Habitat network optimization method that couples landscape and conflict resistance

Abstract

This invention relates to the field of animal ecology technology, and particularly to a habitat network optimization method coupling landscape and conflict resistance, comprising the following steps: S1, data extraction and preprocessing: the data includes: LULC, topography, normalized difference vegetation index, multi-year average precipitation, roads, species distribution, and incident data; S2, calculating the species habitat suitability index using the HSI model and determining the comprehensive weight of each evaluation factor; S3, classifying the normalized species habitat using the natural discontinuity method; S4, selecting habitat evaluation factors and quantifying them using the analytic hierarchy process (AHP); determining the weights of habitat evaluation factors using the entropy weight method; and identifying the optimal ecological source area using the ant colony algorithm; S5, constructing a spatial absorption Markov model. This application enables more scientific and efficient formulation of species conservation strategies, providing a theoretical basis for biodiversity conservation.

Description

Technical Field

[0001] This invention relates to the field of animal ecology technology, and in particular to a habitat network optimization method that couples landscape and conflict resistance. Background Technology

[0002] Habitat fragmentation severely impacts the survival of wildlife worldwide, and identifying habitat networks and restoring habitat connectivity are crucial measures to address this global challenge. The Asian elephant (Elephas maximus), listed as an endangered species by the IUCN and a Class I protected wild animal in China, is a vital flagship species for Asian tropical forest ecosystems and plays a significant role in maintaining regional biodiversity. However, in recent years, habitat fragmentation and human-elephant conflict have intensified, posing a serious threat to the survival of its population.

[0003] Constructing habitat networks can enhance habitat connectivity, promote interspecies exchange and dispersal, and mitigate biodiversity loss. Habitat network identification is a prerequisite for constructing habitat networks, primarily including source area extraction and corridor identification. Related studies often directly use nature reserves as source areas, or employ methods such as natural discontinuity analysis and morphological spatial pattern analysis. However, these methods cannot comprehensively and objectively identify source areas, especially when data distribution is uneven or segmentation is unclear, their effectiveness may be unsatisfactory. Furthermore, when constructing resistance surfaces for large mammals, the impact of animal-human conflict on species dispersal behavior and decreased connectivity is downplayed. Existing research on habitat network identification for large mammals focuses on species' preference for landscape features, without clearly distinguishing and independently considering the connectivity decline caused by conflict. Therefore, the impact of natural landscape resistance and conflict-induced resistance is often confused, weakening the influence of conflict. However, large mammals frequently conflict with humans, and their impact on connectivity may be greater than that of landscape resistance. In conclusion, a method is needed to couple landscape and conflict resistance for more accurate habitat network identification, facilitating the subsequent conservation of large mammals. Summary of the Invention

[0004] The features and advantages of the present invention are set forth in part in the description which follows, or may be apparent from the description, or may be learned by practicing the invention.

[0005] To overcome the problems of existing technologies, this invention provides a habitat network optimization method that couples landscape and conflict resistance, comprising the following steps: S1. Data extraction and preprocessing: The data includes: LULC, topography, normalized differential vegetation index, multi-year average precipitation, roads, species distribution and accident data; S1. Data extraction and preprocessing; the data includes: LULC, topography, normalized differential vegetation index, multi-year average precipitation, roads, species distribution and accident data; S2. Calculate the species habitat suitability index using the HSI model and determine the comprehensive weight of each evaluation factor; S3. The natural breakpoint method is used to classify the normalized species habitats. S4. Select habitat evaluation factors and use the analytic hierarchy process (AHP) to quantify the evaluation factors; determine the weights of the habitat evaluation factors using the entropy weight method; and use the ant colony algorithm to identify the optimal ecological source area. S5. Construct a spatial absorption Markov model to determine the resistance caused by the landscape matrix and the resistance formed by conflict, and then calibrate the landscape resistance parameters and conflict resistance parameters. S6. Extract ecological corridors and evaluate their importance.

[0006] Preferably, the calculation formula of the HSI model is as shown in formula (1). Formula (1), In the formula, For the first One evaluation factor, For the first The weights of each factor, The number of evaluation factors.

[0007] Preferably, the species habitat classification includes optimal habitat, relatively suitable habitat, marginal habitat and non-habitat.

