The Assessment of Biodiversity Changes and Sustainable Agricultural Development in The Beijing-Tianjin-Hebei Region of China

Abstract

In the face of a series of challenges, such as climate change, population growth, and agricultural intensification, as well as the issue of how to promote sustainable development and guarantee food security, biodiversity, with its unique genetic, ecological, and traditional socio-cultural values, has become an important way to solve this dilemma. Urban biodiversity has continued to decline in recent decades due to rapid urbanization. The agroecosystem health of the Beijing-Tianjin-Hebei region, a typical urban agglomeration economic area, is facing a critical situation. Therefore, assessing the potential of ecosystem diversity in the Beijing-Tianjin-Hebei region and exploring the assessment mechanisms and methods of ecosystem health can provide theoretical support for biodiversity conservation and utilization. In this thesis, the overall ecosystem health of the Beijing-Tianjin-Hebei region was assessed based on the land cover data from 1992 to 2022 and the projected land cover data up to 2032, as well as using the habitat quality indicated by the Fragstats and InVEST models and the landscape pattern index, habitat quality, and mean species abundance (MSA) indicators of the GLOBIO module. The main results are as follows: Habitat quality and mean species abundance (MSA) in the Beijing-Tianjin-Hebei region were observed to show a continuous downward trend over 40 years from a landscape level perspective, and landscape fragmentation due to urbanization was the main reason. Habitat loss and habitat degradation caused by landscape fragmentation led to a decline in biodiversity. The spatial distribution of habitat quality in the Beijing-Tianjin-Hebei region is closely correlated with topography and landscape, being higher in the northwest and lower in the southeast, forming a clear spatial pattern that declined from 0.599 to 0.564 between 1992 and 2032. The mean species richness (MSA) value of the Beijing-Tianjin-Hebei region is significantly affected by infrastructure, especially road construction. With the continuous expansion of the road network, the MSA values in the region generally show a decreasing trend from 0.270 to 0.183 between 1992 and 2032. Based on the above results, it is recommended to carry out several aspects of agrobiodiversity conservation and ecosystem restoration.

1. Introduction

Globally, rapid urban growth is destroying areas of biodiversity, while countries with lower levels of political stability and regulatory quality will experience more rapid declines in biodiversity [1]. The Beijing-Tianjin-Hebei region, situated at the core of China’s Bohai Rim, encompasses the Beijing and Tianjin municipalities and the Hebei province [2,3]. This urban agglomeration, also known as China’s “capital city economic circle”, faces challenges due to inherent limitations in natural resources. The intensive development and exploitation activities in recent years have further depleted these resources, resulting in intense socio-economic conflicts and a decline in biodiversity, leading to a reduction in ecosystem function [4]. The Beijing-Tianjin-Hebei region is firmly committed to closely adhering to national policy, as evidenced by the recognition of the importance of ecological health in promoting green development within the three draft proposals of the 14th Five-Year Plan. As part of the coordinated development efforts in progress in Beijing-Tianjin-Hebei, the region is preparing for a significant expansion of urban construction land, which will inevitably lead to a reduction in farmland [5,6]. The challenges facing ecosystems and biodiversity underscore the need to reinforce conservation efforts, implement ecological control measures, and promote sustainable development.

Over the past 40 years, the Chinese government’s measures have played a pivotal role. The evolution of China’s environmental protection measures can be divided into four main stages: Before the 21st century, environmental protection efforts began with the Chinese government on Environmental Protection in 1973. This period saw the establishment of environmental protection as a core state policy and the introduction of China’s initial environmental protection laws, including significant projects like converting farmland to forests and grasslands and safeguarding natural forests [7].

From 2002 to 2012, the 16th National Congress of government introduced new environmental strategies. These included the promotion of sustainable development, the construction of a harmonious socialist society, and an emphasis on environmental protection as a key livelihood issue [8]. Important measures were taken to reduce major pollutant emissions, strengthen pollution control in critical regions, and enhance environmental law enforcement [9].

The 18th Government National Congress in 2012 marked the integration of ecological civilization into China’s socialist agenda, underscoring the construction of ecological civilization as a strategic decision to address environmental challenges [10].

In 2022, President Xi Jinping’s report at the 20th Party Congress focused on strengthening ecosystem diversity, stability, and sustainability, differentiating ecosystem diversity from biodiversity for the first time. This sets the direction for future research on agroecosystem diversity, emphasizing habitat changes due to agricultural activities and conservation efforts for species diversity.

These stages reflect China’s evolving approach to environmental protection, from foundational legal frameworks to integrating ecological considerations into national development strategies. This study aimed to address two key research questions:

Human-induced changes in land use, including alterations in land use, infrastructure development, and intensive agricultural development, negatively affect regional biodiversity. Mean species abundance (MSA) values are used to capture the causal links between these human activities and biodiversity, providing a more precise measurement than traditional methods. Most contemporary international assessments of regional biodiversity rely solely on basic and imprecise measurements of habitat quality. In contrast, in this study, mean species abundance (MSA) values were used to explore regional biodiversity, capturing the causal link between five human-induced land use alterations and biodiversity. The MSA provides an overall estimate of species abundance for the original reference condition.

The balance between urban development and environmental conservation in the Beijing-Tianjin-Hebei region is a critical issue. This question explores the policy trade-offs and resource competitions within the region, highlighting the importance of sustainable policies that promote both urban development and environmental preservation.

This study sought to quantify the impacts of individual and synergistic effects on biodiversity resulting from changes in land use, infrastructure development, and intensive agricultural development. Determining the precise effects of each factor, as well as the combined effects, on biodiversity is crucial. This research provides vital information to assist in mitigating any negative impacts on biodiversity and promoting its conservation. Additionally, this study can aid in the development of effective policies for sustainable land use, infrastructure development, and agriculture to benefit both humans and the environment.

