Abstract
Large-scale ecological restoration programs are increasingly used to mitigate ecosystem degradation and strengthen regional ecological security, yet their management effectiveness depends on whether restoration investments are spatially aligned with ecological urgency, implementation feasibility, and landscape connectivity. Using China’s Mountains–Rivers Projects as a national case, this study developed a spatial prioritization framework to diagnose restoration mismatches and optimize future project allocation. The framework integrated long-term declines in ecosystem quality and key ecosystem services, restoration implementation gaps across cross-administrative natural geographical units, nationally designated ecological protection spaces, relative implementation difficulty, and landscape connectivity indicators. Results showed that degradation in ecosystem quality and services was spatially clustered, with major relative degradation-priority areas concentrated west of the Hu Huanyong Line, particularly in the Tianshan Mountains, Qinghai–Tibet Plateau, Hengduan Mountains, and southwestern China. However, existing Mountains–Rivers Projects were disproportionately distributed in eastern China, leaving several ecologically fragile and functionally critical western regions insufficiently covered. By jointly considering ecological priority and cost feasibility, we identified 24 potential optimized zones covering 1.1174 million km2, of which nine zones were located west of the Hu Huanyong Line and accounted for 69.45% of the total optimized area. The optimized scenario indicated potential structural connectivity gains, with the number of ecological corridors increasing by 18.7%, while dIIC and dPC increased by 31.52% and 45.38%, respectively. These findings suggest that future restoration planning may benefit from strategic rebalancing toward ecologically urgent, cross-boundary, and potentially connectivity-enhancing regions. The proposed framework provides a transferable decision-support tool for improving the spatial efficiency, ecological effectiveness, and management feasibility of large-scale restoration programs.
1 Introduction
Large-scale ecological restoration has become a central environmental management strategy for reversing ecosystem degradation, enhancing ecosystem services, and strengthening ecological security under accelerating climate change and land-use pressures (; ). Restoration programs are increasingly expected not only to recover degraded habitats, but also to improve water regulation, soil retention, vegetation carbon uptake capacity, biodiversity conservation, and the resilience of coupled social–ecological systems (; ; ; Zeller et al., 2012). However, the management effectiveness of large-scale restoration depends critically on where restoration investments are allocated. When restoration projects are spatially concentrated in areas that are easier to implement or administratively convenient, but not necessarily the most ecologically urgent, ecological benefits may be reduced and long-term restoration efficiency may be compromised (; ; Xia et al., 2021). Therefore, identifying priority areas that balance ecological urgency, implementation feasibility, cross-boundary ecological integrity, and landscape connectivity has become a key challenge in environmental management.
Systematic conservation planning and spatial prioritization approaches provide useful tools for allocating limited conservation and restoration resources (Wang X. et al., 2024). These approaches have been widely applied to biodiversity conservation, ecosystem service protection, protected-area planning, and restoration opportunity assessment. Nevertheless, several limitations remain when they are applied to large-scale, government-led ecological restoration programs (Wang X. et al., 2024). First, many prioritization studies emphasize ecological benefits but insufficiently consider implementation costs and management feasibility, resulting in priority areas that may be ecologically important but difficult to implement. Second, restoration planning is often constrained by administrative boundaries, whereas ecological degradation, hydrological processes, potential ecological linkages, and landscape connectivity frequently operate across administrative units. Third, existing assessments commonly focus on single ecological functions, such as vegetation carbon uptake capacity or habitat quality, while large-scale restoration requires a more integrated evaluation of ecosystem quality and multiple ecosystem services. These limitations reduce the ability of spatial planning frameworks to support practical restoration decisions under real-world budgetary, institutional, and governance constraints.
China’s integrated Protection and Restoration of Mountains, Rivers, Forests, Farmlands, Lakes, Grasslands, and Deserts Projects, hereafter referred to as the Mountains–Rivers Projects, provide a major national case for examining these challenges (; ). Guided by the principle that mountains, rivers, forests, farmlands, lakes, grasslands, and deserts constitute a community of life, the Mountains–Rivers Projects aim to promote integrated ecosystem restoration across watersheds, ecological function zones, and regional ecological security barriers (). Since their launch in 2016, 52 projects have been implemented across China, involving watershed rehabilitation, forest and grassland restoration, mine reclamation, soil erosion control, and other ecological restoration measures (; ). These projects have played an important role in improving ecosystem quality and supporting national ecological security. However, as the program expands, its spatial allocation requires further optimization to ensure that future restoration investments are directed toward areas with the greatest ecological urgency and management significance.
A major concern is the potential spatial mismatch between the current distribution of Mountains–Rivers Projects and the distribution of ecological degradation and restoration needs (Zhao and Dai, 2024). In China, the Hu Huanyong Line provides a useful spatial benchmark for distinguishing the densely populated and economically developed east from the sparsely populated but ecologically fragile west. Existing Mountains–Rivers Projects are disproportionately concentrated east of this line, whereas many western regions contain extensive alpine, arid, and mountainous ecosystems with high ecological sensitivity and critical ecosystem service functions (; ). These western regions often face severe declines in ecosystem quality, water conservation, soil retention, and vegetation carbon uptake capacity, but their remoteness, high elevation, harsh environmental conditions, and limited accessibility substantially increase restoration costs (; ). In contrast, eastern regions often have lower implementation costs and stronger management capacity, but restoration needs tend to be more fragmented and strongly affected by intensive human activities. This east–west contrast highlights a fundamental management trade-off: restoration planning must balance ecological necessity with economic feasibility, while avoiding systematic bias toward low-cost but lower-priority areas (; Yuan et al., 2024).
Another unresolved issue is the insufficient integration of Mountains–Rivers Projects with cross-administrative natural geographical units and nationally designated ecological protection spaces (). Ecological processes are not confined to administrative jurisdictions. Watershed connectivity, Mountains–Rivers interactions, potential ecological linkages, and ecosystem service flows often depend on the integrity of natural geographical units that span multiple provinces or regions (). Meanwhile, China has established several categories of important ecological protection spaces, including key ecological function zones, important ecological function zones, biodiversity conservation priority areas, ecologically fragile areas, and major ecological supervision zones (Choi et al., 2022; ). These areas form the spatial foundation of national ecological security. However, if restoration projects are not well aligned with these ecological protection spaces, restoration actions may remain spatially fragmented, ecological corridors may be weakened, and the overall contribution of restoration to national ecological security may be constrained (). Current project implementation gaps are particularly pronounced in western mountainous and aquatic systems, whereas eastern gaps are more scattered and often located in densely populated ecological service areas.
Previous studies on ecological restoration in China have mainly evaluated ecological outcomes, restoration effectiveness, or regional policy performance (; Zhang et al., 2024). While these studies provide valuable evidence, few have systematically examined the spatial mismatch between existing restoration projects and ecological degradation at the national scale. Even fewer have integrated ecosystem quality decline, multifunctional ecosystem service degradation, implementation cost, cross-boundary ecological units, nationally designated ecological protection spaces, and landscape connectivity into a unified spatial optimization framework (Schirpke et al., 2019; Xu et al., 2017). As a result, current knowledge remains insufficient to answer three key environmental management questions: where are the most urgent relative degradation-priority areas; to what extent do existing restoration projects cover these relative degradation-priority areas and strategic ecological spaces; and how can future restoration projects be spatially optimized to maximize ecological benefits while maintaining cost feasibility and network connectivity?