[0008] Preferably, the optimal identification of the ecological source area using the ant colony algorithm includes: The ecological source area probability function is used to reflect the habitat utility of each grid cell based on the information concentration between different grid cells; the ecological source area probability function is shown in formula (2): Formula (2), In the formula, Indicates the first Only ants Select the grid unit used to form the ecological source area at any time. The probability, Represents grid cell Heuristic information, These are the effects of pheromone concentration and heuristic information on ant selection, respectively. The tabu list ensures that ants can only choose from unvisited or accessible cells. Indicates the current grid cell The concentration of pheromones.

[0009] Preferably, the The calculation formula is shown in formula (3): Formula (3), in, This is the habitat utility function. It is the sum of the utility values ​​of all habitats in the study area.

[0010] Preferably, the The calculation formula is shown in formula (4): Formula (4), In the formula, The pheromone evaporation coefficient controls the decay of pheromones. By the Only one ant in the grid cell The amount of pheromone left on the path is related to the quality of the path, that is, it is related to the value of the overall objective function. Among them, the The calculation formula is shown in formula (5): Formula (5), In the formula It is with the central unit The distance; Represents pheromone intensity; It is the overall objective function; the formula for calculating the overall objective function is shown in formula (6). , In the formula, Let be the habitat utility function. It is a function of space compactness; The formula for calculating the spatial compactness function is shown in formula (7): Formula (7), In the formula, A represents the total area of ​​the source region. The perimeter of the source area.

[0011] Preferably, the construction of the spatial absorbing Markov model includes the following steps: S501. Construct a probability matrix P to represent the processes of movement and death. The P matrix can be expressed as: , in, This represents the probability of absorption due to natural death resistance. This represents the absorption probability caused by the resistance of the human-image conflict; in each row, ; S502, Calibration Parameter: Using the reciprocal of the species' habitat suitability... Parameterization; parameterization using the species' annual survival rate. Calibrate conflict drag parameters using the Maxent model. .

[0012] Preferably, the calibration of conflict drag parameters using the Maxent model is... Specifically, the following steps are included: S50201. Select human interference-related factors as conflict resistance evaluation factors. Use correlation analysis and variance inflation factor test to screen these evaluation factors. Input the occurrence points of the selected factors and species in construction land and farmland into the MaxEnt model. Randomly select 75% of the occurrence point data for model training and 25% of the occurrence point data for model validation. Finally, obtain the probability results. S50202: Using the relationship between the Poisson distribution and the complementary log-log link function, the minimum and maximum occurrence rates are converted into the probability of each pixel at each time step.

[0013] Preferably, the extraction of ecological corridors and evaluation of their importance specifically includes the following steps: S601. Calculate the net access rate of space based on the spatial absorption Markov model to obtain habitat network connectivity. S602. Use the minimum cumulative resistance model to extract ecological corridors; S603. Evaluation of the importance of ecological corridors: The importance of ecological source areas in ecological corridors is evaluated by landscape connectivity, and the interaction intensity between two ecological patches is calculated by gravity model to judge the importance of ecological corridors between ecological patches.

[0014] Preferably, the landscape connectivity is calculated using formula (8): Formula (8), In the formula, Total number of plaques; plaque The area; Indicates the species in the patch The maximum probability of diffusion between them Total landscape area; For landscape connectivity, ; The calculation formula for the gravity model is shown in formula (9): Formula (9), In the formula, plaque The strength of their interaction; plaques The resistance value; plaques The area; plaques The cumulative resistance value of the inter-ecological corridor; This represents the maximum cumulative resistance of all ecological corridors.

[0015] The beneficial effects of this invention are: This application introduces conflict resistance to extend the spatial absorption Markov model, coupling landscape resistance and conflict resistance to optimize habitat networks. Traditional habitat network studies often downplay the complex impacts of human disturbance on habitat networks, which may lead to biased or inefficient conservation planning. This study addresses this deficiency, providing a more scientific basis for Asian elephant conservation decisions. By coupling landscape resistance and conflict resistance, this application successfully identified key corridors between Asian elephant ecological homelands. The modeling method in this application has strong scalability. Different absorption states can be introduced based on different species or situations, thereby formulating differentiated conservation strategies and more effectively addressing the trade-offs between multiple threats. Therefore, the habitat network optimization method coupling different resistances can more scientifically and efficiently formulate species conservation strategies, providing a theoretical basis for biodiversity conservation. Attached Figure Description

