylostella), pathogens like bacterial black rot (Xanthomonas campestris pv. campestris) and various Phytophthora spp., have become well established, causing damage to a wide range of ornamentals, field brassicas, and forests and landscape plantings, respectively. It is expected that forest pathogens will be directly affected by climate (Sturrock et al., 2011). For example, Phytophthora root rot (P. cinnamomi) is present in most temperate and subtropical areas, affecting in excess of 1000 plant species. Temperature, moisture and pH each affect pathogen growth and reproduction. Over the past 60 years, this pathogen has been favoured and spread especially in beech (Fagus spp.) and other susceptible hosts, including oak (Quercus spp.), alder (Alnus spp.), maple (Acer spp.), fir (Abies spp.) and pine (Pinus spp.). Temperature rises will allow host range increases. The CLIMEX models suggest increased spread in the UK, coastal Europe and trans-continentally and reductions in the tropics and subtropics (Brasier & Scott, 1994; Brasier, 1996). Similarly, sudden oak death (Ramorum blight, Phytophthora ramorum), is an important invasive microbe which appears to have migrated from the USA into Europe, affecting a wide range of ornamental trees and shrubs and forest species (Webber, 2008); it has recently invaded larch stands (Webber et al., 2010). CLIMEX suggests it will reduce in significance in eastern USA but increase in west coast states like Washington, Oregon and California (Venette & Cohen, 2006; Venette, 2009).

Overall, while temperature increases would significantly raise the severity and spread of plant diseases, precipitation could act as a regulator (Madden et al., 2007). Climate change models are not yet sufficiently sensitive or detailed enough to incorporate estimations of their impact on microbial activity. Extreme weather events, such as excessive rainfall and consequent flooding, are most likely to worsen the incidence of crop pathogens. A major effect of climate warming in temperate zones could be increased winter survival of pests and pathogens and subsequent spring-time damage. Nutrient acquisition is closely associated with overall plant biomass and is strongly influenced by the available root surface area. When climate change alters root exploration in the soil, a restriction of nutrient acquisition follows, leading to stress and reduced growth. Nutrient replacement management will be required where crop spectra change following the effects of rising temperature and increased carbon dioxide availability (Brouder & Volenec, 2007). This will affect disease susceptibility. Spring winds cause substantial damage to woody perennials especially fruit and landscape trees. The damage may not be apparent in the year of the event. With large trees, root damage may take at least one season to cause an effect and then could lead to foliar chlorosis and die-back. This is frequently followed by pest and pathogen invasion which compounds the damage.

The key food crop for at least half of the world's population is rice (Oryza sativa). Reliance is greatest in under-developed and developing nations. Studies of the rice cultivar IR36 simulating yield changes with increasing carbon dioxide levels and rising temperatures have been made using the INFOCROP model for the Tamil Nadu region of India. Crop duration, days to anthesis, leaf area index and dry matter percentage (DMP) all fell, resulting in lower grain yield per m². It is concluded that crop husbandry will need to improve substantially (Srivani et al., 2007) in order to offer any chance of sustaining the food supply. Bangladesh is also a region which is highly vulnerable to the impact of climate change and requires adaptive strategies for risk reduction. Suggestions are that greater use could be made of local plants such as Jatropa curca and Simmondsia chinensis as supplies of biofuel extracted from these oilseeds (Bowe, 2007).

Case study – clubroot caused by Plasmodiphora brassicae Wor

One of the most marked changes in host–pathogen interaction is found with the increasing severity of clubroot disease affecting oilseed rape (canola) in the UK. Plasmodiophora brassicae, the causal agent of clubroot disease in all members of the family Brassicaceae, presents a useful model system from which the implications of climate change on disease incidence and severity might be assessed. This microbe is soilborne and survives over long periods as very robust, well-protected soilborne resting spores. Germination is at least partially triggered by the presence of hexasaccharide root exudate signals from host plants (Mattey et al., unpublished). Thereafter, the soil environment influences survival and the processes of host invasion (Dixon, 2009b ). Key environmental factors which influence the success of P. brassicae as a pathogen are soil moisture and temperature, pH and soil ionic content. This pathogen appears to be immensely capable of developing physiological strains fit to succeed when agronomic systems alter, as in the evolution of virulent strains which are compatible with resistant cultivars or where the crop husbandry environment changes. This has been seen in the UK in relation to disease development on autumn-sown oilseed rape (Brassica napus).