To address these challenges, this study develops a national-scale spatial prioritization framework for optimizing the future layout of China’s Mountains–Rivers Projects. Specifically, we aim to: (1) identify spatial patterns of significant declines in ecosystem quality and three key ecosystem services, namely, vegetation carbon uptake capacity, soil retention, and water conservation; (2) diagnose spatial gaps between current Mountains–Rivers Projects and ecologically important cross-administrative natural geographical units and national ecological protection spaces; (3) construct an ecology–cost prioritization framework to distinguish ecological priority areas from cost-priority areas; and (4) evaluate whether the optimized layout improves landscape connectivity and the integrity of the national ecological security network. By linking ecological degradation diagnosis, cost-sensitive spatial prioritization, cross-boundary governance, and connectivity assessment. This study provides a decision-support framework for improving the spatial efficiency and management effectiveness of large-scale ecological restoration programs. The findings can inform the future allocation of Mountains–Rivers Projects in China and offer broader insights for large-scale restoration planning in other regions facing similar trade-offs between ecological urgency and implementation feasibility.
2 Materials and methods
2.1 Study area
China has implemented two phases of Mountains–Rivers Projects, totaling 52 individual projects and covering approximately 1.525 million km2. Currently, the total area of completed ecological restoration exceeds 8 million hectares, involving 29 provinces nationwide (Figure 1). To evaluate the spatial imbalance of project implementation, all implemented projects were reclassified according to their location relative to the Hu Huanyong Line. The results show that 42 projects are located east of the Hu Huanyong Line, whereas only 10 projects are located west of it. This east–west classification provides a consistent spatial benchmark for examining whether existing restoration investments are aligned with ecological degradation patterns and restoration needs.
FIGURE 1
2.2 Data sources
Temperature, precipitation, and evapotranspiration data were obtained from the China Meteorological Data Service Center. A total of 699 meteorological stations were used for interpolation during 2000–2023. The station observations were interpolated into 1 km gridded datasets using ANUSPLIN, with longitude, latitude, and elevation included as covariates to account for spatial and topographic controls on climate variables. Monthly meteorological variables were interpolated first and then aggregated to annual values. Precipitation was calculated as annual total precipitation, whereas temperature, evapotranspiration, and solar radiation were calculated as annual means. Vegetation indicators, including Normalized Difference Vegetation Index (NDVI), gross primary productivity (GPP), net primary production (NPP), and leaf area index (LAI), were selected to represent ecosystem ecological quality. These datasets, derived from MODIS products, cover 2000–2023 at a spatial resolution of 500 m. NDVI, LAI, and GPP were used to construct the ecosystem quality index, whereas NPP was used as a proxy for vegetation carbon uptake. For biweekly NDVI, maximum value compositing was applied to reduce cloud contamination, followed by spike filtering and temporal compositing. All remote-sensing products were resampled and projected to a common national geospatial reference framework. Edge buffers were removed during mosaicking to reduce boundary noise.
The water conservation and soil retention data for 2000–2023 were obtained from the Resource and Environmental Science Data Center (https://www.resdc.cn/), with a spatial resolution of 500 m.
Land-use/land-cover data, digital elevation model data, road networks, settlement distribution, administrative boundaries, existing Mountains–Rivers Project boundaries, and national ecological protection-space datasets were compiled for spatial overlay, cost approximation, and connectivity analysis. National ecological protection spaces included key ecological function zones, important ecological function zones, biodiversity conservation priority areas, ecologically fragile zones, and major ecological supervision zones.
To reduce scale effects and the modifiable areal unit problem, all raster layers were resampled to 1 km × 1 km. Vector layers were standardized to a 1:100,000 scale where applicable and aligned using a unified national spatial reference framework. For area statistics and spatial comparison, all layers were projected to an equal-area coordinate system. The temporal window was standardized to 2000–2023 for all ecological indicators.
2.3 Ecosystem quality assessment
2.3.1 Estimation of fractional vegetation cover
Fractional Vegetation Cover (FVC) is a crucial indicator for assessing surface vegetation condition and overall ecosystem health. It can be effectively derived from remote sensing data. Based on the pixel dichotomy model, FVC is estimated using the Normalized Difference Vegetation Index (NDVI). This model assumes that the NDVI value of a pixel represents a mixture of two endmembers: pure vegetation and bare soil. The calculation formula is as follows (Equation 1):where FVC denotes the fractional vegetation cover; NDVIveg represents the NDVI value of pure vegetation pixels (95th spatial percentile of NDVI values over the whole period); and NDVIsoil indicates the NDVI value of pixels without vegetation cover (fifth spatial percentile of NDVI values over the whole period). This percentile-based approach accommodates spatial heterogeneity and is more robust than fixed or pixel-level min–max methods, especially for ecosystems with persistent vegetation cover, such as evergreen forests.
2.3.2 Construction of the ecosystem quality index
Ecological conservation is not simply defined and/or measured by vegetation restoration targets such as afforestation coverage, but rather by quantifiable outcomes of vegetation types (e.g., ecosystem quality) or processes (e.g., the recovery of ecosystem function). Ecosystem quality primarily reflects the overall health status of ecosystems and vegetation conditions ().
In this study, we employed a composite index integrating Leaf Area Index (LAI), FVC, and Gross Primary Productivity (GPP) to assess Ecosystem Quality Index (EQI), which follows the national ecological standard of the People’s Republic of China, Technical Specification for Investigation and Assessment of National Ecological Status—Ecosystem Quality Assessment (HJ 1172–2021). To comprehensively assess the ecosystem quality, all three ecosystem indicators (e.g., FVC, LAI, and GPP) were normalized to a 0–1 scale using min–max rescaling. The relative density of each indicator within the same ecosystem type, such as alpine grasslands, deserts, and arid ecosystems, was calculated, and the EQI was then expressed as a weighted average of these three indicators (Equation 2):where LAIk, FVCk, and GPPk represent the relative densities of leaf area index, fractional vegetation cover, and gross primary productivity, respectively.
2.4 Trend analysis
The Theil–Sen median method, also known as the Sen’s slope estimator, is a robust nonparametric approach for detecting monotonic trends in time series data. It is highly resistant to measurement errors and outliers, making it well-suited for analyzing long-term environmental datasets. In this study, ecological degradation was identified at the pixel level using the Theil–Sen slope estimator and the Mann–Kendall significance test. A pixel was defined as degraded only when the Sen’s slope was negative and the Mann–Kendall test was significant at the 95% confidence level (p < 0.05) (Equations 3, 4). Because pixel-level trend tests over large spatial datasets may be affected by serial autocorrelation, tied values, and multiple testing, the resulting decline maps were interpreted as national-scale screening outputs rather than precise pixel-level degradation diagnoses.
