Method for selecting preferential protection species and species set by combining food web stability evaluation

Abstract

The invention relates to a method for selecting a preferential protection species and a species set by combining food web stability evaluation, which comprises the following steps of: acquiring species data, determining a predatory relationship among species, calculating a nutrition level and a food source ratio among the species, and constructing a food web model on the basis of the predatory relationship; priority protection species are screened based on a directed and weighted food web model, wherein the priority protection species comprise key species, umbrella protection species and rare umbrella protection species. The method breaks through the limitation that the screening standards of the preferential protection species are not uniform and the screening method is time-consuming and labor-consuming, and the preferential protection species and the species set are accurately screened, so that the protection effect that lines are carried by points and surfaces are carried by lines is achieved, and the protection benefits of wild animals are greatly improved.

Description

Technical Field

[0001] The present invention relates to the fields of conservation biology and ecology, and in particular to a method for selecting priority species and species sets for protection in combination with food web stability assessment. Background Art

[0002] By scientifically and accurately identifying priority protected species in the ecosystem, we can achieve "big returns with small investments."

[0003] Umbrella species and keystone species are often considered priority species for conservation, a strategy often seen as a "shortcut" to biodiversity conservation. An umbrella species is a species whose conservation benefits its co-occurring species and the ecosystems in which they inhabit. A keystone species is a species with significant ecological impact, with its impact on an ecosystem disproportionate to its abundance and biomass. While these two approaches offer a "shortcut" for biodiversity conservation, current umbrella and keystone species designations still face challenges such as inconsistent selection criteria, subjective and time-consuming selection methods, and a lack of systematic and complementary approaches to selecting priority species. For example, designating the giant panda as an umbrella species often confuses the concepts of flagship and umbrella species. Existing technical solutions, combining infrared cameras with remote sensing, have confirmed that the current nature reserve system designed with the giant panda as an umbrella species does not adequately cover key landscapes for several species. In other words, research suggests that the giant panda's conservation impact as an umbrella species has been overestimated, and that it is more appropriately designated as a flagship species rather than an umbrella species. Therefore, mixing flagship species with umbrella species is a common problem, and it is even more necessary to clarify the differences between the two.

[0004] Therefore, there is an urgent need for a new method that takes a holistic and global perspective to accurately screen priority protected species and priority protected species sets represented by umbrella species and key species. Summary of the Invention

[0005] In order to solve the problems existing in the above-mentioned prior art, the purpose of the present invention is to provide a method for selecting priority protected species and species sets in combination with food web stability assessment. From the perspective of the food web, on the premise of ensuring the stability of the food web, on the one hand, the present invention develops a standardized process for constructing a directed, weighted food web for multi-source data input. The existing technology is limited by the difficulty of field sampling, etc., and the application of stable isotope values combined with the MixSIAR model to determine the limited number of species that can be included in the food web framework. The present invention relies on the established topological "big food web" (number of species >100) construction technology of multi-source data fusion analysis. Through multi-source data collection, it constructs the predation relationship matrix between species, the predation relationship characteristic matrix, the species typical trait matrix, etc. By fitting the predator-prey related traits, the predation relationship between candidate predators and prey is judged, and the feeding ratio between species is quantitatively determined, thereby constructing a directed, weighted "big food web" (number of species >100). On the other hand, the present invention, from the perspective of the food web, pioneers a new method for screening umbrella species and key species, breaking through the limitations of inconsistent screening standards and time-consuming and labor-intensive screening methods in the past, and accurately screening priority protected species and species sets to greatly improve the benefits of wildlife protection.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A method for selecting priority species and species groups for conservation based on food web stability assessment, including:

[0008] Obtaining species data, using the species data to determine predator-prey relationships, trophic level characteristics, and food source ratios among species, and constructing a food web model;

[0009] Priority protected species are screened according to the food web model, and the priority protected species include: keystone species, umbrella species and rare umbrella species.

[0010] Optionally, determining the predation relationship between the species comprises:

[0011] Establish the predator-prey matrix, the predator-prey feature matrix and the all species feature-species matrix respectively:

[0012] A=(a ij ) i×j

[0013] P=(p ik ) i×k

[0014] T=(t kj ) k×j

[0015] Where i is the predator, j is the prey, k is the species characteristics, A is the predator-prey matrix, (a ij ) i×j is the original predator-prey relationship between species, P is the predator-prey feature matrix, (p ik ) i×k is the predator's prey characteristic preference, T is the species characteristic-species matrix, (t kj ) k×j A characteristic of a species;

[0016] Using the predator-prey feature matrix and the all-species feature-species matrix, a fourth matrix is obtained:

[0017] M=P×T=(m' ij ) i×j

[0018] Among them, P is the predator-prey feature matrix, T is the all species feature-species matrix, (m' ij ) i×j is the probability of predation-prey relationship between species;

[0019] The fourth matrix is compared with the predator-prey matrix to determine the minimum threshold for the occurrence of a predator-prey predation relationship. Based on the minimum threshold, the predator's predation preference for species characteristics is mapped to each species to obtain a fifth matrix:

[0020]

[0021] Where θ represents the minimum threshold for the occurrence of predator-prey relationship, (a' ij ) i×j The ultimate predator-prey relationship between species.

