Microbial indicator species screening method supporting water resource scheduling decision

Abstract

The invention relates to a microbial indicator species screening method for supporting a water resource scheduling decision, which is used for screening representative microbial indicators by combining an environmental DNA (eDNA) technology and monitoring the dynamic change of the microbial indicators so as to monitor and evaluate the diffusion range and process of the indicators and provide comprehensive support for the water resource scheduling decision. The blank of related methods is filled, and the ecological benefit, the environmental benefit and the economic benefit of water resource scheduling are comprehensively improved.

Description

Technical Field

[0001] The present invention relates to a method for screening microbial indicator species to support water resources scheduling decisions, belonging to the technical field of environmental monitoring and ecological protection. Background Art

[0002] Inter-basin water resources scheduling is a necessary measure to ensure the water supply safety of water-scarce areas, such as "diverting the Yangtze River water to Taihu Lake", "diverting the Yangtze River water to Chaohu Lake", "diverting the Yangtze River water to Huaihe River", etc. However, all along, there has been a lack of effective decision-making methods to intuitively monitor and evaluate the process and influence scope of water resources scheduling. The tracer method based on natural ions provides a useful reference for evaluating the process of water resources scheduling. However, since the selected natural ions generally exist in both the "raw water" and the "receiving water area" during water diversion, in the later stage of water resources scheduling, it is impossible to accurately display the diffusion process of the diverted raw water in the receiving water area.

[0003] In recent years, the application of environmental DNA (eDNA) technology has provided a new solution for microbial monitoring in water environments. The eDNA technology can non-invasively capture the species and community structure of microorganisms in water bodies by analyzing the genetic material in water samples, especially for those microorganisms that are difficult to culture, and the eDNA technology provides a more accurate detection method. Through the eDNA technology, with the help of methods such as machine learning, the microbial indicator species in the "raw water" and the receiving water area during water diversion are statistically analyzed. This indicator species is different from natural ions. Before water resources scheduling, this indicator species only exists in the raw water. As the water diversion process progresses, this indicator species gradually spreads with the flow of water in the receiving water area. By monitoring the information of microbial indicator species during the water diversion process, the diffusion range and process of the diverted raw water in the receiving water area can be comprehensively, accurately and efficiently monitored and evaluated, providing comprehensive support for water resources scheduling decisions, thereby improving the ecological, environmental and economic benefits of water resources scheduling. Summary of the Invention

[0004] In order to solve the above existing problems, the present invention discloses a method for screening microbial indicator species to support water resources scheduling decisions, and its specific technical solutions are as follows:

[0005] A method for screening microbial indicator species to support water resources scheduling decisions includes the following steps:

[0006] Step 1: Preliminary screening of microbial indicator species

[0007] Before water resources scheduling, environmental DNA analysis is carried out on the microbial communities in the "raw water" and "receiving water area" water samples. Through non-invasive water sample collection, PCR amplification of the target gene region, high-throughput sequencing, and denoising, clustering and classification of the original sequences, operational taxonomic units (OTUs) are generated to determine the composition of the microbial community;

[0008] Step 2: Investigation of the background value of the diffusion range of the diverted raw water

[0009] During the water resources scheduling process, the chloride ion tracing method is adopted to monitor the diffusion process after the "raw water" flows into the "water receiving area". As a tracer element, chloride ions can effectively track the source and flow path of water bodies; by analyzing the change of chloride ion concentration in water bodies, the diffusion range of the "raw water" source in the "water receiving area" and its mixing situation with the local water bodies in the "water receiving area" are revealed;

[0010] Step 3: Confirmation and verification of microbial indicator species

[0011] During Step 2, environmental DNA samples are collected synchronously. Several sampling points are selected to cover the entire "water receiving area". By analyzing the changes in the microbial community characteristics in the environmental DNA, the microbial indicator species preliminarily screened in Step 1, combined with the temporal and spatial changes of chloride ions in Step 2, are divided into several groups of microbial community data and analyzed by the random forest algorithm again to obtain the final microbial indicator species; the finally obtained indicator species are the microorganisms that are unique or significantly enriched after the introduction of the "raw water", and these microorganisms did not exist in the "water receiving area" before the water diversion, which characterizes the mixing degree of the "raw water" and the "water receiving area", helps to determine the species that continuously exist and are stable in the "water receiving area", and thus represents the successful integration of river microbial species;

[0012] Step 4: Application of water resources scheduling decision-making

[0013] Finally, the selected biomarkers can be used to predict the diffusion range and water quality changes of the "raw water" in the "water receiving area", providing data support for ecological risk assessment.

