Intelligent identification method, system and device for influence factors of fish community

Abstract

The invention relates to an intelligent identification method, system and device for a fish community impact factor. The method comprises the following steps: constructing a total sequence by using a fish community and an environment sequence; calculating the Euclidean distance between every two point locations in each sequence to obtain first, second and third Euclidean distance sequences; building a comprehensive factor model according to each Euclidean distance sequence; building a species single-factor model according to the second Euclidean distance sequence and the Euclidean distance sequence of each species between every two point locations; building an environment single-factor model according to the third Euclidean distance sequence and the Euclidean distance sequence of each environment index between every two point locations; substituting the species and environment single-factor model into the comprehensive factor model to obtain a single-factor comprehensive model; and intelligently identifying the fish community influence factors according to the density distribution characteristics of the model coefficients. According to the method, the interaction between the water environment and the fish community is comprehensively considered, and the guiding species causing the fish community structure difference and the environment factors with the difference can be intelligently identified.

Description

Technical Field

[0001] The present invention relates to the field of biocommunity influencing factor identification, and in particular to an intelligent identification method, system and device for fish community influencing factors. Background Art

[0002] The ecological environment inside a lake includes environmental factors, biological factors, and the interaction between the two. The influence between environmental factors and biological factors is mutual. The health of the water ecological environment directly affects the diversity and stability of the aquatic animal community, and the structural state of the aquatic animal community can also reflect the health of the water ecological environment to a large extent. For aquatic ecosystems, the stability of the ecosystem includes both the stability of the water ecological environment, i.e., the physical and chemical environment, and the stability of the aquatic biological community, i.e., the biological environment. Current research on the stability of aquatic ecosystems mainly considers the impact of aquatic biological communities, which may ignore or weaken the impact of water environmental factors and the interaction between the water environment and aquatic organisms. This deviation may lead to an incomplete understanding of the stability and resilience of the ecosystem. As an important part of the aquatic ecological environment, the changes in fish species diversity and community structure directly reflect the health of the aquatic environment. Timely and accurate attention or identification of important species that affect the structure of fish communities and important indicators that cause environmental differences, by observing the distribution and changes of these species and indicators, scientists and environmental managers can timely identify signs of pollution, habitat destruction or other environmental pressures. Based on this, it is necessary to propose an intelligent identification scheme for factors affecting fish communities. Summary of the invention

[0003] The present invention provides a method, system and device for intelligently identifying factors affecting fish communities to solve at least one of the above technical problems.

[0004] The technical solution of the present invention to solve the above technical problem is as follows: an intelligent identification method for fish community influencing factors, comprising:

[0005] S1, collecting fish community data and environmental data at multiple points, obtaining corresponding fish community sequences and environmental sequences, and merging the fish community sequences and environmental sequences to construct an overall sequence;

[0006] S2, respectively calculating the Euclidean distances between any two points in the overall sequence, the fish community sequence and the environmental sequence, and correspondingly obtaining a first Euclidean distance sequence, a second Euclidean distance sequence and a third Euclidean distance sequence;

[0007] S3, building a comprehensive factor model according to the first Euclidean distance sequence, the second Euclidean distance sequence and the third Euclidean distance sequence;

[0008] S4, calculating the Euclidean distance sequence between two points of each species in the fish community sequence, and building a species single factor model according to the second Euclidean distance sequence and the Euclidean distance sequence between two points of each species in the fish community sequence;

[0009] S5, calculating the Euclidean distance sequence between two points of each environmental indicator in the environmental sequence, and building an environmental single factor model according to the third Euclidean distance sequence and the Euclidean distance sequence between two points of each environmental indicator in the environmental sequence;

[0010] S6, substituting the species single factor model and the environment single factor model into the comprehensive factor model to obtain a single factor comprehensive model;

[0011] S7, intelligently identifying the fish community influencing factors according to the density distribution characteristics of the model coefficients in the single factor comprehensive model.

[0012] Based on the above technical solution, the present invention can also be improved as follows.

[0013] Further, in S1, the fish community sequence is represented as:

[0014]

[0015] The environment sequence is expressed as:

[0016]

[0017] The overall sequence is expressed as:

[0018]

[0019] Wherein, F represents the fish community sequence, n represents the total number of species in the fish community, m represents the total number of points, j represents the ordinal number of the species, i represents the ordinal number of the point, and f ji represents the measured value of the jth species at the ith point; E represents the environmental sequence, l represents the total number of environmental indicators, k represents the ordinal number of environmental indicators, e ki represents the measured value of the kth environmental indicator at the ith point; T represents the overall sequence.

[0020] Further, in S3, the comprehensive factor model is expressed as:

[0021] T r ~α×F r +β×E r +ε;

[0022] Among them, T r represents the first Euclidean distance sequence, Fr represents the second Euclidean distance sequence, E r represents the third Euclidean distance sequence, α and β represent the fitting coefficients of least squares fitting, and ε represents the fitting constant of least squares fitting.

