Rangelands-grasslands, shrublands, savannahs, and hot and cold deserts, excluding hyper-arid deserts-cover 33% of the area of Latin America (IPCC 1996, WG II, Section 2.1). Rangeland productivity and species composition are directly related to the highly variable amounts and seasonal distribution of precipitation and only secondarily controlled by other climate variables. Because rangeland systems are driven by extremes, increases in the frequency and magnitude of extreme events (Easterling, 1990) may have a disproportionate effect on them (Westoby et al., 1989). Because of their low productivity, rangeland management units usually are very large, often incorporating considerable spatial heterogeneity (Stafford Smith and Pickup, 1993).
Latin American rangelands sustain pastoralist activities, subsistence farming, and commercial ranching and are a key factor in the economy of many countries (e.g., Mexico, countries of the Central American isthmus, Brazil, Argentina, Uruguay). There are approximately 570 million animal units on the subcontinent, and over 80% of them feed from rangelands (Annex D).
Human activities may bring about more change in rangeland ecosystems than any other forces of global change and may interact strongly with climate change impacts, particularly in tropical and subtropical areas (IPCC 1996, WG II, Section 2.3.3). In Argentine Patagonia, for example, the introduction of unsustainable numbers of sheep along with inappropriate management policies has resulted in major changes in pastureland composition and even desertification (Soriano and Movia, 1986). This process is causing the loss of approximately 1,000 km2 per year (LAC CDE, 1992); overall, 35% of the area's pastureland has been transformed into desert (Winograd, 1995). As a result, the number of sheep decreased by 30% between 1960 and 1988-representing a loss of about US$260 million (Paruelo and Sala, 1992).
Boundaries between rangelands and other biomes are likely to be affected by climate changes both directly (e.g., changes in species composition) and indirectly, through changes in fire regimes, opportunistic cultivation, or agricultural release of less-arid margins of the rangeland territory.
Concerning possible climate impacts of tropical deforestation, Salati and Nobre (1991) pointed out that the large-scale conversion of tropical forests into pastures will likely lead to changes in local climates of the Amazon region. This kind of land-use change increases surface and soil temperatures, the diurnal fluctuation of temperature, and the specific humidity deficit and reduces evapotranspiration-there is less available radiative energy at the canopy level because grass has a higher albedo than forests (Cerri et al., 1995).
Climate characteristics affecting soil moisture conditions, relative humidity, or drought stress-in conjunction with changes in fire and grazing regimes-will have the greatest influence on the boundaries of grassland and woody species. In Mexico, approximately 70% of xerophytic shrublands have been projected to shift their geographical distribution as a result of climate change (Villers, 1995). This projection is roughly in agreement with outputs from MAPSS and BIOME 3 (Neilson, 1995; Haxeltine and Prentice, 1996; see Annex C), although these models suggest less dramatic suface reduction. The overall carrying capacity for herbivores may increase or decrease, depending on the balance between eventual increases in productivity and decreases in plant nutritional value. In some regions, warmer temperatures and increased summer rainfall, with fewer frost days, may facilitate the replacement of temperate grasses by tropical grasses (IPCC 1996, WG II, Section 2.6). Although some laboratory experiments on plants grown individually have shown that C3 (temperate) plants tend to respond more positively than C4 (tropical) plants to elevated CO2, differential effects can be offset or even reversed in the field because traits other than photosynthetic pathways-such as architecture phenology, water, and nutrient-use efficiency-tend to play more decisive roles in the field (Bazzaz and McConnaughay, 1992).
The net effect of elevated CO2 on forage quality also is uncertain. Elevated CO2 usually results in higher C:N ratios in plants. This effect is reflected in increased herbivore consumption and decreased herbivore fitness in laboratory experiments involving insect populations and individual plants (Lindroth, 1996). It is uncertain whether similar patterns would occur in systems involving rangelands and large vertebrate herbivores (D�az, 1995). Any alteration of the standing capacity of grasslands will be economically important, given the scale of livestock production in Latin American tropical and temperate grasslands.
Fire has been a factor in the evolution of grasslands and many other types of rangelands (Medina and Silva, 1990; Eskuche, 1992). Projected substantial increases in the frequency and severity of wildfires (Ottichilo et al., 1991; Torn and Fried, 1992) could lead to vegetation and soil alterations (Ojima et al., 1990).
Temperate grasslands and the animal production depending on them are vulnerable to drought. Therefore, livestock production could be negatively affected by higher temperatures or increased evapotranspiration rates. However, experience has shown that extreme events, such as large-scale floods or drought-erosion cycles, may pose the highest risks (Burgos et al., 1991; Suriano et al., 1992).
Because of their large extent and important capacity to sequester carbon, temperate grasslands play an important role in the maintenance of the composition of the atmosphere. Sala and Paruelo (1997) have estimated the value of maintaining native grasslands to be US$200/ha; they stress that, once grasslands have been transformed into croplands, the reverse process of abandonment of croplands and their slow transformation into native grasslands sequesters only a modest amount of carbon over relatively long periods of time. Fisher et al. (1994) have proposed increasing the carbon sequestration capacity of grasslands through reduced burning frequency, nutritional supplementation of soils, and introduction of deep-rooted grasses and legumes, in combination with controlled stocking rates.
Other reports in this collection