Comprehensive soil and water loss control on the Loess Plateau is a central component of ecological civilization construction in the region and a fundamental measure for ensuring ecological protection and high-quality development in the Yellow River Basin. Through decades of sustained soil and water conservation efforts, vegetation coverage has been significantly restored, soil erosion has been effectively curbed, and the ecological environment has markedly improved on the Loess Plateau, achieving a historic transformation from “yellow” to “green”. Meanwhile, living standards have continued to improve, and the spatial structure of production, living, and ecological functions has been progressively optimized. Nevertheless, soil erosion control on the Loess Plateau is still confronted with a series of serious challenges. These include a significant increase in the frequency of extreme rainfall events and associated erosion disasters; changes in the dominant types and severity of soil erosion; declining vegetation stability and ecosystem sustainability. Additional challenges include intensified wind-blown sediment deposition and channel siltation in the “Jiziwan” section of the Yellow River, insufficient development of technologies for monitoring, quantifying, and enhancing soil and water conservation carbon sinks, and growing impacts of modern high-tech development on traditional soil and water conservation practices. Future efforts should prioritize the quantitative characterization and intelligent monitoring of soil erosion processes, the development of precision-targeted erosion control strategies, the establishment of adaptive technical systems and governance paradigms for soil erosion management, and studies on key technologies for environmental remediation and comprehensive rural revitalization in erosion-prone areas. These efforts will provide critical scientific and technological support for integrated soil erosion control on the Loess Plateau and for ecological protection and high-quality development in the Yellow River Basin.

Realizing the value of ecological products is an important pathway for promoting ecological protection and high-quality development in the Yellow River Basin. Under multi-level governance structures and conditions of information asymmetry, how to design effective incentive mechanisms to enhance genuine ecological investment and improve the efficiency of ecological compensation has become an urgent issue. Based on the realities of ecological governance in the Yellow River Basin, this paper constructs a three-stage game model involving the central government, local governments, and ecological suppliers. It considers three scenarios: the socially optimal benchmark under complete information, conventional performance-based compensation under information asymmetry, and the introduction of an audit accountability mechanism. The paper systematically examines the effects of ecological compensation, local supervision, and audit constraints on ecological product supply and the efficiency of value realization. The results show that:(1)Increasing the intensity of ecological compensation can encourage ecological suppliers to increase genuine ecological investment and expand certified performance as well as the scale of ecological product supply. However, under information asymmetry, it may also simultaneously strengthen incentives for certification bias, resulting in the coexistence of expanded supply scale and insufficient supply authenticity.(2)Local government supervision serves as a key transmission mechanism linking fiscal incentives with genuine ecological improvement. Strengthened supervision can not only reduce certification bias and improve performance identification, but also increase the returns to genuine ecological investment, thereby enhancing the efficiency of ecological product supply.(3)Compared with conventional performance-based compensation, the audit accountability mechanism, when costs are controllable, can strengthen local supervision, increase genuine ecological investment, reduce certification bias, and narrow performance deviations, thereby moving institutional outcomes closer to the socially optimal allocation and alleviating compensation mismatch under information asymmetry.(4)As the basic returns of ecological products increase, social welfare rises overall, while the optimal performance-based ecological transfer payment gradually declines and approaches zero. This indicates that the enhanced revenue-generating capacity of ecological products helps reduce institutional dependence on fiscal compensation, although institutional optimization remains indispensable. The findings of this paper contribute to a deeper theoretical understanding of the value realization mechanism of ecological products and provide policy implications for optimizing the ecological compensation system in the Yellow River Basin and improving a coordinated governance framework that integrates ecological compensation, local supervision, and audit accountability.

