The relationship between land cover and SOC storage is briefly discussed below, with more detailed discussion in Section 4.2.2. Forest soils can store large amounts of carbon that could be released to the atmosphere through deforestation (Houghton et al., 1983). For example, Brown and Lugo (1990) reported that in wet and moist life zones, sites cultivated after deforestation typically lose 60-70 percent of the initial carbon contained in mature forests. In the dry life zone, where initial soil carbon content is only about 50-60 percent of that in wet and moist life zone soils, soils may lose only 14 percent of their initial carbon under cultivation. Carbon may be lost by decomposition, with carbon released as CO2 to the atmosphere, or by erosion-in which case carbon may be deposited somewhere else in the landscape, with only some lost as CO2 (Lal, 1995).
There is ample evidence that when forests are converted to cultivated cropland, the organic layer is depleted, and soil carbon contents and cation exchange capacities can decrease (Detwiler, 1986; Mann, 1986; Schlesinger, 1986; Davidson and Ackerman, 1993). If cleared land is not cultivated or degraded, however, soil carbon amounts are mostly observed not to change over decades (e.g., Lugo and Brown, 1993; Christopher et al., 1997)-although in some studies, soil carbon losses of up to 30 percent have also been reported (e.g., Garc�a-Oliva et al., 1999). The outcome ultimately may depend more on the productivity of the vegetation (and hence nutrient supply and climate) than on vegetation type (Trumbore et al., 1995). Thus, if there is no regular cultivation, the inter-conversion between forest and non-forest vegetation may not have a consistent effect on soil carbon amounts.
Soil organic matter (carbon) content is often related to soil fertility. Forests are distinct in that they develop an organic layer above the mineral soil. This layer generally improves physical (soil aeration, water retention, resistance to erodability, etc.) and biological properties (build-up of soil microorganisms, nutrients, etc.), which enhance the productive capacity of the soil.
The fertility of soils following slash-and-burn has been shown to decrease rapidly (Lal, 1987, 1996). Up to 80 percent of the total carbon and nitrogen content of the soil can be contained in macroaggregates, where it is partly protected from microbial action and thus more securely stored. Burning can reduce the amount of organic carbon associated with macroaggregates by 32 percent, and it can disrupt soil aggregate stabilization by changing the chemical nature of organic carbon (Garc�a-Oliva et al., 1999).
When cultivated lands or soils that previously were low in organic matter are afforested or reforested, there can be substantial increases in the amount of soil organic matter (Ovington, 1959)-as occurs following shifting agriculture in the tropics. In addition, trees and their roots can play an important role in maintaining desirable soil physical properties (Young, 1989). Peatlands are exceptions: When drainage and disturbance to establish trees accelerates decomposition, the loss of carbon can eventually exceed the carbon store created by growing trees (Cannell et al., 1993).
Grassland management may include practices that are beneficial for carbon storage and reduce the adverse environmental impacts of agricultural activity but would not be profitable without carbon compensation (e.g., allowing savanna thickening at the expense of livestock production). Chapter 4 provides more detail on the sustainable development impacts of changes in agricultural practices.
The introduction of forest operations such as harvesting and site preparation in areas that remain as forest have the potential to decrease the amount of soil carbon if litter inputs to the soil are reduced (Liski et al., 1998; Thornley and Cannell, 2000). In most instances, however, detecting changes in soil organic matter observationally may be difficult (Johnson, 1992).
The Parties will have to decide which carbon pools to include as recognized carbon stocks (Section 2.3). If soil carbon pools were not included in the list of carbon stocks, there would be no incentive to maximize carbon storage in this important reservoir. Soil cultivation, for example, could be exercised to stimulate forest growth even though it could reduce soil carbon storage and thus lower total overall site carbon storage. Inclusion of all on-site carbon stocks for carbon accounting purposes would ensure that management options will have to consider effects on all of these pools.
Managing agricultural soils to store more carbon is likely to have ancillary benefits by reducing soil erosion; the use of cover crops, crop rotations, nutrient management, and organic amendments is likely to increase soil fertility and enhance food security for affected populations. Hence, practices such as improved crop productivity and conservation tillage may be warranted independent of their carbon sequestration benefits. As a result, the opportunity costs of these mitigation strategies may be fairly low. Another relevant activity is the removal of land from agricultural production. Such removal may be warranted for land that generates substantial carbon emissions in its current use or has high potential for carbon storage when it is left undisturbed. The opportunity costs for these set-asides depends on the land's relative productive potential in its current use.
Other reports in this collection