Search

Soils of the northern high latitudes store carbon over millennial timescales (thousands of years) and contain approximately double the carbon stock of the atmosphere. Warming and associated permafrost thaw can expose soil organic carbon and result in mineralization and carbon dioxide (CO2) release. However, some of this soil organic carbon may be eroded and transferred to rivers. If it escapes degradation during river transport and is buried in marine sediments, then it can contribute to a longer-term (more than ten thousand years), geological CO2 sink. Despite this recognition, the erosional flux and fate of particulate organic carbon (POC) in large rivers at high latitudes remains poorly constrained. Here, we quantify the source of POC in the Mackenzie River, the main sediment supplier to the Arctic Ocean, and assess its flux and fate. We combine measurements of radiocarbon, stable carbon isotopes and element ratios to correct for rock-derived POC. Our samples reveal that the eroded biospheric POC has resided in the basin for millennia, with a mean radiocarbon age of 5,800 ± 800 years, much older than the POC in large tropical rivers. From the measured biospheric POC content and variability in annual sediment yield, we calculate a biospheric POC flux of 2.2(+1.3)(-0.9) teragrams of carbon per year from the Mackenzie River, which is three times the CO2 drawdown by silicate weathering in this basin. Offshore, we find evidence for efficient terrestrial organic carbon burial over the Holocene period, suggesting that erosion of organic carbon-rich, high-latitude soils may result in an important geological CO2 sink.

The oxidation of organic carbon contained within sedimentary rocks ("petrogenic" carbon, or hereafter OCpetro) emits nearly as much C2 as is released by volcanism, thereby playing a key role in the long-term global C budget. High erosion rates in mountains have been shown to increase OCpetro oxidation. However, these settings also export unweathered material that may continue to react in downstream floodplains. The relative importance of OCpetro oxidation in mountains versus floodplains remains difficult to assess as disparate methods have been used in the different environments. Here, we investigate the sources and fluxes of rhenium (Re) in the Rio Madre de Dios to quantify OCpetro oxidation from the Andes to the Amazon floodplain using a common approach. Dissolved rhenium concentrations (n = 131) range from 0.01 to 63 pmol L-1 and vary depending on lithology and geomorphic setting. We find that >75% of the dissolved Re derives from OCpetro oxidation and that this proportion increases downstream. We estimate that in the Andes, OCpetro oxidation releases 11.2+4.5/-2.8 tC km-2 y-1 of C2, which corresponds to ~41% of the total OCpetro denudation (sum of oxidized and solid OCpetro). A Re mass balance across the Rio Madre de Dios shows that 46% of OCpetro oxidation takes place in the Andes, 14% in the foreland-lowlands, and 40% in the Andean-fed floodplains. This doubling of OCpetro oxidation flux downstream of the Andes demonstrates that, when present, floodplains can greatly increase OCpetro oxidation and C2 release. [ABSTRACT FROM AUTHOR]