Application of eDNA Metabarcoding Technology to Monitor the Health of Aquatic Ecosystems

Abstract

Environmental DNA (eDNA) is DNA isolated from environmental samples. It is distinctly different from genomic DNA, which is extracted directly from biological specimens. eDNA metabarcoding technology is a novel surveillance tool combining eDNA and second-generation high-throughput sequencing technology. Different from conventional approaches and biomonitoring techniques, eDNA metabarcoding technology (eMT) has many advantages, such as promising timeliness and accuracy, lower time consumption, and low cost, and thus is widely used in ecological and environmental monitoring, including that in rivers, lakes, oceans, soils, and sediments. As a tool, eDNA metabarcoding technology supplements the evaluation of environmental qualities by monitoring both the diversity of aquatic biology communities and target species. In addition, it is essential to understand the limitations of eDNA metabarcoding technology in practical applications. As a tool, eDNA metabarcoding technology features high efficiency, providing indicators of environmental health and allowing for the indirect estimation of the impact and extent of water pollution with respect to aquatic ecosystems. It provides new insights for aquatic environment protection.

1. Introduction

Aquatic ecosystems are bolstered by aquatic biodiversity to maintain their function and provide social services [1,2]. In the past few decades, an increase in anthropogenic activities has resulted in a series of environmental problems, such as water pollution, acid rain, ozone hole creation, etc. [3,4]. Under those impacts, the diversity of ecological communities has been impaired [5,6]. Studies have shown [1] that the global population of freshwater organisms in aquatic ecosystems declined by 81% from 1970 to 2012 [1,7,8]. It is very important to figure out how pollution affects our aquatic environment, a task that can be performed with an efficient tracking tool that employs accurate, comparative data and biological indicators. The utilization of biotechnology to alleviate this problem continues to gain popularity and requires more attention as a means of understanding the negative effects of human activity and restoring the balance of nature [8]. Conventional approaches to pollution evaluation usually involve laborious and prolonged physical and chemical assessments, constraining their efficacy and application. The latest advances in environmental DNA (eDNA) have provided valuable alternatives for the detection and surveillance of biodiversity in the environment, offering enhanced sensibility, sensitivity, and efficiency [9]. eDNA metabarcoding is a molecular biology method that combines eDNA analysis with metabarcoding to perform ecological monitoring across all habitats. eDNA metabarcoding technology is used for biodiversity assessment in various environments, involving the extraction and analysis of DNA fragments from environmental samples such as tissues, mucous, secretions, feces, blood, etc. [10]. Though eDNA metabarcoding technology itself cannot directly monitor pollutants, it offers a sensitive tool for monitoring changes in biodiversity, serving as an indicator of environmental health and allowing for an indirect estimation of the impact and extent of aquatic environment pollution [10,11]. eDNA metabarcoding is widely used in Europe, the United States, and other regions/countries in the monitoring of target, rare, and endangered species, as well as in the assessment of the response of aquatic biodiversity to changes in water quality [12].

Previous research has shown that the capability of eDNA metabarcoding technology to identify species quantity is 1.3 times greater than that of conventional approaches [13]. Jiang et al. identified 175 species of fish using eDNA metabarcoding technology, of which only 47 were identified using conventional approaches [14]. Xu et al. identified a total of 10 phyla, 22 classes, 50 orders, 82 families, 108 genera, and 108 species of phytoplankton using eDNA metabarcoding technology [15]. Zhou’s research team monitored the phytoplankton in the Chai River using eDNA metabarcoding technology. Results revealed that the eDNA method detected 108 genera of phytoplankton, which were members of 11 phyla, 41 orders, and 60 families [16]. Song and Liang used eDNA metabarcoding technology to identify planktonic algae from 9 families, 25 orders, 33 families, and 43 genera and showed that total nitrogen, water temperature, river width, and elevation were the key factors influencing the changes in zooplankton communities [17]. This complements the lack of ability in conventional approaches to elucidate the interaction of hydro-ecohydrological variables with ecological community structures with different land types [17]. Therefore, the advantage of eDNA metabarcoding is that it can serve as a powerful complement to traditional large-scale surveys of ecological resources and play an important role in ecological conservation, monitoring, and environmental management. This review provides an overview of the principles and development of eDNA metabarcoding technology, its application in aquatic environment monitoring, its limitations and breakthrough directions, etc. for improved understanding of technologies and tools for river and lake ecological health evaluation and biodiversity assessment.

