Land Use Change in a Typical Transect in Northern China and Its Impact on the Ecological Environment

Abstract

This paper uses seven periods of land use/cover data from 1990 to 2020 to investigate the spatio-temporal features of land use change in a typical transect in northern China. By using the land use transfer matrix, transect analysis, and ecological contribution rate, image interpretation data obtained from the Chinese Academy of Sciences Resource and Environmental Science Data Platform were analyzed using a spatial overlay to quantitatively examine the transect’s land use changes and their impact on the eco-environment. The results indicate that (1) the transect’s land use is dominated by cropland and grassland. (2) Cropland, forest land, and waters experienced significant transitions in 1995 and 2000, which was linked to socio-economic development and policy factors. (3) The total dynamic percentage of land use change is 4.52%, with built-up land and cropland showing the highest change. (4) The transect’s eco-environmental quality (EQ) has significantly declined, with the quality index dropping from 0.3839 to 0.3773. The transformation of cropland to forest land improves the eco-environment, while the transformation to built-up land has negative impacts. Human activities adversely affect the EQ. The findings are promising for leading the development of conserving the eco-environment and supporting the formation of regionally differentiated paths of ecological civilization construction in the transect.

1. Introduction

As a part of global change, land use change (LUC) is an intuitive reflection of the impacts of human intervention on the natural environment. The quantity and form of LUCs directly affect the eco-environment quality (EQ), such as causing regional warming effects [1], triggering extreme climate events [2], threatening habitat quality [3], and so on. The transect approach, which is among the best ways to investigate the connection between terrestrial ecosystems and global change, is also useful for examining LUCs and their effects on the natural environment.

At the International Transect Conference on Global Change and Terrestrial Ecosystem (GCTE) held in Marshell, USA, in 1993, dozens of transects were proposed [4]. The conference attracted the attention of scholars to the transect analysis method. By 1995, the North–East China transect and the North–South China transect in eastern China became two of the fifteen global change terrestrial transects that the International Geosphere-Biosphere Program (IGBP) constructed worldwide [5]. In research on terrestrial transects, international scholars mainly focus on botanical systems [6], climate change [7], elemental analysis [8], biodiversity [9], etc., while exhibiting minimal concerns about how human activity affects land use. Studies of the two domestic transects by Chinese scholars mainly focus on material cycling and utilization in ecosystems [10,11], the impacts of terrestrial ecosystem energy and productivity processes, and the modifications to land use practices and how they affect agricultural ecosystems [12].

The land system constitutes a complex adaptive social ecosystem—a complex ecological process created by the combination of the basic events of short-term land use changes, such as deforestation, and changes that may occur at a broader and longer-term scale [13]. The changes in land systems may bring about many persistently developing challenges. Evidence from existing studies suggests that changes in land use could directly affect regional climate change through the alteration of the physical properties of surface organisms [1]; the alteration of land use could also affect the living environment [3]; and reasonable land use changes could help to improve the ecological environment [14]. Therefore, LUCs have complex effects on the quality of the eco-environment. The in-depth comprehension of LUCs is central to crafting tactics to confront persistent challenges, including addressing energy transition, food security, climate change, and biodiversity loss [13]. In recent years, the effect of LUCs on the eco-environment has resulted in wide-spread concern for domestic and international academics. The effects of LUCs on the ecological environment, ranging from the composition of the atmosphere to surface ecology, are the subject of much Chinese academic research [15], which can be divided into four categories: (1) research into the impact of carbon emission effects [16,17,18] on single natural elements, such as climate and soil environment changes [19,20]; (2) research into the impact on ecosystem service value [21,22]; (3) research into the impact on landscape ecological effects [23,24]; and (4) research into comprehensive regional EQ [25,26,27,28].

According to the literature data retrieved using the title of “land use change” from the Web of Science, approximately 60% of LUC research published in international journals focused on the natural environment. Relevant research in recent years has mainly been carried out from three perspectives: driving factors [29,30], impact analysis [31,32], and the prediction of changes [33,34]. For example, Shuaishuai Jia et al. employed temperature as an indicator in their investigation into LUC influence on the biological environment to confirm the diverse impacts of various land utilization categories and regional variations on climate change [35]. Some scholars studied the LUCs in the Pearl River Delta (PRD) against the background of rapid urbanization, and believe that the urbanization-driven spatial growth in the PRD region had the most far-reaching impact on the decline of regional EQ [36]. Land use change was identified as one of the most significant elements affecting the biological water quality of the Guayas River basin, according to some other researchers [37]. Generally speaking, a substantial portion of previous studies mainly concerned the effects of changing land use on the ecosystem, as seen from a single-factor influence or micro-scale perspective, with less comprehensive regional research on a larger scale [3]. In particular, it should be highlighted that the transect method is applicable for analyzing land use changes with gradient differences. However, the international academic community still pays insufficient attention to transect-method-based studies in terms of LUCs and their effects. Therefore, conducting studies on LUCs and their effects on the ecological systems along transect lines is crucial.

China possesses the third-largest land area globally, and the regional variations in its natural geographies and socio-economic conditions are very significant. Therefore, selecting typical transects with obvious gradient differences to carry out land use research is one of the most effective means of deeply analyzing the spatio-temporal variation patterns of land use/cover, as well as driving forces, resources, and environmental effects in typical regions in China [38]. The transect area centered around Beijing selected in this paper covers the land use types formed through different human activities, including agricultural areas, pastoral areas, agro-pastoral ecotones, forest areas, and urban construction areas. The study region includes China’s second and third types of terrain, making it possible to more fully describe how various land use types have changed over the past three decades and how those changes have affected the natural environment in the transect area. However, existing research did not pay attention to the typicality and research value that are important to this transect. In addition, it is important to note that the related objectives when building an ecological civilization and their emphases vary, given that there are variations in how the ecological environment responds to pressures from different human activities. Therefore, regional differences should be considered when promoting the construction of an ecological civilization. In this space, the transect method can be applied to determine gradient patterns, which is useful in exploring the differentiated paths of ecological civilization construction. Specifically, each region of the transect faces different ecological and environmental pressures.

