Introduction

Land is a vital source upon which human life and production depend. In recent decades, rapid industrialization and urbanization have led to excessive human intervention, significantly impacting habitat quality1. Land use change (LUC), a concentrated manifestation of human activities, is a key driver of habitat quality changes2,3,4,5. It influences the flow processes of materials, energy, and information between habitat patches, affecting regional habitat quality variations6. The exploitation of transitional land resources threatens regional biodiversity and degrades habitat quality7. Therefore, analyzing the impact of LUC on regional habitat holds significant implications for land use planning, biodiversity conservation, and sustainable development.

Habitat quality refers to the ability of an ecosystem to provide a sustainable environment for organisms’ survival and reproduction, reflecting regional biodiversity and ecological service levels. High quality habitats contribute to enhancing regional biodiversity and promote positive interactions between humans, other species, and their ecological environments8,9,10,11,12. It is influenced by natural factors, such as climate and soil types, and human activities, including urban expansion and grassland degradation. Field surveys and model simulations are two prevalent methods for studying habitat quality. Field surveys involve direct investigations of habitats for ecological indicators, which are used to construct assessment systems; however, they are limited by high costs and scalability issues13. In contrast, modelling approaches, such as the Habitat Suitability Index (HSI)14, SoIVES15, IDRISI16, and the InVEST model17, have been widely used. The InVEST model is particularly favored due to its ease of data acquisition and strong visualization and spatial solid analysis capabilities18. Previous studies have leveraged this model to start with land use changes at various scales, including watersheds19, islands20, urban areas21, water bodies and wetlands22,23, analyzing spatiotemporal variations in habitat quality along and their driving mechanisms18. they also utilize CA–Markov model to simulate historical periods of habitat quality changes while employing FLUS model to predict future shifts under various land-use scenarios24,25. However, there remains a relative scarcity of research focusing on the impact of land use changes on habitat quality within plateau lake ecosystems.

The lake basin in central Yunnan is by the close yet relatively independent distribution of five major plateau lakes, forming an essential ecological security barrier between Sichuan and Yunnan. This region possesses abundant water resources, making it a critical strategic reserve. The uniqueness of these lakes’ origins and their natural zonation contribute to rich biodiversity. Diverse land use types, including extensive forest and grassland coverage. These valuable natural resources impart significant economic and ecological value to the lake basin, providing a foundation for the sustainable development of central Yunnan. However, rapid economic growth in the area has triggered ecological issues such as water pollution, encroachment of impervious surfaces, and fragmentation of natural landscapes, which directly threaten habitat quality and ecosystem services26,27. Therefore, considering these five individual lake basins as an integrated entity, analyzing habitat quality of the lake basin in central Yunnan from a holistic perspective, becomes essential and meaningful.

Based on the aforementioned context, this study addresses a significant research gap concerning the impact of LUC on habitat quality in plateau lake ecosystems, particularly in the ecologically significant and rapidly developing central Yunnan region. Focusing on the lake basin of central Yunnan, which are of strategic ecological importance yet subjected to intensive land use changes, this research employes the InVEST model to analyze the evolution patterns of habitat quality over a 30-year period (1990–2020) under varying intensities of land use change. Additionally, it investigates the driving factors affecting habitat quality, thereby providing a theoretical foundation for ecological conservation and sustainable land use development in plateau lake basins.

Materials and methods

Study area

The five major plateau lake basins, including Dianchi Lake, Fuxian Lake, Xingyun Lake, Qilu Lake, and Yangzonghai Lake, are located in the central region of Yunnan Province, China (24.05–25.46°N, 102.48–103.04°E) (Fig. 1). This area encompasses approximately 4510 km2 and is characterized by a subtropical monsoon climate, with an average annual temperature of 15.35 °C and average yearly precipitation of 883.98 mm. The predominant vegetation comprises secondary forests and artificial woodlands, while the topography is characterized by mountains and basins with gentle undulations. Among these lakes, the Dianchi basin is the largest, covering about 2920 km2. Fuxian Lake, known as the source lake for the Pearl River system, maintains a water quality rated as Class I year-round and supports robust tourism development. Xingyun Lake is situated upstream of Fuxian Lake and is connected to it via a river, with its basin containing rich phosphorus mineral resources. Qilu Lake is a typical closed highland subsidence lake characterized by a densely populated basin with a high degree of urbanization. Lastly, Yangzonghai Lake belongs to the southern Panjiang watershed within the Pearl River system, yet it has a relatively small drainage area of only 192 km2. The five major lakes in Central Yunnan are typical plateau fault lakes, with a total water area of 630 km2. The extensive wetlands distributed across their basins play a vital role in maintaining regional water balance, regulating the basin’s climate, and conserving biodiversity. In particular, the Dianchi Lake Basin is characterized by its rich diversity of plant and bird species and serves as a critical habitat for migratory birds, such as the Black-headed Gull. Moreover, the basin hosts 303 species of aquatic plants, whose synergistic interactions with microorganisms contribute significantly to improving water quality.

