Carbon in karst: investigating sources, transport and isotopic fractionation to better inform the interpretation of speleothem climate records of central Texas caves

Speleothems (cave calcite deposits) are a well-recognized terrestrial paleoclimate proxy. The most commonly constructed records from speleothem geochemical data are made from the measurement of stable oxygen isotope ratios (δ18O). This is most often accomplished by the analysis of CO2 on a gas-source IRMS, also allowing for the simultaneous measurement of stable carbon isotope ratios (δ13C). Variations in speleothem δ18O values have long been attributed to changes in temperature and/or precipitation. By contrast, speleothem δ13C values are more difficult to interpret; explanations to account for these variations include both 1) above-cave processes, such as the relative proportion of C3 versus C4 vegetation and climate change, and 2) below-ground processes, such as the amount of prior calcite precipitation associated with epikarst flow paths into the cave. It is the purpose of this study to understand and quantify the factors influencing speleothem δ13C values from soil horizon to ultimate precipitation within a cave, and offer new insight to using these values as proxy information. Our study focuses on active monitoring sites at Inner Space Cavern and Natural Bridge Caverns, both located in the Edwards Plateau of central Texas. We collected soil gas and drip water from each cave every 2-6 weeks. At Natural Bridge Caverns during summer months, mean δ13C values of soil CO2 underlying C4 grasses were -14.0% (±0.3%) and underlying Juniper trees (C3 plants) were -19.1% (±0.2%). Measured δ13C values of drip water DIC collected during the summer averaged -10.9% (±0.08%). Given an expected fractionation factor between CO2 and bicarbonate (the primary species of dissolved inorganic carbon in these waters) of 7.9% at 25°C, the calculated δ13C value of soil gas in equilibrium with the measured drip water DIC is -18.8% (±0.08%), indistinguishable from the δ13C values of soil CO2 beneath juniper trees. We suggest that drip water δ13C values are controlled by juniper root respiration and are largely unaltered by factors suc