长三角城市群特色乡村旅游发展潜力与协同分区研究OA

In the context of rural revitalization,rural tourism has emerged as a pivotal driver of regional development.However,the spatial differentiation patterns and synergistic mechanisms underpinning characteristic rural tourism in the Yangtze River Delta Urban Agglomeration(YRDUA)remain inadequately characterized.To address this gap,we analyzed 1,526 national-level characteristic villages across 213 counties in the YRDUA,leveraging comprehensive datasets encompassing environmental,socioeconomic,and tourism-specific variables.Our methodological framework integrates kernel density estimation,the XGBoost-SHAP machine learning model,and location entropy analysis.Three principal findings emerge:(1)Characteristic villages exhibit a pronounced"denser in the east,sparser in the west"spatial distribution,with resource-type-driven agglomeration.Cultural villages,including traditional villages and historical-culturally famous villages,are concentrated along the Anhui-Zhejiang border,where a deep-rooted cultural heritage prevails.Ecological villages clustered in the mountainous areas of western Zhejiang and southern Anhui underscore the decisive role of natural endowments.Tourism-oriented villages,such as national key rural tourism villages,are predominantly situated near core urban centers(e.g.,Shanghai),reflecting market accessibility as a primary locational determinant.(2)The XGBoost-SHAP model identified elevation,slope,population density,road density,and the number of A-level scenic spots as the five most influential predictors.Crucially,it uncovered significant nonlinear effects and strong interaction effects among these variables,demonstrating that tourism development potential arises not from isolated factors but from the interplay between biophysical constraints and socioeconomic enablers.(3)By integrating the assessments of development potential and location matching degree,we propose a four-zone functional classification for the YRDUA:the Characteristic Development Zone(7.5％),Collaborative Optimization Zone(10.8％),Quality-Efficiency Enhancement Zone(46.5％),and Potential Exploration Zone(35.2％).Each zone is assigned a targeted,evidence-based optimization strategy aligned with its structural attributes and developmental stages.For instance,the Characteristic Development Zone,located proximate to high-demand urban cores,should prioritize short-haul leisure offerings and cross-regional thematic tourism corridors,while the Potential Exploration Zone,endowed with abundant natural resources yet constrained by underdeveloped infrastructure and public services,requires strategic investment to activate latent assets and upgrade foundational service capacity.This study advances the methodological rigor by introducing XGBoost-SHAP to capture complex nonlinearities and interactions,which are limitations inherent to conventional regression-or rule-based approaches,thereby generating generalizable,statistically grounded insights beyond case-specific inference.It further contributes to a unified,empirically calibrated synergy framework for heterogeneous characteristic villages and delivers a spatially explicit,policy-actionable zoning scheme.However,this study had several limitations that warrant further investigation.First,data availability confines our scope to national-level designated villages,omitting local-level inventories and institutional dimensions,including designation categories and policy funding mechanisms,which may shape village trajectories,and therefore affect assessment comprehensiveness.Second,the natural break classification method introduces some subjectivity in threshold selection;therefore,practitioners should apply the zoning results flexibly,informed by local contextual knowledge.Future research should integrate multi-tiered village registries,incorporate institutional variables,and benchmark alternative machine learning models and spatial partitioning techniques to strengthen the robustness and transferability of the findings.