Influence of rocky desertification intensity of karst areas on soil loss at a regional scale

Rocky desertification and water and soil loss are prominent ecological problems in karst area of Southwest China, which restrict the development of social economy. At present, the coupling relationship between rocky desertification and soil loss has not yet systematically proven, and how the interaction of rocky desertification intensity evaluation factors-vegetation coverage, soil layer thickness and bedrock exposure rate affects soil loss is ill-informed. Based on the investigation of the current situation of rocky desertification, the RUSLE model, spatial association index Getis-Ord Gi* analysis, Spearman correlation analysis, geographic detector and other methods were used to calculate the soil loss status in karst area of Guizhou Province(103°36′-109°35′E、24°37′-29°13′N), identify the distribution of cold and hot zones, quantify the relationships between rocky desertification intensity evaluation factors and soil loss, and analyze the interactive effects of factor combination on soil loss. The results showed that: 1) The soil erosion in karst area of Guizhou was dominated by moderate and micro erosion, with a total of 65 469.32 km2, accounting for 58.71% of the total study area, and with an average soil loss rate of 17.69 t/(hm2·a). Erosion hot zones (major prevention and control area) were mainly in the relatively underdeveloped areas in western Guizhou, such as the eastern part of Bijie City, the western part of Anshun City, the central part of Liupanshui City and the junction of Bijie and Liupanshui, and the total area of erosion hot zones was 31 617.18 km2, accounting for 28.35% of the total study area; while erosion cold zones (slight area) were mainly in relatively economically developed areas such as Zunyi in northern Guizhou and Guiyang in central Guizhou, with a total area of 22 533.26 km2, accounting for 20.21%. 2) Soil loss had a significant negative correlation with vegetation coverage and soil layer thickness, and had a significant positive correlation with the exposure rate of bedrock. The correlation coefficients were -0.067, -0.022 and 0.025, respectively. The relationship between rocky desertification and soil loss was a complex nonlinear relationship. Among them, the relationship between vegetation coverage and soil loss was optimal by a cubic curve function. The relationship between soil layer thickness and soil loss could be described by an exponential function. And the arcsine function had the highest goodness of fit between the exposed rate of bedrock and soil loss. 3) The explanatory power of the evaluation factors of rocky desertification intensity on the spatial differentiation of soil loss was the highest by vegetation coverage, followed by bedrock exposure rate and soil layer thickness. The interaction effects of the factor combination on soil loss were all non-linear enhancement, among which the interaction between the exposed rate of bedrock and vegetation coverage had the strongest explanatory power for the spatial difference of soil erosion, playing a leading role in the spatial differentiation of soil erosion. It was followed by the thickness of soil layer and vegetation coverage, with the least explanatory power for exposed rate of bedrock and soil layer thickness. The results provide valuable information for the coordinated prevention and control of water and soil erosion and rocky desertification in karst area.