The Beijing-Tianjin-Hebei region was selected for this study because of its abundant urbanization and its status as China’s national capital region. However, Tianjin and Hebei provinces, which comprise the Beijing periphery, remain understudied and are often overshadowed by the capital. Consequently, there are differences and resource competition between the cities. Moreover, a policy trade-off exists between environmental preservation and urban development [11]. Furthermore, the policy dilemmas faced in this region, which involve balancing ecological conservation and urban expansion, are representative of wider international challenges. Therefore, this study’s findings on managing these trade-offs provide valuable lessons for global urban areas dealing with comparable conflicts between development and environmental stewardship. This study’s international significance is highlighted by presenting the Beijing-Tianjin-Hebei region as a case study for synergistic development strategies. This could inform policy decisions in urban areas worldwide, emphasizing the universal relevance of the findings. The purpose of this study is firstly to bridge the gap between assessing the changes in landscape patterns and biodiversity in the Beijing-Tianjin-Hebei region at the landscape level over the last 30 years (taking the evolution of environmental protection measures in China as the time point) and predicting the trends of biodiversity changes in the Beijing-Tianjin-Hebei region over the next decade. On the basis of the results, relevant measures are proposed to enhance the use of biodiversity for the sustainable development of agriculture in the Beijing-Tianjin-Hebei region and to ensure food and ecological security.

2. Study Area

The Beijing-Tianjin-Hebei area, situated between 36°01′ N and 42°37′ N latitude, and 113°04′ E and 119°53′ E longitude, covers an area of over 216,000 square kilometres and encompasses diverse geomorphological features, including the Dashang Plateau, Yanshan Mountains, Taihang Mountains to the west, foothills, low plains, and coastal plains in the central and southeastern areas (see Figure 1). The Beijing-Tianjin-Hebei region has a predominantly temperate monsoon climate (Köppen classification Dwa) [12]. This climate is characterized by hot, rainy summers and cold, dry winters. The region has a wide range of elevation changes, from the plains along the coast to the mountains inland. The western and northern parts of Beijing are mountainous and can reach altitudes of more than 2000 m, while Tianjin and much of Hebei are flatter and generally less than 100 m above sea level [13]. The Beijing-Tianjin-Hebei region has a diverse range of vegetation, including temperate broadleaf and mixed coniferous forests in the north and grasslands in the south [14]. The region’s forests are home to a variety of trees, such as larch, birch, and oak, which contribute to its rich biodiversity. The area’s major soil types, such as windswept sandy soils, brown loam, sandy soils, and loess soils, reflect the complex topography and climatic conditions of the region [15]. The region’s land use types present a land cover pattern of transitions from coastal wetlands to cropland, forest area, and grassland, with the transition occurring from southeast to northwest [16]. As of 2021, the region has a total population of approximately 110.1 million inhabitants, representing 7.8% of China’s population. Additionally, the gross regional product (GRP) is valued at CNY 9635.59 billion (approximately USD 1339 billion), accounting for 8.5% of the country’s total GDP, as per the China Statistical Yearbook, 2022 [17]. Notably, the Beijing-Tianjin-Hebei Cooperative Development Strategy, also known as the Beijing-Tianjin-Hebei Strategy, was officially proposed as a national strategy in 2014. The primary objective of this strategy is to create an integrated Beijing-Tianjin-Hebei region characterized by improved economic growth, a more rational industrial structure, and a healthier ecological environment [18]. Consequently, the Beijing-Tianjin-Hebei Strategy aims to effectively address the current challenges encountered by the Beijing-Tianjin-Hebei region.

3. Materials and Methods

3.1. The Software Used and Data Sources

Fragstats 4.2, InVEST (InVEST3.9)’s Habitat Quality Plate, and the GLOBIO Plate are all Geographic Information System (GIS 10.8)-based applications used in this article that provide powerful decision support by analysing the effects of surface cover type, land use change, and human activities on habitat quality and biodiversity. Fragstats is a landscape structure quantification software program that assesses the impact of human activities on ecosystems by analysing indicators such as landscape fragmentation, patch size, shape, diversity, and connectivity [19]. These indicators help to understand the spatial distribution of biodiversity and potential threats. InVEST (Integrated Valuation of Ecosystem Services and Trade-offs) is a suite of tools for assessing the value of natural ecosystem services [20]. Its Habitat Quality module focuses specifically on the spatial distribution of ecosystem services and assesses the potential impacts of different management decisions on habitat quality. GLOBIO (Global Biodiversity Model for Policy Support) is a global biodiversity model designed to assess the impacts of land use change, infrastructure development, and climate change on biodiversity [21]. GLOBIO uses GIS data to simulate anthropogenic pressures on biodiversity and to predict future trends in biodiversity.

The data needed in this study include land use data, vegetation data, digital elevation model data, socio-economic data, and basic geographic information data. The sources of specific datasets and pre-processing are as follows:

(1)

Land use data: The land use data of Beijing-Tianjin-Hebei in 1992, 2002, 2012, and 2022 used in this study are from the year-by-year land use classification dataset of China from 1990 to 2022 at a resolution of 30 m, and the data were obtained from the National Glacial and Permafrost Desert Science Data Centre (http://www.ncdc.ac.cn (accessed on 23 June 2023)). In this study, the land use/land cover (LULC) types in Beijing-Tianjin-Hebei were reclassified into 13 land classes according to the research needs, which were construction land, unused land, paddy land, dry land, deciduous broadleaf forest, evergreen broadleaf forest, evergreen coniferous forest, deciduous coniferous forest, natural pasture, mixed forest, man-made farm, shrub from, and water.

(2)

Vegetation data: The normalized difference vegetation index (NDVI) was mainly used in this study. The NDVI for the Beijing-Tianjin-Hebei region is the annual maximum NDVI dataset at 30 m resolution for 1986–2023 in China, which was obtained from the Geo-Remote Sensing Ecological Network (GREN) Scientific Data Registration and Publication System (www.gisrs.cn (accessed on 16 June 2023)).

(3)

Digital Elevation Model (DEM) data: The digital elevation model (DEM) data used in this study were obtained from the Geospatial Data Cloud Platform (http://www.gscloud.cn (accessed on 16 February 2023)), ASTER GDEM, with a resolution of 30 m. The DEM data of the study area were obtained through splicing, projection conversion, cropping, and other processing, in which the slope data were based on the DEM data using the ArcGIS 10.8 software Slope tool.

(4)

Socio-economic data: The 2012 and 2022 POP and GDP spatial distribution kilometre grid datasets used in this study were obtained from the Resource Environment and Science Data Centre of the Chinese Academy of Sciences (http://www.resdc.cn/ (accessed on 17 June 2023)).

(5)

Basic geographic information data: The highway, railway, and river data of the Beijing-Tianjin-Hebei region used in this study were obtained from OSM (Open Street Map, https://openmaptiles.org/languages/zh/#0.33/0/0 (accessed on 16 February 2023)) and calculated with the help of the Euclidean distance method of ArcGIS10.8; the data raster size used in this study is 30 m × 30 m and the reference coordinate system is WGS_1984.

