Fish Diversity Monitored by Environmental DNA in the Yangtze River Mainstream

: Surveys and assessments based on environmental DNA are not only efﬁcient and time-saving, but also cause less harm to monitoring targets. Environmental DNA has become a common tool for the assessment and monitoring of aquatic organisms. In this study, we investigated ﬁsh resources in the Yangtze River mainstream using environmental DNA, and the variations in ﬁsh during two seasons (spring and autumn) were compared. The results showed that 13 species were identiﬁed in spring, and nine species of ﬁsh were identiﬁed in autumn. The ﬁsh with higher eDNA detection were Sinibotia superciliaris , Tachysurus fulvidraco , Cyprinus carpio , Ctenopharyngodon Idella , Monopterus albus , Acanthogobius hasta , Saurogobio dabryi , Oncorhynchus mykiss , Mugil cephalus , Odontamblyopus rubicundus . Seasonal variation between spring and autumn was not signiﬁcant, and the environmental factors had different effects on ﬁsh assemblages during the two seasons. Our study used the eDNA technique to monitor the composition of ﬁsh in the spring and autumn in the Yangtze River mainstream, providing a new technology for the long-term management and protection of ﬁshery resources in the region. Of course, problems such as pollution and insufﬁcient databases are the current shortcomings of environmental DNA, which will be the focus of our future research and study.


Introduction
As the largest river in Asia, and the third largest river in the world, the Yangtze River is rich in resources and has many tributaries and lakes, forming an important economic link connecting the east and the west of China [1]. The variety and output of Yangtze River fish rank first in China [1]. According to statistics, the fishery resources in the Yangtze River mainstream have gradually declined in the past half century [2]. In recent years, the government has gradually implemented relevant fishing bans in key waters such as the mainstream of the Yangtze River and important tributaries. In addition, an accurate understanding of its biodiversity is required to address the current challenges [3]. To appropriately monitor, manage and protect these key regions, it is essential to develop better monitoring approaches for biodiversity [4].
First explicitly proposed by Ogram et al. (1987) for microbial DNA from sediment, the environmental DNA (eDNA) technique introduced new opportunities in the study of microbial diversity [5]. Later, Rondon et al. (2000) used eDNA to assess the genetic and functional diversity of uncultured microorganisms in soil [6]. The eDNA technique is used to identify the target species presented by exploiting DNA molecules shed in the environment (such as water, soil, air, feces or snow) [7]. Due to its limited cost and versatility in numerous contexts, the eDNA technique is expected to become an effective monitoring method [8]. The eDNA technique, with high probability of detection, is especially suitable for fish, and water is the most common sample [9]. eDNA monitoring of aquatic creatures was first used in 2008 to detect the presence of the American bullfrog Rana Catesbeiana in freshwater, and it has since been used extensively to monitor fish species [10]. Since then, eDNA has been used to explore the trajectories of invasive organisms such as American bullfrog Rana Catesbeiana [11], silver carp (Hypophthalmichthys molitrix) and bighead carp (Hypophthalmichthys mobilis) in USA area canals and waterways [12], New Zealand mudsnails Potamopyrgus antipodarum [13], the crayfish Procambarus clarkia [14], quagga Dreissena bugensis and zebra D. polymorpha mussels [15] and golden mussel Limnoperna fortune [16], and to protect the endangered largetooth sawfish Pristis pristis [17], Maugean skate Zearaja maugeana [18] and Plecoptera species Isogenus nubecula [19], as well as to monitor rare species such as the round goby Neogobius melanostomus from the Black Sea [20].
Additionally, eDNA has gradually become more widely used, mainly including the following three aspects. Firstly, it is used for monitoring single fish distribution, including determining habitat connectivity for the migration of fish between the sea and rivers [21]. Sigsgaard et al. (2016) demonstrated that the high-throughput sequencing of seawater eDNA could provide useful estimates of genetic diversity in a whale shark (Rhincodon typus) aggregation [22]. Bakker et al. (2017) used an eDNA metabarcoding approach specifically targeted to infer shark presence, diversity and eDNA read abundance in tropical habitats [23]. In addition, Boussarie et al. (2018) confirmed that eDNA can be used to detect shark diversity by comparison with traditional underwater visual censuses and baited videos [24]. Thalinger et al. (2019) monitored the spawning migrations of potamodromous fish species including Danube bleak Alburnus mento and Vimba bream Vimba vimba via eDNA, which found a strong correlation between daily visual fish counts and downstream eDNA signals [25].
Secondly, eDNA technology is used for multi-species monitoring, including surveys of vertebrate and invertebrate populations [26]. Port et al. (2015) successfully surveyed the vertebrate fauna present along a gradation of diverse marine habitats associated with a kelp forest ecosystem using eDNA [27]. Forsström et al. (2016) demonstrated for the first time that eDNA from an introduced mud crab Rhithropanopeus harrisii could be successfully amplified in aquarium water samples and detected in a brackish water environment [28]. Lanzén et al. (2016) applied DNA metabarcoding of eukaryotic communities to successfully identify changes in sediment communities surrounding oil platforms [29]. O'Donnell et al. (2017) investigated the potential of eDNA to provide information about the breadth of biodiversity present in a tropical marine environment, which illustrated that eDNA can be used to explore diversity beyond taxon identifications [30]. DiBattista et al. (2017) assessed the utility of eDNA as a tool to survey reef fish communities in the Red Sea [31]. Later, DiBattista et al. (2019) also investigated the potential for rapid universal metabarcoding surveys (RUMS) of eDNA in sediment samples to provide snapshots of eukaryotic subtropical biodiversity [32].
Moreover, eDNA approaches have made great achievements in ancient and contemporary ecosystems for the description of past biodiversity and practical present-day conservation [33]. Jeunen et al. (2019) determined the ability of eDNA metabarcoding surveys to distinguish localized signals, adding the evidence that eDNA in marine environments can detect a broad range of taxa that are spatially discrete [34]. With the development of large-scale parallel sequencing, eDNA has been used to conduct quantitative analyses of monitoring targets in a large number of studies [35]. Kimmerling et al. (2018) used high-throughput metabarcoding to quantify the species-level ecology of reef fish larvae, which could assist in coral reef conservation and fishery management efforts [36]. West et al. (2020) also demonstrated the utility of a multimarker metabarcoding approach in capturing multitrophic biodiversity across an entire coral reef atoll, which sets an impor-tant baseline for ongoing monitoring and management [37]. As an eco-friendly tool, eDNA technology can not only be used for fishery monitoring, but might also importantly assist in marine resource conservation and the sustainable exploitation of fisheries [38].
The purpose of this study was to use eDNA techniques to investigate fish resources in the Yangtze River mainstream, as they could provide a new monitoring technology for fish resources in this region. Additionally, we compared the changing seasonal trends between spring and autumn. The results provide a data reference for future studies on the monitoring and protection of fish resources in the Yangtze River. Our research confirms the utility of eDNA technology in resource assessment.

