Nitrogen cycling in a semiarid grassland under extreme rainfall conditions

Courtney M. Currier, School of Life Sciences, Arizona State University, Tempe, AZ, USA, Osvaldo E. Sala, School of Life Sciences and School of Sustainability, Arizona State University, Tempe, AZ, USA

In tandem with soil water availability, biogeochemical mechanisms play a crucial role in terrestrial productivity. After water, nitrogen (N) availability is the most important limiting factor for aboveground net primary productivity. The relationship of foliar δ15N and the amount of rainfall the ecosystem receives is a well-established one. Arid sites have higher δ15N ratios than mesic sites, resulting in a negative relationship between δ15N and mean annual precipitation. The ecological explanation is that accumulated losses of N relative to ecosystem N pools are greater in drier sites. As climate and precipitation variability are expected to increase, especially in arid regions, we asked how prolonged shifts in water availability interact with N cycling in a semiarid grassland. Using natural abundances of stable nitrogen isotopes for dominant plants and soils and rainfall manipulation experiments located at the Jornada Basin LTER (NM, USA), we tested the hypothesis that N cycling will interact with water availability further amplifying the openness of the N cycle through time. We found a mean plant community-level foliar δ15N of 2.7‰. For the dominant plant species, Bouteloua eriopoda and Prosopis glandulosa, we found the relationship for δ15N vs. mean annual precipitation to be significantly positive. Contrary to existing spatial models, foliar δ15N increased with precipitation amount for both plant species (B. eriopoda: y = 0.0075x – 0.74, R2 = 0.22, p < 0.05; P. glandulosa: y = 0.0059x + 1.89, R2 = 0.21, p < 0.05). Two hypotheses explain this trend. 1) We expect that vegetation on the wetter end of our imposed MAP range eventually draw their nitrogen from decreased N pools with recalcitrant sources that reflect enriched isotopic signatures; or 2) Asynchronous N supply and demand results in increased rates of N loss. We also considered the temporal dynamics of our experiments, which imposed directional rainfall manipulations in duration ranging from 1 to 16 years. We explore preliminary results from these experiments and discuss a new hypothesis that temporal trends in foliar δ15N operate at different rates than mean isotope ratios across space.