Flux and influencing factors of CO2 outgassing in a karst spring-fed creek: Implications for carbonate weathering-related carbon sink assessment

Abstract

Whether carbonate weathering on land represents an important terrestrial carbon sink has long been debated as the CO2 sequestered by the dissolution of carbonate may return to the atmosphere through CO2 outgassing from groundwater-feeding surface waters. In headwater area of karst terrain, the CO2 outgassing frequently happens in spring-fed streams, but its flux and influencing factors still remain under-researched. In this study, we conducted a monthly monitoring of hydrochemical and isotopic compositions of stream water along a karst spring-fed creek at Changliushui, SW China that has a steep gradient and in which tufa deposition occurs, in order to explore the process of CO2 outgassing and elucidate the outgassing flux across space and time and its influencing factors. The analyses of downstream geochemical evolution have revealed that two distinguishable types of CO2 outgassing exist along the flow path: (i) outgassing of physically dissolved soil CO2 during the process of carbonate dissolution, as the primary CO2 outgassing, and (ii) an additional outgassing of the CO2 arising from the precipitation of calcite. At a short distance from the spring, the aqueous CO2 escapes from water to a large degree through molecular diffusion driven by high gradient of partial pressure of CO2 (PCO2) between the water and the atmosphere. Such CO2 outgassing concomitantly causes the saturation index with respect to calcite (SIc) to rise, resulting in subsequent calcite precipitation and thus more CO2 outgassing. Spatiotemporal variations in the mass transfer of CO2 and CO2 outgassing flux show that the rates of both types of CO2 outgassing are significantly influenced by hydrodynamic conditions, such as flow discharge and streambed morphology. In particular, lower rates of CO2 outgassing are observed at higher flow discharge, and vice versa. In the studied reach of stream, the primary and additional CO2 outgassing have caused about 21%–35% and 4%–20% of the total soil CO2 sequestered in DIC of the feeding spring water to be released back to the atmosphere, respectively and the higher flow discharge, the less CO2 return. Based on these findings, we postulate that the CO2 outgassing, especially that induced by calcite precipitation should be relatively smaller in streams which have a lower gradient and elevated discharge rate and where the mixing or dilution of tributaries and increased in-stream metabolism occur. This provides more evidence for the previous conclusion that the H2O-carbonate-CO2-aquatic phototroph interaction could represent a net carbon sink at least over short timescales.