Method for predicting seawater quality pollution degree based on sensitive phytoplankton

Abstract

The invention relates to the technical field of regional marine environment evaluation, in particular to a method for predicting the seawater quality pollution degree based on sensitive phytoplankton. Comprising the following steps: collecting a water body sample of surface seawater in an offshore area; dNA in the water body sample is filtered and extracted, and high-throughput sequencing analysis is carried out after PCR amplification; performing species classification annotation on each sequence, and constructing a data matrix according to obtained phytoplankton OUT data; calculating a Spearman correlation coefficient; constructing an inter-species correlation matrix, and selecting proper species to construct a co-occurrence network; exporting edge data and point data, and screening nodes with the maximum betweenness centrality or the highest tight centrality as key species; calculating an organic pollution index (OPI) and carrying out Spearman correlation analysis to determine OPI sensitive phytoplankton; and establishing a prediction model for prediction to obtain a predicted environmental pollution index. The method has the advantages that model complexity and noise influence are reduced; the prediction accuracy reaches 100% when the water quality is seriously polluted.

Description

Technical Field

[0001] The invention relates to the technical field of regional marine environment assessment, and in particular to a method for predicting the degree of seawater pollution based on sensitive phytoplankton. Background Art

[0002] In recent years, the issue of land-based sewage discharge into the sea and the research on its impact on the marine ecological environment have become issues that cannot be ignored worldwide. The nutrients in the sewage from the sewage outlet will affect the water quality of the sea, and then affect the ecological environment of the nearshore waters and even the offshore, posing a threat to the health of the marine ecology. Due to the large amount of land-based organic pollutants discharged into the nearshore waters, the changes in the structure of phytoplankton communities and the frequent occurrence of red tides in local seas, which in turn caused the collapse of the nearshore marine ecosystem, have gradually attracted widespread attention.

[0003] Marine phytoplankton is the most important primary producer in the ocean and the basis for the material cycle and energy flow of the marine ecosystem. The composition of phytoplankton communities adjusts with habitat changes. Studies have shown that the use of phytoplankton community structure, diversity and functional groups in waters can effectively judge the quality of the water environment. Compared with traditional monitoring methods, biological indicators can not only characterize the instantaneous state of water environment quality, but also reflect the potential hazards in the water environment. The organic pollution index (OPI) is an important indicator for assessing the degree of organic pollution in water bodies. It is an indicator for comprehensive evaluation, including chemical oxygen demand (COD), inorganic nitrogen (DIN), phosphate (DIP) and dissolved oxygen (DO). According to the OPI value, the degree of seawater pollution is divided into six categories: <0 means good water quality, 0-1 means good, 1-2 means water quality begins to be polluted, 2-3 means light pollution, 3-4 means moderate pollution, and >4 means severe pollution.

[0004] Indicator organisms refer to a type of organism that has a specific indicative effect on a specific environment or a certain biological group, including organisms sensitive to specific pollutants, and organisms with dose associations in water pollution, soil pollution, or air pollution. Indicator organisms show dynamic changes in growth or abundance as the polluted environment changes, and can show specific responses to polluted environmental elements. The difference in phytoplankton's tolerance to the environment can play the role of environmental indicator and can be used as an indicator organism for measuring water quality. Therefore, OPI-sensitive phytoplankton can be used to indicate the level of offshore organic pollution. At present, studies have reported that bacterioplankton can be used as an indicator organism for organic pollution in seawater, but there are no reports on sensitive phytoplankton indicators of organic pollution in seawater, mainly due to the lack of screening methods for OPI-sensitive bacterioplankton in seawater.

[0005] By establishing a prediction model for OPI sensitive phytoplankton and seawater pollution index, it can help assess the health of seawater ecology and promptly discover potential pollution problems so that preventive measures can be taken before pollution occurs. Therefore, a method for predicting the degree of seawater pollution based on sensitive phytoplankton is developed, which is suitable for water quality evaluation in different environmental water bodies and can provide a reliable basis for water body monitoring and governance.