[0022] Optionally, determining the food source ratio between the species comprises:

[0023] Calculate the average probability of predation-prey relationships between the species under the prediction of species traits:

[0024]

[0025] in, represents the average probability of predator i occurring at prey node j, f ij is the probability of predation relationship under single feature prediction, that is, m' under the corresponding species feature ij value, n is the number of species characteristics selected;

[0026] The relative frequency of the predator-prey relationship between the predator and all prey is determined by the average probability as the food source ratio:

[0027]

[0028] Among them, F ij represents the food source ratio of predator i to prey species j, is the average probability of the predation relationship between predator i and prey k.

[0029] Optionally, determining trophic level characteristics between species includes:

[0030] Tr i =1+∑Tr j ×F ij

[0031] Among them, Tr i represents the trophic level of predator i, Tr j is the trophic level of prey j, F ij is the food source ratio of predator i at predation node j.

[0032] Optionally, selecting the key species according to the food web model includes:

[0033] The food source ratio between species is used as the link weight in the food web model, and the importance of each species in the food web model is ranked using centrality indicators; the centrality indicators include degree centrality, closeness centrality, betweenness centrality and PageRank centrality;

[0034] Species in the food web model are removed in descending order of centrality index or randomly, and in the process of removing the species, the stability of the food web model is calculated, and the optimal centrality index is determined according to the stability, and the key species is determined based on the ranking result of the species according to the optimal centrality index.

[0035] Optionally, calculating the stability of the food web model includes:

[0036] The relative size of the largest connected subgraph is used to determine the extent of damage to the food web model after being attacked:

[0037]

[0038] Where N' is the number of species in the largest connected subgraph after the network is attacked, and N is the total number of species in the initial network when the network is not attacked;

[0039] The global network efficiency is used to measure the efficiency of information transmission between different species in the food web model:

[0040]

[0041] Where E(G) represents the global network efficiency of network G, n is the total number of species, d ij is the shortest path between species.

[0042] Optionally, selecting umbrella species according to the food web model includes:

[0043] Get the species' home range area and use it to update the umbrella strength index:

[0044]

[0045] Among them, TL i is the trophic level of food web node i, K i is the number of nodes adjacent to node i, S i is the sum of the weights of the edges connected to node i, α is a fixed value of 0.5, W(TL) represents the trophic level weight, and W(H) represents the home range area weight;

[0046] The updated umbrella strength index is used to select umbrella species in the food web model, and the species with the highest umbrella strength index ranking is used as the umbrella species.

[0047] Optionally, obtaining the home range area of the species includes:

[0048]

[0049] Where H i is the home range index of species i, X i is the home range area of species i, X min and X max Respectively represent the minimum and maximum home range areas of all species in the food web of all regions.

[0050] Optionally, selecting rare umbrella species according to the food web model includes:

[0051] On the basis of the updated umbrella protection strength index, the species protection level is integrated to obtain the rare umbrella species strength index:

[0052] RUSS(i)=K i (1-α) ×S i α ×(TL i ×W(TL)+PEI i ×W(PEI)+H i ×W(H)

[0053] Where W(PEI) is the weight of the protection effect index;

[0054] The rare umbrella species strength index is used to select rare umbrella species in the food web model, and the species with the highest rare umbrella species strength index is used as the rare umbrella species.

[0055] The beneficial effects of the present invention are:

[0056] The present invention constructs a food web model through a predation matrix based on existing data of different types of species; evaluates the stability of the food web through sequence removal, and determines the "key species" of the food web in combination with the complex network centrality index, determines the "umbrella species" through the "umbrella strength" index, and determines the rare "umbrella species" through the rare "umbrella strength" index, breaking through the problems of previous studies such as the difficulty of collecting data, inconsistent screening standards for priority protected species, and time-consuming and labor-intensive screening methods.

[0057] This invention uses a matrix method to calculate the food source ratio between species and construct a food web, improves the existing "umbrella strength" index and proposes a rare "umbrella strength" indicator and algorithm, providing a more scientific basis for the selection of priority protected species. It helps to improve the scientific nature and effectiveness of biodiversity conservation work while achieving the goal of saving protection funds. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0059] Figure 1 This is a flow chart of a method for selecting priority species and species sets for protection in combination with food web stability assessment according to an embodiment of the present invention.

[0060] Figure 2 Schematic diagram of the food source ratio between species in the Himalayan forest ecosystem according to an embodiment of the present invention;

[0061] Figure 3 A schematic diagram of the trophic levels of various species in the Himalayan forest ecosystem according to an embodiment of the present invention;

[0062] Figure 4 A schematic diagram of a food web of a Himalayan forest ecosystem according to an embodiment of the present invention;

[0063] Figure 5 The impact of topology-based sequence removal on the stability of the food web in the Himalayan forest ecosystem according to an embodiment of the present invention; Figure 5 (a) is the stability of the maximum connected subgraph under relative scale, Figure 5 (b) is the stability under the global efficiency of the network;

[0064] Figure 6 This is a schematic diagram of the results of the species protection strength of the Himalayan forest ecosystem according to an embodiment of the present invention;

[0065] Figure 7 This is a schematic diagram of the results of the rare umbrella protection strength of species in the Himalayan forest ecosystem according to an embodiment of the present invention. DETAILED DESCRIPTION

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0067] This embodiment discloses a method for selecting priority protected species and species sets in combination with a food web stability assessment, including: obtaining species data, using the species data to determine the predator-prey relationships, trophic level characteristics, and food source ratios between species, and establishing a food web model; selecting priority protected species based on the food web model, wherein the priority protected species include: keystone species, umbrella species, and rare umbrella species.

