Method and system for identifying stress sources of freshwater fish communities under external environmental influences

Abstract

The present invention belongs to, but is not limited to, the field of ecological environmental protection technologies, and discloses a method for identifying the main stressors of freshwater fish communities under the influence of climate change and human activities. Data on historical presence points and absence points of the target freshwater fish community are collected, and bioclimatic data and human activity-related data affecting fish distribution are collected to establish a variable set; VIF is used to screen variables without collinearity, machine learning algorithms are used to construct species distribution models for each organism in the fish community, the distribution of the fish community is simulated based on S-SDM, and environmental factors with higher model contribution are screened; according to the machine learning sub-models selected by AUC and TSS, stacking and prediction of the ensemble model are realized; the impacts of climate change, human activities and other stressors on the freshwater fish community are separated; the I e and HCP indicators are used to evaluate the main stressors and their influencing degrees of the freshwater fish community. The present invention identifies the main stressors and their influencing magnitudes from the perspective of the fish community, and solves the problem of biodiversity protection facing multiple multiple stressors.

Description

Technical Field

[0001] The present invention belongs to but is not limited to the field of ecological environmental protection technology, and in particular relates to a method and system for identifying major stress sources of freshwater fish communities under the influence of climate change and human activities. Background Art

[0002] The loss of freshwater biodiversity has always been a serious problem facing ecological protection at home and abroad, and freshwater ecological protection and restoration is an important task of strategic significance at the national level. Human activities have become an important source of stress for China's freshwater ecosystems, and climate change is the main cause of freshwater biodiversity loss and species extinction in China and even the world. Not only that, climate change will intensify the impact of existing stressors on biodiversity. Existing freshwater biological protection strategies mainly target single species and single stressors, and it is difficult to ignore one and lose the other when formulating protection strategies. Moreover, due to the diversity of stressors faced by freshwater organisms, it is difficult to formulate effective targeted measures to restore freshwater ecosystems and protect freshwater biodiversity. On the basis of species distribution models, stacking the collective models of all species and constructing a stacked species distribution model at the community level is an effective way to deal with the problem of coordinating the protection of multiple species. Separating the impacts of climate change and human activities and calculating the impact coefficient of stressors can propose targeted solutions for regional biological ecosystem protection.

[0003] Existing strategies for freshwater ecological protection cannot propose effective measures for community protection from a macro perspective and lack targeted strategies, wasting human and material resources. Therefore, this method for formulating protection strategies for freshwater fish communities to cope with multiple stressors is of great significance. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a method and system for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities.

[0005] The present invention is implemented as follows: a method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities, comprising the following steps:

[0006] Step 1: Collect historical data on the presence and absence of target freshwater fish communities, collect bioclimate data that affect fish distribution and data related to human activities to establish a variable set;

[0007] Step 2: Use the variance inflation factor (VIF) to screen variables that are not collinear, preliminarily select a machine learning algorithm, build a species distribution model for each organism in the fish community, simulate the distribution of fish communities based on the stacked species distribution model (S-SDM), and screen environmental factors with higher contribution to the model;

[0008] Step 3: Based on the two evaluation indicators of the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS), select the machine learning sub-model with superior performance to achieve stacking and prediction of the ensemble model;

[0009] Step 4: Isolate the impacts of climate change, human activities, and other stressors on freshwater fish communities;

[0010] Step 5: Use the stressor influence coefficient (I e ) and habitat colonization potential (HCP) indicators were used to evaluate the main stressors of freshwater fish communities and their impact levels.