Fig. 1
figure 1

Regional overview map of the study area.

Full size image

Data sources and processing

The land use data for the four time periods (1990, 2000, 2010, and 2020) in this study were derived from the China Land Cover Dataset (CLCD) provided by Wuhan University28 (https://zenodo.org/record/5816591). The land use types include nine primary categories, of which this study focuses on seven: Cropland, Forest, Shrub, Grassland, Water, Barren, and Impervious. The dataset spans from 1985 to 2023, with a spatial resolution of 30 m and an overall accuracy of 79.13%, offering advantages in both temporal-spatial resolution and data accuracy. The digital elevation model (DEM) was obtained from the Geospatial Data Cloud (http://www.gscloud.cn/) with a spatial resolution of 30 m × 30 m, from which slope data was derived. Normalized Difference Vegetation Index (NDVI) data were sourced from the MOD13Q1 product available on the NASA website (http://ecocast.arc.nasa.gov/), with a spatial resolution of 250 m × 250 m. This dataset features a long time series and high temporal resolution, meeting the requirements for attributive analysis in this study. Meteorological data, including monthly average temperature and precipitation datasets for China from 1901 to 2020 at a 1 km resolution, were sourced from the National Earth System Science Data Center (http://www.geodata.cn). These datasets have been validated using data from 496 independent meteorological observation stations, ensuring their reliability. Information on Population density, soil types, and Gross Domestic Product (GDP) were obtained from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (http://www.resdc.cn/). These datasets have a long time series and have been widely applied in various fields, including ecological and environmental studies.Furthermore, nighttime light data incorporated two datasets: DMSP-OLS and NPP-VIRS (https://www.ngdc.noaa.gov/eog/download.html). Nighttime light data can indirectly reflect human activities, and these two datasets complement each other in terms of coverage, time span, and data quality, fulfilling the requirements of this study.

The resolution of data influences the reliability of research outcomes.To enhance data analysis precision and consistency, all other datasets were resampled to a resolution of 30 m × 30 m, aligning with the precision of the land use data. This approach improves spatial detail and meets the scale requirements of the study, with all data unified in the Albers Conic Equal Area projection coordinate system to ensure consistency across the analysis.

Methods

Analysis of land use

This study quantitatively analyzes changes in land use over the past 30 years within the research area by employing three main methodologies; the single land use dynamic degree, the comprehensive land use dynamic degree, and the land use transfer matrix. The single land use dynamic degree assesses the magnitude of changes in the number of land use types within the region during a specific period, providing insight into individual categories of land use. In contrast, the comprehensive land use dynamic degree presents a broader overview of land use evolution, reflecting overall changes across various land use types in the study area29. Additionally, the land use transfer matrix plays a crucial role by elucidating both the direction and extent of land use transitions.

(1) Single land use dynamism

$$K = \frac{Ub - Ua}{{Ua}} * \frac{1}{T} * 100\%$$
(1)

where K represents the rate of change in the number of land use types during the study period, Ub-Ua denotes the difference in area of a particular land use type between the initial and final stages of the study, with T representing the duration of the study in years. When K is positive, it indicates an increase in the area of a particular land class, while a negative value for K signifies a decrease in its area.

(2) Comprehensive land use dynamism

$$LC = \frac{{\sum\limits_{i = 1}^{n} {\Delta LUi - j} }}{{2\sum\limits_{i = 1}^{n} {LUi} }} * \frac{1}{T} * 100\%$$
(2)

where LC represents the comprehensive land use dynamism in the study area during the research period, LUi denotes the area of the i-th land use type at the beginning of the study, △LUi-j is the absolute value of the area when the i-th type of land use converted to the j-th type during the study period. T is the duration of the study.

To clarify the evolutionary characteristics of different land classes in geographical space, this study utilized ArcGIS software and employed the transition matrix method to quantitatively analyze the mutual transfer areas and directions between land types from 1990 to 2020, revealing the process of land evolution30.