[0016] The present invention will be described in detail below with reference to the accompanying drawings and examples. The advantages and implementation methods of the present invention will become more apparent from this description. The accompanying drawings are for illustrative purposes only and do not constitute any limitation on the present invention. In the accompanying drawings: Figure 1 This is a flowchart of a habitat network optimization method that couples landscape and conflict resistance in a specific embodiment of the present invention; Figure 2 In this specific embodiment of the invention, the results of identifying the ecological origin of Asian elephants under the scenario of overseas habitat are disregarded; Figure 3 The specific embodiments of the present invention take into account the identification results of Asian elephant ecological origins in the context of overseas habitats; Figure 4 This specific embodiment of the invention does not consider the Asian elephant habitat network in the context of overseas habitats; Figure 5 This invention provides a specific embodiment of the Asian elephant habitat network that takes into account the context of overseas habitats. Detailed Implementation

[0017] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0018] like Figure 1 As shown, this invention provides a habitat network optimization method that couples landscape and conflict resistance. This embodiment aims at optimizing the habitat network of Asian elephants, and the optimization method specifically includes the following steps: S1. Data Extraction and Preprocessing: The data includes: LULC (Luminous Naturalized Land Cover) data, topography, Normalized Differential Vegetation Index (NDVI), multi-year average precipitation, roads, Asian elephant distribution, and incident data. LULC data is sourced from Esri Land Cover, with a spatial resolution of 10 m, covering six LULC types: forest, sparse vegetation, cultivated land, built-up land, bare land, and water bodies. Sentinel-2 images with cloud cover less than 15% during the rubber tree leaf fall and new leaf stages were used to calculate the Normalized Burn Ratio (NBR). The leaf fall images were synthesized using the minimum pixel value, while the new leaf stage images were synthesized using the maximum pixel value. The NBR difference method was used to extract the distribution of rubber forests in the study area. A stratified sampling method based on LULC type was used, combined with high spatial resolution remote sensing imagery from Google Earth™, to identify 3248 sample points. The confusion matrix method was used to evaluate the accuracy of the LULC data including rubber forest types in the study area, and the overall accuracy was calculated to be 83%. Topographic data were derived from ALOS-PALSAR DEM data with a spatial resolution of 12.5 m. Based on this DEM data, the following parameters were calculated using trigonometric functions: slope (slope = arctan(elevation difference / horizontal distance)), aspect (aspect = X-direction slope / Y-direction slope), surface roughness (surface roughness = 1 / cos(slope)), slope position (using the topographic position index), curvature (based on the second derivative of elevation data), distance to valley, and distance to ridge (calculated based on slope position results). These calculated topographic indicators were used as natural habitat evaluation factors for subsequent Asian elephant habitat assessment. The NDVI data used were first calculated based on Sentinel-2 images with cloud cover less than 10% for the year, and then generated using the median composite method. Precipitation data came from the WorldClim database, representing the average annual precipitation from 1970 to 2000. Road data came from OpenStreetMap; population density data came from WorldPop; Asian elephant distribution range data came from the IUCN; and Asian elephant incident data came from the Asian Elephant Research Center of the National Forestry and Grassland Administration, Xishuangbanna Pacific Insurance Co., Ltd., and domestic news data from China. After geocoding all the above data, they were resampled to a spatial resolution of 10 m using the nearest neighbor method.

[0019] S2. Calculate the habitat suitability index for Asian elephants using the HSI model. The calculation formula is shown in Formula 1: Formula 1, In the formula, For the first One evaluation factor, For the first The weights of each factor, To determine the number of evaluation factors, a weighted average of the weights obtained from both the analytic hierarchy process (AHP) and the entropy weight method is calculated, and the final weight of each evaluation factor is determined.

[0020] S3. Using the natural discontinuity method, the normalized species habitats are classified into four levels: optimal habitat (0.65–1), moderately suitable habitat (0.54–0.65), marginal habitat (0.40–0.54), and no habitat (0–0.40).

[0021] S4. Select habitat evaluation factors and use the analytic hierarchy process (AHP) to quantify the evaluation factors; determine the weights of the habitat evaluation factors using the entropy weight method; and use the ant colony algorithm to identify the optimal ecological source area. Since habitat suitability assessment is limited to natural habitats, frequently occurring factors such as climate, topography, and land cover were selected as habitat evaluation factors. Correlation analysis and variance inflation factor tests were used to further screen 13 natural habitat evaluation factors, and these factors were then processed according to their positive / negative attributes. The attributes and descriptions of the selected natural habitat evaluation factors are shown in Table 1. Table 1. Attributes and descriptions of natural habitat evaluation factors.