[0011] Further, the specific steps of step 2 are as follows:

[0012] Step 2.1: Preliminarily select no less than 5 sub-models from the available machine learning model set M as the candidate model set M p ;

[0013] Step 2.2: Determine the factors to be screened, including climate and environmental factors Env c 、Human activity related factors Env h Etc., constitute the factor data set Env to be screened;

[0014] Step 2.3: Calculate the variance inflation factor (VIF) of each factor in Env, and remove environmental factors whose VIF magnitude is higher than that of other factors; the calculation formula of the variance inflation factor (VIF) is as follows:

[0015]

[0016] In the formula, j is the jth environmental factor;

[0017] ---The sum of the squares of the relative changes of this environmental factor and all other environmental factors;

[0018] Step 2.4: Use the remaining environmental factors to simulate the model according to the ensemble model set M p Contribution results, remove the variables with an average contribution of less than 10% of the freshwater fish community as the modeling factor data set Env s .

[0019] Further, the specific steps of step 3 are as follows:

[0020] Step 3.1: Selected model set M p Each model was run ten times;

[0021] Step 3.2: Randomly select 70% of the total presence and absence data as the training set to fit the algorithm, and the remaining 30% is reserved for evaluating the algorithm performance;

[0022] Step 3.3: Evaluate on the evaluation dataset using the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS);

[0023] Step 3.4: When the sub-model set M s After determination, the collective models of each species were stacked to obtain the species distribution model S-SDM (stacked species distribution model) at the community level;

[0024] Step 3.5: Use S-SDM (stacked species distribution model) to simulate and predict different scenarios {C, H} of the target community in the historical period and future period t to obtain the community richness under climate mode i

[0025] Furthermore, the evaluation method in step 3.3 includes: averaging the AUC and TSS of all sub-models to obtain the AUC of the ensemble model e and TSS e , if AUC e >0.85,TSS e >0.75, the ensemble model is considered to have superior performance, and the selected sub-model set is the final sub-model set M s Otherwise, reselect the submodel set M based on the AUC and TSS of the single model p .

[0026] The calculation formula of TSS is as follows:

[0027] TSS=Specificity+Sensitivitg-1 (2)

[0028] In the formula, Specificity is the model specificity, and the calculation formula is:

[0029] Specificity = true negative examples / (true negative examples + false positive examples) (3)

[0030] Sensitivitg is the model sensitivity, and the calculation formula is:

[0031] Sensitivitg = True positive / (True positive + False negative) (4)

[0032] Further, the specific steps of step 4 are as follows:

[0033] Step 4.1: Average the values ​​from n climate models to get a representative future forecast The calculation formula is as follows:

[0034]

[0035] In the formula, i=1,2,......n represents n different climate modes;

[0036] C represents the climate scenario to which the climate factor predicted by the model belongs;

[0037] t represents the corresponding period of the climate factor predicted by the model, generally 2030s, 2050s, 2070s or 2090s;

[0038] H represents the human activity factor for different situations predicted by the model;

[0039] Step 4.2: Isolate the effect of climate factors on community richness; the effect of climate change on community richness is measured by the difference between the model prediction (changing only the climate factor) and the model simulation for the historical period, expressed as ΔR CC , the calculation formula is as follows:

[0040] ΔR CC =R C,t -R (6)

[0041] In the formula, R represents the richness of the community in the historical period;

[0042] Step 4.3: Isolate the impact of human factors on community richness; the impact of human activities on ST richness is measured by the difference between the model simulation of the historical period and the simulation of the historical period projected onto the human activity factor, expressed as ΔR HA , the calculation formula is as follows:

[0043] ΔR HA =R H -R (7)

[0044] Step 4.4: Calculate the combined effects of climate change and human activities on community richness; the combined effect is the combined effect of climate change and human activities on community richness, expressed as ΔR, and the calculation formula is as follows:

[0045]

[0046] Further, the specific steps of step 5 are as follows:

[0047] Step 5.1: Calculate the stressor impact coefficient Ie of the abundance change of each grid. The calculation formula is:

[0048] I e =|ΔR HA | / (|ΔR CC |+|ΔR HA|) (9)

[0049] Step 5.2: Use the habitat colonization potential (HCP) to quantitatively calculate the magnitude and trend of community richness changes;

[0050] Step 5.3: According to the pressure source influence coefficient I e Identify the main stressors in the area and determine the impact of the stressors in the area based on the HCP.