Beijing has undergone significant land use changes since becoming a megacity, and it is often affected by sandstorms. Meanwhile, the grassland area in the northern part of the transect is facing problems such as degradation and desertification. As it contains the estuary of the Yellow River, Shandong Province also faces challenges in water environment management and air pollution control. In this study, taking the belt area from southwest of Shandong to northeast of Inner Mongolia as our research transect, we applied techniques such as a land use transition matrix and the land use dynamic degree to examine features of the LUCs that have arisen in the transect region during the last three decades, and how they have spatially and temporally evolved. For instance, in this study, the impact on the EQ was quantitatively measured on the basis of a meta-analysis, to enrich the findings of the transect research. The research results of this paper have high reference value in controlling cropland loss, sustaining the ecological balance, and promoting the formation of regionally differentiated paths of ecological civilization construction, as in the transect zone studied.

2. Materials and Methods

2.1. Introduction to the Research Area

The transect we studied is situated in the north of China, between longitude 111°09′44″–120°57′50″ E and latitude 35°00′25″–45°05′37″ N. It is centered around Beijing and runs in a northwest–southeast direction, spanning five provincial-level administrative regions; Shandong, Hebei, Beijing, Tianjin, and Inner Mongolia; with a total of 23 municipal districts and 103 county-level administrative regions, covering an area of 223,100 square kilometers. The selected transect has distinct differences in climate, terrain, social and economic development, culture, etc. The transect area also spans various terrain types, including plains, hills, mountains, and plateaus, and crosses warm temperate and mid-temperate zones. The northern section of the transect is predominantly covered by grasslands, the central part is dominated by plains, and the southern part alternates between hills and mountains. The overall terrain is high in the northwest and low in the southeast. There are many natural resources and a variety of ecosystems throughout the transect. The gradients of natural conditions such as terrain and climate differ significantly. Diverse land uses coexist alongside numerous species; therefore, this is an ideal location for carrying out land use change research (see Figure 1). In addition, the socio-economic conditions of the selected research area differ broadly, including the economically developed metropolitan area represented by Beijing, the large area of cultivated land dominated by agricultural production in Shandong and Hebei Provinces, and the agricultural–pastoral intertwined area at the junction of Inner Mongolia and Hebei Province, which collectively reflect LUCs under different socio-economic levels. The study of LUCs along the transect is thus representative of regional variances.

2.2. Data Sources

The data on land use employed in this study originate from the Resource and Environmental Science and Data Center of the Chinese Academy of Sciences (http://www.resdc.cn (accessed on 20 March 2023)), including land use image interpretation data for seven points in time: 1990, 1995, 2000, 2005, 2010, 2015, and 2020. The image interpretation data classify land use/cover into six types: forest, grassland, waters, cropland, built-up land, and unutilized land. Cropland includes paddy land and dry land; forest includes dense forest, scrubland, open forest, and other forest; grassland includes low-, medium-, and high-covered grassland; waters include rivers, canals, marshes, lakes, reservoirs, and low beaches; built-up land encompasses urban areas, rural habitations, industrial zones, and mining sites; and unutilized land includes sandy land, desert, bare land, and so on. The land use data processing and analysis were performed using ArcMAP 10.7.

2.3. Methods

2.3.1. Land Use Transfer Matrix

The first article detailing the land use transfer matrix in China was published in 2002 [39], and since then, this approach has been extensively employed in land use research. Utilizing a land use transfer matrix offers a comprehensive and detailed depiction of the structural attributes of regional land use and cover changes (LUCs), mirroring the trends in LUCs driven by human interventions. A land use transfer matrix is created through the quantitative characterization of system states and the examination of how these states evolve within the system. This study used ArcGIS 10.7, a powerful tool for spatial processing of geographic information, to cross-analyze image interpretation data, and then through Excel’s pivot table processing, the quantitative relationships between land cover types and how those have transformed in different periods were investigated. For detailed data, please refer to the Supplementary Materials at the end of the article.

2.3.2. Dynamic Change in Land Use

Wang Xiulan’s research team was a pioneer in exploring research methodologies for land use dynamic change [40], and our study of the dynamism of land use followed one such approach, which is to analyze the degree of LUC using land use area data, to facilitate an analysis of the factors influencing LUC. This covers all major changes in all kinds of land uses within a research region over a specified timeframe. It serves as a gauge for the area most affected by LUC, and it may also be applied to compare LUCs between the local area and the whole or regional area [41]. This methodology is commonly applied in LUC studies, encompassing both comprehensive and single land use dynamics. The expression of the comprehensive land use dynamics is as follows:

$$\mathrm{LC}={\displaystyle \frac{{{\displaystyle \sum }}_{i=1}^n\Delta L{U}_{i-j}}{{{\displaystyle \sum }}_{i=1}^{n}L{U}_i}}\times {\displaystyle \frac{1}{\mathrm{T}}}\times 100\%$$

(1)

where LC is the comprehensive land use dynamics; LU_i is the extent of i land use type at the initial stage of the research; LU_i–j is the land extent of j land use type’s transformation from i land use type within the study period (denoted by T); and n is the number of land use types [41].

