Distance‐based mixing models of δ 18 N N O 3 − and δ 18 O N O 3 − in a marsh‐lined estuary with multiple, distinct NO 3 − sources (Murderkill Estuary, Delaware, USA)

Abstract The Murderkill Estuary (Delaware, USA) receivesNO 3 −principally from its upland watershed and from a wastewater treatment facility. Due to disparateNO 3 −sources, one‐dimensional salinity‐based mixing models were inadequate for describing distributions ofNO 3 −, δ 15 NN O 3 −, and δ 18 ON O 3 −. Distance‐based mixing models with multiple, spatially‐specified inputs were, therefore, applied to describe conservative mixing of these constituents and determine the extent to which biogeochemical reactions lead to non‐conservative behavior ofNO 3 −. These models closely matched Si observations in both winter and summer, consistent with high wastewater silicate loads and light limitation, and serve to validate modeling parameters for both seasons. A close fit of distance‐based models to estuarineNO 3 −observations suggested a lack of uptake and fractionation in early winter. In the summer, modeled predictions ofNO 3 −, δ 15 NN O 3 −, and δ 18 ON O 3 −diverged from estuarine observations, particularly in the oligohaline and polyhaline regions, consistent with in situ nitrogen cycling or additional sources and sinks. Effluent from an adjacent marsh in the lower estuary containedNO 3 −with low δ 15 NN O 3 −and δ 18 O NO3, low DO and highNH 4 +concentrations in late summer. This data and previous studies of adjacent Delaware Bay suggest that reactions in marshes and Bay waters likely drove the non‐conservative behavior ofNO 3 −and its stable isotopes. Potential uncertainty in watershed discharge, however, limited explicit quantitation ofNO 3 −loss in the estuary. Nonetheless, distance‐based models are useful tools for the study ofNO 3 −, δ 15 NN O 3 −and δ 18 ON O 3 −distributions and cycling patterns in complex marsh‐lined estuaries with multipleNO 3 −inputs.