2. Principles and Development of Environmental DNA Technology

Principles and Development of eDNA Metabarcoding Technology

The eDNA metabarcoding technique is a monitoring tool used for identifying biological communities and their diversities in rivers, soil, lakes, seas, and sediments by extracting DNA from these sites and performing PCR amplification and high-throughput sequencing of specific DNA fragments [8]. This technology was first proposed in the 2000s and applied in the field of environmental microbiology to test DNA sequence fragments to obtain sequence information and to classify microorganisms [7,8]. eDNA metabarcoding technology achieved a precedent for its application in aquatic environmental monitoring in 2008, via the detection of the presence of an invasive species (i.e., the American bullfrog) in European waters [18]. After more than 20 years of rapid development, eDNA metabarcoding technology has continuously been applied in other fields [18,19]. eDNA metabarcoding information may not only be stored and used for future research [20]; it may also be used to detect changes in the distribution of pollution-associated species [20]. The use of metabarcoding in aquatic environments can be used to determine the consequences of contamination [19]. It is also being applied to marine ecological monitoring to assess the impact of pollutants on marine ecosystems [21].

3. Application of Water Ecological Health Monitoring

3.1. Monitoring the Diversity of Aquatic Communities

Aquatic community diversity monitoring is a basis of aquatic environment state assessments. In the past few decades, experts have mainly relied on conventional approaches to identify species, via field sampling surveys, monitoring of individual larger species, and community sampling [22,23]. However, it is extremely difficult to carry out a wide range of surveys of complex aquatic ecosystems. The results of those monitoring data are not immediate accessibility and sufficiently accurate [24]. The timeliness is low as well. In addition, conventional approaches require personnel to have professional knowledge of species identification and classification, which are highly experienced, time-consuming, and labor-intensive [16]. In addition, the accuracy of result identification may be compromised due to phenotypic plasticity across various life stages and the lack of complete species reference libraries [25,26]. Consequently, eDNA metabarcoding technology is used as a supplementary tool to conventional approaches for improved monitoring efficiency and accuracy. Molecular identification via eDNA is a promising method for efficient and accurate biological monitoring. eDNA metabarcoding can be defined as using polymerase chain reaction (PCR) [9,27] with universal primers to amplify mixed DNA samples from the environment followed by high-throughput sequencing to determine the species composition in the specific environment where the samples were collected. For example, the use of traditional fish sample collection methods, such as gill net, cast net, hook and line, and immersion net, may potentially harm and disturb the target organisms and their living environment [28]. eDNA metabarcoding technology represents the latest advanced means to characterize the biodiversity of aquatic communities due to its minimized environmental disturbance [29]. The feasibility of eDNA metabarcoding technology may be demonstrated by comparing with conventional approaches. As shown by researchers, with conventional approaches, 12 species of fish were observed, which is 4 species fewer than with the eDNA metabarcoding technology [30]. Similarly, researchers have used eDNA metabarcoding technology to monitor fish and demonstrated its accuracy in monitoring and assessing the health of aquatic ecosystems [2]. eDNA metabarcoding technology has also been applied to monitor other biological species, such as phytoplankton, zooplankton, amphibians, etc. [11,29,31]. Studies have shown that eDNA metabarcoding technology is used as a tool for evaluating the resilience of aquatic ecosystems and evaluating the quality of restoration through monitoring the biological communities of aquatic ecosystems that are treated, repaired, and restored. Song and Liang [17] applied the technology to determine the variability characteristics of the community structures in the Weihe River under changing environments and evaluated the performance of the restoration of the aquatic environment. Zeng [32] used the technology to monitor green algae and gained a comprehensive understanding of the impacts of green tides on local ecosystems, which promoted the understanding of green tide dynamics, providing a new way for green tide control and ecological restoration and providing important scientific support for sustainable water environmental protection and management [32].

