Spatial variability in cave drip water hydrochemistry: Implications for stalagmite paleoclimate records

Abstract

The identification of vadose zone hydrological pathways that most accurately transmit climate signals through karst aquifers to stalagmites is critical for accurately interpreting climate proxies contained within individual stalagmites. A three-year cave drip hydrochemical study across a spectrum of drip types in Crag Cave, SW Ireland, reveals substantial variability in drip hydrochemical behaviour. Stalagmites fed by very slow drips (< 0.1 ml/min) may best retain information regarding decadal- through millennial-scale climate because the drip sites' diffuse recharge minimizes interferences to the long-term pattern produced by isolated meteorological events. Additionally, hydrological routing shifts did not influence these very slow drips. Intermediate flow regimes (0.1�2 ml/min) are apparently most sensitive to water excess, and may best preserve a paleoseasonality signal because of a combination of rapid stalagmite growth, seasonally responsive drip rates, and minimal interferences from stochastic processes within the aquifer. Stochastic drip-rate variability existed at several high-discharge (> 2 ml/min) sites, apparently unconnected with local meteorological events. Water from these drips was typically undersaturated with respect to calcite, and thus did not result in calcite deposition. Data presented here suggest that drips in this flow regime also experience flow re-routing and blocking, and that any stalagmites developed under such drips are unsuitable as mid- to high-resolution paleoclimate proxies. Most drip sites demonstrated seasonal [Ca2+] and [Mg2+] variability that was probably linked to water excess. Prior calcite precipitation along the flowpath affected the chemistry of slowly dripping sites, while dilution predominantly controlled the water chemistry of the more rapidly dripping sites. This research underscores the importance of understanding drip hydrology prior to selecting stalagmites for paleoclimate analysis and before interpreting any subsequent proxy data.