Model construction and simulation analysis of ecological corridors in landscape environment planning

In this study, the shortest-circuit algorithm was used to extract potential ecological corridors between conservation priority areas. The response and feedback effects of ecological processes such as migration of species between protected areas and movement of material and energy flows to different land use types on the landscape surface can be quantified by the magnitude of resistance. Ecological processes can actually be understood as movements that overcome spatial resistance, which is measured by accessibility, cost distance, and minimum cumulative resistance. On land, resistance does not only refer to the length of distance in space, but landscape cover type and character are also important influences. Different land use types have different impacts on species movement and ecological flow movement, specifically, species are more likely to migrate successfully through wetlands or woodlands, and are at greater risk when passing through urban centers. Quantifying the magnitude of the dangers faced by material-energy flows moving through different media as a value of apparent impedance, material-energy flows in ecosystems have a greater chance of moving from the surface of media with lower impedance and a smaller chance of completing their flow to their destination in media with high impedance, and, in addition, species spontaneously move from the surface of landscapes with low impedance, due to their instinct to avoid hazards. This paper builds on this theoretical foundation to identify potential ecological corridors between conservation priority areas in landscape environmental planning.

The shortest path algorithm [26] is an algorithm for solving the shortest path problem, denoting accessibility, which in the context of species conservation refers to the resistance that a species needs to overcome in order to cross the surface of different types of landscape media from its “source”. A nature reserve consists of three parts: the core zone, the buffer zone and the experimental zone. Material energy or a species has the least resistance to movement within the core zone, followed by the buffer zone, and the experimental zone is the most difficult to move. To determine the shortest path, three key factors need to be found, which are the “source”, the impedance surface and the resistance value. The “source” is the origin of the diffusion of a substance or the migration of a species, and can be representative of the needs of a substance or an organism for its environment. When selecting the impedance surface, attention should be paid to the principle of quantifiable, try to find out all the factors that may produce impedance to the movement process, and assign the value to each factor according to the size of its impedance. The impedance values reflect the magnitude of resistance to species movement caused by different types of landscape surfaces by assigning weights.

The most critical problem in identifying ecological corridors is how to determine the minimum resistance path for material and energy flow, and the optimal path can be found through the shortest path algorithm. The shortest circuit algorithm is a kind of optimization algorithm, which has been widely used. For example, some related studies have proposed specific steps and methods to realize the functional zoning of the Old County Giant Panda Nature Reserve based on the shortest-circuit algorithm. Or, through the shortest-circuit algorithm, the functional zoning of giant panda nature reserve group is carried out by combining the habitat evaluation, and the procedures and methods for optimizing the functional zones of Qinling mountain system nature reserve group are elaborated. In this paper, the shortest circuit algorithm is used to abstractly calculate the flow cost of material and energy flow between patches, and the minimum value area of which is regarded as the optimal corridor or connection.

According to the laws of material flow or species movement, no material energy moves along only a single predefined optimal path, so spatial structural uncertainty needs to be taken into account. At the same time, ecological corridors also have a certain degree of redundancy, because both the optimal and the remaining suboptimal paths between the “source” and the destination may provide potential sites for the movement of matter. In view of these two characteristics of ecological corridors, this paper identifies ecological corridors between two conservation priority areas based on the following computational method.

For any study landscape, it can be considered as an impedance surface, with S(Source) and D(Destination) denoting the “source” and destination of matter movement, respectively, and R(Random) representing any point (not coinciding with S and D) on the surface of the landscape [27]. The minimum cumulative impedance for a substance to move from S to D (via R) on this surface is clearly equal to the resistance of moving first from S to R plus the resistance of moving from R to D, i.e.: (2) Cum_Cost(s→RD)=Cost(S→R)+Cost(R→D)

