Investigating spatial and temporal variability of environmental and biological controls on riparian soil denitrification

Soil denitrification is a critical component of nitrogen (N) cycling on Earth. It is a microbially-mediated process that removes N from soils by reducing nitrate, a highly bioavailable molecule and significant contributor to eutrophication, to gaseous forms of N (dinitrogen gas or nitrous oxide). Although denitrification provides a critical ecosystem service by reducing N inputs to streams and rivers, it is highly variable in time and space, especially in riparian areas. Accordingly, to accurately predict when and where riparian soil denitrification occurs and contributes significantly to N removal, it is critical to characterize variability in the key controls on this process. We therefore addressed the heterogeneity of riparian soil oxygen, a major control on denitrification, using a machine-learning approach to pinpoint suites of soil conditions that lead to contrasting oxygen regimes. Seasonal and spatial variability in denitrifier community composition and soil conditions was then assessed using a cutting-edge DNA sequencing technology. Finally, links between environmental and biological controls on denitrification rates were explored during the spring snowmelt period. The results highlight problems associated with predicting soil oxygen solely based on soil moisture. The findings also confirm that soil redox, soil moisture, and oxygen conditions shape denitrifier community composition, and suggest that site-specific features impact links between the biotic community and denitrification rates.