[0022] The Analytic Hierarchy Process (AHP) was used to quantitatively evaluate habitat assessment factors. The target layer of the hierarchical structure was the suitability of Asian elephant habitat; the criterion layers included climate, topography, and land cover factors; and the indicator layer consisted of the 13 final natural habitat assessment factors. A 1–9 scaling method was used to assess the relative importance of each indicator and construct a natural habitat factor evaluation judgment matrix. Each element in the judgment matrix represents the relative importance of one factor to another, where 1 indicates that both factors are equally important, and 9 indicates that one factor is extremely important. The eigenvalue method was used to solve for the largest eigenvalue of the judgment matrix and its corresponding eigenvector, where each element of the eigenvector represents the weight of each indicator. The sum of the weights is 1, meaning the eigenvector needs to be normalized by dividing each element of the eigenvector by the sum of all its elements to obtain the final weight value. To ensure the rationality of the judgment matrix, a consistency index was calculated. and consistency ratio To determine the consistency of the matrix, the calculation method is shown in Formulas 10 and 11: Formula 10, In the formula, It determines the dimension of the matrix. It is the largest eigenvalue.

[0023] Formula 11,

[0024] In the formula, Is related to matrix dimension The relevant random consistency indicators can be obtained by consulting the Saaty table.

[0025] if If the value is less than 0.1, the judgment matrix of that criterion layer is considered to meet the consistency requirement, reflecting that the decision-makers' judgments are consistent. If there are multiple criterion layers, the judgment matrix of each layer should be calculated separately. value.

[0026] The determination matrix scale and its meaning are explained in Table 2: Table 2. Judgment Matrix Scale and Explanation

[0027] The obtained natural habitat evaluation judgment matrix is ​​shown in Table 3: Table 3 Evaluation and Judgment Matrix of Natural Habitat Factors

[0028] The weights of habitat evaluation factors are determined using the entropy weight method. The specific steps are as follows: First, the raw data is standardized to eliminate the influence of different indicator units. Standardized data makes data from different units and magnitudes comparable. Positive indicators are calculated using formula line 12: Formula 12, The negative indicator is calculated using formula line 13: Formula 13, in, Indicates the first The sample at the th The original values ​​under each indicator This represents the standardized value. and The first The maximum and minimum values ​​of each indicator.

[0029] Secondly, the entropy value is calculated, and after standardization, the information entropy of each habitat evaluation factor is calculated. The method for calculating information entropy is shown in Formula 14: Formula 14, Indicates the first The sample at the th The proportion of each indicator is calculated as shown in Formula 15: Formula 15, in, It is a constant, and is generally taken as... , This represents the number of samples.

[0030] Finally, the weight of each habitat evaluation factor is calculated based on its information entropy using Formula 16: Formula 16, In the formula, The number of habitat evaluation factors. For the first The weights of each habitat evaluation factor.

[0031] To optimize the compactness and overall suitability of the ecological source area pattern, an ant colony algorithm is used for optimal identification of ecological source areas. During this identification process, the ant colony primarily searches based on the differences in information concentration between different grid cells. This concentration actually reflects the habitat utility of each grid cell, i.e., the... Only ants At any given time, a unit is selected to form the probability function of the ecological source area, as shown in Formula 2: Formula 2, In the formula, Indicates the first Only ants Select the grid unit used to form the ecological source area at any time. The probability, Represents grid cell The heuristic information, typically related to the suitability of the grid (or other environmental factors), instructs the ants on the priority of selecting grids. In this application, The calculation method is shown in Formula 3. Formula 3, in, This is the habitat utility function. It is the sum of the utility values ​​of all habitats in the study area. These are the effects of pheromone concentration and heuristic information on ant selection, respectively. The tabu list ensures that ants can only choose from those unvisited or accessible cells. Indicates the current grid cell The pheromone concentration affects the probability of an ant selecting a grid cell, and is updated as the ant makes its selection. The basic update formula is shown in Formula 4: Formula 4, In the formula, The pheromone evaporation coefficient controls the decay of pheromones. By the Only one ant in the grid cell The amount of pheromone left on the path is usually related to the quality of the path (i.e., the total objective function value), and its formula is shown in Formula 5: Formula 5, In the formula It is with the central unit The distance. The variable is used to address the neighborhood influence in site selection. In this application, it represents: if a location's neighboring pixels have already been included in the ecological source area, then that location should be more likely to be selected. This represents the intensity of pheromones. This is the overall objective function, and its calculation method is shown in Formula 6: Formula 6, In the formula, Let be the overall objective function. Let be the habitat utility function. The spatial compactness function is used to optimize the spatial layout of ecological source areas, making them present a compact and orderly pattern. This is achieved by considering differences in natural habitat distribution, population density, and actual species distribution, in order to guide ecological source areas to concentrate in regions with high suitability, low population density, and proximity to Asian elephant distribution areas.