[0006] Existing studies have shown that planktonic bacteria can serve as important indicator organisms for organic pollution in seawater. By monitoring changes in their abundance and population structure, they can effectively reflect the presence of organic pollutants in the water and changes in their concentrations. This method provides new ideas and means for marine water quality monitoring, and can detect and respond to pollution incidents more promptly. At present, planktonic bacteria have made certain progress in indicating organic pollution, but there have been no relevant research reports on sensitive phytoplankton indicators of organic pollution in seawater. This phenomenon shows the limitations of existing technologies. Among the currently reported methods for screening OPI-sensitive planktonic bacteria, Pearson correlation analysis is directly used to select populations whose abundance is significantly correlated with organic pollution. Selecting key populations only by correlation may ignore more complex interactions and mechanisms in ecology and environmental science, and lack a comprehensive understanding of the ecosystem; the "labdsv" package in R software is used to screen sensitive planktonic bacterial communities indicating different levels of organic pollution, and the significant differences between the selected indicator groups and organic pollution are further analyzed by Pearson correlation. However, the functions and parameter settings of the "labdsv" package may be relatively complex, and there may be certain difficulties in learning and using them; the "randomForest" package in R software is used to fit the relative abundance of OTUs to their corresponding OPI values, and the order of importance is sorted. The training and prediction of the random forest model may require larger computing resources, especially for large-scale data sets, which may result in long computing time and high memory usage.

[0007] Through further research, it is expected that some specific phytoplankton species will be found to be highly sensitive to organic pollution and can serve as new indicators. This will not only enrich our understanding of the marine ecological environment, but also provide more comprehensive and diverse tools for marine pollution monitoring, helping to better protect the health and sustainable development of the marine ecosystem. Summary of the invention

[0008] In order to solve the above problems, the present invention provides a method for predicting the degree of seawater pollution based on sensitive phytoplankton.

[0009] The present invention aims to provide a method for predicting the degree of seawater pollution based on sensitive phytoplankton, which specifically comprises the following steps: S1. Collect surface seawater samples in coastal waters; S2. Filter the water samples; extract DNA from the water samples after filtration, perform high-throughput sequencing analysis after PCR amplification; then perform species classification annotation on each sequence, and construct a data matrix based on the obtained phytoplankton operational classification unit data; S3. Calculate the Spearman correlation coefficient through the data matrix; construct the inter-species correlation matrix and select appropriate species to construct the co-occurrence network; S4. Based on the co-occurrence network constructed in step S3, the edge data and point data are exported, and the betweenness centrality and closeness centrality of each node are calculated; the nodes with the largest betweenness centrality or the highest closeness centrality are selected as key species; S5. Evaluate the pollution status of the sea area and calculate the organic pollution index OPI; S6. Perform Spearman correlation analysis between the selected key species and OPI to determine OPI-sensitive phytoplankton; S7. A prediction model is established with the relative abundance of the screened OPI-sensitive phytoplankton as the independent variable and the degree of water pollution at the sampling site as the dependent variable. The prediction model is used to make predictions to obtain the predicted environmental pollution index.

[0010] Preferably, in step S2, PCR amplification is performed using the V4 region of 18S rRNA; BLAST is used to annotate species classification of each sequence in the NCBI database; and the filtration is performed using a suction filtration method, and the filtration is performed using a 0.22 μm microporous filter membrane.

[0011] Preferably, in step S3, p<0.01 is selected and >0.7 to construct a co-occurrence network.

[0012] Preferably, the betweenness centrality calculation formula is as follows:

[0013] In the formula, represents the betweenness centrality, represents the number of paths that pass through node i and are the shortest paths; represents the number of shortest paths linking nodes s and t; The closeness centrality calculation formula is as follows:

[0014] represents close centrality, Represents the average distance from node i to the rest of the points; The formula is as follows:

[0015] In the formula, N represents the number of nodes, Represents the distance between node i and node j.