[0068] Specifically, this embodiment discloses a method for selecting priority protected species and species sets in combination with food web stability assessment, including: constructing a directed, weighted "large food web" (number of species > 100) based on existing species data; and selecting priority protected species and species sets based on the constructed food web.

[0069] Furthermore, data needed to construct the food web are collected based on experiments, books, literature, etc., including predation relationships between species, species traits, species home range area, and species protection level.

[0070] Furthermore, determining predation relationships between species involves establishing a predator-prey matrix, a predator-prey trait matrix, and a species-trait matrix. Using the predator-prey trait matrix and the species-trait matrix, a fourth matrix is obtained. The fourth matrix is compared with the predator-prey matrix to determine the minimum threshold for the occurrence of a predator-prey predation relationship. Based on the minimum threshold, the predator's predation preference for species traits is mapped to each species to obtain a fifth matrix.

[0071] Specifically, construct a predator-prey matrix and perform predator-prey relationship prediction:

[0072] To predict predator-prey relationships between species, a matrix × matrix approach is used. To create the matrix, the present invention converts all feature variables into binary format. The specific process for constructing the matrix is as follows:

[0073] First, establish the predator-prey matrix A, whose dimension is i×j, and the formula is as follows:

[0074] A=(a ij ) i×j

[0075] Where i is the predator and j is the prey. The nutritional interaction a between predator i and prey j is ij Represented by 1 or 0 (1 if predator i preys on j, 0 otherwise), the matrix A represents the documented predation relationships between predators and prey.

[0076] The second step is to establish the predator-prey feature matrix P, whose dimension is i×k, and the formula is as follows:

[0077] P=(p ik ) i×k

[0078] Where k is the species trait. For each species characteristic, a predator-prey characteristic matrix P is established separately. The relationship between predator i and prey characteristics k is p ik The matrix P represents the predator's prey preference, expressed as the relative frequency of predator i preying on prey feature k.

[0079] The third step is to establish the species matrix T of all species characteristics, whose dimension is k×j, and the formula is as follows:

[0080] T=(t kj ) k×j

[0081] For each species’ trait, a species trait-species matrix T is established separately, and the relationship between species trait k and species j is t kj Represented by 1 or 0 (1 if species j has species characteristic k, otherwise 0), the matrix T represents the characteristics of the species.

[0082] The fourth step is to establish the matrix A′ to visualize the predator's predation preference for species characteristics for each species. The formula is as follows:

[0083] M=P×T=(m' ij ) i×j

[0084]

[0085] Where θ represents the minimum threshold for the occurrence of a predator-prey relationship, which is derived from the comparison of matrix M and matrix A, and m' ij with a ij The minimum overlap is the value of θ. For each species characteristic, a matrix A′ can be obtained. The predator-prey relationships that occur under each species characteristic are screened out. On this basis, obviously unreasonable predator-prey relationships are removed and the food web is constructed accordingly.

[0086] Furthermore, determining the food source ratio between species includes: calculating the average probability of predation relationships between species occurring under the prediction of species trait characteristics; and determining the relative frequency of predation relationships between predators and all prey through the average probability as the food source ratio.

[0087] Specifically, we accurately calculate the prey source ratio between species and calculate the prey source ratio based on the probability of predation relationships in the above matrix. First, we calculate the average probability of predation relationships under the prediction of species traits, using the following formula:

[0088]

[0089] Where, represents the average probability of predator i occurring at predator node j, f ij is the probability of predation relationship under single feature prediction (the value is m' under the corresponding species feature ij value), n is the number of selected species characteristics.

[0090] Calculate the relative frequency of predation between predator i and all prey as the food source ratio, the formula is as follows:

[0091]

[0092] Where, F ij represents the food source ratio of predator i to prey species j, is the average probability of the predator-prey relationship between predator i and prey k.

[0093] Accurately calculate the trophic level of species based on the food source ratio between species. The formula is as follows:

[0094] Tr i =1+∑Tr j ×F ij

[0095] Where Tr i represents the trophic level of predator i, Tr j is the trophic level of prey j, F ijis the food source ratio of predator i at predation node j, and the present invention sets the primary producer trophic level to 1.

[0096] Furthermore, the food web model was constructed by combining the predator-prey relationship between food web species, the trophic level characteristics of each node and the food source ratio to construct a quantitative and weighted food web.