[0032] The method for calculating the space compactness function is shown in Equation 7: Formula 7, In the formula, A represents the total area of ​​the source region. The perimeter of the source area is important, as both habitat suitability and habitat patch integrity are crucial for the survival of Asian elephants. All weights are 1.

[0033] S5. Construct a spatial absorption Markov model to determine the resistance caused by the landscape matrix and the resistance formed by conflict, and then calibrate the landscape resistance parameters and conflict resistance parameters. The input values ​​for the number of ant colonies (i.e. the number of ecological source area pixels to be extracted) are 18%, 20%, 25% and 30% of the study area area, respectively. The grid side length is set to 500 m and the total number of iterations is set to 1000. Each ecological source area percentage scheme is repeated 10 times, and the highest value of the overall objective function is taken as the final result.

[0034] Constructing a spatially absorbing Markov model: This involves constructing a probability matrix P to represent the processes of motion and death, which is composed of instantaneous matrices. (e.g., landscape diffusion resistance) and mortality rates between one or more absorption states Composition. This model can distinguish between the behavioral response of animals spreading in the matrix and the risk of death. For a landscape consisting of C pixels, matrix P can be written as: , In the formula, Let C×C be the instantaneous state transition matrix. Let C be a C×1 vector consisting of the transition probabilities from the instantaneous state to the absorption state. It is the zero vector of 1×C. For matrix The first in OK The elements of the column represent the state within a time step. Switch to The probability, 𝑝 𝑖 For matrix The Middle OK The elements of the column (i.e. The (row element), representing the state within a time step. The probability of conversion to absorption; since death still occurs even after the conversion from absorption to absorption, therefore... In each line, =1.

[0035] This model is applied to Asian elephant habitat network optimization to distinguish between resistance caused by the landscape matrix and resistance caused by conflict. The model is extended to multiple absorption states, considering two different absorption states: one caused by resistance due to natural mortality, and the other caused by resistance due to human-elephant conflict. The new P-matrix can then be expressed as: , This matrix is ​​a transition matrix of order (C+2)×(C+2), where, The probability of absorption due to natural death resistance represents the probability of absorption due to human-elephant conflict resistance, i.e., the conflict resistance parameter. In each row, Therefore, this extension can be decomposed into different types of motion absorption and landscape connectivity.

[0036] Based on this matrix, the inverse pair of the instantaneous matrix of Asian elephant habitat suitability is used. Parameterization is performed; based on the annual survival rate of Asian elephants being 0.97, the mortality probability at each time step is 0.97, and this is used for parameterization. Using the Maxent model, the probability of human-elephant conflict is predicted based on Asian elephant conflict points, thereby calibrating the parameters. .

[0037] In this application, landscape resistance represents the willingness and cost of organisms to traverse a specific environment, where low resistance indicates easy dispersal and a tendency for Asian elephants to appear, while high resistance indicates restricted dispersal; conflict resistance represents the mortality rate of species traversing a specific environment, where low resistance indicates low conflict mortality, while high resistance indicates a high conflict incidence.

[0038] Since Asian elephants move at an average speed of 0.35–0.52 km / h, with a time step of 1 hour, a landscape resistance layer can be created by subtracting the natural habitat suitability from 1, in order to calibrate the landscape resistance parameters. .

