HQ refers to the capacity and functionality of the natural ecological environment to provide sustainable living conditions for individual organisms and communities1,2. It serves as a direct indicator of the ecological status of biological habitats and is one of the key indicators of biodiversity3. With ongoing social development, human activities have increasingly disrupted the regional ecological environment, resulting in the degradation of HQ4. Therefore, studying the evolutionary patterns of HQ within a region is crucial for ecological conservation and achieving harmonious coexistence between humans and nature5. Current HQ assessment methods can be classified into two categories based on the research scale. The first category involves small-scale studies focusing on single-species HQ, typically combining field surveys of specific habitats to construct an evaluation system and assess HQ6. The second category comprises larger-scale regional studies7, which primarily rely on remote sensing data, including land use and landscape indices, with modeling approaches to calculate and assess HQ8. Commonly used HQ assessment models include MaxEnt model9, HSI model10, SoLVES model11, and InVEST model12. Among these, the InVEST model has a greater capability to comprehensively consider the benefits and impacts of human activities. It enables the quantitative assessment of regional HQ based on threat factors and sensitivity under land use scenarios13. Due to its advantages, including a mature theoretical foundation, independence from regional species, ease of data acquisition, compatibility with land use type prediction studies, strong spatial visualisation capability, high precision of evaluation results and simple operation and data processing, it is widely used in HQ assessment14. In recent years, many scholars have conducted in-depth studies on regional HQ using InVEST model. Berta Aneseyee et al.15 applied the InVEST model to analyze habitat degradation in the Winike watershed over the past three decades, revealing that habitat degradation has intensified, with increased influence from threat impact factors and habitat sensitivity. Nematollahi et al.16 used InVEST to simulate HQ and degradation in Chaharmahal and Bakhtiari provinces of southwestern Iran, identifying the eastern and northeastern regions of the study area as the most suitable habitats for wild sheep, which were significantly affected by the road network. Broquet et al.17 explored the interaction of land use changes and HQ estimated in the Upper Paraguay River Basin in Brazil, using the InVEST HQ model to estimate spatiotemporal landscape trends and evaluate the potential impact of land use policies for the future. Wu et al.18 employed the InVEST model and entropy method to assess the spatiotemporal dynamics of HQ and urbanization in Yangtze River cities.And they found that HQ consistently declined from 1990 to 2019. Tian et al.19 used the InVEST model to employe to assess habitat quality based on both historical and projected land use data. HQ assessment was validated using species richness of vascular plants estimated from field-sampled specimens. Huo et al.20 taked Xiangyu in the Feng River Basin as the study area, the land use change in 1985 and 2022 was analyzed, and the habitat quality and degradation degree were evaluated by using the InVEST model. Collectively, these studies demonstrate the significant application value of the InVEST model in regional HQ assessment, effectively revealing the spatial and temporal evolution of HQ and its driving factors.

As a key ecological functional area in China and the core water source of the South-to-North Water Diversion Project, the Han River Basin (HRB) has a pristine natural environment and holds ecological strategic importance. However, in recent years, the accelerated pace of urbanization, particularly driven by the implementation of the South-to-North Water Diversion Project, has substantially intensified human interference in the basin’s ecosystem. This has led to profound changes in land use patterns and increased landscape fragmentation. These changes have triggered a series of ecological and environmental issues, such as vegetation degradation and soil erosion21, which seriously threaten the stability and service functions of the basin ecosystem. Since the ecological protection of the HRB is not only crucial to the strategic security of national water resources, but also serves as a fundamental basis for maintaining regional ecological balance and promoting sustainable development, it is of great practical significance to conduct studies on the spatial and temporal evolution of HQ at the basin scale. Although some studies on the HQ of the HRB basin have contributed to ecological conservation efforts in the region, several shortcomings remain: (1) Existing research on the HRB is largely confined to local areas and lacks systematic studies on spatiotemporal patterns of changes in HQ across the entire basin over an extended period. This results in an incomplete understanding of the overall changes in HQ within the basin. (2)There is a lack of systematic analysis of the driving factors behind changes in HQ across the entire basin, including the quantitative impact of human activities, topography and climate change. This makes it difficult to provide precise, effective scientific evidence to enhance ecological and environmental protection in the basin.