[0014] Furthermore, Step 1 is specifically as follows: Screen the microbial species in the water samples of the "raw water" and the "water receiving area", divide the "raw water" and the "water receiving area" into two groups, namely the "raw water group" and the "water receiving area group". Before the water diversion, the microbial community data of the "raw water group" and the "water receiving area group" are analyzed by the random forest algorithm to screen out a group of representative microbial indicator species; by calculating the AUC value, the prediction accuracy of the biomarker for the change of the microbial community under different water area conditions is evaluated; the prediction process can evaluate the stability and accuracy of the random forest algorithm model under different conditions, ensure the reliability of the selected biomarker in practical applications, and the screened microbial indicator species can reflect the differences in the microbial communities of the "raw water" and the "water receiving area", that is, the microorganisms that exist in the "raw water" before the water diversion but do not exist in the "water receiving area".

[0015] Further, the specific analysis of the random forest algorithm in step 1 is as follows: The parameters of the random forest algorithm model are set as the number of classification trees being 500 - 1000, and the number of features selected during each split is the square root of the total number of features; high-discrimination species are selected through AUC verification, with the AUC value ≥ 0.7, and the selected indicator species include microorganisms of the genera g__Vogesella, g__Curvibacter, and g__Anabaena_XPORK15F.

[0016] Further, the specific chloride ion tracing in step 2 is as follows: First, by measuring the change in chloride ion concentration in water samples from different regions of the "water receiving area", the diffusion area of the "raw water" flow is located, and its propagation speed and influence range are speculated. For this purpose, water sample collection will cover different water depths and water area positions, and combined with environmental factors such as flow velocity and water temperature, a water flow diffusion map is drawn, thereby speculating the diffusion direction and speed of the "raw water" source in the "water receiving area".

[0017] Further, the dynamic time synchronization mechanism is included when verifying in 8 groups in step 3: When the monitoring result of the chloride ion concentration fluctuates by ≥ 15%, the eDNA in the corresponding area is analyzed keyly; the correlation coefficient between the microorganism data and the temporal change of the chloride ion concentration reaches above 0.7, and the reads value of the indicator species in the area where the chloride ion drops rapidly needs to increase by ≥ 5 times to be confirmed.

[0018] Further, in step 3, the microbial community data is divided into 8 groups. In the early stage of water transfer, the "water receiving area group" is the 1st group and the "raw water group" is the 2nd group. In the middle stage of water transfer, the "water receiving area group" that changes synchronously with the chloride ion is the 3rd group, the "water receiving area group" that has nothing to do with the chloride ion is the 4th group, and the "raw water group" is the 5th group. In the late stage of water transfer, the "water receiving area group" that changes synchronously with the chloride ion is the 6th group, the "water receiving area group" that has nothing to do with the chloride ion is the 7th group, and the "raw water group" is the 8th group.

[0019] The beneficial effects of the present invention are:

[0020] The method for screening indicator species based on environmental DNA (eDNA) technology of the present invention is used to monitor and track the diffusion range during the water diversion process from "raw water" to the "water receiving area". This method can be widely applied to water resource scheduling, large-scale ecological projects, and water body health monitoring, providing effective data support and ecological risk assessment.

