Seasonal climate signals (1990–2008) in a modern Soreq Cave stalagmite as revealed by high-resolution geochemical analysis

Abstract

High-resolution isotopic and geochemical analyses in a modern (1990–2008) Soreq Cave stalagmite are compared to instrumental records of rainfall and dripwater from the cave, with the aim of determining how seasonal-resolution climate information is transmitted to speleothem geochemistry. In situ, micron-scale analysis of oxygen isotope ratios (δ18O) and trace elements by ion microprobe in combination with a continuous, linear traverse of trace element concentrations by laser-ablation ICP-MS (LA-ICP-MS) allow the definition of geochemical pathways within the cave. Fluorescent banding, imaged by confocal laser fluorescent microscopy (CLFM), as well as δ18O and trace element variations is used to define 18 annual growth increments. Reduced intensity of fluorescent banding and a change in trace element variability reflect the decrease in average rainfall from 628mm/yr (1990–1998) to 433mm/yr (1999–2008). During the wetter period before 1998, Pearson (r-value) and Spearman (ρ-value) correlation coefficients are >0.5 for ion microprobe analyses of the element pairings Sr–Y, Y–P, and Mn–Si. After the transition to the drier period in 1999, a different set of geochemical pairings have r- and ρ-values >0.5, including Mg–δ18O, Mg–Sr, and Sr–Ba. Principal component analysis of data from the adjacent LA-ICP-MS traverse identifies two primary underlying modes of trace element variability. Based on the ion microprobe correlations and principal component analyses, we suggest that a greater seasonal influx of particulate material into the cave during the wetter period (1990–1998) brought about greater P, Cu, Sr, Na, and U variability in the stalagmite. The co-variability of δ18O, Mg, Sr, and Ba is characteristic of the trace element pattern from the drier period (1999–2008) of growth when particulate transport is reduced. These findings support a two-reservoir model of Soreq Cave dripwaters. One reservoir displays a well-mixed “baseline” with a decadal residence time that supplies water to the cave year-round, probably from fine pores or grain-boundary films in the vadose zone. The second reservoir is seasonal rainwater enriched in organic acids, colloids, and small particles and is rapidly transmitted to the cave. Finally, the similar patterns of fluorescence intensity, P, and Cu concentrations support the hypothesis that fluorescent bands in Soreq Cave speleothems are caused by the influx of organic colloids.