Based on the above issues, this study aims to use the InVEST model and relevant data to investigate the long-term spatial and temporal changes in HQ across the entire HRB. The objective is to reveal the spatio-temporal patterns of these changes and, quantitatively explore the impact of various driving factors using the Geodetector. The specific work content includes: (1) studying the spatio-temporal distribution characteristics and evolution trends of HQ in the HRB from 2000 to 2023; (2) using the Geodetector to quantitatively explore the influence of driving factors relating to both human activities and natural processes; (3) providing a scientific basis and policy recommendations to enhance ecological and environmental protection in the basin. This research will provide a comprehensive understanding of the ecological dynamics of the HRB and promote harmonious human-nature coexistence in the region. Furthermore, the findings of this study will serve as valuable references for high-level ecological and environmental protection in similar river basins.

Materials and methods

Study area

As the largest tributary of the Yangtze River, the HRB (30° 10ʹ–34° 20ʹ N, 106° 15′–114° 20′ E) has a total length of 1577 km, with a watershed area of 159,000 km2. It flows through the provinces of Shaanxi, Henan, and Hubei from northwest to southeast, finally joining the Yangtze River in Wuhan. The river originates from the main peak of Bozhong Mountain in the tectonic uplift zone at the southern foot of the Qinling Mountains. Based on its geomorphological and hydrological characteristics, the basin is divided into three major sections. The upstream section from the source to Danjiangkou, which is the core water source area of the South-to-North Water Diversion Project, is characterized by tectonically eroded mountainous terrain. The middle section from Danjiangkou to Zhongxiang crosses the transitional zone of the Nanxiang Basin. The downstream section from Zhongxiang to estuary flows into the alluvial plains, extending deep into the hinterland of the Jianghan Plain22. The terrain of the HRB shows significant step-like changes. In the northern part of the upper reaches, the topography is controlled by the tectonic structure of the southern slopes of the Qinling Mountains, forming high-elevation geomorphic units at around 2000 m. The southern part is affected by the thrust and nappe tectonics of the Daba and Micang Mountains, resulting in the development of deeply incised canyons. In the central region, the Hanshui Valley has formed a series of bead-like fault basins due to tectonic activity. The middle reaches are characterized by the distribution of rolling granite hills, while the lower reaches, except for the tectonic remnants of Dahong Mountain, generally exhibit low-lying terrain23. The basin experiences a subtropical monsoon climate, characterized by mild and humid conditions, with an average annual temperature of 12–16 °C and annual precipitation ranging from 700 to 1400 mm. There are significant spatial variations in sunlight, temperature, and water resources, resulting in distinct vertical zonation (see Fig. 1 for an overview of the study area). Notably, the HRB is densely populated, placing significant pressure on the environmental carrying capacity. It also faces ecological risks such as soil erosion, non-point source pollution, and biodiversity disturbances24,25. Therefore, it is crucial to conducting an in-depth study on the changes in the ecological quality and the driving mechanisms.

Fig. 1
figure 1

Overview of the study area (created by ArcMap, version 10.7, http://www.esri.com/).

Full size image

Study data

This study selected land cover classification data from six time points: 2000, 2005, 2010, 2015, 2020 and 2023. The driver data mainly includes not only information related to human activities, such as land use intensity, population density and night light index; but also natural factors, such as elevation, mean annual temperature and precipitation, and NDVI. This data is divided into six periods: 2000, 2005, 2010, 2015, 2020 and 2023. The sources of this data are detailed in Table 1. The mean annual GDP data were obtained from studies by Kummu et al.26. The 2023 data has not yet been released, so 2022 data was used instead. The annual average NDVI (normalized difference vegetation index) data was mainly extracted using the GEE (Google Earth Engine) platform. All data were downloaded and pre-processed with coordinate system transformation, geometric correction and resampling to 1 km × 1 km before use in analyses.