Traditionally, autumn-sown rape dominates the British crop with a very limited area of spring sown crops. The reasons for this are that autumn sowing fits well with British cereal rotations and the crop suffers only mildly from clubroot. On a few occasions when European Union Common Agricultural Policy (EU-CAP) payments favoured the use of spring rape, the crops were devastated by clubroot, mirroring regular events in mainland Europe. The reason why autumn sown crops escaped clubroot disease development was explained by the late C. Williamson (The Hutton Institute, Dundee, Scotland, pers. comm.) who showed that soils at the time when the crop germinated and formed a rosette stage were cooling and P. brassicae ceased to be active. Between February to early March when temperatures rose sufficiently for the pathogen to resume activity, the host plants were growing already and had formed their components of yield. This resulted in production of profitable crops despite them being grown on infested land. This situation has now changed. Drilling of autumn-sown oilseed rape has now advanced such that it is really a late-summer (late July to early August) activity. The result is that the crop is planted in soil where the temperature is still high enough to allow P. brassicae activity. Additionally, autumn has extended at least to the end of November with the retention of higher soil temperatures accompanied by rain. This fulfils predictions for climate change in Britain that there will be longer, warmer and wetter autumns. As a result, clubroot has now become a major and dominating pathogen of autumn-sown oilseed rape. Severe regular disease outbreaks are now recorded as far north as Aberdeenshire and Morayshire in northern Scotland. Here, 75% of the land is infested with clubroot as a result of past generations of swede (B. napus) production for use as overwintering in situ sheep forage. As a result of changing agronomic and environmental conditions favouring P. brassicae, substantial crop losses are happening, consequently reducing yields and quality of home-produced rape oil (Dixon, 2006).

Environment and plant disease

The factors governing disease outbreaks may be summarized by a simple linear equation as:

Climate influences pathogen (P_E) and host environments (H_E) separately and collectively throughout the period of crop germination, growth and reproduction (Time, T) from infection to host death. This influence will not be static and can vary throughout the lifecycles of both host and pathogen. Those aspects of the biology of each organism most affected by environment can be differentiated and analyzed.

Host effects

The level of resistance to particular pathogens varies with age of the host – juvenile foliage may be more susceptible than adult and vice versa. For example, early season and juvenile brassica foliage is seldom infected by powdery mildew (Erysiphe cruciferarum) whereas maturing summer and autumn foliage becomes devastated (Dixon, 1981, 1984). Density and distribution of the crop influence the likelihood of disease. Plants growing on limited areas and in scattered locations are less likely to be infected compared with those grown in large areas of monoculture. The practice of isolating seed crops such as sugar beet (Beta vulgaris) or potatoes (Solanum tuberosum) away from major areas of ware production greatly reduces the potential for infection. The physiological condition of the host has a marked effect on the incidence of disease. Lush unthickened growth encouraged by the use of excessive nitrogen fertilization may be more easily penetrated than tissues produced with more balanced nutrition. Crop plants vary in their nitrogen requirements; thus those favoured by acidic conditions (calcifuges) are more likely to utilize ammonium nitrogen (NH₃-N) compared with those favoured by alkaline soil pH (calcicoles) which are favoured by nitrate (NO₂-N) nitrogen.

Pathogen effects

The initial level of inoculum has often been unquantifiable. Molecular diagnostics is becoming much more readily available especially for soilborne pathogens; consequently, this impediment is reducing. The impact of this factor was defined by Garrett (1957) as the ‘inoculum potential’ and has been the subject of intense study by Gilligan and colleagues (Gilligan, 2002). Many diagnostic warning systems presuppose that infection begins from the same basic value in each season or automatically assume that infective propagules are always present. Neither of these tenets are necessarily the case especially with perennial plants where the pathogen may reside in an overwintering rootstock (e.g. Verticillium albo-atrum in lucerne (alfalfa, Medicago sativa) (Dixon et al., 1989) or in the body of the plant as in fruit trees (Posnette & Cropley, 1965).