3.2. Fragstats

Based on the land cover data of Beijing-Tianjin-Hebei in 1992, 2002, 2012, and 2022, and the predicted data for 2032, the indices of the landscape level of the Fragstats (v4.2) software were suitable for comparative analyses of the landscape pattern and the change of features in different time phases, in order to reflect the characteristics of the total landscape pattern of the Beijing-Tianjin-Hebei region. The correlation between landscape pattern indices obtained from all models and habitat quality was then analysed by IBM SPSS Statistics 20. This analysis aims to investigate the correlation between landscape type and the level of anthropogenic disturbance. The correct quantification of landscape heterogeneity requires the help of landscape indices [21,22]. These indicators describe landscape composition (diversity and richness of patch types) and configuration (size, shape, distribution of patches, and their distance from each other). Landscape indices are essential for understanding the health of ecosystems, the distribution of biodiversity, and how ecological processes change at different spatial scales. The land cover data of the region were resampled at a resolution of 90 using ArcGIS, and a network analysis method based on Fragstats 4.2 software was used to derive the landscape pattern index analyses of the Beijing-Tianjin-Hebei region for the years 1992, 2002, 2012, 2022, and 2032. The main objectives of resampling the land cover data to 90 m resolution using ArcGIS were to ensure consistency and compatibility between different data sources, improve analysis efficiency, meet the accuracy requirements of specific ecological process analyses, optimize the allocation of computational resources, and adapt to a variety of analytical objectives so as to take into account both detail retention and computational efficiency when dealing with large-scale landscape analyses. Subsequently, five landscape indices (

Table 1. Descriptions of landscape pattern metrics used in this study [6,23,24].

Index	Expression	Ecological Meaning
Edge density (ED)	$\mathrm{ED}=\mathrm{E}/\mathrm{A}$	Describes the ratio of the total length of edges in a patch to the area of the landscape.
Landscape division index (DIVISION)	$\mathrm{DIVISION}=\left[1-{{\displaystyle \sum }}_{i=1}^m{{\displaystyle \sum }}_{j=1}^n{\left[\frac{\mathrm{a}ij}{A}\right]}^2\right]$	Describes the degree of dispersion of spatial distribution between individuals of different patches in the landscape; the larger the value, the more dispersed the patches are and the higher the degree of fragmentation.
Largest patch index (LPI)	$LPI=\frac{max{a}_{jj}}{A}\times 100.$	Describes the dominance of the landscape.
Shannon’s evenness index (SHEI)	$\begin{array}{c}\mathrm{SHEI}=\\ \frac{-{{\displaystyle \sum }}_{i=1}^m\left(P_{ix}\ln P_i\right)}{1_{nm}}\end{array}$	Describes the degree of patch area uniformity used to characterize different types of patches in the landscape.
Shannon’s diversity index (SHDI)	$SHDI=-{{\displaystyle \sum }}_{i=1}^m[\begin{array}{c}{P}_i\times ln{P}_i\end{array}].$	Describes the diversity of patches.

) were used for network analysis. Among them, the increase in ED, DIVISION, SHEI, and SHDI and the decrease in LPI represented the deepening of landscape fragmentation in the region, and vice versa represented the alleviation of landscape fragmentation. Landscape fragmentation leads to a reduction in the total area of the original habitat, resulting in isolated heterogeneous populations, and the process of habitat fragmentation causes a series of changes in biotic and abiotic conditions in the habitat, leading to a decline in habitat quality and biodiversity. Secondly, using the SPSS platform, we effectively processed and analysed a large amount of data in order to explore the statistical relationship between non-landscape pattern indices (ED, DIVISION, LPI, SHEI, and SHDI) and habitat quality. The research methodology included data collection and pre-processing, including the integration of landscape pattern data and habitat quality assessment data from the Beijing-Tianjin-Hebei region into the SPSS platform. Statistical analysis tools in SPSS, including descriptive statistical analyses, correlation analyses, and regression analyses, were used to quantitatively assess the relationship between landscape pattern indices and habitat quality. Correlation analysis helped to determine whether there was a statistically significant linear relationship between variables and the strength of that relationship. This analysis identifies landscape pattern features that are positively or negatively associated with high habitat quality. For example, a positive correlation between a particular pattern index and habitat quality indicates that an increase in that index is accompanied by an increase in habitat quality and vice versa.

3.3. Markov Chain Model

This study is based on land use data, digital elevation model data, and socio-economic data (POP, GDP spatial distribution) of Beijing-Tianjin-Hebei in 1992, 2002, 2012, and 2022. The land use data for 2032 under the natural development scenario were predicted and simulated using the CA-Markov model. The land use data of the Beijing-Tianjin-Hebei region in 2012 and 2022 were analysed using superposition analysis to derive the area transfer matrix of land use types during this period, which was used as the transfer probability matrix. The land use data in 2022 were used as the initial state matrix, and the land use data of the region in 2032 were predicted by simulation prediction under the natural development scenario. Predicting land use changes using Markov models can provide the InVEST model with accurate dynamic predictions of future land use scenarios, thus improving the ability to assess the value of ecosystem services and their changes. This approach improves prediction accuracy, supports multi-scenario analyses, enhances model applicability, facilitates integrated management and planning, and provides important support for sustainable development and ecosystem management. Since the Markov chain is a process without a posteriori effects, which is in line with the characteristics of land use change, the Markov model was used in this study to predict land use demand [15]. The transition probability matrix was calculated using the Markov module in the IDRISI 17.0 software package (Clark Labs, Worcester, MA, USA), and the user’s guide for the software was used as a reference for detailed procedures [25]. Hence, this paper does not delve into the specific procedures and mathematical expressions involved in this process. This approach allowed for the prediction of future land use dynamics in the Beijing-Tianjin-Hebei region, providing valuable insights for land use planning and management in the area.

3.4. Habitat Quality of the InVEST Model

The InVEST^® model (https://naturalcapitalproject.stanford.edu/software/invest (accessed on accessed on 16 June 2023)), an open-source software developed by Stanford University, Minnesota University, the World Wildlife Fund, and the Nature Conservancy, is used to visualize and estimate goods and services from nature on a spatial scale [26]. In this study, the Habitat Quality module of the InVEST model was employed to assess the quality of habitats [27]. By observing differences in land usage or cover, the assessment of ecosystem services generated by the environment can be appreciated [28]. The supply of ecosystem services and the value of the landscape are calculated and expressed, and the methods and trends of landscape change are predicted.