[0097] Food web theory can provide a natural framework for understanding the complex interactions between species, revealing important ecological processes such as material circulation and energy flow, and understanding the stability mechanism of ecosystems against disturbances, thus laying a solid theoretical foundation for conservation practice. However, in conservation practice, the food web contains many species and complex relationships, making it difficult to apply directly. An effective method is that the food web can be replaced by a network, in which species are nodes and connections represent the predator-prey relationship between them. By identifying key nodes that are related to the integrity and stability of the network, these umbrella species and key species with the significance of "leading points to lines and lines to areas" can be accurately screened out, which is crucial to improving the effectiveness of biodiversity conservation.

[0098] Furthermore, selecting key species based on the food web model includes: using the food source ratio between species as the link weight within the food web model, and using centrality indicators to rank the importance of each species in the food web model; centrality indicators include: degree centrality, closeness centrality, betweenness centrality and PageRank centrality; species in the food web model are removed in descending order or randomly according to the centrality indicator, and in the process of removing species, the stability of the food web model is calculated, and the optimal centrality indicator is determined according to the stability, and the key species are determined based on the ranking results of the species based on the optimal centrality indicator.

[0099] Furthermore, calculating the stability of the food web model includes: using the relative size of the largest connected subgraph to determine the extent of damage to the food web model after an attack, and using the global efficiency of the network to measure the efficiency of information transmission between different species in the food web model.

[0100] Specifically, we accurately select priority species for protection—key species—and combine them with food web stability indicators to select "key species." Based on the construction of a food web model, we consider the food source ratio between species as the weighted link within the food web. We use four representative centrality indicators—degree centrality, closeness centrality, betweenness centrality, and PageRank centrality—to rank the importance of species in the food web. We further validate the effectiveness of centrality indicators through topology-based sequence removal experiments to accurately identify key species in the food web.

[0101] Topological removal: Use topological removal to remove sequences. Removal methods are mainly divided into random removal and deterministic removal. Random removal mainly simulates the random disappearance of species in nature, while deterministic removal mainly removes certain species in a targeted manner based on the importance of species. Based on the calculation results of the above four centralities, the species in the food web are sorted, and five schemes are used to determine the priority of species to be deleted: (1) remove species in descending order of degree centrality; (2) remove species in descending order of betweenness centrality; (3) remove species in descending order of closeness centrality; (4) remove species in descending order of PageRank centrality; (5) remove species randomly. In the process of gradually removing species, observe the changes in network stability. Repeat the steps for each sequence removal until no secondary extinction occurs, and then start the next round of sequence removal. When a species loses all its resources, it is considered to be secondary extinct. Primary producers will not experience secondary extinction.

[0102] The following two indicators are used to indicate the stability of the food web during the topological sequence removal process:

[0103] 1) Relative size of the largest connected subgraph:

[0104] The relative size of the maximum connected subgraph refers to the ability of the network to maintain connections between the remaining species when it is damaged. It is used to indicate the degree of damage to the network after an attack and is calculated using the following formula:

[0105]

[0106] Where N' is the number of species in the largest connected subgraph after the network is attacked, and N is the total number of species in the initial network when the network is not attacked.

[0107] 2) Global network efficiency:

[0108] Since global efficiency can reflect the overall connectivity of the network and measure the efficiency of information transmission between different species in the network, this embodiment uses global efficiency to indicate the change in food web stability during species removal. The formula is as follows:

[0109]

[0110] Where E(G) represents the global network efficiency of network G, n is the total number of species, d ij is the shortest path between species.

[0111] Furthermore, selecting umbrella species according to the food web model includes: obtaining the species' home range area, updating the umbrella strength index using the species' home range area, and selecting the umbrella species in the food web model using the updated umbrella strength index.

[0112] Specifically, to accurately select priority protected species, namely "umbrella species", this invention innovatively integrates the species' home range area (HomeRange) characteristics based on the traditional "umbrella species strength" index to update the umbrella strength index to more accurately determine "umbrella species". The specific formula is as follows:

[0113] The species home range area was normalized and the original data was converted into data within a specific range [0-1] to eliminate the dimension and order of magnitude effects. The species home range area index was obtained using the following formula:

[0114]

[0115] Where H i is the home range index of species i, X i is the home range area of species i, X min and X max Respectively represent the minimum and maximum home range areas of all species in the food web of all regions.

[0116] The updated "protection strength" index formula is as follows:

[0117]

[0118] Where, TL i is the trophic level of food web node i, K i is the number of nodes adjacent to node i (degree), S i where α is the sum of the weights of all edges connected to node i (strength) (%), and α is taken as an empirical value of 0.5. W(TL) represents the trophic level weight, and W(H) represents the home range weight. The weights of each indicator are calculated based on the CRITIC weight calculation method.

[0119] The updated umbrella strength index is used to select umbrella species in the food web model, and the species with the highest umbrella strength index ranking is the most suitable species to be the umbrella species.

[0120] Furthermore, selecting rare umbrella species according to the food web model includes: integrating the species protection level on the basis of the updated umbrella strength index to obtain the rare umbrella species strength index, and using the rare umbrella species strength index to select rare umbrella species in the food web model.