In Equations 5–7, y is the time series of each ecological indicator, t is the year, α is the intercept, β is the slope (i.e., the trend rate), and ε is the random error. yi and yj are the value at the ith and jth position, respectively (∀ j > i). The significance of Sen’s slope was further examined with Mann-Kendall test, by calculating Z values.where n is the number of years, sgn () is the symbolic function. The 95% confidence level was used to define statistically significant degradation; pixels with |Z| > 1.96 and a negative Sen’s slope were identified as significantly degraded pixels. All the statistical analyses and significant detections were performed using MATLAB R2024b.
2.5 Spatial prioritization analysis of the Mountains–Rivers Projects
Based on the technical workflow shown in Figure 2, we developed a five-step spatial prioritization and optimization framework for the Mountains–Rivers Projects. This framework was designed to link ecological degradation diagnosis, relative implementation difficulty, spatial gap identification, optimized project allocation, and connectivity-effect evaluation into an integrated decision-support process.
FIGURE 2
2.5.1 Identification of ecological quality and ecosystem service decline
First, we identified areas with significant declines in ecosystem quality and ecosystem service functions. Ecosystem quality was represented by the ecosystem quality index, while ecosystem services included water conservation, soil retention, and vegetation carbon uptake capacity. For each indicator, pixel-level temporal trends during 2000–2023 were quantified using the Theil–Sen slope estimator, and their statistical significance was tested using the Mann–Kendall test.
To examine whether ecological degradation exhibited spatial clustering, we further applied global Moran’s I, local Moran’s I, and Getis–Ord Gi* statistics to the declining trends of ecosystem quality and ecosystem service indicators. Because all ecological indicators were resampled to a regular 1 km × 1 km raster grid, a first-order queen contiguity spatial weights matrix was used, in which each grid cell was considered adjacent to its eight surrounding neighboring cells. The spatial weights matrix was row-standardized before calculating the spatial autocorrelation statistics. Statistical significance was assessed using 999 random permutations, and local clusters were identified at the 95% confidence level.
Global Moran’s I was used to evaluate the overall spatial autocorrelation of ecological degradation, while local Moran’s I and Getis–Ord Gi* were used to identify local clusters and local clusters and high/low-value clusters of declining ecological indicators. These spatial autocorrelation analyses were used to characterize the spatial aggregation pattern of degradation rather than to independently determine the final restoration priority zones. The final priority classification was based on the combined criteria of significant declining trends, degradation magnitude, ecosystem service overlap, implementation gaps, and relative implementation difficulty.
To evaluate whether the spatial pattern of degradation-priority areas was sensitive to threshold selection, we conducted sensitivity tests using the top 30%, 40%, and 50% of pixels with the greatest declining slopes among statistically significant degraded pixels (Supplementary Figure S1). The results showed that the spatial extent of degradation-priority areas decreased under stricter thresholds, but the major spatial patterns remained broadly consistent. Therefore, the main spatial conclusions were not substantially driven by the choice of the top-50% threshold. Because local spatial statistics over large raster datasets may be affected by multiple testing and by the choice of spatial weights matrix, the resulting clusters were interpreted as descriptive spatial aggregation patterns rather than as the sole criterion for restoration priority delineation.
2.5.2 Classification of ecological restoration priorities
Second, ecological restoration priorities were classified by overlaying ecosystem quality decline with the relative degradation-priority areas of the three ecosystem services. Because water conservation, soil retention, and vegetation carbon uptake capacity represent different but equally important ecological functions of the Mountains–Rivers Projects, the three ecosystem services were assigned equal weights in the baseline analysis. This avoids imposing subjective preferences among ecosystem service objectives while maintaining transparency in the prioritization process.
Ecological restoration priorities were classified using two binary dimensions: ecosystem quality decline and ecosystem service degradation. Ecosystem quality decline was identified based on statistically significant negative trends in EQI, whereas ecosystem service degradation was identified based on statistically significant negative trends in water conservation, soil retention, and vegetation carbon uptake capacity. First-priority areas were defined as areas where ecosystem quality decline overlapped with simultaneous degradation of all three ecosystem services. Second-priority areas were defined as areas where ecosystem quality decline overlapped with degradation of one or two ecosystem services. Third-priority areas were areas with ecosystem quality decline but without significant degradation of the three ecosystem services. Fourth-priority areas were areas with degradation of one or more ecosystem services but without significant ecosystem quality decline.
2.5.3 Assessment of relative implementation difficulty and ecology–cost trade-offs
Third, we assessed the relative implementation difficulty of Mountains–Rivers Projects. Because nationally consistent and spatially explicit data on land acquisition cost, labor cost, local fiscal capacity, and long-term maintenance cost were unavailable, we used terrain and accessibility proxies to approximate relative implementation difficulty. Specifically, elevation, slope, distance to settlements, and distance to roads were selected as relative implementation difficulty variables (Supplementary Table S1). Higher elevation, steeper slope, greater remoteness from settlements, and greater distance from roads were assumed to increase project implementation difficulty and cost.
Each cost-related variable was normalized and integrated into a relative cost index. Following previous large-scale restoration prioritization studies, we assumed that fixed costs, such as basic labor and infrastructure inputs, accounted for 70% of total project expenditure, while opportunity-related and accessibility-related costs accounted for 30% (; ). To evaluate the robustness of this assumption, an alternative 80:20 fixed-to-opportunity cost scenario was also tested (Supplementary Figure S2). The consistency between the two scenarios was assessed using spatial overlap and Spearman rank correlation of planning-unit priorities.
The ecological restoration priority layer and relative cost layer were then combined to identify ecology–cost trade-offs. Areas with high ecological priority and high cost were interpreted as strategic restoration zones requiring long-term investment and stronger policy support. Areas with moderate ecological priority and relatively low cost were interpreted as feasible restoration zones where targeted interventions could generate rapid ecological benefits under limited funding.
2.5.4 Identification of current spatial gaps in ecological protection spaces
Fourth, we diagnosed current spatial gaps in the Mountains–Rivers Projects by overlaying existing project boundaries with cross-administrative natural geographical units and national ecological protection spaces, which included key ecological function zones, important ecological function zones, biodiversity conservation priority areas, ecologically fragile zones, and major ecological supervision zones. Overlapping ecological protection categories were merged into a single composite protection-space layer to avoid double counting. Existing Mountains–Rivers Project areas were erased from this composite layer, and the remaining uncovered areas were defined as implementation gaps. Large natural geographical units were split according to their actual uncovered portions rather than being classified as gaps in their entirety. This step was intended to identify ecological blind spots and management gaps where ecological security functions are important but restoration projects are currently insufficient. In particular, the analysis considered the integrity of natural geographical units and the continuity of ecological processes across administrative boundaries, including mountain systems, watersheds, coastal zones, and land–sea transition regions.
2.5.5 Spatial optimization and connectivity-effect evaluation
Finally, we developed an optimized spatial layout for future Mountains–Rivers Projects by integrating ecological restoration priority, relative implementation difficulty, current spatial gaps, and the integrity of natural geographical units. Two optimization scenarios were considered.
