DISENTANGLING SOURCES OF BASEFLOW (AND SALINITY) IN THE HEADWATERS OF AN AGRICULTURALLY-DOMINATED WATERSHED USING MULTIPLE ISOTOPIC AND GEOCHEMICAL TRACERS (Invited Presentation)

The installation of tile drains has been a common practice in the agricultural landscape of the American Midwest for nearly 150 years, yet our understanding of the impact of these drainage modifications on recharge processes, the exchange of water between shallow and deep aquifers, and baseflow generation is incomplete. To address this knowledge gap, a synoptic baseflow sampling campaign was initiated in the Wabash River watershed in 2015. This campaign revealed anomalously high salinity (chloride) coincident with high concentrations of sodium, potassium, strontium, sulfate, and fluoride in three river sites located along the headwaters reach from Fort Recovery, OH to the Forks of the Wabash near Huntington, IN. These observations pose two interrelated questions: 1) what are the sources of the high salinity and TDS, and 2) what insight does the observed salinity provide into baseflow generation in the headwaters of the Wabash River? The three headwaters sites have low ³⁶Cl/Cl ratios ranging from 27.5 (x 10^-15) to 33.4 (x 10^-15) that are not consistent with the ³⁶Cl/Cl of monthly precipitation in 2015. Instead, their ³⁶Cl/Cl ratios are indicative of ³⁶Cl dilution due to mixing with a source of old Cl^-, most likely fertilizers. These three sites also have elevated Cl^-/Br^- ratios and high nitrate concentrations which support an anthropogenic source of salinity, thereby suggesting that baseflow is supported by water that is impacted by agricultural practices, perhaps shallow aquifers. However, the ⁸⁷Sr/⁸⁶Sr ratios of the three headwaters sites range from 0.708530 (+/- 0.000025) to 0.708904 (+/- 0.000025) and are similar to samples of Silurian carbonate bedrock (0.708292 +/- 0.000047) that is located beneath the unconsolidated sediment. In addition, the strontium, sulfate, and fluoride concentrations can largely be explained by dedolomitization and gypsum dissolution, thereby suggesting that baseflow is supported by water from the regional carbonate aquifer. Our evolving conceptual model is that baseflow is supported by a mixture of nitrate-rich groundwater from the shallow aquifer (that is impacted by agricultural practices) and relatively nitrate-poor groundwater from the regional aquifer. The contribution from the regional aquifer is not insignificant, however without the aid of multiple isotopic tracers, the hydrogeologic connection between the headwaters of the Wabash River and the regional Silurian-Devonian carbonate aquifer would have remained masked by solutes derived from anthropogenic sources.