[0021] The present invention combines environmental DNA (eDNA) technology to screen representative microbial indicator species (biomarkers), and monitors the dynamic changes of the microbial indicator species to monitor and evaluate the diffusion range and process of the indicator species, providing comprehensive support for water resource scheduling decisions, making up for the gaps in related methods, and comprehensively improving the ecological, environmental, and economic benefits of water resource scheduling. Description of the Drawings

[0022] Figure 1 is the background value of chloride ions and microbial indicator species before water diversion in the embodiments of the present invention,

[0023] wherein, Figure 1-1 is the background value distribution map of tracer ions before water diversion,

[0024] Figure 1-2 is the background value distribution map of g__0ttowia before water diversion,

[0025] Figure 1-3 is the background value distribution map of g__Solitalea before water diversion,

[0026] Figure 1 -4 is the background value distribution map of g__norank_o__SM1A07 before water diversion,

[0027] Figure 1 -5 is the background value distribution map of g__Anabaena_XP0RK15F before water diversion,

[0028] Figure 1 -6 is the background value distribution map of g__Curvibacter before water diversion,

[0029] Figure 1 -7 is the background value distribution map of g__Vogesella before water diversion;

[0030] Figure 2 is the abundance distribution of chloride ions and microbial indicator species during the early stage of water diversion in the embodiments of the present invention,

[0031] wherein, Figure 2-1 is the abundance distribution map of tracer ions during the early stage of water diversion,

[0032] Figure 2-2 is the abundance distribution map of g__0ttowia during the early stage of water diversion,

[0033] Figure 2-3 is the abundance distribution map of g__Solitalea during the early stage of water diversion,

[0034] Figure 2 -4 is the abundance distribution map of g__norank_o__SM1A07 during the early stage of water diversion,

[0035] Figure 2 -5 is the abundance distribution map of g__Anabaena_XP0RK15F during the early stage of water diversion,

[0036] Figure 2-6 is the abundance distribution map of g__Curvibacter during the early stage of water transfer,

[0037] Figure 2 -7 is the abundance distribution map of g__Vogesella during the early stage of water transfer;

[0038] Figure 3 It is the abundance distribution of chloride ions and microbial indicator species during the middle and late stages of water transfer in the embodiments of the present invention,

[0039] Among them, Figure 3-1 is the abundance distribution map of tracer ions during the middle and late stages of water transfer,

[0040] Figure 3-2 is the abundance distribution map of g__0ttowia during the middle and late stages of water transfer,

[0041] Figure 3-3 is the abundance distribution map of g__Solitalea during the middle and late stages of water transfer,

[0042] Figure 3 -4 is the abundance distribution map of g__norank_o__SM1A07 during the middle and late stages of water transfer,

[0043] Figure 3 -5 is the abundance distribution map of g__Anabaena_XP0RK15F during the middle and late stages of water transfer,

[0044] Figure 3 -6 is the abundance distribution map of g__Curvibacter during the middle and late stages of water transfer,

[0045] Figure 3 -7 is the abundance distribution map of g__Vogesella during the middle and late stages of water transfer.

[0046] Figure 1-3 The points on it are the data collection points, and the letter combinations are the numbers of the collection points, without any other meaning.

[0047] g__0ttowia, g__Solitalea, g__norank_o__SM1A07, g__Anabaena_XP0RK15F, g__Curvibacter, g__Vogesella are all names of microorganisms. They are usually expressed in English in the industry. To maintain scientific rigor, the present invention directly uses the English names of microorganisms to prevent deviations caused by literal translation into Chinese and avoid unnecessary troubles in industry research. Detailed implementation manners

[0048] The present invention will be further clarified below in conjunction with the accompanying drawings and specific implementation manners. It should be understood that the following specific implementation manners are only used to illustrate the present invention and not to limit the scope of the present invention.

