Martens, Dean (USDA-ARS Tucson AZ, USDA-ARS Soutwest Watershed Research Center, Tucson, AZ, 85719; Phone: 520-670-6380 x156; Fax: 520-670-5550; Email: dmartens@tucson.ars.ag.gov)

Rapid Loss of Soil Active C Pools Fuel Depletion of Soil C Following Mesquite Removal in Semiarid Grasslands

D.A. Martens *, M. McClaran

Grasslands or savannas are one of the most widespread biomes on earth, covering about 40% of the terrestrial land surface. The proliferation of woody plants on grasslands has prompted the deployment of brush management techniques in attempts to improve grasslands. Recent evidence suggests that enhanced woody growth results in a major biomass and soil C sink and may provide financial incentives to promote C sequestration. In semiarid grasslands, soil C contents are typically greater beneath mesquite trees than open grasslands, but little is known about the rates of accumulation during tree growth or depletion following removal. We estimated these rates by comparing C content of soil and physical fractions (light, POM, MOM, silt and clay) among three settings: beneath old trees (100 yr, n = 4), 40 yr skeletons of herbicide-treated old trees (n = 2), and open grasslands (n = 6). All areas had sandy loam soils, and were between 950-1100 m elevation on the Santa Rita Experimental Range, Tucson AZ. To determine a soil profile C composition under the different treatments, soil and bulk density samples were taken in 2002 from 0-5, 5-10, 10-30, 30-50 cm depths. From the 0-10 cm soil depth, C accumulation was greater under the old trees (0.91 kg m^-2) compared to the skeletons (0.62 kg m^-2) and the grass sites (0.55 kg m^-2). Physical fractionation with density separation (1.85 g cm^-2) found that the active/passive C ratio was greater under the mesquite (0.86) compared with skeletons (0.64) and grass (0.69) sites. From the 10-30 cm depth, C accumulation was similar under the old trees (0.81 kg m^-2) compared to the skeletons (0.81 kg m^-2), but greater than the grass sites (0.73 kg m^-2). Determination of active and passive C pools found that the active/passive ratio was greater under the mesquite (0.29) compared with skeletons (0.16) and grass (0.26). From the 30-50 cm depth, C content was similar under the old trees (0.78 kg m^-2) compared to the skeletons (0.75 kg m^-2) and the grass sites (0.77 kg m^-2). The active/passive ratio was greater under the mesquite (0.18) compared with stumps (0.09) and grass (0.10). The mesquite sites had a greater soil C content of 0.31 kg C m^-2> compared with the skeletons and 0.45 kg C m^-2 more C than the grass sites in the 50 cm profile, but the majority of the “sequestered” C was in the litter, light and POM fractions. Average rates of C accumulation under the mesquite trees were 5 g C m^-2 yr^-2 (50 cm depth) while average rates of C loss following removal were 8 g C m^-2 yr^-2 (50 cm depth). This rapid rate of decomposition is fueled by the relatively high proportion of active fraction C in the accumulated C beneath old mesquite compared to open grasslands. The results suggest that large-scale removal of mesquite from this ecosystem will result in rapid mineralization of the active C pools. Thus, potential financial incentives for maintaining woody plants would provide an alternative to limit large-scale clearing and the rapid C mineralization, which could act as a positive feedback to climate warming.