Table 1 Data and sources.
Full size table

Study methods

HQ assessment methods

The HQ assessment module of the InVEST model is based on the regional land use status, the relative sensitivity of each habitat type to threat sources and spatial distance. It calculates the threat source coercion distance and weight, the suitability of each landscape and the sensitivity of different threat sources, thus obtaining the HQ index of the study area27. The HQ module of the InVEST model was used to evaluate and analyze the HQ of the HRB from 2000 to 2023. The specific calculation formula is as follows:

$${Q}_{xj}={H}_{j}\left[1-\left(\frac{{D}_{xj}^{Z}}{{D}_{xj}^{z}+{k}_{xj}^{z}}\right)\right]$$
(1)
$${D}_{xj}=\sum_{R=1}^{R}\sum_{y=1}^{{Y}_{r}}\left(\frac{{W}_{r}}{{\sum }_{r=1}^{r}{W}_{r}}\right){r}_{y}{i}_{rxy}{\beta }_{x}{S}_{jr}$$
(2)

In formula (1), Qxj represents the HQ index of land use type j in raster x, with a value range of (0–1); Hj represents the habitat suitability of land use type j; Dxj represents the degree of habitat degradation of land use type j in raster x; k is the half-packing constant; z is the default parameter in the model, which is set to 0.05. In formula (2), Dxj represents the habitat degradation index of raster X in habitat type j. R represents the number of coercive factors; r represents a particular threat factor; Yr represents the number of rasters occupied by coercive factor r in the layer; y represents the raster where coercive factor r is located; Wr represents the weight magnitude of coercive factor r, with a value range of (0–1); ry represents the intensity of coercive factors on each raster; irxy represents the threat of coercive factors in raster y on habitat raster X; βx represents the accessibility level of habitat type patch X; Sjr represents the relative sensitivity of land use type j to different stressors r, with a value range of (0–1), and a value closer to 1 indicates greater sensitivity to stressors.

In general, the greater the distance between the patch and the stress factor, the lower the threat level, and the two are inversely correlated. Thus, the threat irxy, which expresses the threat of the stress factor of raster y to the habitat raster X, can be represented by both linear and exponential functions.

Linear decay function:

$${i}_{rxy}=1-\left(\frac{{d}_{xy}}{{d}_{rmax}}\right)$$
(3)

Exponential decay function:

$${i}_{rxy}=exp\left[-\left(\frac{2.99}{{d}_{rmax}}\right){d}_{xy}\right]$$
(4)

In formulas (3) and (4), drmax represents the maximum threat distance of habitat stress factor r; dxy represents the linear distance between rasters x and y.

According to the InVEST user manual and other similar research results28,29,30,31,32, combined with the actual conditions of the HRB, cropland, rural settlements, urban land and other construction land were identified as habitat threat sources. Their maximum impact distances, weights and attenuation types were determined, as shown in Table 2. Based on this, different habitat suitability values and sensitivity of each habitat to each threat source were set, as shown in Table 3. HQ results for the years 2000, 2005, 2010, 2015, 2020, and 2023 were obtained after inputting the model calculations.

Table 2 Threat sources and related parameters of HQ.
Full size table
Table 3 Habitat suitability and its sensitivity to threat sources.
Full size table

Based on related studies24,25, HQ was classified into five ecological levels using the equidistant grading method: low [0–0.2), lower [0.2–0.4), medium [0.4–0.6), higher [0.6–0.8), and high [0.8–1]).

To better study the trend of HQ in the HRB from 2000 to 2023, the trend of HQ in the HRB over time was derived using the least squares method by performing pixel-by-pixel meta-linear regression33. The formula is:

$$S=\frac{\left(n\times \sum_{i=1}^{n}i\times {Q}_{i}\right)-\left(\sum_{i=1}^{n}i\times \sum_{i=1}^{n}{Q}_{i}\right)}{n\times \sum_{i=1}^{n}{i}^{2}-{(\sum_{i=1}^{n}i)}^{2}}$$
(5)

where S is the slope of the fitted trend and represents the regression coefficient of HQ at the pixel location over the time series, the interannual rate of change; n is the number of samples (with 6 years of data in this study, sample size n = 6), and Qi is the HQ in the ith year. If S > 0, it indicates an increasing trend in HQ over the study period; if S < 0, it indicates a decreasing trend in HQ over the study period. When S = 0, it indicates no change in HQ at the pixel level.

To illustrate the trend of change more clearly, the trend results were classified into five categories using the natural breaks34. The natural breaks algorithm is a widely used classification method. Zongfan et al.35 used natural breaks to divide the desertification monitoring index into non-desertification, mild desertification, moderate desertification, and severe desertification. Chen et al.36 used natural breaks algorithm to categorize the freeze–thaw deterioration areas into five susceptibility levels. Using the natural breaks algorithm, the classified results represent that, S ≤− 0.0694 indicates a significant degradation trend in HQ, − 0.0694 < S ≤− 0.0211 indicates slight degradation, − 0.0211 < S ≤ 0.0146 indicates basic stability, 0.0146 < S ≤ 0.0576 indicates slight improvement, and S > 0.0576 indicates significant improvement.

