Fusion species dispersal ability and ecological network hierarchical construction method and system based on graph theory

Abstract

The application discloses an ecological network hierarchical construction method and system fusing species dispersal ability and graph theory, and comprises the following steps: constructing a habitat undirected graph model according to the spatial overlap relationship of a spatial buffer, and dividing a global habitat network into isolated subnets, series subnets and cluster subnets based on the connected components and node degree characteristics of the habitat undirected graph model; performing differentiated spatial path searching strategies on the isolated subnets, series subnets and cluster subnets, removing the isolated subnets from a calculation sequence, extracting the minimum cost path between adjacent habitat patches in the series subnets as ecological corridors, and constructing a fully connected graph model for the cluster subnets and extracting core trunk corridors based on a minimum spanning tree algorithm; performing attribute fusion and network level division on the habitat nodes of the isolated subnets, series subnets and cluster subnets respectively, merging and grading the extracted ecological corridors and core trunk corridors, and outputting a hierarchical ecological network set.

Description

Technical Field

[0001] This invention relates to the field of ecological network hierarchical construction technology, and more specifically, to a method and system for ecological network hierarchical construction that integrates species dispersal capacity and graph theory. Background Technology

[0002] With the acceleration of global urbanization and the continuous advancement of high-intensity spatial development activities, the spatial pattern of natural habitats is facing increasingly serious trends of fragmentation and islanding. Constructing ecological networks and identifying ecological corridors connecting isolated patches has become a core technical means for territorial spatial planning and ecological restoration. Existing ecological network construction methods typically utilize geographic information systems combined with comprehensive environmental resistance data, employing minimum cost path methods or circuit theory to simulate the optimal migration paths of species between habitat patches. However, such methods generally lack a pre-judgment mechanism based on the actual dispersal capacity of species when processing habitat data on a large scale and with massive numbers of nodes. Their underlying computational logic usually treats all habitat patches in the region as equal nodes and performs pairwise pathfinding globally. When the study area expands and the number of patch nodes increases dramatically, the spatial pathfinding matrix becomes extremely large, forcing the system to perform a large amount of redundant calculations without ecological significance, resulting in extremely high computational time costs and severely restricting the effectiveness of macro-ecological planning and the efficiency of model iteration.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This invention provides a method and system for hierarchical construction of ecological networks that integrates species dispersal ability and graph theory, in order to at least solve the technical problems in the prior art, such as a large amount of invalid and redundant calculations and extremely low computational efficiency in the construction of large-scale ecological networks due to the lack of a prior topology prediction mechanism.

[0005] According to one aspect of the present invention, in order to achieve the above-mentioned objective, a method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory is provided, comprising the following steps:

[0006] Acquire habitat patch vector data and comprehensive environmental resistance data for the target area. Generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold of species. Construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers. Based on the connectivity components and node degree characteristics of the undirected habitat graph model, divide the global habitat network into isolated subnetworks, serial subnetworks, and clustered subnetworks.

[0007] Differentiated spatial pathfinding strategies are implemented for isolated subnets, serial subnets, and clustered subnets. Isolated subnets are removed from the computation sequence. For serial subnets, the minimum cost path between adjacent habitat patches is extracted as ecological corridor. For clustered subnets, a fully connected graph model is constructed and the core backbone corridor is extracted based on the minimum spanning tree algorithm.

[0008] The habitat nodes of isolated subnets, serial subnets and clustered subnets are fused according to their attributes and divided into network levels. The extracted ecological corridors are merged and classified with the core backbone corridors to output a hierarchical ecological network set.

[0009] Furthermore, habitat patch vector data and comprehensive environmental resistance data for the target area are acquired, including:

[0010] Acquire habitat patch vector data and multi-source environmental resistance factor data for the target area. The habitat patch vector data includes a polygon layer representing the core area of ​​the target species' habitat, and the multi-source environmental resistance factor data includes topographic factors, vegetation cover factors, and human activity disturbance factors.

[0011] Check and unify the projection coordinate system of all input data, resample the multi-source environmental resistance factor data to a uniform spatial resolution, use bilinear interpolation for continuous data and nearest neighbor assignment for discrete categorical data.

[0012] Each environmental resistance factor is standardized and assigned a resistance value. The weight of each environmental resistance factor is determined by the analytic hierarchy process or principal component analysis. Based on the standardized resistance factor layers and their corresponding weights under a unified spatial resolution and coordinate system, spatial raster overlay calculation is performed to output a comprehensive environmental resistance raster dataset.

[0013] Furthermore, spatial buffer zones for each habitat patch are generated based on the maximum dispersal distance threshold for species. An undirected graph model of the habitat is then constructed based on the spatial overlap relationships of these buffer zones, including:

[0014] Based on the biological characteristics of the target species, the maximum diffusion distance threshold of the species under unhindered conditions is obtained. Based on the polygons of each habitat patch, the expansion radius of the buffer zone is set to half of the maximum diffusion distance threshold, and a spatial buffer zone is generated outward for each habitat patch polygon.

[0015] The habitat patch is abstracted as a set of nodes in an undirected graph. The buffer layers of all nodes are traversed. The overlap relationship between each pair of buffers is calculated by spatial intersection analysis. If the buffers of any two nodes have spatial intersection, it is determined that the two nodes are connected within the species' maximum dispersal capacity. An undirected edge is created between the two nodes and added to the edge set. The binary adjacency matrix storing the structure of the undirected graph is output.

[0016] Furthermore, based on the connectivity components and node degree characteristics of the habitat undirected graph model, the global habitat network is divided into isolated subnets, serial subnets, and clustered subnets, including:

[0017] The habitat undirected graph is traversed using a depth-first search or breadth-first search algorithm. The nodes and their edges that are connected to each other in the graph are divided into the same subgraph. The independent connected component subgraphs are extracted, and the connectivity degree of each node is calculated.

[0018] Traverse each connected component. If the total number of nodes in the connected component is one, or the connectivity degree of the nodes inside the component is zero, then it is determined to be an isolated subnet and included in the isolated node set.

[0019] If a connected component satisfies the following conditions: the total number of nodes is greater than one, the connectivity degree of all nodes within the component is not greater than two, and there are no closed loops within the network, then it is determined to be a serial subnet and included in the serial network set.

[0020] Connected components that exclude the above two categories, have a total number of nodes greater than two, and at least one node has a connectivity degree greater than or equal to three, or have a closed loop within the network, are classified as clustered subnets and included in the clustered network set.

[0021] Furthermore, the minimum cost paths between adjacent habitat patches are extracted from the serial subnetwork as ecological corridors, including:

[0022] For any subnet in a serial set, the adjacent node pairs with a value of one in the adjacency matrix are extracted. One node in the adjacent node pair is taken as the source point and the other node as the target point. The minimum cost path between adjacent habitat patches is extracted using the comprehensive environmental resistance raster data as the weight, which serves as the ecological corridor connecting the adjacent node pairs.

[0023] Furthermore, a fully connected graph model is constructed for the clustered subnet, and the core backbone corridors are extracted based on the minimum spanning tree algorithm, including:

[0024] Extract all nodes within the clustered subnet to form a node set, establish a complete graph model, calculate the minimum cost path between any two nodes using comprehensive environmental resistance raster data as weights, and use the obtained shortest cumulative cost distance value as the resistance weight of the corresponding undirected edge in the complete graph model to construct a fully connected network containing all potential migration paths.

