VADOSE CO2 GAS MAY BE MORE IMPORTANT IN FORMING CAVES IN EOGENETIC KARST THAN DISSOLUTION FROM MIXING OF VADOSE AND PHREATIC WATER

Most models of cave formation in eogenetic limestones invoke mixing dissolution. This process has been suggested to form caves by mixing of waters with different ionic strengths, each having equilibrated with calcite at a different _pCO₂. Dissolution occurs because reactant concentrations vary linearly but saturation state is controlled by a power law. Mixing dissolution is invoked to form caves where fresh and saline waters mix (e.g. coastal aquifers) or where vadose and phreatic waters mix (e.g. water tables). An alternative explanation for cave formation in eogenetic karst aquifers that has not been as widely considered is that spatially or temporally heterogeneous inputs of vadose CO₂ gas to water at water tables may form caves. To determine if mixing dissolution or inputs of CO₂ gas was driving dissolution at Briar Cave, a water table cave in the eogenetic upper Floridan aquifer in central Florida, we made high resolution time series measurements of specific conductivity (SpC), temperature and meteorological data, and collected synoptic water samples, over one year. SpC, _pCO₂ and degree of undersaturation with respect to calcite increased gradually through summer, when Briar Cave experienced less ventilation by outside air, and decreased through winter, when more ventilation occurred. Broad seasonal trends in SpC, _pCO₂ and calcite undersaturation coincided with changes in the intensity of cave ventilation by outside air rather than with discrete recharge events, suggesting variable inputs of CO₂ associated with variability in cave ventilation were driving dissolution rather than mixing. We hypothesize dissolution occurs when water flows from regions of eogenetic aquifers with low _pCO₂ into regions with elevated _pCO₂. Elevated _pCO₂ would be promoted by fractures connecting the soil zone to water tables. Water flowing from regions of low _pCO₂ into regions of high _pCO₂ would dissolve CO₂ from the atmosphere, reducing pH, and dissolving calcite. Simple geochemical models demonstrate small gradients in _pCO₂ along flowpaths are an order of magnitude more efficient at dissolving limestone than mixing of vadose and phreatic water, which, on its own is unlikely to form caves in realistic time scales.