Geodetector

Geodetector is a new statistical method proposed by Wang and Xu37, mainly used to detect spatial dissimilarity and reveal the influencing factors behind it. The core idea is that when the independent variable significantly influences the dependent variable, there is a certain similarity in their spatial distributions. Based on this theory, the method detects the primary driving factors of spatial dissimilarity38. The calculation formula is:

$$q=1-\sum_{n=1}^{L}{M}_{n}{\sigma }_{n}^{2}/M{\sigma }^{2}$$
(6)

In the formula, q represents the explanatory power of the independent variable for the spatial differentiation of the dependent variable; L is the number of classifications or divisions of the independent variable; Mn and M are the sample sizes within the nth division and the entire region; \({\sigma }_{n}^{2}\) and \({\sigma }^{2}\) are the variances of the nth division and the entire region. The value range of q is [0, 1]. The larger the value of q, the stronger the explanatory power of the independent variable for the spatial differentiation of the dependent variable, and vice versa. Interaction detection is used to identify the interrelationships among different independent variables and evaluate their combined effects to observe whether the joint action of any pair of independent variables increases or decreases the explanatory power for the spatial differentiation of the dependent variable (as shown in Table 4).

Table 4 Result types of two-factor interactions.
Full size table

In this study, we used the factor detection and interaction detection of the Geodetector to analyze a total of five factors selected for human activities (land use intensity,GDP,population density and night light index) and natural factors (mean annual temperature, mean annual precipitation,NDVI and elevation). Factor detection was used to determine the degree to which a single driving factor explains the spatial differentiation of HQ, and interaction detection was employed to determine whether there is any interaction between two driving factors. Finally, the driving factors behind HQ changes in the HRB over the past two decades were identified.

Results

Spatial and temporal distribution characteristics of HQ

The time series analysis from 2000 to 2023 indicates that the HQ index of the HRB initially declined and then gradually rebounded. The mean value fluctuated from 0.461 (in 2000) to 0.453 (in 2023), with an overall decrease of 0.008. Specifically, the period from 2000 to 2015 experienced continuous degradation (a decline of 0.015), followed by a slow recovery after 2015 (an increase of 0.007). By using the equidistant grading method, structural analysis revealed that the overall HQ in the watershed remained at a medium level, but the class structure exhibited signs of degradation. The proportion of low-quality zones continued to increase from 33.66% in 2000 to 35.79% in 2023, an increase of 2.12%. The proportion of lower-quality zones declined from 15.36% in 2000 to 14.6% in 2023, a decrease of 0.76%, while the medium-quality zones decreased from 11.67% in 2000 to 10.63% in 2023, a decrease of 1.04%. Notably, the higher-quality zone increased slightly by 0.22%, from 16.9% in 2000 to 17.12% in 2023.The high-quality zone gradually decreased from 22.41% in 2000 to 21.86% in 2020, and slightly rebounded to 21.87% in 2023(Table 5). This trend is closely associated with the high-intensity land development and ecological conservation activities in the watershed, highlighting the pressing issue that the pace of HQ restoration lags behind the intensity of human intervention.

Table 5 Statistics of HQ grades and areas in HRB from 2000 to 2023.
Full size table

In the spatial dimension, the HQ of the HRB shows a typical three-dimensional topographic control pattern (Fig. 2), with the distribution characteristic of ‘central axis uplift and depressions on both sides’ persisting from 2000 to 2023. Specifically, the high-value areas remain consistently located in the upstream Qinba Mountain ecological barrier zone, the midstream Jingshan-Dahongshan tectonic uplift area, and the downstream Dabie Mountain remnants. These regions are characterized by significant topographic heterogeneity, high forest cover, and low-intensity human activities, forming continuous ecological corridors. The medium-value zone is distributed in a concentric ring pattern around the high-value zone, mainly corresponding to the low mountainous and hilly regions. The low-value zone shows an axial expansion trend, forming an ecologically fragile ‘river valley-basin-plain’ zone along the mainstream of the Han River. This includes the upstream Han River valley, the midstream Nanxiang Basin, and the downstream Jianghan Plain, where the expansion of arable land and the strong ecological impediment effect of urbanization have resulted in a low HQ zone.