[0025] The minimum spanning tree algorithm is introduced. All edges in the fully connected graph are sorted in ascending order of their resistance weights. The edges with the smallest weights are selected and added to the target core network. Each time a new edge is added, a loop topology check is performed. If the newly added edge forms a closed loop with an existing edge in the network, the edge is discarded. The above process is repeated until the target core network contains all the nodes in the clustered subnet and the total number of edges is one less than the total number of nodes. The algorithm terminates when the remaining edges are output as the core backbone corridor of the clustered subnet.

[0026] Furthermore, the habitat nodes of isolated subnets, serial subnets, and clustered subnets are subjected to attribute fusion and network hierarchy division, including:

[0027] The habitat nodes of isolated subnets, serial subnets and clustered subnets are merged, and the nodes are classified according to the patch area attribute and node degree.

[0028] Habitat patches with a patch area greater than a preset threshold or a node degree greater than or equal to three are selected as core ecological source areas;

[0029] Extract nodes with a degree of 2 as ecological stepping stone nodes;

[0030] Isolated nodes with a degree of zero are defined as patches to be repaired.

[0031] Furthermore, the extracted ecological corridors are merged and classified into different levels with the core trunk corridors, including:

[0032] All minimum cost paths extracted from the serial subnet and the core backbone paths extracted from the clustered subnet based on the minimum spanning tree algorithm are merged to form a primary core corridor.

[0033] For a fully connected graph model of a clustered subnet, the remaining connected paths in its edge set that have not been claimed by the minimum spanning tree algorithm are classified into secondary auxiliary corridors.

[0034] Furthermore, topographic factors include elevation raster data and slope raster data, vegetation cover factors include vegetation cover data, and human activity disturbance factors include land use type data and road network data.

[0035] According to one embodiment of the present invention, an ecological network hierarchical construction system integrating species dispersal capacity and graph theory is also provided, comprising:

[0036] The data acquisition module is used to acquire habitat patch vector data and comprehensive environmental resistance data of the target area, generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold of species, construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers, and divide the global habitat network into isolated subnetworks, serial subnetworks and clustered subnetworks based on the connectivity components and node degree characteristics of the undirected habitat graph model.

[0037] The differentiated pathfinding module is used to implement differentiated spatial pathfinding strategies for isolated subnets, serial subnets and clustered subnets. It removes isolated subnets from the calculation sequence, extracts the minimum cost path between adjacent habitat patches as ecological corridors for serial subnets, and constructs a fully connected graph model for clustered subnets and extracts the core backbone corridors based on the minimum spanning tree algorithm.

[0038] The hierarchical output module is used to perform attribute fusion and network hierarchy division on habitat nodes of isolated subnets, serial subnets and clustered subnets respectively, merge the extracted ecological corridors and core backbone corridors and classify them into levels, and output a hierarchical ecological network set.

[0039] In this embodiment of the invention, an upfront topological buffer hierarchical mechanism is used to directly remove isolated nodes from the spatial connectivity calculation sequence, avoiding a large number of invalid calculations with no possibility of connectivity. At the same time, a minimum spanning tree algorithm is introduced for clustered networks to accurately extract core trunk lines to replace the massive complete pathfinding matrix, significantly reducing the number of pathfinding operations on resistance surfaces, which have high time costs. This significantly improves the computational efficiency and iterative capability of large-scale models such as those spanning provinces while ensuring network connectivity. On the other hand, before performing complex environmental resistance calculations, this invention uses the maximum diffusion distance threshold of the target species as a rigid constraint to construct a buffer and generate a preliminary habitat undirected graph. This ensures that each ecological corridor extracted in the end falls within the true physical diffusion limit of the species, eliminating false connectivity that is impossible to achieve in an ecological sense, and making the network construction result both ecologically reasonable and computationally efficient. Attached Figure Description

[0040] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0041] Figure 1 This is a flowchart of a method for constructing a hierarchical ecological network that integrates species dispersal capacity and graph theory according to one embodiment of the present invention;

[0042] Figure 2 This is a flowchart of a method for constructing a hierarchical ecological network that integrates species dispersal capacity and graph theory according to one embodiment of the present invention;

[0043] Figure 3 This is a structural block diagram of an ecological network hierarchical construction system that integrates species dispersal capacity and graph theory according to one embodiment of the present invention. Detailed Implementation

[0044] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0045] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0046] According to an embodiment of the present invention, an embodiment of a method for constructing hierarchical ecological networks that integrates species dispersal capacity and graph theory is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0047] This method embodiment can be executed in an electronic system or similar computing system that includes a memory and a processor. Taking operation on a terminal as an example, the terminal may include one or more processors (processors may include, but are not limited to, processing systems such as central processing units (CPUs), graphics processing units (GPUs), digital signal processing (DSP) chips, microprocessors (MCUs), field-programmable gate arrays (FPGAs), neural network processors (NPUs), tensor processors (TPUs), artificial intelligence (AI) type processors, etc.) and a memory for storing data. Optionally, the terminal may also include transmission devices, input / output devices, and display devices for communication functions. Those skilled in the art will understand that the above structural description is merely illustrative and does not limit the structure of the terminal. For example, the terminal may include more or fewer components than described above, or have a different configuration than described above.

[0048] The memory can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the method for constructing hierarchical ecological networks based on the fusion of species diffusion capacity and graph theory in this embodiment of the invention. The processor executes various functional applications and data processing by running the computer program stored in the memory, thereby realizing the aforementioned method for constructing hierarchical ecological networks based on the fusion of species diffusion capacity and graph theory. The memory may include high-speed random access memory (RAM) and non-volatile memory, such as one or more magnetic storage systems, flash memory, or other non-volatile solid-state memories. In some instances, the memory may further include memory remotely located relative to the processor, which can be connected to a mobile terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks (LANs), mobile communication networks, and combinations thereof.

[0049] A transmission system is used to receive or send data via a network. Specific examples of such networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission system includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission system may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0050] Display devices can be, for example, touchscreen liquid crystal displays (LCDs) and touch displays (also referred to as "touchscreens" or "touch displays"). The LCD allows users to interact with the user interface of the mobile terminal. In some embodiments, the mobile terminal has a graphical user interface (GUI), which allows users to interact with the GUI through finger contact and / or gestures on a touch-sensitive surface. Optional human-computer interaction functions include: creating web pages, drawing, word processing, creating electronic documents, playing games, video conferencing, instant messaging, sending and receiving emails, call interfaces, playing digital video, playing digital music, and / or web browsing, etc. Executable instructions for performing the above human-computer interaction functions are configured / stored in one or more processor-executable computer program products or readable storage media.

