Intelligent fish community stability evaluation method and system based on network model

Abstract

The present invention relates to an intelligent fish community stability evaluation method and system based on a network model, the method comprising optimizing species distribution data based on the impact of a single species on community biodiversity and the impact of interspecific interactions on community structure, obtaining species distribution optimization data and constructing a community association network model; using a PCoA method to quantify the functional diversity overlap of species and construct a functional niche network model; constructing a comprehensive network model based on the community association network model and the functional niche network model; constructing a radar chart based on the community association network model, the functional niche network model and the comprehensive network model, and evaluating network stability based on the radar chart area method. The present invention can better reflect the lack of species interaction relationships caused by incomplete sampling, make the network structure more complex and closer to the actual situation, can deeply explore the interaction relationships between species, and improve the accuracy of evaluation results and the comparability between different biological community networks.

Description

Technical Field

[0001] The present invention relates to the field of fish community stability evaluation, and in particular to an intelligent fish community stability evaluation method and system based on a network model. Background Art

[0002] The ecological network of a biological community refers to a complex network of relationships formed by the interactions between biological populations and between them and the environment in an ecosystem. For example, the ecological network of a fish community includes the predation relationship, competition relationship, symbiotic relationship, etc. between fish, as well as their interactions with aquatic plants, microorganisms, inorganic environment and other factors. At present, there are various methods for constructing ecological networks, but there is a lack of unified evaluation criteria; in practical applications, high requirements are placed on data quality and integrity, and the sampling of species interactions in reality is often incomplete, which may affect the accuracy and comparability of network indicators, and the operability of the research and the universality of the results need to be improved; in addition, in aquatic ecosystems, there is a lack of comparability between ecological networks constructed across time and space, and ecological network analysis relies more on the study of the relationship between food chains and food webs. Summary of the invention

[0003] The present invention provides an intelligent fish community stability evaluation method and system based on a network model to solve at least one of the above technical problems.

[0004] The technical solution of the present invention to solve the above technical problems is as follows: an intelligent fish community stability evaluation method based on a network model, comprising:

[0005] S1, based on the impact of a single species on community biodiversity and the impact of interspecific interactions on community structure, the species distribution data of the fish community is optimized to obtain species distribution optimization data, and a community association network model is constructed based on the species distribution optimization data;

[0006] S2, using PCoA method to quantify the functional diversity overlap of species in fish communities, and constructing a functional niche network model based on the functional diversity overlap;

[0007] S3, constructing a comprehensive community distribution network model based on the community association network model and the functional niche network model;

[0008] S4, constructing a radar chart according to the community association network model, the energy niche network model and the community distribution comprehensive network model, and evaluating the network stability based on the radar chart area method.

[0009] Based on the above technical solution, the present invention can also be improved as follows.

[0010] Further, the S1 is specifically:

[0011] S11, based on the impact of single species on the community biodiversity, calculate the biodiversity influence index of each single species in the fish community;

[0012] S12, based on the impact of interspecific interactions on community structure, the Apriori association rule was used to mine the interspecific association relationships in fish communities, and the importance matrix of the entire species distribution in fish communities was obtained;

[0013] S13, obtaining species distribution data of fish communities, optimizing the species distribution data using the biodiversity influence index and the importance matrix, and obtaining species distribution optimization data;

[0014] S14, calculating the correlation coefficient between every two species in the fish community according to the species distribution optimization data to obtain a correlation coefficient matrix;

[0015] S15, in the correlation coefficient matrix, setting the correlation coefficient whose significance is greater than a preset significance threshold or the correlation coefficient whose significance is less than a preset correlation coefficient threshold to 0, to obtain a significant close correlation coefficient matrix;

[0016] S16, using the R language igraph package to draw a weighted directed graph of the significant close correlation coefficient matrix to obtain a community association network model.

[0017] Furthermore, in S11, the calculation formula for the biodiversity influence index of a single species in a fish community is:

[0018]

[0019] Among them, Index i represents the biodiversity impact index of the i-th species in the fish community, D 0 The biodiversity index of all species in the fish community, D i It represents the biodiversity index of all species remaining in the fish community after removing the i-th species.

[0020] Furthermore, in S12, the Apriori association rule is used to mine the interspecies correlation relationships in the fish community, specifically:

[0021] The Apriori association rule is used to find multiple frequent item sets with different numbers of items appearing in each species in the fish community;

[0022] For frequent item sets with the same number of items, the support of the combination is used to represent the importance of the combination, so as to obtain the support matrix composed of the frequent item sets with the same number of items;

[0023] The support matrices of all frequent itemsets are averaged to obtain the importance matrix of all species distribution in the fish community.

[0024] Further, in S13, the formula for optimizing the species distribution data using the biodiversity influence index and the importance matrix is:

[0025]

[0026] in, Represents the species distribution optimization data, Den m×n Indicates the species distribution data, Index n×1 The biodiversity influence index of each species in the fish community, Im m×n represents the importance matrix; m represents the number of sampling points, and n represents the number of species.