Dynamism refers to fluctuations in the area of a particular land use type in a certain region [40], which are used to measure the rates and magnitudes of area changes in different land types [41]. The calculation formula is as follows:

$${\mathrm{K}}_{\mathrm{i}}={\displaystyle \frac{S_2-S_1}{S_1}}\times {\displaystyle \frac{1}{\mathrm{T}}}\times 100\mathrm{\%}$$

(2)

where K_i represents the dynamic change in the i-th type of land use during the study period, S₁ denotes the initial extent of the i-th type of land use, S₂ corresponds to its extent at the end of the study period, and the duration of the study is represented by T [42].

Given the gradual and cumulative nature of land use changes, a quantitative assessment of these dynamics at defined intervals helps to clarify them. Considering that land use changes in the transect are strongly influenced by socio-economic development, and since the regional economic and social development plan formulated every five years is closely tied to socio-economic development, we use the same time interval as that of the five-year plan to calculate the dynamic changes in land use, based on Equations (1) and (2).

2.3.3. Analysis of the Transect

A transect refers to a strip-like research area selected based on a dominant driving factor, with obvious gradient patterns or differences [43]. Motivated by the international IGBP research program, the transect study methodology has been successfully applied in research on global change [44]. However, despite its potential, few studies on land use change employ the transect method, with existing cases largely focused on administrative regions [45] or river basins [38]. To address the limited application of this approach, a transect was chosen in this study. We based this transect on ecological research on land use gradient changes in sampling plots, centered around Beijing city, with administrative regions at the county level as the basic unit, consisting of 23 municipal regions and 103 county-level administrative divisions, spanning five provincial-level administrative regions: Shandong, Hebei, Beijing, Tianjin, and Inner Mongolia. When we consider the terrain and climate patterns, it can be found that the land use types covered by this transect are abundant and diverse, and that they have clear-cut gradation. There is a significant hierarchical change in land use, forming a spatially stratified pattern of pastoral areas, agricultural–pastoral interlaced zones, and agricultural and forestry areas gradually expanding from north to south, which supports research on LUC. To present the findings of our examination of the spatial pattern features of LUC in the transect, this article is divided into six subregions: Inner Mongolia grassland area, agricultural and pastoral intertwined area, northern Beijing–Hebei forest area, Beijing–Tianjin–Hebei urban area, Hebei–Shandong agricultural area, and central and southern Shandong hilly areas, built upon the gradient pattern of terrain, climate, and land cover in the transect selected. The regional division of the transect is shown in Figure 2.

2.3.4. Evaluation of EQ of Land Use

A land use change is often brought about by socio-economic and natural causes combined, which may harm social and natural environments. For example, in the process of urbanization, many croplands are converted to built-up lands, resulting in significant changes in the surface permeability, surface albedo, types and quantity of vegetation, and biological habitats, which, in turn, affect the regional precipitation, local climate, and biodiversity. Thus, alterations in the way land is utilized may directly affect the local natural ecosystem. At present, the method most commonly used to rate the EQ of land use is the Eco-Environmental Quality Index (EQI) established by Li Xiaowen et al. [43], which employs representative and accurate model parameters. This study applied a meta-analysis to determine the EQI, following the approach displayed in Figure 3.

(1)

Eco-environmental Quality Index (EQI)

The influences of alterations in land use on the EQ in the transect were assessed by calculating the EQI and comprehensive ecological contribution rate. As stated, the value of the EQI was determined through a meta-analysis; the index calculation process is shown in

Table 1. Meta-analysis process.

Author	Ref.	Cropland	Orchard Land	Forest Land	Grassland	Urban and Industrial Land	Water	Unutilized Land
Qin Sigang	[46]	0.325	0.75	0.558	0.4	0.17	0.42	0.11
Gulipostan Batu et al.	[47]	0.275	-	0.6125	0.47	-	0.575	0.315
Chen Kunpeng et al.	[48]	0.317	0.65	0.675	0.3	0.2	0.456	0.15
Cui Jia et al.	[49]	0.275	-	0.6125	0.467	0.183	0.625	0.13
Fu Yingxiu et al.	[50]	0.275	-	0.6125	0.467	0.183	0.625	0.13
Shi Ling et al.	[51]	0.275	0.46	0.633	-	0.06	0.5125	0.353
Wu Yuhong et al.	[52]	0.275	-	0.6125	0.467	0.183	0.625	0.13
Yang Yajuan et al.	[53]	0.28	-	0.67	0.47	0.19	0.64	0.13
Zhang Fangyi et al.	[54]	0.295	-	0.68	0.488	0.022	0.645	0.015
Zhang Yang et al.	[55]	0.275	-	0.6125	0.467	0.183	0.57	0.18
Wang Jiahui et al.	[56]	0.25	-	0.95	0.75	0.2	0.65	0.1
Hou Lei et al.	[57]	0.275	-	0.6125	0.467	0.183	0.625	0.13
Zhang Wanping et al.	[58]	0.275	-	0.6125	0.467	0.183	0.625	0.13
Guo Yanjun et al.	[41]	0.275	-	0.6125	0.467	0.183	0.583	0.05
Zhang Shuhan et al.	[42]	0.27	-	0.76	0.7	0.18	0.54	0.11
Li Xiaowen et al.	[43]	0.275	-	0.6125	0.467	0.183	0.625	0.13
Nan Shengxiang et al.	[59]	0.25	-	0.65	0.45	0.2	0.55	0.01
Yang Shuhe et al.	[60]	0.275	-	0.6125	0.467	0.183	0.625	0.13
Wu Xinhai et al.	[61]	0.275	-	0.6125	0.2	0.183	0.575	0.05
average value		0.28	-	0.65	0.48	0.19	0.59	0.14

. Considering the geographic location of the transect and the reference value of previous cases, 50 papers were retrieved from the China Knowledge Network and Web of Science Core Database for the past three decades concerning the EQI and with China as the study area. Thirteen papers were eliminated as they were not related to our study approach, leaving nineteen studies on the analysis of the EQ according to the land use categories. A meta-analysis of the research area was conducted on the basis of those 19 papers and EQI results were obtained, as shown in

Table 2. Data assignment of EQI for different land types.