Conditions favourable and adverse to the pathogen vary rapidly with changing weather. Warning systems tend to take into account conditions favourable to the pathogen and not those which inhibit successful invasion. Increased sophistication in disease prediction is adding adversity into invasion models. Competition with other microorganisms is also an important but largely an un-quantified factor. The environmental conditions most conducive to an epidemic are also likely to encourage competitive microbial antagonists, a situation which is becoming increasingly appreciated as soil ecology receives the increased attention which it undoubtedly warrants. For example, the relationship between the antagonist Coniothyrium minitans parasitizing species of Sclerotinia has now progressed to practical use as a form of biological control (Whipps et al., 2001).

Environmental effects

These can be differentiated into three aspects: crop climates, local climates and regional climates. Differences between within-crop and local climates depend at least partially on host factors such as canopy density and plant architecture interacting with soil and air moisture and temperature values. Calm sunny days make for the biggest differences between plant and locality climate while cloudy wet days minimize these differences. The latter is the most conducive to pathogen establishment. Gross variations between local and regional weather patterns are caused by factors such as elevation and proximity to large water masses. Again, when the weather is stable, the variations between local and regional effects are small but become very large when conditions are unstable (Dixon, 1981).

Ecosystem effects

There will be effects on ecosystem services which in turn influence the effects of pathogen and host relationships and the eventual outcome in terms of reduced crop yield and quality. Ecosystem effects are cited by Eviner & Likens (2006) as pathogen effects on host survival, physiology, behaviour, and/or reproduction; changes to the life stages of a host vulnerable to a pathogen; the proportion of individuals, and/or biomass infected at a site; spatial extent and distribution of infection; rate of pathogen effects on hosts in relation to their rates of response, and/or recovery by hosts or individuals replacing other lost hosts; functional similarity of infected individuals versus replacements; frequency and duration of pathogen impact.

Colhoun (1973) reviewed the development of scientific understanding and principles underlying the influence of weather on disease development and intensity. Recognition that weather conditions affect plant diseases goes back to Theophrastus (370–286 BC) who realized that cereals grown on elevated land suffered less from rust (Puccinia spp.) than those grown at lower less windy locations. One of the earliest scientific studies was in France (Prevost, 1807) demonstrating that bunt (Tilletia caries) on wheat (Triticum aestivum) is caused by a fungus and disease aetiology relates to the prevailing environment. Subsequently, Hartig (1882) stated that plants contract diseases only when subject to defined conditions so that a predisposition to the tendency for disease results. He recognized seasonal factors and water content as determined by weather as predisposing factors which encouraged disease outbreaks. Analytically, Marshall Ward (1890) considered that any understanding of disease will be unsuccessful unless it includes considerations of the variations in the host-plant and parasite induced by changes in the environment. Changes in disease proneness due to environment and the distinction between the influence of external factors on the establishment of the parasitic relationship and its subsequent course were analyzed by Gäumann (1950). Correlations between environmental factors and increases and decreases in susceptibility were raised by Yarwood (1959). This work brought Darwin's equation of

firmly into focus in the analysis of climate and plant disease interaction. Even beginning to resolve the implications of that step forward had to await the detailed molecular level discoveries which are now emerging.

The identification of individual drivers in the relationship between plant disease and climate change has been developed to a substantial level which interprets factors in the interactions between climate change and plant disease. Factors which are important include for example temperature as it affects pre-inoculation temperature of the host; keeping quality during storage; survival of pathogens; spore germination; incubation period; symptom development and expression; cultivar reaction; response to disease of different host species; the stage of plant development; the stage of development of the fungus or other pathogen; reduction in plant vigour; spore production and discharge; mutation of fungi and other microbes and the effectiveness of fungicides and other agrochemicals. Moisture in both the soil and aerial atmospheres significantly affects the relationship. Similarly, light intensity, spectral composition and photoperiod through their impact on the vigour and reproductive state of the host are environmental factors ultimately affecting plant disease. Finally, the physical, chemical and biological status of soil will substantially affect disease development caused by soilborne microbes (Dixon & Tilston, 2010).

