Discrimination of Nitrogen Sources in Karst Spring Contributing Areas Using a Bayesian Isotope Mixing Model and Wastewater Tracers (Florida, USA)

Abstract

ABSTRACT Many springs in Florida have experienced a proliferation of nuisance algae and alteration of trophic structure in response to increases in nitrate concentration concurrent with rapid population growth and land use intensification beginning in the mid-20th century. While loading targets and remediation plans have been developed by state agencies to address excess nitrogen inputs, further confirmation of the relative contribution of nitrogen sources to groundwater is necessary to optimize the use of resources when implementing projects to reduce nitrogen loads. In the present study, stable isotopes of nitrate and wastewater indicators were used to discriminate sources of nitrogen in wells and springs in central Florida. Sampling was performed in 50 wells at 38 sites and at 10 springs with varying levels of agriculture and urban development. Nitrate isotope values were used to develop Bayesian mixing models to estimate the probability distribution of the contributions of nitrate sources in wells. Prior probabilities for the fractional contribution of each source were adjusted based on land use and density of septic tanks. Sucralose and the Cl:Br mass ratio were used as confirmatory indicators of wastewater sources. In residential areas, mixing model results indicated that fertilizer or mixed fertilizer and wastewater (septic tank effluent and reuse water) were the primary sources, with sucralose detections corresponding to wells with elevated contributions from wastewater. Sources of nitrogen in pasture and field crop areas were primarily fertilizer and manure; however, model posterior distributions of δ15N indicated that manure sources may have been overpredicted. The present study demonstrates the utility of a multi-tracer approach to build multiple lines of evidence to develop locally relevant remediation strategies for nitrogen sources in groundwater.