Threshold-governed nonlinearities and configurational interactions of development activities on cultural landscape resilience in karst traditional villages

In mountainous regions, traditional villages face deepening tensions between cultural landscape conservation and economic development. Various development activities' impacts on Cultural Landscape Resilience (CLR) exhibit complex nonlinearities and uncharted configurational effects. This study examines 636 village-year observations from 106 karst traditional villages in Southwest China (2020–2025), using gradient boosting regression trees and configurational analysis. Results demonstrate threshold-governed nonlinearities in individual activities: tourism investment, infrastructure improvement, road expansion, public service facilities investment, and agricultural facilities investment emerge as high-importance predictors, following four distinct patterns, optimal-intensity saturation, dual-critical-point transition, discrete-threshold activation, and path-dependent lock-in. For combinational effects, tourism and infrastructure investments serve as primary interaction hubs, capable of forming strong interactions with multiple development activities and influencing CLR. Moderate-intensity configurations (M-M, M-L, L-M) generate positive CLR outcomes through synergistic mechanisms, whereas extreme configurations ( L − L , H − L , L − H ) produce antagonistic effects. These findings advance a precision governance framework grounded in dynamic intensity thresholds and evidence-based synergy strategies.