[0049] The following takes the water diversion from the Yangtze River to Lake Taihu (the Yangtze River is the "raw water" and Lake Taihu is the "receiving area") as an example to illustrate the technical solution of the present invention:

[0050] 1. Preliminary screening of microbial indicator species

[0051] Before water resources dispatching, eDNA analysis is carried out on the microbial communities in the water samples of the "raw water" and the "receiving area". Through non-invasive water sample collection, PCR amplification of the target gene region, high-throughput sequencing (such as the Illumina platform), and denoising, clustering, and classification of the original sequences, operational taxonomic units (OTUs) are generated to determine the composition of the microbial communities. The Random Forest algorithm is used to screen the microbial species in the water samples of the "raw water" and the "receiving area". Random Forest is a powerful machine learning method that can handle complex multi-variable data and helps identify those microbial indicators (biomarkers) related to the water source introduction process and its diffusion. By analyzing the microbial community data at different positions and different time points in the "raw water" and the "receiving area", a group of representative microbial indicator species are screened out. These microbial indicator species can not only reflect the differences in the microbial communities between the "raw water" and the "receiving area", that is, the microorganisms existing in the Yangtze River before water diversion but not in Lake Taihu, but also characterize the mixing degree of the Yangtze River water and Lake Taihu water and the influence range on the water quality of Lake Taihu.

[0052] The specific screening of the microbial species in the water samples of the "raw water" and the "receiving area" is as follows: The "raw water" and the "receiving area" are divided into two groups, namely the "raw water group" and the "receiving area group". Before water diversion, the microbial community data of the "raw water group" and the "receiving area group" are analyzed by the Random Forest algorithm to screen out a group of representative microbial indicator species; by calculating the AUC (Area Under Curve) value, the prediction accuracy of the biomarker for the change of the microbial community under different water area conditions is evaluated; the prediction process can evaluate the stability and accuracy of the Random Forest algorithm model under different conditions to ensure the reliability of the selected biomarker in practical applications. The screened microbial indicator species can reflect the differences in the microbial communities between the "raw water" and the "receiving area", that is, the microorganisms existing in the "raw water" before water diversion but not in the "receiving area".

[0053] 2. Investigation of the background value of the diffusion range of the diverted raw water

[0054] The background value of chloride ions monitored during the water diversion from the Yangtze River to Lake Taihu is the average chloride ion concentration in Gonghu Lake and its surrounding waters before the introduction of Yangtze River water. During the water resources regulation process, the chloride ion tracing method is used to monitor the diffusion process after the Yangtze River water flows into Lake Taihu. Chloride ions have high stability and strong solubility. As a tracer element, they can effectively trace the source and flow path of water bodies. By analyzing the change in chloride ion concentration in water bodies, the diffusion range of the Yangtze River water source in Lake Taihu and its mixing with the local water bodies in Lake Taihu can be revealed. Specifically, first, by measuring the change in chloride ion concentration in water samples from different regions of Lake Taihu, the diffusion area of the Yangtze River water flow is located, and its propagation speed and influence range are speculated. For this purpose, water sample collection will cover different water depths and water area positions, and combined with environmental factors such as flow velocity and water temperature, a water flow diffusion map is drawn to speculate the diffusion direction and speed of the Yangtze River water source in Lake Taihu. This process provides a key spatial and temporal background for the subsequent analysis of changes in the microbial community.

[0055] 3. Verification of microbial indicator species

[0056] During the process of Step 2, environmental DNA (eDNA) samples are synchronously collected. 26 sampling points cover the entire Gonghu Lake and parts of Meiliang Lake and Xu Lake. For the microbial indicator species preliminarily screened in Step 1, combined with the spatio-temporal changes of chloride ions in Step 2, the microbial community data are analyzed again by the Random Forest algorithm in 8 groups (pre-water diversion period, Lake Taihu Group 1, Yangtze River Group 2, mid-water diversion period, Lake Taihu Group 3 that changes synchronously with chloride ions, Lake Taihu Group 4 that has nothing to do with chloride ions, Yangtze River Group 5, late-water diversion period, Lake Taihu Group 6 that changes synchronously with chloride ions, Lake Taihu Group 7 that has nothing to do with chloride ions, Yangtze River Group 8) to obtain the final microbial indicator species; the finally obtained indicator species are microorganisms that are unique or significantly enriched after the introduction of the Yangtze River, and these microorganisms did not exist in Lake Taihu before the water diversion. It can also characterize the mixing degree of "raw water" and "receiving water area", and the results will help to determine the species that persist and are stable in Lake Taihu, thus representing the successful integration of river microbial species.