Sample Sites and Collection
Sampling points were set up in the upper reaches of the Yangtze River from Panzhihua to the lower reaches of Nanjing, which covered the Yangtze River mainstream (Figure 1). We collected 10 eDNA samples in spring and 11 eDNA samples in autumn at the same coordinate position (Table 1). Due to the distance between the sampling points, the sample collection of the two quarters was completed in around 8 days. The specific eDNA samples and corresponding sampling site information are shown in Table 1. Before sampling, we sterilized all instruments with ultrapure water and 75% alcohol and used the water near the sampling point to rinse the sampling supplies again; then, we discarded them downstream of the sampling to reduce field sampling pollution. A 1 L water sample from a single site was collected from the surface water layers and mixed in a sterile brown plastic bottle with Urine DNA Locker (Beijing, China). Then, the 1 L water samples were filtered through a 0.22 µm Sterivextex TM-GP filter unit (Millipore, Burlington, MA, USA), and each water sample bottle was filtered twice to ensure adequate collection. In addition, 1 L of water was collected to measure environmental factors. First, 500 mL of water was filtered with a 0.65 µm pore-size polycarbonate microporous membrane. After the membrane was collected, the chlorophyll (Chl) was measured by fluorescence extraction, and the filtered water was used to measure nutrients. A QuAAtro Nutrient Salt Automatic Analyzer (Germany SEAL, Norderstedt, Germany) was used to analyze the nutrient salt, including total nitrogen (TN), total phosphorus (TP), SiO 3 , NH 4 , NO 3 , NO 2 . In the laboratory, we wore gloves to handle samples, in strict accordance with the experimental specifications, and to perform disinfection; in addition, all experimental materials were disinfected and sterilized to prevent contamination. To prevent contamination of the experimental operation, the filter bottles and anti-corrosion peristaltic filter tubes were rinsed with 75% alcohol and ultrapure water after each sample was filtered. We then added DNA preservation buffer (Tiandz Inc., Beijing, China) to the filter unit and filtered three equal volumes of deionized water as the negative control.

