CONSTRAINING THE ONSET OF A RAPID DEGLACIAL CLIMATE CHANGE EVENT IN THE SW US THROUGH HIGH RESOLUTION SPELEOTHEM IMAGING AND ISOTOPIC ANALYSES

Speleothem records spanning intervals of rapid climate change are well suited to constrain the nature and rates of terrestrial responses to such change, and thus may help inform projections of the impacts of 21^st-century warming in drought-prone regions like the SW US. The ~14.7-12.7 ka Bølling-Allerød episode (BA) is globally recognized in a range of deglacial proxy records as a rapid warming “event” associated with glacial retreats, ~20 m SL rise, and changes in wind directions and ocean circulation. The onset of the BA in central Texas is preserved in a 50-cm-long stalagmite (CWN4) as a prominent δ¹⁸O decrease similar to that found in records of meltwater input to the Gulf of Mexico (GoM). Feng et al. (2014) linked δ¹⁸O variations in CWN4 and other SW US speleothems to increased delivery of GoM vs. Pacific-sourced moisture during the BA. Here we assess how central Texas terrestrial environments responded during the onset of the BA, using analyses of growth rate, growth-band stratigraphy, and δ¹⁸O variations in CWN4. CWN4’s age model is constrained by 13 high-precision U-series ages encompassing the BA. Growth band thicknesses, as uniquely imaged by confocal laser scanning fluorescence microscopy, correspond to annual growth rates based on U-series ages, consistent with the inference that the bands are seasonal. Conventional IRMS δ¹⁸O analyses of samples micromilled along the central growth axis reveal the base of the BA as a 1.5‰ negative shift spanning 5.5 mm, with most of the drop occurring within the initial 2.3 mm. The excursion position is replicated by secondary-ion mass spectrometry δ¹⁸O analysis at subannual sampling resolution (10 µm, vs. 50 & 160 µm for IRMS) but at lower precision (± 0.3‰) than IRMS (±0.1‰). Comparison of both data sets shows the negative δ¹⁸O shift was rapid and continuous, without a strong relationship to growth band fabric. Slow growth rates (~15-20 µm/yr) accompanied the main δ¹⁸O decline (spanning ~130±20 yrs) before transitioning to growth rates 3 to 14 times higher (up to ~240 µm/yr) in the overlying 35.7 mm (spanning ~630 yrs). Using growth-band thickness and δ¹⁸O as proxies mainly reflective of moisture amount and moisture source, respectively, we infer that the onset of meltwater influence on the GoM preceded a shift to wetter conditions in central Texas by approximately 130 years.