Fig. 2
figure 2

Spatial distribution of HQ levels in HRB from 2000 to 2023. (created by ArcMap, version 10.7, http://www.esri.com/).

Full size image

Spatial and temporal evolution characteristics of HQ

In order to better explore the spatial and temporal evolution characteristics of HQ in the HRB, the spatial and temporal trends of HQ in the basin were determined using pixel-by-pixel linear regression with the least squares method. The spatial distribution and area statistics of these trends are shown in Fig. 3, from which it can be observed that the area with a significant degradation trend accounting for 3.84%, is the total area, primarily located in the middle and downstream areas. These areas were mainly concentrated around cities and county towns, where urban expansion and other human activities have had a negative impact on HQ. The area with a slight degradation trend accounts for 12.12%, which forms a buffer zone around the significantly degraded area, with the intensity of its degradation gradually weakening as the distance from the built-up area increases. The area with a basically stable trend accounts for 69.31% of the total area, and is primarily concentrated in the upstream Hanshui Valley, the middle reaches of the hilly areas, the Nanxiang Basin, and most parts of the Jianghan Plain in the lower reaches. These regions are mainly characterized by low HQ, consisting of urban core areas or cropland areas. Due to the early and persistent human activities, the changes in HQ are not significant, reflecting a phenomenon of blunted ecological response in fragile zones subjected to prolonged human disturbance. The area with a slight improvement trend accounts for 11.15% and is unevenly distributed in the eastern part of the Qinba Mountain region, particularly in areas where croplands have been converted back to forests, and around the ecological barrier zone of the Danjiangkou Reservoir. The area with a significant improvement trend accounts for 3.58%, located in the upper and middle reaches of the basin. It can be seen that the Danjiangkou Reservoir has shown a trend of significant improvement.And it is the largest and most concentrated area of significant improvement in the HRB, which is closely associated with water source protection efforts in the Danjiangkou Reservoir area. Across the entire basin, degraded area accounts for 15.96% of the total area, which is 1.23% more than the 14.73% of improved areas. The spatial distribution shows a distinct vertical differentiation pattern characterized by ‘concentrated degradation at low elevations, and local improvement at middle and high elevations’.

Fig. 3
figure 3

Change trend of HQ in HRB from 2000 to 2023 (created by ArcMap, version 10.7, http://www.esri.com/).

Full size image

Driving factors of the spatial and temporal evolution of HQ

Analysis of driving factors

The analysis of driving factors using the Geodetector model (Fig. 4) showed that the evolution of HQ from 2000 to 2023 was characterized by a composite natural-human activity driving mechanism. The model test showed that the p-values of all factors passed the 0.01 significance level, confirming the validity of the driving mechanism. Among these factors, the impact of land use intensity on HQ has always been dominant (q-value: 0.514–0.545), with an overall downward trend (although there was a slight rebound in 2023). Elevation is the second most important factor (q-value: 0.422–0.432), with fluctuating changes overall, and it is related to the fact that the higher altitude leads to less human interference. Additionally, the driving influence of temperature fluctuates (q-value: 0.346–0.356), with an overall downward trend. The driving influence of GDP increases year on year (q-value: 0.335–0.378). The q-value of population density undergoes a U-shaped change, decreasing from 0.339 in 2000 to 0.115 in 2010, increasing to 0.281 in 2023.This trend might be related to the large-scale population outflow during the second heightening of the Danjiangkou Dam (2005–2013). The influences of NDVI, precipitation and night-time light index are relatively weaker.The driving influence of NDVI showed fluctuating changes but an overall increasing trend (q-value of 0.052–0.238); the driving influence of precipitation showed a large fluctuation in 2010, related to the extreme precipitation event in the HRB that year (q-value of 0.076–0.142); and the driving efficacy of the night lighting index showed a continuous increase over the 20-year period (0.01–0.117), which is related to the continuous increasing in the area of urbanization.

Fig. 4
figure 4

Investigation results of factors affecting HQ in HRB from 2000 to 2023.

