Zhejiang Province is located in the southern Yangtze River Delta along China’s southeastern coast, spanning 27°02′N to 31°11′N latitude and 118°01′E to 123°10′E longitude (Figure 1). Bordered by the East China Sea to the east, it adjoins Shanghai and Jiangsu Province to the north, Anhui and Jiangxi Provinces to the west, and Fujian Province to the south, with a total area of 105,500 km². As of 2024, the province governs 11 prefecture-level cities and 90 county-level administrative units (including municipal districts and county-level cities), achieving an urbanization rate of 75.5%.

Ranked fourth in China by GDP, Zhejiang represents a highly developed provincial economy with robust county-level development, hosting 17 counties listed among China’s Top 100 Counties. It maintains one of the nation’s lowest urban-rural income disparities, recording a ratio of 1.83 in 2024. The province encompasses 26 mountainous counties, including Chunan, Yongjia, Pingyang, Cangnan, Wencheng, Taishun, Wuyi, Panan, Kecheng, Qujiang, Changshan, Kaihua, Longyou, Jiangshan, Sanmen, Tiantai, Xianju, Liandu, Qingtian, Jinyun, Suichang, Songyang, Yunhe, Qingyuan, Jingning, and Longquan. Primarily situated in southwestern Zhejiang’s hilly and mountainous terrain, these counties exhibit comparatively lower socio-economic development levels than non-mountainous regions, serving as focal areas for Zhejiang’s high-quality development and the construction of a common prosperity demonstration zone. Mountainous counties occupy approximately 45% of the province’s land area and account for nearly 24% of its population.

Land use data for Zhejiang Province (2005, 2010, 2015, 2020) with a spatial resolution of 30 m were obtained from the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, demonstrating over 93% accuracy in primary land use classification (Ning et al., 2018). Surface PM_2.5 concentration data were sourced from Washington University in St. Louis, featuring a spatial resolution of 0.01° × 0.01 ° (Shen et al., 2024). The NDVI dataset was acquired from the PKU GIMMS NDVI database (spatial resolution: 1/12° × 1/12°; temporal resolution: biweekly; temporal coverage: 1982–2022) (Li et al., 2023). County-level socioeconomic statistics were compiled from provincial statistical yearbooks (Zhejiang Statistical Yearbook), national statistical compilations (China Civil Affairs Statistical Yearbook, China Statistical Yearbook for Regional Economy), municipal yearbooks, and county-level statistical bulletins on national economic and social development. Administrative boundary data were provided by the National Geomatics Center of China (Table 1).

Rural systems, as multidimensional complexes, can be categorized into production, living, and ecological spaces, forming the PLEF dominated by PF, LF, and EF (Lu et al., 2024). PF serves as the foundation of rural PLEF. Its scope extends beyond agricultural output to encompass diverse agricultural, industrial, and service activities contributing to local economies and community wellbeing (Achu et al., 2021), specifically comprising agricultural production functions and non-agricultural production functions. Agricultural production functions focus on agricultural product supply capacity, including fundamental capacities like cropland utilization and food security, as well as enhanced capacities through technological advancements such as agricultural efficiency improvements and mechanization. Non-agricultural production functions reflect rural areas’ wealth-generation capabilities beyond agriculture, primarily through supporting service economies and secondary/tertiary industries like rural tourism and handicrafts. LF quantifies rural residents’ living standards through socio-cultural and economic dimensions. It encompasses basic service provision, housing, and infrastructure supporting daily life, along with social welfare systems and safety net implementation (Qin et al., 2024). EF is crucial for maintaining environmental security and sustaining human livelihoods. It involves ecosystem service provision—clean water, air, soil, and biodiversity conservation—enhancing rural climate change resilience and disaster mitigation. Integrating ecological considerations into agricultural practices facilitates production-ecology balance.

PLEF components exhibit mutually reinforcing yet constraining relationships through complex feedback loops. PF provides material foundations for LF while generating ecological negative externalities. LF influences agricultural production patterns and consumes ecological resources. EF underpins agricultural sustainability and community viability. Synergistically, coordinated PLEF development enhances holistic rural sustainability and promotes high-quality development. Thus, rural PLEF emphasizes coordinated symbiosis among functional spaces, advocating scientific development paradigms over environmentally detrimental models. Its objectives include efficient resource utilization, improved resident wellbeing, and ecological conservation through functional harmonization (Figure 2).

