IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group II: Impacts, Adaptation and Vulnerability

13.4.2 Agriculture

Several studies using crop-simulation models and future climate scenarios were carried out in Latin America for commercial annual crops (see Table 13.5). According to a global assessment (Parry et al., 2004), if CO2 effects are not considered, grain yield reductions could reach up to 30% by 2080 under the warmer scenario (HadCM3 SRES A1FI), and the number of additional people at risk of hunger under the A2 scenario is likely to reach 5, 26 and 85 million in 2020, 2050 and 2080, respectively (Warren et al., 2006). However, if direct CO2 effects are considered, yield changes could range between reductions of 30% in Mexico and increases of 5% in Argentina (Parry et al., 2004), and the additional number of people at risk of hunger under SRES A2 would increase by 1 million in 2020, remain unchanged in 2050 and decrease by 4 million in 2080.

More specific studies considering individual crops and countries are also presented in Table 13.5. The great uncertainty in yield projections could be attributed to differences in the GCM or incremental scenario used, the time-slice and SRES scenario considered, the inclusion or not of CO2 effects, and the site considered. Other uncertainties in yield impacts are derived from model inaccuracies and unmodelled processes. Despite great variability in yield projections, some behaviour seems to be consistent all over the region, such as the projected reduction in rice yields after the year 2010 and the increase in soybean yields when CO2 effects are considered. Larger crop yield reductions could be expected in the future if the variance of temperatures were doubled (see Table 13.5). For smallholders a mean reduction of 10% in maize yields could be expected by 2055, although in Colombia yields remain essentially unchanged, while in the Venezuelan Piedmont yields are predicted to decline to almost zero (Jones and Thornton, 2003). Furthermore, an increase in heat stress and more dry soils may reduce yields to one-third in tropical and sub-tropical areas where crops are already near their maximum heat tolerance. The productivity of both prairies/meadows and pastures will be affected, with loss of carbon stock in organic soils and also a loss of organic matter (FAO, 2001b). Other important issues are the expected reductions in land suitable for growing coffee in Brazil, and in coffee production in Mexico (see Table 13.5).

Table 13.5. Future impacts on the agricultural sector.

   Yield impacts (%) 
Study Climate scenario Wheat Maize Soybean Rice Others  
Guyana CGCM1 2020-2040 (2xCO2       -3 Sg: -30 
(NC-Guyana, 2002)  CGCM1 2080-2100 (3xCO2      -16  Sg: -38 
Panama HadCM2-UKHI (IS92c-IS92f)   +9/-34/-21       
 2010/2050/2100 (1xCO2          
Costa Rica +2ºC -15% precip. (1xCO2)        -31  Pt: ↓  
(NC-Costa Rica, 2000)              
 +1.5ºC -5% precip.   +8 to -11   -16 Bn: +3 to -28 
(NC-Guatemala, 2001)  +2ºC +6% precip.   +15 to -11   -20 Bn: +3 to -42 
 +3.5ºC -30% precip.    +13 to -34   -27 Bn: 0 to -66 
Bolivia GISS and UK89 (2xCO2).I   -25       
 Incremental (2xCO2  +50       
 +3ºC -20% precip.       -15   
 optimistic-pessimistic (1xCO2        Pt: +5 to+2*2 
 optimistic-pessimistic (2xCO2        Pt: +7 to+5*2 
 IS92a (1xCO2)*1     -3 to -20     
 IS92a (2xCO2)*1       +12 to +59     
Brazil GISS (550 ppm CO2)  -30  -15  +21      
SESA*3 Hadley CM3-A2 (500 ppm) +9 to +13 -5 to +8 +31 to +45     
(Gimenez, 2006)  Hadley CM3-A2 (500 ppm).I +10 to +14 0 to +2  +24 to +30     
Argentina, Pampas +1/+2/+3ºC (550 ppm CO2).I +11/+3/-4 0/-5/-9  +40/+42/+39     
(Magrin and Travasso, 2002)  UKMO (+5.6ºC) (550 ppm CO2).I -16  -17 +14      
Honduras Hadley CM2 (1xCO2) 2070   -21       
(Díaz-Ambrona et al., 2004) Hadley CM2 (2xCO2) 2070    0        
Central Argentina Hadley CM3-B2 (477ppm)   +21       
(Vinocur et al., 2000;  ECHAM98-A2 (550ppm)   +27       
Vinocur, 2005)  +1.5/+3.5ºC (1xCO2  -13/-17       
 +1.5/+3.5ºC (1xCO2) (2T?)*4    -19/-35        
Latin America HadCM2    -10        
(Jones and Thornton, 2003)  (smallholders)           
Latin America HadCM3 A1FI (1xCO2Cereal yields:   -5 to -2.5 (2020) -30 to -5 (2050)  -30 (2080) 
(Parry et al., 2004) HadCM3 B1 (1xCO2  -10 to -2.5 (2020) -10 to -2.5 (2050) -30 to -10 (2080) 
 HadCM3 A1FI (2xCO2   -5 to +2.5 (2020) -10 to +10 (2050)  -30 to +5 (2080) 
 HadCM3 B1 (2xCO2  -5 to -2.5 (2020) -5 to +2.5 (2050)  -10 to +2.5 (2080) 
Mexico, Veracruz HadCM2 ECHAM4 (2050) Coffee: 73% to 78% reduction in production 
(Gay et al., 2004)     
Brazil, São Paulo +1ºC + 15% precip. Coffee: 10% reduction in suitable lands for coffee 
 +5.8ºC + 15%precip.  97% reduction in suitable lands for coffee 
Costa Rica Sensitivity analysis  Coffee: Increases (up to 2ºC) in temperature would benefit crop yields 

