Assessing the value and sensitivity of ecosystem services based on land use in the middle and lower reaches of the Shiyang River

In response to increasing ecological and environmental challenges in arid areas, it is of great significance to investigate the ecosystem service value (ESV), accompanying the changes in ecological sensitivity for the protection of ecologically vulnerable areas. Our analysis seeks to elucidate the ESV and ecological sensitivity changes in the middle and lower reaches of the Shiyang River to determine the trends and influencing factors of ESV under changing land use patterns. The key findings include: (1) From 1995 to 2020, the ESV in the study area witnessed fluctuations, culminating in an overall decline of 1.249 × 108 yuan. (2) In 2020, sensitivity coefficients (CSs) for ESV were as follows: 0.4335 for grassland, 0.2586 for farmland, and 0.1170 for unused land within the study area. Furthermore, coefficients of improved cross-sensitivity (CICSs) for the reciprocal transformation of farmland, grassland, and unused land were 1.10, 1.18, and 1.54, respectively, indicating the pivotal role of the three land types in driving ESV fluctuations.


Introduction
With rapid socio-economic development, human activities have induced substantial changes in land use, affecting the structure and function of ecosystem services.This has exacerbated ecological and environmental issues, a concern particularly prevalent in developing countries (Barbier andHochard 2016, Cao et al 2021).Ecosystem service value, hereafter referred to as ESV, includes the value of services derived directly or indirectly from ecosystems and their species and are directly or indirectly provided to humans (Shao et al 2018, Jie et al 2020).Through the assessment of ESV, we gain not only a deeper understanding of ecosystems in terms of supply, regulation, support, and cultural functions but also timely insight into the development trend of the ecological environment (Nanshan andJijun 2017, Aanesen andArmstrong 2019).Such insights hold significant implications for the sustainable development of ecosystems.
Over the period, the investigation of ESV and sensitivity has been a major focus of discussion concerning sustainable development.(Keith et al 2017, Kang et al 2022).The concept of the ESV function was first introduced by Costanza (Costanza et al 1997), and it was further refined by the Millennium Ecosystem Assessment in the definition of 'human benefits from ecosystems,' subsequently driving the evolution of ESV research (Assessment 2005).In recent years, global biodiversity has declined, diminishing land ecological quality, and degradation has affected the functionality of ecosystem services (Pan et al 2023).This has heightened global attention to coordinate the need for balanced land use development and efforts to mitigate the adverse impact of economic and social advancement on ecosystems (Yi et al 2017).Arowolo assessed and analyzed the impact of land use on ESV in Nigeria, revealing that single land use types must be compatible with prevailing socio-economic conditions to prevent disparities in economic, social, and ecosystem development.(Arowolo et al 2018).Similarly, in another study conducted in Kolkata (India), ESV evaluation demonstrated how rapid urban digitization contributed to the gradual disappearance of certain geomorphological groups around the city, leading to a significant decline in specific ecosystem service functions (Sannigrahi et al 2020).In response to

Study area
The Shiyang River Basin is located between 36°29′-39°27′N latitude and 101°41′-104°16′E longitude, covering a total basin area of 41,600 km 2 .It is bounded by the Wushao Ridge to the east, the Dahuang Mountain to the west, the Badain Jaran Desert and Tenger Deserts to the north, and the northern foothills of the Qilian Mountains to the south.The basin comprises eight main rivers, running from west to east: Dajing River, Gulang River, Huangyang River, Zamu River, Jinta River, Xiying River, Dongda, and Xida Rivers.The middle and lower reaches of the Shiyang River Basin are located deep within the inland hinterlands, featuring a predominantly flat terrain at an average altitude of around 1,600 m, and the climate of the region is classified as temperate continental arid climate.The region experiences strong solar radiation, abundant sunshine, an annual average temperature of 7.8 °C, high evaporation, and low precipitation, ranging from 150 to 250 mm annually.Particularly in the downstream Minqin area, the annual average precipitation falls below 150 mm.These middle and lower reaches of the Shiyang River are among the most densely populated inland river regions in China and have the highest degree of water and soil resource development and utilization of water and soil resources in China (Liu et al 2021, Jun et al 2022, Li et al 2022).Moreover, irrigated agricultural areas are characterized by intense human activities and often exhibit the most severe ecological and environmental challenges within a river basin.The study areas selected for analyzing the impact of land-use change on the value and sensitivity of ecosystem services include Jinchang District, Yongchang County, Liangzhou District, Minqin County, and Gulang County in the middle and lower reaches of the Shiyang River (figure 1).