[0023] Further, in S4, the species single factor model is expressed as:

[0024]

[0025] Among them, F r represents the second Euclidean distance sequence, n represents the total number of species in the fish community, j represents the ordinal number of the species, Q j represents the Euclidean distance sequence between two points of the jth species in the fish community sequence, a j represents the fitting coefficient using least squares fitting, ε f Represents the fitting constant using least squares fitting.

[0026] Further, in S5, the environmental single factor model is expressed as:

[0027]

[0028] Among them, E r represents the third Euclidean distance sequence, l represents the total number of environmental indicators, k represents the ordinal number of environmental indicators, EQ k represents the Euclidean distance sequence between two points of the kth environmental indicator in the environmental sequence, b k represents the fitting coefficient using least squares fitting, ε e Represents the fitting constant using least squares fitting.

[0029] Further, in S6, the single factor comprehensive model is expressed as:

[0030]

[0031] Among them, ω 1j =α×a j ,ω 1j is the model coefficient in the single factor comprehensive model, indicating the species influence coefficient of the jth species on the ecological environment difference; ω 2k =β×b k ,ω 2k is the model coefficient in the single factor comprehensive model, representing the environmental impact coefficient of the kth environmental indicator on the ecological environment difference; ε′ represents the constant in the single factor comprehensive model.

[0032] Further, the S7 is specifically:

[0033] After taking the absolute value of each species impact coefficient, a density map is drawn according to the density distribution characteristics to obtain a species density map; the species impact coefficient corresponding to the maximum density is extracted from the species density map as the species evaluation threshold; the species corresponding to the species impact coefficient greater than the species evaluation threshold is taken as the main species;

[0034] After taking the absolute value of each environmental impact coefficient, a density map is drawn according to the density distribution characteristics to obtain an environmental density map; the environmental impact coefficient corresponding to the maximum density is extracted from the environmental density map as the environmental evaluation threshold; the environmental index corresponding to the environmental impact coefficient greater than the environmental evaluation threshold is taken as the main environmental index;

[0035] Among them, the main species and the main environmental indicators are the influencing factors of the fish community.

[0036] Furthermore, after obtaining the fish community sequence and the environment sequence in S1, the method further includes: performing standardization processing on the fish community sequence and the environment sequence.

[0037] Based on the above-mentioned intelligent identification method of fish community influencing factors, the present invention also provides an intelligent identification system of fish community influencing factors.

[0038] An intelligent identification system for fish community influencing factors, comprising:

[0039] An overall sequence construction module is used to collect fish community data and environmental data at multiple points, obtain fish community sequences and environmental sequences accordingly, and merge the fish community sequences and environmental sequences to construct an overall sequence;

[0040] A Euclidean distance sequence calculation module, which is used to calculate the Euclidean distances between any two points in the overall sequence, the fish community sequence and the environmental sequence, respectively, and obtain a first Euclidean distance sequence, a second Euclidean distance sequence and a third Euclidean distance sequence accordingly;

[0041] A comprehensive factor model building module, which is used to build a comprehensive factor model according to the first Euclidean distance sequence, the second Euclidean distance sequence and the third Euclidean distance sequence;

[0042] A species single factor model building module, which is used to calculate the Euclidean distance sequence between each species in the fish community sequence and build a species single factor model according to the second Euclidean distance sequence and the Euclidean distance sequence between each species in the fish community sequence;

[0043] An environmental single factor model building module, which is used to calculate the Euclidean distance sequence between each environmental indicator in the environmental sequence and to build an environmental single factor model according to the third Euclidean distance sequence and the Euclidean distance sequence between each environmental indicator in the environmental sequence;

[0044] A single factor comprehensive model building module, which is used to substitute the species single factor model and the environmental single factor model into the comprehensive factor model to obtain a single factor comprehensive model;

[0045] The influencing factor identification module is used to intelligently identify the influencing factors of fish communities according to the density distribution characteristics of the model coefficients in the single-factor comprehensive model.

[0046] Based on the above-mentioned intelligent identification method of fish community influencing factors, the present invention also provides an intelligent identification device for fish community influencing factors.

[0047] A device for intelligently identifying factors affecting fish communities, comprising a processor, a memory and a computer program stored in the memory. When the computer program is executed by the processor, the method for intelligently identifying factors affecting fish communities as described above is implemented.