The Yellow River Basin is both a major ecological barrier and a key grain-producing region in China. Promoting synergy between agricultural ecology and rural revitalization is therefore essential to the national strategy for ecological conservation and high-quality development. Using 199 papers from Peking University Core Journals and CSSCI journals retrieved from CNKI, this study applies CiteSpace to conduct keyword co-occurrence and cluster analyses and then systematically reviews the research trajectory, core issues, and feedback mechanisms in this field. The bibliometric results reveal a multidimensional knowledge structure encompassing basin ecological conservation, rural revitalization practice, spatial optimization and governance, synergy mechanisms, and ecological value realization. Research attention has shifted from separate discussions of ecological governance and rural revitalization toward coupling coordination, driving factors, and the value realization of agricultural ecological products, although systematic investigation of their co-evolution remains insufficient. The review further shows that agricultural ecosystems support rural revitalization through food provision, water retention, soil conservation, biodiversity conservation, and carbon sequestration, while rural revitalization generates positive ecological feedback through ecological restoration and transitions toward green production. Three representative models have emerged: ecological industrialization, industrial ecologization, and ecological compensation and value realization. Their further development, however, is constrained by rigid water-resource limits and inadequate technological adaptation, weak incentives and capacities among agricultural operators for green transition, and fragmented cross-regional and interdepartmental governance. Future research should build an integrated ecological-economic-social framework, strengthen multiscale integration, long-term dynamic analysis, and household-level behavioral research, and promote systemic policy coordination. These efforts can provide theoretical and practical support for the deep integration of ecological conservation and rural revitalization in the Yellow River Basin.

The effect of livelihood transformation on dependence intensity on ecosystem services(DIES)has emerged as a core issue in understanding human-land system evolution, while existing literature has overlooked its nonlinear characteristics. This study selected farmers from three mineral-agricultural composite areas in the middle reaches of the Yellow River Basin as samples, and categorized their farm household livelihood status into non-transformed, agriculture-dominated transformed, and non-agriculture-dominated transformed groups. A random forest model was employed to explore the nonlinear effect of livelihood transformation on DIES. The results showed that:(1)The transformed group had a significantly higher DIES value(0.378±0.134)than the non-transformed group(0.348±0.127), whereas no significant difference existed within the transformed subgroups;(2)Livelihood transformation enhanced the influencing intensity of cultivated land area, residential distance to town, family dependency ratio, labor capacity and participation frequency in collective activities on DIES;(3)The relationships of residential distance to mining area, educational attainment and health status with DIES exhibited obvious heterogeneity between the two groups. Specifically, residential distance to mining area showed a V-shaped correlation with DIES for transformed farmers, and educational attainment presented a senior high school threshold effect. This paper revealed that livelihood transformation did not inevitably reduce farmers reliance on ecosystem services. Affected by the distinctive industrial context of mineral-agricultural composite areas, transformed farmers suffered a higher risk of ecological dependence. This study provides a scientific basis for the formulation of differentiated ecological compensation policies.

The Yellow River Basin serves as a vital ecological barrier and socioeconomic hub in China, and its matching of ecosystem service supply and demand will directly influence regional environmental security and the achievement of high-quality development goals. This study quantified the evolution and matching patterns of supply and demand for five key ecosystem services in the Yellow River Basin using spatiotemporal data, and predicted potential changes in ecological value based on future land use under four different development scenarios: natural development, ecological conservation, economic development, and farmland protection. Thus, a progressive analytical framework, “historical dynamic evolution of supply and demand-future multi-scenario valuation”, was established. The study found that the supply and demand of five ecosystem services displayed varied trends during 2000-2020. The supply of food production and soil conservation increased continuously, and the demand for carbon sequestration, water yield, and food production also showed a steady increase, while the supply of water yield and the demand for soil conservation exhibited a decrease followed by an increase. Five ecosystem services exhibited surpluses across most regions of the Yellow River Basin. While the supply-demand balance for soil conservation improved in some counties in Shaanxi and Shanxi, the lower reaches of the basin faced a risk of carbon sequestration supply-demand imbalance. Among the four development scenarios, the ecological conservation scenario was the most conducive to preserving the provisioning capacity of ecosystem services, with 70.42% of counties experiencing an increase in ecological value under this scenario. The subsequent scenarios were farmland protection, natural development, and economic development, which progressively weakened the provisioning capacity of ecosystem services. Based on a dual perspective that encompasses historical evolutionary patterns and future development projections, this study provides a scientific basis for the sustainable management of the supply-demand relationship for ecosystem services in the Yellow River Basin, thereby enhancing regional ecological security and contributing to high-quality development.

