Environmental controls on nitrogen cycling in northern Chihuahuan desert soils

This dissertation attempts to estimate the magnitude of inputs and losses of nitrogen from soils in the Jornada basin in the Chihuahuan desert of southern New Mexico. An acetylene reduction assay was used to measure nitrogenase activity (N{dollar}sb2{dollar} fixation) by soil microorganisms. N{dollar}sb2{dollar} fixation inputs were low (estimated at {dollar}<{dollar}1 g ha{dollar}rmsp{lcub}-1{rcub} ysp{lcub}-1{rcub}),{dollar} with the highest rates detected in tarbush shrubland and playa soils. Nitrogenase activity was greatest at high levels of soil moisture and carbon availability and inhibited by inorganic nitrogen. A strong linear relationship between the ratio of available nitrogen to phosphorus and nitrogenase activity was not found in Jornada soils.; The production and loss of nitric oxide (NO) by nitrifying bacteria was strongly regulated by soil moisture and NH{dollar}sb4sp+{dollar} availability. Nitrifiers responded rapidly to watering (10 minutes). NO losses were highest from NH{dollar}sb4sp+{dollar}-fertilized black grama grassland soils (69 ng cm{dollar}rmsp{lcub}-2{rcub} hsp{lcub}-1{rcub}).{dollar} In situ measurements at 15 sites within the Jornada basin revealed low NO emission from untreated soils, with higher losses detected after rain (max. 4 ng cm{dollar}rmsp{lcub}-2{rcub} hsp{lcub}-1{rcub}).{dollar} Soil moisture derived from runon may sustain NO emission from playa soils for several days after rainfall.; Grassland soils showed net N{dollar}sb2{dollar}O consumption, and N{dollar}sb2{dollar}O fluxes were not affected by increased soil moisture availability. Watering had no immediate effect on N{dollar}sb2{dollar}O fluxes from soils between or adjacent to shrubs, but increased production under shrubs after 4 daily 1-cm watering treatments. Fluxes were also higher under shrub canopies than between shrubs 1-hr after a 1-cm watering treatment. The highest N{dollar}sb2{dollar}O fluxes from wet soils in the field (max. 0.4 ng cm{dollar}rmsp{lcub}-2{rcub} hsp{lcub}-1{rcub}){dollar} were an order of magnitude lower than past estimates of denitrification in lab-incubated soils.; Extrapolated estimates of N trace inputs and losses for the Jornada LTER suggest that soil nitrogen fixation ({dollar}<{dollar}1 g N ha{dollar}rmsp{lcub}-1{rcub} ysp{lcub}-1{rcub}){dollar} is negligible compared to atmospheric deposition.; Mean atmospheric deposition (1988-93) is 0.8 kg N ha{dollar}rmsp{lcub}-1{rcub} ysp{lcub}-1{rcub}{dollar} in dryfall and 1.7 kg ha{dollar}rmsp{lcub}-1{rcub} ysp{lcub}-1{rcub}{dollar} in wetfall. N{dollar}sb2{dollar}O losses from Jornada soils are higher than NO and NH{dollar}sb3{dollar} losses, using the flux estimates derived from lab-incubated soils. Annual N trace gas losses are higher from shrubland ecosystems than from grassland (6.2 vs. 3.9 kg N ha{dollar}rmsp{lcub}-1{rcub} ysp{lcub}-1{rcub}).{dollar} Total N gas losses from the basin may have increased over the past century, due to the expansion of shrubland.; A conceptual model of N trace gas losses from a desert basin is proposed which relates N trace gas emission to soil resource availability. During rain storms, soil organic matter, inorganic nutrients and fine soil particles are removed from upland areas in overland flow. The underlying hypothesis of the model is that soil N cycling--and therefore N trace gas emission--is highest in topographic depressions, where soil moisture and nutrients accumulate. The relative emission of N{dollar}sb2{dollar}O and NO may be higher in low-lying areas, due to the limited diffusion of NO in fine-textured soils.