I = Irrigated crops; precip. = precipitation; *1 Values correspond to soybean sowing in winter and summer for 2010 and 2020; *2 Increases every 10 years. *3 SESA= South East South America; *4 2T?: doubled variance of temperature. Bn: bean, Sg: sugar cane, Pt: potato.

In temperate areas, such as the Argentinean and Uruguayan Pampas, pasture productivity could increase by between 1% and 9% according to HadCM3 projections under SRES A2 for 2020 (Gimenez, 2006). As far as beef cattle production is concerned, in Bolivia future climatic scenarios would have a slight impact on animal weight if CO2 effects are not considered, while doubling CO2 and increases of 4°C in temperature are very likely to result in decreases in weight that could be as much as 20%, depending on animal genotype and region (NC-Bolivia, 2000).

Furthermore, the combined effects of climate change and land-use change on food production and food security are related to a larger degradation of lands and a change in erosion patterns (FAO, 2001b). According to the World Bank (2002a, c), some developing countries are losing 4-8% of their GDP due to productive and capital losses related to environmental degradation. In drier areas of Latin America, such as central and northern Chile, the Peruvian coast, north-east Brazil, dry Gran Chaco and Cuyo, central, western and north-west Argentina and significant parts of Mesoamerica (Oropeza, 2004), climate change is likely to lead to salinisation and desertification of agricultural lands. By 2050, desertification and salinisation will affect 50% of agricultural lands in Latin America and the Caribbean zone (FAO, 2004a).

In relation to pests and diseases, the incidence of the coffee leafminer (Perileucoptera coffeella) and the nematode Meloidogyne incognita are likely to increase in future in Brazil’s production area. The number of coffee leafminer cycles could increase by 4%, 32% and 61% in 2020, 2050 and 2080, respectively, under SRES A2 scenarios (Ghini et al., 2007). According to Fernandes et al. (2004), the risk of Fusarium head blight incidence in wheat crops is very likely to increase under climate change in south Brazil and Uruguay. The demand for water for irrigation is projected to rise in a warmer climate, bringing increased competition between agricultural and domestic use in addition to industrial uses. Falling watertables and the resulting increase in the energy used for pumping will make the practice of agriculture more expensive (Maza et al., 2001). In the state of Ceará (Brazil), large-scale reductions in the availability of stored surface water could lead to an increasing imbalance between water demand and water supply after 2025 (ECHAM scenario; Krol and van Oel, 2004).