A Prediction–Interaction–Driving Framework for Ecosystem Services Under Climate Change and Human Activities: A Case Study of Zoigê Plateau

Under climate change and human activities, ecosystem service (ES) research lacks systematic approaches and scientific depth. This study develops a comprehensive framework integrating advanced models to predict ESs, analyze interactions, identify key drivers, and assess spatial effects on the Zoigê Plateau. The results indicate the following: (1) From 2000 to 2020 and across three 2040 scenarios, water conservation (WC) improves, while carbon storage (CS) and habitat quality (HQ) decline, leading to overall ES degradation. Core ES areas face rising degradation risks from 9% to 29% under increasing environmental stress (SSP119 to SSP585). (2) ES importance follows HQ > CS > SC > WC, with bivariate interactions outperforming single-factor effects. Future scenarios show weakened interactions, correlating with higher ecological stress, indicating ES stability risks. (3) Land use (>40% explanatory power) is the primary driver, while urban expansion, slope, evapotranspiration, and precipitation contribute (6–12%). (4) ES drivers showed weak spatial patterns from 2000 to 2020 but became more stable under future scenarios, suggesting stronger environmental control. This study provides a methodological paradigm for ES analysis and supports ecological planning in alpine wetland–grassland regions.