[0057] 4. Application in water resources regulation decision-making

[0058] Finally, the screened biomarkers can be used to predict the diffusion range of the Yangtze River water source in Lake Taihu and water quality changes, and further provide data support for ecological risk assessment. These biomarkers also provide a scientific basis for ecological monitoring and environmental protection measures in future similar water resources regulation projects.

[0059] The present invention takes the distribution change of microbial indicator species in the water diversion of Gonghu Lake as an example to illustrate the specific application of the present invention, and Figure 1-3 shows the distribution change of microbial indicator species in Gonghu Lake at different water diversion stages.

[0060] In this embodiment,Figure 1 is the background value of microbial indicator species before water diversion (August 21st), see Figure 1 , before water diversion (August 21st), the average chloride ion concentration range in Gonghu Lake and its surrounding waters was 39.2 - 42.8 mg / L, and the average concentration was 41.4 mg / L. The spatial distribution of chloride ion concentration in each area of Gonghu Lake was relatively uniform, and the overall difference in tracer ion concentration was not significant.

[0061] Before water diversion (August 21st), g__Vogesella, g__Curvibacter, and g__Anabaena_XPORK15F were used as special indicator species during the Yangtze River water diversion to Taihu Lake. Among them, g__Vogesella belongs to the Chromobacteriaceae family, g__Curvibacter belongs to the Comamonadaceae family, and g__Anabaena_XPORK15F belongs to the Nostocaceae family. All of them showed the tracer potential of being ecologically friendly, stable in nature, and highly sensitive.

[0062] Figure 2 is the abundance distribution of chloride ions and microbial indicator species during water diversion (September 3rd) (log 10 Reads), see Figure 2 , during water diversion (September 3rd), the average chloride ion concentration in Gonghu Lake and its surrounding waters was 37.4 mg / L, showing a generally decreasing trend compared with before water diversion. The chloride ion concentration in the water body in the area of Wangyu Estuary and within the range from the estuary to Xidong Waterworks was significantly lower than that in other areas, indicating that the water from the Yangtze River had spread to the entire estuary area and had migrated to the north bank of Gonghu Lake, having a greater impact on the location of Xidong Waterworks.

[0063] During water diversion (September 3rd), g__Ottowia mainly aggregated near the entrance of Wangyu River into the lake, the diffusion area of g__Solitalea spread to the vicinity of Xidong Waterworks, g__norank_o__SM1A07 migrated to most of the lake areas of Gonghu Lake and Meiliang Lake, the migration path of g__Anabaena_XPORK15F was similar to that of g__norank_o__SM1A07, and both g__Curvibacter and g__Vogesella showed the migration characteristics of spreading from Gonghu Lake to Xuhu Lake.

[0064] Figure 3 is the abundance distribution of chloride ions and microbial indicator species in the middle and late stages of water diversion (September 12th) (log 10 Reads), see Figure 3 , in the middle and late stages of water diversion (September 12th), the chloride ion concentration in the northern lake area of Gonghu Lake was significantly lower than that in other areas, indicating that the water from the Yangtze River mainly migrated to the north bank of Gonghu Lake, affecting the area near Renzi Port and having limited impact on the middle and southern waters.

[0065] In the middle and late stages of water transfer (September 12th), g__Ottowia and g__Solitalea were only detected at a few monitoring points in Gonghu Lake. g__norank_o__SM1A07 mainly gathered in the northern part of Gonghu Lake and spread towards Meiliang Bay. g__Anabaena_XPORK15F had spread to the entire area of Gonghu Lake and Meiliang Bay and further spread towards Xu Lake. g__Curvibacter and g__Vogesella could be detected at most monitoring points in Gonghu Lake and its surrounding waters.