InVEST model

The habitat quality assessment for graduate research primarily relies on the Habitat Quality module within the InVEST model. The principle behind this assessment is to consider different land types as human disturbance factors, establish their relationship with threat sources, determine the sensitivity of different habitat types to these threats, and assess the impact distance between them. This process culminates in the generation of a habitat quality assessment map. A higher habitat quality index indicates greater biodiversity, while a lower index signifies stronger anthropogenic disturbances and reduced biodiversity levels.

In this model, the degree of habitat quality degradation is quantified by evaluating the sensitivity of threat sources and the intensity of threats to habitat quality. The specific equation is as follows:

$$D_{xj} = \sum\limits_{r = 1}^{R} {\sum\limits_{y = 1}^{Yr} {\left( {\frac{W_r}{{\sum\limits_{r = 1}^{R} {W_r} }}} \right)} } r_yi_{rxy}\beta _xS_{jr}$$
(3)

where R represents the number of threat factors; Wr is the weight of threat factor r; Yr is the number of grids of threat factor r; ry represents the threat factor value of grid y; irxy represents the impact of threat factor ry of grid y on grid x; βx denotes the accessibility of threat factors to grid x; Sjr represents the sensitivity of different habitat types j to threat factor r.

Within the model, the spatial decay of threat factors is characterized by two types of functions: linear and exponential decay.

If linear:

$$i_{rxy} = 1 - \left( {\frac{d_{xy}}{{d_{r\max }}}} \right)$$
(4)

If exponential:

$$i_{rxy} = \mathit{exp} \left( {\frac{ - 2.99d_{xy}}{{d_{r\max }}}} \right)$$
(5)

Where dxy represents the straight-line distance between grid x and grid y; drmax is the influence range of threat factor r.

The calculation formula for habitat quality assessment is as follows:

$$Q_{xj} = H_j\left[ {1 - \left( {\frac{{D{}^{z}_{xj}}}{{D{}^{z}_{xj} + k^{z} }}} \right)} \right]$$
(6)

where \(Q_{xj}\) represents the habitat quality index of grid x in land use type j; Hj represents the habitat suitability of land use type j; Dxj represents the habitat degradation degree of grid x in land use type j; k is the half-saturation constant; and z is the normalized constant, which are default parameters of the model.

In this study, three land use types—Cropland, Barren, and Impervious—were selected as stress factors. The rationale for this selection is as follows: Cropland is a land use category resulting from the combined influence of human activities and natural processes. Notably, Cropland occupies the largest area within the lake basin of central Yunnan, posing a potential threat to the balance of natural ecosystem. Impervious surfaces represent typical artificial land covers, directly reflecting human-induced disturbances to the natural environment. Additionally, Barren areas in the study region primarily include wastelands and bare lands, which exhibit poor natural substrate conditions. Theoretically, Impervious surfaces exert the most significant impact on habitat quality, followed by Cropland, with Barren areas having the least influence. In the habitat quality assessment model, the sensitivity of different stress factors varies, with values closer to 1 indicating higher sensitivity of the land use type to the threat factor. Therefore, based on the actual conditions of the study area and referencing relevant literature25,31,32,33, the parameters for the habitat quality module were appropriately configured (Tables 1 and 2).

Table 1 Threat factors, their maximum impact distance, and weight.
Full size table
Table 2 The sensitivity of habitat types to each threat factor.
Full size table

GeoDetector

The Geographical Detector is a statistical model designed to explore spatial heterogeneity and identify driving factors in geographical contexts34. In this study, we primarily utilized the factor detector within the geographical detector framework to quantitatively analyze the influence mechanisms of natural and non-natural factors on the spatial evolution of habitat quality. Lake basins represent complex ecosystems that integrate natural resources, socio-economic activities, and water resources. Typically, various factors such as topographic characteristics, soil type diversity, vegetation coverage, fluctuations in temperature and precipitation, and human activities can significantly impact the ecological environment quality within a basin.