The methods and trends for the calculation and expression of ecosystem service supply and landscape value, and the prediction of landscape change, can facilitate the evaluation of biodiversity value and change trends within the landscape [28]. The assessment of the landscape’s habitat quality involves inputting data such as LULC (land use/land cover) maps, LULC sensitivity toward each threat, spatial data on the distribution and density of each threat, and the spatial location of protected areas. These inputs are used to determine the GLOBIO section of the InVEST model for the Beijing-Tianjin-Hebei region across various time points across 1992, 2002, 2012, 2022, and 2032 [2,29]. The ecosystem services produced by the environment can be assessed by looking at differences in land use or vegetation cover, while biodiversity is closely linked to the production of ecosystem services. Patterns of biodiversity are inherently spatial and can therefore be estimated by analysing land use and land cover (LULC) maps and threats to species’ habitats. The model assesses biodiversity in the study area by modelling habitat quality and rarity as proxies for biodiversity The formula for computing habitat quality (HQ) is as follows:

$$Q_{xj}=H_j[1-(\frac{D_{xj}^2}{D_{xj}^2+k^z})]$$

where Q_xj represents the HQ of pixel x in habitat type j, H_j denotes the habitat suitability score for habitat type j, and D_xj is the weighted average of all threat levels of pixel x in habitat type j. Additionally, k represents a half-saturation constant (0.5 in this paper), and z denotes a constant (3.9 in InVEST). The HQ ranges from 0 to 1, with values close to 1 indicating high regional biodiversity [30]. This methodology provides valuable insights into the habitat quality and biodiversity trends in the Beijing-Tianjin-Hebei region.

In this study, cultivated land, constructed land, unused land, and road threat factors were identified as habitat threat factors in the land classification system. The maximum impact distances of the threat factors, along with their respective weights and the degrees of habitat sensitivity to these factors, were determined based on previous studies [31,32,33], the model guidebook, and the characteristics of the study area (Table 1 and

Table 2. Sensitivity of habitat types to each threat.

Habitat Type	Habitat Suitability	Cultivated Land	Construction Land	Unused Land	Railroad	Freeway	First Grade Highway	Second Grade Highway
Cultivated land	0.4	0	0.7	0.2	0.1	0.2	0.1	0.1
Forest	1	0.6	0.8	0.3	0.6	0.7	0.6	0.4
Grassland	0.9	0.5	0.6	0.4	0.3	0.5	0.4	0.2
Water	0.75	0.5	0.65	0.3	0.5	0.6		0.4
Unused land	0.1	0.1	0.2	0	0.1	0.2	0.1	0.1
Construction land	0	0	0	0	0	0	0	0

). Technical abbreviations are explained when first used to ensure clarity and understanding for readers. These determinations provide a comprehensive understanding of the impact of various threat factors on habitat quality within the study area.

Using the SPSS platform, we efficiently processed and analysed a large amount of data to explore the statistical relationship between non-landscape pattern indices (ED, DIVISION, LPI, SHEI, and SHDI) and habitat quality. The research methodology involved data collection and pre-processing, including integrating landscape pattern data and habitat quality assessment data from the Beijing-Tianjin-Hebei region into the SPSS platform. Statistical analysis tools, including descriptive statistical analysis, correlation analysis, and regression analysis, were used in SPSS to quantitatively assess the relationship between landscape pattern indices and habitat quality. The correlation analysis helped to determine whether there was a statistically significant linear relationship between the variables and the strength of this relationship. This analysis identifies landscape pattern features that positively or negatively correlate with high habitat quality. For instance, a positive correlation between a specific pattern index and habitat quality indicates that an increase in that index is accompanied by an increase in habitat quality, and vice versa.

3.5. InVEST Model GLOBIO

The InVEST GLOBIO model is based on the mean species abundance (MSA), which is assigned values within a range of 0 to 1. The MSA represents the average proportional change in the abundance of individual species at a site compared to the original flora at the same site. The number of species in this context refers to the number of populations. Comparing MSA values across scenarios allows for an assessment of the impact of human-induced changes on biodiversity. The GLOBIO model measures human impacts on biodiversity, measured by mean species abundance (MSA). Mean species abundance is an improvement over the more traditional species–area curve approach for two reasons. First, it gives aggregate estimates of species densities, not just species presence, which is important for representing true diversity since presence alone gives limited information about population viability. Second, it relates more than habitat area to changes in biodiversity by including information about the impact of fragmentation and threats from infrastructure. The Habitat Quality module of InVEST and the GLOBIO module differ in their land use type requirements, mainly because of their different objectives, scales of analysis, and methodologies. InVEST focuses on the detailed assessment of biodiversity and habitat quality in localized areas and requires more detailed land use data. GLOBIO, on the other hand, targets biodiversity loss assessment at the global scale, using a wider range of land use classifications. This reflects the fundamental differences between the two in terms of the range of applications, accuracy requirements, and considerations of data availability and processing power, allowing each to provide appropriate analyses in different areas. Therefore, to ensure consistency, we reclassified the original land use/land cover (LU/LC) maps [34].

Stressors reduce the MSA in a multiplicative manner. In the InVEST GLOBIO model, stressors include land use/land cover (LU), excess atmospheric nitrogen (N) deposition, proximity to infrastructure (mainly roads; I), fragmentation (F), and climate change (CC); however, to adjust for land use change, the InVEST model omits the N deposition and climate change terms because these factors remain constant regardless of the land change scenario [21,35]. The GLIBIO section of the InVEST model was used to calculate the mean species abundance (MSA) for the Beijing-Tianjin-Hebei region in 1992, 2002, 2012, 2022, and 2032, and the MSA was calculated as follows:

$${MSA}_i=MS{A}_{L{U}_i}·MS{A}_{N_i}·MS{A}_{I_i}·MS{A}_{F_i}·MS{A}_{C{C}_i}$$

We assessed land cover changes, including fragmentation and infrastructure development, and their impacts on biodiversity in the Beijing-Tianjin-Hebei region in 1992, 2002, 2012, and 2022. Projections of the reclassified land use map for 2032 were made using the Markov model. We used previous land cover data, with MSALU values related to management intensity or human use intensity from the literature, the model guide, and the characteristics of the study area, with derived mean values for various land types. For example, the MSALU value for water was set to 0 because MSA values were calculated only for terrestrial ecosystems. Paddy fields were defined as those with “more intensive agriculture” due to their irrigation and frequent fertilization [28,34,36] (

Table 3. Threat data.