[0121] Specifically, to accurately select priority protected species—rare "umbrella species"—this proposal integrates the concept of species protection level (including China's nationally protected wildlife, IUCN Red List species, and CITES appendix species) into the "umbrella strength" index, proposing for the first time a rare "umbrella strength" index to quantitatively identify rare "umbrella species." The proposal quantifies species protection level (including China's nationally protected wildlife, IUCN Red List species, and CITES appendix species) into a Protective Effect Index (PEI).

[0122] The protection effect index formula is as follows:

[0123] PEI i =S CHINAi ×S IUCNi ×S CITESi

[0124] Where S CHINAi 、S IUCNi 、S CITESi The weighted averages of the protection levels of China's key protected wild animals, the weighted averages of the protection levels of IUCN Red List animals, and the weighted averages of the protection levels of CITES Appendix animals for all species in the food web with species i as the top predator are listed. The formula for the rare "umbrella strength" index is as follows:

[0125]

[0126] Where W(PEI) is the weight of the protection effect index, and the weight of each indicator is calculated using the CRITIC weight calculation method.

[0127] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0128] like Figure 1 As shown, this embodiment takes the selection of priority protected species and species sets in the food web of the Himalayan forest ecosystem as an example. It can be understood that the ecosystem food web evaluation and optimization method of the present invention can also be applied to other ecosystems.

[0129] Step 1: Targeting the Himalayas, collect multi-source data necessary for constructing food webs based on experiments, books, and literature, including information on predator-prey relationships between species, species traits, species home ranges, and species conservation status. See Table 1 for the basic information collected.

[0130] Table 1 Information on species in the Himalayan forest ecosystem

[0131]

[0132]

[0133]

[0134] Step 2: Construct a predator-prey matrix and predict the predator-prey relationship:

[0135] In order to predict the predation relationship between species, a matrix × matrix method is used. To create the matrix, the present invention converts all feature variables into binary format. For the classification characteristics of species order, family, and diet, data (0 or 1) is directly assigned according to the presence / absence of each classification characteristic of the species. For species weight as a quantitative characteristic, according to the body size standard of Chinese terrestrial mammals, mammals are divided into three levels: small mammals, medium mammals and large mammals. On this basis, the present invention further divides mammals of each body size level into large, medium and small based on the upper and lower limits of the species body size among all species. Finally, Chinese terrestrial mammals are divided into 9 categories based on the weight characteristics, and then converted into binary variables, and the matrix is constructed accordingly. The specific process is as follows:

[0136] The first step is to establish the predator-prey matrix A, whose dimension is i×j, and the formula is as follows:

[0137] A=(a ij ) i×j

[0138] Where i is the total number of predators and j is the total number of prey. The trophic interaction a between predator i and prey j is ij Represented by 1 or 0 (1 if predator i preys on j, 0 otherwise), the matrix A represents the documented predation relationships between predators and prey.

[0139] The second step is to establish the predator-prey feature matrix P, whose dimension is i×k, and the formula is as follows:

[0140] P=(p ik ) i×k

[0141] Where k is the species trait. For each species characteristic, a predator-prey characteristic matrix P is established separately. The relationship between predator i and prey characteristics k is p ik The matrix P represents the predator's prey preference, expressed as the relative frequency of predator i preying on prey feature k.

[0142] The third step is to establish the species feature-species matrix T, whose dimension is k×j, and the formula is as follows:

[0143] T=(tkj ) k×j

[0144] For each species’ trait, a species trait-species matrix T is established separately, and the relationship between species trait k and species j is t kj Represented by 1 or 0 (1 if species j has species characteristic k, otherwise 0), the matrix T represents the characteristics of the species.

[0145] The fourth step is to establish the matrix A′ to visualize the predator's predation preference for species characteristics for each species. The formula is as follows:

[0146] M=P×T=(m' ij ) i×j

[0147]

[0148] Where θ represents the minimum threshold for the occurrence of a predator-prey relationship, which is derived from the comparison of matrix M and matrix A, and m' ij with a ij The minimum overlap is the value of θ. For each species characteristic, a matrix A′ can be obtained. The predator-prey relationships that occur under each species characteristic are screened out. On this basis, obviously unreasonable predator-prey relationships are removed and the food web is constructed accordingly.

[0149] Step 3: Accurately calculate the prey source ratio between species, and calculate the prey source ratio based on the probability of predation relationship in the above matrix. First, calculate the average probability of predation relationship under the prediction of species characteristics, the formula is as follows:

[0150]

[0151] Where, represents the average probability of predator i preying on species j, f ij The probability of a predator-prey relationship occurring under a single characteristic (i.e., order, family, trophic level, body weight) is taken as m' under the corresponding species characteristic. ij value), n is the number of selected species characteristics.

[0152] The relative frequencies of all prey-predator relationships of predator i are calculated to calculate the food source ratios among species in the Himalayan forest ecosystem, such as Figure 2 , the formula is as follows:

[0153]

[0154] Where, F ij represents the food source ratio of predator i to prey species j, is the average probability of the predation relationship between predator i and prey k.