Land Use Type	Eco-Environmental Quality Index Value of Each Land Type
cropland	0.28
forest land	0.65
grassland	0.48
waters	0.59
built-up land	0.1833
unutilized land	0.14

.

Based on the above, the EQI can be computed in an alternative manner, as detailed below:

$$\mathrm{EV}={{\displaystyle \sum }}_{i=1}^n{S}_i\times R_i÷S$$

(3)

where the EQI for period t is represented by EV; S_i is the land extent of i land use type in time period t; R_i is the EQI value of i land use type; S is the aggregate land area of the study extent; and n is the number of land use types.

(2)

Ecological contribution rate of land use change

The ecological contribution rate is used to quantify the changes in regional EQ resulting from the changes in land use types [62,63]. If a land use type with a low EQ value shifts to a land use type with a high value, the local ecological setting undergoes an enhancement, and vice versa. The expression is as follows:

$$\mathrm{LEI}={\mathrm{EV}}_2-{\mathrm{EV}}_1\times {\displaystyle \frac{\mathrm{LC}}{\mathrm{S}}}$$

(4)

where LEI is the ecological contribution rate; EV₁ and EV₂ are the EQIs of land use types at the initial stage and the completion of the research period; LC stands for the altered land use extent; and S represents the gross extent of the study region.

3. Results

3.1. Spatial–Temporal Characteristics of Land Use Change

3.1.1. Characteristics of Temporal Evolution

(1)

Dynamic changes in land use

Between 1990 and 2020, there was a comprehensive shift in land use within the transect (of 4.52%, measured by Formula (1)), amounting to a significant change. Nonetheless, throughout this period, the major land uses within the transect were cropland and grassland (see

Table 3. Area changes in different land types from 1990 to 2020.

	1990	1995	2000	2005	2010	2015	2020	Dynamic Change
grassland	82,371.17	81,765.84	81,750.37	80,885.81	81,109.01	79,077.67	79,022.42	−4.24%
cropland	94,825.56	89,045.62	92,984.77	91,730.78	91,100.16	88,528.94	87,839.38	−7.95%
built-up land	13,094.69	14,718.54	14,984.37	16,361.00	17,025.71	21,960.16	22,693.72	42.30%
forest land	19,960.39	24,504.68	20,149.27	20,174.90	20,132.48	20,550.72	20,521.81	2.74%
waters	5318.91	5008.90	5549.09	5512.54	5509.04	5196.04	5253.47	−1.25%
unutilized land	7534.96	8062.10	7750.86	8504.91	8216.11	7909.67	7892.39	4.53%

Area Unit: km².

). Satellite imagery of LUC over the last three decades revealed that the coverage of these two land types was constantly above 74%. In 1990, the sum area of cropland and grassland was 79.42% of the transect’s total area, and by 2020, the total extent of the two land types has shrunk only slightly to 74.75%. The main changes in land use were the conversion of cropland and grassland into urbanized areas and unused land. From 1990 to 2020, the areas of built-up land, forest land, and unutilized land increased to varying degrees, while the areas of cropland, grassland, and waters were reduced. The extent of built-up land experienced the greatest net increase, at 9599.03 square kilometers, and the amount of cropland dropped the most, by 6986.18 square kilometers. Cropland and built-up land both saw considerable changes, with built-up land undergoing the largest dynamic shift (73.30%), suggesting that a sizable portion of other land types was converted to built-up land. The dynamic change in cropland was −7.37%, making it the most drastically transformed land type in the transect. The area of waters did not change much within the study period, essentially maintaining a relative balance.

The line graph in Figure 4 depicts changes in the extent of each land use type every five years; this was utilized to assess the shifts in land use. This graph revealed that the dynamism of different land use types peaked at different time points. Cropland underwent the most significant change between 1990 and 2000, with an increase in overall dynamism followed by a reduction, producing an overall increase in cropland area. Specifically, in the period from 1990 to 1995, cropland experienced a notable decline, showing a dynamic change of −1.22%. Then, between 1995 and 2000, it underwent a modest increase, with a dynamic change of 0.88%. Following that, from 2000 to 2020, the cropland area consistently decreased, initially at an accelerating rate, which then gradually slowed. These adjustments were directly tied to the period’s urbanization and economic growth. In recent years, the Chinese government has formulated many regulations on cropland protection to address the issues of rapid industrialization and urbanization for agricultural land. These policies have reduced the rate of cropland’s transformation to other land types to an extent, but nonetheless, cropland area continues to decrease. In the future, strengthening cropland protection will be pivotal to realizing Chinese-style modernization in the transect area.

The area of grassland generally demonstrated a decreasing trend, but with changes significantly fluctuating. The grassland area somewhat increased in the period from 2005 to 2010, with a positive dynamic change, while the dynamic changes were all negative in the other time periods. Among them, the dynamic change in grassland reduction was the highest, which was −0.50% in the period from 2010–2015. Since 1990, amidst the swift expansion of both population and livestock numbers, the grassland in Inner Mongolia has degraded continuously. Moreover, in the years before and after 2000, the ecological environment of grassland was further deteriorated through the impacts of drought and other serious natural disasters. Since 2000, the government of the Inner Mongolia Autonomous Region has adopted a series of measures to address this harm, such as prohibiting and resting grazing, but with limited effects in protecting the grassland.

The dynamism of built-up land was consistently positive and demonstrated fluctuations. The extent of built-up land continuously increased in the three decades studied, with the lowest dynamic degree of 0.36% in the period from 1995 to 2000, and the highest dynamic change of 5.80% in the period from 2010 to 2015, indicating that the built-up land expanded rapidly in the transect area during these periods.