One of the most compelling pieces of evidence for the impact of atmospheric change on plant disease come from Fitt et al. (2011) using detailed analyses of samples taken over an extended period from the Rothamsted Broadbalk Field experiment which started early in the 19th century. Preserved wheat samples from the 170-year-old Broadbalk winter wheat experiment at Rothamsted offer a unique insight into changes in the long-term prevalence of pathogens. Samples of DNA from two septoria pathogens, Mycosphaerella graminicola (Septoria tritici) (speckled leaf blotch) and Phaeosphaeria nodorum (Stagnospora nodorum) (glume blotch, leaf blotch), were amplified from grain and straw. Phaeosphaeria nodorum is seedborne whereas M. graminicola is not. The long-term data series showed M. graminicola was common in the mid-19th century but very rare in the first three-quarters of the 20th century, while P. nodorum showed an approximately opposite trend. Over the 170-year period, the incidences of these two pathogens were closely correlated with sulphur dioxide emissions over England and Wales. This suggests that atmospheric concentration of sulphur dioxide emissions were responsible for the changing balance between the two pathogens.

Evidence from other sectors of the biosphere supports the contention that fungal pathogens will thrive as climate changes (Hof et al., 2011). Amphibian populations are declining at faster rates than is seen with other vertebrate groups probably as a result of the pathogenic fungal disease chytridiomycosis caused by Batrachochytrium dendrobatidis. Climatic change is cited as part at least of the cause of this decline. Recently Fischer et al. (2012) produced compelling evidence of the dangers coming from emerging infectious diseases caused by fungi which increasingly threaten animal, human, plant and ecosystem health encouraged by climate change.

Crop loss assessment

The analysis of the effects of climate change on disease and food security demands an ability to assess crop losses. Early studies are summarized by Zadocks (1985). He suggested that defining what is at risk in crops attacked by pathogens in a period of climate change is a matter for speculation since this is placing past events into a foresight context. This means different motivations from farmers, politicians and scientists will be the drivers. Is ‘loss’ an economic (reduced financial return per unit area) or pathological value (reduced yield per unit area) or a series of increased costs due to crop protection resulting from an epidemic (increased costs per unit area)? Demand for agricultural commodities is normally inelastic with prices falling in times of oversupply and rising steeply in times of scarcity. It may well be that the farmer benefits from an increased disease pressure because that causes price rises. Even this can vary depending on motivation – in the developing world, the added-value from reducing disease risk is food security from starvation rather than a monetary reward. While for horticulturists either in the developed world or supplying its supermarkets, risk is physiological in terms of reduced quality as opposed to losses in gross yield. Here the presence of a small blemish which may be a failed infection (ghost spotting of tomatoes by unsuccessful Botrytis cinerea, grey mould for example) is sufficient to render the produce unacceptable and hence cause total crop rejection (Dixon, 1981, 1984). A framework for understanding risks to crop health as affected by climatic and agronomic changes is proposed by Savary et al. (2011). These authors emphasize the crucial need for scientifically valid data which quantifies change and the attendant risks in order to set priorities for crop protection.

The forms of disease-induced loss have been identified by Garrett et al. (2006) who provide a detailed review of the potential for change at different ecological and physiological levels. Thus, climate change-induced alterations to the form and structure of individual plants where, for example, differing plant structures (architecture) result in denser host growth and hence increased canopy humidity, can encourage some plant pathogens. It is already seen that temperature directly affects yield. For example, rice yield in the Philippines is estimated to decline by 1% for each 1 °C increase in minimum temperature during the dry season (Peng et al., 2004). Molecular changes have been linked with physiological effects, e.g. stomatal closure, inhibition of leaf growth, changes to plant architecture and shifts in root to shoot ratios (Chaves et al., 2003). Plant resistance may also be affected by temperature as described by Browder & Eversmeyer (1986) for interactions of temperature and time with some Puccinia recondite: Triticum corresponding gene pairs. They showed that host-pathogen pairs responded differently to varying temperature ranges, in particular pairings which tended to produce low infection rates at specific temperatures and time periods of exposure to these temperatures. Alterations at population level mean that changing climate may well affect wild genotypes and land races which are sources of resistance and other genes used for breeding programmes. Loss of genetic diversity might result in shorter useful lives for resistance genes.

Pathogens and vectors will change with alterations in their availability and patterns of colonization; the balance between different pathogens and their importance may shift as identified by Fitt et al. (2011). Effects may be to bring pathogens into contact with a different array of hosts and provide opportunities for hybridization which had not existed previously, as for example, expansion in the host range of Phytophthora cinnamomi in Europe (Bergot et al., 2004). This is already a pathogen with a huge host range across many taxonomic host groupings. In North America, needle blight (Dothistroma septosporum), is spreading northwards with increasing temperatures and precipitation (Woods et al., 2005) which may be related to climate change.