[0048] The beneficial effects of the present invention are as follows: the present invention provides an intelligent identification method, system and device for fish community influencing factors, based on fish community and environmental indicator data, digitally includes an ecological model of overall factors of biological and environmental factors to quantify the comprehensive influence of biological and environmental factors, which can effectively avoid the deviation of incomplete understanding of ecosystem stability and resilience due to ignoring or weakening the influence of water environment factors and the interaction between water environment and aquatic organisms; then build a species single factor model and an environmental single factor model of local points respectively, and quantify the differences in ecological environment from top to bottom; finally, a single factor comprehensive model is used to intelligently diagnose important species and environmental indicators that affect the differences in fish community structure and water environment based on the density distribution characteristics of the model coefficients. By more comprehensively considering the interaction between water environment and fish community, the present invention can intelligently identify the guiding species and environmental factors that cause differences in fish community structure, effectively enhance the understanding of scientists and environmental managers on the integrity of ecosystem stability and resilience, and has certain guiding significance for the protection and management of aquatic ecosystems. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flow chart of an intelligent identification method of fish community influencing factors of the present invention;

[0050] Figure 2 It is the density distribution map of the impact of species and environmental indicators on ecological environment differences in the example;

[0051] Figure 3 The present invention is a structural block diagram of an intelligent identification system for fish community influencing factors. DETAILED DESCRIPTION

[0052] 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.

[0053] like Figure 1 As shown, an intelligent identification method for fish community influencing factors includes:

[0054] S1, collecting fish community data and environmental data from multiple points, obtaining corresponding fish community sequences and environmental sequences, and merging the fish community sequences and the environmental sequences to construct an overall sequence.

[0055] S2, respectively calculating the Euclidean distances between any two points in the overall sequence, the fish community sequence and the environmental sequence, and correspondingly obtaining a first Euclidean distance sequence, a second Euclidean distance sequence and a third Euclidean distance sequence.

[0056] S3: building a comprehensive factor model according to the first Euclidean distance sequence, the second Euclidean distance sequence and the third Euclidean distance sequence.

[0057] S4, calculating the Euclidean distance sequence between each species in the fish community sequence and building a species single factor model based on the second Euclidean distance sequence and the Euclidean distance sequence between each species in the fish community sequence and the fish community sequence.

[0058] S5, calculating the Euclidean distance sequence between two points of each environmental indicator in the environmental sequence, and building an environmental single factor model according to the third Euclidean distance sequence and the Euclidean distance sequence between two points of each environmental indicator in the environmental sequence.

[0059] S6, substituting the species single factor model and the environment single factor model into the comprehensive factor model to obtain a single factor comprehensive model.

[0060] S7, intelligently identifying the fish community influencing factors according to the density distribution characteristics of the model coefficients in the single factor comprehensive model.

[0061] Each step is described in detail below.

[0062] S1 is the overall sequence of construction:

[0063] The aquatic environment includes the physical environment and the biological environment. In order to better highlight the actual situation of the aquatic environment, fish community data (the measured values ​​of each species in the fish community at a point) and environmental data (the measured values ​​of each environmental index in the environment at a point) are used as the performance of the characteristics of the point. For the aquatic environment characteristics of the same point, the overall and local methods are used from top to bottom to express the differences in the ecological environment. The local method is to use fish community data to represent the biological environment of the local point, and the environmental index data to represent the physical environment of the local point; the overall method is to use fish community data and environmental index data as a whole to represent the environmental characteristics of the aquatic ecology as a whole, and the overall sequence can be constructed.

[0064] Fish community data is used to represent the biological environment of a local point. The measured values ​​of each species in the fish community at each point (which can be abundance data or biomass data) constitute a fish community sequence; the fish community sequence is expressed as:

[0065]

[0066] Wherein, F represents the fish community sequence, n represents the total number of species in the fish community, m represents the total number of points, j represents the ordinal number of the species, i represents the ordinal number of the point, and f ji Represents the measured value of the j-th species at the i-th point.

[0067] Environmental index data is used to represent the physical environment of a local point. The measured values ​​of each environmental index at each point in the environment constitute an environmental sequence; the environmental sequence is expressed as:

[0068]

[0069] Wherein, E represents the environmental sequence, l represents the total number of environmental indicators, k represents the ordinal number of environmental indicators, and e ki Represents the measured value of the kth environmental indicator at the i-th point.

[0070] The fish community data and environmental index data are taken as a whole to represent the environmental characteristics of the aquatic ecology as a whole, thereby constructing an overall sequence; the overall sequence is expressed as:

[0071]

[0072] Wherein, T represents the overall sequence; [] ′ Indicates transpose.

[0073] In order to eliminate the impact of data dimension on the calculation results, the fish community sequence and environmental sequence need to be standardized before analysis. The formula is as follows:

[0074]

[0075] Where x represents a sequence, represents the mean of the sequence x, σ represents the standard deviation of the sequence x, and Z represents the standardized sequence x.

[0076] Therefore, after obtaining the fish community sequence and the environmental sequence, the method further includes: performing standardization processing on the fish community sequence and the environmental sequence.