The dynamic changes in forest land fluctuated first and then stabilized. During the research period, the extent of forest land increased rapidly, with a dynamic change of 4.55%. However, between 1995 and 2000, the extent of forest land declined swiftly, with a dynamic change of −3.55%. Since entering the 21st century, China has adopted a series of policies to protect forest land and has achieved phased results. Since 2000, the dynamic change in forest land has thus had less fluctuation, and the forest land extent has slightly increased.

In general, the area of waters in the transect showed a decreasing trend, and the dynamic changes fluctuated significantly. Specifically, in the period from 1990 to 1995, the area of waters decreased, with a dynamic change of −1.17%; from 1995 to 2000, the water extent increased, with a dynamic change of 2.16%; and between 2000 and 2020, the area of waters fluctuated continuously, with limited decrease and increase. The dynamic changes in unutilized land remained between −1% and 2%, with the area increasing in general.

Overall, other than rare exceptions with the built-up land and grassland, the magnitudes of dynamic changes in the studied land categories were found to have declined over time.

(2)

Transformation of land use types

Based on Equation (3), and using the overlay tool in ArcGIS 10.7, we analyzed the land use vector data [64] for the seven periods in the transect area for their “intersection”, and the land use conversion matrix was then derived, as presented in

Table 4. Land use transfer matrix from 1990 to 2020 (km²).

1990	2020						Total	Transformation
	Grassland	Cropland	Built-Up Land	Forest Land	Waters	Unutilized Land
grassland	74,311.93	3559.09	828.91	1426.25	255.09	1968.53	82,349.79	8037.87
cropland	2015.41	78,261.47	11,776.62	1482.56	1211.10	64.71	94,811.87	16,550.40
built-up land	127.00	3447.93	9253.02	71.51	182.28	9.82	13,091.57	3838.55
forest land	849.67	1173.03	417.25	17,413.88	76.17	16.87	19,946.86	2532.98
waters	335.61	970.59	305.64	102.11	3448.05	158.36	5320.36	1872.31
unutilized land	1341.15	368.71	88.15	11.15	57.25	5667.40	7533.80	1866.40
total	78,980.77	87,780.82	22,669.58	20,507.45	5229.94	7885.69	223,054.24	--
transform-to	4668.84	9519.35	13,416.56	3093.57	1781.89	2218.29	--	--

. This table shows that the transect’s land use types changed dramatically during the 30 years studied. Six of the land use types transformed with each other, but with different transformation characteristics. Among them, the built-up land and grassland mainly impacted change rates. The built-up land’s transformation changed from a relatively slow increase before 2010 to a rapid increase after 2010. Meanwhile, grassland’s transformation changed from a relatively slow decrease before 2010 to a rapid decrease, and then it tended to level off. Beyond them, the unutilized land and waters showed a reversal trend in the long run, where both exhibited an increase first, followed by a decrease. When taking five years as a cycle, cropland, forest land, and waters all had two turning points, in 1995 and 2000. Cropland and waters showed a “decrease–increase–decrease” pattern, while forest land showed an “increase–decrease–leveling off” pattern. The unutilized land, in general, experienced one transition, where the turning point from an increase to a decrease appeared in 2005 (see Figure 4). In sum, there was a notable transformation of land use in the transect area over the studied period.

From the perspective of “transform-from”, lots of cropland, unutilized land, and forest land were transformed from grassland within the study period. Their areas transformed from grassland accounted for 4.32%, 2.39%, and 1.73% of the aggregate grassland area in 1990, respectively. The main directions of cropland transformation were built-up land and grassland, comprising 12.42% and 2.13%, respectively. The main transformation direction of built-up land was cropland, accounting for 26.34%. Forest land was mainly transformed into cropland and grassland, comprising 5.88% and 4.26%, respectively. Waters were mainly transformed into cropland and grassland, comprising 18.24% and 6.31%, respectively. The unutilized land, meanwhile, was mainly transformed into grassland, with the transformed area accounting for 17.80%. The land areas transformed from all types of land in the 30 years in an order of decreasing magnitude were as follows: cropland > grassland > built-up land > forest land > waters > unutilized land.

When viewed from the “transform-to” perspective, the land types transformed into grassland in the transect were arranged in descending order from most to least as follows: cropland > unutilized land > forest land > waters > built-up land. Among them, the extent of grassland transformed from cropland was 2015.41 km², constituting 43.17% of the total area of transformed grasslands, mostly from abandoned lands scattered on grasslands.

The land types transformed into cropland arranged in descending order were as follows: grassland > built-up land > forest land > waters > unutilized land, among which the grassland and built-up land were the main types. The area of these two kinds of land transformed into cropland accounted for 73.61% of the total area of the cropland transformed. The variation curve in Figure 4 demonstrates that cropland had a land use transition in 1995, transitioning from a reduction in 1990–1995 to an increase in 1995–2000. It can be seen further by combining those findings with Figure 5 that the areas where the grassland was transformed into cropland were mainly distributed in the agro-pastoral ecotone, and the regions where cropland was reclaimed from built-up land were predominantly concentrated in rural settlements in Hebei and Shandong. This indicates that China’s cropland occupation, compensation balance policy, and built-up land increment and decrement linkage policy were effective in stabilizing cropland.

The types of land transformed into built-up land arranged in descending order were as follows: cropland > grassland > forest land > waters > unutilized land. The extent of cropland transformed into built-up land was close to the initial extent of the built-up land in 1990, indicating that construction activities were the main cause of cropland transformation in the transect. The land types transformed into forest land were mainly cropland and grassland, with the area of these two types of lands transformed into forest land accounting for 94.03% of the aggregate extent of forest land transformed. When analyzing Figure 4, it can be found that this change was strongly associated with the policy aimed at reverting cropland to forests, released in 2002. A series of policies released subsequently also promoted the recovery of forest land area.