[0039] Calibration of conflict drag parameters using the MaxEnt model This refers to the absorption probability caused by the conflict between humans and images. The specific steps are as follows: Human disturbance-related factors were selected as conflict resistance evaluation factors. These factors included the proportion of frequently active LULC types (including rubber plantation frequency, building frequency, and farmland frequency) within each 500×500 m pixel, as well as the distance from each pixel to these LULCs (including the frequency to rubber plantation, building, and farmland) and population density. Correlation analysis and variance inflation factor tests were used to screen these evaluation factors. The selected factors and the occurrence points of Asian elephants in built-up areas and farmland were input into the MaxEnt model. 75% of the occurrence point data were randomly selected for model training, and 25% of the occurrence point data were used for model validation, ultimately yielding probability results.

[0040] Using the relationship between the Poisson distribution and the complementary log-log link function, the minimum and maximum occurrence rates are converted into probabilities per pixel per time step, given a region. and a location Due to the incidence It follows a Poisson distribution, and the incidence rate is calculated as shown in Formula 17: Formula 17, In the formula, It is an estimated quantity per unit area on a log-linked scale. The area covered by the number of conflicts. The calculation method is shown in Formula 18: Formula 18, In the formula, Number of conflicts.

[0041] The probability of a conflict can be derived using complementary log-log link functions, as shown in Equation 19. Equation 15 is shown below: Formula 19, S6. Extract ecological corridors and evaluate their importance. The specific steps are as follows: S601. Calculate the net access rate of space based on the spatial absorption Markov model to obtain habitat network connectivity. Through the basic matrix Calculate the net access rate of the space, where, It is an identity matrix with an access rate of The first of the matrix An element, representing the individual's location. Spread to location Total time spent. Net visits are represented by location. Exercise to and The net movement probability is proportional to the product of the time difference at a certain location and the instantaneous state transition probability, thus yielding Formula 20, which is shown below: Formula 20, In the formula, It is a matrix Element.

[0042] Net access rate can quantify the connectivity of individuals in a landscape through a specific grid based on a specific start and end location.

[0043] S602. Using the minimum cumulative resistance model, ecological corridors are extracted. The extraction formula is shown in Formula 21: Formula 21, In the formula, It is an unknown positive function that reflects the positive correlation between the minimum resistance at a point in space and its distance to all sources and the characteristics of the landscape base surface; As the source To space unit distance, Representing spatial units The resistance coefficient is calculated; the normalized connectivity is subtracted from 1 as the resistance surface to obtain the minimum cost path from the source to the target patch. After eliminating the repetitive corridors that cross the ecological source area, the target ecological corridor is obtained.

[0044] S603. Evaluating the Importance of Ecological Corridors. The importance of ecological source areas within ecological corridors is evaluated using landscape connectivity, which is calculated using Formula 8: Formula 8, In the formula, Total number of plaques; plaque The area; Indicates the species in the patch The maximum probability of diffusion between them (considering only one direction). Total landscape area; For landscape connectivity, 0 < <1, The value increases with increasing landscape connectivity.

[0045] The importance of patch removal for maintaining landscape connectivity is calculated using Equation 22: Formula 22, This indicates the removal of the patch from the landscape. value; This indicates the importance of removing patches in maintaining landscape connectivity; the higher the value, the greater the contribution of the patch to the overall landscape.

[0046] The interaction strength between two ecological patches was calculated using a gravity model to assess the importance of ecological corridors between the patches. The calculation method is shown in Formula 9. Formula 9, In the formula, plaque The strength of their interaction; plaques The resistance value; plaques The area; plaques The cumulative resistance value of the inter-ecological corridor; This represents the maximum cumulative resistance of all ecological corridors. The higher the value, the more important the corridor.