Data collection
The primary datasets employed in this study include land use data, Digital Elevation Model (DEM) elevation data, and statistics on grain crop area and grain yield.The land use data, spanning from 1995 to 2020, were collected from the China Land Cover Dataset, collaboratively compiled by Professors Yang and Huang from Wuhan University (Yang and Huang 2021).This study combined the land classification data from three land adjustment surveys with the specific conditions of the middle and lower reaches of the Shiyang River, categorizing the land into eight types: farmland, forest land, grassland, construction-use land, mining land, waters, wetland, and unused land.DEM elevation data were retrieved from the Geospatial Data Cloud website (http://www.gscloud.cn).Grain crop area and yield data, ranging from 1995 to 2020, were mainly derived from the statistical yearbooks of Gansu Province and the statistical yearbook of relevant cities and counties of the Shiyang River Basin.

Changes in land use area
The utilization of a land-use transfer matrix can elucidate both the direction and magnitude of area changes in various land-use types over a specific time frame (Long et al 2022).It is a tool to describe the process of land use change, and shows the transfer relationship between different land use types in the form of a matrix.The Sankey diagram, which serves as a visual representation of this land-use transfer matrix, allows for a more intuitive and clear understanding of the directions and magnitudes of mutual transfers between different land-use categories.Where S represents the area of land use area; n denotes the number of different land uses at the beginning of the study; i is the type of land use at the beginning of the study; j signifies the type of land use at the end of the study.The land-use transition matrix is a square matrix in which each element represents the area of land use types transferred from one land-use type to another.The rows and columns of the matrix represent different land use types, and each element (i, j) in the matrix represents the area transferred from type i to type j.

ESV
In this study, we combined the findings from previous research (Yu et al 2022) with the specific conditions of the Shiyang River region to refine the economic valuation of the equivalent factor for individual ecosystem services.The calculation formula is presented as follows: Where X represents the economic value equivalent of a single ecosystem in the middle and lower reaches of the Shiyang River; EA signifies the economic value of a single ESV equivalent factor in China; y′ denotes the average grain yield of farmland in the middle and lower reaches of the Shiyang River Basin from 1995 to 2020; y represents the average grain yield of farmland in China from 1995 to 2020.Arable land yield is a commonly used economic indicator that can be easily compared with other economic activities to better demonstrate the importance of ecosystems to people, governments, businesses, and the public.Moreover, it can reflect the productivity of the land, and the value of the services provided by the ecosystem can be measured by comparing it with the yield of cultivated land.This helps to assess the contribution and impact of ecosystems on land resources.Most importantly, the importance of land to ecosystems is self-evident, which helps people pay more attention to the conservation and sustainable use of ecosystems, so as to promote ecological compensation mechanisms and promote ecosystem protection and restoration.So in this study, we collected data on the average grain yield of farmland in both the national and Shiyang River Basin contexts from 1995 to 2020.Specifically, the average grain yield of farmland in China from 1995 to 2020 was recorded as 5017.05kg.In contrast, the average grain yield of farmland in the middle and lower reaches of the Shiyang River Basin for the same period was at 3796.4 kg•hm −2 .The economic value equivalent factor of a single ecosystem service in the middle and lower reaches of the Shiyang River Basin was calculated using Formula (2) and was found to be 339.83.This value was then combined with Xie's proposed ESV equivalent scale (Xie et al 2008) to derive the value coefficient of ecosystem services specific to the middle and lower reaches of the Shiyang River, as detailed in table 1.
Formula of ecosystem service value Where ESV represents the total value of services for regional ecosystems; VC i denotes the value coefficient of ecosystem services for the i category; Si signifies the area of the i lot.