The land types transformed into water were mainly cropland and grassland, of which cropland transformed into water comprised 22.76% of the whole water area in 1990. Meanwhile, the main land type transformed into unutilized land was grassland. There were 1968.53 km² of grassland transformed into unutilized land within the 30 years studied, indicating that the grassland degradation in the transect was relatively serious. The degraded area of grassland was mainly distributed in the agricultural–pastoral intertwined area.

Overall, the extent of transformation of the six land use types could be arranged in the following descending order: built-up land > cropland > grassland > forest land > unutilized land > waters, suggesting that the transect’s land use shift was largely driven by urban and rural building, particularly fast urbanization.

3.1.2. Spatial Characteristics of LUC

Let us now consider subregions. The Inner Mongolia grassland zone refers to the range within the Inner Mongolia Autonomous Administrative Region in the north of the transect, featuring grassland as the predominant land use. This is also the main location where unutilized land is found. The intertwined agricultural–pastoral area refers to the zone where agriculture and animal husbandry coexist in a blend, situated at the junction of Inner Mongolia and Hebei Province, where grassland, cropland, and built-up land are interlaced evenly and distributed in a uniformly scattered pattern without an evident clustering center. The forest zone in the north of Beijing and Hebei refers to the mountainous area in the north of Hebei and around the Yanshan Mountains, where forest land dominates all year round, with a large and concentrated woodland distribution. The Beijing–Tianjin–Hebei urban zone refers to the core region of the Beijing–Tianjin–Hebei city cluster in the transect, where the built-up land distribution is relatively dense, with a conspicuously agglomerative diffusion pattern within the Beijing and Tianjin administrative regions. The Hebei–Shandong agricultural zone refers to the area south of Hebei and north of Shandong in the transect, where cropland dominates and the built-up land is evenly distributed in a dotted pattern, mainly developing agricultural production. The central and southern Shandong hilly zone refers to the hilly area in Shandong Province in the south of the transect. Relative to other zones, the land types in the southern hilly zone of central Shandong are diverse and relatively balanced, with cropland, built-up land, forest land, and the water area distributed in a staggered manner, without obvious spatial agglomeration.

Using the ArcGIS intersection analysis tool, a map was produced illustrating the spatial pattern of LUCs, as depicted in Figure 6. Analyzing the LUCs in each zone revealed that changes were mainly in the aspects detailed herein.

The land use transformation in the Inner Mongolia grassland zone was mainly a mutual transition between unutilized land and grassland, which ultimately resulted in a transformation of unutilized land into grassland. The predominant shift in land utilization at the agro-pastoral interface was a reciprocal conversion between grassland and cropland, with significant decreases in cropland and waters. The land use transformation in the forest zone north of Beijing and Hebei was dominated by a transition between grassland and forested land, with grassland significantly reduced, forest land expanded, and waters such as the Miyun Reservoir significantly expanded. There was a notable increase in built-up land in each administrative region of the Beijing–Tianjin–Hebei urban area. The most significant rise was observed in the regions encircling Beijing. The land use transformation in the Hebei–Shandong agricultural zone consisted largely of the transition of cropland to built-up land, with built-up land significantly increased and evenly distributed. The land use transformation in the central and southern Shandong hilly zone was mainly a transition of grassland to cropland. Finally, there was a substantial aggregation of built-up land in principal urban areas, such as Jinan, Zibo, and Qingdao.

Modifications to the land use structure across the studied period were represented with a change chord diagram (Figure 7), which revealed that changes in cropland and grassland were the main LUCs between 1990 and 2000, and that the transformation of land use was chiefly characterized by variations in cropland and built-up land from 2000 to 2015. During the period of 2015–2020, the changes in cropland and built-up land were particularly clear. Most of the reduced cropland changed to built-up land.

3.2. Ecological Impact Analysis of Land Use Change

3.2.1. Comprehensive Analysis of the Eco-Environmental Quality

(1)

Eco-environmental Quality Index

Table 5. Changes in EQI.

Year	Eco-Environmental Quality Index	Change Compared to the Previous Node
1990	0.3839	--
1995	0.3895	0.0055
2000	0.3830	−0.0064
2005	0.3812	−0.0019
2010	0.3812	0.0001
2015	0.3777	−0.0036
2020	0.3773	−0.0003

displays the calculated EQI values at the seven distinct time points selected, which were derived using Equation (3). It can be seen that, during the 30 years from 1990 to 2020, the EQ of the transect decreased to 0.3773 from 0.3839 (Figure 8), and the overall EQ deteriorated. Combined with the temporal and spatial features of land use, between 1990 and 1995, the EQI of the transect increased to 0.3895 from 0.3839, and the EQI somewhat improved. In general, the EQI declined in the 25 years from 1995 to 2020. Although there was a slight increase in 2010, it continued to decline thereafter, and it decreased to 0.3773 in 2020. When combined with the findings in Table 3, it can be surmised that the large increase in built-up land and large reduction in grassland and cropland were notable causes of the decline in the EQ in the transect area.

