TRACING CHLORIDE SOURCES IN URBAN WATERWAYS IN AUSTIN, TEXAS USING STABLE CHLORINE ISOTOPES

Salt concentrations in Austin-area streams that recharge the Barton Springs Segment of the Edwards Aquifer have been increasing over the past two decades. The aquifer discharges at springs comprising habitat for endangered species sensitive to water quality. Chemical deicers are not commonly used in the region, unlike watersheds in the northern U.S., suggesting input from alternative salt sources (e.g., wastewater, fertilizer). Furthermore, the limited degree to which Na and Cl concentrations are diluted during intervals of high stream discharge, suggest the presence of salt sources that are readily weathered during rain events. To investigate the possible salt inputs to these urban waterways, we use chlorine isotopes to characterize potential chloride inputs. Thirty samples, including anthropogenic waters (i.e., wastewater, municipal drinking water, swimming pool water), natural waters (i.e., cave drip water and stream water from protected and rural watersheds), and raw anthropogenic materials (i.e., pool chlorinators, fertilizers), were analyzed on a ThermoElectron MAT 253 isotope ratio mass spectrometer. All samples were analyzed with respect to Standard Mean Ocean Chloride with an uncertainty of ±0.2‰. Stream water from protected and rural watersheds has δ³⁷Cl values from -0.2 to -0.3‰. Treated and untreated wastewater and septic water range from -1.0 to -0.3‰; there was no systematic offset between treated and untreated wastewater. Swimming pool water, bleach, and municipal drinking water have values that range from 0.0 to +0.7‰. Fertilizers are the most variable of the materials sampled, ranging from -1.8 to +12.7‰, which suggests that variable amounts of fractionation occur during the manufacturing process. Urban waterways have δ³⁷Cl values of -0.6 to +0.4, variability outside of analytical uncertainty. The sample from the area with greatest urban density has a value of +0.4‰. One urban stream had a low δ³⁷Cl value (-0.6‰) during dry conditions and higher δ³⁷Cl value (0.1‰) during wetter conditions, suggesting chloride inputs vary with changing hydrological conditions. Our preliminary results suggest that variations in δ³⁷Cl values in the Austin-area waterways are small relative to the large amount of variation that exists in δ³⁷Cl values of potential chloride inputs, especially fertilizers.