[0016] Preferably, in step S5, the organic pollution comprehensive index method is used to calculate the organic pollution index OPI according to the measured environmental factors chemical oxygen demand, inorganic nitrogen, inorganic phosphorus and dissolved oxygen; OPI calculation formula:

[0017] Where, OPI is the organic pollution index; To actually measure the chemical oxygen demand of seawater; To actually measure inorganic nitrogen in seawater; To actually measure inorganic phosphorus in seawater; To actually measure the dissolved oxygen in seawater; , , D and They are the standard value of chemical oxygen demand of seawater, the standard value of inorganic nitrogen of seawater, the standard value of inorganic phosphorus of seawater and the standard value of dissolved oxygen of seawater.

[0018] Preferably, a nonlinear kernel support vector machine method is used to establish a prediction model, and the prediction model can be expressed as:

[0019] In the formula, is the predicted environmental pollution index; is the number of support vectors; Each support vector The weights are obtained through the training process; is the kernel function, which represents the input vector With support vector The similarity between .

[0020] Preferably, the kernel function is a radial basis function kernel, and the expression is as follows:

[0021] In the formula, is the input vector With support vector The squared Euclidean distance, is a free parameter.

[0022] Preferably, in step S7, the predicted environmental pollution index is compared with the seawater pollution level indicated by the OPI environmental index obtained in step S5 to obtain the accuracy of the prediction model.

[0023] Compared with the prior art, the present invention can achieve the following beneficial effects: The present invention combines the centrality index of important nodes in complex networks to improve the method of screening sensitive phytoplankton, which can effectively reduce the computational complexity and time complexity. The time complexity of this method is O(n)+O(n+m). Compared with the prior art, the method of the present invention can significantly reduce the time complexity of the algorithm, and compared with the Pearson correlation method used alone, after introducing two centrality indexes, it can more comprehensively evaluate the importance of nodes and reduce the complexity and noise impact of the model. Compared with the currently reported quantitative prediction method of water pollution degree, the accuracy of the present invention can reach 100% when the water quality is seriously polluted. Moreover, on a small sample data set, the support vector machine (SVM) model is more effective and can reduce the complexity and noise impact of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 The present invention provides a flowchart of a method for predicting the degree of seawater pollution based on sensitive phytoplankton.

[0025] Figure 2 This is a phytoplankton co-occurrence network diagram provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the following description, the same modules are represented by the same reference numerals. In the case of the same reference numerals, their names and functions are also the same. Therefore, the detailed description thereof will not be repeated.

[0027] 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 accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and do not constitute a limitation of the present invention.

[0028] The present invention provides a method for predicting the degree of seawater pollution based on sensitive phytoplankton, which specifically comprises the following steps: S1. Collect surface seawater samples in coastal waters; S2. Filter the water samples; extract DNA from the water samples after filtration, perform high-throughput sequencing analysis after PCR amplification; then perform species classification annotation on each sequence, and construct a data matrix based on the obtained phytoplankton operational taxonomic unit (OUT) data; Specifically, PCR amplification was performed using the V4 region of 18S rRNA; BLAST was used to annotate species classification of each sequence in the NCBI database; the filtration was performed using a suction filtration method, and the filtration was performed using a 0.22 μm microporous filter membrane, and the filtered membrane was used for eDNA macrobarcode sequencing to analyze the composition of phytoplankton communities; S3. Calculate the Spearman correlation coefficient (Spearman coefficient) through the data matrix; construct the inter-species correlation matrix and select p < 0.01 and >0.7 species to construct a co-occurrence network; S4. Based on the co-occurrence network constructed in step S3, the edge data and point data are exported, and the betweenness centrality and closeness centrality of each node are calculated; the nodes with the largest betweenness centrality or the highest closeness centrality are selected as key species.

[0029] The calculation formula of betweenness centrality is as follows:

[0030] In the formula, represents the betweenness centrality, represents the number of paths that pass through node i and are the shortest paths; Represents the number of shortest paths linking nodes s and t.

[0031] The formula for calculating closeness centrality is as follows:

[0032] represents close centrality, Represents the average distance from node i to the rest of the points; The formula is as follows:

[0033] In the formula, N represents the number of nodes, Represents the distance between node i and node j.

