Nitrate leaching in soil: Tracing the NO3− sources with the help of stable N and O isotopes

Legumes increase the plant-available N pool in soil, but might also increase NO3− leaching to groundwater. To minimize NO3− leaching, N-release processes and the contribution of legumes to NO3− concentrations in soil must be known. Our objectives were (1) to quantify NO3−-N export to >0.3 m soil depth from three legume monocultures (Medicago x varia Martyn, Onobrychis viciifolia Scop., Lathyrus pratensis L.) and from three bare ground plots. Furthermore, we (2) tested if it is possible to apply a mixing model for NO3− in soil solution based on its dual isotope signals, and (3) estimated the contribution of legume mineralization to NO3− concentrations in soil solution under field conditions. We collected rainfall and soil solution at 0.3 m soil depth during 1 year, and determined NO3− concentrations and δ15N and δ18O of NO3− for >11.5 mg NO3−-N l−1. We incubated soil samples to assess potential N release by mineralization and determined δ15N and δ18O signals of NO3− derived from mineralization of non-leguminous and leguminous organic matter.Mean annual N export to >0.3 m soil depth was highest in bare ground plots (9.7 g NO3−-N m−2; the SD reflects the spatial variation) followed by Medicago x varia monoculture (6.0 g NO3−-N m−2). The O. viciifolia and L. pratensis monocultures had a much lower mean annual N export (0.5 and 0.3 g NO3−-N m−2). The averaged NO3−-N leaching during 70 days was not significantly different between field estimates and incubation for the Medicago x varia Martyn monoculture.The δ15N and δ18O values in NO3− of rainfall (δ15N: 3.3±0.8‰; δ18O: 30.8±4.7‰), mineralization of non-leguminous SOM (9.3±0.9‰; 6.7±0.8‰), and mineralization of leguminous SOM (1.5±0.6‰; 5.1±0.9‰) were markedly different. Applying a linear mixing model based on these three sources to δ15N and δ18O values in NO3− of soil solution during winter 2003, we calculated 18–41% to originate from rainfall, 38–57% from mineralization of non-leguminous SOM, and 18–40% from mineralization of leguminous SOM.Our results demonstrate that (1) even under legumes NO3−-N leaching was reduced compared to bare ground, (2) the application of a three-end-member mixing model for NO3− based on its dual isotope signals produced plausible results and suggests that under particular circumstances such models can be used to estimate the contributions of different NO3− sources in soil solution, and (3) in the 2nd year after establishment of legumes, they contributed approximately one-fourth to NO3−-N loss.