Land Utilization, Landscape Pattern, and Ecological Efficiency: An Empirical Analysis of Discrimination and Overlap from Suining, China

Abstract

The rapid urbanization in recent decades has highlighted the impact of rural land utilization, which further affects the spatial structure and efficiency of rural ecosystems. Optimizing the structure of urban green infrastructure is an effective way to alleviate the fragmentation of rural landscapes, coordinate the relationship between rural development and ecosystem services, and ensure sustainable rural development. The purpose of this paper is to provide a clear direction for the optimization of construction for the sustainable development of rural green infrastructure (RGI). This study provides a new method for RGI identification and analysis by taking Suining County, a typical rural area on the North China Plain, as an example. Morphological spatial pattern analysis (MSPA) was used to distinguish different space scenery elements in RGI and combine them with land utilization elements, so as to obtain two types of overlapping degree data in each village and town. We further combined the overlapping degree data with ecological efficiency indicators to evaluate the spatial structure construction priorities of different land use components in the RGI system. The results show that the MSPA and ecological efficiency analysis method proposed in this paper are conducive to qualitative and quantitative analysis of the relationship between land use type, spatial structure, and ecological efficiency in the RGI system and are suitable for the construction of a green infrastructure network. This method can be used to better understand the spatial distribution and priority of green infrastructure networks to achieve sustainable rural development on the North China Plain.

1. Introduction

In recent years, the rapid urbanization process seen worldwide has led to an increasing demand for land resources, which not only causes changes in rural land utilization and cover but also triggers serious ecological crises [1,2]. More specifically, the change in area and spatial distribution for different types of ecosystems in the region resulted from the scale of and structural changes in land utilization, which affects the spatial structure of rural ecosystems, reduces the efficiency of rural ecosystems [3,4], and impedes the sustainable development of rural areas.

Through a review of the current situation with regard to urbanization, it can be seen that the rapid urbanization process is characterized by “Focusing on the Capital, Pursing the Profit Maximization, and Following the Value Characteristic of Local Government’s Unilateral Pursuit on GDP”, as it is still reliant on the sacrifice of rural space resources to “Geography-Physics” [5]. The transformation of highly intensive land development and utilization caused by the entry of highly polluting enterprises with high energy consumption and the expansion of rural cultivated land accompanied by deforestation and reclamation have jointly changed the land utilization composition, which is one of the main reasons for the change in the ecological environment [6]. The type, pattern, and intensity of land utilization have either a direct or indirect impact on ecosystem service functions [7,8]. The evolution of land utilization exerts an influence on the structure and function of ecological space, thus resulting in the erosion, fragmentation, and isolation of ecological patches. This has an immediate impact on sustainable development and the regional scenery pattern, which further affects ecosystem efficiency [9].

As a green space network composed of natural, semi-natural, and artificial scenery, rural green infrastructure (RGI) plays a significant role in the life support of an ecosystem [10], not only meeting the needs for development, but also preserving the natural rural environment. With the geometry and connectivity of image components as the target, geographic information system analysis was carried out to automatically identify the existing patches and corridors. Providing a new approach to identifying the key elements of a network structure, morphological space pattern analysis (MSPA) has been widely practiced in the construction, identification, and evaluation of GI networks over the past few years [11,12,13,14]. When conducting research on combining MSPA and GI, scholars have paid more attention to the structural connectivity, aesthetic function [15,16,17,18,19,20,21], planning, and design of GI [22]. Usually, they treat the GI system as an entire ecological space for analysis [20,23,24]; after analyzing GI with MSPA, all land use types in GI are often combined into one, and areas such as the core area and the bridge area are extracted and then analyzed as a new system. In addition, different forms of internal land utilization were insufficiently distinguished. This kind of analysis ignores the potential impact caused by the internal land utilization structure and the spatial structure. MSPA can clearly distinguish various spatial landscape elements in GI and has great potential in the differentiation and comparison of different spatial landscape elements. Therefore, this study of GI within different land use types and spatial landscape elements was carried out, introducing RGI spatial structure and land form as two aspects for the determination of the overlapping degree index and the comprehensive consideration and study of land use types, the ecological spatial structure, and ecological efficiency, because the inner link between these factors is advantageous to a multi-dimensional understanding of the change trend in the regional ecological environment. Moreover, this is of great significance to improving the cognitive perspective of relevant research on the rural ecological environment and the optimization and adjustment of regional land use and the ecological pattern.

As one of the important factors affecting GI [25,26,27] in the current GI evaluation system, the area proportion of each land type is closely associated with the ecological space composition of GI, determining the ecological efficiency of GI. The contradiction between rural construction and ecological environment protection is particularly prominent in urban areas with advanced urban–rural integration. With MSPA adopted to calculate the land utilization and space scenery overlap of RGI in different villages and towns, these data were used to quantitatively analyze the relationship between land utilization composition and spatial structure in an RGI system, with ecological efficiency as the primary goal of the article. To this end, ENVI, ARCGIS, and MSPA were used to extract the main land utilization elements and space scenery elements of RGI from remote sensing image data, and to perform superposition analysis on them, so as to obtain two types of overlapping degree data from each village and town. Afterwards, the correlation analysis of the overlapping degree index and ecological-efficiency-related indicators of each village and town was carried out to quantitatively and qualitatively analyze the relationship between land use type, spatial structure, and ecological efficiency in the RGI system. This allowed timely adjustments to be made to various land utilization and distribution types in the current RGI network system for sustainable development. In this study, Suining County, Xuzhou is taken as the study object, as a rapidly urbanized region in China and highly representative of the rapid-urbanization areas on the North China Plain. By introducing the overlapping degree of land utilization and space scenery, and comparing it between land utilization types and space scenery elements of the GI system in different villages, this study aims to solve the following problems: (1) What is the correlation between the ecological efficiency of RGI and its land utilization composition and spatial structure? (2) In the context of new urbanization, how can we determine the construction orientation of various land utilization and space scenery elements in RGI?

