UNGULATE VS. LANDSCAPE CONTROL OF SOIL C AND N PROCESSES IN GRASSLANDS OF YELLOWSTONE-NATIONAL-PARK

Within large grassland ecosystems, climatic and topographic gradients are considered the primary controls of soil processes. Ungulates also can influence soil dynamics; however the relative contribution of larg e herbivores to controlling grassland soil processes remains largely u nknown. In this study, we compared the effects of native migratory ung ulates and variable site (''landscape'') conditions, caused by combine d climatic and topographic variability, on grassland of the northern w inter range of Yellowstone National Park by determining soil C and N d ynamics inside and outside 33-37 yr exclosures at seven diverse sites. Sites included hilltop, slope, and slope bottom positions across a cl imatic gradient and represented among the driest and wettest grassland s on the northern winter range. We performed two experiments: (1) a 12 -mo in situ net N mineralization study and (2) a long-term (62-wk) lab oratory incubation to measure potential N mineralization and microbial respiration. Results from the in situ experiment indicated that avera ge net N mineralization among grazed plots (3.8 g N . m(-2).yr(-1)) wa s double that of fenced, ungrazed plots (1.9 g N . m(-2).yr(-1)). Mean grazer enhancement of net N mineralization across sites (1.9 g N . m( -2).yr(-1)) approached the maximum difference in net N mineralization among fenced plots (2.2 g N . m(-2).yr(-1)), i.e., the greatest landsc ape effect observed. Furthermore, ungulates substantially increased be tween-site variation in mineralization; grazed grassland, 1 SD = 2.2 g N . m(-2).yr(-1), fenced grassland, 1 so = 0.85 g N . m(-2).yr(-1). I n the long-term incubation, potential microbial respiration and net N mineralization were positively related to total soil C and N content, respectively. There was greater variation in potential respiration and net N mineralization early in the incubation, when labile material wa s processed, compared to late in the incubation, when more recalcitran t substrate was processed, suggesting that between-site variation in l abile organic matter was greater than that of recalcitrant material. H erbivores improved the organic matter quality of soil, increasing the labile fractions and reducing the recalcitrant fractions. Grazers redu ced C respired/N mineralized ratios, an index of microbial N immobiliz ation, by an average of 21%. However, the largest landscape influence on the immobilization index was 13-fold greater than the grazer effect . Given that the greatest landscape influence on in situ net mineraliz ation (2.2 g N . m(-2).yr(-1)) was similar to the average grazer impac t on that rate(1.9 g N . m(-2).yr(-1)), we hypothesize that the landsc ape effect on field N availability was primarily caused by variation i n microbial immobilization, while the grazing effect was primarily due to stimulation of gross mineralization. These results indicate that t he relative importance of ungulates in controlling soil N cycling may be more important than previously suspected for grasslands supporting large herds of migratory ungulates, and that the dominant mechanisms u nderlying the landscape and ungulate influences on soil mineral fluxes may differ.