Ecosystem sensitivity analysis
In this study, sensitivity is employed to reflect the extent to which ESVs rely on changes in ESV coefficients over time.It serves as a measure of the applicability of ESVs and the reliability of assessment results (Akber et al 2018).The calculation formula is provided below: Where CS represents the sensitivity coefficient of ESV; ESV j signifies the adjusted ecosystem service value factor (typically adjusted by ±50%); ESV i denotes the value of initial ecosystem services; VC ik and VC jk represent the value coefficients of ecosystem services before and after the adjustment of land category k, respectively.When |CS|<1, it indicates that the value coefficient of ecosystem services for land types was stable.The evaluation results were deemed credible, but the accuracy of the evaluation results was less accurate, leading to lower credibility.
The CICS depicts the improved cross-sensitivity of ESV to changes in land area between different land types.It quantifies the influence on ESV resulting from changes in the ESV induced by net area transformations between two specific types of land (Yang et al 2023).The calculation formula is as follows: Where CICS kl represents the coefficient of improved cross-sensitivity; VC k denotes the value coefficient of the ecological service for the k land type; VC l signifies the value coefficient of the ecological service for the l land type; ΔAkl is the net conversion area from land type k to land type l; ΔPESV represents the total change in the value of ecological services over a specific period.

Land use changes
In table 2 and figure 2, we have provided an overview of the prevailing land types in the middle and lower reaches of the Shiyang River.These land types majorly include unused land, grassland, and farmland, collectively constituting approximately 95% of the total land area.The unused land claims the largest share, covering approximately 53% of the total area.It is primarily concentrated in the middle and southern regions of the river basin, covering Minqin County in the lower reaches and scattered irregularly throughout the central and western regions.Specifically, these unused lands are typically adjacent to the Tengger Desert and the Badain Jaran Desert.Grassland is the next most prominent land type, contributing 23% of the total area, and is mostly situated in the southern and western edges of the basin, with sporadic distribution in the northern and eastern regions.These areas are in proximity to the Qilian Mountains with abundant water resources, primarily characterized by extensive grasslands, and have exceptional ecological environments (Lei 2022).In contrast, the northern and eastern regions predominantly feature artificially cultivated grasslands as a result of various ecological restoration measures, including sand fixation afforestation, grass planting, regeneration of sandy vegetation, and the cultivation of shrub stubble.Farmland accounts for approximately 19% of the total area, with the majority concentrated in the central and northern oasis plain areas.It has been observed that in the evolving landscape between 1995 and 2020 (figure 3), there was a consistent decrease in unused land, which declined by 489.22 km 2 during this period, with an average annual decrease of 19.57km 2 .The grassland area experienced a reduction of 583.63 km 2 , with an average annual decline of 23.35 km 2 .Farmland expanded by 856.41 km 2 , marking the most substantial net change, characterized by an increase-decrease-increase trend.Construction-use land and mining land witnessed respective increases of 140.77 km 2 and 155.73 km 2 .Conversely, forest land showed a relatively smaller net change, decreasing by 58.64 km 2 over the course of 25 years.Wetland experienced a modest net change, initially decreasing and then rebounding, resulting in a net area decrease of 24.7 km 2 .Waters displayed a pattern of initial increase followed by a subsequent decrease, with a total net increase of only 3.29 km 2 , which was not statistically significant.
The overall decline in grassland, forest land, unused land, and wetland is nearly balanced by the simultaneous increase in farmland, construction-use land, and mining land within the study area.This phenomenon can be attributed to rapid population growth, socio-economic development, and urbanization in the Shiyang River Basin from 1995 to 2020.These factors led to the conversion of unused land, as well as the reduction of grassland and forest land, to meet the increasing demand for farmland, urban and rural expansion, and industrial and mining zones.
Between 2010 and 2015, however, there was a decrease in farmland area.Meanwhile, an expansion of grassland and forest land was observed because of measures implemented to protect the ecological environment.Specifically, the 2010 Key Management Plan for the Shiyang River Basin in Gansu Province advocated for the restoration of farmland to forest and grassland (Qiao et al 2020).