Since the movement of substances in an ecosystem is bi-directional and equivalent, that is, the magnitude of the resistance to moving from R to D is equal to the resistance to moving from D to R. Therefore, the minimum cumulative impedance for a substance to move from S through R to D on the surface of this landscape is equal to the sum of the resistance values for the substance to move from S and D to R, respectively. That is: (3) Cost(R→D)=Cost(D→R) (4) Cum_Cost(S→RD)=Cost(S→R)+Cost(D→R)

cost1, cost2 represent the minimum cumulative impedance from patch 1 and patch 2 to all points on that surface, respectively. The raster summation operation of cost1 and cost2 in ArcGIS, i.e., cost1+2, refers to the minimum cumulative impedance value between patch 1 and patch 2, the minimum of which is an optimal path between patch 1 and patch 2. Since material movement and species migration have spontaneity, their choice of paths is not unique, so some sub-optimal paths may also become potential ecological corridors. In this paper, the areas in cost1+2 that are less than or equal to a certain value (a specific threshold) are selected, i.e., the corridor between patch 1 and patch 2 is obtained. The red solid line takes the width of a single raster and indicates the optimal path from patch 1 to patch 2, which has the smallest cumulative impedance value, while the surrounding black line contains part of the suboptimal path („ a certain threshold value). In practice, the optimal path or the concatenation of the optimal and suboptimal paths can be selected by choosing the appropriate threshold according to the specific situation of the study area and different needs.

The ecological patch layer contains multiple (more than two) objects, the attribute records of the layer are sorted according to the objects, and its data type is the surface data in vector, which is managed and processed by vector data processing tools. Ecological patches are areas with important ecological functions or high sensitivity and other areas that need to be protected in a key way, as a source and connecting target point for establishing ecological networks.

The objects in the resistance factor layer are divided into multiple types, and the attribute records of the layer are sorted according to the type, and its data type is the surface data in vector, which is managed and processed by the vector data processing tool. Resistance factors are the factors that have a hindering effect on the ecological connection, such as the land use type, the distribution of transportation roads on land, and the marine functional zoning in the ocean.

The main function of the patch segmentation module is to separate the patches in the input ecological patch layer, i.e., each patch is separated into a layer file, and the record is stored in the newly created patch folder, which is provided to be called for the establishment of costly and least-cost paths.

The main function of the resistance assignment module is to assign different resistance values to the input resistance layers according to the different types of factors, which is the degree of obstruction when establishing ecological connections between patches. The completed resistance layer is converted into raster data for the establishment of cost and cost direction.

The main function of this module is to calculate the minimum cost path from the source patch to the target patch. The minimum cost path starts from the source patch and ends at the target patch, the width of the path is one image element, measured in each cost unit, and the cost direction is used to determine the minimum cost path between two two patches (sources).

The ecological corridor network module is to vectorize the minimum cost paths between ecological patches, splice them to form an ecological network, and output them in the form of a vectorized linear network. The ecological node module is to identify the intersections of the paths in the ecological network (intersections are bottleneck points in the ecological network) and output them as point data.

The planning area of Hanjiang Eco-city covers “seven towns on both sides of a river” in Shiyan Utopia, including the whole area of Chadian Town, Qingshan Town, and part of Chengguan Town, Liupi Town, Yangxipu Town, Tanjiayuan Town and Anyang Town, with a planning area of about 894.23km². Hanjiang Eco-city has a good ecological background, dominated by woodland, scrubland and wetland types, with the main stream of the Han River and tributaries such as the Shending River, Si River, Weir River, Banghao River, Liujiagou, Yangjia River, Juyu River, Xiayu River and Jujia River. The area of the ecological red line is 215.24 km², accounting for 24.07% of the area of Hanjiang Eco-city. It includes the Han River Gengjia Yazi Water Source Primary Reserve, Tanjiwan Reservoir Water Source Primary Reserve, Qinglongshan Dinosaur Egg Fossil Complex National Nature Reserve, Hubei Danjiangkou Reservoir Area National Wetland Nature Park, and Shiyan Utopia Lake National Wetland Nature Park.

