Sources of dissolved carbon in large rivers: Insights from coupled 13C-14C in the upper Changjiang (Yangtze) River

Abstract

Export of dissolved carbon from rivers connects terrestrial and oceanic carbon cycling, and represents a key component of global carbon budgets. In this study, we measured the concentrations of major ions, the sulfur isotopic composition of sulfate, and the stable (δ¹³C_DIC) and radioactive (Δ¹⁴C_DIC) isotopic composition of the dissolved inorganic carbon (DIC) in the upper Changjiang (Yangtze) River, China. Their spatio-temporal variabilities give constrains on carbon cycling across catchments with diverse physiographic conditions (e.g. climate, hydrology, geology, land-use change). Daily DIC fluxes were modeled using a hydro-chemical approach and showed significant temporal variations agreeing with the measured data, and supporting a conservative mixing behavior in river channel, similar to variations in anions and cations concentrations. However, the modeled (δ¹³C_DIC) were always higher (up to 3.1‰, and 0.8‰ on average) than the measured values. In addition, Δ¹⁴C_DIC would suggest a much lower carbonate contribution (8 to 33%) than concentration mixing relationship (40 to 64%). From the chemical weathering reaction front to the riverine transport, DIC can be processed with little net change in concentration but isotopic exchange and it can also exchange with the soil and/or atmospheric CO₂ as an open system. The openness parameter (ψ) uses the Δ¹⁴C_DIC to estimate the extent of such exchange, ranging from 0 (fully closed system) to 1 (fully open system) and varied from 0.26±0.10 to 0.86±0.06. ψ showed a strong relationship with drainage mean elevation, likely to represent the elevation-induced climate variation (i.e. precipitation and temperature) but also the proportion of cropland in the drainage area. All of these parameters are correlated and point toward a biological impact on ψ. In addition, modern DIC yield, estimated from Δ¹⁴C_DIC, is positively correlated to ψ. This study demonstrates that Δ¹⁴C_DIC measurements can greatly underestimate the contribution of carbonate-derived carbon because of exchange (estimated by ψ) especially in high biological-productivity areas.