[0027] Further, the S2 is specifically:

[0028] S21, using the PCoA method to reduce the dimensionality of the functional characteristic data of multiple samples of each species in the fish community, and obtaining multiple sample data of each species in the fish community;

[0029] S22, extracting the principal component data of the first two axes from each sample data of each species in the fish community, corresponding to the coordinate value of each sample data in the two-dimensional plane coordinate system;

[0030] S23, regarding the coordinate values ​​of all sample data of the same species in the fish community in the two-dimensional plane coordinate system as a sample point set of the species, drawing a polygonal convex hull of the sample point set, and calculating the area of ​​the polygonal convex hull;

[0031] S24, calculating the overlapping area of ​​the polygonal convex hulls of every two species in the fish community, and calculating the ratio of the overlapping area of ​​the polygonal convex hulls of every two species in the fish community to the polygonal convex hull areas of the corresponding two species, thereby obtaining the overlapping degree of functional diversity of every two species in the fish community;

[0032] S25, using the R language igraph package to draw a weighted directed graph of the functional diversity overlap between every two species in the fish community, and obtain a functional niche network model.

[0033] Further, the S3 is specifically:

[0034] S31, optimizing the community association network model to obtain a community association network optimization model; wherein the formula for optimizing the community association network model is:

[0035]

[0036] Specifically, represents the community association network optimization model, G 1 represents the community association network model;

[0037] S32, optimizing the functional niche network model to obtain an optimized functional niche network model; wherein the formula for optimizing the functional niche network model is:

[0038]

[0039] Specifically, represents the functional niche network optimization model, G 2 represents the functional niche network model;

[0040] S33, constructing an initial model of the community distribution comprehensive network according to the community association network optimization model and the functional niche network optimization model; wherein the construction formula of the initial model of the community distribution comprehensive network is:

[0041]

[0042] Specifically, G represents the initial model of the community distribution integrated network;

[0043] S34, adjusting the initial model of the community distribution comprehensive network to obtain a community distribution comprehensive network model; wherein the formula for adjusting the initial model of the community distribution comprehensive network is,

[0044]

[0045] Specifically, Represents the comprehensive network model of community distribution.

[0046] Further, the S4 is specifically:

[0047] S41, based on the multi-dimensional ecological network stability evaluation index, obtaining a first multi-dimensional index value from the community association network optimization model, obtaining a second multi-dimensional index value from the functional niche network optimization model, and obtaining a third multi-dimensional index value from the community distribution comprehensive network model;

[0048] S42, combining the first multi-dimensional index value, the second multi-dimensional index value, and the third multi-dimensional index value together for standardization to obtain a first multi-dimensional standardized index value, a second multi-dimensional standardized index value, and a third multi-dimensional standardized index value;

[0049] S43, averaging the standardized index values ​​of corresponding dimensions in the first multi-dimensional standardized index value, the second multi-dimensional standardized index value, and the third multi-dimensional standardized index value to obtain an index average value of the multi-dimensional ecological network stability evaluation index;

[0050] S44, sorting the average values ​​of the multi-dimensional ecological network stability evaluation indicators in ascending order to obtain the arrangement order of the multi-dimensional ecological network stability evaluation indicators;

[0051] S45, respectively reordering the first multi-dimensional index value, the second multi-dimensional index value, and the third multi-dimensional index value according to the arrangement order of the multi-dimensional ecological network stability evaluation index;

[0052] S46, respectively draw radar charts according to the reordered first multidimensional index value, the reordered second multidimensional index value, and the reordered third multidimensional index value, to obtain a first radar chart, a second radar chart, and a third radar chart accordingly, and respectively calculate the area of ​​the first radar chart, the area of ​​the second radar chart, and the area of ​​the third radar chart, and evaluate the network stability according to the area of ​​the first radar chart, the area of ​​the second radar chart, and the area of ​​the third radar chart.

[0053] Furthermore, the multi-dimensional ecological network stability evaluation indicators include: network scale, network connectivity, average connectivity, average clustering coefficient and connection.

[0054] On the basis of the above-mentioned intelligent fish community stability evaluation method based on network model, the present invention also provides an intelligent fish community stability evaluation system based on network model.

[0055] An intelligent fish community stability evaluation system based on a network model comprises a processor, a memory and a computer program stored in the memory. When the computer program is executed by the processor, the intelligent fish community stability evaluation method based on a network model as described above is implemented.

[0056] The beneficial effects of the present invention are as follows: the intelligent fish community stability evaluation method and system based on the network model of the present invention, based on the consideration of the species community distribution association network, based on the influence of a single species on the community biodiversity and the influence of interspecific interactions on the community structure, uses the influence index and the Apriori association rule to intelligently optimize the fish community ecological network structure model, so that it can better restore the more complex association relationship between actual species on the basis of the sampled species, and more vividly reflect the community characteristics; build a community association network model based on the optimized association relationship; and use the PCoA method to quantify the overlap of species functional diversity, and build a functional niche network model; based on the community association network model and the functional niche network model, construct a comprehensive network model of community distribution; and use the radar map area to evaluate the network stability, quantify the difference characteristics of different spatiotemporal networks, realize the comparative analysis between the community's cross-spatial ecological network models, and improve the accuracy of the evaluation results and the comparability between different networks. In addition, the present invention has strong expansibility, is suitable for the stability evaluation of network models in more scenarios, and provides a basis for the evaluation and comparison of ecosystem stability in different environments or different periods of the same environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a flowchart of the intelligent fish community stability evaluation method based on the network model of the present invention;

[0058] Figure 2 is a schematic diagram of an exemplary radar chart;

[0059] Figure 3 Schematic diagram of community association network model constructed for species distribution data;

[0060] Figure 4 Schematic diagram of the community association network model constructed for species distribution optimization data;