[0051] Another object of the present invention is to provide a method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities, and a system for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities, comprising:

[0052] The variable set module is established to collect historical data on the presence and absence of target freshwater fish communities, and to collect bioclimate data that affect fish distribution and data related to human activities to establish a variable set;

[0053] The environmental factor screening module is used to screen variables without collinearity using the variance inflation factor (VIF), preliminarily select the machine learning algorithm, build the species distribution model of each organism in the fish community, simulate the distribution of fish communities based on the S-SDM stacked species distribution model (S-SDM), and screen environmental factors with higher contribution to the model;

[0054] The model stacking and prediction module is used to select the machine learning sub-model with superior performance based on the two evaluation indicators of the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS), and realize the stacking and prediction of the ensemble model;

[0055] an impact separation module to separate the effects of climate change, human activities, and other stressors on freshwater fish communities;

[0056] The main stress source and impact assessment module is used to use the stress source impact coefficient (I e ) and habitat colonization potential (HCP) indicators were used to evaluate the main stressors of freshwater fish communities and their impact levels.

[0057] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities.

[0058] Another object of the present invention is to provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities.

[0059] Another object of the present invention is to provide an information data processing terminal, which includes the identification system for the main stress sources of freshwater fish communities under the influence of climate change and human activities.

[0060] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0061] First, the method of the present invention for identifying the main stressors of freshwater fish communities under the influence of climate change and human activities can identify the main stressors and the magnitude of their impact from the perspective of fish communities, thereby solving various biodiversity conservation problems facing multiple stressors.

[0062] The present invention provides a method for identifying the main stressors of freshwater fish communities under the influence of climate change and human activities, which solves the problem that the existing technology cannot accurately distinguish and evaluate the impact of multiple environmental stressors on fish communities. Through systematic steps, from data collection, variable screening, model construction to impact analysis, comprehensive identification and quantitative analysis of stressors of freshwater fish communities are achieved. This method not only improves the prediction accuracy of fish community changes, but also provides a scientific basis for formulating protection and management strategies.

[0063] First, the present invention establishes a variable set by collecting historical existence and missing point data of target freshwater fish communities, as well as bioclimatic data and human activity-related data that affect fish distribution. The diversity and completeness of this data provide a solid foundation for the accuracy of subsequent models. By using the variance inflation factor (VIF) to screen variables that do not have collinearity, the reliability and scientificity of the model are further improved. The constructed stacked species distribution model (S-SDM) can simulate the distribution of fish communities and screen out environmental factors that contribute more to the model, thereby accurately locating the main stressors.

[0064] Secondly, the present invention uses two evaluation indicators, the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS), to select machine learning sub-models with superior performance and realize the stacking and prediction of the ensemble model. This process not only ensures the efficiency and accuracy of the model, but also ensures the stability and reliability of the model results through multiple runs and evaluations. By separating the effects of climate change, human activities and other stressors on freshwater fish communities, the independent and superimposed effects of each factor are accurately measured, providing a reliable basis for scientific evaluation.

[0065] Finally, the present invention quantitatively evaluates the main stressors of freshwater fish communities and their impact levels through stressor impact coefficients and habitat colonization potential (HCP) indicators. This evaluation method not only provides the impact size and trend of specific stressors, but also helps identify the main environmental stressors in a specific area, thereby providing key data support for the formulation of protection and management measures. Compared with traditional methods, the present invention has achieved significant technical progress in data processing, model construction and result analysis.

[0066] The present invention successfully solves the problem of the existing technology that it is difficult to accurately identify and evaluate the impact of multiple environmental stressors on freshwater fish communities through systematic data processing and model building methods. Its significant technical progress is reflected in the improvement of prediction accuracy and model reliability, providing a scientific basis and practical tools for the protection and management of freshwater ecosystems. The promotion and application of this method will have a far-reaching impact on the cause of freshwater ecological protection.