[0077] S2~S6 are used to establish the ecological environment difference model:

[0078] The ecological environment of different points is formed by the biological factors and abiotic factors of the points, forming the unique performance of each point. In order to better show the ecological environment characteristics of each point and quantify the ecological environment differences between points, the present invention selects the spatial Euclidean distance to quantify the ecological environment differences of each point.

[0079] The difference in ecological environment between the two points is mainly caused by the difference in biological factors and environmental factors. The difference caused by biological factors is caused by the difference in fish community structure; while the difference caused by environmental factors is caused by the difference in environmental indicators. Therefore, the difference in ecological environment between the two points can be divided into two levels: the first level is the comprehensive factor model that expresses the overall ecological environment, which is the difference when biological factors and environmental factors are used as a whole to represent the characteristics of each point; the second level is the single factor model that expresses the ecological environment of the local point, which is the distribution difference between species in biological factors and the performance difference of each environmental indicator in environmental factors.

[0080] By building a comprehensive factor model, we can obtain the overall comprehensive impact of biological factors and environmental factors on the overall differences; through the analysis of single factor models, we can evaluate the individual impact of each species on the differences in biological factors between locations and the individual impact of each environmental indicator on the differences in environmental factors, and quantify the impact of single factor indicators on fish community structure or water environment differences; based on the density distribution of single factor indicator influence coefficients, we can intelligently and accurately identify the main species that cause differences in biological factors and the main environmental indicators that cause differences in environmental factors, and calculate from the bottom up the impact of the main species and main environmental indicators on the overall ecological environment differences.

[0081] S2 is the Euclidean distance calculated for each sequence (environmental sequence, fish community sequence and overall sequence).

[0082] First, the Euclidean distance sequence of each sequence is calculated as follows, and the sequence length is r = C 2 m (m is the total number of points):

[0083] The Euclidean distance between two points in the overall sequence is the first Euclidean distance sequence: Tr ;

[0084] The Euclidean distance between two points in the fish community sequence is used to obtain the second Euclidean distance sequence: F r ;

[0085] The Euclidean distance between two points in the environment sequence gives the third Euclidean distance sequence: E r ;

[0086] Specifically, when calculating the second Euclidean distance sequence F r When calculating the third Euclidean distance sequence E, each column of the fish community sequence F is used as the characteristic value of each point, and the Euclidean distance between two points is calculated based on the characteristic value; that is, a point is represented by a vector containing n species, and then the Euclidean distance between each vector is calculated. r When calculating the first Euclidean distance sequence T, each column of the environmental sequence E is used as the characteristic value of each point, and the Euclidean distance between two points is calculated based on the characteristic value; that is, a point is represented by a vector containing l environmental indicators, and then the Euclidean distance between each vector is calculated. r When , each column of the overall sequence T is used as the characteristic value of each point, and the Euclidean distance between two points is calculated based on the characteristic value; that is, a point is represented by a vector containing n species and l environmental indicators, and then the Euclidean distance between each vector is calculated.

[0087] S3 is to build a comprehensive factor model:

[0088] T was calculated using the Pierce correlation coefficient r and F r , T r and E r 、F r and E r The correlation coefficient of is significant. When the correlation coefficient is significant, the applicability of this method is better. The comprehensive factor model is built as follows:

[0089] T r ~α×F r +β×E r +ε;

[0090] Among them, T r represents the first Euclidean distance sequence, F r represents the second Euclidean distance sequence, E r represents the third Euclidean distance sequence; α and β represent the fitting coefficients of least squares fitting, that is, α represents the comprehensive impact of biological factors on ecological environment differences, and β represents the comprehensive impact of environmental factors on ecological environment differences; ε represents the fitting constant of least squares fitting.

[0091] S4 is to build a single factor model - species single factor model:

[0092] For the fish community sequence F, there are certain differences in the distribution of each species between the points. It is precisely because of these differences that there are differences in biological factors at each point. For the difference between two points of a single species, the absolute value of the difference between the measured values ​​of the species at the two points is used to represent it. Assume that the measured values ​​of the jth species at the i-th point and the i+1-th point are f ji and f ji+1 , then the difference between the species at these two points is:

[0093] Q j,i-i+1 =|f ji -f ji+1 |;

[0094] From this, we can get the Euclidean distance sequence Q between two points of the j-th species j , whose length is r. Therefore, the Euclidean distance sequence between each species can be expressed as:

[0095] [Q1,Q2,…,Q j ,…,Q n ];

[0096] The species single factor model constructed in this way is expressed as:

[0097]

[0098] Among them, F r represents the second Euclidean distance sequence, n represents the total number of species in the fish community, j represents the ordinal number of the species, Q j represents the Euclidean distance sequence between two points of the jth species in the fish community sequence, a j represents the fitting coefficient using least squares fitting, ε f Represents the fitting constant using least squares fitting. a j Characterize the individual impact of the jth species on the differences in biological factors. The solution process uses the stepwise regression method and selects the optimal variables to be included in the model based on the AIC criterion. The a j That is, the individual influence of the jth species on the differences in biological factors between sites, and the species variables included in the model are the species factors that cause the differences in biological factors between sites.