(2)

Spatial pattern of EQ

Using the natural breakpoint method, the EQ calculated was divided into four grade zones: a low-quality region, medium–low-quality region, medium–high-quality region, and high-quality region, and the spatial distribution of the EQ in the transect area in the period from 1990 to 2020 was determined, as shown in Figure 9. In this figure, the darker the color, the worse the EQ. The figure demonstrates that except for the Inner Mongolia grassland area, which had smaller changes, the overall EQ grade of all areas in the transect declined. (In the Inner Mongolia grassland zone, the ecological environment was constantly at a high quality, but the increase in unutilized land during the three decades studied led to an expansion of the scattered low-quality areas.) The overall EQ of the agricultural–pastoral intertwined area declined most clearly, decreasing from high-quality grade domination in 1990 to medium–high-quality classification domination in 2020, and the low-quality classification areas expanded significantly. The main changes in the forest zone north of Beijing and Hebei were that, in the northwest area, the high-quality classification zone declined to a medium–high-quality classification zone, and the medium–high-quality classification zone declined to a medium–low-quality classification zone. Meanwhile, in the Beijing–Tianjin–Hebei urban zone, the EQ declined by one grade. The area of the low-quality grade zone expanded significantly as compared with the medium–low-quality grade zone in the initial stage. Overall, the EQ was greatly impacted by the widespread development of built-up area. As built-up land expanded, the EQ of some areas declined from a high-quality grade zone to a low-quality grade zone. Similar to the Beijing–Tianjin–Hebei urban areas, the EQ of the Hebei–Shandong agricultural zone and the central and southern Shandong hilly zone declined by one grade, dropping from high-quality classification domination to medium–high-quality classification domination. Finally, as a result of the sprawling urbanization and expansion of rural habitation, the low-quality classification regions became fragmented and dispersed, appearing in a scattered dot pattern.

3.2.2. Analysis of the Contribution of LUC to the Eco-Environment

Improvement and degradation of the regional EQ occurred simultaneously, and the eco-environmental effects of transformations of two types of land uses could balance each other out to some degree, promoting a relative balance of the EQ in the study area. Here, the eco-environmental contribution rate of LUC and its proportion in the transect area in the period from 1990 to 2020 were obtained through Equation (4) (

Table 6. Eco-environmental contribution in the period from 1990 to 2020.

Land Use Transformation Leading to Ecological Environment Improvement				Land Use Transformation Leading to Ecological Deterioration
Land Use Change	Area (km2)	Contribution Rate	Proportion	Land Use Change	Area (km2)	Contribution Rate	Proportion
CR-F	1482.56	0.00246	20.90%	CR-CO	11,776.62	−0.05059	27.81%
CR-G	2015.41	0.00181	15.35%	F-CR	1173.03	−0.01928	10.60%
CR-W	1211.10	0.00168	14.30%	F-G	849.67	−0.00642	3.53%
G-F	1426.25	0.00109	9.24%	G-CR	3559.09	−0.03162	17.38%
G-W	255.09	0.00013	1.07%	G-CO	828.91	−0.01093	6.01%
U-CR	368.71	0.00023	1.97%	G-U	1968.53	−0.02973	16.34%
U-G	1341.15	0.00204	17.37%	F-CO	417.25	−0.00865	4.76%
CO-G	127.00	0.00017	1.44%	W-CR	970.59	−0.01337	7.35%
U-W	57.25	0.00012	0.98%	W-G	335.61	−0.00164	0.90%
CO-CR	3447.93	0.00149	12.70%	W-CO	305.64	−0.00552	3.04%
CO-F	71.51	0.00015	1.27%	W-U	157.18	−0.00314	1.73%
CO-W	182.28	0.00033	2.82%	CR-U	64.71	−0.00040	0.22%
U-F	11.15	0.00003	0.22%	F-W	76.17	−0.00020	0.11%
U-CO	88.15	0.00002	0.15%	F-U	16.87	−0.00038	0.21%
W-F	102.11	0.00003	0.23%	W-U	1.18	−0.00002	0.01%
total	12,187.65	0.01177	1	CO-U	9.82	−0.00002	0.01%
--	--	--		total	22,510.86	−0.18192	1

Notes: CR refers to cropland, CO refers to built-up land, G refers to grassland, F refers to forest land, W refers to waters, and U refers to unutilized land.

). The transformation of cropland to forest land contributed most to the improvement in the EQ (20.9%), followed by the restoration of unused land to grassland (17.37%) and of cropland to grassland (15.35%). These findings indicate that reverting cropland to grassland and forest, together with introducing greenery to unutilized areas, is crucial in enhancing the local ecosystem. Among the land use transformation types leading to a deterioration in the EQ, the transition of cropland to built-up land (27.81%) and that of grassland to cropland (17.38%) had the greatest impacts. Generally speaking, the influence of land use variations on the EQ in the transect area was more of producing deterioration than improvement, with a comprehensive contribution rate of −0.00659 to the ecological environment. The transformations of cropland, grassland, and built-up land were most critical for the EQ of the transect area. To sustain a high EQ in the transect area, it is crucial to control the excessive expansion of cities into cropland and that of cropland into grassland. Although the reclamation of grassland as cropland ensures there is a sufficient area of cropland, it reduces the EQ of the region. Grassland is relatively rich in species, while there are less diverse species in cropland—that is to say, grassland can provide habitats for a wider range of animals. This is conducive to increasing species richness and heightening the stability of the ecological environment. Therefore, limiting and slowing the enlargement of built-up land and managing the transformation relationship between cropland and grassland are essential for food security and ecological environment security.

4. Discussion

We found in this study that the land use in the transect area explicitly transitioned, and the regional LUC was significant, especially the change in cropland and grassland in the agricultural–pastoral intertwined area and the conversion of cropland and built-up land in Hebei and Shandong Provinces, indicating that China’s balance policy of cropland occupation and compensation and its policy linking the rise and decline of urban and rural built-up lands contributed substantially in stabilizing the area of cropland. The above findings are consistent with the relevant research conclusions of Tao Zefu [65], Long Hualou [66], and Liao Liuwen [67], among others. Furthermore, we found that the types of LUCs were highly complex, manifesting in changes in land types and shifts in spatial patterns. Concerning the transformation of land types, the area of built-up land steadily increased over the three decades studied, with this growth only slowing between 2015 and 2020. The main impetus for the growth of urbanized regions was the conversion of cropland to built-up land. Over the studied period, the cropland area experienced a significant decline, with us consistently registering negative growth over the three decades. This did, however, have a gradually decreasing rate, and we found that since the beginning of the 21st century, China’s policies aimed at protecting cropland have achieved some success. However, it is crucial to continue enforcing these policies to curb the ongoing reduction in cropland and to establish a robust foundation for food security. In terms of spatial patterns, the overall configurations across the six sub-areas did not undergo significant changes. However, notable transformations occurred among various land types within these sub-areas, particularly in the agricultural–pastoral intertwined area, where the EQ deteriorated most markedly.