The ecological-priority scenario aimed to maximize ecological protection benefits without first considering implementation cost constraints. Under this scenario, priority was given to areas with severe ecological quality decline, multiple ecosystem service degradation, and high ecological security importance.
The cost-priority scenario aimed to maximize ecological benefits under limited funding. Under this scenario, priority was given to areas with lower relative implementation difficulty while retaining ecological restoration value. This scenario reflects the practical need to improve restoration feasibility and cost-effectiveness.
The final optimized layout was generated by balancing ecological urgency and implementation feasibility while maintaining regional coordination between western and eastern China. This integrated optimization framework was designed to avoid two management risks: excluding ecologically urgent but high-cost western regions, and over-concentrating restoration investment in low-cost areas with relatively lower ecological urgency.
To evaluate the potential connectivity effects of the optimized layout, we constructed terrestrial ecological networks under two scenarios: the current project scenario, represented by the implemented Mountains–Rivers Projects, and the optimized project scenario, represented by the potential future layout proposed in this study. The resistance-surface-based network analysis was used for scenario comparison rather than as an independent validation of the prioritization framework. Existing project areas were therefore treated as the current implementation baseline. A composite resistance surface was developed using land-use type, slope, elevation, distance to roads, and distance to rivers (Supplementary Table S2). Based on this resistance surface, Circuitscape, Linkage Mapper, and Conefor were used to identify ecological corridors and calculate landscape connectivity indicators, including the delta Integral Index of Connectivity and delta Probability of Connectivity (; ; ). The same resistance surface was used for both scenarios. Source patches were defined using a minimum area threshold of 100 km2. dIIC was calculated using a 50 km distance threshold, and dPC was calculated using a negative exponential distance-decay function. Connectivity results were interpreted as structural and potential connectivity rather than direct evidence of species migration or gene flow. By comparing ecological corridors, dIIC, and dPC between the current and optimized scenarios, we assessed whether the proposed optimized layout indicated potential gains in structural connectivity and spatial coherence of the national ecological security network.
3 Results
3.1 Spatial patterns of ecological quality and ecosystem service decline
We focused exclusively on degraded ecosystems, defined as pixels exhibiting a statistically significant declining trend (Figure 3). The areas with declining ecological quality were primarily distributed across the Tianshan Mountains, the southwestern Tibetan Plateau, the Hunshandake Sandy Land, the Qilian Mountains, and the Hengduan Mountains. The areas with declines in the vegetation carbon uptake capacity were concentrated in southern China, including Fujian, Taiwan, and Hainan, as well as in the Tianshan Mountains, southern Tibet, and southern Yunnan. Reductions in the soil retention function predominantly occurred in western China, especially across the southwestern Tibetan Plateau, the Tianshan Mountains, and the Hengduan Mountains. In contrast, the declines in water conservation were more spatially extensive, affecting large parts of southern China, the Loess Plateau, and nearly the entire Qinling Mountain region.
FIGURE 3
To distinguish relatively stronger degradation within statistically degraded pixels, the top 50% with the greatest declining trends were defined as relative degradation-priority areas. Although the degradation of the three ecosystem services generally showed a broad westward intensification pattern, their relative degradation-priority areas differed substantially in spatial extent and overlap (Figure 4). Areas with simultaneous declines in vegetation carbon uptake capacity, soil retention, and water conservation were mainly concentrated in the Hengduan Mountains and parts of eastern Yunnan, accounting for only 0.3% of all degraded areas nationwide. In comparison, dual-function and single-function degradation areas accounted for 14.3% and 85.5% of degraded regions, respectively. Among the dual-function degradation types, the combined decline in vegetation carbon uptake capacity and water conservation was the largest and was mainly distributed in southern Tibet, Xishuangbanna, and Fujian Province.
FIGURE 4
3.2 Ecology–cost prioritization of restoration areas
To identify restoration areas that are both ecologically urgent and practically feasible, we integrated ecological restoration priority with relative implementation difficulty (Figure 5). The ecological-priority layer was first derived from the overlap between ecosystem quality decline and ecosystem service degradation, representing the urgency of restoration intervention (Figure 5a). First-priority areas were defined as regions where ecosystem quality decline coincided with simultaneous degradation of vegetation carbon uptake capacity, soil retention, and water conservation. These areas were limited in spatial extent and were mainly located in central Yunnan, but they represent the highest relative restoration priority under the adopted criteria because they combine structural ecosystem degradation with multifunctional service loss. Second-priority areas, where ecosystem quality decline overlapped with single- or dual-function degradation, were mainly distributed in the eastern and southwestern Qinghai–Tibet Plateau and the Tianshan Mountains. Third-priority areas were mainly located in central Inner Mongolia and parts of the Hengduan Mountains, where ecosystem quality declined but ecosystem service functions remained relatively intact. Fourth-priority areas occupied the largest area and were dominated by single-function degradation, especially water conservation decline.
FIGURE 5
The cost-priority layer was then constructed to represent the relative feasibility of project implementation (Figure 5b). Areas with lower relative costs were mainly distributed in eastern and central China, where terrain conditions are generally less complex, accessibility is higher, and project implementation is more feasible. In contrast, high-cost areas were concentrated in western China, particularly in the Qinghai–Tibet Plateau, the Tianshan Mountains, the Hengduan Mountains, and other remote alpine or mountainous regions. These regions have high ecological significance but face greater implementation constraints because of high elevation, steep terrain, remoteness, and limited infrastructure.
By overlaying the ecological-priority and cost-priority layers, we obtained an integrated ecology–cost prioritization map (Figure 5c). The integrated results reveal a clear spatial trade-off between ecological urgency and implementation feasibility. High ecological-priority and high-cost areas were mainly concentrated in western China, indicating regions where restoration is ecologically necessary but financially and technically more challenging. These areas should be regarded as strategic restoration zones requiring long-term investment, ecological compensation, adaptive monitoring, and process-oriented restoration. In contrast, areas with relatively low implementation costs and moderate ecological priority were mainly distributed in eastern, southern, and central China. These regions are more suitable for targeted and cost-efficient restoration interventions that can rapidly improve ecosystem service supply and reduce land-use conflicts.
3.3 Analysis of current spatial gaps in Mountains–Rivers Projects
Beyond ecological degradation and implementation costs, the current project layout was evaluated against cross-administrative natural geographical units and nationally designated ecological protection spaces (Figure 6). These included key ecological function zones, important ecological function zones, biodiversity conservation priority areas, ecologically fragile areas, and major ecological supervision zones. These areas constitute the spatial foundation of China’s ecological security framework and provide important reference units for restoration planning. The gap analysis showed that implementation gaps were more pronounced in mountainous and aquatic systems west of the Hu Huanyong Line. In western China, areas not covered by existing Mountains–Rivers Projects were mainly concentrated in large-scale arid alpine zones and plateau mountainous regions with fragile ecosystems and harsh natural conditions. These included water conservation and biodiversity conservation zones in the Qinghai–Tibet Plateau, the Pamir Plateau, and the southern foothills of the Altai Mountains, as well as water conservation and soil conservation zones in the southwestern Yunnan–Guizhou Plateau and the northwestern Hengduan Mountains.