[0034] S5. The pollution status of the sea area was assessed using the organic pollution comprehensive index method, and the organic pollution index (OPI) was calculated based on the measured environmental factors chemical oxygen demand (COD), inorganic nitrogen (DIN), inorganic phosphorus (DIP), and dissolved oxygen (DO).

[0035] OPI calculation formula:

[0036] Where, OPI is the organic pollution index; To actually measure the chemical oxygen demand of seawater; To actually measure inorganic nitrogen in seawater; To actually measure inorganic phosphorus in seawater; To actually measure the dissolved oxygen in seawater; , , D and They are the standard value of chemical oxygen demand of seawater, the standard value of inorganic nitrogen of seawater, the standard value of inorganic phosphorus of seawater and the standard value of dissolved oxygen of seawater; In a specific embodiment, 2mg / L; 0.2mg / L; D 0.015mg / L; It is 6mg / L.

[0037] The degree of seawater pollution is divided into six categories according to the OPI value: <0 means good water quality, 0-1 means better, 1-2 means the water quality is beginning to be polluted, 2-3 means light pollution, 3-4 means moderate pollution, and >4 means severe pollution.

[0038] S6. Perform Spearman correlation analysis between the selected key species and OPI to determine OPI-sensitive phytoplankton.

[0039] S7. Using the relative abundance of the selected OPI-sensitive phytoplankton as the independent variable and the water pollution degree at the sampling site as the dependent variable, a prediction model is established; prediction is performed through the prediction model to obtain a predicted environmental pollution index; In a specific embodiment, a prediction model is established using a support vector machine method with a nonlinear kernel (such as radial basis function RBF). The prediction model can be expressed as:

[0040] In the formula, is the predicted environmental pollution index; is the number of support vectors; Each support vector The weights of , which are obtained through the training process; is the kernel function, which represents the input vector With support vector The similarity between .

[0041] The kernel function is the radial basis function (RBF):

[0042] In the formula, is the input vector With support vector The squared Euclidean distance, is a free parameter.

[0043] In this step, the predicted environmental pollution index is compared with the seawater pollution level indicated by the OPI environmental index obtained in step S5 to obtain the accuracy of the prediction model.

[0044] Example 1 This embodiment provides a method for predicting the degree of seawater pollution based on sensitive phytoplankton (see flowchart for details) Figure 1 ), specifically including the following steps: S1. Collect surface seawater samples in the coastal waters of Dalian, a total of 11 samples; collect 3L of surface water at each preset sampling station, and immediately transport the collected samples back to the laboratory within 24 hours through cold chain transportation for follow-up research; 1L of water samples was used for the determination of environmental physical and chemical indicators. Use a portable water quality analyzer to measure the dissolved oxygen (DO) in the surface seawater on-site, and determine the chemical oxygen demand (COD), inorganic nitrogen (DIN) and inorganic phosphorus (DIP) in seawater in accordance with GB17378-2007 marine monitoring specifications.

[0045] S2. Take 2L of collected water sample and extract DNA after vacuum filtration. After PCR amplification, use 18S rRNA high-throughput sequencing analysis, use BLAST to annotate species classification for each sequence in the NCBI database, and construct a data matrix based on the obtained phytoplankton OTU data; filter through a 0.22μm microporous membrane during filtration, and use the membrane for eDNA macrobarcode sequencing to analyze the phytoplankton community composition.

[0046] S3. Import the obtained data matrix into Rstudio software and calculate the spearman coefficient; construct the species correlation matrix and construct the co-occurrence network according to the "Hmisc" package; screen for p<0.01 and Species (nodes) with >0.7 eventually construct a co-occurrence network. Among them, the p-value is the significance level, a parameter used to determine the results of hypothesis testing. Usually, a p-value less than 0.05 is used as a criterion for judging whether the correlation is significant. The smaller the p-value, the more significant the result; the R value (also known as the correlation coefficient) is an indicator to measure the correlation between two variables. Its value range is between -1 and 1, where 0 means no correlation, 1 means a completely positive correlation, and -1 means a completely negative correlation.