[0061] Figure 5 This is a schematic diagram of the comprehensive network model of community distribution;

[0062] Figure 6 Schematic diagram of the radar charts of the community association network model, functional niche network model and community distribution comprehensive network model. DETAILED DESCRIPTION

[0063] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0064] like Figure 1 As shown in the figure, the intelligent fish community stability evaluation method based on the network model includes:

[0065] S1, based on the impact of a single species on community biodiversity and the impact of interspecific interactions on community structure, the species distribution data of the fish community is optimized to obtain species distribution optimization data, and a community association network model is constructed based on the species distribution optimization data;

[0066] S2, using PCoA method to quantify the functional diversity overlap of species in fish communities, and constructing a functional niche network model based on the functional diversity overlap;

[0067] S3, constructing a comprehensive community distribution network model based on the community association network model and the functional niche network model;

[0068] S4, constructing a radar chart according to the community association network model, the energy niche network model and the community distribution comprehensive network model, and evaluating the network stability based on the radar chart area method.

[0069] In the method of the present invention, the method of the present invention uses the Apriori association rule to deeply explore the potential relationship between species, and builds a community association network model, which can better reflect the lack of interaction between species caused by incomplete sampling in the actual ecosystem, making the network structure more complex and closer to the actual situation; at the same time, the community distribution comprehensive network model constructed by the present invention not only considers the distribution characteristics of species and quantity between species, but also incorporates the distribution characteristics of functional characteristics, which can deeply explore the interaction relationship between species in the community; in addition, the present invention uses a two-dimensional radar map area to comprehensively evaluate the network characteristics, effectively making up for the deficiencies that different evaluation methods may have differences, while quantifying the difference characteristics of different spatiotemporal networks, realizing the comparative analysis between community cross-spatial ecological network models, and improving the accuracy of the evaluation results and the comparability between different networks. The method of the present invention has strong scalability, is suitable for the stability evaluation of network models in more scenarios, and provides technical support for the evaluation and comparison of ecosystem stability in different environments or different periods of the same environment.

[0070] The following is a detailed introduction to each step:

[0071] The S1 specifically includes the following S11 to S16:

[0072] S11, based on the impact of single species on the community biodiversity, calculate the biodiversity influence index of each single species in the fish community.

[0073] Biodiversity is a key indicator of ecosystem health and function. Communities with high biodiversity have more complex species interaction networks, including predation, parasitism, symbiosis and competition. Therefore, understanding the impact of a single species on the biodiversity of the entire community can help us better understand the community and the ecosystem in which it is located. We use the Biodiversity Impact Index to assess the impact of a single species on biodiversity.

[0074] Biodiversity impact index: refers to the impact of a single species on the diversity of the entire community. The calculation method is: the diversity of all species in the community is used as a reference, and the diversity of the remaining community after removing a certain species changes with the original community diversity. Specifically, the calculation formula for the biodiversity impact index of a single species in a fish community is:

[0075]

[0076] Among them, Index i represents the biodiversity impact index of the i-th species in the fish community, D 0 The biodiversity index of all species in the fish community, D i It represents the biodiversity index of all species remaining in the fish community after removing the i-th species.

[0077] S12, based on the impact of interspecific interactions on community structure, the Apriori association rule was used to mine the interspecific correlation relationships in fish communities and obtain the importance matrix of the entire species distribution in fish communities.

[0078] In addition to the influence of a single species on community structure, the influence of interspecific interactions on the community is also multifaceted, including predation, competition, symbiosis, parasitism and other interactions. We use an importance matrix to evaluate the impact of interspecific interactions on community structure, which can effectively explore the associations between species.

[0079] Importance Matrix Im m×n :The more times a single species or several species appear together, the more important the species or the combination of species is.

[0080] In S12, the Apriori association rule is used to find the frequent item sets of each species, including single frequent item sets, 2 frequent item sets, 3 frequent item sets, etc., with reference to the frequency of occurrence of each species at the sampling point and the frequency of co-occurrence with other species at the sampling point; for frequent item sets with the same number of items, the support of each combination is used to represent the importance of the combination, and the support matrix composed of the frequent item sets of the number of items is obtained as the overall support of the frequent item sets of the number of items; the support matrices of all item sets are averaged to obtain the importance matrix of the entire species distribution. If 0 appears on the diagonal of the importance matrix, it is replaced by the support of a single species.

[0081] Specifically: Frequent item sets represent combinations of different species. For example, fish in a lake (including carp, crucian carp, yellow catfish, Culter albus and grass carp) are sampled and 8 sampling points are selected. A single frequent item set represents a set containing one type of species (frequent item sets consisting of one type of species are all single frequent item sets), such as {carp} is a single item set, {carp} is also a single item set, {yellow catfish} is also a single item set, etc.; a 2-item frequent item set represents a set containing two types of species, such as {carp, carp} is a 2-item frequent item set, {carp, yellow catfish} is also a 2-item frequent item set, etc.; a 3-item frequent item set represents a set containing three types of species, such as {carp, carp, Culter albus} is a 3-item frequent item set, etc. It can be seen from this that an n-item frequent item set represents a set containing n types of species.

[0082] Support is a reused concept in Apriori association rules. For example, for the two-item set {carp, crucian carp}, the probability (joint probability) that carp and crucian carp appear at the same point reflects the importance of a certain rule; assuming that there are a total of 8 points, 5 of which have carp and crucian carp respectively, then the overall support of the combination {carp, crucian carp} is 5 / 8.