3.2. Monitoring of Target Species

In addition, eDNA metabarcoding technology is used to monitor the community dynamics in aquatic ecosystems in real time, breaking through the limitations of the traditional ecological survey methods in time and space [28]. The application of eDNA metabarcoding technology for sampling ballast water from ships has revealed invasive species [27]. To confirm the accuracy and effectiveness of eDNA analysis, a research team extracted DNA from ship ballast water and harbor water that had navigated the Great Lakes of North America and found two invasive species, zebra mussels and zebra donkey mussels, in ship ballast water [27,33,34]. In addition, eDNA metabarcoding technology is of great significance in monitoring the early stages of invasive alien species, which can be used to support early warning to reduce the risk of the community formation and further spread of invasive species and can also be used to detect the controlling effect of invasive species. Wang et al. successfully monitored the crown-of-thorns starfish, an invasive alien species, by eDNA metabarcoding technology [34], which supported the development of new strategies for controlling the early invasion and prevention of invasive alien species. Osathanunkul et al. [35] monitored jellyfish and provided a scientific basis for environmental risk management. Phytoplankton are key producers of food webs, and they are essential for maintaining the health of aquatic ecosystems. However, under certain circumstances, the outbreak proliferation of specific algal species may be harmful to the aquatic environment and living organisms. Therefore, it is necessary to monitor aquatic phytoplankton for early warning. Chen et al. monitored harmful algal blooms in the coastal waters of Beibuwan Bay [36], resulting in successfully warning of outbreaks of algal blooms. Manaff et al. applied the eDNA metabarcoding technology for large-scale investigation of harmful microalgae species in the southwestern waters of the South China Sea, supporting the prevention of harmful microalgae outbreaks in advance [37]. Tracking climate-change-induced biological invasions may also be achieved by using metabarcoding analysis of archived natural eDNA samplers. Junk et al. monitored mussel communities via eDNA metabarcoding technology and reconstructed the invasion trajectory of Austrominius modestus [38]. The related research is well illustrated for invasive species, a major threat to global biodiversity, whose colonization trajectory is often cryptic and poorly characterized.

Pollution of aquatic ecosystems has caused the endangerment of biological species, such as rare species. Therefore, it is crucial to monitor endangered and protected species. Due to their scarcity and the difficulty of identifying appropriate sampling locations, it is even more challenging to identify and monitor these aquatic organisms by conventional approaches [39]. However, non-invasive molecular methods may aid in the estimation, protection, and management of these species. eDNA metabarcoding technology is such a method. A research team has demonstrated the effectiveness of eDNA metabarcoding in monitoring endangered freshwater fish [40]. In addition, they verified that absolute quantification of species abundance can be achieved based on the eDNA metabarcoding technology [40]. Researchers have used this tool to monitor freshwater fish species for environmental assessment and quantitative analyses [29,41]. The eDNA metabarcoding technology has potential for monitoring and assessing the abundance of rare and endangered aquatic species.

3.3. Monitoring of Aquatic Organisms in Response to Environmental Stresses

Currently, fragile aquatic ecosystems are under great pressure, and the self-purification ability of many aquatic ecosystems is declining [32]. Protecting the ecological health of aquatic environments is crucial in ensuring the sustenance of the inhabiting flora and fauna. While eDNA metabarcoding technology itself does not directly test pollution conditions, its offers a sensitive tool for monitoring changes in biodiversity, serving as an indicator of environmental health and allowing for indirect estimation of the impact and extent of pollution in ecosystems [42]. Using eDNA appears to be a substitute for conventional monitoring techniques in assessing the ecological health of impacted aquatic environments and shows several prospects for development. The eDNA metabarcoding technique was used to monitor certain aquatic communities with low tolerance to environmental factors and indirectly assess the condition of aquatic environments [39]. It successfully showed the effects of pollutants, such as ammonia nitrogen, on biological communities. For example, Yang et al. monitored zooplankton communities by eDNA metabarcoding technology and successfully diagnosed the baseline values of environmental factors for the ecological barriers of Taihu Lake based on the effects of ammonia nitrogen on zooplankton communities in the watershed [42]. In addition, several studies have successfully applied the eDNA metabarcoding method to monitor and evaluate the effects of structural stress and potential negative impacts of heavy metals on biological communities for pollution evaluation [43]. With eDNA metabarcoding technology, the monitoring of aquatic biological communities complements the biological information lacking in the traditional monitoring methods. Yang et al. explored the responses of zooplankton to different environmental factors in the Taihu Lake basin, showing that the effect of nutrients (such as nitrogen) on zooplankton was greater than those of other environmental factors [42]. Meanwhile, in order to verify the effectiveness of eDNA metabarcoding technology in monitoring the responses of aquatic communities to environmental factors, Li et al. revealed human impacts on macroinvertebrate communities in polluted headwater streams with eDNA metabarcoding technology [44]. Results showed that the community compositions and structures in headwater streams had unique and significant differences under human impacts. Five-day biological oxygen demand (BOD₅) and ammonia nitrogen were the key variables explaining the variation in community structures [44]. eDNA metabarcoding technology is an important tool to diagnose the key factors that drive the quality of the water body towards an unhealthy state, which provides an important basis to addressing pollution in watersheds. Additionally, by applying eDNA metabarcoding for predicting anthropogenic pollution in rivers, Li et al. observed that the relative abundances of indicative operational taxonomic units (OTUs) were significantly correlated with nutrient levels [45]. These OTU data could be used to predict the nutrient status, with up to 79% accuracy on testing data sets, providing a novel approach to predicting the pollution status of rivers by using eDNA data [45]. Collectively, the eDNA metabarcoding technology may indirectly indicate the risk of certain areas and evaluate the restoration status and ecological risk in water bodies. It also has a good applicability to indirectly assess the quality of aquatic environments. The potential of eDNA metabarcoding technology to monitor the responses of aquatic organisms to environmental factors is also promising.

