Influence of topography and human activity on erosion in Yunnan, SW China

Abstract

Schmidt et al. present an interesting study in which they aim to evaluate the effects of human activity on long-term apparent erosion rates calculated from measurements of in-situ 10Be collected from river sediments in a nested sampling scheme across three watersheds in Yunnan, China. The authors choose three watersheds differing in size across covarying gradients of elevation, precipitation, channel steepness and land use. The authors present the results of correlation and regression analyses of in-situ 10Be-derived long-term apparent erosion rates calculated using the CRONUS-Earth online calculator with topographic (basin average slope, normalized channel steepness and relief/elevation), climatic (mean annual precipitation), and land use/land cover variables.Through this analysis, the authors present results which suggest that in-situ 10Be derived apparent erosion rates are related predominantly to topographic variables in two C1 of the three watersheds (Yongchun and Weiyuan) which are not dominated by human land-use), while apparent erosion rates are not correlated with topographic variables (and instead correlated with area in agricultural land use) for a third watershed (Nankai) which was heavily dominated by human land-use. Further regression analysis suggests that human land-use may play a secondary but not controlling role in the Weiyuan watershed (which lies in the middle of the land use gradient from most intensive human land-use (Nankai) to least intensive human land use (Yongchun).The authors conclude that in-situ 10Be apparent erosion rates can be significantly affected by human land-use at the basin scale and suggest caution to researchers attempting to derive long term apparent erosion rates in watersheds which have undergone significant human land-use change, are not large enough to buffer the effects of land-use change on sediment composition, and are heavily disturbed.

Characterizing landscape‐scale erosion using 10Be in detrital fluvial sediment: Slope‐based sampling strategy detects the effect of widespread dams

“…Large dams with high trap efficiencies [ Brune , ] by design impede the flow of water and sediment; thus, the interpretation of 10 Be concentrations as basin‐scale erosion rates, when measured in sediment collected from channels below such dams, is widely known to be uncertain [ Schmidt et al ., ]. The uncertainty arises because the sampled material may not represent that carried into the reservoir from upstream.…”

“…Samples of river sediment are considered to have 10 Be concentrations representative of background, temporally‐integrated erosion rates, because in most environments the upper meter or so of hillslope material moving into river channels is mixed by physical and biological stirring [e.g., Jungers et al ., ]. These data have been used to understand the relationship between erosion rate and climate, tectonic setting, and lithology [e.g., Matmon et al ., ; Portenga and Bierman , ; Schmidt et al ., ; Kirchner et al ., ; Schaller et al ., ]. Because the method is both expensive and time consuming, some regional studies rely on a limited number of 10 Be measurements made in subbasins and averaged, sometimes using area weighting, to generate a landscape‐scale, average erosion rate [e.g., Lupker et al ., ].…”

Spatial and temporal replicability of meteoric and in situ10Be concentrations in fluvial sediment

“…Over centuries to millennia, this would lead to a decrease in 10 Be i concentration on slopes and an apparent increase in erosion rates for agricultural areas compared to non‐agricultural areas (Von Blanckenburg et al ., ). Prior research suggests that this is the case in the more heavily agricultural parts of our study region (Schmidt et al ., ).…”

“…Because many landscapes in China have been disturbed by natural tectonic and climatic driven surface processes as well as human impacts on the landscape, it is a challenging place to apply the 10 Be method and so provides, in some ways, a ‘worst case’ test of replicability; however, China is not the only highly disturbed landscape in which cosmogenic nuclides have been measured. Many regions where 10 Be has been used for geomorphic analysis are tectonically active (Lupker et al ., ; West et al ., ), steep (Portenga et al ., ), subject to extreme rainfall events (Foster and Anderson, ) and/or have experienced widespread land use change (Hewawasam et al ., ; Reusser and Bierman, ; Schmidt et al ., ). By performing spatial and temporal replicate analyses in a setting where many factors are likely to increase the temporal and spatial heterogeneity of analyte concentration in sediment, our results illustrate the limitations of detrital sediment studies and inform sampling strategies not only for cosmogenic nuclides but also for all geochemical detrital sediment analyses.…”

“…Similarly, increased precipitation, particularly concentrated in the monsoon season (Yang et al ., ), could also increase the likelihood of debris flows and landslides, episodically altering the isotopic concentration of channel sediment (Niemi et al ., ; Binnie et al ., ; Savi et al ., ). Upstream agricultural land use increases modern sediment yield relative to long‐term rates of sediment generation (Hewawasam et al ., ; Schmidt et al ., ) through increased linear erosional features or widespread sheetwash (Von Blanckenburg et al ., ). Such rapid sediment transfer mechanisms could increase sediment yield from one part of the basin only during certain times of the year, thus seasonally affecting the isotopic concentration of in channel sediment.…”