FIGURE 6
In eastern China, implementation gaps were more scattered and mainly located in ecologically important but densely populated regions with intensive economic activities. These included the Hulunbuir grassland ecological function area, the northern agro-pastoral ecotone, estuarine and coastal protection areas in the Yangtze River Delta and Pearl River Delta, and coastal wetland conservation areas in the Beibu Gulf. Overall, western gap areas were dominated by large-scale and relatively intact natural ecosystems, where restoration is constrained by harsh environmental conditions and high costs. Eastern gap areas were mainly small-to medium-scale ecological units with high ecosystem service demand, where restoration is constrained by intensive human disturbance and competing land-use pressures.
3.4 Spatial optimization of future Mountains–Rivers Projects
By integrating ecosystem degradation, ecological restoration urgency, implementation costs, cross-administrative natural geographical units, and nationally important ecological protection spaces, we identified 24 potential optimized zones for future Mountains–Rivers Projects, covering a total area of 1.1174 million km2 (Figure 7). The optimized layout substantially strengthened the representation of western ecological security areas. 9 optimized zones were located west of the Hu Huanyong Line, covering approximately 0.81 million km2 and accounting for 69.45% of the total optimized area. Among these western zones, the Sanjiangyuan grassland–meadow–wetland region alone accounted for 42.07% of the western optimized area. The remaining 15 optimized zones were located east of the Hu Huanyong Line and were generally smaller in area, reflecting more fragmented restoration opportunities in densely populated regions.
FIGURE 7
According to the ecology–cost classification in Figures 7a–d, the ecology-first zones were mainly distributed in remote western and northern regions, including the Qinghai–Tibet Plateau, northwestern Yunnan, the Altai–Pamir region, Hulunbuir, Xilingol, and arid desert–grassland transition zones. These areas generally have high ecological urgency and high restoration costs, but they are irreplaceable for maintaining national water security, biodiversity conservation, grassland stability, and desertification control. Therefore, they require long-term, process-oriented restoration and sustained monitoring rather than short-term engineering intervention. In contrast, the cost-first zones were mainly located in eastern, southern, coastal, and river-basin regions, such as the Guangdong–Hong Kong–Macao Greater Bay Area, Beibu Gulf coastal wetlands, Dabie Mountains, Danjiangkou and its upstream region, the Yangtze River Delta, Poyang Lake, the middle Huai River, the Yellow River Delta wetland, and Taiping Lake.
Connectivity metrics suggested potential network-level effects of the optimized layout, but these results should be interpreted as changes in structural connectivity rather than independent evidence of ecological process improvement (Figure 8). Compared with the current scenario, the optimized scenario increased dIIC and dPC by 31.52% and 45.38%, respectively, indicating that the proposed layout may enhance the potential contribution of restoration patches to the overall ecological network and increase the probability of structural connectivity among patches. Spatially, the optimized scenario shifted the connectivity pattern from one dominated by a few large patches toward a more distributed configuration supported by multiple restoration nodes and corridor linkages. Some areas showed lower relative dIIC values after optimization, which may reflect a redistribution of network importance from isolated large patches to a broader set of connected restoration zones rather than an overall weakening of connectivity. The dPC pattern showed more widespread increases after optimization, particularly in newly added small and medium-sized zones located near corridor intersections. These zones, including the Wuliang and Laoshan Mountains, Dabie Mountains, and Danjiangkou upstream region, may function as potential stepping-stone nodes that improve structural links among larger ecological source areas. Overall, the optimized layout indicates a potential improvement in the spatial configuration and structural connectivity of future Mountains–Rivers Projects, but the results should be interpreted with caution because connectivity metrics are sensitive to differences in source definition, patch area, patch number, and spatial distribution between scenarios.
FIGURE 8
4 Discussion
4.1 Ecological quality and ecosystem service decline in restoration prioritization
This study reveals a clear spatial mismatch between the current distribution of the Mountains–Rivers Projects and areas with high ecological degradation risk. Although the implemented projects have achieved important restoration outcomes in several regions, some areas experiencing significant declines in ecosystem quality and ecosystem service functions remain insufficiently covered (; ). This mismatch indicates that the current project layout has not fully responded to the spatial heterogeneity of ecological degradation, especially in ecologically fragile western and southwestern regions (; Xu et al., 2022a).
Our results show that declines in ecosystem quality, vegetation carbon uptake capacity, soil retention, and water conservation were spatially clustered rather than randomly distributed. This suggests that restoration demand is concentrated in specific ecological zones where climate stress, terrain constraints, land-use pressure, and ecosystem fragility interact (Xu et al., 2022b). Notably, areas with simultaneous degradation of vegetation carbon uptake capacity, soil retention, and water conservation accounted for only 0.3% of degraded land. Although limited in area, these multifunctional relative degradation-priority areas are strategically important because they indicate a higher risk of ecological instability and ecosystem service loss (). Therefore, they may warrant priority consideration in future restoration planning.
The ecology–cost scenario analysis further confirms this trade-off. An ecology-priority strategy would concentrate restoration efforts in western China, especially in the Qinghai–Tibet Plateau, Tianshan Mountains, Hengduan Mountains, and arid desert–grassland transition zones. These areas are ecologically critical but costly to restore because of high elevation, complex terrain, remoteness, and limited infrastructure (). By contrast, a cost-priority strategy would shift restoration toward more accessible eastern and central regions, whererelative implementation difficulty are lower and management capacity is stronger (). Therefore, neither an ecology-only nor a cost-only strategy is sufficient. Future Mountains–Rivers Projects should adopt an integrated ecology–cost approach that protects high-urgency western ecosystems while also using low-cost eastern areas to achieve targeted and efficient restoration benefits.
The identification of the central Yunnan region as the only first priority restoration area reflects the ecological significance and vulnerability of this mountainous biodiversity hotspot (; Zhang et al., 2025). However, the mismatch between the current restoration project deployment (mainly focused on fourth priority areas in the eastern region) and the ecologically vulnerable western regions indicates that the current project allocation does not fully meet the ecological needs. This misalignment has been noted in previous national-level recovery plans, that is, projects are preferentially implemented in areas with better accessibility and lower costs, rather than in the areas with the highest ecological urgency (). Our scenario analysis demonstrates that purely ecological priority strategies concentrate efforts in remote western regions such as the Tibetan Plateau and the Tianshan Mountains, while cost-minimization strategies shift focus to more accessible eastern areas. These contrasting patterns highlight a long-standing conflict in restoration planning - balancing ecological benefits and cost efficiency. This has been reported in large-scale restoration contexts such as Brazil’s Atlantic Forest and Africa’s Great Green Wall (; ). The proposed integrated ecology–cost strategy offers a more spatially equitable approach, ensuring that high-priority but cost-effective areas, particularly in eastern lowlands, are not overlooked.