Temperature governs the rate of reproduction of fungi and longer seasons which allow more pathogen generations to mature will extend opportunities for pathogen evolution. Similarly, as pathogen populations increase, so the rate of evolution of novel virulence genes may increase. Altered temperature regimes may offer greater scope for overwintering of sexual stages thereby accelerating gene recombination and opportunities for the development of more aggressive pathogen strains. Changes occur to host physiological relationships, for example, where elevated carbon dioxide concentration can increase the pathogen load on C₃ grasses due perhaps to increased leaf longevity and photosynthetic rate (Mitchell et al., 2003; Mitchell, 2006). Decreases in species diversity in grassland swards may exacerbate this effect. Other greenhouse gases such as ozone (O₃) have varying effects on fungal growth and reproduction and also damage the host plant which could alter susceptibility to disease. Alternatively, infection by fungi can predispose host plants to increased damage from other groups of pathogens (von Tiedemann & Firsching, 2000). Precipitation and its interaction with disease outbreaks and severity have been long studied with fungi and as a result, prediction becomes increasingly possible. Reduced rather than increased precipitation appears to be a more common result of climate change. Drought stress and pathogen induced disease may have additively debilitating effects. This has been postulated particularly for viruses such as Maize dwarf mosaic virus (Olsen et al., 1990) and Beet yellows virus (Clover et al., 1999).

Disease management

Disease control increasingly focuses on the use of Integrated Pest Management (IPM). Climate change will seriously affect efficacy of elements of IPM, such as pathogen resistance in cultivars, efficacy of pesticides, the efficiency of bio-control agents and opportunities for the use of husbandry controls. Bio-control agents are increasing in importance for field horticultural crops (e.g. fruit diseases) and have a dominant role in protected cropping as the availability of pesticides is reduced and consumers demand residue-free fresh produce. Risk models of the invasion of alien pathogens depend on using variables such as temperature, rainfall and humidity as major components. The outcomes of such models affect trade restrictions between countries e.g. the effects of the presence of Karnal bunt (Tilletia indica) on trade. These models rely on inputs which have substantial degrees of uncertainty. The non-linear nature of data and the presence of thresholds in the relationships between climate and epidemiology make it difficult to obtain sufficient information while adaptation by plants and their pathogens provide further complicating factors. Overlaying these difficulties with climate change prediction models which are only coarsely defined currently makes for difficulties in achieving robust and reliable predictions. Resistance durability (sensu Johnston), ‘resistance that remains effective during prolonged and widespread use in an environment favourable to the disease’ (Johnston, 1984), and increasingly favourable and prolonged environments will tend to reduce durability and encourage adaptation. Increased variability of weather due to climate change (Schiermeier, 2011) adds extra uncertainty in dealing with diseases and achieving food security. Large-scale changes in host and pathogen distribution will only be studied by remote sensing (Scherm et al., 2000). Manipulation of husbandry offers a sustainable means for minimizing or completely controlling pathogens. Each step from crop planning through to harvest and subsequent processing and distribution is amenable to being used for disease control. Location and distribution of crops from the macro-regional to the micro-holding scales will influence the vulnerability of plants to invasion, colonization and destruction by pathogens. Employing husbandry that is fit-for-purpose ensures that all resources are used efficiently and produces healthy crops which have minimal environmental impact (Dixon & Tilston, 2010), thereby mitigating the effects of climate change.

This approach has been summarized in the requirement ‘To develop credible and useful climate-change predictions, plant disease models themselves must capture a thorough quantitative understanding of disease epidemiology and their reliability in disease forecasting must be proven through rigorous testing and validation’ (Shaw, 2009); and by Sturrock et al. (2011) who identify that management will require monitoring, forecasting, planning and mitigation for diseases. This applies to food crops and with equal force to forests and landscapes which are critical refuges for biodiversity (Dixon & Margerison, 2009), a central component in the earth's biogeochemical systems and a source of ecosystem services essential for human well-being (Shvidenko et al., 2005). They are identified as a major component as carbon sinks in the mitigation of global climate change (Solomon et al., 2007). Forests have declined by 40% in 300 years due to human activities, largely being utilized for agricultural land (Shvidenko et al., 2005). Now less than one-third of the earth's land area is forested. It is expected that most pathogens will be able to migrate to regions where the climate is compatible for them at rates faster than their tree hosts. Climate change will affect the lifecycles and biological synchrony of many forest trees and their pathogens. The result will be changes to the distribution and phenology, such as bud-break in trees, spore release by pathogens, and insect activities where these serve as vectors. This could significantly affect disease incidence and severity.

