A multiple cave deposit assessment of suitability of speleothem isotopes for reconstructing palaeo-vegetation and palaeo-temperature

The suitability of speleothems for interpreting palaeoclimate is typically determined by using either the Hendy Test, overlapping analysis or long-term cave environment monitoring. However, in many cases, these methods are not applicable, because a speleothem lacks clearly traceable layers for the Hendy Test, it is difficult to obtain an overlapping speleothem nearby, or long-term cave monitoring is impractical. The authors propose a multiple cave deposit approach to assess the suitability of speleothems for palaeoclimate study. Speleothems collected from two sites within Raccoon Mountain Cave, Tennessee (USA) exhibit remarkable spatial variation (δ¹³C: -10·3‰ to -2·2‰) over a relatively short distance (ca 260 m). Drip water δ¹⁸O values exhibit a seasonal precipitation signal at Site 1 and an annual signal at Site 2. Combining field observations, water isotope analysis and trace-element data, the authors propose that the speleothem formation at Site 1 and Site 2 tapped distinct sources of CO₂: (i) CO₂ derived from overlying soils for Site 1; and (ii) limestone dissolved inorganic carbon induced by ground water dissolution for Site 2. Using fresh cave deposits (modern speleothem) δ¹³C (100% C3 vegetation) as an analogue, a simple model was developed to estimate land surface vegetation for speleothems. The speleothem formation temperature estimated using fresh cave deposit δ¹⁸O values generally reflects the mean annual temperature in this region. This study indicates that spatial variations in carbon isotopes could be caused by different carbon sources dominating in different parts of the cave, which should be taken into consideration by researchers when using speleothem δ¹³C values to reconstruct temporal palaeo-vegetation changes. This study demonstrates a practical sampling strategy for verifying suitability of speleothems for palaeo-vegetation and palaeo-temperature reconstructions by analysing multiple cave deposits, especially for cases in which the Hendy Test, parallel sampling and long-term monitoring of cave environment are not feasible.