4.2 Improving cross-boundary coordination and ecological network connectivity
A key contribution of this study is that it shifts the spatial optimization of the Mountains–Rivers Projects from an administratively segmented perspective toward a framework based on natural geographical units, ecological protection spaces, and landscape connectivity. Ecological processes such as hydrological regulation, soil retention, and ecosystem service flows are not confined by administrative boundaries (Wang et al., 2019). However, restoration projects are often planned and implemented within administrative jurisdictions, which may weaken cross-regional coordination and reduce restoration effectiveness.
The gap analysis shows clear east–west differences in current project coverage. In western China, implementation gaps are mainly distributed in large-scale alpine, arid, grassland, desert, and plateau mountain ecosystems, including the Qinghai–Tibet Plateau, the Pamir Plateau, the southern foothills of the Altai Mountains, the Yunnan–Guizhou Plateau, and the Hengduan Mountains. These regions generally have high ecological sensitivity, low natural recovery capacity, harsh environmental conditions, and limited accessibility. Restoration in these areas should emphasize long-term ecological protection, natural regeneration, grazing regulation, desertification control, ecological compensation, and adaptive monitoring ().
In eastern China, the gaps are more fragmented and are mainly located in densely populated regions with high ecosystem service demand, such as estuarine wetlands, coastal zones, agro-pastoral transition zones, and river delta areas. These areas face stronger land-use conflicts and human disturbance (Zhao and Dai, 2024). Therefore, restoration in eastern and coastal regions should focus on targeted interventions, including wetland restoration, river corridor rehabilitation, ecological buffer construction, pollution control, and habitat reconnection.
The optimized layout identified 24 future implementation zones covering 1.1174 million km2. Among them, nine zones were located west of the Hu Huanyong Line and accounted for 69.45% of the total optimized area. This indicates that the proposed framework partially corrects the current east-biased distribution of restoration investment and improves the representation of western ecological security barriers. Meanwhile, the 15 optimized zones east of the Hu Huanyong Line remain important for coastal protection, flood regulation, water supply security, and ecosystem service provision in densely populated regions.
The optimized scenario indicated potential gains in structural ecological connectivity. After optimization, the number of ecological corridors increased by 18.7%, while dIIC and dPC increased by 31.52% and 45.38%, respectively. These improvements suggest that the optimized projects not only expand restoration coverage but also improve the spatial organization of restoration areas (Wang et al., 2019). The increase in dIIC indicates stronger contribution of optimized zones to overall network integrity, while the increase in dPC reflects improved potential connectivity among ecological source areas and corridor nodes. Therefore, future restoration performance should not be evaluated only by restored area or vegetation recovery, but also by whether projects improve ecological corridors, structural connectivity and spatial coherence of ecological networks.
4.3 Management implications and limitations
The findings provide several implications for improving the environmental management effectiveness of the Mountains–Rivers Projects. First, future project allocation should give greater weight to ecological degradation risk and multifunctional ecosystem service decline. Areas where ecosystem quality degradation overlaps with declines in multiple ecosystem services should be prioritized, even when their spatial extent is limited. Second, restoration planning should balance ecological urgency and cost feasibility. High-urgency western regions require sustained funding, technical support, ecological compensation, and long-term monitoring, while lower-cost eastern regions are suitable for targeted and rapid restoration interventions. Third, project planning should better align with natural geographical units, watersheds, ecological protection redlines, biodiversity conservation priority areas, and major ecological supervision zones. This would help strengthen cross-boundary coordination and improve the coherence of national ecological security networks.
Several limitations should also be acknowledged. First, although ANUSPLIN was used to generate nationally consistent 1 km climate surfaces and cross-validation was conducted to evaluate interpolation performance, station density is spatially uneven across China. In particular, western China contains extensive plateau and mountainous regions where meteorological stations are relatively sparse and terrain complexity is high. This may increase interpolation uncertainty and affect the representation of local climatic gradients. Therefore, the climate-related results should be interpreted as national-to regional-scale background information rather than site-level climatic diagnosis. Future studies could further improve climatic inputs by integrating denser station observations, reanalysis datasets, and multi-source climate data fusion. Second, the prioritization framework relies on equal weighting and threshold-based classification of ecological indicators. Although this improves transparency and comparability at the national scale, it may not fully capture regional differences in restoration objectives. Future studies could incorporate expert knowledge, stakeholder preferences, and systematic conservation planning tools such as Zonation, Marxan, or Prioritizr to improve optimization robustness. Third, implementation feasibility was represented by relative difficulty proxies, including slope, elevation, and accessibility. These variables capture broad-scale restoration feasibility but do not fully represent actual economic costs, land tenure, local fiscal capacity, opportunity costs, labor costs, or long-term maintenance costs. The 70:30 and 80:20 cost-feasibility scenarios should therefore be interpreted as heuristic assumptions rather than empirically calibrated cost structures. Fourth, the ecological indicators used in this study mainly emphasize vegetation structure, productivity, and ecosystem service functions. Although landscape connectivity was evaluated after optimization, biodiversity-specific indicators such as species richness, habitat suitability, population viability, and functional diversity were not directly incorporated into the prioritization model. Future research should integrate biodiversity monitoring data, species distribution models, and habitat fragmentation metrics to support more comprehensive restoration planning.
This study provides a scenario-based national screening framework for supporting the future spatial planning of Mountains–Rivers Projects. The proposed optimized zones should be interpreted as potential priority areas identified from national-scale ecological degradation, implementation feasibility, spatial gaps, and connectivity considerations, rather than as final investment or implementation boundaries. Before practical implementation, local feasibility, land tenure, restoration type, stakeholder acceptance, biodiversity value, empirical cost data, and governance capacity should be further assessed.
5 Conclusion
This study developed a national-scale spatial prioritization framework to support the optimization of China’s Mountains–Rivers Projects by integrating ecosystem degradation, ecosystem service decline, implementation cost, cross-administrative ecological gaps, nationally important ecological protection spaces, and landscape connectivity. The results showed that declines in ecosystem quality and ecosystem services were spatially clustered, with high-urgency restoration needs mainly concentrated in western and southwestern China. However, existing Mountains–Rivers Projects remain disproportionately distributed in eastern China and are insufficiently aligned with several ecologically fragile and functionally critical regions. This mismatch indicates that future restoration planning should move beyond administratively convenient project allocation and more directly respond to ecological degradation risk.
By jointly considering ecological urgency and cost feasibility, this study identified 24 potential optimized zones covering 1.1174 million km2. Nine of these zones were located west of the Hu Huanyong Line and accounted for 69.45% of the total optimized area, indicating that the proposed framework partially corrects the current east-biased distribution of restoration investment. The optimized scenario indicated potential structural connectivity gains, with the number of ecological corridors increasing by 18.7%, while dIIC and dPC increased by 31.52% and 45.38%, respectively. These results suggest that, under the scenario assumptions, the proposed layout may improve the structural coherence of the ecological network, but the optimized zones should be interpreted as national-scale screening outputs rather than final project boundaries.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
XX: Conceptualization, Formal Analysis, Funding acquisition, Software, Supervision, Visualization, Writing – original draft. DL: Investigation, Project administration, Writing – original draft. FJ: Data curation, Resources, Writing – original draft. XM: Resources, Validation, Writing – original draft. NL: Data curation, Validation, Writing – original draft. JL: Methodology, Visualization, Writing – original draft. CZ: Funding acquisition, Resources, Writing – original draft. JQ: Resources, Validation, Writing – original draft. KZ: Project administration, Visualization, Writing – original draft.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by Natural Science Foundation of Jiangsu Province (BK20240277), Jing-Jin-Ji Regional Integrated Environmental Improvement-National Science and Technology Major Project (2026ZD1210004), the Science and Technology Fundamental sources Investigation Program of China (2023FY100101), the Special Fund of the Chinese Central Government for Basic Scientific Research Operations in the commonweal Research Institute (GYZX250107).