[0083] Each element in the support matrix can represent the average support of the combination of the species corresponding to the row and the species corresponding to the column containing the position. For example, there are five kinds of fish in the lake, namely carp, crucian carp, yellow catfish, black carp and grass carp. For all three frequent item sets {carp, crucian carp, yellow catfish}, {carp, yellow catfish, black carp}, {crucian carp, carp, black carp}, the element in the first row and second column of the matrix can be represented as the average support of the combination of carp and crucian carp at the same time. Since carp and crucian carp appear in both combination 1 and combination 3, it is (3 / 8+1 / 8) / 2=2 / 8. Therefore, the support of all three frequent item sets is shown in Table 1 below:

[0084] Table 1: Support of all three frequent itemsets

[0085]

[0086]

[0087] Therefore, the support matrix composed of all three frequent item sets is:

[0088]

[0089] The construction principles of the support matrices of other single frequent itemsets, two frequent itemsets, and four frequent itemsets are similar to the construction principle of the support matrix composed of three frequent itemsets mentioned above. Then, the support matrices of all frequent itemsets are averaged to obtain the importance matrix of all species distribution in the fish community.

[0090] After obtaining the importance matrix, the elements on the diagonal may be 0. For example, the element corresponding to the behavior of grass carp and the column of grass carp is 0. In this case, the element needs to be replaced by the probability of grass carp appearing at 8 points. For example, if it appears 2 times, then it is 2 / 8.

[0091] S13, obtaining species distribution data of fish communities, optimizing the species distribution data using the biodiversity influence index and the importance matrix, and obtaining species distribution optimization data.

[0092] The original species distribution data of fish communities (i.e., species abundance data) are affected by many factors, such as the ecological niche of species, environmental conditions, human activities, population dynamics, and data standardization and analysis methods. Here, the community abundance data is optimized by considering the interaction between single species and species. The biodiversity influence index (Index) of the impact of a single species on the overall community diversity is used to calculate the biodiversity impact of a single species on the overall community diversity. n×1 ), showing the importance matrix of the inter-species connections (Im m×n ) for the original species distribution data ((Den m×n ) to adjust and obtain species distribution optimization data The formula for optimizing the species distribution data using the biodiversity impact index and the importance matrix is:

[0093]

[0094] in, Represents the species distribution optimization data, Den m×n Indicates the species distribution data, Index n×1 The biodiversity influence index of each species in the fish community, Im m×n represents the importance matrix; m represents the number of sampling points, and n represents the number of species.

[0095] S14, calculating the correlation coefficient between every two species in the fish community according to the species distribution optimization data to obtain a correlation coefficient matrix.

[0096] The calculation formula of the correlation coefficient is:

[0097]

[0098] X i and Y i They represent the optimized species distribution data of the two species in the fish community at the i-th sampling point, and They represent the average values ​​of the optimized data of species distribution of two species in fish communities at all sampling points. The correlation coefficient matrix is ​​a symmetric matrix.

[0099] S15, in the correlation coefficient matrix, the correlation coefficient whose significance is greater than the preset significance threshold (0.05) or the correlation coefficient whose significance is less than the preset correlation coefficient threshold (0.3) is set to 0, to obtain a significant close correlation coefficient matrix.

[0100] S16, using the R language igraph package to draw a weighted directed graph of the significant close correlation coefficient matrix to obtain a community association network model.

[0101] For example, the correlation coefficient between crucian carp and yellow catfish is 0.8, and through the significance test, the significance of the correlation coefficient between crucian carp and yellow catfish is p<0.05, then there will be a directed line segment from crucian carp to yellow catfish, and the weight on the line segment is marked as 0.8. In this way, a weighted directed graph is drawn for the significant close correlation coefficient matrix.

[0102] The S2 is specifically:

[0103] S21, using the PCoA method to reduce the dimensionality of the functional characteristic data of multiple samples of each species in the fish community, and obtaining multiple sample data of each species in the fish community;

[0104] S22, extracting the principal component data of the first two axes from each sample data of each species in the fish community, corresponding to the coordinate value of each sample data in the two-dimensional plane coordinate system;

[0105] S23, regarding the coordinate values ​​of all sample data of the same species in the fish community in the two-dimensional plane coordinate system as a sample point set of the species, drawing a polygonal convex hull of the sample point set, and calculating the area of ​​the polygonal convex hull;

[0106] S24, calculating the overlapping area of ​​the polygonal convex hulls of every two species in the fish community, and calculating the ratio of the overlapping area of ​​the polygonal convex hulls of every two species in the fish community to the polygonal convex hull areas of the corresponding two species, thereby obtaining the overlapping degree of functional diversity of every two species in the fish community;

[0107] S25, using the R language igraph package to draw a weighted directed graph of the functional diversity overlap between every two species in the fish community, and obtain a functional niche network model.

[0108] The increase of functional diversity in aquatic ecosystems can improve the stability and productivity of ecosystems because it allows species to have more adaptive strategies and niche choices when facing environmental changes. In aquatic ecosystems, functional diversity may affect community stability through different mechanisms. Functional diversity overlap refers to the similarity of functional traits among different species in a community, that is, niche overlap among species, that is, multiple species perform the same or similar ecological functions, which can serve as an insurance mechanism to increase the resistance of ecosystems to disturbances, thereby maintaining the stability of ecosystems.

