The effect of climate and vegetation variation on monthly sediment load in a karst watershed

The variation in monthly sediment load is greatly influenced by climate and vegetation changes which are generally coupled and interconnected. However, few studies have quantified the relative magnitudes of the direct and indirect influences of climate variability and vegetation dynamics on monthly sediment load. The objective of this study was to decouple the impacts of climate and vegetation changes on the monthly sediment load in a karst watershed using partial least squares-structural equation modeling (PLS-SEM). Monthly sediment load, runoff, climatic factors (temperature and precipitation), and vegetation data (NDVI, EVI, and LAI) were collected from 2003 to 2017. Results indicated that climatic factors and vegetation dynamics explained 73% of the variation in monthly runoff, while climatic factors, vegetation dynamics, and runoff accounted for 62% of the monthly sediment load changes. Climate change can not only directly alter vegetation and hydrological characteristics, but also indirectly affect hydrological characteristics by changing vegetation dynamics, thus affecting the cooperative processes of runoff and sediment load. Specifically, runoff (direct effect = 0.84) and precipitation (indirect effect = 0.67) had a positive impact on monthly sediment load, while antecedent temperature (indirect effect = −0.02), temperature (indirect effect = −0.03), antecedent precipitation (indirect effect = −0.01), and vegetation dynamics (direct effect = −0.12; indirect effect = 0.07) had a negative impact. Compared with traditional methods, PLS-SEM can more completely describe the relationship among the observed variables and other potential influencing variables and may provide relevant references for eco-hydrological management and land resource optimization in karst watersheds.