Evaluation of isotopic fractionation error on calculations of marine-derived nitrogen in terrestrial ecosystems

Pacific salmon (Oncorhynchus spp.) transport nitrogen (N) from oceans to inland ecosystems. Salmon delta15N is higher than delta15N expected in terrestrial plants, so linear two-source mixing models have commonly been used to quantify contributions of marine-derived N (MDN) to riparian ecosystems based on riparian plant delta15N. However, isotopic fractionation potentially contributes to error in MDN estimates by changing delta15N of salmon-derived N appearing in soil and plants. We used a simulation model to examine potential effects of fractionation on MDN estimates. We also measured changes in delta15N and delta13C as N and carbon (C) moved from bear feces into soil, and compared MDN estimates using three different estimates for the marine endmember of a linear mixing model. Simulation demonstrated that fractionation during soil N losses could lead to large overestimations of MDN when delta15N of salmon tissue is used as the marine endmember. delta15N of bear feces was significantly enriched (by 1.9 per thousand) relative to salmon tissue, but did not change during movement of feces-derived N into soil. In contrast, delta13C decreased by 1.9 per thousand between salmon and bear feces and declined an additional 4.2 per thousand during movement into soil. We propose a new method for estimating the delta15N of the marine endmember that accounts for isotope fractionation occurring as marine N is cycled in soil. This method uses the proportional difference in soil 15N content between reference and spawning sites to calculate the marine endmember delta15N.