Spatial-temporal evolution analysis of ESVs
According to figure 4, the ESV in the study area exhibited a fluctuating trend (decreasing → increasing → decreasing → increasing) from 1995 to 2020, decreasing from 7,133.2 million CNY to 7,008.3 million CNY.And the overall annual decline rate is 0.0351 × 108CNY/a.The expansion of farmland between 2000 and 2005 led to a rise in ESV, thus demonstrating a positive trend.Meanwhile, there was an increase in ESV from 2010 to 2015, attributed to policies favoring the restoration of farmland to forest and grassland, which boosted the ESVs specifically for these land types.The sustained growth in farmland, forest land, and wetland areas from 2015 to 2020 contributed to a positive trend in ESV, reflecting the significant efforts to improve the ESV in the region.Among the various land types within the middle and lower reaches of the Shiyang River, grassland had the highest total ESV and underwent the most significant change over the 25-year period, reduced by a total of 231.1 million CNY.Farmland ranked second in terms of overall ESV importance (table 3).
Moreover, wetlands exhibited larger variability in their ESV, with the most substantial decline occurring during the period from 1995 to 2000, resulting in a total decrease of 116.4 million CNY.This specific timeframe coincided with the rapid development of the agricultural sector, leading to the continuous encroachment of wetlands with limited economic benefits.Throughout the study period, the ESV reduction of wetlands was due to the synergistic effect of rapid economic and social development and the shortage of natural resources.During the period from 2000 to 2020, due to the rapid development of agricultural economy, wetlands were converted  to farmland, and then due to the implementation of ecological and environmental measures, such as returning farmland to forest and grassland, farmland was turned to grassland, but due to the arid area of Northwest China, the natural shortage of water resources led to the gradual transformation of some grasslands into tidal land, and finally wetlands into unused land.The reason for this is that the rapid economic and social development and the shortage of natural resources during the study period synergistically affect the ESV of wetlands.In contrast, the value of other ecosystem services remained relatively stable, all falling within the range of 50 million CNY.
The ranking of each ecosystem service type's value in the middle and lower reaches of the Shiyang River Basin, ordered from highest to lowest, is shown in table 4: soil conservation > biodiversity maintenance > climate regulation > hydrological regulation > gas regulation > provision of aesthetic landscapes > food production > raw material production.
Specifically, a significant portion of the ESV was attributed to the support and regulatory functions, with the values of soil conservation and biodiversity maintenance services decreasing by 14.5 million CNY and 26.3 million CNY, respectively.However, it is worth highlighting that the fluctuations in the value of various ecosystem services within the regulatory function did not reach the levels observed in 1995.This indicates the need for further strengthening of ecological restoration efforts.
Based on the empirical data collected from the study area, the assessment of ESVs in the middle and lower reaches of the Shiyang River for the period from 1995 to 2020 involved the classification of these values into three distinct levels: low, medium, and high, corresponding to value ranges of [0, 5), [5,15), and [15, 50], respectively.Figure 5 represents the distribution of ESVs within the study area.Specifically, the values are evenly distributed across the region, predominantly falling within the median-to low-value areas.This even distribution can be attributed to the substantial presence of farmland, grassland, and unused land in the watershed, which collectively comprise the largest land-use categories and exhibit considerable spatial overlap.
In contrast, high-value areas are scattered throughout the study area.These high-value zones are primarily scattered in the central farmland regions, areas characterized by extensive grassland cover, and some river beach wetlands situated in the northeastern part of the study area (Huang and Ma 2013).