2. Materials and Methods

2.1. Study Area

Located in the north of Jiangsu Province, China, 33°40′ N–34°10′ N, 117°31′ E–118°10′ E, Suining County is a plain area in Xuzhou city. It covers an area of 1769.2 km², 1666.0 km² of which is plain area. There are a few low mountains and hills located in the northwest, west, and southwest, with an average altitude of 28.3 m. Suining County shows a slight slope from northwest to southeast, featuring a land climate in the form of a monsoon warm temperate zone, favorable natural conditions, four distinct seasons, an abundance of rainfall in summer, and a scarcity of rainfall and snowfall in winter. The location and jurisdiction of each town are shown in Figure 1, with all the villages and towns comprising Suining County (excluding Suicheng town) taken as the study object due to their high urbanization rate. On the one hand, Suining County has the typical characteristics of the North China Plain—flat terrain, convenient transportation, and numerous rivers and lakes. It has unique advantages in terms of land use and social and economic development. However, on the other hand, the North China Plain covers only 3 percent of China’s total land area and is home to nearly a quarter of the country’s population. The rural areas of the North China Plain are generally characterized as having a “large population but little land resources”, and a strong demand for the development of urbanization. Suining County is typical of this characterization. The following contradictions between its ecological environment and social development are becoming increasingly prominent due to the demand for development brought about by rapid urbanization:

• In recent decades, there has been a severe conflict arising between the economic benefits created by rapid urbanization and rural ecological benefits. The integrity and systematic nature of the local rural ecosystem have been jeopardized by insufficient ecological protection awareness, the highly fragmented RGI in Suining County, and the increasingly reduced and broken RGI matrix and connectivity between patches;
• Although the local land utilization composition has been seriously damaged as a result of the changed land utilization type caused by urbanization expansion, it still plays an important role in preserving the biodiversity and the connectivity of the rural ecological structure.

Considering the importance attached to rural land resource protection, the construction of an RGI network is essential for the process of rapid urbanization. The spatial data used in this study were sourced from the administrative division map of Suining County, Xuzhou in 2018, geographic information data, 2018 TM/OLI image data, and 30-m-levelled DEM data downloaded from a geographic space data cloud platform.

2.2. Method

The following works were conducted in this study. With the details shown in Figure 2, they will be introduced in the following sections:

• Firstly, the distribution of land utilization types was determined, and the constituent elements of RGI land utilization in the study area were extracted;
• Secondly, the space scenery elements of the RGI in the study area were extracted, and the core area, loop area, and bridge area according to MSPA were determined;
• Thirdly, overlapping analysis was conducted between space scenery elements and land component elements in RGI, so as to obtain two types of overlapping data: the Overlapping Degree of Space Scenery (ODSS) and the Overlapping Degree of Land Utilization (ODLU);
• Fourthly, the ecosystem service efficiency of RGI was analyzed;
• Fifthly, the correlation between land utilization composition and ecological efficiency of the RGI in Suning County was evaluated through correlation analysis by using overlapping data. Finally, unitary linear regression analysis was performed on the ecosystem service efficiency indicator and each overlapping degree indicator in order to further explore the deeper quantitative relationship between them.

2.2.1. Stage 1: Extracting the Land Composition Elements of RGI in the Study Area

The primary task in Stage 1 was to determine the type of land distribution in the study area. Through a combination of radiometric calibration, atmospheric correction, and the seamless mosaic and clipping of the TM/OLI image by ENVI5.1 data, the image data of Suining County were obtained and the different land types were identified by applying the multispectral overlapping and visual interpretation method. Then, the confusion matrix was adopted to ensure the accuracy of the classification results. Finally, the land utilization status map of Suining County was built. By revising the classification results through a field investigation and relevant data, a total of five categories of land utilization were identified: cultivated land, woodland, water body, bare land, and construction land.

In general, there are seven different land types in GI [28], including public green space, garden land, woodland, water body, cultivated land, grassland, and other green land. Located in the north of Jiangsu Province, cultivated land, woodland, and water represent the most important ecological resources in Suining County. Therefore, the woodland, cultivated land, and water bodies in the current land utilization map were extracted as the main elements of RGI land in the study area for analysis in ARCGIS, and were treated as the foreground of the binary grid map. Then, with the bare land and construction land as the background, the binary data were obtained and converted into geo-tiff format, thus building the binary grid map.

2.2.2. Stage 2: Extracting RGI Space Scenery Elements in the Study Area

Based on MSPA, the spatial structure of RGI was analyzed by using the binary image of land utilization presented by Guidos Toolbox software. Then, the scenery pattern was analyzed by identifying the spatial position of each foreground pixel (cultivated land, woodland, or water body). According to a previous summary [29,30] on the effectiveness of expert advice and the use of a corridor with different widths, eight neighborhoods with an edge width of 2 (60 m) were set in the resolution ratio. Based on the above practice, the spatial structure of RGI was divided into seven space elements, including the core area, bridge area, loop area, islet area, edge area, perforation area, and branch area. To maintain the connectivity of the RGI network, the core areas, loop area, and bridge area were extracted as important space scenery elements in this study.

The core area, loop area, and bridge area were extracted as three types of space scenery element of the RGI in the study area for analysis and were treated as the foreground of the binary grid map in ARCGIS. Taken as the background to form binary data, the other four types of elements were converted into geo-tiff format to build the binary grid map.