Food supply

It is apparent that climate change is adversely influencing the growth and reproduction of crops especially in areas of the world least able to cope with an increasingly insecure food supply (Lumpkin, 2011). Additionally, the scale of crop and post-harvest damage caused by plant pathogens is also accelerating since the changed climates benefit the activities of disease-causing microbes. The potential impact of pathogens in relation to yield reductions was identified by Scherm et al. (2000) and Scherm (2004) as due to their short generation times, high reproductive rates and efficient dispersal mechanisms. While there are some inherent deficiencies in using a modelling approach in order to examine the impact of climate change on pathogens and consequential reductions in food supply, it does at least offer a means of implementing phyto-sanitary control of invading aliens and understanding the earlier appearance of established microbes during the growing season (Gregory et al., 2009). Increased variability of weather events is expected to make pathogen attacks more aggressive and unpredictable and hence increase the difficulties of employing Integrated Pest Management (IPM) systems for their control. Establishing robust food security policies nationally, regionally and worldwide would be greatly aided if it were possible to include scientifically valid predictions and evaluations of the impact of climate change on pathogen behaviour (Ingram et al., 2008).

Two major forces will affect the capability to provide food security (Crute & Muir, 2011). Firstly, the abilities of pathogens to evolve tolerant strains capable of eroding genetic resistance and tolerating pesticides, and secondly, the rate of environmental change which is a consequence of climatic shifts. The speed, magnitude and variation between locations of change affect the sustainability of food production. Both affect abilities to adjust to previously un-encountered biotic and abiotic stresses (Gornall et al., 2010; Jaggard et al., 2010). Opportunities for the exploitation of biological control of pathogens are increasing as greater knowledge accumulates regarding plant defence systems, perception of the environment by plants and arrival of damaging microbes (Lucas, 2011). Each process involves signalling and transduction molecules which are potentially capable of being formulated as means of biological control (biopesticides). Benign microbes which far outnumber malignant forms potentially offer a substantial contribution towards sustainable crop production in the face of climate change (Dixon & Tilston, 2010), especially where their use is integrated with husbandry-based controls.

The greatest losses caused by microbes and other forms of food waste in developed economies take place after the farm-gate, in the marketing chain and during consumer use. Here, up to 30% of all food produced is wasted. This starkly contrasts with underdeveloped and developing nations where the greatest source of waste is before the farm-gate in the field and in crop storage and processing (Hodges et al., 2011) as direct consequences of microbial and pest damage.

Previously, food crises were solved by the Green Revolutions of the 1960s and 1970s, where productivity was expanded at a faster rate than population growth and increasing consumption. Now, science and technology lags behind rising populations and increasingly sophisticated consumption. Grappling with the inequality between supply and demand for food as both calories and nutritional content cannot be solved solely by a top-down approach through the action of governments, however well-meaning they may be. In the past 18 months, several non-governmental organizations (NGOs) and others have united in an attempt to increase food production without damaging wildlife biodiversity (Clay, 2011). This includes groups such as the Global Harvest Initiative (www.globalharvestinitiative.org), Sustainable Agriculture Initiative (www.saiplatform.org) and the World Wildlife Fund (WWF) (www.wwf.org). These organizations have identified what they term as ‘food wedges’ i.e. areas where capacity building will have a substantial effect in bringing demand and supply into balance without increasing damage to the environment and biodiversity. These wedges are genetics; improving husbandry techniques; increased husbandry efficiency which includes reducing the impact of pathogens; rehabilitating degraded land; reducing waste, especially food waste; reducing excessive consumption; preserving intellectual property rights; and conserving carbon in soil by increasing soil organic matter content. The need is to double the genetic potential of the 10–15 main calorie-supplying crops using the same land area as currently, simply in order to meet the rising demand with little slack in the system for contingent natural and man-made disasters.