Sensitivity analysis of ESV
The sensitivity index of the ESV in the middle and lower reaches of the Shiyang River remains less than 1(table 5), signifying the credibility of our results.The magnitude of the CS index follows a specific sequence: grassland > farmland > unused land > wetland > forest land > waters > construction-use land > mining land.Throughout the period from 1995 to 2020, there were large shifts in the CS values.Specifically, the CS of farmland increased, indicating an increased impact of its ESV coefficient on the overall ESV.In contrast, the CSs of grassland, unused land, wetland, and forest land all decreased during this period, suggesting a diminished influence of ESVs associated with these land types.Furthermore, it is noteworthy that the CSs of construction-use land, mining land, and waters all remained consistently low, each registering at less than 0.015.This observation implies that the changes in land use within these three categories had minimal effects on the overall ESV in the middle and lower reaches of the Shiyang River.

Ecosystem cross-sensitivity analysis
Our findings highlighted the CICS between farmland, grassland, and unused land, demonstrating the level of sensitivity of the total ESV to transformations among these land types.The highest CICS was observed among these categories, signifying those alterations between farmland, grassland, and unused land had the most substantial impact on the overall ESVs (figure 6).Specifically, the maximum CICS values were recorded for the transitions between farmland and grassland, farmland and unused land, and grassland and unused land, registering at 1.1, 1.18, and 1.54, respectively.For instance, the peak CICS for the transition from farmland to grassland occurred between 2010 and 2015, reflecting an increase in grassland area due to the implementation of measures to convert farmland to grassland, consequently enhancing the ESV.Similarly, the maximum CICS for farmland to unused land conversions was observed between 2000 and 2005, driven by a substantial demand for farmland and significant conversions from unused land to farmland, resulting in ESV growth.In addition, the  peak CICS for grassland to unused land transformations occurred between 2005 and 2010, primarily attributed to relatively stable local land class changes, resulting in minor ESV fluctuations.Specifically, from 2015 to 2020, the maximum CICS was associated with the transformation from unused land to wetland (figure 6(e)).This was due to the significant difference in the ESV coefficients between unused land and wetland, making ESV highly sensitive to the conversion from unused land to wetland despite small ESV changes.Furthermore, forest land, farmland, grassland, and unused land displayed comparatively higher CICS values compared to other land types, indicating that changes in forest land area also had a significant impact on ESV variations.In contrast, the absolute CICS values for farmland, construction-use land, forest land, grassland, industrial and mining land, and water area transformations were minimal.This indicates that the ESV exhibited lower sensitivity to conversions involving these specific land types (figure 6).

Land use change
The middle and lower reaches of the Shiyang River Basin, situated within the arid expanse of northwest China, experience challenging environmental conditions characterized by scarce precipitation, substantial evaporation rates, and scarcity of water resources.These environmental factors significantly influence land use, impacting the evolution and function of ecosystems (Xu et al 2020).From 1995 to 2020, an obvious pattern emerged in land use trends within this region, described as the 'three increases and three decreases.'Specifically, there was an increase in farmland, construction-use land, and mining land.At the same time, grassland, unused land, and forest land exhibited a decreased pattern, consistent with earlier research conducted by (Ngabire et al 2022, Wen et al 2023).Among the 'three increases,' the expansion of irrigated farmland primarily results from rapid population growth and accelerated agricultural economic development.In addition, during this research period, the process of urbanization intensified (Jia et al 2016) due to adjustments in industrial structures and enhanced economic growth.This led to an expansion in the area of construction-use land and mining land.In the 'three decrease' category, grassland experienced the most significant decrease in area, primarily due to the expansion of farmland (figure 3).Furthermore, unused land exhibited a continual reduction owing to the influence of the 'three increases' in land types.The decrease in ecological land, such as forestland and grassland, is a direct result of changing climatic conditions that have increased the proportion of dry ditches, wind-eroded land, and residual hills in the watershed, leading to a gradual decline in the proportion of ecological land.However, in recent years, the rate of reduction has gradually decelerated, owing to the positive impact of artificial grasslands in the watershed.This innovative practice has effectively crusted the soil, providing favorable conditions for vegetation growth in the middle and lower reaches of the river (Jun et al 2022).