[0066] The following specifically describes the specific process of diverting water into Gonghu Lake:

[0067] I. Establishment of background values before water transfer (corresponding to Figure 1 )

[0068] Before the start of the water transfer project (August 21st), a background value survey was conducted on Gonghu Lake and its surrounding waters. Water samples were collected from 26 points, and the microbial community was analyzed by eDNA technology. At the same time, the chloride ion concentration was measured. The random forest algorithm was used to screen for differential microorganisms between the raw water (Yangtze River) and the water receiving area (Taihu Lake). The parameters were set as 500 - 1000 classification trees and the number of features was taken as the square root of the total features. Species with AUC≥0.7 were screened.

[0069] Background value of chloride ions: The average concentration in Gonghu Lake was 41.4 mg / L, and the spatial distribution was uniform.

[0070] Microbial indicator species: g__Vogesella (Chromobacteriaceae), g__Curvibacter (Comamonadaceae), and g__Anabaena_XPORK15F (Nostocaceae) were determined as the unique species in the raw water of the Yangtze River.

[0071] II. Monitoring during the water transfer process (corresponding to Figure 2 , September 3rd)

[0072] After introducing the Yangtze River water, the dynamic changes of chloride ion concentration and microbial abundance were monitored simultaneously. When the chloride ion concentration decreased by ≥15% (the average concentration decreased from 41.4 mg / L to 37.4 mg / L), eDNA sampling and sequencing were triggered. The microbial abundance was quantified by the log 10 Reads value, and the diffusion path was analyzed in combination with the change of chloride ions.

[0073] Distribution of chloride ions: The concentration was the lowest in the area from Wangyu Estuary to Xidong Water Plant, indicating that the Yangtze River water had spread to the north bank of Gonghu Lake.

[0074] Microbial migration characteristics: g__Ottowia gathered at Wangyu Estuary; g__Solitalea spread to Xidong Water Plant; g__norank_o__SM1A07 and g__Anabaena_XPORK15F migrated towards Meiliang Lake.

[0075] III. Verification in the Middle and Late Stages of Water Diversion (corresponding to Figure 3 , September 12)

[0076] In the middle and late stages of water diversion, the chloride ion and microbial abundance are continuously monitored. Species with strong synchronization with the diffusion of raw water are screened (the growth of reads value ≥ 5 times).

[0077] Chloride ion distribution: The concentration in the northern lake area continues to decline, and the central and southern parts are less affected.

[0078] Microbial dynamics: g__Anabaena_XPORK15F spreads to Meiliang Bay and Xuhu; g__Curvibacter and g__Vogesella are detected throughout Gonghu Lake.

[0079] IV. Decision-making Application

[0080] Through the above data, species such as g__Anabaena_XPORK15F are screened out as reliable biomarkers, and combined with the machine learning model to predict the diffusion path of raw water, providing a basis for optimizing the water resources scheduling plan (such as adjusting the water diversion intensity or scope).

[0081] Enlightened by the ideal embodiments of the present invention as described above, through the above description, relevant staff can completely make various changes and modifications without departing from the technical idea of this invention. The technical scope of this invention is not limited to the content in the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A method for screening microbial indicator species to support water resource scheduling decisions, characterized in that, Including the following steps: Step 1: Preliminary screening of microbial indicator species Before water resources allocation, environmental DNA analysis is carried out on the microbial communities in the "raw water" and "receiving area" water samples. Through non-invasive water sample collection, PCR amplification of target gene regions, high-throughput sequencing, and denoising, clustering, and classification of the original sequences, operational taxonomic units (OTUs) are generated to determine the composition of the microbial communities; Step 2: Investigation of the background value of the diffusion range of the diverted raw water During the water resources allocation process, the chloride tracer method is used to monitor the diffusion process after the "raw water" flows into the "receiving area". Chloride, as a tracer element, can effectively trace the source and flow path of the water body; by analyzing the change in chloride concentration in the water body, the diffusion range of the "raw water" source in the "receiving area" and its mixing with the local water body in the "receiving area" are revealed; Step 3: Confirmation and verification of microbial indicator species During Step 2, environmental DNA samples are collected synchronously. Several sampling points are selected to cover the entire "receiving area". By analyzing the changes in the characteristics of the microbial communities in the environmental DNA, the microbial indicator species preliminarily screened in Step 1, combined with the spatio-temporal changes of chloride in Step 2, are divided into several groups of microbial community data and analyzed by the random forest algorithm again to obtain the final microbial indicator species; the final obtained indicator species are the microorganisms that are unique or significantly enriched after the introduction of the "raw water", and these microorganisms did not exist in the "receiving area" before the water diversion, which characterizes the mixing degree of the "raw water" and the "receiving area", helps to determine the species that can persist and be stable in the "receiving area", and thus represents the successful integration of river microbial species; Step 4: Application of water resources allocation decision-making Finally, the selected biomarkers can be used to predict the diffusion range and water quality changes of the "raw water" in the "receiving area", providing data support for ecological risk assessment.