THREAT	MAX_DIST	WEIGHT	DECAY
Cultivated land	8	0.6	linear
Construction land	10	1	exponential
Unused land	6	0.3	linear
Railroad	6	0.5	linear
Freeway	8	0.6	linear
Roads1	7	0.5	linear
Roads2	5	0.4	linear

).

The normalized difference vegetation index (NDVI) was used to define pasture coverage, with values ranging between −0.20 and 0.92. We normalized the derived NDVI values between 0 and 1. The normalized NDVI subtracted from 1 was masked by the LU/LC grid for natural forests and plantations to determine the pasture area for InVEST GLOBIO. Pasture areas were defined as grasslands with a proportion of pasture greater than or equal to 50% [37,38].

We considered intensive agriculture as facility farming, and we calculated the proportion of intensive agriculture by dividing the area of facility farming in Beijing-Tianjin-Hebei by the area of arable land in Beijing-Tianjin-Hebei [39]. The proportion of intensive agriculture in 2002 was calculated as 0.04, and the proportions of intensive agriculture in 2012, 2022, and 2032 were calculated as 0.1, 0.2, and 0.3, respectively. The data required for the calculation of the proportion of intensive agriculture are based on the expected attainment targets of China’s 14th Five-Year Plan and annual statistical yearbooks and government publications for each province [17,24,40]. We did not calculate the proportion of intensive agriculture in the calculation of average species richness for 1992 and 2002 because the InVEST model sets the proportion of intensive agriculture to a minimum of 0.1. Moreover, in the 21st century, before the slow development of the Party Central Committee in 2004, China established the first facility agriculture since the reform and opening up on the “three issues of agriculture“ of the first document, and facility agriculture began to develop comprehensively thereafter [41].

The InVEST GLOBIO model is used to analyse forest fragmentation using the Forest Fragmentation Quality Index (FFQI). The FFQI is calculated by considering how many of a forest’s neighbouring cells are also forested. The FFQI estimates the relative effect of fragmentation with a Gaussian smoothing function. This treats habitat patches that are separated by only very small patches of infrastructure or non-habitat as less fragmented than habitat patches separated by wider distances (refer to the guide for specific MSA fragmentation effects at different patch sizes). The impact of the MS_AI depends on the calculation of the distance between ecosystems and anthropogenic categories, as detailed in the user’s guide [32]. Therefore, detailed procedures and mathematical expressions are not discussed in this paper (

Table 4. LDD LU/LC conversion to GLOBIO LU/LC and remaining MSAs affected by land use.

LDD LU/LC Class	GLOBIO LU/LC Classes	MSALU
Construction land	Built-up areas	0.5
Unused land	Primary vegetation	1
Paddy field	All agriculture (Intensive agriculture)	0.1
Dryland	All agriculture (Low-input agriculture)	0.3
Deciduous broadleaf forest	Primary vegetation	1
Evergreen broadleaf forest	Primary vegetation	1
Evergreen needleleaf forest	Primary vegetation	1
Deciduous needleleaf forest	Primary vegetation	1
Natural rangeland	Livestock grazing	0.7
Mixed forest	Secondary forest	0.5
Artificial pasture	Man-made pastures	0.1
Shrublands	Primary vegetation	1
Water	N/A	N/A

).

4. Results

4.1. Land Use Space and Landscape Pattern Changes

The analysis of changes in land use space and landscape patterns is aimed at studying the changes in land landscape types and understanding the status of the Beijing-Tianjin-Hebei ecological environment as well as biodiversity from a macroscopic point of view by means of the landscape indices of landscape composition (diversity and abundance of patch types) and configuration (size, shape, distribution of patches, and their distances from each other). Based on the land use data from Beijing-Tianjin-Hebei in 1992, 2002, 2012, and 2022, this study predicted and simulated the land use data in 2032 using the CA-Markov model under the natural development scenario (see Figure 2).

The southeastern plains of the Beijing-Tianjin-Hebei region mainly consist of arable land and both urban and rural land dedicated to construction. Construction land dominates the central areas of Beijing and Tianjin. The Yanshan Mountains in the northeast are characterized by forested and unused land, representing the concentrated distribution area of the forests in the region. The Bashang area is mainly composed of arable land and grassland, whereas the intermountainous basins in the northwest and the Taihangshan Mountains are characterized by arable and unused land.

The land use structure data indicate that between 1992 and 2032, arable land was the predominant land use in the Beijing-Tianjin-Hebei region, with forest and grassland following closely behind. During this period, there was a general decline in forestland and grassland, as well as the conversion of arable land to construction land and the conversion of water bodies to arable land. This has led to a reduction in landscape heterogeneity and an increase in fragmentation. The trends for all landscape indices are consistent between 1992 and 2032, indicating that the landscape pattern of the study area is undergoing directional transformation.

In 1992, agricultural land (arable is equated with agricultural land in this study) was dominant, with a relatively low proportion of construction land. During this time, the “33211” project was executed by the Communist Party of China’s Central Committee, which included a directive to manage all major pollutant emissions for the Bohai Sea area [42]. While land for construction declined in the Bohai Sea region, it increased in other ares. In the same year, the State Council issued the Opinions on Further Improving the Policies and Measures of Returning Cultivated Land to Forestry document, setting in motion the full-scale launch of the project to return cultivated land to forests.

From 2012 to 2022, the development of an ecological civilization was proposed as a crucial component of green development [43]. This period saw significant progress and discernible outcomes resulting from a series of consequential policies and measures formulated by the Party Central Committee and the State Council [44]. Premier Li Keqiang proposed the integration of the Beijing-Tianjin-Hebei region toward synergistic sustainable development during this time. Environmental protection policies led to an increase in forested land and the continuous promotion to return farmland to forests [44].

Certainly, the implementation of far-reaching land improvement policies has led to a surge in the efficient and cost-effective utilization of construction land in urban areas, resulting in positive development. However, despite the data being from 2022, the local ecology remains fragile, and issues of land degradation still persist. It is imperative to increase efforts to fortify ecological infrastructure and halt the trend of land degradation.

