Extremely arid (<100 mm annual precipitation) deserts in Latin America (i.e., the Chihuahuan, Sonoran, Peruvian, Atacama, Monte, Patagonian deserts) occupy a large proportion of the region (see Figure 6-4); they have significant species richness and a high degree of endemism (IPCC 1996, WG II, Section 4.4.2). On the other hand, their contribution to the region's primary productivity and C and N pools is extremely low. Because they are driven by discrete events (e.g., rainfall) that occur at irregular intervals, deserts are described as pulse-driven ecosystems (IPCC 1996, WG II, Section 3.3.2); soil instability, dune systems, and shifting boundaries are some of their typical features. In some of the driest coastal deserts (e.g., in northern Chile and Peru), fog banks (camanchaca) are particularly frequent and provide the largest, if not the entire, moisture supply in most years (IPCC 1996, WG II, Figure 3-1 and Section 3.3.4). Harvesting of camanchaca-using frames with vertical nylon threads, either static or rotating at very low speeds-is a common activity among communities along the coast; such harvesting provides enough water to maintain some coastal rangelands and woodlands, which support seasonal grazing by sheep and goats.
Specific information on the vulnerability of Latin America's extreme deserts is rather poor, and further research is required. As with all extreme deserts, however, these systems already experience wide fluctuations in rainfall and are adapted to coping with the consequences of extreme conditions. Initial changes associated with climate change are unlikely to create conditions significantly outside the present range of variation (IPCC 1996, WG II, Section 3.2).
Changes in precipitation projected under a set of climate scenarios assembled for sensitivity studies in the Second Assessment Report (SAR) (Greco et al., 1994) (e.g., a 25% increase in the present mean of 2 mm/yr rainfall) are unlikely to produce major ecosystem changes. Regarding changes in temperature, increases projected in the Greco et al. (1994) scenarios typically are in the range of 0.5-2.0�C, with greater increases in the summer. A rise of 2�C without an increase in precipitation would increase potential evapotranspiration by 0.2-2 mm per day (IPCC 1996, WG II, Sections 3.3.2, 4.2.1).
Human-induced desertification at the boundary between arid and extremely arid areas has the potential to counteract any ameliorating effects of climate change on most deserts, unless appropriate management actions are taken. Even if wetter conditions were to prevail, their effects may be overridden by pressure from resource exploitation (IPCC 1996, WG II, Section 4.4.2). Other arid and semi-arid areas in Latin America also suffer from desertification; in Mexico, for example, water erosion alone affects 85% of the territory (Gligo, 1995). Vulnerability studies undertaken in Mexico under the USCSP (Gay-Garc�a and Ruiz Suarez, 1996) show that drought in central Mexico could be severe, affecting various states in the area (such as Michoacan, where more than 50% of the land surface is highly vulnerable to desertification). Similar trends are expected in Jalisco, Colima, Nayarit, Queretaro, Hidalgo, and Guanajuato.
Other reports in this collection