Cave monitoring in the Peruvian Andes reveals monsoon climate preserved in speleothem calcite

Abstract

Speleothem paleoclimate records from the Peruvian Andes have been interpreted to reflect the strength of the South American monsoon. While these interpretations have been verified through comparison with other regional and global climate records, the mechanics of the cave environment that facilitate the preservation of this signal with such consistency remain unstudied. Here, we present four years of environmental data from Huagapo and Pacupahuain cave, and one year from Antipayarguna cave. The data reveal that the cave environment is very stable with little to no change in temperature and 100% relative humidity year-round. This stability in cave air is juxtaposed to the monsoonal drip water pulse that increases drip rates over 40 times on average across all seven monitored drip sites. Compared to the amount weighted precipitation average δ18Oprecip value, the cave drip water δ18ODW values are evaporatively 18O enriched during infiltration through the soil/epikarst. As the monsoonal precipitation pulse fades and drip rates decrease, changes in the drip water chemistry (trace elements Mg/Ca and Sr/Ca, dissolved inorganic carbon δ13CDW, and δ18ODW values) indicate that prior calcite precipitation (PCP) drives the trace element and δ13CDW variability. The δ13Cc and δ18Oc values of farmed slide calcite are highly variable. However, high drip rate and lower cave air pCO2 during the monsoon combine to increase calcite precipitation rates. This causes speleothem records from these caves to be weighted toward annual monsoon conditions. Calcite isotope values from actively growing stalagmite tops support this finding. These results suggest that speleothems from these caves are sensitive to changes in monsoon precipitation amount, because it determines the duration of the monsoon drip water pulse, and therein, the extent of dry season PCP. Further, these data indicate that heterogeneity in the dolomitic limestone massif causes offsets between the carbon isotopes and trace metal concentrations between the caves, highlighting the need to normalize these datasets when chronology-stacking these proxies.