[0109] This method uses the principal coordinate analysis (PCoA) method to reduce the high-dimensional fish community functional characteristic data (the rows of this data represent species, and the columns represent the attributes of the species, including attributes of multiple dimensions such as total length, body length, relative size of the mouth cleft, and ratio of pectoral fin to tail fin), so that the community data can visualize the separation of samples in two-dimensional or three-dimensional space. The present invention extracts the data of the first two axes as the coordinate values ​​of each sample in the two-dimensional plane coordinate system XOY; the sample data of the same species is regarded as a set of sample points, and the polygonal convex hull composed of the sample point set in the two-dimensional plane and its corresponding convex hull area are calculated; if the polygonal convex hulls of two species overlap, the proportion of the overlapping area to the corresponding polygonal convex hull area of ​​the two species is calculated to obtain the functional diversity overlap matrix. For example, the convex hull area formed by the point set composed of the sample points of species a is τ a , the convex hull area formed by the point set composed of the sample points of species b is τ b , the overlapping area of ​​the two convex hulls is τ ab , then the overlap degree S (including S a and S b )for:

[0110]

[0111] Among them, S a S is the percentage of the overlapping area of ​​the convex hulls of species a and b to the convex hull formed by the set of sample points of species a. b It is the percentage of the overlapping area of ​​the convex hulls of species a and b to the convex hull formed by the set of points consisting of the sample points of species b.

[0112] It should be noted that if the polygonal convex hulls of two species do not overlap, the overlapping area of ​​the convex hulls τ ab is 0, then the overlap S is 0 (S a and S b are all 0).

[0113] In the process of drawing a weighted directed graph of the functional diversity overlap between each two species in the fish community using the R language igraph package, for example, species a and species b have overlap, S a =0.4, S b =0.5, then there will be two line segments between species a and b. The weight of the line segment from species a to species b is 0.5, and the weight of the line segment from species b to species a is 0.4. The network relationship between species a and b is determined, and the functional niche network model can be obtained.

[0114] The species in the community and the interaction between them constitute a complex network structure, which is significantly affected by the diversity of species distribution and the diversity of functional characteristics. The two are quantitatively modeled using the community association network model and the functional niche network model. Since the higher the overlap of species functional niches, the stronger the correlation between the two species; however, species with no overlapping functional niches cannot indicate that there is no correlation between the two species. This correlation is also affected by resources, niches, environment, etc. In addition, unrelated species may still remain unique or reduce competition between them due to functional overlaps such as niche differentiation, resource utilization differences, and ecological service differences; related species may form different niches according to the principle of competitive exclusion, so the overlap of functional niches may have a downward trend to some extent. Since the community structure is not only affected by its own species, but also by the interactions between species and between species and the environment, the present invention only considers the interactions between communities and weakens the impact of the environment. Therefore, a comprehensive network model of community distribution can be obtained, which is doubly affected by the community association network model and the functional niche network model. S3 of the present invention is the construction of a comprehensive network model of community distribution, and S3 is specifically:

[0115] S31, optimizing the community association network model to obtain a community association network optimization model; wherein the formula for optimizing the community association network model is:

[0116]

[0117] Specifically, represents the community association network optimization model, G 1 represents the community association network model;

[0118] S32, optimizing the functional niche network model to obtain an optimized functional niche network model; wherein the formula for optimizing the functional niche network model is:

[0119]

[0120] Specifically, represents the functional niche network optimization model, G 2 represents the functional niche network model;

[0121] S33, constructing an initial model of the community distribution comprehensive network according to the community association network optimization model and the functional niche network optimization model; wherein the construction formula of the initial model of the community distribution comprehensive network is:

[0122]

[0123] Specifically, G represents the initial model of the community distribution integrated network;

[0124] S34, adjusting the initial model of the community distribution comprehensive network to obtain a community distribution comprehensive network model; wherein the formula for adjusting the initial model of the community distribution comprehensive network is,

[0125]

[0126] Specifically, Represents the comprehensive network model of community distribution.

[0127] The S4 is specifically:

[0128] S41, based on the multi-dimensional ecological network stability evaluation index, obtaining a first multi-dimensional index value from the community association network optimization model, obtaining a second multi-dimensional index value from the functional niche network optimization model, and obtaining a third multi-dimensional index value from the community distribution comprehensive network model;

[0129] S42, combining the first multi-dimensional index value, the second multi-dimensional index value, and the third multi-dimensional index value together for standardization to obtain a first multi-dimensional standardized index value, a second multi-dimensional standardized index value, and a third multi-dimensional standardized index value;

[0130] S43, averaging the standardized index values ​​of corresponding dimensions in the first multi-dimensional standardized index value, the second multi-dimensional standardized index value, and the third multi-dimensional standardized index value to obtain an index average value of the multi-dimensional ecological network stability evaluation index;

[0131] S44, sorting the average values ​​of the multi-dimensional ecological network stability evaluation indicators in ascending order to obtain the arrangement order of the multi-dimensional ecological network stability evaluation indicators;

[0132] S45, respectively reordering the first multi-dimensional index value, the second multi-dimensional index value, and the third multi-dimensional index value according to the arrangement order of the multi-dimensional ecological network stability evaluation index;

[0133] S46, respectively draw radar charts according to the reordered first multidimensional index value, the reordered second multidimensional index value, and the reordered third multidimensional index value, to obtain a first radar chart, a second radar chart, and a third radar chart accordingly, and respectively calculate the area of ​​the first radar chart, the area of ​​the second radar chart, and the area of ​​the third radar chart, and evaluate the network stability according to the area of ​​the first radar chart, the area of ​​the second radar chart, and the area of ​​the third radar chart.