[0067] Second, the present invention successfully solves multiple technical problems in the identification of freshwater fish community stressors in the prior art through a series of main parameters, algorithms and mathematical models, and has achieved significant technical progress. First, the present invention uses historical existing point and missing point data, as well as bioclimatic data and human activity-related data to establish a comprehensive set of variables. The variance inflation factor (VIF) is used to screen variables that do not have collinearity, ensuring the stability and scientificity of the model. This process overcomes the problem of model inaccuracy caused by variable collinearity in traditional methods, and significantly improves the accuracy and efficiency of variable screening.

[0068] Secondly, the present invention introduces a stacked species distribution model (S-SDM) and combines it with a machine learning algorithm to construct a fish community distribution model. The sub-models with superior performance are screened by two evaluation indicators, the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS), to achieve the stacking and prediction of the ensemble model. This method overcomes the problem of insufficient prediction ability of a single model in the prior art and significantly improves the prediction accuracy and stability of the model.

[0069] In terms of separating influencing factors, the present invention proposes an innovative separation method to independently analyze the impact of climate change, human activities and other stressors on freshwater fish communities. By calculating the changes in community richness under climate patterns and using the prediction results of historical time period models, the independent impacts of various factors are separated. This method effectively solves the problem of difficulty in distinguishing the independent impacts of multiple stressors in traditional methods and significantly enhances the ability to identify environmental stressors.

[0070] Finally, the present invention quantitatively evaluates the main stressors of freshwater fish communities and their impact levels through stressor impact coefficients and habitat colonization potential (HCP) indicators. This quantitative evaluation method not only provides detailed stressor impact sizes and trends, but also helps identify the main environmental stressors in specific areas, thereby providing key data support for the formulation of protection and management measures. Compared with traditional methods, the present invention has achieved significant technological advances in data processing, model construction, and result analysis, significantly improving the scientificity and practicality of freshwater ecosystem protection and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 This is a flow chart of a method for identifying major stress sources of freshwater fish communities under the influence of climate change and human activities provided by an embodiment of the present invention.

[0072] Figure 2 This is a richness distribution map of the schizothorax subfamily community in the upper reaches of the Yangtze River provided by an embodiment of the present invention.

[0073] Figure 3 This is a diagram of changes in the richness of the subfamily of schizothorax under the influence of climate change in the upper reaches of the Yangtze River provided by an embodiment of the present invention.

[0074] Figure 4 This is a diagram of the changes in the richness of the subfamily of schizothorax under the influence of water resource development in the upper reaches of the Yangtze River provided by an embodiment of the present invention.

[0075] Figure 5 This is a graph of the impact coefficients of stress sources on the schizothorax subfamily community under climate change and water resource development in the upper reaches of the Yangtze River provided by an embodiment of the present invention.

[0076] Figure 6 This is a structural diagram of a system for identifying major stress sources of freshwater fish communities under the influence of climate change and human activities provided by an embodiment of the present invention.

[0077] Figure 7 This is a graph of changes in river network connectivity in the upper reaches of the Yangtze River around 2000, provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0078] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0079] like Figure 1 As shown, an embodiment of the present invention provides a method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities, comprising the following steps:

[0080] Step 1: Collect historical data on the presence and absence of target freshwater fish communities, collect bioclimate data that affect fish distribution and data related to human activities to establish a variable set;

[0081] Step 2: Use the variance inflation factor (VIF) to screen variables that are not collinear, preliminarily select a machine learning algorithm, build a species distribution model for each organism in the fish community, simulate the distribution of fish communities based on the stacked species distribution model (S-SDM), and screen environmental factors with higher contribution to the model;

[0082] Step 3: Based on the two evaluation indicators of the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS), select the machine learning sub-model with superior performance to achieve stacking and prediction of the ensemble model;

[0083] Step 4: Isolate the impacts of climate change, human activities, and other stressors on freshwater fish communities;

[0084] Step 5: Use the stressor influence coefficient (I e ) and habitat colonization potential (HCP) indicators were used to evaluate the main stressors of freshwater fish communities and their impact levels.