Full size image

Analysis of factors interaction

From 2000 to 2023, the effect values (q-values) of all two-factor interactions on HQ significantly exceeded those of the single-factor independent effects (Fig. 5), indicating the existence of a systematic synergistic enhancement mechanism among the driving factors in the study area. Their modes of action can be categorized into two typical forms: two-factor enhancement and nonlinear enhancement. From 2000 to 2023, the interaction of land use intensity ∩ elevation continued to dominate the spatial variance of HQ (0.61–0.628), but the overall trend was a gradual decrease. This was followed by the interaction of land use intensity ∩ temperature (0.599–0.615), which also showed a gradual decrease and was overtaken by the interaction between land use intensity ∩ GDP in 2023. The interaction of land use intensity ∩ GDP gradually decreased from 0.612 in 2000 to 0.597 in 2015, before gradually increasing to 0.602 in 2023, which made it the second most important interaction factors after land use intensity ∩ elevation.The interaction of land use intensity ∩ population density (0.543–0.593), first decreased and then increased.Meanwhile, the interaction between land use intensity and the other factors were ranked in order. Land use intensity, the main manifestation of human activity, has a profound influence on changes in HQ in the watershed. Elevation, the fundamental topographic feature, establishes the spatial framework for HQ through gradient differences. Based on this pattern, temperature regulates vegetation growth in the watershed through inter-annual fluctuations, while GDP exerts spatial coercion through density gradients.The influence of population is primarily manifested in changes to land use patterns, which are driven by population migration. Additionally, it is also reflected in the impact on HQ resulting from the exploitation and use of natural resources.The composite superposition of these elements results in significant spatial variation in HQ, with higher HQ in the middle and lower on both sides. These factors combine and overlap to produce a significant spatial enhancement of HQ differentiation. This finding quantitatively reveals the synergistic mechanism of multi-scale drivers and provides a basis for predicting the dynamics of habitat evolution under the dual pressures of urbanization and climate change.

Fig. 5
figure 5

Research results of interactive factors of HQ in HRB from 2000 to 2023.

Full size image

Discussion

Comparison of HQ with other river basins

This study reveals that the HQ in the HRB has remained moderate from 2000 to 2023, with an initial decline followed by recovery. This finding is consistent with previous studies from Xu et al.39, in the middle and lower reaches of the Yangtze River and Xia Ying et al.40 in the Four Lakes region of the Hanjiang Plain. These findings provide spatial corroboration and validate the evolutionary trend of HQ in this basin. When compared with research findings on HQ in other basins, such as Bian et al.41 studied the Yangtze River basin, the results indicated that the overall HQ in the Yangtze River basin was relatively good (0.736–0.733) from 2000 to 2020, showing a declining trend with an overall decrease of 0.408%. The increasing human activity in the Yangtze River basin poses a significant threat to habitat restoration. Song et al.42 assessed the HQ of the Yellow River Basin from 2000 to 2020. The results showed that the overall level of HQ in the basin was relatively good (0.6849–0.7001), with a trend of initial decrease followed by increase in habitat degradation. Land use conversion had the most significant impact on HQ. HQ in the HRB was significantly lower than in the two larger basins. This is primarily because the Yangtze and Yellow Rivers span vast areas with large regions of excellent ecological quality in their upstream source areas.Compared with rivers within the same watershed, the study by Duan et al.43 on the evolution of HQ in the Jialing River basin from 2000 to 2020 showed a steady decline in HQ indices. Specifically, the indices for 2000, 2010 and 2020 were 0.5401, 0.5338 and 0.5084, respectively. This decline suggests a highly negative correlation between HQ and land use intensity. Similarly, Hong et al.44 assessed HQ in the Xiangjiang River basin and found that it decreased from 0.770 to 0.757 in 1990 to 2020, indicating a downward trend over the 30-year period. HQ in the HRB is lower than in the Jialingjiang and Xiangjiang basins. This is closely related to the fact that the Hanjiang River basin includes major grain-producing areas such as the Hanzhong Basin, the Nanyang Basin and the Jianghan Plain, where human activity is more intense and habitat destruction is more severe.