Systematically analyzing the spatiotemporal interaction between urbanization and typical ecosystem services in the lower Yellow River is of great practical value for implementing the major national strategy of the Yellow River Basin and constructing the basins ecological security pattern. This study took 27 counties in the Shandong section of the Yellow River Basin as the study area. Based on multi-source data from 2000 to 2020, we employed the InVEST model, Pearson correlation analysis, coupling coordination degree model, and Tapio decoupling analysis method to systematically reveal the bidirectional association characteristics, static coordination levels, and dynamic evolutionary relationships between urbanization and ecosystem services. Through the superposition analysis of coupling coordination degree and decoupling effects, a typological classification of county-level development paths was achieved. The results indicated that:(1)From 2000 to 2020, the comprehensive urbanization index of the study area increased from 0.184 to 0.282, forming a spatial pattern with Jinan as the core and gradient decreasing outward, among which the land urbanization index showed the most prominent increase.(2)During the study period, the comprehensive ecosystem service index showed an overall upward trend. Specifically, the average water yield increased by 42.94%, exhibiting a spatial distribution of “low in the northeast and high in the central-south”; the average food production registered a cumulative increase of 41.77%, maintaining a pattern of “high in the southwest and low in the northeast”; total carbon storage decreased by 4.76%, primarily driven by the conversion of cultivated land to construction land.(3)Pearson correlation analysis and coupling coordination degree analysis showed that urbanization and ecosystem services had an overall negative correlation relationship. The average annual negative correlation coefficient between land urbanization and food production and carbon storage was approximately -0.700. The coupling coordination degree between the two was dominated by basic coordination, with the coordination imbalance in the core urban area of Jinan and some rapid urbanization expansion areas continuing to intensify.(4)During the study period, decoupling states were mainly non-ideal states such as strong negative decoupling and weak negative decoupling. The zoning type was dominated by “coordination-negative decoupling”, indicating that although most counties maintained a basic coordination state, urbanization development has led to obvious ecosystem service degradation. In the future, based on different zoning types, differentiated spatial control and quality improvement strategies should be implemented respectively in various regions of the Shandong section of the Yellow River Basin, to achieve orderly and coordinated advancement of regional urban development and ecological environmental protection.

Wetland ecosystem service valuation is essential for quantitatively assessing ecological functions and informing better resource management and conservation policies. In this study, a combination of valuation methods, including the market value method, replacement cost method, pollutant reduction value method, and travel cost method, was used to assess 13 ecosystem service functions of the Xiaohaierkou wetland across three broad categories: provisioning, regulating, and social services. The results showed that the wetland provides an annual total ecosystem service value of approximately CNY 374 million, with regulating services contributing the largest share at about 81.3%. Among these, water purification, climate regulation, and water storage and supply were particularly important in supporting regional ecological security and improving local living conditions. Provisioning services accounted for about 5.8% of the total value, mainly through aquatic product production and plant resource supply, indicating the wetlands relatively stable productive capacity. Social services contributed about 12.9%, reflecting substantial value in recreation and tourism, science education, cultural communication, and the promotion of public physical and mental well-being. These findings highlight the broad benefits of urban wetlands in public service provision and ecological civilization development. The study further suggests that the wetland regulating capacity and social service functions could be enhanced through refined hydrological management, stronger vegetation maintenance, and improved functional zoning. Overall, this research not only provides a systematic valuation of the ecosystem services supplied by the Xiaohaierkou wetland, but also offers a scientific basis for its conservation, adaptive management, and the realization of its ecological value.

The migration pattern of river shorelines is one of the important bases for comprehensive riparian management. Taking the main stream of the Datong River on the northeastern margin of the Qinghai-Tibet Plateau as the study object, this research selected two reaches: the natural shoreline reach(Qilian section)dominated by water conservation functions, and the engineered shoreline reach(Menyuan section)dominated by comprehensive agricultural management functions. Based on the digital shoreline analysis system(DSAS)model, combined with multi-source remote sensing and hydrological reanalysis data, this study quantitatively analyzed the evolutionary characteristics of alpine river shorelines under different management regimes from 2016 to 2025, and proposed a differentiated riparian zonal management model based on “nature conservation-engineering protection”. The results showed that:(1)bank erosion on the left bank of the Datong River was significantly higher than that on the right bank: the average migration rate of the left bank was -0.87 m/a, while that of the right bank was -0.21 m/a.(2)shoreline migration intensity exhibited significant spatial differentiation(p<0.01). The maximum hydraulic energy intensity in the Menyuan section was three times that in the Qilian section, but its average geomorphic activity(1.29 m/a)was 36% lower than that in the Qilian section(2.03 m/a), showing a characteristic of “high driving force but low response”.(3)Human intervention reduced the average geomorphic activity by 0.74 m/a, verifying the effectiveness of the differentiated zonal management model. The findings clarify that retaining geomorphic freedom in nature conservation functional zones to conserve water resources, and strengthening boundary constraints in areas with intensive human activities to ensure safety, provide support for refined management of alpine watersheds.