2.2.3. Stage 3: Analyzing the Overlapping Degree between Space Scenery Elements of RGI and Land Utilization Elements

In research combining MSPA and GI, many scholars have explored the impact of changes on the proportion of land utilization composition and space scenery elements in GI [5,23,31,32]. In addition, some scholars have introduced an overlapping area between land utilization elements and space scenery elements as an indicator to explore the interaction among them and the scenery indicator [24]. The overlapping area and its proportion between land utilization elements and space scenery elements require careful consideration to be given to the potential impact between them, which is one of the factors worth considering for RGI planning. In the present study, the town was taken as a unit to conduct overlapping analysis for three types of space scenery elements in the core area, loop area, and bridge area, as well as the three types of land utilization elements in the land utilization status map, respectively. In this way, the overlapping area and proportion can be obtained for the three types of space scenery elements and the three types of land utilization elements in each village and town. Furthermore, the overlapping degrees of both the space scenery and land utilization type were introduced into the study to analyze the overall status of ecological space in individual villages and towns.

(1)

The Overlapping Degree of Space Scenery (ODSS) is aimed at revealing the impact of the overlapping degree on ecosystem service efficiency from the perspective of the spatial structure. The formula of such an impact is expressed as follows:

$${\mathrm{ODSS}}_{\mathrm{i}-\mathrm{j}}=\frac{{\mathrm{S}}_{\mathrm{i}-\mathrm{j}}}{{\mathrm{S}}_{\mathrm{i}}}$$

(1)

where S_i−j represents the overlapping area between space scenery element type i and land utilization type j, and S_i refers to the area of the corresponding space scenery element. The value of ODSS ranges between 0 and 100%. The larger the value, the higher the proportion of the corresponding RGI land utilization type in space scenery elements of the village and town. For example, the overlapping degree of core cultivated land in a town is expressed as ODSS_{core-cultivated land} = S_{core-cultivated land}/S_core.

(2)

The Overlapping Degree of Land Utilization (ODLU) aims to demonstrate the impact of the overlapping degree on ecosystem service efficiency from the perspective of the land utilization structure. The formula of such an impact is expressed as follows:

$${\mathrm{ODLU}}_{\mathrm{j}-\mathrm{i}}=\frac{{\mathrm{S}}_{\mathrm{i}-\mathrm{j}}}{{\mathrm{S}}_{\mathrm{j}}}$$

(2)

where S_i−j represents the overlapping area between space scenery element type i and land utilization type j, and S_j denotes the area of the corresponding land utilization element. The value of ODLU ranges between 0 and 100%. The larger the value, the higher the proportion of the corresponding spatial space scenery elements in the RGI land type of the village and town, and the easier it is to convert such an RGI land utilization type into the corresponding spatial scenery elements. For example, the overlapping degree of core cultivated land in a town is expressed as ODSS_{cultivated land-core} = S_{core-cultivated land}/S_{cultivated land}.

2.2.4. Stage 4: Analysis of the Efficiency of the RGI Ecosystem

Affecting the efficiency of RGI ecosystems, the quality of the ecological space pattern of villages and towns is one of the key factors worth considering for the evaluation of RGI planning. The scenery pattern indicator can be used to reflect the pattern of ecological space and protection status. By introducing the common scenery pattern indicator into the evaluation of ecological space protection, the overall situation with regard to the ecological space of villages and towns was evaluated in this study [33,34,35], and the following three scenery pattern indicators of significant weight and relatively independent ecological meaning [5] were adopted to analyze the efficiency of RGI ecosystem services.

(1)

Landscape Shape Indicator (LSI)

In general, the LSI is intended to measure the complexity of the shape and evolutionary trend of ecological patches, the formula of which is expressed as follows:

$$\mathrm{LSI}=\frac{0.5\mathrm{L}}{\sqrt{\mathrm{S}}}$$

(3)

The value of the LSI exceeds 1 and there is no upper limit. L represents the perimeter of patches in the study area, and S refers to the total area of ecological patches in the study area. When there is only one regular patch, the LSI takes the minimum value of 1, and increases with the complexity of the patch shape and the degree of deviation from the regular shape.

(2)

Mean Patch Area (MPS)

The MPS is not only indicative of the patch density and difference in ecological space in the study area, but is also directly reflective of the average patch size. Different levels of plaques tend to have different MPS values. The lower the level, the lower the value of the MPS. The degree of fragmentation tends to be higher for those low-level patches.

$$\mathrm{MPS}=\frac{\mathrm{S}}{\mathrm{N}}$$

(4)

where S represents the total area of patches in the study area, and N refers to the total number of patches in the study area.

(3)

Contagion (CONTAG)

There is a space attribute associated with CONTAG, which is commonly used to indicate the agglomeration degree and contagion trend shown by different types of patches. The higher the value of the contagion degree, the greater the agglomeration degree of the main patch types in the study area and the higher the connectivity. Conversely, this suggests that the more complex the type of patch, the lower the degree of agglomeration for the same type of patch, and the higher the degree of scenery fragmentation.

$$\mathrm{CONTAG}=\left\{1+\frac{{{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{m}}{{\displaystyle \sum }}_{\mathrm{k}=1}^{\mathrm{m}}\left[{\mathrm{p}}_{\mathrm{i}}\left(\frac{{\mathrm{g}}_{\mathrm{ik}}}{{{\displaystyle \sum }}_{\mathrm{k}=1}^{\mathrm{m}}{\mathrm{g}}_{\mathrm{ik}}}\right)\right]·\left[\ln {\mathrm{p}}_{\mathrm{i}}\left(\frac{{\mathrm{g}}_{\mathrm{ik}}}{{{\displaystyle \sum }}_{\mathrm{k}=1}^{\mathrm{m}}{\mathrm{g}}_{\mathrm{ik}}}\right)\right]}{2\ln \left(\mathrm{m}\right)}\right\}\times 100$$

(5)

where p_i represents the proportion of patch type i to the total patch area in the study area, g_ik denotes the number of nodes between patch type i and patch type k based on the double method, and m represents the number of patch types, including the patch types on the scenery boundary. The value of CONTAG ranges between 0 and 100. A small value of CONTAG means that most of the scenery comprises small patches. Conversely, a large value of CONTAG suggests that there are main patches with closer connectivity in the scenery.