The influence of Land use and elevation on HQ in the HRB

Similar to Yangtze, Yellow and Jialing River basins, human-induced changes in land use patterns are the dominant factors influencing HQ in the HRB. Additionally, elevation factors also play a significant role. The urban areas of the 13 prefectures within the basin (including core cities such as Hanzhong, Nanyang and Xiangyang, each with populations exceeding 5 million) continue to expand. Meanwhile, the proportion of farmland remains high at 35%, which has a significant impact on regional HQ24. Conversely, the steep terrain of the Qinling-Daba Mountain Range in the upper reaches of the basin and the hilly areas in the middle reaches forms a natural barrier that significantly reduces human accessibility. This results in high vegetation coverage and low human disturbance, ultimately forming high-quality habitat core zones. In contrast, the spatial accessibility provided by the flat terrain of low-altitude alluvial regions such as the Hanzhong Basin, Nanyang Basin, and Jianghan Plain has facilitated urbanization, resulting in consistently low HQ from 2000 to 2020. These factors led to a sustained decline in HQ from 2000 to 2015. Meanwhile, the Danjiangkou Reservoir, a critically important source of drinking water within the HRB, is the main water source for the Central Route of the South-to-North Water Diversion Project, making it a key site for protection. Since the project began operating in December 2014, an ecological compensation mechanism has been implemented in the basin. This includes afforestation projects in the upstream Danjiangkou Reservoir area, as well as the establishment of river and lake ecological flow protection systems in the middle and lower reaches. Consequently, these measures have increased vegetation coverage and reduced soil erosion, effectively mitigating the negative impacts of the water diversion project, such as the prolonged dry season45. Ultimately, this 'pressure-response’ dynamic balance mechanism has ultimately facilitated the recovery of HQ since 2015. As a result, this finding provides crucial temporal evidence for optimizing the ecological safety landscape in areas affected by large-scale water diversion projects.

HQ protection and management in the HRB

HQ, as a key indicator reflecting the ability of regional ecological background to support human survival and development, shows a deteriorating trend in the HRB, a national key ecological functional area. The continuous year-by-year increase in the area of low HQ underscores the urgency of ecological protection and restoration. Based on the coupled ‘terrain-climate-human’ driving mechanism revealed by Geodetector, it is recommended to build a three-level synergistic management system: Firstly, a resilient urban development regulations mechanism should be established. Specifically, a composite system comprising blue-green corridors, small wetlands and three-dimensional greening should be established in built-up areas. This approach will be essential for the development of ecological garden cities. Moreover, timely ecological restoration measures should be carried out in areas where HQ has significantly deteriorated, particularly in regions with higher development intensity. Secondly, efforts should be intensified to strengthen the precise restoration of engineering disturbances. For mining areas, a stepwise restoration approach that integrates terrain remodeling, soil reconstruction and vegetation reconstruction should be operationalized. Additionally, 500-m-wide ecological isolation zones should be constructed on both sides of transportation corridors. Finally, the differentiated terrain gradient management should be implemented. In the Qinling-Daba Mountains Ecological Barrier Zone with an elevation of over 800 m, it is essential to strictly implement the ecological protection red line system, strengthen the natural forest protection project, and establish a protected area system with national parks as the core. For the hilly transition zone with an elevation of 200–800 m, it is advisable to implement comprehensive erosion gully management projects to reduce soil erosion and improve ecological functions. In the plain urban agglomeration areas where the elevation is below 200 m, measures such as tree planting and grass planting should be adopted to ensure a gradual increase in the proportion of ecological land. Through the integrated ‘spatial control-process regulation-engineering restoration’ strategy, the HQ index in the watershed can be improved year by year, ultimately establishing the region as a key ecological security hub for the Yangtze River Economic Belt.

Conclusions

This study applied the InVEST model to assess the spatial and temporal evolution characteristics of HQ in the HRB, using the Geodetector model to reveal the key drivers of HQ. The main conclusions are as follows:

  1. (1)

    The HQ in the HRB from 2000 to 2023 exhibited a trend of initial decline followed by recovery, and overall decreased 0.008. The HQ index remained at a medium level. The spatial distribution showed a spatial distribution pattern of ‘high in the middle, low on both sides’ in space.

  2. (2)

    Over the past 20 years, 69.31% of the HQ has remained relatively stable in HRB. Improvement areas accounted for 14.73%, while degradation areas accounted for 15.96%, 1.23% higher than the improvement areas. The spatial distribution shows a vertical differentiation pattern of ‘concentrated degradation at low elevations and localized improvement at middle and high elevations’.

  3. (3)

    Over the past 20 years, the evolution of HQ in the HRB has been primarily driven by land use intensity and elevation, with temperature and GDP also playing significant roles. This coupling effect of topography, climate, and human factors ultimately has given rise to a dual-mode driving pattern characterized by‘ human-induced stress at low elevations and natural regulation at high elevations’.