4. Limitations and Breakthrough Directions

4.1. Lack of Standardized Monitoring Processes

Despite the rapid developments in the eDNA metabarcoding technology and its advantages over conventional approaches, there are shortcomings in this method, which are related to factors such as sampling environment, sampling range, target species, etc. [29]. Currently, there is a lack of a complete set of standardized operating procedures for eDNA technology with high versatility. For example, different laboratories carry out individual eDNA research with personalized parameter selection, resulting in potential inconsistency, hindering the integration and comparison of cross-laboratory monitoring data [46]. In addition, for different lineages of organisms, the selection of universal primers has not yet reached agreements in the application to different types of environmental samples, making the results from different studies lack comparability [47]. In addition, most of the studies are relatively scattered. There are inconsistent conclusions and a lack of systematic analysis and comparison, which is not conducive to the standardization of the protocols of the eDNA metabarcoding technology.

4.2. eDNA Sample Source, Collection

Sample source and collection problems exist in the natural environment, where eDNA samples are subject to abiotic and biotic influences such as water flow, pH, temperature, ultraviolet light, molecular motion, etc., making pertinent sample collection challenging [26,48]. Samples are collected from multiple forms of eDNA composed at different times or in different organismal states, and during parallel experiments, eDNA that has been degraded cannot be detected properly, reducing the accuracy of the results [49]. eDNA is also subject to the action of water currents. eDNA at the original site may be carried by water currents to new sites [50]. All these phenomena may lead to false negative results for the original locus and false positive results for the new locus. False negatives may also occur when the eDNA concentration is below the PCR detection limit [51].

There are different methods for eDNA collection. Scholars have conducted research specifically on eDNA collection methods, including filtration, centrifugation, chemical solution, precipitation, etc., all of which can have a certain impact on the eDNA extraction rate, affecting the persuasiveness of the final results [51,52]. When the eDNA content in a sample is too low, repeated filtering and concentration before eDNA extraction are usually required for better sequencing efficiency [53]. In addition, it is uncertain how environmental factors affect the rate of production and degradation of eDNA, which indirectly influences the unbiased characterization of biological communities in the environment [10,52].

4.3. Pitfalls in Quantification and Relative Abundance

There are limitations in the primer design and selection. One example is the cytochrome c oxidase subunit I (COI) metabarcoding primers, as Hajibabae et al. revealed that COI metabarcoding primer choice affects the richness and recovery of indicator taxa in freshwater systems [38,54]. Amplicon choice and PCR replicates did not explain any significant variation in beta diversity among samples [55]. Sample choice affected the results. The Earth Microbiome Project, held at the Argonne National Laboratory, specifically discussed sample selection and acquisition [56].

As PCR is a sensitive method, there are a number of considerations that need to be made, especially when working with mixed-community samples [56]. Evans et al. tested the quantification of mesocosm fish and amphibian species diversity with eDNA metabarcoding. eDNA sampling and metabarcoding approaches’ estimation of relative abundance of species from read abundance may not agree with traditional approaches [57]. Additionally, eDNA concentrations are correlated with regional biomass. Salters et al. studied Atlantic cod in oceanic waters and the results revealed that haddock (Melanogrammus aeglefinus) and saithe (Pollachius virens) were present at biomass levels one order of magnitude lower than those of cod [58]. Norway pout (Trisopterusesmarkii), whiting (Merlangius merlangus), and blue whiting (Micromesistius poutassou) was present at two orders of magnitude lower. Poor cod (Trisopterus minutus) and silvery pout (Gadiculus argenteus thori) were three orders of magnitude lower [57].