The ecological space scenery elements as obtained through MSPA were reclassified. The core area, bridge area, and loop area with high ecological value were classified as foregrounds, and other elements were classified as the backgrounds in ARCGIS. Then, they were inputted into Fragstats4.2, so as to calculate the scenery pattern indicator for the selected indicators [27].

2.2.5. Stage 5: Analysis of the Land Utilization Composition–Ecological Efficiency Correlation of the RGI

Correlation analysis was conducted to explore the relationship between the ecological efficiency and land utilization composition of the RGI network, and to quantify the impact of land utilization composition on ecological efficiency for the RGI system. In terms of correlation analysis, there are two main measurement variables involved:

(1)

ODSS represents the proportion of land utilization in all of the space scenery elements of RGI, revealing the composition of land utilization elements in the core area, bridge area, and loop area. ODLU represents the proportion of space scenery elements in all land utilization elements of RGI, reflecting the different roles that each type of land plays in the space scenery pattern and their ability to transform into different space scenery elements. With these two types of overlapping degree as dependent variables, data support can be provided for analyzing which types of land are more likely to become the components of patches, corridors, and other scenery structures, which is conducive to exploring the impact of changes in space scenery on ecological efficiency.

(2)

In terms of ecosystem service efficiency, LSI, MPS, and CONTAG were introduced. With size complexity, diversity, agglomeration degree, spread trend, and method of ecological patch measurement as variables, a study was conducted to reveal the internal relationship between land utilization types, ecological space structure, and ecological efficiency.

In SPSS, the correlation coefficient matrix thermodynamic diagram was first drawn by taking the two types of overlapping degree elements of each town as dependent variables, and the three indicators of ecosystem service efficiency as variables for multivariable regression analysis. Then, the combinations that passed the significance test in the multivariable regression analysis were adopted to further conduct univariate linear regression analysis on the ecosystem service efficiency indicators and each overlapping degree indicator. In this way, the internal quantitative relationship was further explored.

3. Results

3.1. Analysis of Space Scenery Elements of RGI Network MSPA in the Study Area

The results show that the MSPA of grid data as conducted by applying Guidos Toolbox is capable of quickly identifying the space scenery elements, including the core, loop, and bridge in the RGI of each town. On this basis, seven different space scenery elements were obtained for fifteen towns in Suining County (Figure 3).

Through the statistical analysis of MSPA results, the RGI composition status was obtained for different towns in Suining County (

Table 1. The proportion of three types of space scenery elements of MSPA in each town.

Area (ha)/Proportion	Proportion of Core	Proportion of Loop	Proportion of Bridge
Gaozuo	62.17%	5.55%	18.48%
Gupi	69.22%	4.57%	17.25%
Guanshan	76.78%	5.29%	9.93%
Lanshan	72.78%	4.26%	15.42%
Liji	77.12%	4.51%	9.28%
Liangji	58.23%	6.38%	22.35%
Lingcheng	67.31%	5.37%	17.73%
Qing’an	68.09%	4.53%	18.25%
Qiuji	69.70%	5.42%	17.03%
Shaji	60.28%	5.17%	20.12%
Shuanggou	70.16%	5.96%	10.98%
Taoyuan	72.85%	5.45%	13.11%
Wangji	68.70%	4.97%	16.38%
Weiji	70.08%	5.71%	15.36%
Yaoji	72.85%	4.92%	14.31%

), of which the core area represents the most significant component of the ecological matrix in RGI. Seven towns, including Liji Town and Guanshan Town, show a large proportion of core areas, exceeding 70%. The core area in Liangji town shows the lowest proportion, of only 58.23%, indicating the stark differences in the number of main ecological substrates and patches of RGI between different towns. Regarding the results of the bridge area, there is a large proportion in most towns. There are 10 towns, including Weiji Town and Lansha Town, reaching over 15%. Among them, Shaji Town and Liangji Town exceeded 20%, with only Liji Town, Guanshan Town, and Shuanggou Town showing a low level, of approximately 10%. The bridge area represents the main structure of the RGI corridor, with most of the bridge area distributed across the core area, indicating a high level of connectivity in most of the RGI in the villages and towns in Suining County. This creates more favorable ecological conditions for material exchange and energy flow. Notably, there is a slight difference in the proportion of loop area elements between different villages and towns, ranging from 4% to 7%.

As can be seen from Figure 3 and Table 1, there is a large proportion of core areas in the villages and towns across the study area, ranging from 58% to 78%. Among them, the villages and towns with a high proportion of core areas concentrate in the southwest and northwest of the study area, and there is a relatively low proportion in the east. The proportion of bridge area varies significantly, with the gap between the highest and the lowest being in excess of 140%. Specifically, the eastern and the central villages and towns show a high proportion of bridge area, and the southwest villages and towns show the lowest proportion.

Overall, these findings conform to a morphological distribution from the east to the northwest, which leads to a satisfactory overall scenery connectivity of the county from east to west. Despite the rich and superior ecological substrates and patches in the villages and towns in the southwest, they lack radiation to other villages and towns in the county, which opposes the flow of material and energy in the core area. In summary, the villages and towns in the west of Suining County are generally diverse in the ecological matrix and patches, whereas this is relatively insufficient in villages and towns in the east. There is a close scenery connectivity shown from east to west. The villages and towns in the southwest show diversity in the ecological matrix and patches, despite the insufficient radiation capacity.