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fenvs.2026.1809482/full#supplementary-material
References
1
AslanC. E.BrunsonM. W.SikesB. A.Epanchin-NiellR. S.VelozS.TheobaldD. M.et al (2021). Coupled ecological and management connectivity across administrative boundaries in undeveloped landscapes. Ecosphere12 (1), e03329. 10.1002/ecs2.3329
2
Álvarez-RomeroJ. G.MillsM.AdamsV. M.GurneyG. G.PresseyR. L.WeeksR.et al (2018). Research advances and gaps in marine planning: towards a global database in systematic conservation planning. Biol. Conserv.227, 369–382. 10.1016/j.biocon.2018.06.027
3
BakerD. J.GastonK. J.MetcalfeK.MacleanI. M. D. (2025). Systematic conservation planning for nature recovery. Bioscience75, 472–489. 10.1093/biosci/biaf030
4
BaróF.PalomoI.ZulianG.VizcainoP.HaaseD.Gómez-BaggethunE. (2016). Mapping ecosystem service capacity, flow and demand for landscape and urban planning: A case study in the Barcelona metropolitan region. Land Use Policy57, 405–417. 10.1016/j.landusepol.2016.06.006
5
BrancalionP. H. S.MeliP.TymusJ. R. C.LentiF. E. B.M. BeniniR.SilvaA. P. M.et al (2019). What makes ecosystem restoration expensive? A systematic cost assessment of projects in Brazil. Biol. Conserv.240, 108274. 10.1016/j.biocon.2019.108274
6
CaoY.YangR.CarverS. (2020). Linking wilderness mapping and connectivity modelling: a methodological framework for wildland network planning. Biol. Conserv.251, 108679. 10.1016/j.biocon.2020.108679
7
ChangB.ChenB.ChenW.XuS.HeX.YaoJ.et al (2024). Analysis of trade-off and synergy of ecosystem services and driving forces in urban agglomerations in northern China. Ecol. Indic.165, 1–12. 10.1016/j.ecolind.2024.112210
8
ChazdonR. L. (2008). Beyond deforestation: restoring forests and ecosystem services on degraded lands. Science320, 1458–1460. 10.1126/science.1155365
9
ChenC.ParkT.WangX.PiaoS.XuB.ChaturvediR. K.et al (2019). China and India lead in greening of the world through land-use management. Nat. Sustain2, 122–129. 10.1038/s41893-019-0220-7
10
ChenX.YuL.DuZ.XuY.ZhaoJ.ZhaoH.et al (2022). Distribution of ecological restoration projects associated with land use and land cover change in China and their ecological impacts. Sci. Total Environ.825, 153938. 10.1016/j.scitotenv.2022.153938
11
ChenW.GuT.XiangJ.LuoT.ZengJ.YuanY. (2024). Ecological restoration zoning of territorial space in China: an ecosystem health perspective. J. Environ. Manage364, 121371. 10.1016/j.jenvman.2024.121371
12
ChoiC.-Y.ShiX.ShiJ.GanX.WenC.ZhangJ.et al (2022). China’s ecological conservation redline policy is a new opportunity to meet post-2020 protected area targets. Conservation Letters15 (2), e12853. 10.1111/conl.12853
13
DicksonB. G.AlbanoC. M.AnantharamanR.BeierP.FargioneJ.GravesT. A.et al (2019). Circuit-theory applications to connectivity science and conservation. Conserv. Biol.33, 239–249. 10.1111/cobi.13230
14
FuM.WangJ.ZhuY.ZhangY. (2023). Evaluation of the protection effectiveness of natural protected areas on the Qinghai—Tibet Plateau based on ecosystem services. International Journal of Environmental Research and Public Health20 (3), 2605. 10.3390/ijerph20032605
15
HarrisL. R.HolnessS. D. (2023). A practical approach to setting heuristic marine biodiversity targets for systematic conservation planning. Biol. Conserv.285, 110218. 10.1016/j.biocon.2023.110218
16
HeywoodV. H. (2019). Conserving plants within and beyond protected areas - still problematic and future uncertain. Plant divers.41, 36–49. 10.1016/j.pld.2018.10.001
17
HollK. D.HowarthR. B. (2001). Paying for restoration. Restor. Ecol.8, 260–267. 10.1046/j.1526-100x.2000.80037.x
18
JiangC.ZhangH.ZhangZ. (2018). Spatially explicit assessment of ecosystem services in China's loess Plateau: patterns, interactions, drivers, and implications. Glob. Planet. Change161, 41–52. 10.1016/j.gloplacha.2017.11.014
19
KoenE. L.EllingtonE. H.BowmanJ. (2019). Mapping landscape connectivity for large spatial extents. Landsc. Ecol.34, 2421–2433. 10.1007/s10980-019-00897-6
20
KongX.FuM.ZhaoX.WangJ.JiangP. (2022). Ecological effects of land-use change on two sides of the Hu huanyong line in China. Land Use Policy113, 105895. 10.1016/j.landusepol.2021.105895
21
KoolJ. T.MoilanenA.TremlE. A. (2012). Population connectivity: recent advances and new perspectives. Landsc. Ecol.28, 165–185. 10.1007/s10980-012-9819-z
22
LiM.LiangD.XiaJ.SongJ.ChengD.WuJ.et al (2021). Evaluation of water conservation function of danjiang river basin in qinling Mountains, China based on InVEST model. J. Environ. Manage286, 112212. 10.1016/j.jenvman.2021.112212
23
LiW.MigliavaccaM.ForkelM.DenissenJ. M. C.ReichsteinM.YangH.et al (2022). Widespread increasing vegetation sensitivity to soil moisture. Nat. Commun.13, 3959. 10.1038/s41467-022-31667-9
24
LiangJ.HeX.ZengG.ZhongM.GaoX.LiX.et al (2018). Integrating priority areas and ecological corridors into national network for conservation planning in China. Science of the Total Environment626, 22–29. 10.1016/j.scitotenv.2018.01.086
25
LichtenbergD.KreuzbergE.von DürckheimK.ReevesB.BerningR.SnakenborgS. K.et al (2025). Landscape connectivity for biodiversity conservation: a mammal-based multi-species corridor approach for the eden to addo corridor initiative, South Africa. Biodivers. Conservation34, 3933–3953. 10.1007/s10531-025-03140-8
26
LiuJ.PeiX.ZhuW.JiaoJ. (2023). Scenario modeling of ecosystem service trade-offs and bundles in a semi-arid valley basin. Sci. Total Environ.896, 166413. 10.1016/j.scitotenv.2023.166413
27
LiuY.WangJ.WuZ.LiS.DaiW. (2024). Identification of biodiversity priority conservation areas in China by integrating genetic, species and ecosystem diversity. Biol. Conserv.300, 110854. 10.1016/j.biocon.2024.110854