Pitfalls are present in relative abundance estimation using eDNA metabarcoding [43]. The use of PCR in metabarcoding studies can introduce biases, which can therefore skew diversity estimates [59]. Several PCR biases can alter taxa abundance estimation, including annealing temperature and template secondary structures, and primer–template mismatches, which are taxa-specific [43]. In particular, PCR cycle number and primer amplification affect diversity in sequencing, causing copy number variation of the target loci and amplicon size [56]. Kelly et al. showed, in their eDNA metabarcoding simulations, a strong effect of the number of PCR cycles on estimates of biodiversity [56]. Increasing the number of PCR cycles resulted in decreases in both richness and Shannon diversity. And the trend and severity of the declines depended upon the distribution of amplification efficiencies [52,57].

4.4. Bias in the Annotation of Operable Classification Units

To some extent, the eDNA metabarcoding results depend on the annotation databases. Incomplete databases may result in a large number of unannotated sequences, resulting in a large amount of unusable genetic information [50,60]. Ideal OTUs should only include sequences from one species, and sequence clustering classification is generally performed by clustering OTUs by artificially setting classification thresholds between species, which may lose sequences from different species with sequence similarity higher than the classification threshold [61]. Operationally annotated taxon classes (OTUs) are typically classified with threshold similarity set at 97% [50,62]. Identification of OTUs often relies on existing sequences in relevant databases, and when relevant information about the species does not exist in the databases or the information is erroneous, the results of their species identification are not sufficiently accurate [57].

Upon the establishment of more complete reference libraries in the future, eDNA metabarcoding technology would become one of the most accurate tools for biological identification. A source of error in sequence assignment may be that the most similar sequence to the query in the reference library is not from the actual species to which the query belongs [63]. In addition to human error (incorrect identification of samples used for reference library sequences), there are other errors [62]. During data processing, it is common to perform denoising procedures to detect errors and clustering analyses to combine correct ‘parent’ sequences, while clustering may combine erroneous sequences into meaningful organisms [53]. Denoising is necessary to reduce data sets, simplify collection, and compare merits in future, long-term biomonitoring applications [63]. This approach may not be applicable when the test sequence has highly variable genetic markers such as COI since COIs will lose a lot of species information by applying denoising. Therefore, there are recommendations to perform clustering first, denoising in OTUs by creating an accurate background of sequence variation and abundance bias, followed by a final abundance filtering step [61]. The use of eDNA metabarcoding technology also depends on databases. If there are new species in nature, these species have not yet been described and included in databases. The use of eDNA metabarcoding technology would become limited. In addition, when their eDNA is detected, they cannot be recognized and taken into account, since, unlike molecular phylogenetic methods, eDNA metabarcoding may only be used to determine the presence or absence of a given organism in an existing classification [10]. The potential of eDNA metabarcoding technology in the identification of species affiliations of new organisms needs to be further explored and enhanced.

4.5. Limited Application Purpose

Metabarcoding technology has been demonstrated to have numerous advantages when it comes to the monitoring of aquatic ecosystems. These include the fact that it is not limited by time and space, no disturbance is caused to the aquatic ecological environment, no destructive damage is done to organisms, and it can be used to monitor rare and endangered species [17,27]. However, it is important to note that there are certain limitations in terms of the monitoring purpose. Metabarcoding technology is primarily used for the monitoring of species, and it cannot directly tap into the metabolic function and metabolic potential of organisms [43]. Metagenomics [23] also relies on high-throughput technology to monitor microbial communities in order to understand microbial metabolic functions and provide a reference for the next step of environmental pollution regulation and remediation [64]. Therefore, metagenomics can be used as a supplementary tool to understand the biological metabolic processes of species in aquatic ecosystems stimulated by environmental stress and provide more complete information on the health of aquatic ecological environments.

5. Prospect

Despite the existing limitations in the eDNA metabarcoding technology, via future technical optimization, its wider application in the field of environmental ecological assessment is expected. Especially, with the increasing popularity of big data techniques, eDNA metabarcoding technology has ushered in new opportunities for automation, intelligence, and diversification of the monitoring process. By integrating eDNA metabarcoding technology with drone sampling and automatic analysis instruments, the whole process from sample collection to sequencing analysis can be automated, which would significantly reduce the risk of human contamination and improve the accuracy of sequencing data. Combined with machine learning technology, eDNA metabarcoding technology can achieve more intelligent and efficient matching analysis. In addition, it may further allow the construction of mesh topology models with environmental factors. This not only deepens the understanding of community diversity in response to environmental stresses but also provides strong support for the protection of aquatic ecosystems through close correlation with water quality indicators.