To systematically assess the water pollution characteristics and ecological risks of the Dawen River, the largest tributary in the lower reaches of the Yellow River basin, this study established 10 representative sampling sites at key cross-sections along the mainstem. These sites span critical zones including the headwaters, midstream urban zones, and downstream confluence with the Yellow River, which were strategically selected based on industrial distribution patterns and hydrological dynamics within the watershed. Liquid chromatography-tandem mass spectrometry(LC-MS/MS)was employed to conduct multi-matrix pollution characterization targeting the following priority contaminants: four antibiotic classes(Sulfonamides/SAs, Tetracyclines/TCs, Quinolones/QNs, and Macrolides/MLs), as well as Nitrophenols(NPs)and Polyfluoroalkyl Substances(PFAS). The combined ecological risk of various pollutants were quantitatively evaluated using the risk quotient(RQ)model, by integrating data on environmental fate and ecotoxicological properties of the contaminants. The study revealed significant differences in the level of ecological risks among four categories of antibiotics in the mainstem of the Dawen River: sulfonamides(SAs)and macrolides(MLs)exhibited elevated detection concentrations with higher RQs. In contrast, tetracyclines(TCs)and quinolones(QNs)showed substantially lower contamination levels with RQs consistently below 0.1, indicating negligible ecological risks below acceptable thresholds. Per- and polyfluoroalkyl substances(PFAS)exhibited low overall concentrations across sampling sites, and detected PFAS levels reached low-risk thresholds, demanding heightened vigilance toward their bioaccumulation potential. In contrast, nitrophenols(NPs)demonstrated consistently negligible concentrations with risk quotients(RQs)of less than 0.01, indicating no significant ecological risks under the current contamination condition.

To screen low-Cd-accumulating wheat cultivars suitable for severely Cd-polluted farmlands, 11 dominant wheat cultivars widely planted in Shandong Province were selected as test materials for this study. A field experiment was carried out in a typical severely Cd-contaminated area. Cd concentrations in wheat roots, straws and grains were measured; bioconcentration factors(BCF)and translocation factors(TF)were calculated, and a comprehensive evaluation was conducted by integrating grain yield. The results showed that grain Cd concentrations of all cultivars ranged from 0.364 to 0.904 mg·kg^-1, all exceeding the maximum permissible limit of 0.1 mg·kg^-1 specified in Chinese national standard GB 2762-2022. Grain Cd BCF differed significantly(p<0.05)among cultivars, ranging from 0.105 to 0.252. Systematic cluster analysis, with grain Cd BCF and yield each weighted at 50%, divided the 11 cultivars into three categories. Group I was high-quality cultivars including Denghai 206(grain BCF=0.105, theoretical yield=495.0 kg·667 m^-2), Jimai 22(grain BCF=0.121, theoretical yield=492.5 kg·667 m^-2)and Shannong 47(grain BCF=0.157, theoretical yield=519.2 kg·667 m^-2), which combined low Cd accumulation and high yield. Group II contained medium cultivars: Yannong 1212, Jimai 44, Jimai 55, Yannong 215, Shannong 48 and Yannong 999. Group III included inferior cultivars Shannong 28 and Shannong 38, and Shannong 38 exhibited the maximum grain Cd BCF(0.252). Thus, Group I cultivars should be preferentially planted in Cd-polluted farmlands. This study demonstrated that Denghai 206, Jimai 22 and Shannong 47 are priority-promotable high-yield and low-Cd-accumulating cultivars for the research region, which can reduce grain Cd concentrations by 59.7%, 54.2% and 34.8%, respectively. Nevertheless, single application of low-Cd cultivars cannot realize safe wheat production under severe Cd pollution. An integrated safe utilization technology system combining low-Cd cultivars with agronomic regulation and soil remediation is necessary to effectively reduce the risk of Cd overstandard in wheat grains. This study provides basic data and scientific basis for safe utilization of contaminated farmland and optimization of cropping structure.

In response to the critical technological demands for judicial appraisal of eco-environmental damage within the framework of building an ecological civilizations legal system, this study focused on methodological innovation and practical application in the assessment of environmental damage caused by illegal mining cases in the Yellow River Basin. Through empirical analysis of the appraisal process in typical cases, this study proposed an “all-elements and multi-scale” assessment framework and constructed a comprehensive evaluation system covering the dimensions of mineral resources, soil environment, water environment, biota, ecosystem service functions, and potential secondary disaster hazards. It discussed the key challenges and corresponding solutions in the appraisal process. This research provides scientific support for the reform of the eco-environmental damage compensation system in the Yellow River Basin and holds significant practical value for improving the technical standard system of environmental judicial appraisal.

