Attribution analysis of soil degradation using change vector analysis and the geographical detector from 2010 to 2020 on Hainan Island

Assessing soil degradation and determining its influences is crucial for the protection of terrestrial ecosystems. The present study focused on Hainan Island (HN), China, from 2010 to 2020, using soil organic carbon (SOC) and cation exchange capacity (CEC) to create an assessment system ((S) over right arrow) for soil degradation based on change vector analysis. A geographical detector was used to determine the mechanism of soil degradation on HN. The total area of soil degradation in the study period was 673.69 km(2), being concentrated in the northeast. Areas of soil degradation in forest, cultivated land, construction land, grassland, and unutilized land measured 439.88 km(2), 184.06 km(2), 25.44 km(2), 22.56 km(2), and 1.75 km(2), respectively. The single factors of elevation (DEM), least distance to a point of interest in residential areas (LDP), and variation in mean annual temperature (VTem) were significantly correlated with soil degradation in cultivated land and grassland, forest, and construction land, respectively. However, paired factors exhibited greater explanatory power. The pairs of DEM boolean AND LDR (DEM + least distance to point of interest of road and railroad, LDR), LDP boolean AND LDR, and VTem boolean AND VGDP (VTem + variation in GDP density) had the most significant correlations with soil degradation in cultivated land and grassland, forest, grassland, and construction land, respectively (p < 0.01). The soil resources of HN should be protected in multiple ways, such as by expanding ecological protection zones in high-altitude grasslands and forests. Moreover, urban development goals should be clearly defined to limit rapid increases in road and settlement construction to protect cultivated land and constructed land.