Low retention of added nitrogen at realistic levels of simulated nitrogen deposition in the Northern Chihuahuan Desert

Aridlands are expected to be sensitive to even low levels of nitrogen (N) deposition, yet most experiments apply N in amounts that greatly exceed deposition estimates. We performed a low, but realistic N addition (i.e., based on N deposition estimates) at three adjacent Chihuahuan Desert semi-arid grassland sites. For four years, we applied 2 and 4 kg N ha⁻¹ yr⁻¹. We also included two additional treatments: a labile carbon (C) addition to test whether pre-existing N deposition had already altered our study ecosystem, and a supplemental water treatment prompted by drought during the first two years. In the field, we measured plant responses (cover, community composition, diversity, foliar C and N), soil ecosystem properties (pH, C, N, and P pools), microbial biomass, and extracellular enzyme activities. Additionally, we performed two short-term laboratory incubations: one measuring net N mineralization, net nitrification, and net ammonification, and another measuring nitric oxide (NO) emissions. Overall, our N inputs rarely drove significant changes in our field measures (including plant and soil N pools). Supplemental water did not induce any N addition effects, presumably due to a concurrent drought. However, despite above-average summer rainfall in the final year, N addition responses were still not apparent. In contrast, our incubations demonstrated that N enrichment can suppress net N mineralization and nitrification, while increasing NO emissions. Our field and incubation findings suggest that aridlands may be insensitive to low N deposition levels. This insensitivity may be associated with N losses, which could limit N retention in these ecosystems.