Notably, land management work is highly valued by both authorities in China. There is continuous improvement and refinement of land management laws and regulations, reflecting the commitment to addressing ecological concerns and sustainable land use practices.

The Bohai Rim region is considered the third growth pole of regional development in China, following the Pearl River Delta and Yangtze River Delta regions [45]. The rapid expansion of townships in the Beijing-Tianjin-Hebei region, a significant grain production area in North China, has resulted in an increase in urban land use and arable land protection [46]. However, this has also led to a decline in patch space connectivity and increased landscape fragmentation due to conflicting objectives.

Under the natural development scenario, the Markov transfer matrix provides a quantitative analysis of the recharge source and flow direction between various types of land use. It also measures the likelihood of land use type conversion and predicts a significant increase in construction land and a considerable decrease in arable land and water by 2032. This transformation indicates a shift from arable land to construction land and grassland, with a slight increase in forestland at the expense of arable land and grassland. Additionally, there has been an increase in grassland at the expense of forestland and arable land to some extent. These changes underscore the evolving landscape dynamics and the need for sustainable land use planning to mitigate ecological impacts.

The data in Table 2 reflect the landscape fragmentation within the Beijing-Tianjin-Hebei region, demonstrating a decreasing trend in patch numbers and changes in landscape dominance and diversity indices over time. This study analysed the correlation between landscape pattern and habitat quality using the SPSS platform. It examined the correlation between habitat quality and landscape pattern indices of 14049 grid cells and obtained statistical data for the years 1992, 2002, 2012, 2022, and 2032 in the Beijing-Tianjin-Hebei region (

Table 5. Landscape index changes for 1992, 2002, 2012, 2022, and 2032.

Year	Index	Mean	Standard Deviation	Correlation Coefficient	Significance
1992	ED	14.35938144	8.954525214	0.361 **	0.00
	DIVISION	0.264707311	0.136499919	0.263 **	0.00
	LPI	80.77653105	10.57623656	−0.282 **	0.00
	SHDI	0.391552075	0.19740407	0.238 **	0.00
	SHEI	0.466639465	0.193177471	0.160 **	0.00
2002	ED	26.77697357	14.53832237	0.435 **	0.00
	DIVISION	0.327418713	0.148856045	0.289 **	0.00
	LPI	76.93705207	11.43965846	−0.286 **	0.00
	SHDI	0.480559341	0.207449064	0.307 **	0.00
	SHEI	0.538625453	0.180570009	0.085 **	0.00
2012	ED	27.19431639	13.72559616	0.357 **	0.00
	DIVISION	0.335529341	0.142936431	0.189 **	0.00
	LPI	76.31449417	11.01591754	−0.185 **	0.00
	SHDI	0.491221455	0.200768398	0.220 **	0.00
	SHEI	0.55372941	0.177987466	−0.029 **	0.00
2022	ED	27.43567592	13.22122537	0.303 **	0.00
	DIVISION	0.343426979	0.140323342	0.138 **	0.00
	LPI	75.73118869	10.82098927	−0.134 **	0.00
	SHDI	0.503521907	0.199098943	0.188 **	0.00
	SHEI	0.562945778	0.175130039	−0.109 **	0.00
2032	ED	27.03093668	12.67919913	0.263 **	0.00
	DIVISION	0.342712214	0.139643878	0.096 **	0.00
	LPI	75.78397998	10.74031103	−0.089 **	0.00
	SHDI	0.502271679	0.199143176	0.161 **	0.00
	SHEI	0.567104514	0.180149811	−0.156 **	0.00

ED: edge density; DIVISION: landscape division index; LPI: largest patch; Index SHDI: Shannon’s evenness index; SHEI: Shannon’s diversity index. ** indicates a highly significant correlation.

). Table 5 demonstrates that the correlation between habitat quality and landscape pattern characteristics in the Beijing-Tianjin-Hebei region is consistently moderate or low. Only LPI exhibits a negative correlation with habitat quality, while the remaining indicators exhibit a positive correlation. All correlation analysis results are statistically significant. The landscape patch index (LPI) consistently decreased from 1992 to 2032, indicating the decreasing dominance of the largest landscape patches due to urban development expansion and policies converting farmland into forested areas. This decline underscores the impact of these policies on landscape composition [47,48].

The ED scores exhibited a rapid upward trend from 1992 to 2002, signifying the high level of human activity in these areas during this decade, resulting in higher density and fragmentation at the edge of the patches, which has remained stable since 2002 due to the demand for urban and rural planning and construction activities in the area [49]. The DIVISION scores showed a consistent upward trend, reflecting increased patch fragmentation between individuals of different patches, with increased separation and improved segmentation of each patch. The Shannon–Wiener diversity index (SHDI) initially increased between 1992 and 2012, indicating increased landscape fragmentation due to rapid urban construction [50].

The continuing upward trend in SHEI scores and SHDI scores suggests that land use is rich, fragmentation is high, and its unqualified information content is high. The area ratios of different patch types in the landscape are converging, reflecting a more balanced weighting of patch areas. This implies that the richness of patches in the study area is becoming more homogeneous. This change is attributed to the expansion of built-up land, especially the conversion of arable and unused land to built-up land.

Considering the present protection of basic farmland, the future distribution of habitats is expected to become more stable, leading to an increase in the diversity index and reduced regional fragmentation. These changes reflect the ongoing urbanization process in the Beijing-Tianjin-Hebei region and emphasize the need for sustainable land use planning to mitigate further landscape fragmentation [51].

4.2. Analysis of Spatial Changes in Habitat Quality

The habitat quality index values ranged from 0 to 1, with higher values indicating better habitat quality. These values were categorized into four zones: low quality (0–0.4), lower quality (0.4–0.75), higher quality (0.75–0.94), and high quality (0.94–1) [52]. The 1992–2032 habitat quality map of the Beijing-Tianjin-Hebei region displayed significant spatial heterogeneity, with an overall distribution of “low in the southeast, high in the northwest” (see Figure 3). The Yanshan and Taihang Mountains exhibited exceptionally high-quality habitats, while the southern and southeastern plains showed comparatively lower habitat quality [53]. The average habitat quality in the Beijing-Tianjin-Hebei region continues to show a decreasing trend from 1992 to 2032. The average habitat quality in the Beijing-Tianjin-Hebei region was about 0.599 in 1992, declined to 0.598 in 2002, 0.588 in 2012, decreased rapidly to 0.571 in 2022, and is predicted to decline further to 0.564 in mid-2032 projected under the natural scenario.