[0155] Step 4: Accurately calculate the trophic level of species. Calculate the trophic level based on the food source ratio between species. The formula is as follows:

[0156] Tr i =1+∑Tr j ×F ij

[0157] Where Tr i represents the trophic level of predator i, Tr j is the trophic level of prey j, F ij is the food source ratio of predator i at predation node j. The present invention sets the primary producer trophic level to 1 and obtains the trophic level of each species in the food web of the Himalayan forest ecosystem, such as Figure 3 , where the overall trophic level of each consumer spans from 2.00 to 4.55, with tiger having the highest trophic level (4.55).

[0158] Step 5: Construct a food web model, combining the predator-prey relationship between species in the food web, the trophic level characteristics of each node and the food source ratio to construct a quantitative and weighted food web, such as Figure 4 .

[0159] Step 6: Select key species based on centrality indicators. The concept of shortest path is often involved in the process of using centrality indicators to determine key species. This example refers to the shortest path concept summarized by Dijkstra (1959) and Barrat et al. (2007), using the interspecies feeding ratio as the link weight to calculate the shortest path between species. The formula is as follows:

[0160] dw(i,j)=min(1 / w ih +……+1 / w hi )

[0161] Where dw(i,j) refers to the shortest distance from species i to species j, and w ih is the weight of species i. In this embodiment, the feeding ratio between species is regarded as the link weight.

[0162] Degree centrality is the most direct indicator of the centrality of each species in a network, reflecting the ability of a species to influence other species in the network. The higher the degree centrality of a species, the more important it is in the network. It is calculated using the following formula:

[0163]

[0164] Among them, D i is the degree centrality of species i, K irepresents the degree of species, and n represents the number of species in the food web.

[0165] Betweenness centrality is an indicator used to quantitatively analyze a species' ability to control information exchange within a network. The higher a species' betweenness centrality, the stronger its ability to control information exchange within the network and the more critical it is to maintaining network stability. It is calculated using the following formula:

[0166]

[0167] Among them, B i is the betweenness centrality of species i, σ vj are all the shortest paths between species v and species j, σ vj (i) is the number of times the shortest path between species pair v and j passes through species i.

[0168] Closeness centrality is an indicator of a species' dominance in transmitting information within a network. The higher a species' closeness centrality, the faster it can spread network information to other species. In other words, the closer it is to the center of the network, the more important it is within the network. It is calculated using the following formula:

[0169]

[0170] Among them, C i is the closeness centrality of species i, d ij is the shortest path length between species i and species j.

[0171] PageRank centrality was proposed by one of Google's founders as a metric for measuring the importance of web pages on the Internet. This algorithm is primarily based on the hyperlink relationships between web pages. In other words, the importance of a web page depends not only on how many other web pages link to it, but also on the importance of the web pages that link to it. In short, PageRank believes that the importance of a web page is determined by the number and quality of other web pages that point to it, and is calculated using the following formula:

[0172]

[0173] Among them, PR(i) is the PageRank centrality of species i, j is a species connected to species i, k j,out It refers to the out-degree of species j, λ is the damping coefficient, and the empirical value is 0.85.

[0174] According to the centrality index results, as shown in Table 2, leopard has the highest degree centrality (0.714) and closeness centrality (0.726), clouded leopard has the highest betweenness centrality (0.008), and tiger has the highest PageRank centrality (0.263).

[0175] Table 2 Results of centrality index of species in Himalayan forest ecosystems

[0176]

[0177]

[0178] By analyzing the effect of topological removal on the stability of the food web, the most appropriate centrality index is selected: sequence removal is performed using topological removal. Removal methods are mainly divided into random removal and deterministic removal. Random removal mainly simulates the random disappearance of species in nature, while deterministic removal mainly removes certain species in a targeted manner based on the importance of the species. Based on the calculation results of the above four centralities, the species in the food web are sorted, and five schemes are used to determine the priority of species to be deleted: (1) remove species in descending order of degree centrality; (2) remove species in descending order of betweenness centrality; (3) remove species in descending order of closeness centrality; (4) remove species in descending order of PageRank centrality; (5) remove species randomly. In the process of gradually removing species, the changes in network stability are observed. Each sequence removal step is repeated until no secondary extinction occurs, and then the next round of sequence removal is started. When a species loses all its resources, it is considered to be secondary extinct. Primary producers will not experience secondary extinction.

[0179] The following two indicators are used to indicate the stability of the food web during the topological sequence removal process:

[0180] 1) Relative size of the largest connected subgraph:

[0181] The relative size of the maximum connected subgraph refers to the ability of the network to maintain connections between the remaining species when it is damaged. It is used to indicate the degree of damage to the network after an attack and is calculated using the following formula:

[0182]

[0183] N' is the number of species in the largest connected subgraph after the network is attacked, and N is the total number of species in the initial network when the network is not attacked.

[0184] 2) Global network efficiency:

[0185] Since global efficiency can reflect the overall connectivity of the network and measure the efficiency of information transmission between different species in the network, this embodiment uses global efficiency to indicate the change in food web stability during species removal. The formula is as follows:

[0186]

[0187] Where E(G) represents the global network efficiency of network G, n is the total number of species, d ij is the shortest path between species.