[0047] See also Figure 2The figure shows the phytoplankton co-occurrence network. The node colors represent phytoplankton at different levels. The node size represents the degree of connectivity between species. The thickness of the line between nodes represents the correlation between nodes. The thicker the line, the stronger the correlation between the two nodes. The more lines connecting the nodes, the more critical the node is in the community. In the figure, Diatomea represents Bacillariophyta, Chlorophyta represents Chlorophyta, and Streptophyta represents Streptophyta.

[0048] S4. Based on the co-occurrence network constructed in step S3, the edge data and point data are exported, and the betweenness centrality and closeness centrality of each node are calculated; the nodes with the largest betweenness centrality or the highest closeness centrality are selected as key species.

[0049] The key species selected by close centrality are Syndiniales (OTU140), the key species selected by betweenness centrality is Thalassiosira salina ( Thalassiosira ,OTU1094).

[0050] S5. The pollution status of the sea area was assessed using the organic pollution comprehensive index method, and the organic pollution index (OPI) was calculated based on the measured environmental factors chemical oxygen demand (COD), inorganic nitrogen (DIN), inorganic phosphorus (DIP), and dissolved oxygen (DO).

[0051] OPI calculation formula:

[0052] Where, OPI is the organic pollution index; To actually measure the chemical oxygen demand of seawater; To actually measure inorganic nitrogen in seawater; To actually measure inorganic phosphorus in seawater; To actually measure the dissolved oxygen in seawater; , , D and They are the standard values ​​of chemical oxygen demand, inorganic nitrogen, inorganic phosphorus and dissolved oxygen in seawater, respectively; they are quoted from the Seawater Quality Standard (GB 3097-1997), and the values ​​are the first category seawater quality standards in the standard. 2mg / L; 0.2mg / L; D 0.015mg / L; It is 6mg / L.

[0053] The specific data of the 11 groups of samples are shown in Table 1.

[0054] Table 1 Environmental factors and OPI data

[0055] The degree of seawater pollution is divided into six categories according to the OPI value: <0 means good water quality, 0-1 means good water quality, 1-2 means the water quality is beginning to be polluted, 2-3 means light pollution, 3-4 means moderate pollution, and >4 means severe pollution.

[0056] S6. Perform Spearman correlation analysis between the selected key species and OPI to determine the OPI-sensitive phytoplankton. Syndiniales (p<0.01)(r=-0.791) can be used as a sensitive phytoplankton indicating organic pollution in seawater.

[0057] S7. Using the relative abundance of the selected OPI-sensitive phytoplankton as the independent variable and the water pollution level at the sampling site as the dependent variable, a prediction model was established for prediction; The model is established using the support vector machine method with a nonlinear kernel (such as radial basis function RBF). The model can be expressed as:

[0058] In the formula, is the predicted environmental pollution index; is the number of support vectors; Each support vector The weights of , which are obtained through the training process; is the kernel function, which represents the input vector With support vector The similarity between .

[0059] One of the commonly used kernel functions is the radial basis function kernel (RBF):

[0060] in, It can be regarded as the squared Euclidean distance between two eigenvectors. is a free parameter.

[0061] The prediction results are shown in Table 2.

[0062] Table 2 Prediction results based on support vector machine model

[0063] In Table 2, S1~S11 represent the sampling stations in summer, and W1~W11 represent the sampling stations in winter. It can be seen from Table 2 that the measured and predicted pollution indexes of the sites with serious water pollution are the same.

[0064] In this step, the predicted environmental pollution index is compared with the seawater pollution level indicated by the OPI environmental index obtained in step S5 to obtain the accuracy of the prediction model (Table 3).

[0065] Table 3 Accuracy of predicted pollution index

[0066] It was found that the model accuracy reached 100% when the water quality was severely polluted.

[0067] The prediction method of the present invention has been verified at 22 sampling points in the coastal waters of Dalian, and the prediction accuracy of the model reaches 100% when the water quality is seriously polluted.