The terrestrial portion of the lake basin in central Yunnan is predominantly characterized by mountainous areas and lake basins, with significant variations in soil types and a relatively dense population. Consequently, this study selected seven natural factors—Land use type, Elevation, Slope, NDVI, Soil type, Temperature, and Precipitation—and three non-natural factors—Population density, DMSP-OLS, and GDP—as independent variables, with the habitat quality index serving as the dependent variable. The NDVI and Population density were classified into eight categories using the natural breaks method, while the remaining variables were classified into seven categories. A 1 km × 1 km grid was established, with grid center points as sampling locations, and all data were discretized into categorical variables. The primary objective was to detect the spatial heterogeneity of the dependent variable Y (habitat quality) and quantify the explanatory power of independent variables X (influencing factors) on the spatial heterogeneity of Y, measured by the q-value, which quantifies the strength of these relationships. The specific calculation formulas for the q value are expressed as follows:

$$\begin{gathered} q = 1 - \frac{{\sum\limits_{h = 1}^{L} {N_{h} \sigma_{h}^{2} } }}{{N\sigma^{2} }} = 1 - \frac{SSW}{{SST}} \hfill \\ SSW = \sum\limits_{h = 1}^{L} {N_{h} \sigma_{h}^{2} ,SST = N\sigma^{2} } \hfill \\ \end{gathered}$$
(7)

where q represents the measure of spatial distribution of habitat quality for different influencing factors, where q ∈ [0,1]. A larger q value indicates a more substantial explanatory power of a specific influencing factor on habitat quality. L is the sample size of influencing factors, N and \(\sigma^{2}\) represents the number of units in the study area and the variance of habitat quality, respectively. Nh represents the sample size of layer h, \(\sigma_{h}^{2}\) represents the variance of habitat quality in that layer. Additionally, SSW and SST are the sum of within-layer squares and the total sum of squares, respectively.

Habitat contribution rate of land use transition

The transformation of land use types can cause changes in the corresponding regional habitat quality. Therefore, the habitat contribution rate is used to quantify the response of habitat quality to the evolution of land use in the lake basin of central Yunnan.

$$Q = \frac{{(Q_{t - 1} - Q_{t} ) \times S_{t} }}{S} \times 100\%$$
(8)

Q represents the habitat contribution rate of land use transformation; Qt-1 and Qt respectively represent the habitat quality index at the beginning and end of a particular land use type transformation. St represents the area of changing land cover; while S denotes the total area of the study area.

Results

Analysis of spatiotemporal changes in land use

Land use quantity variation

The lake basin in central Yunnan is predominantly characterized by cropland and forestland, which collectively account for approximately 70% of the total basin area. During the study period, the area of cropland decreased by 611.35 km2, resulting in a reduction of 13.31% in its proportion. In contrast, forestland and impervious surfaces showed an increasing trend, with expansions of 546.46 km2 and 267.41 km2, respectively, corresponding to growth rates of 11.9% and 3.77%. Grassland area experienced a modest increase of 1.35% between 1990 and 2010. Meanwhile, the areas of shrubland, bare land, and water bodies maintained relative stability (Fig. 2).

Fig. 2
figure 2

Proportion of land use type area in the lake basin of central Yunnan.

Full size image

Analysis of land use spatial evolution

The chord diagram effectively illustrates the transfer directions and quantitative characteristics of different land use types in the central Yunnan lake basin during the study period (Fig. 3). Across the three periods of 1990–2000, 2000–2010, and 1990–2020, the directions and types of land use transfers were relatively consistent, with cropland, forestland, and grassland exhibiting the highest transfer intensities and frequent interconversions. The ranking of land use types by transferred-out area was as follows: cropland > grassland > forestland > shrubland > water bodies > bare land > impervious surfaces. Cropland experienced the most significant changes, with 840.83 km2 transferred out and 229.48 km2 transferred in, resulting in a net reduction of 611.35 km2. Forestland and grassland received 649.7 km2 and 340.61 km2, respectively, primarily from cropland, which is associated with the Grain for Green Program initiated in the lake basin of central Yunnan in 200035. Impervious surfaces mainly expanded from cropland and grassland, with transferred-in areas of 147.41 km2 and 24.25 km2, respectively. Shrubland, bare land, and water bodies showed minimal conversions with other land types, indicating weaker transfer intensities.

Fig. 3
figure 3

Chord diagram of land use transitions in the lake basin of central Yunnan.

Full size image

This study provides a spatial visualization of 13 transition types with significant transfer intensity across different periods in the research area, offering deeper insights into the spatial characteristics of land use type transitions (Fig. 4).

Fig. 4
figure 4

Chart of significant land use transfer types in the lake basin of central Yunnan (1: Cropland; 2: Forest; 3: Shrub; 4: Grassland; 5: Water; 8: Impervious).