[0047] Using the above method to identify the ecological homelands of Asian elephants, if only ecological homelands within China are considered, the identification results are as follows: Figure 2 As shown, the ecological source areas are mainly concentrated in two districts and counties of Pu'er, three cities and counties of Xishuangbanna, and three ecological source areas within two counties of Lincang. Short corridors exist between the ecological source areas of Lincang and those in other parts of China, such as... Figure 4 As shown; the environmental carrying capacity of the ecological source area is Head / 102 km 2 Because Asian elephants exhibit seasonal transboundary migration, and in order to further rationally construct ecological corridors to accommodate the growth of Asian elephant populations, it is also necessary to identify ecological source areas in Myanmar and Laos bordering China. This is to facilitate the formulation of better species conservation strategies and, consequently, to adapt to the growth of Asian elephant populations. Following the methods described above, when considering extant habitats, such as... Figure 3In addition to the aforementioned areas, the China-Laos cross-border ecological source area is located in the region bordering Mengla County in Xishuangbanna with Luang Namtha and Phongsaly provinces in Laos. This area is the main distribution area for the Mengla Asian elephant population. The China-Myanmar cross-border ecological source area is located in the region bordering Cangyuan County in Lincang City with Kunlong and Lashio townships in Myanmar. This area is the main distribution area for the Nan Gunhe Asian elephant population. The corridors between the China-Myanmar cross-border ecological source area and the ecological source areas within China are mostly long corridors, such as... Figure 5 As shown; ecological source patches located in the China-Laos cross-border area and ecological source patches in Xishuangbanna and Pu'er. Higher, with the increase in the proportion of ecological source areas, the ecological source patches in the China-Myanmar cross-border area Gradually increasing; short corridors near the Pu'er patches in Xishuangbanna Larger and more important. The environmental carrying capacity of the ecological source area (i.e., per 100 km²) 2 The ecological source area can support a higher number of Asian elephants, reaching [amount not specified]. Head / 102km 2 Therefore, in order to maintain the long-term stability and security of Asian elephant populations, the construction of Asian elephant habitat networks should take into account overseas habitats and improve the connectivity of habitat networks.

[0048] This application introduces conflict resistance to extend the spatial absorptive Markov model, coupling landscape resistance and conflict resistance to optimize habitat networks. Traditional habitat network studies often downplay the complex impacts of human disturbance on habitat networks, which may lead to biased or inefficient conservation planning. This study addresses this deficiency, providing a more scientific basis for Asian elephant conservation decisions. By coupling landscape resistance and conflict resistance, this application successfully identified key corridors between Asian elephant place of origin. The results show that the superposition of landscape and conflict resistance often has a greater restrictive effect on connectivity. Conflict-induced mortality reduces connectivity between place of origin, thereby weakening corridor function and reducing the likelihood of species migration and gene exchange, turning these areas into "bottlenecks" or "disconnects." This indicates that considering the impact of conflict on connectivity in habitat network optimization can provide a more scientific reference for the construction of Asian elephant habitat corridors. Ignoring conflict resistance or failing to clearly distinguish the roles of the two types of resistance may lead to underestimating the negative impact of human disturbance on habitat network connectivity, resulting in erroneous decisions. This embodiment uses Asian elephants as a case study, but since wild animals often conflict with humans, this method and findings are applicable to other wild animals. The modeling approach presented in this application is highly scalable. Different absorption states can be introduced based on different species or circumstances, thereby enabling the development of differentiated conservation strategies that more effectively address the trade-offs between multiple threats. Therefore, the habitat network optimization method coupled with different resistance levels can more scientifically and efficiently formulate conservation strategies, providing a theoretical basis for biodiversity conservation.

[0049] The preferred embodiments of the present invention have been described above with reference to the accompanying drawings. Those skilled in the art can implement the present invention in various modifications without departing from its scope and spirit. For example, a feature shown or described in one embodiment can be used in another embodiment to obtain yet another embodiment. The above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. All equivalent changes made based on the description and drawings of the present invention are included within the scope of the present invention.

Claims

1. A habitat network optimization method that couples landscape and conflict resistance, characterized in that, It comprises the following steps: S1, data extraction and preprocessing; the data comprises LULC, terrain, normalized difference vegetation index, multi-year average precipitation, road, species distribution and accident data; S2, using HSI model to calculate species habitat suitability index, determining the comprehensive weight of each evaluation factor; S3, using natural breakpoint method to grade the normalized species habitat; S4, selecting habitat evaluation factors, using analytic hierarchy process to quantitatively evaluate the habitat evaluation factors; determining the weight of the habitat evaluation factors through entropy weight method; and using ant colony algorithm to identify the optimal ecological source; S5, constructing a spatial absorption Markov model to determine the resistance caused by landscape matrix and the resistance formed by conflict, and then calibrating the landscape resistance parameter and the conflict resistance parameter; S6, extracting ecological corridors and evaluating their importance.

2. The habitat network optimization method of coupling landscape and conflict resistance according to claim 1, characterized in that, The calculation formula of the HSI model is shown in formula (1) Equation (1) wherein is the number of the evaluation factor, is the weight of the factor, is the number of evaluation factors.

