Biogeochemical and ecosystem properties in three adjacent semiarid grasslands are resistant to nitrogen deposition, but sensitive to edaphic variability

Background/Question/Methods

Drylands, which constitute >40% of Earth’s land area, have naturally low nitrogen stocks and are predicted to be sensitive to modest increases in reactive nitrogen availability, but direct evidence that atmospheric nitrogen deposition will have sustained effects on dryland ecosystems is sparse and conflicting. We used data from three long-running nitrogen deposition simulation experiments and a complementary laboratory incubation experiment to address three fundamental questions about how nitrogen inputs affect dryland ecosystems: 1) What are the short- and long-term biogeochemical consequences of nitrogen inputs?; 2) Do these consequences depend on soil moisture availability?; and 3) Does soil texture modify the effects of nitrogen inputs and/or soil moisture availability? In 2011, we established three study sites along a soil texture gradient in Arches National Park, USA with plots receiving 0, 2, 5, or 8 kg N ha-1 annually (n = 5). A suite of soil biogeochemical metrics was assessed over the long- and short-term and we compared foliar chemistry, soil extracellular enzyme activities, heterotrophic respiration rates, and nitrogen trace gas fluxes in the fertilization plots at select intervals during the study period (2011-2019). Finally, we conducted a laboratory incubation to measure the effects of soil moisture on heterotrophic respiration rates.

Results/Conclusions

We identified some short-term effects in situ, but no lasting consequences of added nitrogen for any of the metrics measured. In the incubation experiment, soil moisture treatments had independent effects on heterotrophic respiration rates but did not modify the short-term effects of added nitrogen. In contrast to nitrogen treatments, subtle differences in soil texture were associated with large differences in biogeochemical cycling. Our results oppose the common prediction that coupled dryland biogeochemical cycles are highly sensitive to nitrogen. Instead, they suggest that available nitrogen inputs are lost from some drylands before they can be immobilized or physically bound to soil, and that fine scale edaphic heterogeneity is a key driver of dryland biogeochemical and ecosystem responses.