LAUSR

Abstract Karst systems are useful for examining spatial and temporal variability in Critical Zone processes because they provide a window into the subsurface where waters have interacted with vegetation, soils, regolith, and bedrock across a range of length and timescales. These hydrologic pathways frequently include the precipitation of speleothems, which provide long-term archives of climate and environmental change. Trace element ratios in speleothems (Mg/Ca, Sr/Ca, Ba/Ca) have the potential to provide information about past changes in rainfall and infiltration, but controls on them can be complex and their interpretation must be based on an understanding of the modern cave system. Here we integrate observations of surface conditions, bedrock, soil, and drip water chemistry of Blue Spring Cave in Tennessee, USA with the reactive transport model CrunchTope, which we have calibrated for karst systems to investigate the primary controls on trace element variations in cave seepage waters. We find that measured drip water Mg/Ca and Sr/Ca are captured within the model through variable amounts of limestone dissolution followed by precipitation of secondary calcite that happens within the cave rather than the host limestone. However, strong spatial controls on drip water Mg/Ca and Sr/Ca likely reflect seepage water interactions with variable amounts of diagenetic phases in the host rock. In contrast, Ba/Ca values are consistent across the cave and vary with effective rainfall, suggesting that this parameter may be the most consistent metric for limestone dissolution and prior calcite precipitation and can act as a proxy for rainfall and infiltration in this cave system. Our findings emphasize the importance of evaluating spatial heterogeneity in cave drip waters and outline a novel modeling approach for determining the dominant controls on drip water chemistry in support of the interpretations of paleoclimate records.