The Zhangbei Dashang Grassland in the northwestern region constituted a significant portion of the region with relatively higher habitat quality, while the Bohai Bay area experienced a shift from mostly lower habitat quality to areas of low habitat quality between 2012 and 2022. The spatial distribution of topography and geomorphology aligns with habitat quality values, with mountainous terrain in the northeastern and southwestern sectors exhibiting mostly high-quality habitats with minimal human impact. This is largely attributed to economic growth and the subsequent increase in the demand for land intended for infrastructure construction, which has resulted in habitat loss [54].

The conversion of farmland to forest since 2002 has influenced the dominance of woodland and grassland habitat types, promoting high-quality habitats. However, the expansion of construction land from 1992 to 2022, driven by economic growth and urban expansion, has had a coercive impact on adjacent cropland habitats, exacerbating the declining quality of the habitat.

Regions experiencing rapid economic development, such as the Tianjin Binhai New Area, have experienced a significant decline in habitat quality due to the demand for land for infrastructure construction, which has led to habitat loss. This decline is less pronounced in mountainous regions with relatively sluggish economic development. The predicted land cover data suggest a consistent decline in habitat quality in the future landscape of 2032 under the natural development scenario, with advancements in higher-, medium-, and lower-quality areas. The urban landscape has compressed and divided the surrounding habitats, resulting in decreased habitat stability and increased fragmentation in high-quality regions.

4.3. Analysis of Spatial Changes in MSA Values

The mean species abundance (MSA) values from 1992 to 2032 indicate a decline in undisturbed areas across all ecosystems in the Beijing-Tianjin-Hebei region, driven by the increasing demand for food, energy, timber, and infrastructure due to urbanization (see Figure 4). Climate change, nitrogen deposition, fragmentation, and infrastructure are expected to intensify the effects of these factors on biodiversity [21,34].

The expansion of road networks has had the most significant impact on MSA values in the region, with all MSA values declining as roads expand. The grassland regions in the northwest showed moderate MSA values. In contrast, the woodland areas along the Yanshan and Taihang Mountains, positioned at higher elevations and less accessible, exhibited higher MSA values. These areas preserve biodiversity effectively due to their steep topography and the absence of human settlements and agricultural development.

The average MSA value decreased from approximately 0.270 in 1992 to 0.237 in 2002, representing a decrease of approximately 12.2%. Subsequently, the average MSA values for 2012 and 2022 were 0.243 and 0.223, respectively. Assuming natural development, MSA is projected to decrease to 0.183 by 2032. At the same time, the standard deviation of species richness (MSA) has been on a decreasing trend, which represents a continuous reduction in the heterogeneity of species richness.(See

Table 6. Analysis of species richness (MSA) in Beijing-Tianjin-Hebei from 1992 to 2032.

Year	MSAmax	MSAmean	MSAstddev
1992	1	0.270	0.300
2002	1	0.237	0.278
2012	1	0.243	0.291
2022	1	0.223	0.276
2032	1	0.183	0.265

)

The considerable decrease in the MSA value from 1992 to 2002 was related to urbanization and the increase in intensive agriculture. However, a series of environmental protection measures resulted in a reduced average MSA value, contributing to the preservation and protection of biodiversity.

5. Discussion

(1)

In this study, the methodology of biodiversity assessment has been explored in depth: Firstly, the Beijing-Tianjin-Hebei region, as a representative urban agglomeration of the Capital Economic Circle, can be a good place to explore the impact of urbanization development and environmental protection measures on biodiversity. Focusing on the assessment of biodiversity in the Beijing-Tianjin-Hebei region using the InVEST model, and in particular its GLOBIO component, this study fills an important gap in global biodiversity assessment practice. By providing clear results and selecting specific policy milestones to disaggregate and discuss observed changes, this study not only enhances our understanding of the impacts of policy on agrobiodiversity conservation, but also illustrates the critical role of targeted policy interventions. This approach provides new perspectives for integrating ecological modelling and policy analysis to address global biodiversity challenges, emphasizing the important contribution of informed decision-making to biodiversity conservation efforts.

This study further analysed the relationship between habitat quality and landscape pattern characteristics in the Beijing-Tianjin-Hebei region. The results showed that such correlations generally existed at moderate or low levels, indicating complex interactions between landscape patterns and habitat quality in this region. A positive correlation was found between most landscape pattern indicators and habitat quality. This suggests that habitat quality generally improved as the values of these indicators increased. However, it should be noted that a negative correlation was found between LPI and habitat quality. This suggests that larger continuous habitat patches may not necessarily lead to an overall improvement in habitat quality in the Beijing-Tianjin-Hebei region. This could be because large patches reduce habitat diversity or increase the vulnerability of habitat edges. All correlation analyses yielded statistically significant results, indicating a coupled relationship between the landscape pattern index and the habitat quality index. This statistical significance not only confirms the existence of interactions between these variables but also provides a solid foundation for further research on how landscape pattern affects habitat quality. Through these analyses, we can gain a better understanding and assessment of how human activities and natural processes interact to affect the structure and function of regional ecosystems, and how these effects, in turn, impact biodiversity and habitat stability.

(2)

Comparison of the findings of this study with those of similar studies: Firstly, the results of the habitat module runs in the InVEST model are broadly consistent with the trends of similar studies, with a continuing downward trend in ecosystem health and biodiversity over the last 30 years. Most studies on biodiversity assessment have indeed used habitat quality indices rather than biodiversity indices, as seen in the literature [51,54,55], with a predominant focus on habitat area within the InVEST-HQ model. Importantly, the assumption in this study was that all indigenous species and habitats within the study area were impacted by these threats while potentially ignoring threats from other biological factors. This study of ecosystem diversity in the Beijing-Tianjin-Hebei region explored how landscape area factors such as habitat quality and landscape pattern indices affect ecosystem diversity. Additionally, this study considered the average species richness, which integrates the impacts of different levels or intensities of each pressure on biodiversity to assess biodiversity at the site. This approach ensures objectivity and clarity in biodiversity assessment. In the future, it may be worthwhile to consider selecting data for the GLBIO section to update land cover data in greater detail, which would aid in the assessment of agrobiodiversity.