[0051] Figure 1 This is a flowchart of a method for constructing hierarchical ecological networks that integrates species dispersal capacity and graph theory according to one embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0052] Step S110: Obtain habitat patch vector data and comprehensive environmental resistance data for the target area. Generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold. Construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers. Based on the connectivity components and node degree characteristics of the undirected habitat graph model, divide the global habitat network into isolated subnetworks, serial subnetworks, and clustered subnetworks. The specific content is as follows:

[0053] In step S110, acquiring habitat patch vector data and comprehensive environmental resistance data for the target area is a fundamental prerequisite for implementing this method. First, it is necessary to acquire basic geospatial data of the target study area, which serves as the input for subsequent ecological network construction. Specifically, habitat patch vector data for the target area is acquired. This data is a polygonal layer containing the core habitat area of ​​the target species. The source can be land use or land cover data interpreted from remote sensing imagery, or species habitat distribution data determined by species distribution models or field surveys. Each polygonal patch in this vector layer represents a spatial unit suitable for species to inhabit, forage, or reproduce. Simultaneously, multi-source environmental resistance factor data must also be acquired. These factors characterize the surface spatial heterogeneity faced by species when moving between patches, including but not limited to continuous raster data of topographic factors such as elevation and slope, continuous raster data of vegetation cover factors such as vegetation cover degree, and discrete categorical or vector data of human activity disturbance factors such as land use type and road network. Because the aforementioned data originates from different sensors and data formats, their projected coordinate systems and spatial resolutions often differ. Therefore, it is necessary to perform unified spatial benchmark processing on all input data. This includes checking and unifying the projected coordinate systems of all input data, for example, converting them to an equal-area projected coordinate system to ensure the accuracy of subsequent distance calculations. Simultaneously, for multi-source environmental resistance factor data, it is necessary to resample to a unified spatial resolution to achieve complete pixel alignment. During the resampling process, differentiated resampling algorithms are required for different types of data. For continuous data such as elevation and slope, bilinear interpolation is used to ensure a smooth data transition. For discrete category data such as land use types, nearest neighbor assignment is used to maintain the integrity of the original category attributes' boundaries. After unifying the spatial benchmark, it is also necessary to standardize and assign resistance values ​​to each environmental resistance factor. A dimensionless resistance value is set using expert scoring combined with actual habitat requirements. The larger the resistance value, the stronger the hindering effect of the environmental factor on the dispersal of the target species. The weight allocation of each environmental resistance factor is determined using the analytic hierarchy process (AHP) or principal component analysis (PCA). Finally, based on the standardized drag factor layers and their corresponding weights under a unified spatial resolution and coordinate system, a spatial raster weighted overlay calculation is performed to output the final comprehensive environmental drag raster dataset. This dataset quantitatively characterizes the comprehensive degree of obstruction to species movement at each spatial location within the target area.

[0054] After completing the basic data preparation, this method innovatively introduces prior knowledge of species dispersal capacity as a pre-constraint for network construction. First, based on the biological characteristics of the target species, such as migratory ability, daily activity radius, or physiological limits recorded in the literature, the maximum dispersal distance threshold of the species under ideal, unhindered conditions is obtained or preset. This threshold represents the maximum straight-line spatial distance that the species can traverse under ideal conditions, serving as a rigid basis for subsequently determining whether there is potential connectivity between patches. Subsequently, based on the obtained habitat patch vector data, a spatial buffer zone is generated outward for each habitat patch polygon. To ensure accurate determination of topological relationships, the expansion radius of the buffer zone is set to half of the maximum dispersal distance threshold. The geometric principle behind this setting is that if the shortest Euclidean distance between the boundaries of two habitat patches is less than or equal to the maximum dispersal distance threshold of the species, then their buffer zones (with a radius of half the threshold) will inevitably overlap or intersect spatially; conversely, if there is no spatial overlap between the buffer zones, it indicates that the shortest spatial distance between the two patches has exceeded the species' limit dispersal capacity, and direct connectivity between the species is ecologically impossible.

[0055] After generating spatial buffers for each habitat patch, an undirected graph model of the habitat is constructed based on the spatial overlap relationships between the buffers. Specifically, all habitat patches within the study area are abstracted into an undirected graph model in graph theory, where all habitat patches constitute the set of nodes in the undirected graph, and the connectivity relationships between nodes constitute the set of edges. To achieve this mapping, the buffer layers of all nodes are traversed, and the overlap relationships between pairwise buffers are calculated through spatial intersection analysis. For any two nodes, if their corresponding buffers have spatial intersection (i.e., the area of ​​the intersecting polygon is greater than zero), then the two habitat patches are determined to be connected within the species dispersal limit, and an undirected edge is created between the two nodes and added to the edge set; if there is no spatial intersection, no edge is created. Through the above topological determination process, the computer outputs a binary adjacency matrix to store the undirected graph structure. A matrix element of 1 represents that the two habitat patches in the corresponding row and column are spatially connected within the dispersal threshold, and a matrix element of 0 represents that there is no connection. This adjacency matrix fully records the topological connectivity of the entire habitat patch network in numerical form, laying the data foundation for subsequent network feature calculations.

[0056] After constructing the undirected graph model of habitats, it is necessary to further calculate the topological characteristic indicators of the network to quantitatively describe the spatial relationships and connectivity attributes of each habitat patch in the network. First, a depth-first search or breadth-first search algorithm is used to traverse the entire undirected graph, dividing the interconnected nodes and their edges into the same subgraph, ultimately extracting several independent connected component subgraphs. Since there are no edges between different connected components, it means that species cannot naturally disperse across different connected components; therefore, each connected component is a relatively independent diffusion unit in an ecological sense. Second, the connectivity degree of each node in the undirected graph is calculated. This degree is defined as the number of edges directly connected to that node, which can be obtained by summing the corresponding rows of the adjacency matrix. The node degree intuitively reflects the number of other patches that each habitat patch can directly connect to in the spatial network, and is a key indicator for quantifying the topological importance of a patch in the network.

[0057] After obtaining the degree information of each connected component and each node, a fully automatic network morphology classification is implemented according to topological logic rules, dividing the global habitat network into three subnet types with clear topological morphology labels. Traversing each connected component, if the total number of nodes in any connected component is one, or the connectivity degree of the nodes within that component is zero, it is determined to be an isolated subnet, indicating that the habitat patch cannot establish physical or potential spatial connections with any other patch within the maximum species dispersal distance threshold. If the total number of nodes in a connected component is greater than one, and the connectivity degree of all nodes within that component is no greater than two, and there are no closed loops within the network, it is determined to be a serial subnet, indicating that the habitat patches within this subnet are arranged in a simple linear series, the species dispersal path within this subnet is unique, and there are no complex path selections or redundant branches. After excluding the two cases mentioned above, if a connected component satisfies the following conditions: the total number of nodes is greater than two, and at least one node within the component has a connectivity degree greater than or equal to three (i.e., there is a hub node where multiple paths intersect), or there is at least one closed topological loop within the subnetwork, then it is determined to be a clustered subnetwork. This indicates that habitat patches within the subnetwork are densely distributed and intertwined, exhibiting complex network connectivity characteristics, and species face multiple alternative paths when moving between patches. Through the above classification, the originally singular global habitat network is accurately divided into three subsets with clear topological relationships: isolated, sequential, and clustered. This provides a precise hierarchical calculation basis for subsequently implementing differentiated spatial routing strategies.

[0058] Step S120 involves implementing differentiated spatial pathfinding strategies for isolated subnets, sequential subnets, and clustered subnets. Isolated subnets are removed from the computation sequence. For sequential subnets, minimum-cost paths between adjacent habitat patches are extracted as ecological corridors. For clustered subnets, a fully connected graph model is constructed, and core backbone corridors are extracted based on the minimum spanning tree algorithm. The specific details are as follows:

[0059] In step S120, after the topological hierarchical division of the global ecological network is completed, differentiated spatial pathfinding strategies are executed for different types of subnets to avoid the waste of computing power caused by uniform global pathfinding for all patches in the traditional method.

[0060] For nodes classified as isolated subnets, the criteria indicate that these habitat patches have no spatially connected adjacent patches within the maximum species dispersal distance threshold. Ecologically, species cannot spread from these patches to any other patches, thus lacking a realistic basis for constructing ecological corridors. For these nodes, this method directly removes all nodes within the isolated node set from the spatial cost distance calculation sequence, allocating no memory space or pathfinding computational power to them, thereby avoiding invalid computations from the outset.

