Lacustrine cave carbonates: Novel archives of paleohydrologic change in the Bonneville Basin (Utah, USA)

Abstract

Records of past changes in lake levels and lake water isotopic composition in closed basins provide key insights into past variations in the hydrological cycle; however, these records are often limited by dating precision and temporal resolution. Here we present data from lacustrine cave carbonates, a previously unexplored class of carbonates that comprise a promising new archive of past hydrologic changes in the Bonneville Basin of the northeastern Great Basin (USA). These dense carbonates precipitated within caves, crevices, and other protected spaces flooded by Lake Bonneville during its highstand in the last glacial period. We report on deposits in Cathedral and Craners caves, located ∼50km apart at similar elevations approximately 100m above the modern Great Salt Lake and almost 200m below Lake Bonneville’s highstand shoreline. Carbonates from the two caves show similar chronologies, mineralogical transitions, isotopic compositions, and uranium concentrations. These findings suggest that lacustrine cave carbonates record changes in lake level and in the isotopic composition and chemistry of lake water. Importantly, the deposits can be precisely dated by U–Th methods, providing the first records of Lake Bonneville’s water balance changes tied to precise U–Th ages. Close agreement between paired U–Th and calibrated 14C ages in the deposits suggests a minimal (<200a) carbon reservoir effect in the lake and allows 14C dating to be used for age control in portions of the deposits less suitable for U–Th dating.We use ages for the onset and cessation of lacustrine cave carbonate deposition to offer new constraints on past changes in lake level and the carbonate saturation state of lake water. We also present precisely dated, high-resolution oxygen and uranium isotope records from the deposits. Within a first phase of deposition reflecting the lake’s transgression between 26 and 18ka, our isotopic data suggest a large influx of freshwater during Heinrich Stadial 2. A hiatus in deposition beginning at 18.2±0.3ka may be the result of freshening related to the lake’s overflow. Calcite deposition resumes at Cathedral Cave at 16.4±0.2ka, suggesting that basin overflow had ceased by this time and that the lake’s calcite saturation state had increased, and δ18O values increase markedly after 15.9ka, consistent with drying at this time. These data imply that the lake’s deglacial regression began well before the Bølling warming. Cessation of this second phase of deposition at 14.7±0.2ka, coincident with the Bølling warming, may reflect the lake’s drop below Cathedral Cave’s elevation. A final stage of aragonite deposition between 13.8 and 13.6ka differs from the underlying facies and may not reflect deposition from lake waters filling the cave.