Full size image

From 1990 to 2010, the northern shore of Dianchi Lake experienced substantial land use changes, dominated by conversions from cropland to forestland (115 km2), cropland to grassland (235 km2), and grassland to forestland (104 km2). Other transition types were less intensive and sporadically distributed across the study area. Between 2010 and 2020, the most prominent transition was from grassland to cropland, covering 172.65 km2 and evenly distributed throughout the lake basin in central Yunnan. A notable conversion from grassland to impervious surfaces (44.92 km2) was concentrated in the Caohai area near Dianchi Lake. Additionally, numerous transitions outwards from cropland and grassland into other land categories widely spread along each lake’s shoreline.

Over the entire study period (1990–2020), 43 distinct land use transition types were identified, with the most intensive transitions involving cropland and grassland shifting into other categories. The regions surrounding each lake basin, characterized by higher elevations, mountainous terrain, and minimal human presence, are predominantly covered by extensive forested and grassy areas. In these zones, the primary transitions involve grassland shifting to forestland and cropland. In contrast, the basin regions surrounding the lakes, which are urbanized and agriculturally active, exhibit intense human activity-driven transitions. This human impact leads to predominant transition types, such as cropland converting into forestland or grassland, as well as transformations toward impervious surfaces. Notably, the conversion of cropland to impervious surfaces is concentrated in the flat basin regions of each watershed, with the most significant aggregation observed in the Dianchi Lake basin.

Evaluation of land use dynamics

According to formulas (1) and (2), the dynamic degree of both single and comprehensive land use in the study area from 1990 to 2020 was calculated for four time periods (Table 3). The comprehensive land use dynamic degree in the lake basin of central Yunnan showed a decreasing trend during the study period, declining from 0.74% to 0.37%. Notably, before 2010, the comprehensive land use dynamic degree exceeded the overall dynamic degree for 1990–2020 (0.53%), indicating that land development activities were relatively intense between 1990 and 2010. From the perspective of single land use dynamics, the rate of change in cropland exhibited a downward trend, suggesting a gradual decrease in arable area over the study period. This decline can be attributed to significant conversions between cropland and other types over three decades, resulting in substantial transfer areas; thus, its dynamic degree is presented as negative. Conversely, forestland demonstrated an increasing rate of change; however, this growth rate decreased by 2.05% during the period from 2010 to 2020 compared to that from 1990 to 2000, indicating a slowdown in forestland expansion within the study area. The change rate for shrubland displayed an initial decline followed by an increase trend, while the grassland trend was opposite those observed for shrubland. Waterbody experienced considerable fluctuations between 2000 and 2010. Additionally, bare land and impervious surfaces had significantly increased change rates respectively reaching 23.01% and 10.68% between year 2010-2020, 2000-2010.

Table 3 Land use dynamics for different periods in the lake basin of central Yunnan.
Full size table

The spatiotemporal evolution characteristics of habitat quality

The habitat quality assessment results for four years in the study area were obtained through the InVEST model, categorizing them into five levels: low (0–0.3), relatively low (0.3–0.8), moderate (0.7–0.8), relatively high (0.8–0.9), and high (0.9–1.0). The areas and percentages of habitat quality under each level were statistically analyzed (Table 4).

Table 4 Area and Percentage change of habitat quality grade in the lake basin of central Yunnan (km2, %).
Full size table

The average habitat quality values for the four periods in the lake basin of central Yunnan were 0.610, 0.661, 0.685, and 0.676, respectively, showing an overall trend of first increasing and then decreasing, with peak habitat quality reached in 2010. Each year predominantly exhibited relatively low and high-grade distributions; during the study period, the area classified as relatively low-grade decreased by 604.68 km2, accounting for a reduction of 13.17%. Conversely, the area classified as high grade increased by 546.46 km2, reflecting a growth rate of 11.90%. The area categorized as moderate level showed a trend of first rising and then falling. Additionally, there was an increase of 3.79% in low-grade areas, while relatively high-grade areas experienced a decrease of 0.37%.

The perspective of spatial distribution of habitat quality in the study area reveals that the habitat quality in the lake basin of central Yunnan exhibits a spatial pattern characterized by an increase from the lakeshore towards surrounding areas for each lake basin (Fig. 5). This gradient indicates a close relationship between habitat quality and geographical elevation, where low-quality habitats are primarily situated in densely populated areas adjacent to the lakes, resulting in increased human activities that compromise environmental integrity. Regions exhibiting low habitat quality are closely associated with impervious surfaces, primarily centered around the northern shore of Dianchi Lake. Other significant concentrations of low-quality habitats are observed along the northern shore of Fuxian Lake, the southwestern corner of Xingyun Lake, and the periphery of Qilu Lake’s water body. These areas typically exist at lower elevations, which facilitates agricultural and socio-economic development. However, such developments lead to frequent land-use activities that threaten the environmental quality within these basins, contributing to habitat degradation.