[0085] The specific steps of step 2 are as follows:

[0086] Step 2.1: Preliminarily select no less than 5 sub-models from the available machine learning model set M as the candidate model set M p ;

[0087] Step 2.2: Determine the factors to be screened, including climate and environmental factors Env c 、Human activity related factors Env h Etc., constitute the factor data set Env to be screened;

[0088] Step 2.3: Calculate the variance inflation factor (VIF) of each factor in Env, and remove environmental factors whose VIF magnitude is higher than that of other factors; the calculation formula of the variance inflation factor (VIF) is as follows:

[0089]

[0090] In the formula, j is the jth environmental factor;

[0091] ---The sum of the squares of the relative changes of this environmental factor and all other environmental factors;

[0092] Step 2.4: Use the remaining environmental factors to simulate the model according to the ensemble model set Mp Contribution results, remove the variables with an average contribution of less than 10% of the freshwater fish community as the modeling factor data set Env s .

[0093] The specific steps of step 3 are as follows:

[0094] Step 3.1: Selected model set M p Each model was run ten times;

[0095] Step 3.2: Randomly select 70% of the total presence and absence data as the training set to fit the algorithm, and the remaining 30% is reserved for evaluating the algorithm performance;

[0096] Step 3.3: Evaluate on the evaluation dataset using the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS); the calculation formula for TSS is as follows:

[0097] TSS=Specificity+Sensitivitg-1 (11)

[0098] In the formula, Specificity is the model specificity, and the calculation formula is:

[0099] Specificity = true negative examples / (true negative examples + false positive examples) (12)

[0100] Sensitivity is the model sensitivity, and the calculation formula is:

[0101] Sensitivity = True positive / (True positive + False negative) (13)

[0102] The AUC and TSS of all sub-models are averaged to get the AUC of the ensemble model. e and TSS e , if AUC e >0.85,TSS e >0.75, the ensemble model is considered to have superior performance, and the selected sub-model set is the final sub-model set M s Otherwise, reselect the submodel set M based on the AUC and TSS of the single model p ;

[0103] Step 3.4: When the sub-model set determines M s Finally, the collective models of each species were stacked to obtain the species distribution model S-SDM (stacked species distribution model) at the community level;

[0104] Step 3.5: Use S-SDM (stacked species distribution model) to simulate and predict different scenarios {C, H} of the target community in the historical period and future period t to obtain the community richness under climate mode i

[0105] The specific steps of step 4 are as follows:

[0106] Step 4.1: Average the values ​​from n climate models to get a representative future forecast The calculation formula is as follows:

[0107]

[0108] In the formula, i = 1, 2, ... n represents n different climate modes;

[0109] C represents the climate scenario to which the climate factor predicted by the model belongs;

[0110] t represents the corresponding period of the climate factor predicted by the model, usually 2030, 2050, 2070 or 2090;

[0111] H represents the human activity factor for different situations predicted by the model;

[0112] Step 4.2: Isolate the effect of climate factors on community richness; the effect of climate change on community richness is measured by the difference between the model prediction (changing only the climate factor) and the model simulation for the historical period, expressed as ΔR CC , the calculation formula is as follows:

[0113] ΔR CC =R C，t -R (15)

[0114] In the formula, R represents the richness of the community in the historical period;

[0115] Step 4.3: Isolate the impact of human factors on community richness; the impact of human activities on ST richness is measured by the difference between the model simulation of the historical period and the simulation of the historical period projected onto the human activity factor, expressed as ΔR HA , the calculation formula is as follows:

[0116] ΔR HA =R H -R (16)

[0117] Step 4.4: Calculate the combined effects of climate change and human activities on community richness; the combined effect is the combined effect of climate change and human activities on community richness, expressed as ΔR, and the calculation formula is as follows:

[0118]

[0119] The specific steps of step 5 are as follows:

[0120] Step 5.1: Calculate the stressor impact coefficient I for each grid cell’s abundance change e , the calculation formula is:

[0121] I e =|ΔR HA | / (|ΔR CC |+ΔR HA |) (18)

[0122] The evaluation method is shown in Table 1;

[0123] Step 5.2: Use the habitat colonization potential (HCP) to quantitatively calculate the magnitude and trend of community richness changes; the definition of HCP is shown in Table 2;

[0124] Step 5.3: According to the pressure source influence coefficient I e Identify the main stressors in the area and determine the impact of the stressors in the area based on the HCP.