[0061] For connected components classified into serial subnets, habitat patches within these subnets are arranged in a simple linear series. The connectivity between adjacent nodes is fully recorded by the aforementioned adjacency matrix, and the dispersal path of a species within this subnet is unique. For any subnet in the serial set, the aforementioned adjacency matrix is ​​directly called to extract adjacent node pairs with a value of 1. These node pairs indicate that two habitat patches have spatial connectivity within the maximum dispersal distance threshold of the species. Subsequently, using one node in the adjacent node pair as the source point and the other as the target point, and using the integrated environmental resistance raster data as the spatial cost weight, a raster pathfinding algorithm is used to calculate the cumulative cost distance between adjacent nodes. The raster path with the lowest cumulative cost is then vectorized into a spatial linear feature, thus extracting the minimum cost path connecting adjacent habitat patches. This path is the ecological corridor connecting the adjacent node pair.

[0062] For connected components classified into clustered subnets, habitat patches within these subnets are densely distributed and intertwined, with complex connectivity relationships between nodes, presenting species with multiple alternative paths when moving between patches. For this type of subnet, we first extract all nodes within the clustered subnet to form a set, assuming that species can theoretically migrate between any two nodes within this set, and construct a complete graph model accordingly. Using integrated environmental resistance raster data as weights, we calculate the minimum cost path between any two nodes, and use the resulting shortest cumulative cost distance as the resistance weight of the corresponding undirected edge in the graph model. This constructs a fully connected network containing all potential migration paths, which quantitatively characterizes the minimum resistance cost required for diffusion between any two habitat patches within the cluster. After obtaining the fully connected network, this method introduces a minimum spanning tree algorithm to filter and eliminate a large number of redundant paths in the fully connected network. All edges in the fully connected graph are sorted in ascending order of their resistance weights. Edges with the lowest weights are then added to the target core network. A cyclic topology check is performed each time a new edge is added. If the newly added edge forms a closed loop with an existing edge, it is discarded. This optimization and verification process is repeated until the target core network contains all nodes within the subnet and the total number of edges is exactly one less than the total number of nodes. The remaining edges represent the lowest total resistance cost path required for a species to traverse all patches within the cluster. These edges serve as the low-cost core backbone corridor output for the cluster-type subnet, ensuring full connectivity of all habitat patches within the cluster while eliminating computational redundancy caused by multiple pathfinding at the lowest resistance cost.

[0063] Step S140 involves performing attribute fusion and network hierarchy division on the habitat nodes of isolated subnets, serial subnets, and clustered subnets, respectively. The extracted ecological corridors are then merged and hierarchically divided with the core backbone corridors to output a hierarchical ecological network set. The specific details are as follows:

[0064] In step S140, after completing the spatial pathfinding and corridor extraction of the aforementioned subnets at each topological level, this method further integrates the habitat nodes and ecological corridors of each type of subnet, performs attribute fusion and constructs a multi-level evaluation system, and finally outputs a complete set of hierarchical ecological networks.

[0065] For the hierarchical processing of habitat nodes, firstly, all habitat nodes in the previously divided isolated subnets, serial subnets, and clustered subnets are merged to form a unified set of habitat patch nodes across the entire domain. Based on this, a comprehensive hierarchical classification is implemented for all nodes according to the physical attributes of each habitat patch and its connectivity characteristics within the network topology. Specifically, regarding patch physical attributes, the geometric area of ​​each habitat patch is extracted as an evaluation index, reflecting the patch's habitat carrying capacity and its potential to accommodate target species. Regarding network topology attributes, the connectivity degree of each node is extracted as an evaluation index, reflecting the patch's topological position and structural importance within the overall spatial network. Based on these two dimensions of evaluation indices, all habitat nodes are divided into three levels. Habitat patches with an area greater than a preset threshold, or habitat patches with a node connectivity degree of three or higher, are designated as core ecological source areas. A patch with an area greater than the preset threshold indicates a high species carrying capacity, enabling it to maintain a stable population size. A node connectivity degree of three or higher indicates that the patch is located at a hub where multiple paths intersect in the network topology, and its connectivity directly affects the accessibility of surrounding patches. Meeting either of these criteria qualifies it as a crucial core ecological source area within the region. Nodes with a connectivity degree of two or higher are extracted and designated as ecological stepping stone nodes. These patches are mainly distributed within serial subnetworks or on the edges of clustered subnetworks. Although they are not as large or hub-like as core source areas, they serve as indispensable resting places and transit stations during species migration, ensuring the continuity of long-distance dispersal paths. Isolated nodes with a connectivity of zero are defined as patches to be restored. These patches cannot achieve gene exchange with other patches under the current conditions because the spatial distance to the nearest neighbor exceeds the threshold of the species’ maximum dispersal capacity. They should be the priority targets for improvement in subsequent ecological restoration projects.

[0066] For the hierarchical processing of ecological corridors, the ecological corridors extracted for different types of subnets are merged, and differentiated level labels are assigned according to the functional positioning and importance of the corridors in the network. Specifically, the minimum cost paths between all adjacent patches extracted from the serial subnets and the core backbone paths extracted from the clustered subnets based on the minimum spanning tree algorithm are spatially merged and jointly classified into first-level core corridors. This set of corridors represents the backbone paths necessary to ensure connectivity of all non-isolated nodes in the entire region under the current comprehensive environmental resistance conditions, and is the lowest-cost framework for maintaining basic gene exchange and population continuation. Meanwhile, for the fully connected graph model constructed in the clustered subnet, the remaining connected paths in its edge set that were not claimed by the minimum spanning tree algorithm are extracted and divided into secondary auxiliary corridors. Although the cost of a single resistance is higher than that of the primary core corridor, their existence in the network provides rich redundant alternatives for species migration. When a primary core corridor fails due to sudden environmental changes or human interference, the secondary auxiliary corridor can serve as an alternative path to maintain the overall connectivity of the network, thereby significantly enhancing the structural resilience and anti-interference ability of the ecological network in the face of uncertain environmental stress.

[0067] After completing the node and corridor classification operations described above, the classified node and corridor data are integrated within a unified spatial reference framework to output a classified ecological network spatial database containing different protection priorities. This classified ecological network not only clearly depicts the core ecological source areas and primary core corridor framework that must be prioritized for protection within the region, but also identifies the spatial locations of isolated patches requiring targeted ecological restoration. Furthermore, it marks secondary auxiliary corridors and stepping stone nodes that can improve network redundancy. This provides a scientific, systematic, and operable decision support basis for the delineation of ecological protection red lines in territorial spatial planning, the timing of ecological corridor construction projects, and the differentiated management of protection areas at different levels.

[0068] Based on steps S110 to S140 above, in this embodiment of the invention, an isolated node is directly removed from the spatial connectivity calculation sequence through a pre-established topological buffer hierarchical mechanism, avoiding a large number of invalid calculations with no possibility of connectivity. At the same time, a minimum spanning tree algorithm is introduced for clustered networks to accurately extract core trunk lines to replace the massive complete pathfinding matrix, significantly reducing the number of pathfinding operations on resistance surfaces, which have high time costs. Under the premise of ensuring network connectivity, the computational efficiency and iterative capability of large-scale models such as cross-provincial regions are significantly improved. On the other hand, before performing complex environmental resistance calculations, this invention constructs a buffer and generates a preliminary habitat undirected graph by using the maximum diffusion distance threshold of the target species as a rigid constraint condition. This ensures that each ecological corridor extracted in the end falls within the true physical diffusion limit of the species, eliminating false connectivity that is impossible to achieve in an ecological sense, so that the network construction result has both ecological rationality and computational efficiency.