[0134] Ecological network stability evaluation indicators usually include multiple aspects. The present invention selects network scale (total number of nodes, N), network connectivity (total number of connections, L), average connectivity (average number of connections per node, average K), average clustering coefficient (degree of node clustering, average S), connection (proportion of links realized in all possible links, Con) and other indicators for evaluation, and these indicators are all positive indicators. Among them:

[0135] Network size (total number of nodes, N) refers to the number of all nodes in a network. In different network types, nodes can represent different entities, here representing different species; in ecological networks, the total number of nodes may affect the connectivity, diversity, and stability of the network. For example, the greater the number of nodes, the more possible ecological interactions and pathways, which may increase the robustness and resistance to disturbances of the network.

[0136] Network connectivity (total number of connections, L) refers to the total number of connections between all nodes in the network. The total number of connections in a network is one of the key indicators for evaluating the structural characteristics of a network, which affects the connectivity, efficiency, and robustness of the network. For example, a network with a high number of connections may have strong redundancy, and the network can still maintain its function even if some connections fail.

[0137] Average connectivity (average number of connections per node, average K) refers to the average number of connections each node in the network has. In an ecological network, the average number of connections may affect the network's resistance to failures or attacks. A network with low connectivity may be more susceptible to local failures, while a network with high connectivity may have better redundancy and robustness.

[0138] The average clustering coefficient (the degree to which nodes are clustered, average S) refers to the ratio of the number of edges that actually exist between the neighbor nodes of a node to the number of edges that may exist between them. If all neighbors of a node are connected to each other, the clustering coefficient of the node is 1, indicating a high degree of clustering.

[0139] Connectivity (Con) refers to the proportion of links that are realized among all possible links. It is an indicator to measure the density of network connections, which can reflect the degree of interconnection between nodes in the network. The connectivity index can reflect the integrity of the network. If Con is close to 1, it means that the connection in the network is very dense and almost all possible connections have been realized.

[0140] The ecological network stability evaluation index includes multi-dimensional evaluation indexes, which is not conducive to the overall evaluation of the network and the comparative analysis of different networks. Therefore, the present invention uses the area of ​​the two-dimensional radar chart constructed by each evaluation index value to comprehensively evaluate the network characteristics. This method can integrate the multi-dimensional evaluation indexes into a comprehensive index to evaluate the stability of the network based on the ecological network stability evaluation index data through the radar chart area method, thereby quickly and comprehensively evaluating the network characteristics to assist management decisions.

[0141] In addition, the radar chart drawing rules are as follows: Figure 2 As shown in the figure, in a two-dimensional plane, for a single network diagram (community association network optimization model or functional niche network optimization model or community distribution comprehensive network initial model), since the network stability evaluation here includes the above five indicators (the five indicators after sorting are arranged in counterclockwise order, A corresponds to the indicator with the smallest average value, and E corresponds to the indicator with the largest average value), it is necessary to draw a regular pentagon; the center of the regular pentagon is taken as the origin of the coordinates, the line connecting the origin of the coordinates and one of the vertices of the pentagon (A) is taken as the X-axis, and the axis perpendicular to the line is taken as the Y-axis. A plane rectangular coordinate system is drawn, and the value of the first indicator is taken as the horizontal coordinate of the indicator. Therefore, the horizontal axis of the point is marked as the actual value corresponding to the indicator, and the vertical coordinate is 0; then, according to the geometric relationship, the coordinate values ​​of other indicators are calculated in turn to obtain the corresponding coordinate points of each indicator value as shown in the figure. Figure 2 As shown, O is the coordinate origin of the regular pentagon, and rays OA, OB, OC, OD, and OE are the reference lines for the values ​​of the five indicators. 1 、x 2 、x 3 、x 4 and x 5 They represent the actual values ​​of the five indicators respectively; according to the coordinate points of each point, the pentagon is split into 5 triangles; Heron's formula is used to calculate the areas of the 5 triangles respectively; and then the sum of the areas of the 5 triangles is obtained, which is the total area of ​​the radar chart.

[0142] The following is an explanation with specific examples.

[0143] In this example, the biological community is the fish community, and the species in the fish community include crucian carp, Wheatear fish, club fish, Ziling goby, Boyi goby, spring carp, Xingkai silver carp, Western Taishen and small yellow croaker.

[0144] (1) By sampling 30 sampling points, the species distribution data of fish communities Den m×n As shown in Table 2 below:

[0145] Table 2: Species distribution data of fish communities

[0146]

[0147] (2) Biodiversity impact index of fish communities n×1 As shown in Table 3 below:

[0148] Table 3: Biodiversity Impact Index

[0149]

[0150]

[0151] (3) Importance matrix of fish community Im m×n As shown in Table 4 below:

[0152] Table 4: Importance Matrix

[0153]

[0154] (4) Optimization of species distribution data of fish communities As shown in Table 5 below:

[0155] Table 5: Species distribution optimization data

[0156]

[0157]

[0158] (5) The overlap of functional diversity of fish communities is shown in Table 6:

[0159] Table 6: Functional diversity overlap

[0160]

