Effects of individual bunchgrasses on potential C and N mineralization of longleaf pine savanna soils1

WEST, J. B. AND L. A. DONOVAN. (Department of Plant Biology, University of Georgia, Athens, GA 30602). Effects of individual bunchgrasses on potential C and N mineralization of longleaf pine savanna soils.3 J. Torrey Bot. Soc. 131:120-125. 2004.-A recent conceptual model of grasslands in the US argues that, because of the discontinuous cover of plants in systems strongly limited by precipitation, the presence or absence of individual plants has significant effects on soil processes, with relatively small effects of species differences. In systems not strongly limited by precipitation, resource limitations are thought to vary in space and time, vegetation is more continuous, and species differences are relatively more important. We ask whether the model can be applied to grass species effects on potential net C and N mineralization in a southeastern US savanna ecosystem. These savannas have very sandy soils, strong soil resource limitations, and discontinuous plant cover, even though they receive 1200 mm yr-' rainfall. Based on the discontinuous herbaceous vegetation, an extension of the model would predict that native perennial bunchgrasses would have strong plant presence effects, and small or no species identity effects on these soil processes. Soils were sampled in a paired fashion, directly under a plant (either Aristida stricta, Schizachyrium scoparium, or Andropogon ternarius) and in adjacent unvegetated locations, and aboveground biomass was collected. Net C-min was significantly higher under plants compared to unvegetated locations, but there were also significant species identity effects, with the greatest rates observed under A. stricta. This pattern is likely explained by the greater biomass of A. stricta, because net C-min was positively related to biomass. For net N-min, there were neither plant presence nor species identity effects. There was, however, a positive relationship between net C-min and net N-min for both S. scoparium and A. ternarius soils, but not for A. stricta, suggesting a subtle, but potentially important, difference among species in their effects on N cycling. The results suggest that individual grasses have significant effects on soil processes in this system, but that the conceptual model developed for grasslands may not generalize to the effects of grasses on the soils of these savannas.