Nitrogen retention in semiarid ecosystems of the United States central grasslands region

Human activities have increased the availability of nitrogen (N) to terrestrial and aquatic ecosystems. Since excess N often leads to NO3 - leaching, soil acidification, and changes in plant species composition, identifying natural sinks for anthropogenic N remains one of the greatest challenges in biogeochemistry. N retention in grasslands has received relatively little attention, due largely to the fact that N deposition in these systems is below levels of N input in most temperate forest ecosystems. However, semiarid grasslands of the U.S. provide good model systems in which to address questions of N retention because they occur in a region where broad climatic patterns influence large scale trends in soil organic matter (SOM) storage. SOM may have an important influence over N retention because it acts as a reduced carbon source for biochemical reactions that stabilize N.; In this dissertation I address the influence of broad and local scale trends in SOM content and composition upon N retention in the laboratory and in the field. I used 15N as a tracer to (1) estimate N immobilization for a series of soils collected from across a SOM gradient, and (2) to estimate N retention in field plots constructed across this gradient. I evaluated our understanding of these regional patterns in SOM content, with particular attention to soil C:N ratios, using the CENTURY SOM model.; I found that regional patterns in soil C content accounted for the largest proportion of variance in N immobilization and N retention. Fine scale spatial variability associated with discontinuous plant cover and local scale differences in soil texture had significant, but much less clear, influences upon N immobilization and N retention than did soil C content. Soil C content alone accounted for 60% and 56% of the variance in estimates of N immobilization and N retention, respectively. Simulation exercises suggest that our understanding of SOM dynamics cannot explain observed trends in soil C:N ratios. This may be due to the CENTURY model's overestimation of an ecosystem's capacity to utilize excess N, and a lack of information regarding trends in SOM composition, belowground production and root turnover.