The conceptual framework of this study is illustrated in Figure 3. First, the research area was delineated, and multi-source data were collected and preprocessed. Subsequently, a rural PLEF evaluation index system was constructed, with weights determined using a comprehensive evaluation method. This was followed by an analysis of spatial evolution characteristics and spatial autocorrelation patterns of rural PLEF from 2005 to 2020. Finally, the influencing factors of each functional dimension and their interactive effects were systematically investigated.

Overall, the average rural PLEF development index of mountainous counties in Zhejiang Province increased from 0.88 in 2005 to 1.56 in 2020, with a total growth of 0.67. The contributions to this growth ranked as follows: LF (46.05%) >EF (32.58%) > PF (21.36%). The average growth of PLEF development index in ecological conservation counties (0.63) was significantly smaller than that in leapfrog development counties (0.72). The spatial distribution of rural PLEF development index evolved toward equilibrium. In 2005, high-value areas were concentrated in five ecological conservation counties; by 2010, high-value zones expanded significantly; in 2015, a “multi-core” spatial pattern emerged; by 2020, the distribution became more balanced, with 16 counties exceeding a PLEF index of 1.5, including nine ecological conservation counties (Figure 4).

The average PF development index increased from 0.19 in 2005 to 0.34 in 2020, transitioning from low to medium levels. Structurally, the score ratio of agricultural production functions to non-agricultural production functions shifted from 57.80:42.20 in 2005 to 48.95:51.05 in 2020, indicating a significant rise in non-agricultural contributions. Spatially, all counties were at low PF levels in 2005; by 2010, five counties reached medium levels; by 2020, 15 counties achieved medium or high PF levels (10 of which were leapfrog development counties), though spatial imbalances remained pronounced.

The average LF development index rose from 0.14 in 2005 to 0.46 in 2020, advancing from low to medium levels. Structurally, the ratio of life support function to social welfare function shifted from 60.26:39.74 in 2005 to 75.14:24.86 in 2020, driven by sustained growth in rural disposable income. Spatially, all counties remained at low LF levels from 2005 to 2010; by 2015, 25 counties reached medium levels; by 2020, two leapfrog development counties attained high LF levels.

The average EF development index improved from 0.54 in 2005 to 0.77 in 2020, progressing from medium to high levels. Structurally, the ratio of ecological provisioning function to environmental conservation function shifted from 44.27:55.73 in 2005 to 37.23:62.77 in 2020, reflecting strengthened environmental protection measures. Spatially, 12 counties had high EF levels in 2005, increasing to 22 counties by 2020, demonstrating continuous ecological improvements (Figure 5).

Based on the rural PLEF development index scores in 2005, 2010, 2015, and 2020, the global Moran’s I values were calculated using the spatial autocorrelation analysis tool in GeoDa software, with results shown in Table 3. All global Moran’s I values were positive, with Z-scores significantly exceeding the upper limit of the uniform distribution interval and P-values below 0.05, passing the 5% significance test. These findings indicate significant positive spatial correlation in the rural PLEF development index of Zhejiang’s mountainous counties, demonstrating spatial clustering. The global Moran’s I values declined and stabilized over time, suggesting reduced spatial dependency among mountainous counties and an emerging spatial distribution pattern characterized by “small clusters with broad dispersion”.

The local spatial autocorrelation analysis was conducted by plotting the rural PLEF development index scores on the horizontal axis against spatial lag values on the vertical axis, generating a scatter plot (Figure 6). Statistically significant points were categorized into four cluster types (Table 4). High-High (HH) clusters, located in the first quadrant, are characterized by both focal and neighboring counties exhibiting high PLEF development index, indicating localized positive spatial correlation. From 2005 to 2020, the number of counties in HH clusters within Zhejiang’s mountainous counties gradually decreased, reflecting weakened spatial agglomeration. Low-High (LH) clusters, situated in the second quadrant, feature focal counties with low PLEF indices surrounded by neighboring counties with high indices, demonstrating spatial negative correlation and functioning as “depression zones” adjacent to high-value areas. SY county consistently remained an LH cluster throughout the study period. Low-Low (LL) clusters, occupying the third quadrant, exhibit low PLEF index in both focal and neighboring counties, representing localized spatial autocorrelation. Between 2005 and 2020, the spatial distribution of LL clusters shifted from western to southeastern regions. High-Low (HL) clusters, located in the fourth quadrant, are defined by focal counties with high PLEF indices adjacent to neighboring counties with low indices, indicating spatial negative correlation. The number of counties in HL clusters decreased over time, reaching zero by 2020.