Fig. 5
figure 5

Spatial distribution of habitat quality in the lake basin of central Yunnan.

Full size image

In contrast, high-grade habitat quality regions are predominantly located around mountainous areas adjacent to individual lake basins. Notably, concentrated zones can be identified in the northern part of Dianchi Basin, the northwestern region of Fuxian Basin, and the southern section of Qilu Basin. These high-quality habitats benefit from reduced human disturbance, allowing for greater biodiversity and ecosystem stability. Relatively low-grade habitat quality regions are most concentrated at each lake’s shoreline, exhibiting a broad radiation range that mirrors the influence of localized human pressures. Conversely, relatively high-grade habitats tend to cluster around lake water bodies themselves, taking advantage of the natural water sources. Areas classified with moderate habitat quality are sporadically distributed throughout the entire lake basin of central Yunnan, illustrating a mixed environmental condition, which can fluctuate based on ongoing land-use changes and conservation efforts.

Analysis of habitat quality drivers in the lake basin of central Yunnan

In this study, the factor detector in the GeoDetector was used to calculate the impact of various factors on habitat quality in the lake basin of central Yunnan. The driving factor detection results (significance level p < 0.001, as shown in Table 5) revealed the following order of impact: land use type (X9) > NDVI (X3) > elevation (X1) > slope (X2) > temperature (X6) > soil type (X8) > DMSP-OLS (X4) > GDP (X10) > population density (X5) > precipitation (X7).

Table 5 Detection of q values for different impact factors in the lake basin of central Yunnan.
Full size table

Land use type emerged as the primary determinant, exerting a far greater impact on the spatial differentiation of habitat quality compared to other factors. NDVI, elevation, and slope followed as significant drivers. NDVI serves as an indicator of vegetation growth, while elevation and slope represent key topographic attributes. These three factors are intricately linked and collectively influence the spatial patterns of habitat quality within the basin. The lake basin is predominantly characterized by mountainous terrain, featuring high elevations, sparse human settlement, and dense vegetation cover, which collectively contribute to superior habitat quality. Conversely, GDP, population density, and precipitation demonstrate lower explanatory power regarding spatial differentiation in habitat quality. This is because habitat quality is a multifaceted outcome influenced by the interplay of numerous factors, whereas GDP and population density, which primarily reflect economic development levels, do not maintain a direct or proportional relationship with ecological conservation efforts. Within the lake basin in central Yunnan, lakes cover a substantial area and play a critical role in regulating regional climate and maintaining ecological health. As a result, the influence of temperature and precipitation on the spatial differentiation of habitat quality is relatively minor. While soil type variations lead to differences in surface vegetation and land use patterns, which in turn influence habitat quality. However, compared to the more direct impacts of land use, NDVI, and elevation, its effect is relatively minor. This highlights the complex interplay of multiple factors in determining habitat quality and underscores the importance of land use planning and vegetation management in enhancing ecological conservation and habitat quality in central Yunnan.

Impact of LUC on habitat quality

Based on the analysis of habitat quality driving forces, it is evident that land use types have the highest impact on habitat quality. To measure the effects of land use changes on habitat quality, a calculation of the contribution rate of land use transformation that affect habitat quality was conducted (Table 6). The results indicate that between 1990 and 2020, the improvement in habitat quality in the lake basin of central Yunnan is substantially attributed to the transition from agricultural land, represented by cropland, to ecological land, specifically forestland and grassland. The cumulative contribution rate from these transitions reached an impressive 91.89%, highlighting the significant positive impact of converting cropland to more sustainable land uses. Contrarily, the succession process from ecological land back to agricultural production areas has emerged as a primary driver for the decline in habitat quality within the study area. This reverse transition encompasses a significant total area of 215.32 km2, contributing over 67% to the decline in habitat quality. The loss of ecological integrity due to the reversion of forestland and grassland to agricultural uses underscores the challenges posed by changing land use priorities and the pressures of agricultural expansion. Furthermore, the replacement of cropland and grassland areas with impervious surfaces has further exacerbated degradation in habitat quality.

