Anthropogenic addition of bioavailable nitrogen to the biosphere is increasing1,2 and terrestrial ecosystems are becoming increasingly nitrogen-saturated3, causing more bioavailable nitrogen to enter groundwater and surface waters4,5,6. Large-scale nitrogen budgets show that an average of about 20–25 per cent of the nitrogen added to the biosphere is exported from rivers to the ocean or inland basins7,8, indicating that substantial sinks for nitrogen must exist in the landscape9. Streams and rivers may themselves be important sinks for bioavailable nitrogen owing to their hydrological connections with terrestrial systems, high rates of biological activity, and streambed sediment environments that favour microbial denitrification6,10,11. Here we present data from nitrogen stable isotope tracer experiments across 72 streams and 8 regions representing several biomes. We show that total biotic uptake and denitrification of nitrate increase with stream nitrate concentration, but that the efficiency of biotic uptake and denitrification declines as concentration increases, reducing the proportion of in-stream nitrate that is removed from transport. Our data suggest that the total uptake of nitrate is related to ecosystem photosynthesis and that denitrification is related to ecosystem respiration. In addition, we use a stream network model to demonstrate that excess nitrate in streams elicits a disproportionate increase in the fraction of nitrate that is exported to receiving waters and reduces the relative role of small versus large streams as nitrate sinks.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $3.90 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Funding for this research was provided by the National Science Foundation. We thank N. Ostrom for assistance with stable isotope measurements of N2 and N2O, and W. Wollheim for initial development of the model that we modified to estimate denitrification rates from field data. We thank M. Mitchell, B. Roberts and E. Bernhardt for their comments on earlier versions of the paper. We thank the NSF LTER network, US Forest Service, National Park Service and many private landowners for permission to conduct experiments on their lands. Partial support to P.J.M. during manuscript preparation was provided by the US Department of Energy, Office of Science, Biological and Environmental Research under contract with UT-Battelle.
Author Contributions P.J.M. coordinated the stream 15N experiments and analysed the compiled experimental data sets. A.M.H. and G.C.P conducted the stream network modelling. P.J.M., A.M.H. and G.C.P. wrote major portions of the manuscript. S.K.H. established sampling protocols and coordinated the 15N analysis of dissolved N2 samples. Except for A.M.H., all authors listed to J.R.W. were joint project Principal Investigators and contributed to the conceptual and methodological development of the project and analysis of data. Authors listed from C.P.A. to S.M.T. coordinated field experiments and analysed data from one or more biomes. All authors discussed the results and commented on the manuscript.
This file contains Supplementary Methods and Supplementary Discussion with additional references; Supplementary Figures and Legends 1-3; and Supplementary Tables 1-3. The Supplementary Methods include more detailed 15N experiment methods and stream network model development. The Supplementary Discussion presents NO3- removal over a standardized stream reach, effects of land use and stream size on uptake rates, and limitations of the data and stream network modeling results. The Supplemental Figure 1 shows study site locations, Figure 2 shows the stream network model structure, and Figure 3 shows the location and hydrography of the Little Tennessee River network used in model simulations. The Supplementary Table 1 presents physical and chemical characteristics and NO3- uptake and denitrification rates determined for all streams in the 15N experiments, Table 2 includes definitions of model terms, and Table 3 presents the methods used to derive model parameters.