Overall, agricultural labor productivity (C₃) and agricultural mechanization level (C₄) were the dominant influencing factors for PF, per capita electricity consumption of rural residents (C₉) and per capita disposable income of rural residents (C₈) for LF, and forest coverage rate (C₁₃) and mean NDVI (C₁₄) for EF. Temporal analysis revealed shifts in dominant drivers across years. For PF in 2005, the primary influencing factor was agricultural mechanization level (C₄), which shifted to per capita output value of agriculture-related services (C₅) by 2020, indicating a transition from agricultural production focus to non-agricultural production emphasis. For LF, the main factor evolved from per capita electricity consumption of rural residents (C₉) to proportion of rural residents receiving minimum living allowance (C₁₁), reflecting a shift from basic life support to social welfare prioritization. For EF, the dominant factor transitioned from forest coverage rate (C₁₃) to mean NDVI (C₁₄), though ecological provisioning remained the core functional driver (Figure 7).

The interaction detection results revealed nonlinear enhancement and bifactor enhancement effects, indicating that all interactive factors significantly amplified their explanatory power on the spatial heterogeneity of rural PLEF compared to individual factors (Figure 8). This demonstrates that the spatial differentiation of PLEF arises from the synergistic effects of multiple factors. In 2005, interactions involving per capita disposable income of rural residents (C₈), forest coverage rate (C₁₃), and mean NDVI (C₁₄) exhibited q-values exceeding 0.7, with the strongest interaction observed between agricultural labor productivity (C₂) and forest coverage rate (C₁₃) (q = 0.973). By 2010, interactions linked to forest coverage rate (C₁₃) and mean NDVI (C₁₄) dominated (q > 0.7), peaking in the interaction between forest coverage rate (C₁₃) and fertilizer application intensity (C₁₅) (q = 0.979). In 2015, agricultural labor productivity (C₃) emerged as a key interactive factor (q > 0.7), while the interaction between the Proportion of rural residents receiving minimum living allowance (C₁₁) and forest coverage rate (C₁₃) showed the highest explanatory power (q = 0.970). By 2020, the Proportion of rural residents receiving minimum living allowance (C₁₁) became the primary factor influencing spatial heterogeneity, with the strongest interaction observed between rural per capita arable land area (C₁) and fertilizer application intensity (C₁₅) (q = 0.931).

This study developed a PLEF evaluation system using multi-source data, establishing a research framework for measuring and analyzing the evolutionary patterns of rural PLEF at the county level. However, due to the inherent complexity of rural functions and limitations in data availability at the county scale, the indicator design does not comprehensively cover all aspects of PLEF. While spatial data such as land use, PM_2.5 concentration, and NDVI were incorporated, the majority of indicators remain statistics-based. Although statistical reporting standards are consistent across counties, absolute data accuracy cannot be fully guaranteed. Future research should integrate expanded data sources, including field surveys, POI data, and UAV monitoring, to enhance the evaluation system’s precision and holistically capture the functional evolution of mountainous PLEF.

Mountainous areas, covering over 60% of China’s land area, represent a quintessential geographical feature. Existing studies indicate an overall upward trend in PF and LF, while EF exhibit significant variability, particularly fluctuating in industrialized regions (Wang YF. et al., 2024). Economically developed coastal areas generally demonstrate improvements in EF (Zhang HZ. et al., 2024), aligning with the findings of this study conducted in the southeastern coastal region of China.

Regarding the influence of topographic factors on mountainous rural PLEF, prior research suggests that EF dominate at high elevations, whereas PF and LF are more prevalent in low-gradient areas (Zhang J. et al., 2024). Other studies highlight the adverse effects of topographic complexity on LF (Zhang W. et al., 2024). This study analyzed correlations between topographic factors (elevation and slope) and PLEF subsystem development levels in 2020, revealing no significant correlation between PF and topography, a significant negative correlation between mean elevation and LF, and significant positive correlations between mean elevation/slope and EF (Figure 9).

