Climate is the single most important determinant of agricultural productivity, primarily through its effects on temperature and water regimes (Oram, 1989). For example, the physiographic boundaries of principal biomes are determined by mean annual temperature (Mooney et al., 1993) and soil water regimes. Climate change is therefore expected to alter the biophysical environment of growing crops and to influence biomass productivity and agronomic yields (Rosenzweig and Hillel, 1998).
Positive effects may be associated with the fertilization effects of CO2 enrichment, increases in the duration of growing seasons in higher latitudes and montane ecosystems, and possible increase in soil water availability in regions with an increase in annual precipitation. Each 1°C increase in temperature may lead to a 10-day increase in the growing season in northern Europe and Canada (Carter et al., 1991). The CO2 fertilization effect is real (Allen, 1994; Wittwer, 1995; Allen and Amthor, 1995; Allen et al., 1996). However, the net positive effect may be moderated by other factors, such as the effective rooting depth and nutrient availability. Further, the productivity per unit of available water is expected to rise by 20% to 40% (van de Geijn and Goudriaan, 1996).
Negative effects of projected climate change on agriculture may be due to increases in respiration rate as temperature rises with attendant decreases in net primary productivity (NPP) (Abrol and Ingram, 1996); increases in the incidence of pests and diseases; shortening of the growing period in some areas; decrease in water availability as rainfall patterns change; poor vernalization; and increased risks of soil degradation caused by erosion and possible decline in SOC concentration. In contrast, the widespread incidence of drought in the United States in 1998 has been attributed to El Nino-related climate change (Bernard, 1993). The yield of rice has been estimated to decrease by 9% for each 1°C increase in temperature (Kropff et al., 1993). Phillips et al. (1996), using the explicit planetary isentropic coordinate (EPIC) model to examine the sensitivity of corn and soybean yields to climate change, projected a 3% decrease in both corn and soybean yields in response to a 2°C increase in temperature from a baseline precipitation level. However, a 10% precipitation increase balanced the negative effect of a 2°C temperature increase.
Rosenzweig et al. (1996) predicted the potential impact of climate change on citrus and potato production in the United
States. Their simulated treatments included combinations of three increased temperature regimes (+1.5, +2.5, and +5.0°C) and three levels of atmospheric CO2 concentrations (440, 530, and 600 ppm). Citrus production may shift slightly northward in the southern states, but yields may decline in southern Florida and Texas. Fall potato production may be vulnerable to increased temperature in the northern states. Parry and Carter (1998) assessed the effects of increase in temperature on crop yields while maintaining the direct effects of CO2 and precipitation held at current levels. They observed that average crop yields (wheat, rice, soybean, and maize) show a positive response (7% to 15%) to warming of +2°C and a negative response (-2% to -10%) to warming of +4°C. The effect is neutral in the range of 2°C to 4°C, whereas beyond this range, crop yields begin to decline.
These are global averages, however, and are unlikely to apply at regional or national levels. For example, a +2°C warming may have a strong positive effect in Canada or North America but a strong negative effect in Pakistan and India (Parry and Carter, 1998). In some sensitive ecoregions (Pakistan, Mexico, Egypt), crop yield declines could be 30% to 45% (Rosezweig and Parry, 1994). The effects of climate change on crop yields may be more negative at lower latitudes and generally positive at middle and high-middle latitudes. Further, crop growth is more affected by extremes of weather than by averages. The annual average changes in temperature or precipitation used in most predictive models do not reflect the short-term effects of so-called extreme events — droughts, floods, freezes, or heat waves.
Was this article helpful?