HYDROTHERMAL WATER-ROCK REACTION MODELING WITH MICROBIAL CONSIDERATIONS: RABBIT CREEK AREA, YELLOWSTONE NATIONAL PARK, WY

Water-rock reactions at depth are the main control on aqueous hydrothermal chemistries of hot springs and other thermal features. Thermophilic microbes living in the hydrothermal system are a secondary control on aqueous hydrothermal chemistries and are expected to have increasing influence as spring temperatures decrease. The Rabbit Creek area of Yellowstone National Park (YNP) is an ideal case study for investigating the geologic and biologic controls on aqueous hydrothermal chemistries due to the proximity of the drill core Y-5 to geochemically diverse hydrothermal features (17.3⁰C to 92.2⁰C ± 0.1⁰C; field pH values 6.50 to 9.60 ± 0.05). The modeling program EQ3/6 was used to compare expected and predicted hydrothermal aqueous chemical speciations of hydrothermal features in the Rabbit Creek area. The expected aqueous chemical speciations were found using EQ3 to speciate initial measured concentrations of dissolved major ions and trace elements from each thermal feature. EQ3/6 was used to predict local water-rock reactions by interacting local meteoric water with the summarized mineralogy of the altered rhyolitic tuff of drill core Y-5. The EQ3/6 model was cooled, depressurized, and calibrated to the aqueous chemistry of a proximal, near boiling spring expected to have negligible microbial activity (based on low extracted DNA yields of ~5 ng DNA/g of sediment). The calibrated EQ3/6 water-rock reaction model was further cooled to the measured temperature of each hydrothermal feature analyzed to predict changes in aqueous chemical speciation. Speciated chemistries of springs were generally similar to predictions from modeled water-rock interactions, but differences increased as temperatures cooled. The EQ3/6 predictions for most springs showed deficiencies in silica, aluminum, and sulfur compared to EQ3 speciations, which could be improved by adding H2S (g) to the system and allowing for supersaturation in the models. Variations in calcium in the thermal features were expected to be a function of plagioclase remaining unsaturated and being more variable in the Y-5 subsurface than the other minerals. Discrepancies in pH between field measurements, EQ3 speciated pH, and EQ3/6 predicted pH values for each feature represent disequilibrium. Part of the disequilibrium in cooler features (\u3c75⁰C) could be due to microbial influence, but sampling time lags and occasional changes from single-stage to continuous steam separation in the Y-5 area most likely imparted the strongest control on pH values. Archaea and bacteria were identified from DNA extracted from the sediment of the hot spring to which the EQ3/6 model was initially calibrated. The microbial community includes H2S (g) producers, sulfur oxidizers, nitrate reducers, arsenite oxidizers, and iron reducers, in addition to a few key organisms that metabolize central carbons or oxidize hydrogen. It is not unlikely that similar organisms inhabit the other thermal features analyzed in this study, but further work would be necessary to constrain those specific microbial communities. An outlier spring at ~74⁰C showed the most unexpected geochemistry and is interpreted as having greater shallow meteoric influence as well as probably more microbial influence. Grizzly Pool (17.3⁰C ± 0.1⁰C, pH of 9.60 ± 0.05) has the least hydrothermal input of the analyzed features and is more affected by evaporation and microbial influence than any other feature