The digital elevation model (DEM) in this study was obtained from the Computer Network Information Center of the Chinese Academy of Sciences (Geospatial Data Cloud Platform), with a spatial resolution of 30 m. The land use data were obtained from the National Science and Technology Basic Conditional Platform-National Earth System Science Data Sharing Platform-Yangtze Delta Science Data Center. Normalized vegetation index (NDVI) was obtained from the Computer Network Information Center of the Chinese Academy of Sciences (Geospatial Data Cloud Platform MOD13Q1 dataset, collected from January 2023 to December 2023, with a spatial resolution of 200 m. The annual average was calculated based on the 12-month NDVI products. The nighttime lighting data were obtained from NOAA National Oceanic and Atmospheric Administration (NOAA) National Centers for Environmental Information (NCEI) of the United States of America, with the type of stabilized lighting imagery and the spatial resolution of 1km × 1km.The data of geologic hazard potential and dangerous points were obtained from the Geologic Hazard Prevention and Control Program of Hanjiang Eco-city for the year of 2023, and were obtained by vector digitization.The data of PM2.5 were obtained from the General Environmental Monitoring Station of China (GEMS), and the collection time was from January 2023 to December 2023. Based on the daily average PM2.5 concentration values of the seven state-controlled air quality monitoring points in Hanjiang Eco-city (the whole area of Chadian Town and Qingshan Town and Chengguan Town, Liupi Town, Yangxipu Town, Tanjiayuan Town and Anyang Town), the monthly average concentration values were calculated. In turn, the annual average PM2.5 concentration values of the seven monitoring sites were summarized. After a series of data preprocessing work, such as projection transformation, cropping, resampling, etc., the above data were uniformly converted into 100m × 100m raster data.

Based on the circuit theory model to identify the ecological land in the Hanjiang Eco-city area, the results of the analysis of land use types with different ecological land importance levels in the Hanjiang Eco-city were obtained as shown in Figure 2. The comprehensive evaluation results of ecological land importance showed that the area of very important ecological land was 352.36 km², with an area share of 39.41%. The land use types in the area are dominated by water bodies, cultivated land and garden land. Among the ecologically important land in rank 4 and 5, the area of garden land, water bodies, and cultivated land in Hanjiang Eco-city accounted for 22.54% and 21.36%, 32.41% and 26.54%, 26.94% and 34.79%, respectively. The area of extremely important ecological land in water resources security is about 86.42 km², mostly important water systems and large reservoirs in the study area, mainly including the main stream of the Han River and the Shending River, Si River, Weir River, Gucheng Lake, Jinniu Lake, Convenience Reservoir, Zhongshan Reservoir and other areas. The region has sufficient surface water, which is an important water conservation function area for storing surface runoff in the region and preventing water and drought disasters. In addition, the extremely important ecological land area for biodiversity conservation in the Han River Eco-city is 156.32 km², which is mainly located in the hilly areas with a cluster of mountains. These areas have sufficient light and heat conditions and diverse landscape types, making them key areas for biodiversity conservation.

Figure 2.

The results of the importance of different ecological lands

The ecological patches with an area larger than 4 km² and distributed in a concentrated and continuous manner were selected as the ecological source sites for the construction of ecological corridors in the Hanjiang Eco-city, and 21 ecological source sites were finally obtained in the study area, with an area of 100.15 km², which accounted for 1.12% of the total land area of the whole area of the Hanjiang Eco-city. The land use type composition analysis of the ecological source land in Hanjiang Eco-city is shown in Table 1. In terms of land use type composition, the identified ecological source land in Han River Eco-city is dominated by water system, followed by forest land and garden land, with the proportion of the area to the total area of the ecological source land being 69.42%, 24.96% and 3.96%, respectively. Superimposing the ecological source areas with the protected areas in the national ecological red line of Hanjiang Eco-city, it can be seen that the ecological source areas basically cover the national and provincial nature reserves within the ecological source areas, and the identification results are reasonable.