[0188] The results of the topology-based sequence removal on the stability of the food web in the Himalayan forest ecosystem are as follows: Figure 5 (a)-(b) Sequence removal based on degree centrality and PageRank centrality leads to the fastest collapse of the food web. Therefore, from the perspective of food web stability, degree centrality and PageRank centrality results can more accurately represent the keystone species ranking results and are more valuable for reference. Based on these results, we believe that leopards and tigers are more suitable as keystone species in this food web.

[0189] Figure 5 Effects of topology-based sequence removal on the stability of food webs in Himalayan forest ecosystems, where DC represents degree centrality, CC represents closeness centrality, BC represents betweenness centrality, and PR represents PageRank centrality.

[0190] Step 7: Quantitatively select umbrella species. Based on the food web, species with different spatial requirements are connected through predator-prey relationships. The umbrella species are quantitatively determined using the umbrella strength index, which integrates four factors: the number (degree) of interspecies connections, the strength (strength) of connections, the trophic level (TL), and the home range area of the species. The formula is as follows:

[0191]

[0192] Where H i is the home range index of species i, X i is the home range area of species i, X min and X max Respectively represent the minimum and maximum home range areas of all species in the food web of all regions.

[0193] The updated "protection strength" index formula is as follows:

[0194]

[0195] Where, TL i is the trophic level of food web node i, Ki is the number of nodes adjacent to node i (degree), Si is the sum of the edge weights (strength) connected to node i (%), and α is taken as an empirical value of 0.5. W(TL) represents the trophic level weight, and W(H) represents the home range weight. The weights of each indicator were calculated based on the CRITIC weighting method.

[0196] According to the calculation results of umbrella strength index, the umbrella species in the food web are determined, such as Figure 6The results showed that the leopard had the highest umbrella strength index (14.32), followed by the tiger (12.95), making it suitable as a candidate "umbrella species" for the forest ecosystem in the Himalayas.

[0197] Step 8: Use the "Rare Umbrella Strength" index to quantitatively identify rare "umbrella species." The proposal quantifies the species' protection levels (including China's nationally protected wildlife, IUCN Red List species, and CITES Appendix species) into a Protective Effect Index (PEI).

[0198] The protection effect index formula is as follows:

[0199] PEI i =S CHINAi ×S IUCNi ×S CITESi

[0200] Where S CHINAi 、S IUCNi 、S CITESi are the weighted averages of the protection levels of China’s key protected wild animals, the weighted averages of the protection levels of IUCN Red List animals, and the weighted averages of the protection levels of CITES Appendix animals for all species in the food web with species i as the top predator (Table 3).

[0201] Table 3 Wildlife protection level weight classification

[0202]

[0203] The formula for the rare "protection strength" index is as follows:

[0204]

[0205] Where W(PEI) is the weight of the protection effect index, and the weight of each indicator is calculated using the CRITIC weight calculation method.

[0206] According to the calculation results of the umbrella strength index, rare "umbrella species" in the food web are determined, such as Figure 7 The results showed that the leopard had the highest rarity umbrella intensity index (14.60), followed by the tiger (12.54), making it suitable as a candidate rare "umbrella species" in the Himalayan forest ecosystem.

[0207] Based on the above results, we believe that the priority conservation species set for the Himalayan forest ecosystem is leopard and tiger.

[0208] The present invention has developed a standardized process for constructing directed, weighted food webs for multi-source data input. Due to the difficulty of field sampling, the number of species that can be included in the food web framework determined by applying stable isotope values in combination with the MixSIAR model is limited. The present invention has created a topological "big food web" construction technology based on multi-source data fusion analysis. Through multi-source data collection, the predation relationship matrix between species, the predation relationship characteristic matrix, the species typical trait matrix, etc. are constructed. By fitting the predator-prey related traits, the predation relationship between candidate predators and prey is judged, and the feeding ratio between species is quantitatively determined, thereby constructing a directed, weighted "big food web" (number of species>100).

[0209] The present invention focuses on the hot issues of umbrella species selection methods and potential assessment in the field of biodiversity conservation. The traditional methods for selecting umbrella species are mainly qualitative selection or relying on a large amount of survey data within the species habitat for evaluation. For example, Seddon and Leech suggested that conservation planners should follow 7 criteria when selecting umbrella species. Sattler et al. believe that species selected as umbrella species are usually so-called "flagship species", but the above methods have low accuracy, high cost and considerable limitations. Therefore, there is an urgent need for new methods for accurately selecting umbrella species to gather key species, exert the umbrella effect and enhance conservation benefits. The food web theory is a classic theory in ecology that reflects interspecific relationships and is a channel for material and energy flows. Starting from the perspective of the food web, based on the crucial predator-prey relationship between species, based on the core framework of the directed weighted food web, and combined with the network centrality theory, this invention updates the "Umbrella species strength" index that comprehensively considers the number of connections between species, connection strength, trophic level, and home range area to measure the species umbrella effect. At the same time, on the basis of the "umbrella strength" index, it innovatively incorporates the concept of species protection level (including China's national key protected wild animals, IUCN Red List animals, and CITES Appendix animals), and proposes the rare "umbrella strength" index for the first time to quantitatively determine the rare "umbrella species".