[0069] The present invention provides an embodiment of an ecological network hierarchical construction method that integrates species dispersal capacity and graph theory to acquire habitat patch vector data and comprehensive environmental resistance data of a target area. This includes: acquiring habitat patch vector data and multi-source environmental resistance factor data of the target area; the habitat patch vector data includes a polygonal layer representing the core area of ​​the target species' habitat; the multi-source environmental resistance factor data includes topographic factors, vegetation cover factors, and human activity disturbance factors; checking and unifying the projection coordinate system of all input data; resampling the multi-source environmental resistance factor data to a unified spatial resolution; using bilinear interpolation for continuous data and nearest neighbor assignment for discrete categorical data; standardizing and assigning resistance values ​​to each environmental resistance factor; determining the weight of each environmental resistance factor using analytic hierarchy process (AHP) or principal component analysis; and performing spatial raster overlay calculations based on the standardized resistance factor layers and their corresponding weights under a unified spatial resolution and coordinate system to output a comprehensive environmental resistance raster dataset.

[0070] Furthermore, spatial buffers for each habitat patch are generated based on the maximum dispersal distance threshold of the species. An undirected graph model of the habitat is constructed based on the spatial overlap relationship of the spatial buffers. This includes: obtaining the maximum dispersal distance threshold of the species under no-resistance conditions based on the biological characteristics of the target species; setting the expansion radius of the buffer to half of the maximum dispersal distance threshold based on the polygons of each habitat patch; generating spatial buffers outward from the polygons of each habitat patch; abstracting the habitat patch as a set of nodes in an undirected graph; traversing the buffer layers of all nodes; calculating the overlap relationship between pairs of buffers through spatial intersection analysis; if the buffers of any two nodes have spatial intersection, it is determined that the two nodes are connected within the species' maximum dispersal capacity, and an undirected edge is created between the two nodes and added to the edge set; and outputting a binary adjacency matrix that stores the structure of the undirected graph.

[0071] Furthermore, based on the connectivity and node degree characteristics of the habitat undirected graph model, the global habitat network is divided into isolated subnetworks, serial subnetworks, and clustered subnetworks. This includes: traversing the habitat undirected graph using depth-first search or breadth-first search algorithms, grouping interconnected nodes and their edges into the same subgraph, extracting independent connected component subgraphs, and calculating the connectivity degree of each node; traversing each connected component, if the total number of nodes in the connected component is one, or the connectivity degree of the nodes within the component is zero, it is determined to be an isolated subnetwork and included in the isolated node set; if the total number of nodes in the connected component is greater than one, and the connectivity degree of all nodes within the component is no greater than two, and there are no closed loops within the network, it is determined to be a serial subnetwork and included in the serial network set; connected components that exclude the above two cases, have a total number of nodes greater than two, and at least one node has a connectivity degree greater than or equal to three, or there are closed loops within the network, are determined to be clustered subnetworks and included in the cluster network set.

[0072] Furthermore, the minimum cost path between adjacent habitat patches is extracted as an ecological corridor for any subnet in the series-type subnet. This includes: for any subnet in the series-type set, extracting adjacent node pairs with a corresponding value of one in the adjacency matrix, taking one node in the adjacent node pair as the source point and the other node as the target point, and using the comprehensive environmental resistance raster data as the weight, extracting the minimum cost path between adjacent habitat patches as an ecological corridor connecting adjacent node pairs.

[0073] Furthermore, a fully connected graph model is constructed for the clustered subnet, and the core backbone corridors are extracted based on the minimum spanning tree algorithm. This includes: extracting all nodes within the clustered subnet to form a node set, establishing a complete graph model, calculating the minimum cost path between any two nodes using comprehensive environmental resistance grid data as weights, and using the shortest cumulative cost distance as the resistance weight of the corresponding undirected edge in the complete graph model to construct a fully connected network containing all potential migration paths; introducing the minimum spanning tree algorithm, arranging all edges in the fully connected graph in ascending order according to their resistance weights, and sequentially selecting the edge with the smallest weight to add to the target core network. Each time a new edge is added, a cyclic topology check is performed. If the newly added edge forms a closed loop with an existing edge in the network, the edge is discarded. The above process is repeated until the target core network contains all nodes within the clustered subnet and the total number of edges is one less than the total number of nodes, at which point the algorithm terminates. The remaining edges are output as the core backbone corridors of the clustered subnet.

[0074] Furthermore, habitat nodes in isolated, serial, and clustered subnets are subjected to attribute fusion and network hierarchy division, including: merging habitat nodes in isolated, serial, and clustered subnets; classifying nodes according to their patch area attributes and node degree; selecting habitat patches with a patch area greater than a preset threshold or a node degree greater than or equal to three as core ecological sources; extracting nodes with a node degree of two as ecological stepping stone nodes; and defining isolated nodes with a node degree of zero as patches to be restored.

[0075] Furthermore, the extracted ecological corridors are merged and classified into levels with the core backbone corridors, including: merging all minimum cost paths extracted from the serial subnets and the core backbone paths extracted from the clustered subnets based on the minimum spanning tree algorithm as first-level core corridors; for the fully connected graph model of the clustered subnets, the remaining connected paths in its edge set that were not claimed by the minimum spanning tree algorithm are classified as second-level auxiliary corridors.

[0076] Furthermore, topographic factors include elevation raster data and slope raster data, vegetation cover factors include vegetation cover data, and human activity disturbance factors include land use type data and road network data.

[0077] like Figure 2 This is another method for constructing hierarchical ecological networks that integrates species dispersal capacity and graph theory according to one embodiment of the present invention. The method includes the following steps:

[0078] Step 201: Obtain habitat patch vector data and comprehensive environmental resistance data for the target area, construct a unified spatial reference and align the spatial resolution, and output the basic analysis dataset.

[0079] Polygonal layers of lakes in Northeast China were acquired as habitat patch vector data, resulting in 471 lake habitat patches with areas ranging from 0.22 to 2053.96 square kilometers. Considering the physiological and ecological characteristics of amphibians (e.g., preference for humid environments and low dispersal capacity), four key environmental factors—land cover type, elevation, slope, and vegetation index (NDVI)—were selected as indicators for constructing comprehensive environmental resistance data. NDVI data was calculated using the spectral reflectance of the red and near-infrared (NIR) bands in Landsat 8 imagery, using the following formula: .

[0080] The aforementioned habitat patch vector data and environmental factor data were uniformly transformed and projected to the Albers_Conical_Equal_Area coordinate system. Simultaneously, the spatial resolution of all factors was unified to 30m, and the basic analysis dataset was output.

[0081] Based on relevant literature on amphibian behavior and the actual habitat characteristics of the study area, the resistance values ​​of each factor were standardized and divided into 5 levels (1 representing minimum resistance, and 5 representing maximum resistance). The specific grading criteria are as follows:

[0082] Land cover type: paddy fields, lakes, reservoirs, ponds, and beaches are assigned a value of 1; grasslands and unused land are assigned a value of 2; forest land is assigned a value of 3; dry land is assigned a value of 4; impermeable surfaces are assigned a value of 5.

[0083] Elevation: 0-100m is assigned a value of 1; 100-200m is assigned a value of 2; 200-300m is assigned a value of 3; 300-500m is assigned a value of 4; 500-2691m is assigned a value of 5.