In a further move towards improving the sustainability of traded crops, Clay (2011) suggests particularly in relation to African producers (but there is no reason why this should not apply elsewhere) that retailers and brand-named companies which purchase commodities like sugar, milk, coffee, cocoa or palm oil should also purchase carbon that the farmer has sequestered or avoided releasing during crop production. He suggests a goal of 1 billion tonnes of carbon forgone being sold by producers per year by 2030. This idea is being developed by the Dutch government, in collaboration with companies such as Unilever, Nutreco and Rabobank. Finally, Clay (2011) very realistically identifies that no one single approach will be satisfactory, that relying on the ‘magic bullet’ of a second Green Revolution will not work. This conclusion is supported by Crute & Muir (2011); there is no single solution to delivering much increased productivity from food production systems while reconciling this with the need to maximize sustainable efficiency of resource use and diminish environmental impact. Combining aspects of organic production with best practices of conventional intensive agriculture currently offers a beneficial means of raising food supply with minimally adverse impacts on the environment. Organic systems alone do not offer a solution to the problem of feeding an increasing human population because of their inherently lower yields (Seufert et al., 2012).

There is little doubt that scientists can deliver such solutions if provided with the right tools and skills. What is in doubt is whether there is sufficient political awareness of the scale of the problem and the diminishing time left within which to achieve substantially new approaches to food production. This requires political awareness on an international scale which is prepared to set aside national interests for the benefit of the whole planet. Food production must increase substantially but agriculture's ‘foot-print’ must shrink (Foley et al., 2011). These authors advocate firstly halting agricultural expansion. Their data demonstrate that between 1985 and 2005 there has been a net redistribution of cropped lands towards the tropics. This erroneous change is at odds with the need for mitigating climate change. Agriculture is the largest single use of the Earth's terrestrial surface (38%) (Ramankutty et al., 2008; Anon., 2011b ). Currently, agricultural expansion largely involves dramatic losses to tropical forests which damage sustainability and erode natural biodiversity with at best limited economic gains (Lambin & Meyfroidt, 2011). The potential for benefiting from forest protection is contained in the proposed Reducing Emissions from Deforestation and Degradation Programme (REDD) (Kremen et al., 2008). Further means for closing the yield gaps (defined as ‘the difference between crop yields observed at any given location and the crop's potential yield at the same location given current agricultural practices and technologies’) are advocated. To that Foley et al. (2011) describe the benefits which can be derived from genetic improvements, improved resource management and combining the advantages of precision and organic agriculture. Particular advantages can be derived from more efficient and sustainable use of water, nutrients and agrochemicals. They identify the ‘Goldilocks Syndrome’ where some regions use far too much and others far too little of a particular resource with few managing to achieve the correct dose (Jaggard et al., 2010). Excess resource use is a particular problem in China, northern India, USA and western Europe. Some 10% of the world's croplands account for 32% of the surplus nitrogen and 40% of the surplus phosphorus used. Foley et al. (2011) suggest targeting ‘hotspots’ of low resource efficiency for improvement. These authors also suggest that current resources can be used more effectively by changing agricultural systems and dietary decisions. Moving from crop production which targets supplying animal feedstuffs as a priority, and recognizing that bio-energy crops are not a sensible route towards the supply of fuel for motor vehicles and other non-food cropping would make far more sustainable use of limited land resources. They also highlight the huge amounts of waste which exist in pre- and post-farm gate chains. Much of the science and technology needed is already available – what is lacking is an effective cadre capable of translating this knowledge into practical solutions. Regrettably, worldwide governments in a time of relative abundance over the past generation have all but eliminated the supply of science-based specialists endowed with the competence, understanding and skills with which to interpret this knowledge into guidance for practical farmers. Re-addressing this deficiency will require substantial changes in the manner by which agricultural science education is delivered.

The overriding need is addressing the causes of climate change. This means reducing gaseous emissions into the atmosphere. The biggest contributor is the emission of carbon dioxide derived from fossil fuels. Other gases such as methane, nitrous oxide and ozone-depleting substances coming largely from non-fossil fuels also contribute to the climate-change problem. These have shorter lifetimes in the atmosphere than carbon dioxide and hence their reduction would have a more rapid effect on the problem (Montzka et al., 2011). Equally, there is a requirement for education which allows women to control their own bodies and reproductive processes such that fewer but healthier and thriving children are added to the world's population (Sulston, 2011).

Notes

This paper was a symposium contribution at the joint meeting held with the Canadian Phytopathological Society and Plant Canada entitled ‘Plant adaptation to environmental change’ in Halifax, Nova Scotia, July 2011.