3.2. Analysis of RGI Network Overlapping Degree in the Study Area

The results show that there are nine different types of ODSS and ODLU in the RGI network that can be obtained by analyzing the overlapping degree for three types of space scenery elements and three types of land utilization elements. In addition, the overlapped part was extracted to obtain the ratio of the overlapped part to the corresponding space structure and land utilization elements in ARCGIS. Then, data analysis was conducted for these results (

Table 2. The proportion of three types of space scenery elements of MSPA in each town.

ODSS	ODSScore-cultivated land (%)	ODSScore-woodland (%)	ODSScore-waters (%)	ODSSloop-cultivated land (%)	ODSSloop-woodland (%)	ODSSloop-waters (%)	ODSSbridge-cultivated land (%)	ODSSbridge-woodland (%)	ODSSbridge-waters (%)
Gaozuo	66.71	33.11	0.18	68.95	30.65	0.4	69.84	29.78	0.38
Gupi	66.36	29.92	3.73	65.42	33.80	0.78	65.64	33.44	0.92
Guanshan	79.99	19.76	0.25	72.00	27.42	0.59	73.6	25.35	1.04
Lanshan	67.91	32.07	0.02	65.31	34.59	0.10	68.79	31.11	0.09
Liji	75.47	24.51	0.02	71.48	28.45	0.07	71.82	27.95	0.22
Liangji	67.83	32.14	0.03	65.1	34.73	0.17	70.13	29.73	0.14
Lingcheng	83.67	16.09	0.24	74.35	25.50	0.15	76.00	23.66	0.34
Qing’an	70.97	21.56	7.48	71.76	27.89	0.35	71.85	27.75	0.39
Qiuji	83.04	16.80	0.16	75.53	24.33	0.14	75.99	23.61	0.40
Shaji	75.78	24.08	0.14	70.59	29.16	0.25	69.76	29.82	0.42
Shuanggou	74.82	23.55	1.63	70.87	28.61	0.52	72.09	26.78	1.13
Taoyuan	73.49	26.47	0.04	68.80	31.04	0.16	71.76	28.08	0.16
Wangji	79.57	19.69	0.74	71.34	28.50	0.16	73.06	26.74	0.19
Weiji	67.83	31.73	0.44	66.16	33.48	0.36	62.05	35.72	2.23
Yaoji	70.32	25.55	4.13	72.11	27.08	0.82	72.20	26.66	1.14

and

Table 3. The proportion of three types of space scenery elements of MSPA in each town.

ODLU	ODLUcultivated land-core (%)	ODLUcultivated land-loop (%)	ODLUcultivated land-bridge (%)	ODLUwoodland-core (%)	ODLUwoodland -loop (%)	ODLUwoodland-bridge (%)	ODLUwaters-core (%)	ODLUwaters-loop (%)	ODLUwater body-bridge (%)
Gaozuo	66.71	33.11	0.18	68.95	30.65	0.4	69.84	29.78	0.38
Gupi	66.36	29.92	3.73	65.42	33.80	0.78	65.64	33.44	0.92
Guanshan	79.99	19.76	0.25	72.00	27.42	0.59	73.6	25.35	1.04
Lanshan	67.91	32.07	0.02	65.31	34.59	0.10	68.79	31.11	0.09
Liji	75.47	24.51	0.02	71.48	28.45	0.07	71.82	27.95	0.22
Liangji	67.83	32.14	0.03	65.1	34.73	0.17	70.13	29.73	0.14
Lingcheng	83.67	16.09	0.24	74.35	25.50	0.15	76.00	23.66	0.34
Qing’an	70.97	21.56	7.48	71.76	27.89	0.35	71.85	27.75	0.39
Qiuji	83.04	16.80	0.16	75.53	24.33	0.14	75.99	23.61	0.40
Shaji	75.78	24.08	0.14	70.59	29.16	0.25	69.76	29.82	0.42
Shuanggou	74.82	23.55	1.63	70.87	28.61	0.52	72.09	26.78	1.13
Taoyuan	73.49	26.47	0.04	68.80	31.04	0.16	71.76	28.08	0.16
Wangji	79.57	19.69	0.74	71.34	28.50	0.16	73.06	26.74	0.19
Weiji	67.83	31.73	0.44	66.16	33.48	0.36	62.05	35.72	2.23
Yaoji	70.32	25.55	4.13	72.11	27.08	0.82	72.20	26.66	1.14

).

In terms of ODSS, cultivated land usually accounts for the absolute majority of space scenery elements of RGI in Suining County, followed by woodland, with the overall proportion of water bodies being low. There is a high degree of overlap between cultivated land and the three main space scenery elements, ranging from 65% to 84%. On the one hand, this suggests that cultivated land plays an extremely important role in three main space scenery elements. On the other hand, this implies a lack of diversity in the composition of the main space scenery elements of villages and towns in Suining County, which means an inadequate ability to withstand ecological risks. The distribution range of ODSS_{loop-woodland} and ODSS_{bridge-woodland} shows a relatively high level of concentration, while ODSS_{core-woodland} varies significantly. Lingcheng Town is the lowest, reaching only 16.09%, while Gaozuo Town is the highest, reaching 33.11%. That is to say, the gap is more than doubled, suggesting the relatively balanced woodland composition of the main ecological corridor and the internal connection zone of the matrix. By contrast, there is a significant difference in the proportion of woodland in the main ecological matrix of each village and town. Except for Qing’an Town, with reservoirs, and Yaoji Town, with an extensive watershed distribution, the space scenery overlapping degree of water bodies in other villages and towns falls below 1% in most cases. That is to say, the proportion of water bodies is low in the main space scenery elements of each village and town, which indicates room for significant improvement.