To investigate the changes in soil nutrients and their effects on bacterial communities in alpine grasslands along an elevational gradient, this study examined alpine meadows at elevations ranging from 3 900 m to 4 500 m in the Yushu Tibetan Autonomous Prefecture, Qinghai Province. High-throughput sequencing was employed to analyze the community structure and diversity characteristics of soil bacteria, and redundancy analysis(RDA)was used to explore their relationships with soil nutrients. The results showed that soil bacterial communities across different elevations were primarily composed of 13 dominant phyla, with Pseudomonadota, Actinomycetota, and Acidobacteriota being the main dominant groups. The Ace, Chao1, and Shannon indices of soil bacteria exhibited a unimodal pattern along the elevational gradient, reaching relatively high levels at 4 200 m, whereas the Simpson index showed a U-shaped trend; however, the overall differences were not statistically significant. Community structure analysis revealed that the composition of the soil bacterial community at 4 500 m differed considerably from that at other elevations and harbored a greater number of unique species. Correlation and redundancy analyses indicated that soil nutrient factors significantly influenced bacterial community structure, with total nitrogen and elevation demonstrating strong explanatory power. Functional prediction results suggested that the core metabolic functional composition of the soil bacterial communities remained relatively stable across different elevations. Overall, the changes in soil bacterial communities in alpine grasslands are primarily driven by soil nutrient conditions, exhibiting a response pattern characterized by variable structure yet relatively stable function.

Accurately elucidating the distribution patterns of suitable habitats and assessing habitat suitability are critical for the conservation of endangered wildlife. In this study, we integrated data from field surveys, literature reviews, and public databases to compile 213 distribution points for the great bustard(Otis tarda dybowskii) in Henan province, which were subsequently filtered using ENMTools to retain 82 spatially independent occurrence points and 10 environmental variables for modeling. The results indicated that the highly suitable area was 0.21×10⁴ km², accounting for 1.29% of the total study area, while the moderately suitable area was 0.19×10⁴ km², making up 1.18%. The primary environmental factors influencing the model predictions were temperature seasonality, annual mean temperature range, distance to water sources, and the coefficient of variation in precipitation. This study clarifies the spatial distribution characteristics of suitable habitats for overwintering Otis tarda dybowskii in this area, providing a scientific basis and practical reference for the graded conservation of the species wintering habitats, ecological restoration in the Yellow River Basin, and the establishment of a conservation paradigm for flagship species.

Driven by global warming, the runoff process in the source region of the Yellow River(YRSR)exhibited complex spatiotemporal responses. To reveal the future trends in climate and runoff evolution of this region, and to further assess its riverine ecological flow risks, this study employed the SWAT hydrological model, driven by a CMIP6 multi-model ensemble, to simulate future runoff changes from 2025 to 2100 under three shared socioeconomic pathway(SSP)scenarios; on this basis, the ecological flow thresholds were determined by integrating the monthly secondary minimum(maximum)method, the modified Tennant method, and the indicators of hydrologic alteration-range of variability approach(IHA-RVA)method, to evaluate the degree to which future runoff met ecological water demands. The results indicated the following:(1)The SWAT model demonstrated strong applicability in the YRSR(with Nash-Sutcliffe Efficiency>0.8 during both calibration and validation periods), and the bias-corrected multi-model ensemble data reliably drove runoff simulations.(2)The future climate exhibited a “warmer and wetter” trend characterized by increased precipitation fluctuations and continuous temperature rise, with the maximum increases in maximum and minimum temperatures reaching 6.1 ℃ and 7.0 ℃, respectively.(3)Future runoff increased significantly(by 32.17% to 58.26%)with a more concentrated intra-annual distribution, and the largest interannual fluctuation occurred under the SSP2-4.5 scenario.(4)The reliability of meeting future ecological water demands showed significant seasonal differences. The ecological base flow in winter and spring could generally be secured; however, the proportion of runoff exceeding the maximum ecological water demand in summer and autumn increased substantially(exceeding 90% in some months under the SSP5-8.5 scenario), indicating an intensified risk of high-flow ecological stress during the flood season. These findings provide a scientific basis for the rational allocation of regional water resources and aquatic ecological protection.