[0084] Slope: 0-5 degrees is assigned a value of 1; 5-10 degrees is assigned a value of 2; 10-15 degrees is assigned a value of 3; 15-30 degrees is assigned a value of 4; and greater than 30 degrees is assigned a value of 5.

[0085] Vegetation Index (NDVI): Since high vegetation cover provides better cover and moisture retention, the closer the vegetation index is to 1, the lower the resistance. The following formula is used for standardized conversion: In the formula, Assign a resistance value to the NDVI factor. This is the Normalized Difference Vegetation Index (NDVI). The calculation results ensure that the higher the NDVI value, the closer the drag value is to 0.

[0086] After determining the individual factor assignments, the Analytic Hierarchy Process (AHP) was introduced to calculate the weights of the four resistance factors as 0.29, 0.16, 0.19, and 0.36, respectively. The following formula was used for raster element calculation: Where R is the comprehensive environmental resistance value of the spatial grid cell; n is the total number of environmental elements; For the first The weight of each environmental element, For the first Each environmental element is assigned a standardized resistance surface value in the grid cell. This yields comprehensive environmental resistance data with a value range of 0-20.

[0087] Step S202: Generate a spatial buffer based on the maximum dispersal distance threshold of the species, construct an undirected habitat graph model through the spatial overlap relationship of the buffer, and calculate the connected components and node degree.

[0088] Based on field surveys and literature review, representative major amphibian species in Northeast China were identified, such as the Chinese toad (Bufogargarizans), the flower-backed toad (Strauchbuforaddei), the Chinese forest frog (Ranachensinensis), and the black-spotted side-fence frog (Pelophylaxnigromaculatus). Taking into account the biological migration limits and water requirements of these species, the maximum dispersal distance threshold for the target species in this embodiment was determined. The threshold was set at 1500m. Based on this threshold, spatial analysis tools were used to generate spatial buffer zones outward from each of the 471 lake habitat patches extracted in step S1. To accurately determine the limit connectivity probability between any two patches, the expansion radius of each buffer zone was set to half of the maximum diffusion distance, i.e. .

[0089] Next, we construct an undirected graph model in graph theory. Map 471 lake patches to a set of nodes in an undirected graph. Perform spatial intersection analysis on 471 patches corresponding to a 750m buffer. If the patch With plaque The buffer zones have spatial overlap (i.e., indicating the actual shortest span distance between the two lakes). If species theoretically possess the ability to directly cross boundaries, then an undirected edge is created in the graph model. After topology determination, the system's underlying layer outputs a... Binary adjacency matrix of dimension Matrix elements This indicates that the patches in the two lakes are spatially connected within the diffusion threshold. This indicates no connection.

[0090] Based on the adjacency matrix A generated above, the connected components and node degrees are calculated. First, a graph traversal algorithm (such as depth-first search) is used to find interconnected subgraphs in the overall undirected graph. Finally, the 471 lake nodes are divided into k physically isolated sets of connected components. In this context, species cannot naturally disperse across different connected components. Next, for each lake node in the set... The connectivity number is calculated by summing the corresponding rows of the adjacency matrix A. This quantifies the number of other lakes that each lake can directly connect to within a 1500m diffusion threshold.

[0091] Step S203: Based on the connectivity components and node degree characteristics output by the graph model, construct network morphology determination rules to divide the global habitat network into isolated, serial, and clustered subnetworks.

[0092] This step is based on the undirected graph containing 471 lake nodes and the set of k connected components output in step S2. and the degree of each lake node It iterates through each connected component and classifies the network topology according to the set judgment rules. Specifically, for any connected component... (The total number of lake nodes inside is) The division rules and judgment logic are as follows:

[0093] Isolated Subnetwork: If connected components Total number of lake nodes Or the degree of a lake node within that component. It is then classified as an isolated subnet and assigned to the isolated node set. This indicates that the shortest distance from this lake to any other body of water exceeds the amphibian dispersal threshold of 1500m. For amphibians that are extremely dependent on aquatic environments, living in this lake would prevent them from migrating to other lake patches.

[0094] Serial Subnetwork: If connected components Satisfy the total number of lake nodes Furthermore, the degree of all lake nodes within this component satisfies... If there are no closed topological loops within the network, it is divided into a series subnet and assigned to a series lake set. This represents a group of lakes distributed linearly or in chains, where the dispersal paths of amphibians migrating within the lake chain are unique.

[0095] Clustered Subnetwork: If connected components Satisfy the total number of lake nodes And the degree of at least one pivot lake exists within this component. Lakes with multiple intersecting connecting paths, or lake groups containing closed cycles, are classified as clustered lakes and included in a clustered lake set. .

[0096] Using the algorithm described above, 471 lakes in Northeast China were accurately divided into three subsets with clear topological relationships: isolated, connected, and clustered. It contains 266 lakes. Including 145 lakes, It includes 60 lakes.

[0097] Step S204: Execute differentiated spatial pathfinding strategies: remove isolated nodes from the computation sequence; extract minimum cost paths for serial subnets; generate fully connected graphs for clustered subnets and extract core corridors based on minimum spanning trees.

[0098] Based on the lake topology subset output in step S203 and the comprehensive environmental resistance data generated in step S201, different pathfinding algorithms are implemented for different lake networks. Specifically:

[0099] For isolated lake collections : Directly remove it from the iterative sequence of CostDistanceComputation, and no longer allocate memory to it to calculate the global path across the resistance surface, thereby avoiding meaningless computational consumption.

[0100] For a series of lakes : with adjacent lakes For the source point, Using the comprehensive environmental resistance data output in step S1 as the target point, the spatial cost weight is determined. A raster pathfinding algorithm (such as Dijkstra's algorithm) is applied to calculate the cumulative cost distance between adjacent lakes. The raster with the lowest cumulative cost is vectorized into a spatial linear feature, thus generating a series of ecological corridors connecting the lake chain.

[0101] For clustered lakes First, establish a complete graph model within the subnet of this cluster. Using the comprehensive ecological resistance surface as weight, the minimum cost path (LCP) between any two lakes within the subnet is calculated, and the resulting shortest cumulative resistance cost is used as the spatial weight of the undirected edge. This constructs a fully connected network model containing all potential migration paths. To eliminate high-cost redundant connections, the Minimum Spanning Tree (MST) algorithm is introduced (Kruskal's algorithm is chosen in this example). All edges in the fully connected graph are weighted according to their resistance. The nodes are sorted in ascending order from smallest to largest. The connection with the lowest weight is added to the core network sequentially. Each time a new connection is added, a cyclic topology check is performed; if a new connection forms a closed loop with an existing network, it is discarded. This optimization process continues until the core network contains all N lake nodes in the cluster, and the total number of connections reaches exactly N-1, at which point the algorithm terminates. The N-1 optimal paths retained are the low-cost core backbone corridors that ensure full connectivity of the lake cluster and minimize total resistance. Amphibians migrating through these core backbone corridors can reach any lake with minimal resistance cost.

[0102] Step S205: Perform attribute fusion and level classification on the habitat nodes divided in step S203 and the corridors of each topological level extracted in step S204, establish an evaluation system for nodes and corridors, and output a hierarchical ecological network set.

[0103] The three types of lakes divided in step S203 (i.e.) , and The data was merged based on the area attributes and node degrees of the 471 lakes in the set. An evaluation system for lake habitat nodes was established.