Table 6 The primary land use change types and contribution rate that affect habitats in the lake basin of central Yunnan.
Full size table

To further clarify the impact of LUC on habitat quality, a superimposed analysis of habitat quality results from 1990 to 2020 was conducted, followed by spatial representation (Fig. 6). By comparing the land use change maps with the habitat quality change maps for each period, it is evident that regions with significant land use changes also experienced alterations in habitat quality. Between 1990 and 2010, the area of improved habitat quality exceeded that of degraded areas, indicating an overall enhancement in habitat quality. However, from 2010 to 2020, a decline in habitat quality was observed. Over the entire period from 1990 to 2020, areas, where habitat quality remained unchanged, accounted for 68.74%, predominantly located around lake water bodies; conversely, degraded habitats comprised 9.23% of the total area and were primarily concentrated along the northern shore of Dianchi Lake Basin; while areas exhibiting improved habitat quality reached 22.04%, mainly distributed around individual lake basins. As indicated in Fig. 4, the improvement in habitat quality between 1990 and 2010 can be attributed to substantial conversions of cropland into forested and grassland areas. The degradation observed during the period from 2010 to 2020 is largely due to the ongoing expansion of impervious surfaces.

Fig. 6
figure 6

Habitat quality change diagram in the lake basin of central Yunnan.

Full size image

Discussion

Habitat quality plays a crucial role in indicating regional biodiversity; thus, assessing the habitat quality of an area contributes to the conservation of biodiversity36. When evaluating habitat quality based on the InVEST model, it simultaneously considers stressors threatening the habitats, the maximum distance of stressors, the geographical spatial distance, and the sensitivity of each habitat to stress. Therefore, the results of habitat quality obtained from this model computation are relatively reliable37.

Habitat quality is the result of complex interactions among multiple factors, making its dynamics inherently intricate. This study underscores that land use change is the dominant driver influencing habitat quality, whereas socio-demographic factors such as GDP and population density exhibit limited explanatory capacity. This limitation arises because socio-demographic factors are fundamentally tied to land; they indirectly influence regional habitat quality through the exploitation of natural resources and modifications to land use patterns and spatial configurations. Moreover, land use change, NDVI, and elevation have a more pronounced impact on habitat quality, thereby overshadowing the effects of socio-demographic factors. While socio-demographic factors reflect regional economic conditions, strategically guided economic development can serve as a strong foundation for ecological conservation, fostering a synergistic balance between ecological sustainability and economic prosperity38. Future research should adopt a holistic perspective by incorporating a wider array of natural and non-natural factors to refine the accuracy and depth of attribution analysis. Additionally, habitat quality exhibits significant spatial heterogeneity39. Consequently, the impact of land use change on habitat quality in this study also demonstrates distinct spatial differentiation characteristics.

habitat quality in the the lake basin of central Yunnan was generally favorable overall, showing an upward trend from 1990 to 2010 and a slight decline from 2010 to 2020. Over the 30-year period from 1990 to 2020, 22.04% of the study area experienced notable improvements in habitat quality. As depicted in Fig. 3, this evolutionary trajectory is strongly correlated with land use changes. The primary driver behind the enhancement in habitat quality was the large-scale conversion of agricultural production land to ecological land, particularly during the 1990–2010 period. This era was marked by intensified land development activities, coupled with the implementation of key ecological forestry projects in Yunnan Province, such as mountain closure and afforestation initiatives, which significantly contributed to environmental protection efforts. As a result, ecological spaces characterized by forests and grasslands expanded significantly, while agricultural production areas, predominantly farmland, continued to diminish. Detailed analyses reveal that 343.49 km2 and 325.54 km2 of farmland were converted to forest and grassland, respectively, with an additional 268.86 km2 of grassland transitioning into forest. These substantial increases in natural habitats have been pivotal in driving the improvement of habitat quality40,41. During the study period, areas exhibiting enhanced habitat quality were predominantly concentrated in high-altitude mountainous and hilly regions. This spatial distribution aligns with patterns observed in numerous other regions42,43, as such areas are typically situated farther from human activity centers and retain extensive coverage of native vegetation, fostering favorable conditions for habitat regeneration and quality improvement.

However, approximately 9.23% of the lake basin area in central Yunnan has experienced notable habitat quality degradation, particularly during the period from 2010 to 2020. This ecological decline is primarily attributed to accelerated urbanization and economic development in the region, driven by the implementation of urban master plans formulated by the municipal governments of Kunming and Yuxi. These strategic plans outlined comprehensive urban development strategies for counties adjacent to major lakes, including Dianchi Lake and Fuxian Lake.