[0068] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the disclosure of the present invention can be performed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and this document does not limit this.

[0069] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting the degree of seawater pollution based on sensitive phytoplankton, characterized in that: The specific steps include: S1. Collect surface seawater samples in coastal waters; S2. Filter the water samples; extract DNA from the water samples after filtration, perform high-throughput sequencing analysis after PCR amplification; then perform species classification annotation on each sequence, and construct a data matrix based on the obtained phytoplankton operational classification unit data; S3. Calculate the Spearman correlation coefficient through the data matrix; construct the inter-species correlation matrix and select appropriate species to construct the co-occurrence network; S4. Based on the co-occurrence network constructed in step S3, the edge data and point data are exported, and the betweenness centrality and closeness centrality of each node are calculated; the nodes with the largest betweenness centrality or the highest closeness centrality are selected as key species; S5. Evaluate the pollution status of the sea area and calculate the organic pollution index OPI; S6. Perform Spearman correlation analysis between the selected key species and OPI to determine OPI-sensitive phytoplankton; S7. A prediction model is established with the relative abundance of the screened OPI-sensitive phytoplankton as the independent variable and the degree of water pollution at the sampling site as the dependent variable. The prediction model is used to make predictions to obtain the predicted environmental pollution index.

2. The method for predicting the degree of seawater pollution based on sensitive phytoplankton according to claim 1, characterized in that: In the step S2, PCR amplification is performed using the V4 region of 18S rRNA; BLAST is used to annotate species classification of each sequence in the NCBI database; and the filtration is performed using a suction filtration method, and the filtration is performed using a 0.22 μm microporous filter membrane.

3. The method for predicting the degree of seawater pollution based on sensitive phytoplankton according to claim 1, characterized in that: In step S3, p<0.01 is selected and >0.7 to construct a co-occurrence network.

4. The method for predicting the degree of seawater pollution based on sensitive phytoplankton according to claim 1, characterized in that: The betweenness centrality calculation formula is as follows: In the formula, represents the betweenness centrality, represents the number of paths that pass through node i and are the shortest paths; represents the number of shortest paths linking nodes s and t; The closeness centrality calculation formula is as follows: represents close centrality, Represents the average distance from node i to the rest of the points; The formula is as follows: In the formula, N represents the number of nodes, Represents the distance between node i and node j.

5. The method for predicting the degree of seawater pollution based on sensitive phytoplankton according to claim 1, characterized in that: In step S5, the organic pollution comprehensive index method is used to calculate the organic pollution index OPI according to the measured environmental factors chemical oxygen demand, inorganic nitrogen, inorganic phosphorus and dissolved oxygen; OPI calculation formula: Where, OPI is the organic pollution index; To actually measure the chemical oxygen demand of seawater; To actually measure inorganic nitrogen in seawater; To actually measure inorganic phosphorus in seawater; To actually measure the dissolved oxygen in seawater; , , D and They are the standard value of chemical oxygen demand of seawater, the standard value of inorganic nitrogen of seawater, the standard value of inorganic phosphorus of seawater and the standard value of dissolved oxygen of seawater.

6. The method for predicting the degree of seawater pollution based on sensitive phytoplankton according to claim 1, characterized in that: The prediction model is established using the support vector machine method with a nonlinear kernel. The prediction model can be expressed as: In the formula, is the predicted environmental pollution index; is the number of support vectors; Each support vector The weights are obtained through the training process; is the kernel function, which represents the input vector With support vector The similarity between .

7. The method for predicting the degree of seawater pollution based on sensitive phytoplankton according to claim 6, characterized in that: The kernel function is a radial basis function kernel, and the expression is as follows: In the formula, is the input vector With support vector The squared Euclidean distance, is a free parameter.

8. The method for predicting the degree of seawater pollution based on sensitive phytoplankton according to claim 1, characterized in that: In step S7, the predicted environmental pollution index is compared with the seawater pollution level indicated by the OPI environmental index obtained in step S5 to obtain the accuracy of the prediction model.

Images