ESV
Previous research in arid regions has highlighted the profound influence of land type changes on ecosystem service functions and subsequently on ESVs, including elements like the water cycle and carbon cycle (He et al 2021).The ESV of grassland in our study area ranked highest (table 3) for several reasons.Firstly, grassland occupied a substantial portion of the study area, even though it decreased in multiple land area transfers between 1995 and 2020.Despite these changes, grassland still covered 22.93% of the study area (table 2).Secondly, the ESV coefficient (VC) associated with grassland was significantly high (table 1).
In contrast, the ESV of forest land, wetland, and unused land showed an overall decreasing trend.This trend can be attributed to the rapid pace of social and economic development, leading to the transformation of these land types into farmland, construction-use land, and mining land.Consequently, the area dedicated to these land use types decreased during the study period, affecting ESV.It is worth noting that climate change can influence land habitat quality, alter the distribution of land species to some extent (Song et al 2021), and subsequently affect the land ESV.
On the other hand, the ESV of farmland generally exhibited a positive trend, consistent with findings from Jun (Jiang et al 2010, Jun et al 2022).The increase in farmland ESV can primarily be attributed to the conversion of forestland, grassland, and unused land into farmland.The expansion of farmland and construction-use land, along with the decrease in ecological land within the study area, reflects the outcomes of ecological governance policies implemented in the Shiyang River Basin since 2000.These policies, such as the reforestation of farmland and grassland and the reduction of farmland, have not achieved the desired results, as studied by the Sun (Sun et al 2018, Wang et al 2019).Unfortunately, these governance efforts could not achieve the proposed goals, resulting in further reductions in forested and grassland areas and the overall ESV.Furthermore, when examining the types of ESV functions (table 4), we observed that only the function related to food production increased during the period from 1995 to 2020, while the remaining functions experienced an overall decrease, reflecting the overexploitation of land due to the rapid growth of the agricultural economy.This overuse poses a potential threat to the stability and sustainable development of the ecological environment in the river basin.

Ecosystem sensitivity
Grassland, specifically CS grassland, emerged as the dominant land type, demonstrating the highest sensitivity to changes in the ESV of grassland.CS farmland and CS unused land also exhibited larger sensitivity, consistent with the findings of Liu (Liu et al 2014).Cross-sensitivity analysis of ecosystems reveals that transformations between farmland, grassland, and unused land exhibit the strongest sensitivities, which can be attributed to several factors.Firstly, these three land types collectively occupy a substantial portion of the study area, making them integral components of the research landscape.Secondly, over the study period, the per capita GDP increased from 44,000 CNY in 1995 to 39,000 CNY in 2020, with per capita farmland area increasing from 0.42 hm 2 /person in 1995 to 0.56 hm 2 /person.This signifies the impact of rapid social and economic development, which drove increased demand for farmland.(Sun and Li 2017).Consequently, the expansion of farmland largely encroached upon grassland and unused land, rendering the value of ESV most sensitive to changes in these land types.
The rapid agricultural and socio-economic development in the middle and lower reaches of the Shiyang River Basin intensified the circulation of different land groups (Wang et al 2012), leading to fluctuations in the value of regional ecosystem services (ESV).Sensitivity analysis of ESV identified farmland, grassland, and unused land as the primary influencing factors in the study area.The negative circulation of these three land types had a certain hindering effect on the increase of ESV growth, affecting the development of the regional ecological environment.In contrast, positive circulation of land types had the opposite effect, enhancing ESV and promoting sustainable development of regional ecological development.
Arunyawat's study further showed the importance of balanced socio-ecological development through appropriate land-use management (Arunyawat and Shrestha 2018).Therefore, river basin management departments should actively and comprehensively consider urban and rural integrated development, food security, and ecological security (Jin et al 2014) to ensure stability, diversity, and sustainability of regional ecosystem services.Given the current ecological environment status of the ecological environment in the middle and lower reaches of the Shiyang River Basin, it is suggested that when formulating policies in the future, we should not only prioritize relatively stable areas with high ESV like grassland and forest land but also appropriately reduce farmland which plays an important role in food production, while focusing on production efficiency.
Aligning with the goals of 'carbon peak and carbon neutrality,' future land-use policies should be scientifically formulated to strictly implement the basic farmland and ecological protection red lines.This includes careful planning for farmland development areas, improving construction-use land efficiency, and enhancing the protection of grassland, forest land, wetland, and other ecological land.Only through the establishment of a robust ecological development pattern, promotion of a virtuous cycle of ecosystems, and enhancement of ecosystem service functions can we achieve sustainable development of the socio-economic and ecological development in the middle and lower reaches of the Shiyang River Basin.
Nevertheless, it is important to acknowledge certain limitations in our land type classification of land types, which only divided the land into eight categories without considering habitat quality.Future research should account for secondary land classification, habitat quality, and the importance of ecosystem service functions to provide more precise reference points for relevant practices.