3. The habitat network optimization method of coupling landscape and conflict resistance according to claim 1, characterized in that, The species habitat grading comprises optimal habitat, relatively suitable habitat, marginal habitat and non-habitat.

4. The habitat network optimization method of coupling landscape and conflict resistance according to claim 1, characterized in that, The optimal identification of the ecological source through the ant colony algorithm comprises: An ecological source probability function is used to reflect the habitat utility of each grid unit according to the information concentration between different grids; the ecological source probability function is shown in formula (2): Equation (2) In the formula, Indicates the first Only ants Select the grid unit used to form the ecological source area at any time. The probability, Represents grid cell Heuristic information, These are the effects of pheromone concentration and heuristic information on ant selection, respectively. The tabu list ensures that ants can only choose from unvisited or accessible cells. Indicates the current grid cell The concentration of pheromones.

5. The habitat network optimization method of coupling landscape and conflict resistance according to claim 4, characterized in that, The The calculation formula is shown as formula (3): Equation (3) wherein, is the habitat utility function, is the sum of all habitat utility values for the study area.

6. The habitat network optimization method of coupling landscape and conflict resistance according to claim 4, characterized in that, The The calculation formula is shown as formula (4): Equation (4) In the formula, The pheromone evaporation coefficient controls the decay of pheromones. By the Only one ant in the grid cell The amount of pheromone left on the path is related to the quality of the path, that is, it is related to the value of the overall objective function. In the formula, the formula of the calculation of the is shown in formula (5). Equation (5) In the formula is the distance from the center unit ; represents the pheromone intensity; is the total objective function; the calculation formula of the total objective function is shown as formula (6) , wherein is a habitat utility function, is a spatial compactness function; The calculation formula of the spatial compactness function is shown in formula (7): Equation (7) wherein A is the total area of the source, is the perimeter of the source.

7. The habitat network optimization method of coupling landscape and conflict resistance according to claim 6, characterized in that, The construction of the spatial absorption Markov model comprises the following steps: S501, constructing a probability matrix P to reflect the movement and death process, and the P matrix can be expressed as: , wherein, represents the absorption probability due to natural death resistance, represents the absorption probability due to man-land conflict resistance; in each row, ; S502, calibrate parameters: use inverse of habitat suitability of species to parameterize; parameterize with annual survival rate of species ; use Maxent model to calibrate conflict resistance parameter .

8. The habitat network optimization method of coupling landscape and conflict resistance according to claim 7, characterized in that, The conflict resistance parameter is calibrated by using a Maxent model Specifically comprising the following steps: S50201, selecting human disturbance related factors as conflict resistance evaluation factors, using correlation analysis and variance inflation factor test to screen these evaluation factors, inputting the screened factors and the occurrence points of the species in the construction land and farmland into a MaxEnt model, randomly selecting 75% of the occurrence point data for training the model, and using 25% of the occurrence point data for model verification, and finally obtaining a probability result; S50202, using the relationship between Poisson distribution and complementary log-log link function to convert the minimum and maximum occurrence rate into the probability of each pixel per time step.

9. The habitat network optimization method of coupling landscape and conflict resistance according to claim 1, characterized in that, The extraction of the ecological corridors and the evaluation of their importance specifically comprises the following steps: S601, calculating the spatial net access rate based on the spatial absorption Markov model to obtain habitat network connectivity; S602, using a minimum cumulative resistance model to extract ecological corridors; S603, evaluating the importance of the ecological corridors: using landscape connectivity to evaluate the importance of the ecological source in the ecological corridors, and using a gravity model to calculate the interaction strength between two ecological patches to evaluate the importance of the ecological corridors between the ecological patches.

10. The habitat network optimization method of coupling landscape and conflict resistance according to claim 9, characterized in that, The landscape connectivity is calculated by formula (8): Equation (8) In the formula, Total number of plaques; plaque The area; Indicates the species in the patch The maximum probability of diffusion between them Total landscape area; For landscape connectivity, 0 < <1; The calculation formula of the gravity model is shown in formula (9): Equation (9) wherein is the patch interaction strength between; is the resistance value of patch ; is the area of patch ; is the cumulative resistance value of the ecological corridor between patch ; is the maximum value of the cumulative resistance of all ecological corridors.

Images