[0099] S5 is to build a single factor model - environmental single factor model:

[0100] For the environmental sequence E, there are certain differences in the distribution of each environment between points. It is precisely because of these differences that there are differences in environmental factors at each point. For the difference between two points of a single environment, the absolute value of the difference between the measured values ​​of the environmental index at the two points is used to represent it. Assume that the measured values ​​of the kth environmental index at the i-th point and the i+1-th point are e ki and e ki+1 , then the difference of the environmental index at these two points is:

[0101] EQ k,i-i+1 =|e ki -e ki+1 |;

[0102] From this, we can get the Euclidean distance sequence EQ between two points of the kth environmental index k , whose length is r.

[0103] Therefore, the Euclidean distance sequence between each pair of environmental indicators can be expressed as:

[0104] [EQ1,EQ2,…,EQ k ,…,EQ n ];

[0105] Environmental single factor model

[0106] The environmental single factor model constructed in this way is expressed as:

[0107]

[0108] Among them, E r represents the third Euclidean distance sequence, l represents the total number of environmental indicators, k represents the ordinal number of environmental indicators, EQ k represents the Euclidean distance sequence between two points of the kth environmental indicator in the environmental sequence, b k represents the fitting coefficient using least squares fitting, ε e represents the fitting constant using least squares fitting. k Characterize the individual impact of the kth environmental indicator on the difference of environmental factors. The solution process adopts the stepwise regression method and selects the optimal variables to be included in the model according to the AIC criterion. The b k That is, the individual impact of the kth environmental indicator on the differences in environmental factors between locations, and the environmental variables included in the model are the environmental factors that cause the differences in environmental factors between locations.

[0109] S6 is to build a single factor comprehensive model:

[0110] The main factors of the ecological environment difference between two points include biological factors and environmental factors. The biological factors are affected by the species distribution of each point, and the environmental factors are affected by the performance of environmental indicators of each point. Therefore, by combining the comprehensive factor model, the species single factor model, and the environmental single factor model, we can obtain the comprehensive impact of a single factor (species indicator or environmental indicator) on the overall ecological environment difference and obtain a single factor comprehensive model. The process is as follows:

[0111] Comprehensive Factor Model:

[0112] T r ~α×F r +β×E r +ε;

[0113] Species single factor model:

[0114]

[0115] Environmental single factor model:

[0116]

[0117] The species single factor model F r , Environmental single factor model E r Substitute into the comprehensive factor model T r , and obtain the single factor comprehensive model:

[0118]

[0119] After simplification, we get:

[0120]

[0121] Among them, ω 1j =α×a j ,ω 1j is the model coefficient in the single factor comprehensive model, indicating the species influence coefficient of the jth species on the ecological environment difference; ω 2k =β×b k ,ω 2k is the model coefficient in the single factor comprehensive model, representing the environmental impact coefficient of the kth environmental indicator on the ecological environment difference; ε′ represents the constant in the single factor comprehensive model.

[0122] S7 is an important factor for intelligent identification:

[0123] Based on the single factor comprehensive model, the influence coefficient sequence ω of species on ecological environment differences was extracted respectively. 1j , the coefficient sequence of environmental indicators on ecological environment differences ω 2kAfter taking the absolute values ​​of the two sequences, we draw a density map based on the density distribution characteristics, and intelligently and accurately identify the driving factors that affect the structure of fish communities, including major species and major environmental indicators; we extract the value ω corresponding to the maximum density. 10 and ω 20 , as the evaluation threshold; the influence coefficient of species on ecological environment differences is greater than ω 10 The species with the largest number of species are the main species, and the influence coefficient of environmental indicators on ecological environment differences is greater than ω 20 The environmental indicators are the main environmental indicators.

[0124] Therefore, the S7 is specifically:

[0125] After taking the absolute value of each species impact coefficient, a density map is drawn according to the density distribution characteristics to obtain a species density map; the species impact coefficient corresponding to the maximum density is extracted from the species density map as the species evaluation threshold; the species corresponding to the species impact coefficient greater than the species evaluation threshold is taken as the main species;

[0126] After taking the absolute value of each environmental impact coefficient, a density map is drawn according to the density distribution characteristics to obtain an environmental density map; the environmental impact coefficient corresponding to the maximum density is extracted from the environmental density map as the environmental evaluation threshold; the environmental index corresponding to the environmental impact coefficient greater than the environmental evaluation threshold is taken as the main environmental index;

[0127] Among them, the main species and the main environmental indicators are the influencing factors of the fish community.

