Climate Change 2001:
Working Group II: Impacts, Adaptation and Vulnerability
Other reports in this collection Change in land use

Table 15-4: Percentage change in U.S. supply of cereals under various constraints, by climate change scenario (Schimmelpfennig et al., 1996).
No adaptation
With adaptation

Drought may increase in the southern Prairies, and production areas may shift northward in Canada. In assessing the potential for expansion to areas in northern Canada (i.e., north of 55°N and west of 110°W) and Alaska, Mills (1994) identified 57 Mha of potentially arable land (class 1-5, based on Canada Land Inventory criteria) with agricultural potential for use in either annual cropping or perennial forage systems. This estimate drops to 39 Mha when climatic limitations are imposed but under a scenario of doubled atmospheric CO2, increases to 55 Mha with an accompanying improvement of land class to class 3. Similar outcomes—expansion of agricultural land, especially expansion of the zone suitable for corn and soybean production—are expected for northern areas of eastern Canada (Brklacich et al., 1997a). Other case studies conducted in the southern portion of the Mackenzie basin in northwestern Canada show that two different climate-change scenarios would relax the constraints imposed by a short and cool frost-free season but that drier conditions and accelerated crop development would offset the potential gains of a warmer climate (Brklacich et al., 1996, 1997b).

Southern regions growing heat-tolerant crops such as citrus fruit and cotton might benefit from reduced incidence of killing frosts resulting from a change in climate (Miller and Downton, 1993; Mearns et al., 2000). Results of simulations without CO2-induced yield improvement indicate that production of citrus fruit would shift northward in the southern United States, but yields may decline in southern Florida and Texas because of excessive heat during the winter (Rosenzweig et al., 1996).

Mexican agriculture appears to be particularly vulnerable to climate-induced changes in precipitation because most (about 85%) of its agricultural land is classified as arid or semi-arid. Recent national assessments of the impacts of climate change indicate that the northern and central regions of Mexico are most vulnerable in the agricultural sector (Conde, 1999) and that in these regions, the area of land that is unsuitable for rainfed maize production would expand under climate change (Conde et al., 1997). On average, more than 90% of losses in Mexican agriculture are caused by drought (Appendini and Liverman, 1995). Using five GCM-based scenarios, it was estimated that potential evaporation may increase by 7-16% and the annual soil moisture deficit could increase by 18-45% in important maize-growing regions in eastern Mexico (Liverman and O'Brien, 1991). Rising levels of CO2 can have the greatest relative beneficial impacts when water is limited. Therefore, rising CO2 may be expected to have a significant positive impact because so much of Mexican crops are water-limited and rising CO2 enhances water-use efficiency (see Chapter 14). Crop yields and adaptation

Depending on existing conditions, global warming and CO2 enrichment can have positive or negative effects on crop yields. It is believed that yield increases in mid- and high latitudes are caused by positive physiological effects of CO2, longer growing season, and amelioration of the effects of cold temperature on growth. Decreases in yield could result from shortening of the growing period, reduced water availability, and/or poor vernalization.

Estimates of the impacts of climate change on crops across North America vary widely (see Table 15-3). In some studies, the impacts range from nearly total crop failure for wheat and soybeans at one U.S. site to wheat yield increases of 180-230% for other sites in the United States and Canada (Brklacich et al., 1994; Rosenzweig et al., 1994). Recent modeling efforts indicate that the impacts on yields for many crops grown under dryland conditions, even without adaptation, is positive (Reilly et al., 2000). Threshold limits associated with temperature increases may be important. Rosenzweig et al. (1995) report generally positive crop yield responses to temperature increases of 2°C, but yield reductions occurred at increases of more than 4°C. Modeled yield results that include the direct physiological effects of CO2 are substantially different from those that do not account for such effects (Fischer et al., 1996).

Although it is known that the distribution and proliferation of weeds, crop diseases, and insects are determined to a large extent by climate, most crop modeling efforts have not thoroughly accounted for potential impacts of climate change and variability on pest populations and ranges. Interactions between crops and pests under changing climate conditions will be very complex and are difficult to predict because elevated CO2, warmer temperatures, and increased climate variability would alter the relationships between crops, weeds, and insects significantly. Higher temperatures and warmer winters could reduce winterkill of insects as well as broaden the range of other temperature-sensitive pathogens (Rosenzweig et al., 2000). Increases in the incidence of extreme weather events could reduce the efficacy of pesticide applications and result in more injury to nontarget organisms (Patterson et al., 1999).

Modeling studies of changes in crop production show strong regional effects, with some areas suffering significant losses compared to other regions—suggesting that climate change may affect the comparative advantage of agricultural production regions within North America. For example, in scenarios investigated for the U.S. National Assessment (Reilly et al., 2000), the lake states, mountain states, and Pacific region showed gains in production, whereas the southeast, delta, southern Plains, and Appalachia generally lost. The economic impact of these changes in crop production as a result of climate change is considered to be mostly beneficial to society as a whole. The effects are largely detrimental to producers because the overall positive effect on production leads to decreasing prices. Thus, climate change is beneficial for foreign trade surplus and for consumers. Analyses of the economic effects of various climate change scenarios on the welfare of consumers and producers in the United States show that agricultural welfare strictly increases with 1.5 °C warming, but further warming reduces the benefit at an increasing rate (Mendelsohn et al., 1999). Additional precipitation is strictly beneficial (Adams et al., 1999).

The costs and benefits of climate change must be evaluated concurrently with behavioral, economic, and institutional adjustments brought about by climate change. These adjustments occur at different levels. For example, farm-level adaptations can be made in plant and harvest dates, crop rotations, selections of crops and crop varieties for cultivation, water consumption for irrigation, use of fertilizers, and tillage practices. At the market level, prices are a strong signal to adapt as farmers make decisions about land use and which crops to grow.

Current economic studies of climate change that include farm- and/or market-level adjustments suggest that the negative effects of climate change on agriculture probably have been overestimated by studies that do not account for adjustments that will be made. This may be caused by the ability of the agricultural production community to respond with great flexibility to a gradually changing climate. Typically, extreme weather poses a significant challenge to individual farming operations that may lack the spatial diversity and financial resources of large, integrated, corporate enterprises with production capabilities in one or more areas.

Simulation modeling using four GCM-based scenarios showed that U.S. cereal production decreases by 21-38% when farmers continue to do what they are now doing (i.e., no adaptation) (see Table 15-4). When scenarios that involve adaptation by farmers are used, decreases in cereal production are not as large and the adaptations are shown to offset the initial climate-induced reduction by 35-60% (Schimmelpfennig et al., 1996; Segerson and Dixon, 1999).

Other reports in this collection