[0125] Table 1 Pressure source influence coefficient I e Evaluation Method

[0126]

[0127] Table 2 Definition of habitat colonization level

[0128]

[0129]

[0130] ΔR stands for ΔR CC , ΔR HA Or ΔR, the specific range of variation depends on the circumstances.

[0131] like Figure 6 As shown, an embodiment of the present invention provides a method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities, and a system for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities, including:

[0132] The variable set module is established to collect historical data on the presence and absence of target freshwater fish communities, and to collect bioclimate data that affect fish distribution and data related to human activities to establish a variable set;

[0133] The environmental factor screening module is used to screen variables without collinearity using the variance inflation factor (VIF), preliminarily select the machine learning algorithm, build the species distribution model of each organism in the fish community, simulate the distribution of fish communities based on the S-SDM stacked species distribution model (S-SDM), and screen environmental factors with higher contribution to the model;

[0134] The model stacking and prediction module is used to select the machine learning sub-model with superior performance based on the two evaluation indicators of the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS), and realize the stacking and prediction of the ensemble model;

[0135] an impact separation module to separate the effects of climate change, human activities, and other stressors on freshwater fish communities;

[0136] The main stress source and impact assessment module is used to use the stress source impact coefficient (I e ) and habitat colonization potential (HCP) indicators were used to evaluate the main stressors of freshwater fish communities and their impact levels.

[0137] The upper reaches of the Yangtze River ( Figure 2 ) The controlled area is about 1 million km 2 , located at 90°32′~111°33′E, 24°42′~35°55′N, with a total length of 4511km. It is mainly affected by the East Asian monsoon, the South Asian monsoon and the Qinghai-Tibet Plateau terrain, and shows significant regional climate characteristics. Since freshwater species face multiple stressors, it is very difficult to coordinate the protection of multiple species. It is of great significance to formulate effective and targeted community-level protection measures. Due to the concentrated development of water conservancy facilities such as dams and hydropower stations in the upper reaches of the Yangtze River, it also has a high degree of aquatic biodiversity. Fish of the subfamily Schizothorinae are rare fish unique to the upper reaches of the Yangtze River. Therefore, the present invention provides a method for formulating a protection strategy for freshwater fish communities to cope with multiple stressors. Taking 20 species of fish of the subfamily Schizothorinae in the upper reaches of the Yangtze River as an example, a stacked species distribution model SSDM is established to separate the impact of climate change and human water resource development on river network connectivity, and use the stressor impact coefficient I e HCP assesses stressors and develops strategies at the level of habitat colonization. The specific process is as follows:

[0138] In the entire upper reaches of the Yangtze River, historical presence data of 20 species of schizothorax from 2001 to 2020 were collected, and a random method was used to generate missing point data of 10 times the presence points, which were merged into a total data set. Input the merged data set, the modeling factor set Env, and the candidate model set M pThe stacked species distribution model SSDM was debugged, and temperature seasonality, precipitation seasonality, human impact index and river network connectivity coefficient were selected as model prediction variables according to the contribution results. Artificial neural network (ANN), classification tree analysis (CTA), flexible discriminant analysis (FDA) and multivariate adaptive regression splines (MARs) were selected as sub-models to form an ensemble model according to the values ​​of the area under the receiver operating characteristic (ROC) curve (AUC) and the true skill statistic (TSS). The 2050 period of the SSP585 scenario under the ACCESS-CM2 climate model was predicted, and the effects of temperature seasonality and precipitation seasonality were taken as the impacts of climate change ( Figure 3 ) and the influence of river network connectivity coefficient is taken as the influence of human activities ( Figure 4 ) and were classified using the habitat colonization potential HCP. The stressor impact factor I e Assess the main stressors in the upper Yangtze River Basin ( Figure 5 ).