Recently, the issue of food security has attracted much attention at home and abroad. One of the key challenges in ensuring food security in China is effectively utilizing restricted land resources to satisfy the dietary needs of its residents [68]. We found in this study that the conversion of cropland into grassland, a type of LUC, significantly contributed to enhancing the EQ in the transect area, comprising 15.35% of the ecological improvement contribution rate. However, under the dual contexts of “maintaining the crucial threshold of 1.8 billion mu of arable land” and “a comprehensive and sustainable food perspective” in China, which are aimed at establishing a harmonious balance between ecological security and food security, there is a need to enhance the research on agricultural systems and grassland agriculture. In the 20th century, under the leadership of Mr. Ren Jizhou, grassland science and grassland agriculture research was gradually explored. However, the development models of grassland agriculture vary in different regions. For example, grassland agriculture in the agro-pastoral ecotone [69] still faces great problems, which stand in the way of achieving sustainable development. Therefore, further research on how to confront the relationship between LUC and agricultural structural transformation is needed. In addition, to maintain the EQ of the transect area at a high level, it is vital that we avoid the excessive conversion of cropland into urban development, and grassland into cropland. Urban development efforts need to be fully shifted to revitalization and utilization of the stock spaces and organic renewal, though this places new demands on urban development.

In addition, it is imperative to consider the variations among regions when comprehensively implementing practices aimed at constructing an ecological civilization. In this study, given the transect’s wide reach and the diverse human activities that impacted it, we found that land use changes differed significantly between regions, which highlights that different focuses and pathways of construction of ecological civilizations are required in different regions. In other words, given the disparities in economic development and ecological governance agendas among regions, it is crucial to uphold the principle of locality-specific adaptation, tailored to each area’s specific needs and conditions. A practical and feasible route to the construction of an ecological civilization should be formulated through a close study of the specific reality of the region at the stage of top-level design by the central government [70]. For example, the ecological quality of the agricultural–pastoral intertwined area of this transect experienced a significant decline, highlighting the need for strict management of excessive cropland reclamation, alongside the protection of forest land and waters. In the Beijing municipal district, built-up land notably expanded, particularly into the surrounding areas, revealing a pressing challenge: controlling the sprawl of urban development. This situation underscores the point that regions with varying land use characteristics require tailored approaches to constructing an ecological civilization. Accordingly, we found distinct geographical disparities in urban renewal efforts and cropland protection strategies across these regions. To support future efforts, this study’s analysis of LUCs’ temporal and spatial patterns in a typical transect should serve as a guide for implementing differentiated approaches in China toward the development of an ecological society.

However, there were some shortcomings to this study. Firstly, the accuracy of the EQI values determined should be improved through future studies based on more than the 19 papers we utilized, and the EQI should be studied more widely across the vast land area of China. Secondly, despite the significant differences in land use changes in different sections of the transect, the current research primarily focused on analyzing the entire transect, and further detail can be gained in future studies by applying a narrower field of study. Moreover, we only considered the six major categories of land use; these six types can be subdivided into subcategories in future research, which we did not attempt due to limitations in our image data. Finally, the regional ecosystem is a vast, open system, and the change in the EQ is affected by many factors. The influence of LUC on the ecosystem is significant and cannot be ignored, but other factors affecting the EQ should also be considered in further research.

5. Conclusions

In this study, a typical transect in the north of China was selected for a quantitative analysis on the characteristics of LUC and their impacts on the EQ in the period from 1990 to 2020. We found that the land use in the transect underwent a clear dominant morphological transition, though the land use structure of the transect was dominated throughout by cropland and grassland, with the sum of their areas always more than 74% of the total area of the transect. Nonetheless, there were differences in the features of LUCs throughout the transect area and over time.

In the temporal dimension, the degree of dynamism concerning built-up land and cropland was the highest, and their changes exhibited the greatest magnitudes of all land types studied. Among those changes, the growth rate of built-up land significantly decreased in 2010, and we surmised that China’s policy of controlling the disorderly increase in built-up land introduced in 2010 played a role. Cropland, meanwhile, experienced the greatest reduction, with the dynamic change in cropland being negative over the latter 20 years of the study period. In response, we propose that close attention should be paid to the transect area to strictly control the reduction in cropland.

In the spatial dimension, the geographical distribution of land use within the study region revealed distinct stratification. The Inner Mongolia grassland zone mainly saw mutual conversion between unused land and grassland, which shows that the problem of grassland degradation in the region still deserves our attention. The most pronounced expansion of built-up land was observed within the Beijing–Tianjin–Hebei urban agglomeration. Meanwhile, as a whole, the distribution of all types of LUCs in the southern part of the transect area was found to be relatively uniform; we suggest that this area needs to be studied at a more detailed level.

In the EQ evaluation, we found that the EQ of the transect area decreased notably. Over the research period, LUCs had dual effects on ecosystem services, with both favorable and unfavorable outcomes. However, the negative effects of LUC on the EQ in the transect area surpassed the positive effects. Going forward, we assert that to improve the EQ, it is necessary to enhance the macro control over land use and better coordinate the relationship between ecological environmental protection and regional land use.