Conclusion
In the context of green and sustainable development of global ecosystems, quantifying the quantification of ecosystem service values (ESV) and the sensitivity of geotype changes proves beneficial for understanding the intricate relationship between land-use changes and ecosystem services.In this study, we quantified the sensitivity of ESV to ecosystems and ESV changes among various land types in the Shiyang River Basin in an arid region of northwest China over the study period.
Our analysis revealed a fluctuating trend in ESV-decreasing, increasing, decreasing, and increasing again.This pattern suggests that the regional ecological environment has the possibility of gradually evolved toward a more favorable trend.However, it is crucial to note that ESV decreased from 7,133.2 million CNY in 1995 to 7,008.3 million CNY in 2020.This signifies that although the regional ecosystem has improved during this period, it has not yet fully recovered to its previous level.Therefore, future efforts should focus on strengthening ecological protection.Furthermore, during our study period, we found that the ESV was most sensitive to transformations between farmland, grassland, and unused land.In contrast, it exhibited lower sensitivity to transformations involving farmland, construction-use land, forest land, grassland, mining land and waters.
Based on these findings, we recommend several key actions for the future.Firstly, we suggest that the expansion of farmland should be carefully controlled to maintain a sustainable balance.Secondly, there should be a concerted effort to enhance the efficiency of construction-use land and mining land.Lastly, the protection of ecological land, such as grassland, forest land, and wetland, should be prioritized to ensure the long-term health and stability of the regional ecosystem.
Overall, our study highlights the crucial need for ongoing monitoring and protection of ecosystem services in the Shiyang River Basin and similar ecologically challenged areas.By implementing our recommendations into action and promoting sustainable land-use practices, we can work towards a harmonious balance between socio-economic development and environmental preservation.

Figure 1 .
Figure 1.Geographical location of the middle and lower reaches of the Shiyang River.

Figure 2 .
Figure 2. Land use transfer in the middle and lower reaches of the Shiyang River basin from 1995 to 2020.

Figure 3 .
Figure 3. Sanji diagram of land use transfer in the middle and lower reaches of the Shiyang River Basin from 1995 to 2020.

Figure 4 .
Figure 4.The trend of the total ESV value.

Figure 5 .
Figure 5. Spatial Pattern of ESV per Unit Area in the Middle and Lower Reaches of the Shiyang River Basin from 1995 to 2020.

Figure 6 .
Figure 6.Cross-Sensitivity System of Ecosystem Services for Land Use Change in the Middle and Lower Reaches of the Shiyang River.

Table 1 .
Value coefficient of ecosystem services in the middle and lower reaches of Shiyang River (CNY•hm −2 •a −1 ).
H Tao et alTable2.Land use types in the middle and lower reaches of the Shiyang RiverBasin (1995Basin ( -2020)).

Table 3 .
ESV by land use type in the middle and lower reaches of the Shiyang River Basin, 1995-2020 (10 8 CNY).

Table 4 .
Composition of ESV in the middle and lower reaches of the Shiyang River Basin, 1995-2020 (10 8 CNY).

Table 5 .
Value sensitivity of ecosystem services in the middle and lower reaches of the Shiyang River Basin, 1995-2020.