In this study, in order to scientifically assess and rank the relative importance of core patches in ecological source site selection, two indices, the Index of Overall Connectivity (IIC) and the Index of Probable Connectivity (PC), were introduced for in-depth analysis. All cores were ranked according to their combined importance dL (average of dIIC and dPC) in the overall network. The connectivity analysis was performed using Conefor 2.6 software. In this study, ecological patches with a core area greater than 4 were selected as the study object, and detailed landscape connectivity calculations and analyses were performed on the identified 21 ecological patches using Conefor 2.6 software. Based on the calculation results, the ecological patches with a comprehensive importance score dL higher than the threshold value of 2 and ranked in the top 11 were identified as important ecological source sites of the core, while the remaining patches were regarded as general ecological source sites. The results of the landscape connectivity analysis of the important ecological source sites in the study area of Hanjiang Eco-city are shown in Table 2, which reveals that the distribution of ecological source sites in the study area is not homogeneous after in-depth exploration of the connectivity of these important ecological source sites. In particular, ecological source area number 3 occupies 19.70% of the total area of all ecological source areas with its significant area advantage, becoming the largest ecological patch. The ecological patches 3 (dL=59.62) and 17 (dL=45.06), centered on the Han River Gengjiayazi Water Source Primary Reserve and Hubei Danjiangkou Reservoir Area National Wetland Nature Park, together constitute a large ecological plate with a large scale and high landscape connectivity. Such high connectivity creates extremely favorable conditions for species migration and dispersal, and is of key significance for the conservation of rare plant and animal populations around the Han River Eco-city. Patch 13 is a section of the Han River mainstem watershed in the southeastern part of the study area, which covers an area of 10.35 km² and has the fourth highest overall importance, filling in the gap of ecological patches in the southeastern part of the study area, and also has a greater impact on the overall connectivity of the ecological network. The remaining more important ecological patches such as 6 (dL=4.04) and 18 (dL=5.14) are the Si River tributary basins located in the northeastern part of the study area as well as the mountainous areas in the central-eastern part of the study area, respectively. However, the northern and northwestern parts of the study area suffer from a lack of ecological source areas, and the landscape connectivity in these areas is poor. This situation directly leads to a certain degree of weakening of the stability and integrity of the ecological network of green spaces in the region. Therefore, in the subsequent planning of the ecological corridor network model, more ecological stepping stones should be set up in these areas to promote the habitat and migration of organisms so as to improve the level of landscape connectivity.

Graph theory [28] is used to define a network structure as a set of comprehensible spatial configurations (a finite set of nodes, a finite set of links, and a combination of rules) to depict its topological quantities, patterns, and relationships. In order to validate the efficacy of the ecological corridor model constructed by simulation, this paper utilizes network closure, line point rate, network connectivity and cost as indicators for evaluating the structure of the ecological corridor network model. By comparing the changes of the four indicators, the effectiveness of ecological corridor network optimization can be verified and further impacts on landscape environment planning practices can be realized.

The results of the effectiveness analysis of the ecological corridor network structure model obtained from the simulation analysis of this paper and the original basic ecological corridor network are shown in Table 3. The α index of the ecological corridor network structure optimized by the method proposed in this paper reaches 0.7852, which is 32.57% higher than that of the basic ecological corridor, and is the best target for improvement. The β-index reflects the number of external connections of the source nodes, and the values of both networks are greater than 1 (1.9814 and 2.3654), indicating a complex ecological network structure. The line point rate (19.38%) and network connectivity (9.15%) of the optimized ecological network also increased. The C-index had a slight decrease (-1.05%) in the optimized ecological corridor network structure, indicating that it will become easier to modify the ecological network structure manually. Therefore, the construction and optimization of the ecological corridor network model should not only pursue the protection of ecological sources, but also strengthen the connections in the structure to improve the spatial performance.