Storage and export of soil carbon and mineral surface area along an erosional gradient in the Sierra Nevada, California

摘要

Steep soil-mantled hillslopes are thought to be important sources of sediments and organic carbon (OC) to rivers. Minerals in these sediments may protect OC from decomposition, yet the significance of such interactions in steep upland soils remains poorly constrained particularly in relation to erosion rates. We examined a tributary basin draining to the Middle Folk Feather River in California, where knickpoint migration has created a series of hillslopes with erosion rates varying over an order of magnitude (35 to 250 mm kyr(-1)). This setting provides a unique opportunity to study soil OC pools and their erosional exports as a function of changes in erosion rates. Soil OC inventories were 37% lower at rapidly eroding sites relative to slowly eroding sites. This difference was driven by coarse rock contents as rapidly eroding soils had more rock fragments, limiting their capacities to store OC. Although clay contents in soils were negatively correlated with erosion rates, the total mineral specific surface area remained relatively invariant. Based on secondary phyllosilicate minerals present in the studied soils and our field observations of saprocks, we suggest that this discrepancy may have originated from different clay mineralogy (types and abundance) associated with different degrees of deep subsurface chemical weathering. Across the erosion gradient, the radiocarbon age of mineral associated organic matter (MOC) in saprock varied by a factor 2 (from 1045 to 2110(14)C years), while soil turnover times estimated from soil thickness and erosion rates varied from 17 to 5.4 kyr. At the site eroding at the fastest rate, the soil turnover time approaches the C-14 age of MOC, suggesting erosion can potentially limit the timescale over which MOC is replaced. We found that organic matter generally covered & 50% of the total mineral surface. The remaining OC-free mineral surface area, once eroded, may thus have a significant, and to date unquantified, capacity to adsorb additional organic matter, which may act as a long-term atmospheric carbon sink.