[0161] (6) Build community association network models, functional niche network models and community distribution comprehensive network models. Figure 3The community association network model constructed for the original species distribution data, Figure 4 Community association network models constructed for species distribution optimization data; Figure 3 In the community association network model constructed by the original species distribution data Den shown in the figure, due to the limitation of sampling data, the correlation between Ziling goby and western smelt and other fish is not significant or the correlation is very weak; after data optimization, in Figure 4 Data shown are optimized by species distribution In the constructed community association network model, these two fish have established internal connections with other fish. Figure 4 On the basis of this, when the functional diversity factor is further considered, the community distribution comprehensive network model constructed is as follows Figure 5 As shown in Figure 2, the community distribution integrated network model becomes more complex, and the interaction relationship between the two species is no longer equal, but there are certain differences. For example, the interaction relationship between the Pacific smelt and the The effect of is very small and can be almost ignored. The impact on Pacific smelt was negative, with an impact coefficient of 0.61. The community distribution integrated network model better reflects the lack of species interaction caused by incomplete sampling in the actual ecosystem, making the network structure more complex and closer to the real situation.

[0162] (7) The area represented by the radar chart constructed by the network stability evaluation index is a reference index for comparing different network models at the same level, which effectively improves the accuracy of the evaluation results and the comparability between different networks. When this evaluation index is used to evaluate and compare the stability of ecosystems in different water environments or in different periods of the same water environment, it provides a technical basis for comparative analysis between community cross-temporal and spatial ecological network models. The evaluation index values ​​of each network model are shown in Table 7 below:

[0163] Table 7: Evaluation index values ​​of network models

[0164]

[0165] (8) The radar charts of the community association network model, functional niche network model and community distribution comprehensive network model constructed after standardizing the evaluation index values ​​of the network models in Table 7 are shown in Figure 2. Figure 6 As shown, from Figure 6It can be seen that the network connectivity, average connectivity, average clustering coefficient and connection index of the community association network model, functional niche network model and community distribution comprehensive network model all show an upward trend; the radar map areas of the community association network model, functional niche network model and community distribution comprehensive network model are 2.82, 3.25 and 3.74 respectively, showing a gradual upward trend, indicating that the network performance of the community association network model, functional niche network model and community distribution comprehensive network model is gradually improving, and the correlation between the nodes of each species is gradually strengthening. Therefore, the community distribution comprehensive network model can better reflect the correlation between the various species in the fish community.

[0166] On the basis of the above-mentioned intelligent fish community stability evaluation method based on network model, the present invention also provides an intelligent fish community stability evaluation system based on network model.

[0167] An intelligent fish community stability evaluation system based on a network model comprises a processor, a memory and a computer program stored in the memory. When the computer program is executed by the processor, the intelligent fish community stability evaluation method based on a network model as described above is implemented.

[0168] The intelligent fish community stability evaluation method and system based on the network model of the present invention considers the species community distribution association network, based on the influence of a single species on the community biodiversity and the influence of interspecific interactions on the community structure, and uses the influence index and Apriori association rule to intelligently optimize the fish community ecological network structure model, so that it can better restore the more complex association relationship between actual species on the basis of the sampled species, and more vividly reflect the community characteristics; build a community association network model based on the optimized association relationship; and use the PCoA method to quantify the overlap of species functional diversity, and build a functional niche network model; according to the community association network model and the functional niche network model, construct a comprehensive network model of community distribution; and use the radar map area to evaluate the network stability, quantify the difference characteristics of different spatiotemporal networks, realize the comparative analysis between the community cross-spatial ecological network models, and improve the accuracy of the evaluation results and the comparability between different networks. In addition, the present invention has strong expansibility, is suitable for the stability evaluation of network models in more scenarios, and provides a basis for the evaluation and comparison of ecosystem stability in different environments or different periods of the same environment.

[0169] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent fish community stability evaluation method based on a network model, characterized in that: include: S1, based on the impact of a single species on community biodiversity and the impact of interspecific interactions on community structure, the species distribution data of the fish community is optimized to obtain species distribution optimization data, and a community association network model is constructed based on the species distribution optimization data; S2, using PCoA method to quantify the functional diversity overlap of species in fish communities, and constructing a functional niche network model based on the functional diversity overlap; S3, constructing a comprehensive community distribution network model based on the community association network model and the functional niche network model; S4, constructing a radar chart according to the community association network model, the energy niche network model and the community distribution comprehensive network model, and evaluating the network stability based on the radar chart area method; The S1 is specifically: S11, based on the impact of single species on the community biodiversity, calculate the biodiversity influence index of each single species in the fish community; S12, based on the impact of interspecific interactions on community structure, the Apriori association rule was used to mine the interspecific association relationships in fish communities, and the importance matrix of the entire species distribution in fish communities was obtained; S13, obtaining species distribution data of fish communities, optimizing the species distribution data using the biodiversity influence index and the importance matrix, and obtaining species distribution optimization data; S14, calculating the correlation coefficient between every two species in the fish community according to the species distribution optimization data to obtain a correlation coefficient matrix; S15, in the correlation coefficient matrix, setting the correlation coefficient whose significance is greater than a preset significance threshold or the correlation coefficient whose significance is less than a preset correlation coefficient threshold to 0, to obtain a significant close correlation coefficient matrix; S16, using the R language igraph package to draw a weighted directed graph of the significant close correlation coefficient matrix to obtain a community association network model; The S2 is specifically: S21, using the PCoA method to reduce the dimensionality of the functional characteristic data of multiple samples of each species in the fish community, and obtaining multiple sample data of each species in the fish community; S22, extracting the principal component data of the first two axes from each sample data of each species in the fish community, corresponding to the coordinate value of each sample data in the two-dimensional plane coordinate system; S23, regarding the coordinate values ​​of all sample data of the same species in the fish community in the two-dimensional plane coordinate system as a sample point set of the species, drawing a polygonal convex hull of the sample point set, and calculating the area of ​​the polygonal convex hull; S24, calculating the overlapping area of ​​the polygonal convex hulls of every two species in the fish community, and calculating the ratio of the overlapping area of ​​the polygonal convex hulls of every two species in the fish community to the polygonal convex hull areas of the corresponding two species, thereby obtaining the overlapping degree of functional diversity of every two species in the fish community; S25, using the R language igraph package to draw a weighted directed graph of the functional diversity overlap between every two species in the fish community, and obtain a functional niche network model.