2. The method for screening microbial indicator species for supporting water resources dispatching decisions according to claim 1, wherein The specific content of Step 1 is as follows: Screen the microbial species in the "raw water" and "receiving area" water samples, divide the "raw water" and "receiving area" into two groups, namely the "raw water group" and the "receiving area group". Before water diversion, the microbial community data of the "raw water group" and the "receiving area group" are analyzed by the random forest algorithm to screen out a group of representative microbial indicator species; by calculating the AUC value, evaluate the prediction accuracy of the biomarker for the change of the microbial community under different water area conditions; the prediction process can evaluate the stability and accuracy of the random forest algorithm model under different conditions to ensure the reliability of the selected biomarker in practical applications. The screened microbial indicator species can reflect the differences in the microbial communities between the "raw water" and the "receiving area", that is, the microorganisms that exist in the "raw water" before water diversion but do not exist in the "receiving area".

3. The method for screening microbial indicator species for supporting water resources dispatching decisions according to claim 1, wherein, The specific analysis of the random forest algorithm in step 1 is as follows: the model parameters of the random forest algorithm are set to 500 - 1000 classification trees, and the number of features selected for each split is the square root of the total number of features; high-discrimination species are selected through AUC verification, with an AUC value ≥ 0.7, and the indicator species screened out include microorganisms of the genera g__Vogesella, g__Curvibacter, and g__Anabaena_XPORK15F.

4. The method for screening microbial indicator species for supporting water resources dispatching decision according to claim 1, characterized in that The specific chloride ion tracer in step 2 is as follows: First, by measuring the change in chloride ion concentration in water samples from different regions of the "water receiving area", the diffusion area of the "raw water" flow is located, and its propagation speed and influence range are speculated. For this purpose, water sample collection will cover different water depths and water area positions, combined with environmental factors such as flow velocity and water temperature, to draw a water flow diffusion map, so as to speculate the diffusion direction and speed of the "raw water" water source in the "water receiving area".

5. The method for screening microbial indicator species for supporting water resources dispatching decisions according to claim 1, wherein The dynamic time synchronization mechanism is included in the 8-group verification in step 3: when the monitoring result of the chloride ion concentration fluctuates by ≥ 15%, the eDNA in the corresponding area is analyzed key; the correlation coefficient between the microbial data and the temporal change of the chloride ion concentration reaches above 0.7, and the reads value of the indicator species in the area where the chloride ion rapidly decreases needs to increase by ≥ 5 times to be confirmed.

6. The method for screening microbial indicator species for supporting water resources dispatching decisions according to claim 1, wherein In step 3, the microbial community data are divided into 8 groups. In the early stage of water transfer, the "water receiving area group" is the 1st group and the "raw water group" is the 2nd group. In the middle stage of water transfer, the "water receiving area group" that changes synchronously with chloride ions is the 3rd group, the "water receiving area group" that has nothing to do with chloride ions is the 4th group, and the "raw water group" is the 5th group. In the late stage of water transfer, the "water receiving area group" that changes synchronously with chloride ions is the 6th group, the "water receiving area group" that has nothing to do with chloride ions is the 7th group, and the "raw water group" is the 8th group.

Images