In terms of ODLU, there is usually a low overlapping degree between the three main space scenery elements and the woodland of RGI in Suining County, despite a slightly higher degree of overlapping with a water body compared with woodland. Cultivated land remains the most common land form in most villages and towns. However, in terms of the overlapping degree of the bridge area, the ODLU_{waters-bridge} of nearly half of the villages and towns shows no significant difference or even reaches a higher level than ODLU_{cultivated land-bridge}. There is also a persistently low degree of overlapping between woodland and the three types of space secrecy elements in most villages and towns. In only three towns, namely Lanshan, Liangji and Taoyuan, the degree of bridge overlapping exceeds 10%, indicating the low conversion rate of woodland for the three types of main space scenery elements. In addition, only a few villages and towns play a relatively important role in the ecological corridor. There remains a high overlapping degree between cultivated land and the three main space scenery elements. However, the ODLU_{cultivated land-bridge} is lower than 50% in seven towns, including Liangji and Liji. In addition, their ODLU_{waters-bridge} is in a leading position, with a maximum of more than 45%, indicating the relatively low conversion efficiency of cultivated land as a bridge area in nearly half of the villages and towns, despite the relatively high conversion rate of water as a bridge area. Compared with the cultivated land, water bodies show a slight difference of less than 8 percentage points, or a higher conversion rate of the bridge area, in six towns, including Weiji and Shaji. The ODLU_{waters-bridge} in Liangji and Liji demonstrates the greatest advantage over the ODLU_{cultivated land-bridge}, with the gap reaching 45.36% and 55.78%, respectively. It can be seen from that mentioned above that, even though the cultivated land represents the main body in terms of the land utilization composition in Suining County, water bodies continue to show a similar or even higher conversion rate in nearly half of the villages and towns after transformation into the main ecological corridor, indicating that water bodies play a crucial role in the composition of the ecological corridor.

3.3. Analysis of the Service Efficiency of the RGI Ecosystem in the Study Area

According to the results, the service efficiency level of the RGI ecosystem in each village and town can be efficiently analyzed through FRAGSTATS to calculate the scenery pattern indicator of the selected indicators. Then, data analysis of the results can be conducted to obtain the indicator status of the RGI scenery pattern in different villages and towns in Suining County (

Table 4. RGI scenery pattern indicator of each town.

	LSI	MPS (ha)	CONTAG%
Gaozuo	93.27	0.34	38.73
Gupi	119.17	0.37	39.03
Guanshan	97.68	0.44	49.6
Lanshan	140.73	0.36	41.70
Liji	89.35	0.38	45.66
Liangji	109.68	0.32	38.11
Lingcheng	84.82	0.36	49.61
Qing’an	128.18	0.37	40.33
Qiuji	113.06	0.36	50.61
Shaji	89.14	0.33	41.28
Shuanggou	101.81	0.38	42.11
Taoyuan	107.6	0.37	44.18
Wangji	125.42	0.35	45.00
Weiji	126.56	0.36	41.64
Yaoji	148.98	0.36	41.58

).

The vast majority of the villages and towns studied show a high level of LSI. A total of ten towns exceed 100, of which five exceed 120. Comparatively, Yaoji Town and Lanshan Town exceed 140, while Shaji Town and Lingcheng Town show a low level of LSI, being below 100, indicating a relative diversity in the morphology and distribution of ecological patches in each town, despite some unevenness.

The difference in the MPS indicator among different towns was found to be clearly insignificant, ranging from 0.32 to 0.38. Guanshan Town was the only location in which a significantly higher level was reached (0.44). This suggests a high degree of homogeneity in the level of patch fragmentation within the RGI of villages and towns.

According to the CONTAG indicator, the overall situation of all towns remains relatively balanced. The contagion indicator of each town falls within the range of 38–51%. Among all of the towns studied, Qiuji Town performed best, reaching the highest level of 50.61%. By contrast, the performance of Liangji Town, Gaozuo Town, and Gupi Town was relatively poor, falling below 40%, which indicates that the connectivity of the main ecological patches of RGI in each village and town was relatively balanced, and the ecological space network shows close connectivity.

3.4. Analysis of the Correlation between Land Utilization Composition and Ecological Efficiency of the RGI Network in the Study Area

Through a regression analysis in SPSS, it was found that, when LSI and MPS are taken as variables, the dependent variables fail the significance test, regardless of ODSS or ODLU (after the adjustment, R2 ≪ 1, significance ≫ 0.1). Therefore, it can be judged that there is no significant correlation between LSI and MPS and the two overlapping degree indicators. When CONTAG is taken as the independent variable and the two overlapping degree indicators are treated as the dependent variables, the analytical results pass the significance test with a high level of significance. After the adjustment, R2 reaches 0.808 and 0.946, respectively, while the significance is far less than 0.1. This leads to a significant correlation between CONTAG and the two overlapping degree indicators. By taking account of the analytical results obtained by taking the regression of the CONTAG indicator as a variable, the following conclusions can be reached.

(1)

Based on the analysis of the space scenery overlapping degree (Figure 4), it can be seen that there is a significant positive correlation between the size of the CONTAG indicator of the three scenery elements of core area, loop area, and bridge area and the overlapping degree of cultivated land, reaching +90.7%, +73.9%, and +67.3%, respectively. There is also a strong and weak negative correlation between woodland and water bodies and the overlapping degree of cultivated land, respectively. According to the above results, the higher the proportion of cultivated land within the three main space scenery elements of RGI, the greater the degree of contagion. That is to say, there are dominant ecological patch types with close connectivity in the village, with a higher level of connectivity shown by the dominant patch.