[0139] The contribution results show that the 20 species of Schizothorinae in the upper reaches of the Yangtze River are mainly affected by temperature seasonality, precipitation seasonality, human impact index and river network connectivity coefficient. Figure 3 It shows that the schizothorax subfamily fishes in the Minjiang River basin are most seriously affected by climate change. Figure 4 It shows that the fishes of the subfamily Schizothorinae in the Jialing River basin are most seriously affected by human water resource development. Figure 5 This indicates that, in addition to the Three Gorges Basin, aquatic ecological restoration work in other basins should be focused on restoring the connectivity of the river network.

[0140] The upper reaches of the Yangtze River have always been understood as one of the areas with the most serious water resource development (see figure below), but from the perspective of protecting freshwater fish, especially the schizothoracinae of this example, restoring connectivity in the upper reaches of the Yangtze River may not be effective in restoring fish abundance. Therefore, unnecessary manpower and material resources may be wasted, especially the flood control and power generation benefits brought by cascade hydropower development may be reduced.

[0141] The embodiment of the present invention proposes a method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities, and in the embodiment, it is determined that the main stress source of Schizothorax in the Three Gorges region is climate change rather than water resource development ( Figure 5 ), which can improve the efficiency of the protection of the schizothorax subfamily and avoid the waste of manpower and material resources. Figure 7 This is a diagram showing the changes in river network connectivity in the upper reaches of the Yangtze River around 2000.

[0142] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for identifying the main stressors of freshwater fish communities under the influence of climate change and human activities, characterized in that: The following steps are involved: Step 1: Collect historical data on the presence and absence of target freshwater fish communities, collect bioclimatic data that affect fish distribution and data related to human activities to establish a variable set; Step 2: Use the variance inflation factor (VIF) to screen variables that do not have collinearity, use a machine learning algorithm to build a species distribution model for each organism in the fish community, simulate the distribution of fish communities based on the stacked species distribution model (S-SDM), and screen the environmental factors whose contribution to the model is higher than 10%. The S-SDM is specifically the Stacked Species Distribution Model. Step 3: Based on the two evaluation indicators of the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS), select the machine learning sub-model with superior performance to achieve stacking and prediction of the ensemble model; Step 4: Isolate the impacts of climate change and anthropogenic stressors on freshwater fish communities; Step 5: Use the stressor influence factor I e and habitat colonization potential (HCP) indicators to evaluate the main stressors of freshwater fish communities and their impact levels; The specific steps of step 2 are as follows: Step 2.1: Preliminarily select no less than 5 sub-models from the available machine learning model set M as the candidate model set M p ; Step 2.2: Determine the factors to be screened, including climate and environmental factors Env c 、Human activity related factors Env h , forming the factor data set Env to be screened; Step 2.3: Calculate the variance inflation factor value VIF of each factor in Env, and remove environmental factors whose VIF magnitude is higher than that of other factors; the calculation formula of the variance inflation factor value VIF is as follows: In the formula, j---the jth environmental factor; ---The sum of the squares of the relative changes of this environmental factor and all other environmental factors; Step 2.4: Use the remaining environmental factors to simulate the model according to the ensemble model set M p Contribution results, remove the variables with an average contribution of less than 10% of the freshwater fish community as the modeling factor data set Env s ; The specific steps of step 3 are as follows: Step 3.1: Selected model set M p Each model was run ten times; Step 3.2: Randomly select 70% of the total presence and absence data as the training set to fit the algorithm, and the remaining 30% is reserved for evaluating the algorithm performance; Step 3.3: Evaluate on the evaluation dataset using the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS); Step 3.4: When the sub-model set M is determined, the collective model of each species is stacked to obtain the species distribution model S-SDM at the community level; Step 3.5: Use S-SDM to simulate and predict different scenarios {C, H} of the target community in the historical period and future period t to obtain the community richness under climate mode i The evaluation method in step 3.3 includes: averaging the AUC and TSS of all sub-models to obtain the AUC of the ensemble model e and TSS e , if AUC e >0.85,TSS e >0.75, the ensemble model is considered to have superior performance, and the selected sub-model set is the final sub-model set M s Otherwise, reselect the submodel set M based on the AUC and TSS of the single model p ; The calculation formula of TSS is as follows: TSS=Specificity+Sensitivity-1 In the formula, Specificity is the model specificity, and Sensitivity is the model sensitivity; The specific steps of step 5 are as follows: Step 5.1: Calculate the stressor impact coefficient I for each grid cell’s abundance change e , the calculation formula is: I e =|ΔR HA | / (|ΔR CC |+|ΔR HA |) In the formula, ΔR HA The difference between the impact of human activities on ST abundance using the model simulation for the historical period and the simulation with the model projected onto the human activity factor for the historical period; ΔR CC The difference between the prediction of the historical model and the simulation of the historical model is used to represent the impact of climate change on community richness; Step 5.2: Use the habitat colonization potential (HCP) to quantitatively calculate the magnitude and trend of community richness changes; Step 5.3: According to the pressure source influence coefficient I e Identify the main stressors in the area and determine the impact of the stressors in the area based on the HCP.