[0128] In summary, the method of the present invention has the following advantages:

[0129] (1) Based on the data of fish communities and environmental indicators, the method of the present invention divides the ecological and environmental differences of sampling points into overall differences (comprehensive factor model) and local differences (species single factor model and environmental single factor model) from top to bottom, which can more completely reflect the interaction between aquatic organisms and water environment in aquatic ecosystems.

[0130] (2) The species single-factor model and the environmental single-factor model quantified the impact of single-factor indicators on the differences in fish community structure and aquatic environment, laying the foundation for the intelligent identification of the main species and main environmental indicators of the differences in fish community structure and aquatic environment.

[0131] (3) The single-factor comprehensive model quantifies the impact of various indicators on the overall ecological environment from the bottom up, and relies on the density distribution characteristics of the model coefficients to realize the intelligent identification of the factors affecting the fish community structure, effectively avoiding the differences in evaluation conclusions due to lack of experience or inadequate understanding, and improving the universality of the model.

[0132] The method of the present invention is described below with specific examples:

[0133] Environmental factors and biological factors in aquatic ecosystems interact and influence each other. This example uses some physical and chemical indicators of water bodies as environmental factors, and all fish caught in the survey and sampling as biological factors. The two together constitute the environmental factors and biological factors that affect the distribution characteristics of fish communities. The data uses the abundance data and environmental data of fish communities in the ecosystem of the same water area and the same period. After calculation, we get:

[0134] (1) The comprehensive factor model is:

[0135] T r =0.64×F r +0.73×E r +0.37

[0136] It can be seen that from the comprehensive factor model, the factor with the greatest impact on the ecological environment difference is environmental factors, with an impact coefficient of 0.73, followed by fish communities of biological factors, with an impact coefficient of 0.64. This coefficient more intuitively and accurately shows the impact of aquatic organisms and water environment in aquatic ecosystems on the overall ecological environment.

[0137] (2) Single Factor Model

[0138] Species single factor model:

[0139] F r =-1.94 wheatear fish+0.80 Japanese shrimp+0.54 crucian carp+0.43 +0.48 Coilfish

[0140] +0.57 scorpionfish-1.48 scorpionfish+1.18 bigfin scorpionfish+0.16 redfin scorpionfish

[0141] +0.80 Short-jawed Coilfish

[0142] Environmental single factor model:

[0143] E r =0.13PO4+0.24Chl.a+0.25COD+0.28DO+0.28TN+0.32NH4

[0144] +0.33SD+0.36NO3+0.37NO2+0.37WT+0.39TP

[0145] +0.40COND

[0146] The species single factor model and the environment single factor model show the relative impact of each indicator on local differences, and quantify the differences in local ecological environment from top to bottom. The fish impacts are ranked from largest to smallest: wheatear fish, mullet, bigfin mullet, Japanese shrimp, short-jawed anchovy, scorpionfish, crucian carp, scorpionfish, and redfin bream; the environmental indicators from largest to smallest are COND (conductivity), TP (total phosphorus), WT (water temperature), NO2 (nitrogen dioxide), NO3 (nitrate nitrogen), SD (transparency), NH4 (nitrite nitrogen), TN (total nitrogen), DO (dissolved oxygen), COD (permanganate index), Chl.a (chlorophyll a), PO4 (orthophosphate). Therefore, researchers can clearly understand the relative impact of each indicator in the ecological environment difference, which is conducive to improving the overall understanding of the stability and resilience of the ecosystem composed of fish communities and their surrounding environment.

[0147] (3) Single factor comprehensive model

[0148] F r 、E r Bring in T r , a single factor comprehensive model was obtained, and the model coefficients are shown in Table 1 below.

[0149] Table 1: Model coefficient table

[0150]

[0151] The density distribution diagram of the impact of species and environmental indicators on ecological environment differences constructed by model coefficients is shown in the figure below: Figure 2 shown.

[0152] From the single factor comprehensive model and density distribution map, the influence of various species on fish communities in biological factors varies greatly, among which the positive influence is greater on short jawed anchovy, Japanese shrimp, and big fin bream; the negative influence is greater on whitehead and intersex; the influence of various factors in environmental factors is not much different, among which NO2, NO3, WT, TP, and COND have greater influence. Therefore, the main species with greater impact on ecological environment differences are short jawed anchovy, Japanese shrimp, big fin bream, whitehead, and intersex, and the main environmental indicators are NO2, NO3, WT, TP, and COND.

[0153] Therefore, in actual business, scientists and environmental managers need to focus on and study the diversity, quantity, distribution and niche characteristics of species such as short-jawed anchovy, Japanese shrimp, bigfin silver carp, wheatear fish and inter-underfish, so as to grasp the main biological object data of community structure changes in a timely and accurate manner; for the water environment indicators such as NO2, NO3, WT, TP and COND, it is necessary to pay attention to their changes, which is conducive to timely identification of signs of pollution, habitat destruction or other environmental pressures, and has a good guiding significance for effectively evaluating and monitoring the health of the water ecological environment.