[0210] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A method for selecting priority species and species sets for conservation based on food web stability assessment, characterized in that: include: Obtaining species data, using the species data to determine predator-prey relationships, trophic level characteristics, and food source ratios among species, and constructing a food web model; Priority protected species are screened according to the food web model, and the priority protected species include: keystone species, umbrella species and rare umbrella species.

2. The method for selecting priority species and species sets for protection in combination with food web stability assessment according to claim 1, characterized in that: Determining the predator-prey relationship between the species involves: Establish the predator-prey matrix, the predator-prey feature matrix and the all species feature-species matrix respectively: A=(a ij ) i×j P=(p ik ) i×k T=(t kj ) k×j Where i is the predator, j is the prey, k is the species characteristics, A is the predator-prey matrix, (a ij ) i×j is the original predator-prey relationship between species, P is the predator-prey feature matrix, (p ik ) i×k is the predator's prey characteristic preference, T is the species characteristic-species matrix, (t kj ) k×j A characteristic of a species; Using the predator-prey feature matrix and the all-species feature-species matrix, a fourth matrix is obtained: M=P×T=(m' ij ) i×j Among them, P is the predator-prey feature matrix, T is the all species feature-species matrix, (m' ij ) i×j is the probability of predation-prey relationship between species; The fourth matrix is compared with the predator-prey matrix to determine the minimum threshold for the occurrence of a predator-prey predation relationship. Based on the minimum threshold, the predator's predation preference for species characteristics is mapped to each species to obtain a fifth matrix: Where θ represents the minimum threshold for the occurrence of predator-prey relationship, (a' ij ) i×j The ultimate predator-prey relationship between species.

3. The method for selecting priority species and species sets for protection in combination with food web stability assessment according to claim 1, characterized in that: Determining the feeding ratios between the species involved: Calculate the average probability of predation-prey relationships between the species under the prediction of species traits: in, represents the average probability of predator i occurring at prey node j, f ij is the probability of predation relationship under single feature prediction, that is, m' under the corresponding species feature ij value, n is the number of species characteristics selected; The relative frequency of the predator-prey relationship between the predator and all prey is determined by the average probability as the food source ratio: Among them, F ij represents the food source ratio of predator i to prey species j, is the average probability of the predation relationship between predator i and prey k.

4. The method for selecting priority species and species sets for protection in combination with food web stability assessment according to claim 1, characterized in that: Characteristics of trophic levels that determine differences between species include: Tr i =1+∑Tr j ×F ij Among them, Tr i represents the trophic level of predator i, Tr j is the trophic level of prey j, F ij is the food source ratio of predator i at predation node j.

5. The method for selecting priority species and species sets for protection in combination with food web stability assessment according to claim 1, characterized in that: Selecting the key species according to the food web model includes: The food source ratio between species is used as the link weight in the food web model, and the importance of each species in the food web model is ranked using centrality indicators; the centrality indicators include degree centrality, closeness centrality, betweenness centrality and PageRank centrality; Species in the food web model are removed in descending order of centrality index or randomly, and in the process of removing the species, the stability of the food web model is calculated, and the optimal centrality index is determined according to the stability, and the key species is determined based on the ranking result of the species according to the optimal centrality index.

6. The method for selecting priority species and species sets for protection in combination with food web stability assessment according to claim 5, characterized in that: Calculating the stability of the food web model involves: The relative size of the largest connected subgraph is used to determine the extent of damage to the food web model after being attacked: Where N' is the number of species in the largest connected subgraph after the network is attacked, and N is the total number of species in the initial network when the network is not attacked; The global network efficiency is used to measure the efficiency of information transmission between different species in the food web model: Where E(G) represents the global network efficiency of network G, n is the total number of species, d ij is the shortest path between species.

7. The method for selecting priority species and species sets for protection in combination with food web stability assessment according to claim 1, characterized in that: Umbrella species selected according to the food web model include: Get the species' home range area and use it to update the umbrella strength index: Among them, TL i is the trophic level of food web node i, K i is the number of nodes adjacent to node i, S i is the sum of the weights of the edges connected to node i, α is a fixed value of 0.5, W(TL) represents the trophic level weight, and W(H) represents the home range area weight; The updated umbrella strength index is used to select umbrella species in the food web model, and the species with the highest umbrella strength index ranking is used as the umbrella species.

8. The method for selecting priority species and species sets for protection in combination with food web stability assessment according to claim 7, characterized in that: Obtaining the home range area of the species includes: Where H i is the home range index of species i, X i is the home range area of species i, X min and X max Respectively represent the minimum and maximum home range areas of all species in the food web of all regions.

9. The method for selecting priority species and species sets for protection in combination with food web stability assessment according to claim 7, characterized in that: Rare umbrella species selected according to the food web model include: On the basis of the updated umbrella protection strength index, the species protection level is integrated to obtain the rare umbrella species strength index: RUSS(i)=K i (1-α) ×S i α ×(TL i ×W(TL)+PEI i ×W(PEI)+H i ×W(H)) Where W(PEI) is the weight of the protection effect index; The rare umbrella species strength index is used to select rare umbrella species in the food web model, and the species with the highest rare umbrella species strength index is used as the rare umbrella species.

Images