Regarding driving factors, all 17 factors exhibited either nonlinear enhancement or bifactor enhancement effects on rural PLEF, indicating mutually reinforcing interactions among factors. The driving mechanisms of rural PLEF are illustrated in Figure 10. During the 2005–2010 period, rural PLEF underwent infrastructure development and initial integration. Centered on the “Thousand Villages Demonstration, Ten Thousand Villages Renovation” project (referred to as the “Ten-Thousand Villages Project”), this phase prioritized rural infrastructure construction and environmental remediation. PF was dominated by traditional agriculture, with preliminary planning of grain production functional zones laying the foundation for subsequent modern agricultural development. LF focused on improving basic rural residential conditions, including gradual upgrades to electricity and water supply systems, while EF initially emphasized increasing forest coverage. From 2010 to 2015, rural PLEF transitioned toward ecological transformation and systemic enhancement. Guided by the “Beautiful Countryside Construction” initiative, this stage emphasized integrated development of ecological habitats, economies, environments, and cultures. PF shifted to modern agricultural parks, significantly boosting agricultural labor productivity. LF prioritized equal access to urban-rural public services, further improving rural living conditions. EF strengthened systematic governance and resource recycling, fostering green development trends. The 2015–2020 period marked high-quality development and multi-sectoral synergy. Driven by the Rural Revitalization Strategy, this phase promoted coordinated development of rural PLEF. Agricultural production advanced through digital transformation, while secondary and tertiary industries flourished. LF emphasized narrowing urban-rural disparities, with increased focus on livelihood security for low-income rural populations. EF prioritized holistic governance and resilience enhancement through sustained ecological restoration efforts.

This study primarily investigated the evolution and influencing factors of rural PLEF in mountainous counties but did not conduct comparative analyses with non-mountainous counties, which could further elucidate the distinctive characteristics of mountainous PLEF. Future research should strengthen interregional comparisons of PLEF and optimization strategies for rural PLEF. Additionally, while this study focused on the explanatory power of individual indicators within the evaluation system, it did not quantify the contributory weights of factors or explore potential driving effects from other variables. Future studies could systematically investigate multidimensional drivers of rural PLEF development and subsystem performance, including but not limited to policy formulation, population migration, land-use changes, economic development, and technological innovation.

Since 2003, Zhejiang Province has implemented the “Ten-Thousand Villages Project” for over two decades, alongside a series of policies targeting rural construction and urban-rural integration in mountainous counties. These initiatives have significantly improved rural ecological environments and livability (Lin and Hou, 2023; Wang et al., 2019), effectively narrowed regional disparities, and promoted the gradual enhancement of rural PLEF development levels (Zhou et al., 2017). In 2023, Zhejiang’s experience with the “Ten-Thousand Villages Project” was nationally promoted. The 2024 Central No. One Document of China emphasized “learning and applying the development philosophy, methodologies, and implementation mechanisms embedded in the ‘Ten-Thousand Villages Project’.” The evolution of rural PLEF in Zhejiang’s mountainous counties has been synergistically driven by rural development policies, as follows:

First, the “Beautiful Countryside” and common prosperity policies have enhanced rural LF. This study found that LF contributed most significantly to PLEF improvement in Zhejiang, attributable to the “Ten-Thousand Villages Project’s” continuous upgrades to rural infrastructure and public services, alongside common prosperity policies that expanded farmers’ income channels and safeguarded basic welfare. Second, ecological prioritization and green development policies have strengthened rural EF. Guided by the “lucid waters and lush mountains are invaluable assets” philosophy, Zhejiang has advanced ecological restoration, pollution control, and resource recycling. Mountainous counties have adopted the “Ecological County” strategy to accelerate green agricultural development (Zhao et al., 2019). Third, industrial cultivation and targeted assistance policies have boosted PF. Focusing on whole-industry-chain development and technological empowerment, mountainous counties have upgraded PF toward high-value-added and specialized sectors. A classified development strategy categorizes mountainous counties into two types: leapfrog development counties focusing on eco-industries and technological innovation, and ecological conservation counties prioritizing distinctive advantageous industries. The “One County, One Policy” framework promotes high-quality development, while strategies like “Mountainous Zhejiang” and “Mountain-Sea Collaboration” have spatially balanced PLEF development across regions.

However, challenges remain in Zhejiang’s mountainous PLEF development. First, PF require further enhancement. As rural industrial revitalization is the cornerstone of rural revitalization, this study identifies PF as the weakest link in mountainous PLEF. Urgent measures—including technological innovation, institutional reforms, and targeted assistance—are needed to improve agricultural productivity and accelerate modernization (Hu et al., 2023; Zhao and Li, 2022). Second, ecological conservation counties lag behind leapfrog development counties in PLEF performance. Prioritizing the development of ecological conservation counties is critical for Zhejiang’s goal of building a common prosperity demonstration zone.