[Spatiotemporal Impacts of Cropland Landscape Patterns on Soil Retention in Rapidly Urbanizing Regions： Evidence from...

Soil erosion poses a significant threat to agricultural productivity and China's food security. Investigating the spatiotemporal heterogeneity of cropland landscape patterns and their impact on soil retention is crucial for the integrated protection of cropland in terms of "quantity， quality， and ecology." Focusing on the Wuhan Metropolitan Area （WMA）， a rapidly urbanizing region， this study employs landscape metrics， the InVEST model， and the geographically and temporally weighted regression （GTWR） to systematically analyze the dynamic relationship between cropland landscape patterns and soil retention from 2000 to 2020. The results show that： ① During the study period， cropland in WMA decreased by approximately 148 900 hm² （a 4.97% reduction）， accompanied by declining landscape diversity， increasing morphological complexity， and fragmentation. ② In 376 watersheds （accounting for 95.7% of the total）， the soil retention capacity per unit area declined， though the overall spatial pattern-lower in plains and higher in hilly and mountainous areas-remained. ③ Among the 11 observed landscape variables， regression coefficients exhibited both positive and negative effects. The interspersion and juxtaposition index （IJI）， patch density of dryland （PD_H）， and area-weighted mean shape index of dryland （AWMSI_H） most significantly enhanced soil retention， while the largest patch index of paddy field （LPI_S）， largest patch index of dryland （LPI_H）， and aggregation index of dryland （AI_H） exerted the strongest inhibitory effects. ④ The influence intensity of these variables varied over time， with some-such as the patch density of paddy field （PD_S）， aggregation index of paddy field （AI_S）， PD_H， and AWMSI_H-even reversing the coefficients in certain watersheds， indicating a threshold in their effects. ⑤ To mitigate soil erosion risks， the central and western plains should address paddy field fragmentation and declining connectivity， while hilly and mountainous areas should optimize cropland patch scale and integrate natural habitats to avoid oversimplification and excessive spatial aggregation. By introducing a spatiotemporal dynamic analysis framework， this study elucidates the complex effects of landscape pattern characteristics on cropland soil retention， offering valuable insights for similar regional research.