Rising CO2 and land use change amplify the increase in terrestrial and riverine export of dissolved organic carbon over the past four decades.

ABSTRACT

The lateral transport of dissolved organic carbon (DOC) from land to rivers and oceans is a significant but overlooked component of the global carbon cycle. However, there are still large uncertainties in the magnitude and trend of global DOC export fluxes, as well as their response to environmental change. In this study, several simulations were conducted using a developed land surface model that considered riverine DOC transport and anthropogenic disturbance to investigate the terrestrial DOC loading and riverine DOC export, and to quantify the relative contributions of natural and anthropogenic factors over the past four decades (1981-2016). These factors include climate change, nitrogen deposition, land use change, atmospheric CO2 concentration, anthropogenic water regulation, and fertilizer and manure application. Results showed that the average global annual terrestrial DOC loading was about 432.30 ± 53.59 Tg C yr-1, and rivers exported about 209.73 ± 36.58 Tg C yr-1 to oceans over the past four decades. Simultaneously, a significant increase in terrestrial DOC loading and riverine DOC export fluxes (3.36 Tg C yr-2 and 2.99 Tg Cyr-2, p < 0.01) was found, which increased by 26.88 % and 47.02 %, respectively. According to our factorial analysis, the interannual variability in DOC fluxes in most regions was mainly attributed to climate change and contributed more than 60 % of the long-term increase. In addition, rising atmospheric CO2 and land use change amplified the increase in terrestrial DOC loading, with the area dominated by the two factors expanding from 7.94 % in the 1980s to 23.84 % in the 2010s, and riverine DOC export showed a similar pattern, which may be related to the increased soil DOC sources. Anthropogenic water regulation and nitrogen addition have led to a slight increase in DOC fluxes, which should not be ignored, otherwise carbon fluxes may be underestimated.