Assessing carbon dynamics in semiarid ecosystems : Balancing potential gains with potential large rapid losses

Photosynthesis and respiration are the largest fluxes into and out of the biosphere (Molles 1999). Consequently, small changes in these fluxes can potentially produce large changes in the storage of carbon in the biosphere. Terrestrial carbon fluxes account for more than half of the carbon transferred between the atmosphere and the earth's surface (about 120 GigaTons/year), and current stores of carbon in terrestrial ecosystem are estimated at 2060 GigaTons. Increasing attention is being focused on the role of managing and sequestering carbon in the terrestrial biosphere as a means for addressing global climate change (IGBP, 1998; U.S. Department of Energy, 1999). Terrestrial ecosystems are widely recognized as a major biological scrubber for atmosphereic CO{sub 2} and their ability to finction as such can be increased significantly over the next 25 years through careful manipulation. The potential for terrestrial carbon gains has been the subject of much attention (Dixon et al., 1994; Masera et al. 1997; Cao and Woodward, 1998; DeLucia et al. 1999). In contrast to other strategies for reducing net carbon emissions, terrestrial sequestration has the potential for rapid implementation. Strategies that focus on soil carbon are likely to be effective because in addition to being a storage pool of carbon, soil carbon also improves site productivity through improving soil quality (e.g., water retention and nutrient availability). The carbon pool in soils is immense and highly dynamic. The flux of carbon into and out of soils is one of the largest uncertainties in the total mass balance of global carbon (NRC, 1999; La1 et al., 1998; Cambardella, 1998). Reducing these uncertainties is key to developing carbon sequestration strategies. Soil carbon pools have been greatly depleted over recent centuries, and there is potential to increase storage of carbon in these soils through effective land management. Whereas carbon in vegetation can be managed directly through land use, carbon in soils generally must be managed indirectly through manipulation of vegetation and nutrients. Land management as well as climate changes have the potential to increase soil carbon, but also could trigger large soil carbon losses. Recently, the importance of accounting for countervailing losses in assessing potential amounts of terrestrial carbon that can be sequestered has been highlighted (Schlesinger, 1999; Walker et al., 1999). Realistic assessment of terrestrial carbon sequestration strategies must consider net results of an applied strategy, not simply projected carbon gains. In addition, large, rapid losses of carbon resulting from carbon management strategies could exacerbate the global warming rather than mitigating it. Such potential losses include rapid loss of carbon in vegetation due to fire and rapid loss of soil carbon triggered by reductions in ground cover (e.g., fire, drought). Therefore, strategies for terrestrial carbon sequestration must determine how to increase terrestrial carbon while minimizing the risk of large-scale catastrophic losses. Our objectives in this paper are to (1) highlight approaches that are being considered in terms of terrestrial carbon sequestration, (2) highlight case studies for which large losses of carbon may occur, and (3) suggest future directions and application for terrestrial carbon sequestration