28
LouH.ShiX.RenX.YangS.CaiM.PanZ.et al (2024). Limited terrestrial carbon sinks and increasing carbon emissions from the Hu line spatial pattern perspective in China. Ecol. Indic.162, 112035. 10.1016/j.ecolind.2024.112035
29
LuH.DingY.ZhangJ.WuW.XuD. (2025). Carbon reduction effect of comprehensive land consolidation and its configuration paths at the township level: a case study of Zhejiang Province, China. J. Environ. Manage373, 123855. 10.1016/j.jenvman.2024.123855
30
MargulesC. R.PresseyR. L. (2000). Systematic conservation planning. Nature405, 243–253. 10.1038/35012251
31
MotaF. M. M.Lima-SantosD.Kenedy-SiqueiraW.OkiY.FernandesG. W. (2025). Climate-driven distribution shifts of invasive earthworm species in a river basin affected by mining tailings. Environ. Monit. Assess.197, 879. 10.1007/s10661-025-14196-y
32
OuyangZ.ZhengH.XiaoY.PolaskyS.LiuJ.XuW.et al (2016). Improvements in ecosystem services from investments in natural capital. Science352, 1455–1459. 10.1126/science.aaf2295
33
PengJ.XuD.XuZ.TangH.JiangH.DongJ.et al (2024). Ten key issues for ecological restoration of territorial space. Natl. Sci. Rev.11, nwae176. 10.1093/nsr/nwae176
34
PuJ.ZhaoX.HuangP.GuZ.ShiX.ChenY.et al (2022). Ecological risk changes and their relationship with exposed surface fraction in the karst region of southern China from 1990 to 2020. J. Environ. Manage323, 116206. 10.1016/j.jenvman.2022.116206
35
SchirpkeU.CandiagoS.Egarter ViglL.JägerH.LabadiniA.MarsonerT.et al (2019). Integrating supply, flow and demand to enhance the understanding of interactions among multiple ecosystem services. Science of the Total Environment651, 928–941. 10.1016/j.scitotenv.2018.09.235
36
StrassburgB. B. N.IribarremA.BeyerH. L.CordeiroC. L.CrouzeillesR.JakovacC. C.et al (2020). Global priority areas for ecosystem restoration. Nature586, 724–729. 10.1038/s41586-020-2784-9
37
SunH.WeiJ.HanQ. (2022). Assessing land-use change and landscape connectivity under multiple green infrastructure conservation scenarios. Ecol. Indic.142, 109236. 10.1016/j.ecolind.2022.109236
38
SvenssonJ.BubnickiJ. W.JonssonB. G.AnderssonJ.MikusińskiG. (2020). Conservation significance of intact forest landscapes in the Scandinavian Mountains green belt. Landsc. Ecol.35, 2113–2131. 10.1007/s10980-020-01088-4
39
WangP.WuX.HaoY.WuC.ZhangJ. (2019). Is southwest China drying or wetting? Spatiotemporal patterns and potential causes. Theor. Appl. Climatol.139, 1–15. 10.1007/s00704-019-02935-4
40
WangX.WangX.ZhangX.ZhouJ.JiaZ.MaJ.et al (2024). Ecological barriers: an approach to ecological conservation and restoration in China. Ambio53, 1077–1091. 10.1007/s13280-024-01988-8
41
XiaM.JiaK.ZhaoW.LiuS.WeiX.WangB. (2021). Spatio-temporal changes of ecological vulnerability across the Qinghai-Tibetan Plateau. Ecol. Indic.123, 107274. 10.1016/j.ecolind.2020.107274
42
XuW.XiaoY.ZhangJ.YangW.ZhangL.HullV.et al (2017). Strengthening protected areas for biodiversity and ecosystem services in China. Proceedings of the National Academy of Sciences of the United States of America114 (7) 1601–1606. 10.1073/pnas.1620503114
43
XuX.JiaoF.LiuH.GongH.ZouC.LinN.et al (2022a). Persistence of increasing vegetation gross primary production under the interactions of climate change and land use changes in northwest China. Sci. Of Total Environ.834, 155086. 10.1016/j.scitotenv.2022.155086
44
XuX.LiuJ.JiaoF.ZhangK.YeX.GongH.et al (2022b). Ecological engineering induced carbon sinks shifting from decreasing to increasing during 1981-2019 in China. Sci. Total Environ.864, 161037. 10.1016/j.scitotenv.2022.161037
45
YuanL.GengM.LiF.XieY.TianT.ChenQ. (2024). Spatiotemporal characteristics and drivers of ecosystem service interactions in the Dongting Lake basin. Sci. Total Environ.926, 172012. 10.1016/j.scitotenv.2024.172012
46
ZellerK. A.McGarigalK.WhiteleyA. R. (2012). Estimating landscape resistance to movement: a review. Landsc. Ecol.27, 777–797. 10.1007/s10980-012-9737-0
47
ZhangY.ZhaoX.GongJ.LuoF.PanY. (2024). Effectiveness and driving mechanism of ecological restoration efforts in China from 2009 to 2019. Sci. Of Total Environ.910, 168676. 10.1016/j.scitotenv.2023.168676
48
ZhangZ.FangF.JiQ.YaoY.YuL.FeiL. (2025). Ecological management zoning based on the ecosystem service supply–demand balance and the driving processes of the influencing factors in China. Environ. Sustain. Indic.27, 100771. 10.1016/j.indic.2025.100771
49
ZhaoZ.DaiE. (2024). Vegetation cover dynamics and its constraint effect on ecosystem services on the Qinghai-Tibet Plateau under ecological restoration projects. J. Environ. Manage356, 120535. 10.1016/j.jenvman.2024.120535
Summary
Keywords
ecological restoration, ecosystem function, landscape connectivity, management feasibility, Mountains–Rivers projects, spatial prioritization
Citation
Xu X, Lin D, Jiao F, Mu X, Lin N, Liu J, Zou C, Qiu J and Zhang K (2026) Rebalancing large-scale ecological restoration investments: a national spatial prioritization framework for China’s mountains–rivers projects. Front. Environ. Sci. 14:1809482. doi: 10.3389/fenvs.2026.1809482
Received
18 February 2026
Revised
08 July 2026
Accepted
20 July 2026
Published
26 August 2026
Volume
14 - 2026
Edited by
Fanglei Zhong, Minzu University of China, China
Reviewed by
Gowhar Meraj, American University of Sharjah, United Arab Emirates
Jinhua Liu, Shandong Jianzhu University, China
Updates
Copyright
© 2026 Xu, Lin, Jiao, Mu, Lin, Liu, Zou, Qiu and Zhang.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Xiaojuan Xu, kattyxiaoj@163.com; Jie Qiu, 1669733424@qq.com; Kun Zhang, 2652667020@qq.com
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.