Streams are hotspots for nitrogen (N) processing in the landscape, and the degree to which they remove N from the water column should relate to rates of carbon (C) cycling. Here we examine results from the two Lotic Intersite Nitrogen eXperiments (LINX), which were coordinated multi-site studies examining factors controlling inorganic N uptake in streams draining multiple North American biomes. We then relate N cycling results to metabolism metrics reflecting activity of the carbon cycle. In LINXI, we conducted 6-week 15N-ammonium tracer additions in 11 relatively pristine streams and for LINXII, we selected 72 streams from 8 biomes and varying land use (e.g. reference, urban, and agriculture) and used 24-hr 15N-nitrate tracer additions to quantify nitrate removal. Streams were small (discharge <300 L s-1) and spanned a range of nitrate concentrations, geomorphology, and metabolism. We explored how gross primary production (GPP) and community respiration (CR) differentially controlled assimilatory nitrate uptake and denitrification.
Results/Conclusions
In LINXI, ammonium uptake was unrelated to whole-stream metabolism, likely due to the small GPP range across sites. In contrast, estimated biological N demand based on C metabolism correlated well with measured N uptake. We found evidence for metabolic compensation (e.g. where GPP was high, CR was low) resulting in relatively constant rates of N uptake across the 11 streams; therefore factors changing the balance between autotrophic and heterotrophic metabolism indirectly controlled N uptake. Subsequently we asked whether streams draining agricultural and urban land use would also exhibit metabolic compensation, or would the linkage between metabolism and N cycling be decoupled due to metabolic variation and increased nutrient inputs found at LINXII sites influenced by human land use. Results from LINXII 15N-nitrate additions allowed the partitioning of assimilatory nitrate uptake from denitrification. Using structural equation modeling we found that autotrophic assimilatory demand dominated nitrate removal and was best predicted by GPP, while denitrification was related to CR. Although human land use did not affect N uptake directly, land use increased nitrate availability and GPP indirectly influenced nitrate uptake and denitrification respectively. Small streams can have high N transformation rates, and metrics reflecting the C cycle were the overriding biological variables predicting N cycling rates. Therefore, actions altering C cycling will indirectly influence N cycling. The 15N tracer approach has been particularly useful because it allows for the partitioning of multiple components of the N cycle, which appear to differentially relate to components of the C cycle.