The rapid expansion of urban areas has resulted in a substantial increase in impervious surfaces, which has emerged as a major driver of habitat quality deterioration44,45. Specifically, significant land use changes have been observed, with 147.41 km2 of cropland and 24.25 km2 of grassland being converted into impervious surfaces, posing serious threats to local ecosystems46. This conclusion is similar to the research results on the impact of land use change in the Urban Agglomeration of central Yunnan on ecosystem service value47. Spatial analysis reveals that degraded habitat areas are predominantly located in flat terrains characterized by robust economic activities, high population density, and intensive agricultural production. Intensive anthropogenic activities in these regions have resulted in habitat fragmentation, thereby disrupting the natural flow of materials, energy, and information among different ecological patches48,49. Furthermore, the substantial conversion of ecological land types, especially forestland and grassland, into cropland. Historical data from 1990 to 2020 indicate that approximately 126.8 km2 of grassland and 88.52 km2 of forestland have been transformed into cropland. This transformation has been a major contributor to localized habitat quality deterioration. Existing studies have identified southwest China as a typical ecologically fragile region, where the encroachment of agricultural production spaces into ecological areas has had an even more pronounced impact than urban expansion50,51. Moreover, research indicates that cropland, subject to substantial human interference, generates agricultural non-point source pollution, leading to the degradation of native vegetation and further exacerbating habitat quality decline52,53.

Formulating socioeconomic policies from a social-ecological systems perspective is essential for maintaining sustainable development between society and ecosystems. In 2024, the territorial spatial master plans (2021–2035) of Kunming and Yuxi cities underscored the ongoing commitment to the integrated management of mountains, rivers, forests, farmlands, lakes, and grasslands. These plans outline strategies such as delineating ecological protection redline zones, optimizing the spatial structure of agricultural and urban development areas, safeguarding the ecological security of lake regions, enhancing the stability of regional ecosystems, and fostering harmonious coexistence between humans and nature. In this context, it is crucial not only to improve the habitat quality of degraded areas in the central Yunnan′s lake basins but also to sustain the ecological environment in areas where habitat quality has already been enhanced. By capitalizing on the inherent advantages of lake wetlands, concerted efforts should be directed toward strengthening the basin’s ecosystem service functions, thereby fostering long-term ecological resilience and sustainability. This holistic approach ensures the preservation of ecological integrity while supporting socioeconomic development aligned with environmental stewardship.

This study utilized the geographic detector method to investigate the driving factors affecting habitat quality. However, it should be noted that the geographic detector is primarily based on static data analysis and assumes linear relationships among variables. Future research could be enhanced by incorporating additional statistical methods and dynamic analytical models to provide a more comprehensive explanation of the driving mechanisms affecting habitat quality in the basin. Furthermore, while this study concentrated on the influence of LUC on habitat quality, the unique characteristics of the central Yunnan′s lake basins—as plateau lake systems with pronounced vertical zonation—present opportunities for deeper exploration. Future research could adopt a more granular perspective by evaluating the role of topographic attributes and altitudinal gradients, thereby unraveling finer-scale variations in habitat quality and providing richer insights into these complex ecosystems. Such an approach would not only enhance the precision of findings but also contribute to more targeted conservation strategies.

Conclusions

This study evaluated the impact of LUC on habitat quality in the lake basin of central Yunnan between 1990 and 2020. Land use change was obvious, cropland has been the predominant land use type but has steadily decreased, while forested areas and impervious surfaces have increased during this period. There has been substantial conversion among land types, primarily from cropland to grassland and forestland, and the expansion of impervious surfaces resulted from encroachment upon both cropland and grassland. Habitat quality in the lake basin improved as a whole over the studied period, with ralatively low and high-grade habitats dominating, ralatively low grade habitat areas shrinking, and high-quality regions expanding. LUC) emerged as the primary factor influencing habitat quality, accounting for 91.89% of the contribution rate to habitat quality improvement predominantly due to shifts from cropland to forestland or grassland. Conversely, encroachment from cropland and impervious surfaces was responsible for 94.12% of the contribution rate to habitat quality degradation. A close connection exists between land use patterns and habitat quality distribution, particularly in high mountainous areas and cropland distribution areas of individual basin, as well as around impervious surfaces near Dianchi Lake. To enhance habitat quality, the study recommends prioritizing reforestation efforts in degraded areas, adopting conservation tillage practices in agricultural zones, and strengthening the protection of forest and grassland regions to preserve biodiversity. Furthermore, stringent regulation of urban expansion is crucial to curb habitat fragmentation and ensure sustainable development. These measures aim to balance ecological preservation with socioeconomic development, fostering resilience in the region’s ecosystems while supporting long-term environmental and human well-being.