[0104] Core ecological source area: Based on the actual distribution of lake area from 0.22 to 2053.96 km², the natural discontinuity method was used to define the threshold, setting the core habitat area threshold at 14 km². Areas larger than 14 km², or those with high topological connectivity, were selected. Lake patches. An area greater than 14 km² indicates a lake with strong habitat carrying capacity, capable of supporting a large amphibian population; while a degree greater than or equal to 3 indicates the lake is located at a crucial intersection of a water system network. Meeting either of these criteria is sufficient to designate it as a core source area for regional amphibians.

[0105] Ecological stepping stone node: Extracting connectivity degree satisfies Furthermore, these are lake nodes whose area does not meet the criteria for core source areas. These small and medium-sized lakes are widely distributed on the edges of serial or clustered subnetworks, providing not only necessary resting, foraging, and refuge sites for the long-distance dispersal of amphibians, but also serving as crucial stepping stones to maintain the overall network connectivity baseline.

[0106] Patches to be repaired: Collections of lakes with a connectivity degree of 0 (i.e., isolated lakes) The nodes of these lakes are defined as patches to be restored. Amphibians living around these lakes cannot migrate to other lakes and can be the focus of subsequent ecological restoration projects.

[0107] The ecological corridors extracted in step S204 are merged, and the following corridor classification is performed:

[0108] Primary core corridor: This involves spatially merging all minimum cost paths (LCPs) between adjacent lakes extracted from the serial subnet in step S4, and the core backbone paths extracted from the clustered subnet using the minimum spanning tree (MST) algorithm. This level of corridor connects all non-isolated water bodies within the region and represents the lowest-cost core corridor necessary to ensure basic gene exchange among amphibian populations in Northeast China under the existing combined resistance conditions.

[0109] Secondary auxiliary corridors: for the clustered fully connected graph model constructed in step S4 The remaining minimum-cost paths not selected by the minimum spanning tree algorithm are extracted and classified as secondary auxiliary corridors. Although these corridors are not optimal in terms of resistance cost, they provide amphibians with crucial alternative migration routes, significantly enhancing their structural resilience to sudden environmental disturbances.

[0110] After completing the above-mentioned hierarchical operations on nodes and corridors, output hierarchical node (core ecological source area, ecological stepping stone node, isolated patch) and hierarchical corridor (first-level core corridor and second-level auxiliary corridor) data.

[0111] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0112] This invention also provides an ecological network hierarchical construction system that integrates species dispersal capacity and graph theory. This system is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0113] Figure 3 According to one embodiment of the present invention, an ecological network hierarchical construction system integrating species dispersal capacity and graph theory is provided, comprising:

[0114] The data acquisition module 301 is used to acquire habitat patch vector data and comprehensive environmental resistance data of the target area, generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold of species, construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers, and divide the global habitat network into isolated subnets, serial subnets and clustered subnets based on the connectivity components and node degree characteristics of the undirected habitat graph model.

[0115] The differentiated pathfinding module 302 is used to execute differentiated spatial pathfinding strategies for isolated subnets, serial subnets and clustered subnets. It removes isolated subnets from the calculation sequence, extracts the minimum cost path between adjacent habitat patches as ecological corridors for serial subnets, and constructs a fully connected graph model for clustered subnets and extracts the core backbone corridors based on the minimum spanning tree algorithm.

[0116] The hierarchical output module 303 is used to perform attribute fusion and network hierarchy division on habitat nodes of isolated subnets, serial subnets and clustered subnets respectively, merge the extracted ecological corridors and core backbone corridors and classify them into levels, and output a hierarchical ecological network set.

[0117] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0118] According to one embodiment of the present invention, an electronic device is also provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the above-described method for constructing an ecological network hierarchy that integrates species dispersal capacity and graph theory.

[0119] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0120] Step S1: Obtain habitat patch vector data and comprehensive environmental resistance data for the target area. Generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold of species. Construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers. Based on the connectivity components and node degree characteristics of the undirected habitat graph model, divide the global habitat network into isolated subnetworks, serial subnetworks, and clustered subnetworks.

[0121] Step S2: Implement differentiated spatial pathfinding strategies for isolated subnets, serial subnets, and clustered subnets. Remove isolated subnets from the calculation sequence. Extract the minimum cost path between adjacent habitat patches as ecological corridors for serial subnets. Construct a fully connected graph model for clustered subnets and extract the core backbone corridors based on the minimum spanning tree algorithm.

[0122] Step S3 involves performing attribute fusion and network hierarchy division on the habitat nodes of isolated subnets, serial subnets, and clustered subnets, merging and classifying the extracted ecological corridors with the core backbone corridors, and outputting a hierarchical ecological network set.

[0123] According to one embodiment of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the storage medium is located to execute the above-described method for constructing an ecological network hierarchy that integrates species dispersal ability and graph theory.

[0124] Optionally, in this embodiment, the storage medium may be configured to store a computer program for performing the following steps:

[0125] Step S1: Obtain habitat patch vector data and comprehensive environmental resistance data for the target area. Generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold of species. Construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers. Based on the connectivity components and node degree characteristics of the undirected habitat graph model, divide the global habitat network into isolated subnetworks, serial subnetworks, and clustered subnetworks.

[0126] Step S2: Implement differentiated spatial pathfinding strategies for isolated subnets, serial subnets, and clustered subnets. Remove isolated subnets from the calculation sequence. Extract the minimum cost path between adjacent habitat patches as ecological corridors for serial subnets. Construct a fully connected graph model for clustered subnets and extract the core backbone corridors based on the minimum spanning tree algorithm.

[0127] Step S3 involves performing attribute fusion and network hierarchy division on the habitat nodes of isolated subnets, serial subnets, and clustered subnets, merging and classifying the extracted ecological corridors with the core backbone corridors, and outputting a hierarchical ecological network set.

[0128] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0129] According to one embodiment of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the above-described method for constructing hierarchical ecological networks that integrates species diffusion capacity and graph theory.

[0130] Optionally, in this embodiment, the above-mentioned computer program product can be configured as a computer program that performs the following steps:

[0131] Step S1: Obtain habitat patch vector data and comprehensive environmental resistance data for the target area. Generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold of species. Construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers. Based on the connectivity components and node degree characteristics of the undirected habitat graph model, divide the global habitat network into isolated subnetworks, serial subnetworks, and clustered subnetworks.

[0132] Step S2: Implement differentiated spatial pathfinding strategies for isolated subnets, serial subnets, and clustered subnets. Remove isolated subnets from the calculation sequence. Extract the minimum cost path between adjacent habitat patches as ecological corridors for serial subnets. Construct a fully connected graph model for clustered subnets and extract the core backbone corridors based on the minimum spanning tree algorithm.

[0133] Step S3 involves performing attribute fusion and network hierarchy division on the habitat nodes of isolated subnets, serial subnets, and clustered subnets, merging and classifying the extracted ecological corridors with the core backbone corridors, and outputting a hierarchical ecological network set.

[0134] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0135] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0136] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The system embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of units or modules may be electrical or other forms.