2. The intelligent fish community stability evaluation method based on network model according to claim 1 is characterized in that: In S11, the calculation formula for the biodiversity impact index of a single species in a fish community is: Among them, Index i represents the biodiversity impact index of the i-th species in the fish community, D0 represents the biodiversity index of all species in the fish community, and D i It represents the biodiversity index of all species remaining in the fish community after removing the i-th species.

3. The intelligent fish community stability evaluation method based on network model according to claim 1 is characterized in that: In S12, the Apriori association rule is used to mine the interspecies correlation relationship in the fish community, specifically: The Apriori association rule is used to find multiple frequent item sets with different numbers of items appearing in each species in the fish community; For frequent item sets with the same number of items, the support of the combination is used to represent the importance of the combination, so as to obtain the support matrix composed of the frequent item sets with the same number of items; The support matrices of all frequent itemsets are averaged to obtain the importance matrix of all species distribution in the fish community.

4. The intelligent fish community stability evaluation method based on network model according to claim 1 is characterized in that: In S13, the formula for optimizing the species distribution data using the biodiversity influence index and the importance matrix is: in, Represents the species distribution optimization data, Den m×n Indicates the species distribution data, Index n×1 The biodiversity influence index of each species in the fish community, Im m×n represents the importance matrix; m represents the number of sampling points, and n represents the number of species.

5. The intelligent fish community stability evaluation method based on network model according to claim 1 is characterized in that: The S3 is specifically: S31, optimizing the community association network model to obtain a community association network optimization model; wherein the formula for optimizing the community association network model is: Specifically, represents the community association network optimization model, G1 represents the community association network model; S32, optimizing the functional niche network model to obtain an optimized functional niche network model; wherein the formula for optimizing the functional niche network model is: Specifically, represents the functional niche network optimization model, G2 represents the functional niche network model; S33, constructing an initial model of the community distribution comprehensive network according to the community association network optimization model and the functional niche network optimization model; wherein the construction formula of the initial model of the community distribution comprehensive network is: Specifically, G represents the initial model of the community distribution integrated network; S34, adjusting the initial model of the community distribution comprehensive network to obtain a community distribution comprehensive network model; wherein the formula for adjusting the initial model of the community distribution comprehensive network is, Specifically, Represents the comprehensive network model of community distribution.

6. The intelligent fish community stability evaluation method based on network model according to claim 5 is characterized in that: The S4 is specifically: S41, based on the multi-dimensional ecological network stability evaluation index, obtaining a first multi-dimensional index value from the community association network optimization model, obtaining a second multi-dimensional index value from the functional niche network optimization model, and obtaining a third multi-dimensional index value from the community distribution comprehensive network model; S42, combining the first multi-dimensional index value, the second multi-dimensional index value, and the third multi-dimensional index value together for standardization to obtain a first multi-dimensional standardized index value, a second multi-dimensional standardized index value, and a third multi-dimensional standardized index value; S43, averaging the standardized index values ​​of corresponding dimensions in the first multi-dimensional standardized index value, the second multi-dimensional standardized index value, and the third multi-dimensional standardized index value to obtain an index average value of the multi-dimensional ecological network stability evaluation index; S44, sorting the average values ​​of the multi-dimensional ecological network stability evaluation indicators in ascending order to obtain the arrangement order of the multi-dimensional ecological network stability evaluation indicators; S45, respectively reordering the first multi-dimensional index value, the second multi-dimensional index value, and the third multi-dimensional index value according to the arrangement order of the multi-dimensional ecological network stability evaluation index; S46, respectively draw radar charts according to the reordered first multidimensional index value, the reordered second multidimensional index value, and the reordered third multidimensional index value, to obtain a first radar chart, a second radar chart, and a third radar chart accordingly, and respectively calculate the area of ​​the first radar chart, the area of ​​the second radar chart, and the area of ​​the third radar chart, and evaluate the network stability according to the area of ​​the first radar chart, the area of ​​the second radar chart, and the area of ​​the third radar chart.

7. The intelligent fish community stability evaluation method based on network model according to claim 6 is characterized in that: The multi-dimensional ecological network stability evaluation indicators include: network scale, network connectivity, average connectivity, average clustering coefficient and connection.

8. An intelligent fish community stability evaluation system based on a network model, characterized in that: It comprises a processor, a memory and a computer program stored in the memory. When the computer program is executed by the processor, the intelligent fish community stability evaluation method based on the network model as described in any one of claims 1 to 7 is implemented.

Images