2. The method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities as claimed in claim 1, characterized in that: The specific steps of step 4 are as follows: Step 4.1: Average the values ​​from n climate models to get a representative future forecast The calculation formula is as follows: In the formula, i = 1, 2, ... n represents n different climate modes; C represents the climate scenario to which the climate factor is predicted by the model; t represents the corresponding period of the climate factor predicted by the model, which is 2030, 2050, 2070 or 2090; H represents the human activity factor for different situations predicted by the model; Step 4.2: Isolate the effects of climate factors on community richness; the impact of climate change on community richness is measured by the difference between the model prediction and the model simulation for the historical period, expressed as ΔR CC , the calculation formula is as follows: ΔR CC =R C,t -R In the formula, R represents the richness of the community in the historical period; Step 4.3: Isolate the impact of human factors on community richness; the impact of human activities on ST richness is measured by the difference between the model simulation of the historical period and the simulation of the historical period projected onto the human activity factor, expressed as ΔR HA , the calculation formula is as follows: ΔR HA =R H -R Step 4.4: Calculate the combined effects of climate change and human activities on community richness; the combined effect is the combined effect of climate change and human activities on community richness, expressed as ΔR, and the calculation formula is as follows:

3. A method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities as claimed in any one of claims 1 to 2, and a system for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities, characterized in that: include: The variable set module was established to collect historical data on the presence and absence of target freshwater fish communities, and to collect bioclimate data that affect fish distribution and data related to human activities to establish a variable set; Environmental factor screening module, used to screen variables without collinearity using variance inflation factor (VIF), preliminarily select machine learning algorithms, build species distribution models for each organism in fish communities, simulate fish community distribution based on S-SDM stacking species distribution model (S-SDM), and screen environmental factors with high contribution to the model; The model stacking and prediction module is used to select the machine learning sub-model with superior performance based on the two evaluation indicators of the area under the receiver operating characteristic curve (AUC) and the true skill statistic (TSS), and to achieve the stacking and prediction of the ensemble model; an impact separation module to separate the effects of climate change, human activities, and other stressors on freshwater fish communities; The main stress source and impact assessment module is used to use the stress source impact coefficient I e The HCP index and habitat colonization potential were used to evaluate the main stressors of freshwater fish communities and their impact levels.

4. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities as described in any one of claims 1 to 2.

5. A computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the method for identifying the main stress sources of freshwater fish communities under the influence of climate change and human activities as described in any one of claims 1 to 2.

6. An information data processing terminal, characterized in that: The information data processing terminal includes the identification system for the main stress sources of freshwater fish communities under the influence of climate change and human activities as described in claim 3.

Images