[0137] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0138] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0139] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0140] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for constructing hierarchical ecological networks that integrates species dispersal capacity and graph theory, characterized in that, Includes the following steps: Acquire habitat patch vector data and comprehensive environmental resistance data for the target area. Generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold of species. Construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers. Based on the connectivity components and node degree characteristics of the undirected habitat graph model, divide the global habitat network into isolated subnetworks, serial subnetworks, and clustered subnetworks. Differentiated spatial pathfinding strategies are implemented for the isolated subnets, the serial subnets, and the clustered subnets. The isolated subnets are removed from the calculation sequence. For the serial subnets, the minimum cost path between adjacent habitat patches is extracted as an ecological corridor. For the clustered subnets, a fully connected graph model is constructed and the core backbone corridor is extracted based on the minimum spanning tree algorithm. The habitat nodes of the isolated subnet, the serial subnet, and the clustered subnet are respectively subjected to attribute fusion and network hierarchy division. The extracted ecological corridors and the core backbone corridors are merged and classified into levels to output a hierarchical ecological network set.

2. The method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory according to claim 1, characterized in that, Acquire habitat patch vector data and comprehensive environmental resistance data for the target area, including: Acquire habitat patch vector data and multi-source environmental resistance factor data for the target area. The habitat patch vector data includes a polygon layer representing the core area of ​​the target species' habitat. The multi-source environmental resistance factor data includes topographic factors, vegetation cover factors, and human activity disturbance factors. Check and unify the projection coordinate system of all input data, resample the multi-source environmental resistance factor data to a unified spatial resolution, use bilinear interpolation for continuous data and nearest neighbor assignment for discrete category data. Each environmental resistance factor is standardized and assigned a resistance value. The weight of each environmental resistance factor is determined by the analytic hierarchy process or principal component analysis. Based on the standardized resistance factor layers and their corresponding weights under the unified spatial resolution and coordinate system, spatial raster overlay calculation is performed to output a comprehensive environmental resistance raster dataset.

3. The method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory according to claim 2, characterized in that, Spatial buffer zones for each habitat patch are generated based on a maximum species dispersal distance threshold. An undirected habitat graph model is then constructed based on the spatial overlap relationships of these buffer zones, including: Based on the biological characteristics of the target species, the maximum diffusion distance threshold of the species under unobstructed conditions is obtained. Based on the polygons of each habitat patch, the expansion radius of the buffer zone is set to half of the maximum diffusion distance threshold, and a spatial buffer zone is generated outward for each habitat patch polygon. The habitat patch is abstracted as a set of nodes in an undirected graph. The buffer layers of all nodes are traversed. The overlap relationship between each pair of buffers is calculated by spatial intersection analysis. If the buffers of any two nodes have spatial intersection, it is determined that the two nodes are connected within the species' maximum dispersal capacity. An undirected edge is created between the two nodes and added to the edge set. The binary adjacency matrix storing the structure of the undirected graph is output.

4. The method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory according to claim 3, characterized in that, Based on the connectivity components and node degree characteristics of the aforementioned habitat undirected graph model, the global habitat network is divided into isolated subnets, serial subnets, and clustered subnets, including: The habitat undirected graph is traversed using a depth-first search or breadth-first search algorithm. The nodes and edges that are connected to each other in the graph are divided into the same subgraph. The independent connected component subgraphs are extracted, and the connectivity degree of each node is calculated. Traverse each connected component. If the total number of nodes in the connected component is one, or the connectivity degree of the nodes inside the component is zero, then it is determined to be an isolated subnet and included in the isolated node set. If a connected component satisfies the following conditions: the total number of nodes is greater than one, the connectivity degree of all nodes within the component is not greater than two, and there are no closed loops within the network, then it is determined to be a serial subnet and included in the serial network set. Connected components that exclude the above two categories, have a total number of nodes greater than two, and at least one node has a connectivity degree greater than or equal to three, or have a closed loop within the network, are classified as clustered subnets and included in the clustered network set.

5. The method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory according to claim 4, characterized in that, Extracting the minimum cost path between adjacent habitat patches as ecological corridors from the serial subnet includes: For any subnet in a series-type set, the adjacent node pairs with a corresponding value of one in the adjacency matrix are extracted. Taking one node in the adjacent node pair as the source point and the other node as the target point, and using the comprehensive environmental resistance raster data as the weight, the minimum cost path between adjacent habitat patches is extracted as the ecological corridor connecting the adjacent node pairs.

6. The method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory according to claim 4, characterized in that, A fully connected graph model is constructed for the clustered subnet, and the core backbone corridors are extracted based on the minimum spanning tree algorithm, including: Extract all nodes within the clustered subnet to form a node set, establish a complete graph model, calculate the minimum cost path between any two nodes using the comprehensive environmental resistance grid data as weights, and use the obtained shortest cumulative cost distance value as the resistance weight of the corresponding undirected edge in the complete graph model to construct a fully connected network containing all potential migration paths. The minimum spanning tree algorithm is introduced. All edges in the fully connected graph are arranged in ascending order of their resistance weights. The edges with the smallest weights are selected and added to the target core network. Each time a new edge is added, a cyclic topology check is performed. If the newly added edge forms a closed loop with an existing edge in the network, the edge is discarded. The above process is repeated until the target core network contains all the nodes in the clustered subnet and the total number of edges is one less than the total number of nodes. The algorithm terminates when the remaining edges are output as the core backbone corridor of the clustered subnet.

7. The method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory according to claim 1, characterized in that, The habitat nodes of the isolated subnet, the serial subnet, and the clustered subnet are respectively subjected to attribute fusion and network hierarchy division, including: The habitat nodes of the isolated subnet, the serial subnet, and the clustered subnet are merged, and the nodes are classified according to their patch area attributes and node degree. Habitat patches with a patch area greater than a preset threshold or a node degree greater than or equal to three are selected as core ecological source areas; Extract nodes with a degree of 2 as ecological stepping stone nodes; Isolated nodes with a degree of zero are defined as patches to be repaired.

8. The method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory according to claim 1, characterized in that, The extracted ecological corridors are merged and classified into different levels with the core trunk corridors, including: All minimum cost paths extracted from the serial subnet and the core backbone paths extracted from the clustered subnet based on the minimum spanning tree algorithm are merged to form a primary core corridor. For the fully connected graph model of the clustered subnet, the remaining connected paths in its edge set that have not been claimed by the minimum spanning tree algorithm are divided into secondary auxiliary corridors.

9. The method for constructing hierarchical ecological networks integrating species dispersal capacity and graph theory according to claim 2, characterized in that, The topographic factors include elevation raster data and slope raster data; the vegetation cover factors include vegetation cover data; and the human activity disturbance factors include land use type data and road network data.

10. A hierarchical ecological network construction system integrating species dispersal capacity and graph theory, characterized in that, include: The data acquisition module is used to acquire habitat patch vector data and comprehensive environmental resistance data of the target area, generate spatial buffers for each habitat patch based on the maximum dispersal distance threshold of species, construct an undirected habitat graph model based on the spatial overlap relationship of the spatial buffers, and divide the global habitat network into isolated subnetworks, serial subnetworks and clustered subnetworks based on the connectivity components and node degree characteristics of the undirected habitat graph model. The differentiated pathfinding module is used to execute differentiated spatial pathfinding strategies on the isolated subnet, the serial subnet, and the clustered subnet. It removes the isolated subnet from the calculation sequence, extracts the minimum cost path between adjacent habitat patches as ecological corridors for the serial subnet, and constructs a fully connected graph model for the clustered subnet and extracts the core backbone corridors based on the minimum spanning tree algorithm. The hierarchical output module is used to perform attribute fusion and network hierarchy division on the habitat nodes of the isolated subnet, the serial subnet and the clustered subnet respectively, merge the extracted ecological corridors and the core backbone corridors and classify them into levels, and output a hierarchical ecological network set.

Images