Temperature and altitudinal influence on karst dripwater chemistry: implications for regional-scale palaeoclimate reconstructions from speleothems

The reconstruction of robust past climate records from speleothems requires a prior understanding of the environmental and hydrological conditions that lead to speleothem formation and the chemical signals encoded within them. On regional-scales, there has been little quantification of the dependency of cave dripwater geochemistry on meteorology (net infiltration, temperature), environmental and geographical factors (elevation, latitude, soil activity, vegetation cover, atmospheric aerosol composition) and geological properties of the aquifer (lithology, porosity and thickness). In the present study, we analysed over 200 karst waters collected in 11 caves of the Trentino region (NE Italy). The caves span sub-humid Mediterranean to cold-humid temperate climates and infiltration elevations (Z<i>_inf</i>) ranging from 355 to 2400 m a.s.l., corresponding to infiltration mean annual temperatures (MAT<i>_inf</i>) between 12 and 0 °C. Since all the caves developed in pure carbonate rocks, soil pCO₂ is found to be the main factor controlling the carbonate dissolution. For this reason, the parameters controlling the carbonate-carbonic acid system and calcite saturation state (SICC) are directly correlated with the MAT<i>_inf</i>, which influences the vegetation zones and eventually the production of CO₂ in the soil. SI_CC linearly depends on MAT<i>_inf</i> (SI_CC = 0.09 MAT<i>_inf</i> - 0.4) and SI_CC = 0 is reached at Z<i>_inf</i> = 1.66 km a.s.l., corresponding to a MAT<i>_inf</i> = 4.4 °C. This point identifies the "<i>speleothem limit</i>" defined here as the elevation (or corresponding MAT<i>_inf</i>) above which no sparitic speleothem precipitation usually occurs. We demonstrate that due to temperature-forced changes in the soil and vegetation and subsequently SI_CC, the speleothem limit shifts to higher altitudes during maximum interglacial conditions. Speleothems from high altitude caves (1.5-2.5 km a.s.l.) thus can identify optimum interglacial periods. By contrast, speleothems formed at lower altitudes are better suited as archives of hydrological proxies. At altitudes below 1.2 km a.s.l., prior calcite precipitation (PCP) modifies percolating waters, particularly during periods of reduced infiltration. We introduce the use of the SiO₂/Ca and SO₄/Ca ratios in cave waters to complement Mg/Ca and Sr/Ca ratios as markers of PCP. SO₄ and SiO₂ are derived from atmospheric deposition and siliciclastic minerals in the soil zone, rather than carbonate host rocks (as in the case of Mg and Sr). By combing shifts to higher Mg/Ca, SiO₂/Ca and SO₄/Ca ratios along their characteristics PCP lines, we improve the robustness of the interpretation that this resulted from increasing PCP, rather than incongruent calcite dissolution (ICD). Our method permits the quantification of PCP between 0% and 40% for low elevation cave waters. This novel approach has important implications for speleothem-based paleoclimate studies where the distinction between PCP and ICD can be ambiguous and, in combination with Mg/Ca and Sr/Ca ratios, permits the quantification of net infiltration and/or rainfall amount from speleothem records