[0154] Based on the above-mentioned intelligent identification method of fish community influencing factors, the present invention also provides an intelligent identification system of fish community influencing factors.

[0155] like Figure 3 As shown, an intelligent identification system for fish community influencing factors includes:

[0156] An overall sequence construction module is used to collect fish community data and environmental data at multiple points, obtain fish community sequences and environmental sequences accordingly, and merge the fish community sequences and environmental sequences to construct an overall sequence;

[0157] A Euclidean distance sequence calculation module, which is used to calculate the Euclidean distances between any two points in the overall sequence, the fish community sequence and the environmental sequence, respectively, and obtain a first Euclidean distance sequence, a second Euclidean distance sequence and a third Euclidean distance sequence accordingly;

[0158] A comprehensive factor model building module, which is used to build a comprehensive factor model according to the first Euclidean distance sequence, the second Euclidean distance sequence and the third Euclidean distance sequence;

[0159] A species single factor model building module, which is used to calculate the Euclidean distance sequence between each species in the fish community sequence and build a species single factor model according to the second Euclidean distance sequence and the Euclidean distance sequence between each species in the fish community sequence;

[0160] An environmental single factor model building module, which is used to calculate the Euclidean distance sequence between each environmental indicator in the environmental sequence and to build an environmental single factor model according to the third Euclidean distance sequence and the Euclidean distance sequence between each environmental indicator in the environmental sequence;

[0161] A single factor comprehensive model building module, which is used to substitute the species single factor model and the environmental single factor model into the comprehensive factor model to obtain a single factor comprehensive model;

[0162] The influencing factor identification module is used to intelligently identify the influencing factors of fish communities according to the density distribution characteristics of the model coefficients in the single-factor comprehensive model.

[0163] Based on the above-mentioned intelligent identification method of fish community influencing factors, the present invention also provides an intelligent identification device for fish community influencing factors.

[0164] A device for intelligently identifying factors affecting fish communities, comprising a processor, a memory and a computer program stored in the memory. When the computer program is executed by the processor, the method for intelligently identifying factors affecting fish communities as described above is implemented.

[0165] The present invention discloses an intelligent identification method, system and device for fish community influencing factors. Based on fish community and environmental index data, the ecological model of the overall factors including biological and environmental factors is digitized to quantify the comprehensive influence of biological and environmental factors, which can effectively avoid the deviation of incomplete understanding of ecosystem stability and resilience due to ignoring or weakening the influence of water environment factors and the interaction between water environment and aquatic organisms; then, a species single factor model and an environmental single factor model of local points are respectively constructed to quantify the differences in ecological environment from top to bottom; finally, a single factor comprehensive model is adopted to intelligently diagnose the important species and environmental indicators that affect the differences in fish community structure and water environment based on the density distribution characteristics of the model coefficients. The present invention can intelligently identify the guiding species and environmental factors that lead to differences in fish community structure by more comprehensively considering the interaction between water environment and fish community, effectively improve the understanding of scientists and environmental managers on the integrity of ecosystem stability and resilience, and has certain guiding significance for the protection and management of water ecosystems.

[0166] 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 identification method for fish community influencing factors, characterized in that: include: S1, collecting fish community data and environmental data at multiple points, obtaining corresponding fish community sequences and environmental sequences, and merging the fish community sequences and environmental sequences to construct an overall sequence; S2, respectively calculating the Euclidean distances between any two points in the overall sequence, the fish community sequence and the environmental sequence, and correspondingly obtaining a first Euclidean distance sequence, a second Euclidean distance sequence and a third Euclidean distance sequence; S3, building a comprehensive factor model according to the first Euclidean distance sequence, the second Euclidean distance sequence and the third Euclidean distance sequence; S4, calculating the Euclidean distance sequence between two points of each species in the fish community sequence, and building a species single factor model according to the second Euclidean distance sequence and the Euclidean distance sequence between two points of each species in the fish community sequence; S5, calculating the Euclidean distance sequence between two points of each environmental indicator in the environmental sequence, and building an environmental single factor model according to the third Euclidean distance sequence and the Euclidean distance sequence between two points of each environmental indicator in the environmental sequence; S6, substituting the species single factor model and the environment single factor model into the comprehensive factor model to obtain a single factor comprehensive model; S7, intelligently identifying the fish community influencing factors according to the density distribution characteristics of the model coefficients in the single factor comprehensive model.

2. The intelligent identification method of fish community influencing factors according to claim 1 is characterized in that: In S1, the fish community sequence is represented as: The environment sequence is expressed as: The overall sequence is expressed as: Wherein, F represents the fish community sequence, n represents the total number of species in the fish community, m represents the total number of points, j represents the ordinal number of the species, i represents the ordinal number of the point, and f ji represents the measured value of the jth species at the ith point; E represents the environmental sequence, l represents the total number of environmental indicators, k represents the ordinal number of environmental indicators, e ki represents the measured value of the kth environmental indicator at the ith point; T represents the overall sequence.