(3)

Policy recommendations related to enhancing the use of agricultural biodiversity to promote sustainable agricultural development in Beijing, Tianjin, and Hebei, and to ensure food and ecological security. Firstly, an evolutionary outlook for landscape fragmentation regional planning: The expansion of urbanization and construction in the Beijing-Tianjin-Hebei region, along with the increase in the proportion of intensive agriculture and the implementation of economic collaboration policies, has led to a reduction in landscape heterogeneity and an amplification of fragmentation. This has resulted in severe damage to biodiversity, especially in regions strategically significant for social and economic development. However, the shift from an agriculture-based economy and recent policies promoting ecological civilization, as well as the integration of sustainable development in the Beijing-Tianjin-Hebei region, have helped mitigate pressure on biodiversity in the study area. The aim is to balance economic and social development with biodiversity conservation. Conservation programs under the current policy focus on preserving forest coverage, designating additional protected areas, and rehabilitating degraded regions. However, the dilemma between social and economic development and environmental protection should be considered, and public awareness of sustainable development and market-based conservation programs can increase.

Problems posed by declining habitat quality and mean species abundance (MSA): By comparing the application of the InVEST model and its GLOBIO panel in the Beijing-Tianjin-Hebei region with other urban economic zones around the globe, common challenges such as ecological vulnerability and land degradation, as well as opportunities for advancing biodiversity conservation and sustainable agricultural practices, were identified. This study highlights the importance of sustainable land use and ecological improvement in the region, as a decrease in habitat quality also represents an increase in individual ecological threat factors. The future is bound to create many challenges in enhancing land use and ecological improvement. Comprehensive measures are needed in future development to counteract habitat reduction, promote regional spatial redevelopment, introduce green agricultural techniques, and promote achievable urban growth. Future strategies should focus on restoring contaminated agricultural land, improving the quality of arable land, transitioning from monoculture to mixed cropping, and progressively mainstreaming biodiversity conservation. Together, these efforts will pave the way for a more ecologically resilient and environmentally sustainable development model that can be adapted and implemented in different environments across the globe. In our exploration of habitat quality in the Beijing-Tianjin-Hebei region, the range of habitat and vegetation types across the landscape and their degraded state were observed, and biodiversity was tangentially related to ecosystem service functions. Our discussion not only focuses on quantitative changes in habitat quality and species richness, but also takes into account the complex interactions between biodiversity conservation and the maintenance of local traditional culture. The decrease in average species richness represents a decline in local treasured plants and animals, and the enhancement of germplasm conservation is recommended in future biodiversity conservation measures. Innovations in ecotourism and environmental safety and security, strengthened socio-economic infrastructure, and more effective natural resource management should be promoted.

6. Conclusions

This chapter uses a combination of Fragstats software, the InVEST model, and the GLOBIO method for an in-depth analysis of the ecological and environmental changes in the Beijing-Tianjin-Hebei region, specifically including changes in landscape patterns, habitat quality, and species diversity. A general assessment of the ecological and environmental conditions in the Beijing-Tianjin-Hebei region was given from a landscape level perspective, specifically the following:

(1)

Landscape fragmentation and changes in Shannon’s diversity index: It was found that the urbanization process in the Beijing-Tianjin-Hebei region has significantly exacerbated the fragmentation of the landscape, in which the ED (edge density), DIVISION (Landscape Delineation Index), SHDI (Shannon’s Homogeneity Index), and SHEI (Shannon’s diversity index) showed a sustained upward trend during the period of 1992 to 2022, and especially in the decade after 1992, during which the rise sharply increased. The LPI (largest patch) is the opposite. This fragmentation is mainly due to changes in land use within the region, especially land conversion behaviour that connects or separates different types of patches, leading to a decrease in the biodiversity index. It is worth noting that the policy of “returning farmland to forests and grasslands” implemented since 2002, though aimed at ecological restoration, has increased the fragmentation and complexity of arable land to a certain extent, further affecting biodiversity. However, in the natural development scenario projected for 2032, these indicators stop rising and falling, and landscape fragmentation is reduced, thanks to the environmental protection policy measures implemented in recent years.

(2)

Correlation between habitat quality and landscape features: The correlation between habitat quality and landscape patterns in the Beijing-Tianjin-Hebei region was also explored based on data from 1992 to 2032. The analyses showed that most indicators were positively correlated with habitat quality, while the landscape patch index (LPI) was negatively correlated with it. Meanwhile, the spatial correlation between habitat quality and landscape pattern characteristics in the Beijing-Tianjin-Hebei region gradually weakened after 2002. This was due to the fact that although the landscape fragmentation in the Beijing-Tianjin-Hebei region had been mitigated, the fragmented landscapes were gradually converted from construction land to natural landscapes, and the fast-developing construction land was still threatening the surrounding habitats, which led to the gradual decrease in the spatial correlation between the habitat quality and the landscape pattern index (LPI).

(3)

Spatial distribution of habitat quality: This study also found that the spatial distribution of habitat quality in the Beijing-Tianjin-Hebei region has a close correlation with the topography and landscape and has shown a decreasing trend over the past 40 years, from 0.599 to 0.564 between 1992 and 2032. Overall, the quality of the habitats in the northwestern part of the region is high, and the quality of the habitats in the southeastern part of the region is low, which creates a clear spatial pattern. The formation of this pattern is closely related to the uneven rate of economic development in the region, and the acceleration of economic development is usually accompanied by the overexploitation of natural resources, leading to a decline in habitat quality. Urban expansion not only compresses the original habitat, but also weakens habitat connectivity and stability by fragmenting the habitat landscape.

(4)

Calculation of average species richness (MSA) and influencing factors: The results of the analysis show that the average species richness has been decreasing from 0.270 to 0.183 over the past 40 years, and the MSA value was calculated by defining the proportion of arable land under facility-based agriculture in order to estimate the proportion of intensive farming, taking into account the national conditions of China. The MSA value in the Beijing-Tianjin-Hebei region is significantly affected by infrastructure, especially road construction. As the road network continued to expand, the MSA values in the region generally showed a decreasing trend. In addition, the MSA values of grasslands in the northwestern part of the region are relatively average, while the higher-elevation Yanshan and Taihangshan regions and remote woodland regions have higher MSA values, indicating that the ecological environment in these regions is relatively favourable. The steep topography limits the range of human activities, thus protecting the local ecological environment to a certain extent.