(2)

Based on the analysis of the overlapping degree of land utilization types, it can be seen (Figure 5) that there is not only a significant positive correlation between the CONTAG indicator and the overlapping degree of the core area, but also an insignificant positive correlation with the overlapping degree of the loop area within the cultivated land type. This suggests that within the RGI system, it can play a greater role in improving the ecological service efficiency if more cultivated land can be transformed into the core matrix and the internal connecting zone of the matrix. From the perspective of woodland type, there is a significant positive correlation between the overlapping degree of the CONTAG indicator and the loop area. That is to say, in terms of improving the overall efficiency of the ecological service for villages and towns, the woodland is more extensively transformed into the internal connection zone of the matrix, with a better performance produced given the diverse internal land composition of the main ecological matrix. From the perspective of water body type, however, there is an insignificant positive correlation between the overlapping degree of the CONTAG indicator and the bridge area, and significant negative correlations between the overlapping degree of the core area and loop area. That is to say, if more water bodies can be transformed into the main ecological corridor, then they can play a better role in the improvement of the ecological service efficiency. However, a poor ecological service efficiency will result from the transformation of more water bodies into main substrates, patches, and corridors.

In conclusion, cultivated land can create more ecological value in terms of concentrated and contiguous areal space elements. However, a fragmented scenery distribution plays no role in strengthening value supply. The smaller the proportion of fragmented cultivated land in villages and towns, the better the outcome. Woodland can create more ecological value only when it is transformed into the internal connecting zone of the matrix to diversify its land utilization composition. Therefore, the suggestions for woodland optimization are focused on how to improve the unitary composition in the main ecological matrix and supplement its internally fragmented area. When zonal space elements are dominant, more water bodies are transformed into the main ecological corridor, which increases ecological value. However, an areal distribution and a fragmented point distribution exert no promoting effect on value supply. Therefore, it has been suggested to strengthen the construction of banded watersheds and enhance the connectivity of water bodies in the following optimization, with a limit imposed on the fragmentation and areal elements of artificial interventions as appropriate, including fishponds and reservoirs.

(3)

By taking the value of CONTAG as the independent variable and each overlapping degree as a dependent variable, and considering the univariate linear regression analysis, it can be seen that there is a significant linear correlation when the three overlapping degree indicators, including ODSS_{core-cultivated}, ODSS_{core-woodland}, and ODSS_{loop-cultivated}, are taken as the dependent variables. After the adjustment, R2 reaches 0.808, 0.656, and 0.512, respectively, with the significance falling below 0.01. Therefore, a significant correlation exists between CONTAG and these three overlapping degree indicators. The analysis of the results obtained from linear regression is shown in Figure 5. The expression is presented as follows:

CONTAG = 0.62ODSS_{core-cultivated land} − 2.35

(6)

CONTAG = 0.916ODSS_{loop-cultivated land} − 20.822

(7)

CONTAG = −0.577ODSS_{core-woodland} + 20.822

(8)

It can be seen from the above that there is a significant positive correlation between the value of CONTAG and both ODSS_{core-cultivated land} and ODSS_{loop-cultivated land}, with the positive correlation coefficient reaching 0.62 and 0.547, respectively. However, there is also a significant negative correlation between the value of CONTAG and ODSS_{core-woodland}, with the negative correlation coefficient reaching −0.577. Therefore, taking Suining County as an example, the centralized and large-scale construction of cultivated land shall be strengthened in future optimization and construction operations in order to improve ODSS_{core-cultivated land} and ODSS_{loop-cultivated land} to a higher level than 76.38% and 71.86%, respectively. Meanwhile, it is necessary to carefully consider woodland construction that aims to supplement ecological vacancies and restrict ODSS_{core-woodland} to less than 22.13% as much as possible. On this basis, it is possible to improve the ecosystem service efficiency in the local area. The value of CONTAG exceeds 45%, ranking among the top 30% in local villages and towns.

3.5. Optimization and Adjustment Strategy of Land Utilization and the Ecological Pattern of the RGI Network in the Study Area

In conclusion, the following optimized construction of cultivated land shall be carried out in a centralized and large-scale manner, and cultivated land shall be utilized as much as possible to fill the vacancy in the main ecological matrix. The smaller the proportion of fragmented cultivated land in villages and towns, the better the results. As for the proportion of cultivated land in the main ecological substrates and connecting zones, it is predicted to reach 76.38% and 71.86%, respectively. It is also necessary to focus the construction of woodland on how to improve the unitary composition for the main body of the ecological matrix and supplement its internal fragmented areas. However, a reasonable limit shall be imposed on such supplementary construction, and the proportion of woodland in the main ecological matrix shall be restricted to a level lower than 22.13%. As for the optimization and construction of water bodies, it is necessary to strengthen the construction of banded watersheds and enhance the connectivity of water bodies. Those villages and towns with water bodies shall be encouraged to rely on water bodies to create more diverse recreational and agricultural spaces, provide better ecological and regulatory services, partially reduce manual intervention into natural water bodies, and apply a limit on fragmented and areal elements as appropriate, including fish ponds and reservoirs.

4. Discussion

4.1. Analysis of Research Conclusions

Currently, agriculture remains the dominant system for the development of villages and towns in Suining County, with cultivated land accounting for a vast proportion of the total area. The state of large-scale construction and ecological utilization is satisfactory. Cultivated land has become the most significant component of the three types of space scenery elements, which has led to an increase in the proportion of cultivated land in the main ecological corridors, ecological substrates, and their internal connection zones, thus improving the ecosystem service efficiency, as suggested by the multiple regression analysis of ODSS–CONTAG. The negative correlation between the overlapping degree of woodland and water bodies and the CONTAG indicator is attributable to the relatively small scale of land utilization, low utilization efficiency, and the fragmentation of the spatial distribution. In addition, the woodland and water bodies play a major role in providing regulation services. They are outperformed by cultivated land in terms of supply service capability, which may explain the negative correlation obtained in the regression analysis of ODSS.

