China’s Cultural Ecological Protection Zones (CEPZs) are a distinctive policy instrument intended to safeguard intangible cultural heritage through the “integrated” protection of cultural practices and their supporting socioecological environments. However, there remains limited robust causal evidence on whether CEPZs generate measurable ecological co-benefits and whether such benefits come with landscape-structure trade-offs. Using a county-level panel covering 454 counties from 2006 to 2023, we evaluate CEPZs’ impacts on ecosystem quality and landscape patterns through a multi-period DID design with a rich set of socio-environmental controls. Ecosystem quality is proxied by a satellite-derived NDVI, NDWI, and NPP, while landscape pattern outcomes are captured by a composite landscape pattern connectivity index (LPCI) derived from multi-metric landscape configuration indicators. We further test mechanisms using mediators constructed from Morphological Spatial Pattern Analysis (MSPA) (core habitat proportion) and built-up land proportion, and examine heterogeneity by local governance capacity and region. The results show that CEPZ designation significantly increases the NDVI (β = 0.002, p < 0.01) but significantly reduces the LPCI (β = −0.004, p < 0.01). The average effects on the NDWI and NPP are statistically insignificant. Mechanism tests reveal two countervailing pathways: CEPZs increase the share of core habitat (β = 0.009, p < 0.01), which is positively associated with the NDVI, while simultaneously expanding built-up land (β = 0.012, p < 0.01), which offsets greening and drives fragmentation. Heterogeneity analyses suggest that ecological gains are amplified where independent CEPZ management agencies exist and are stronger in western/central China. These findings provide causal evidence that biocultural governance can yield “greening” co-benefits but may undermine landscape integrity unless development pressures are spatially regulated and local institutional capacity is strengthened.
Yang et al. (Fri,) studied this question.