Understanding recreational ecosystem service supply-demand mismatch and social groups' preferences: Implications for urban–rural planning.

• Transferable model is adapted to spatially explicit recreational ecosystem service (RES). • RES supply increase, while demand decrease exponentially from urban to rural areas. • Females, older, and local groups show higher natural landscape preferences. • High-income, younger, and non-local people pursued longer travel distances. • Urban–rural and social groups' preferences facilitate multiscale landscape planning. Recreational ecosystem service (RES) supply and demand are fundamentally influenced by urbanization and are closely related to residents' well-being. Nevertheless, how socio-economic attributes affect spatial RES demand and preferences and can be integrated into urban–rural planning is still unclear. Taking the Beijing-Tianjin-Hebei urban agglomeration region of China as an example, based on survey and spatial data, we applied the Recreation Opportunity Spectrum and a new transferable Quasi-Poisson regression model to examine spatial RES supply and demand. Then, a scaling approach was adopted to interpret the urban–rural spatial patterns of RES. The results showed that the RES supply values exhibited either quadratic relationships or no particular patterns in most cities along urban–rural gradients, while the RES demand values monotonically decreased with an exponential decay or power law. The RES imbalance with higher demand (43%) in urban areas, the supply–demand balance (37%) in urban–rural fringes, and the supply sufficiency (64%) in rural areas were dominant. Divergent RES demand and preferences were identified across genders, ages, income, city scales, and household statuses. Females showed higher preferences for water, grassland, and agricultural landscapes than males; older groups showed higher preferences for natural landscapes but traveled less distance for RES; and high-income groups showed a lower frequency of visits but pursued longer travel distances. We suggest that at both the regional-city and local-community levels, more adaptive and elaborate landscape planning and design should be implemented by integrating urban–rural heterogeneity and the specific population's preferences into RES management. [ABSTRACT FROM AUTHOR]