The value of CONTAG was taken as the independent variable, and the six overlapping degrees of water bodies in ODSS and ODLU were treated as dependent variables in order to perform the consistency test, with the results obtained from univariate linear regression analysis showing no significant correlation between the overlapping degree of water bodies in the RGI network and the value of CONTAG in Suining County. To a large extent, this results from the fact that the water bodies in Suining County are subjected to artificial interventions, thus resulting in the unevenness of the spatial structure distribution. Different from woodland and cultivated land, the spatial structure of water bodies varies significantly among villages and towns. There is usually a lack of sufficient scientific unified planning guidance for the construction of surface elements and fragmentation elements, including reservoirs and fishponds, which are significantly affected by human intervention. Therefore, with Suining County as an example, it is difficult to further quantify the relationship between the overlapping degree of water bodies and ecosystem efficiency.

4.2. Implications and Limitations of Research Methods

In the present study, the correlation between land utilization composition and ecological efficiency of RGI in Suining County was transformed into an intuitive thermal map, with the relevant indicators of “overlapping degree” and “ecosystem service efficiency” adopted as quantitative indicators to evaluate the priority of RGI construction in terms of land structure. The multivariable regression analysis carried out in the study was conducive not only to exploring the internal relationship between land utilization composition, ecological space structure, and ecosystem efficiency, but also to understanding the evolutionary trend of regional ecological environment from various angles.

There are, however, also some limitations to the study. Firstly, all the villages and towns taken as the research object of this study are located in Suining County, which constrains the generalizability of some of the analytical results and conclusions. Secondly, the selected ecosystem service efficiency indicators are not fully representative. It is thus necessary to reach more comprehensive conclusions through using a better method to study more cases and introduce diversified ecosystem service efficiency indicators, especially for generating further quantitative conclusions related to water bodies. Finally, in previous studies where the MSPA method is used, the main analysis is focused on the extraction of the core area and bridge area. In this study, however, the loop area was also included as a research object based on the ecological service efficiency of providing migration corridors for species in the patch. In future studies, it will also be necessary to pay attention to other space scenery elements, especially those mainly composed of fragmented ecological corridors, which can be improved and incorporated into the branch areas of the main ecological corridors.

4.3. The Guiding Significance of Research Results for Decision-Makers

The Returning Farmland to Forest Program (RFFP) has been implemented in China for more than two decades to ameliorate the adverse effects of the long-term, relatively extensive rural production methods, but its effectiveness is highly controversial. Researchers and officials have claimed that many forests are of inferior quality [36], or that there are problems with regard to compensation for land taken out of cultivation [37]. The RFFP does not significantly optimize the status of local ecosystems [38], and may even have unexpected impacts on cultivated land [39]. In recent years, with the impact of high-speed urbanization, China has vigorously promoted the intensification of rural land construction, which has brought about a large number of changes in land use types. In addition, to cope with the potential food crisis caused by COVID-19, China has started to implement a new policy of returning forests to farmland. These two points will inevitably lead to a lack of environmental consideration in future construction, and the main research purpose of this paper is to provide valuable suggestions for the spatial layout of land planning, so as to avoid the risks of ecological environment destruction and landscape fragmentation as much as possible.

From the findings of this paper, we highlight the following observations and suggestions for rural RGI development. We should recognize that the large-scale construction of cultivated land is a good guide for the development of plain agriculture and an effective measure to improve landscape fragmentation, and that “isolated” areas, such as wastelands within major ecological substrates, are good options for complementing and expanding existing cultivated land and woodland, allowing us to achieve both economic and ecological benefits. Ecological protection strategies should focus on land planning measures to protect cultivated land that has formed a broad ecological matrix, to as great a degree as possible. In terms of ecology, the larger the area, the higher the value of the cultivated land. Although final decisions on future rural landscape patterns will largely depend on land use regulations and the objectives of the relevant authorities, the analytical methods used in this study are useful in identifying specific optimization options for construction.

Land planning and policy must be evidence-based. Our methodology ultimately supports the adjustment and integration of the two elements in the Suining County RGI. From this point of view, the research results should be considered in the choice of future planning tools for rural development on the North China Plain, such as the Suining County Territorial Spatial Planning tool. Agricultural areas identified as more vulnerable to future urban shocks could be incorporated into conservation systems to at least prevent land use changes. Finally, the degree of protection of these areas can also be determined through the participation of stakeholders in the master plan update.

5. Conclusions

With Suining County as an example, the correlation between land utilization composition and ecological efficiency of the local RGI network was analyzed by adopting the MSPA model, introducing overlapping degree elements, and taking account of indicators of ecosystem service efficiency. The results show the following. First, in the internal RGI system of Suining County, ecosystem service efficiency can be improved by increasing the proportion of cultivated land in the main ecological corridors, ecological substrates, and their internal connecting zones. Second, cultivated land can create more ecological value in the concentrated and contiguous areal space elements, with its value supply being affected in the fragmented scenery distribution. Third, forest land can play a more effective ecological role when it is transformed into the internal connecting zone of the matrix in order to diversify its land utilization composition. Fourth, when zonal space elements are dominant, more water bodies can be transformed into a main ecological corridor, which increases ecological value. However, its value supply can be negatively affected by the distribution of concentrated and contiguous area elements and the distribution of fragmented point scenery.

As revealed by the case study, the described method is effective in revealing the correlation between land utilization types, the overlapping degree of space scenery, and ecosystem service efficiency in an objective and efficient way. In addition, suggestions can be made for the optimization of the construction of the RGI by considering the analysis results. Therefore, the method is applicable to providing effective technical support for RGI planning and construction, which are essential for the implementation of a sustainable rural development strategy.