Spatiotemporal dynamics and driving mechanisms of grassland landscape fragmentation in China’s karst regions

Grassland fragmentation is a defining feature of global terrestrial ecosystem degradation. In China’s karst regions, particularly the Yunnan-Guizhou Plateau and the Qinghai-Tibet Plateau, the dynamic evolution of grasslands is shaped by the interaction between natural processes and human activities. The investigation of mechanisms underlying landscape pattern changes is therefore of both theoretical and practical significance. However, existing studies on grassland fragmentation in China’s karst regions lack comprehensive, large-scale, and long-term analysis. Using land use data from 1990 to 2020, we applied ArcGIS and Fragstats along with land use transfer matrices, landscape pattern indices, trend analysis, and the geographical detector method to reveal the spatiotemporal patterns and driving factors behind grassland fragmentation in karst landscapes. Results show that: (1) From 1990 to 2020, China’s karst grassland experienced a pattern of net loss characterized by a rapid decline initially, followed by gradual stabilization, with a total net loss of 101,942 km2. (2) Grassland fragmentation demonstrated significant spatial variation. The Yunnan–Guizhou Plateau and the Loess Plateau transitional zone have the most complex patch shapes and the highest patch density, whereas the Qinghai–Tibet Plateau experienced lower fragmentation because of natural geographic barriers. (3) Quantitative analysis with the geographical detector model shows that natural factor interactions continue to be the main explanation. However, human activities are becoming more influential, signaling a shift in driving forces from “Natural Dominance” to “Natural-Human Co-Dominance.” Insights into grassland distribution patterns and the causes of fragmentation in China’s karst regions provide valuable guidance for conserving and restoring grassland ecosystems.