To elucidate the physical mechanisms by which permafrost degradation affected water conservation functions in alpine regions, and to address the cognitive limitations arising from the prevalent simplification of permafrost as a static categorical variable and the confinement of geographic detector analyses to the global scale, this study took the Three-River Headwaters Region as the study area. By coupling the InVEST model with the geographic detector, and introducing active layer thickness as a dynamic physical attribute of permafrost, we quantitatively assessed the spatiotemporal patterns of water conservation amount during 1996-2020, and conducted a comparative analysis of driving mechanisms between the permafrost zone and the seasonally frozen ground zone through a subregional approach. The results indicated that the total water conservation amount in the study area exhibited a weak and non-significant increasing trend, with a growth rate of 0.226×10⁸ m³/a, and a spatial pattern of “high in the southeast and low in the northwest”, while a decreasing trend was observed in the northwestern permafrost zone. Precipitation emerged as the dominant factor governing the spatial differentiation of water conservation, with q values ranging from 0.699 to 0.772; the independent explanatory power of permafrost type was the weakest(q values: 0.048-0.107)but showed an increasing trend, and its interactive effect with precipitation was significantly enhanced(q values: 0.728-0.788). Notable differences in driving mechanisms were found between subregions: in the permafrost zone, precipitation played an absolutely dominant role(q values: 0.749-0.818), and the interaction between active layer thickness and precipitation exhibited a strongly synergistic enhancement(q values: 0.767-0.829), confirming the amplifying effect of the impermeable permafrost table on precipitation-runoff responses; in the seasonally frozen ground zone, a “precipitation-freeze-thaw dual-core” driving mode was identified, where active layer thickness rose to become an independent core driving factor(q values: 0.211-0.322), yet its interaction with precipitation was relatively the weakest among all precipitation combinations(q values: 0.714-0.756), indicating a significant decoupling between precipitation and freeze-thaw processes in terms of spatial driving effects. Meanwhile, during 2006-2010 and 2011-2015, the observed runoff increases were markedly higher than the increases in water conservation, with ratios of 1.75 and 1.16, respectively, suggesting signs of “hydrological decoupling”. This paper shows that permafrost influenced water conservation functions through two distinct pathways: “background aquiclude modulation” and “active layer driving”; progressive permafrost degradation might weaken the water storage and regulation capacity of alpine ecosystems, posing potential risks to the stability of water supply from the “Asian Water Tower”.

Jinan, a central city in the Yellow River Basin and a renowned “City of Springs,” is characterized by a distinctive spring-centered water ecosystem. A comprehensive assessment of the spatiotemporal changes in its surface water connectivity, identification of the key influencing factors, and development of scientifically sound and feasible optimization pathways are of considerable significance for enhancing water circulation, improving water quality, and promoting the restoration of spring-centered ecosystems. Based on the geomorphological pattern characterized by mountains in the south and water systems in the north, this study established a three-dimensional evaluation framework integrating structural, process, and functional connectivity. Multi-source data, including remote sensing interpretation and geospatial analysis, were used to systematically reveal the spatiotemporal evolution and driving mechanisms of surface water connectivity in Jinan from 1985 to 2022, and multi-scale optimization strategies were subsequently proposed. In practical terms, the concept of sponge city development was incorporated into the design of restoration pathways, and a coordinated optimization strategy for spring conservation and water supply in northern karst cities was proposed, providing a technical framework for addressing the imbalance between water supply and demand. The results showed that surface water connectivity in Jinan exhibited a fluctuating upward trend over the past four decades. Spatially, connectivity was generally higher in the southern mountainous areas than in the northern urban areas. Influenced by agricultural water conservancy projects, the transitional zone generally developed a positive cycle of structural optimization and functional improvement. Ecological restoration projects alleviated the degradation of urban water connectivity, although the fragmentation of the water network caused by urban expansion remained evident. Quantitative attribution analysis using geodetector indicated that GDP and annual precipitation were the dominant factors explaining the spatial differentiation of connectivity, and that human activities had become the principal driver of connectivity evolution. Based on these findings, differentiated restoration strategies were proposed, including mountain closure for afforestation and the construction of stepped infiltration basins to restore groundwater in the southern karst region; river dredging and ecological-weir construction to reconstruct water corridors in the northern plain; and the promotion of sponge-city projects and smart management in built-up areas to enhance blue-green coordination.