3. The intelligent identification method of fish community influencing factors according to claim 1 is characterized in that: In S3, the comprehensive factor model is expressed as: T r ~α×F r +β×E r +e; Among them, T r represents the first Euclidean distance sequence, F r represents the second Euclidean distance sequence, E r represents the third Euclidean distance sequence, α and β represent the fitting coefficients of least squares fitting, and ε represents the fitting constant of least squares fitting.

4. The intelligent identification method of fish community influencing factors according to claim 3 is characterized in that: In S4, the species single factor model is expressed as: Among them, F r represents the second Euclidean distance sequence, n represents the total number of species in the fish community, j represents the ordinal number of the species, Q j represents the Euclidean distance sequence between two points of the jth species in the fish community sequence, a j represents the fitting coefficient using least squares fitting, ε f Represents the fitting constant using least squares fitting.

5. The intelligent identification method of fish community influencing factors according to claim 4 is characterized in that: In S5, the environmental single factor model is expressed as: Among them, E r represents the third Euclidean distance sequence, l represents the total number of environmental indicators, k represents the ordinal number of environmental indicators, EQ k represents the Euclidean distance sequence between two points of the kth environmental indicator in the environmental sequence, b k represents the fitting coefficient using least squares fitting, ε e Represents the fitting constant using least squares fitting.

6. The intelligent identification method of fish community influencing factors according to claim 5 is characterized in that: In S6, the single factor comprehensive model is expressed as: Among them, ω 1j =α×a j ,ω 1j is the model coefficient in the single factor comprehensive model, indicating the species influence coefficient of the jth species on the ecological environment difference; ω 2k =β×b k ,ω 2k is the model coefficient in the single factor comprehensive model, representing the environmental impact coefficient of the kth environmental indicator on the ecological environment difference; ε′ represents the constant in the single factor comprehensive model.

7. The intelligent identification method of fish community influencing factors according to claim 6 is characterized in that: The S7 is specifically: After taking the absolute value of each species impact coefficient, a density map is drawn according to the density distribution characteristics to obtain a species density map; the species impact coefficient corresponding to the maximum density is extracted from the species density map as the species evaluation threshold; the species corresponding to the species impact coefficient greater than the species evaluation threshold is taken as the main species; After taking the absolute value of each environmental impact coefficient, a density map is drawn according to the density distribution characteristics to obtain an environmental density map; the environmental impact coefficient corresponding to the maximum density is extracted from the environmental density map as the environmental evaluation threshold; the environmental index corresponding to the environmental impact coefficient greater than the environmental evaluation threshold is taken as the main environmental index; Among them, the main species and the main environmental indicators are the influencing factors of the fish community.

8. The intelligent identification method of fish community influencing factors according to claim 1 is characterized in that: After obtaining the fish community sequence and the environment sequence in S1, the method further includes: performing standardization processing on the fish community sequence and the environment sequence.

9. An intelligent identification system for fish community influencing factors, characterized in that: include: An overall sequence construction module is used to collect fish community data and environmental data at multiple points, obtain fish community sequences and environmental sequences accordingly, and merge the fish community sequences and environmental sequences to construct an overall sequence; A Euclidean distance sequence calculation module, which is used to calculate the Euclidean distances between any two points in the overall sequence, the fish community sequence and the environmental sequence, respectively, and obtain a first Euclidean distance sequence, a second Euclidean distance sequence and a third Euclidean distance sequence accordingly; A comprehensive factor model building module, which is used to build a comprehensive factor model according to the first Euclidean distance sequence, the second Euclidean distance sequence and the third Euclidean distance sequence; A species single factor model building module, which is used to calculate the Euclidean distance sequence between each species in the fish community sequence and build a species single factor model according to the second Euclidean distance sequence and the Euclidean distance sequence between each species in the fish community sequence; An environmental single factor model building module, which is used to calculate the Euclidean distance sequence between each pair of environmental indicators in the environmental sequence, and build an environmental single factor model according to the third Euclidean distance sequence and the Euclidean distance sequence between each pair of environmental indicators in the environmental sequence; A single factor comprehensive model building module, which is used to substitute the species single factor model and the environmental single factor model into the comprehensive factor model to obtain a single factor comprehensive model; The influencing factor identification module is used to intelligently identify the influencing factors of fish communities according to the density distribution characteristics of the model coefficients in the single-factor comprehensive model.

10. An intelligent identification device for fish community influencing factors, characterized in that: The method comprises a processor, a memory and a computer program stored in the memory, wherein when the computer program is executed by the processor, the method for intelligently identifying the factors affecting fish communities as described in any one of claims 1 to 8 is implemented.

Images