DNA Extraction and PCR
We used a DNeasy Blood and Tissue Kit (Qiagen, Shanghai, China) to extract total eDNA, following the manufacturer's protocol. At the same time, Nanodrop was used to quantify the DNA, and the DNA extraction quality was detected by 1.2% agarose gel electrophoresis. The conserved region of 12S rRNA was amplified by PCR using the primers MiFish-F (5 -ACACTCTTTCCCTACACGACGCTCTTCCGATCTNNNNNNGTCGGTAAA ACTCGTGCCAGC-3 ) and MiFish-R (5 -GTGACTGGAGTTCAGACGTGTGCTCTTCCGAT CTNNNNNNCATAGTGGGGTATCTAATCCCAGTTTG-3 ) [39]. Three PCR replicates were performed for each sample, and the details of library construction were as follows: all three reaction volumes were 25 µL, including 5 µL 5 × buffer, 2 µL dNTP (2.5 mM), 1 µL (10 uM) forwardprimer, 1 µL (10 uM) reverse primer, 1 µL DNA Template, 14.75 µL ddH2O and 0.25 µL fast pfu DNA polymerase. The thermal-cycle profile was for 12S: 98 • C for 2 min; 35 cycles of 98 • C for 15 s, 55 • C for 30 s and 72 • C for 30 s; and 72 • C for 5 min. The negative control was set, which was used to detect microbial contamination from sources such as the environment and reagents, and any sample group with bands amplified by the negative control could not be used for subsequent experiments. The PCR products were quantified by a microplate reader (BioTek, FLx800, Invitorgen, San Diego, CA, USA) as the quantification instrument. The fluorescence reagent was the Quant-iT PicoGreen dsDNA Assay Kit (Invitorgen, San Diego, CA, USA).

Library Preparation and Sequencing
The sequencing library was prepared using the TruSeq Nano DNA LT Library Prep Kit (Illumina, San Diego, CA, USA). Firstly, the sequence ends of the amplified products were repaired by the End Repair Mix 2 in the kit. We added an "A" base to the 3 end of the DNA sequence to prevent the DNA fragment from self-linking and ensure that the target sequence could be connected to the sequencing adapter. A sequencing adapter containing a library-specific tag (index sequence) was added to the 5 end, so that the DNA molecule could be immobilized on the Flow Cell. Secondly, the BECKMAN AMPure-XP Beads were used to purify the library system after adding adapters; then, we performed PCR amplification on the above-mentioned DNA fragments with adapters, and used BECKMAN AMPure XP Beads to purify the library-enriched products again. Finally, the library was subjected to final fragment selection and purification by 2% agarose gel electrophoresis. The sequencing platform was Illumina Novaseq 6000. Before sequencing, we used the Agilent High Sensitivity DNA Kit to perform quality inspection on the library on the Agilent Bioanalyzer. The qualified library had only a single peak and no linker before proceeding to the next step. The NovaSeq sequencer was used for pair-end sequencing, and the corresponding reagent was the NovaSeq 6000 SP Reagent Kit (500 cycles).

Bioinformatic Statistics and Fish Identification
The raw sequencing data were saved in FASTQ format. The sequence clustering statistics were mainly performed with QIIME 2 2019.4, while the OTU clustering procedure followed the Vsearch (v2.13.4) pipeline described previously (https://github.com/torognes/ vsearch/wiki/VSEARCH-pipeline, 8 August 2021) [40,41]. Briefly, raw sequence data were demultiplexed using the demux plugin, followed by primer cutting with the cutadapt plugin. Sequences were then merged, filtered and dereplicated using the fastq_mergepairs, fastq_filter and derep_fulllength functions in Vsearch [42]. All the unique sequences were then clustered at 98% (via cluster_size), followed by chimera removal (via uchime_denovo). Lastly, the non-chimera sequences were re-clustered at 95% to generate out representative sequences and an OTU table. Through the statistics of the flattened OTU table, the specific composition table of the fish communities in each sample at each classification level was obtained. Then, alpha diversity metrics (Chao1) [43], observed species, Shannon [44], Simpson [45], Faith's PD [46], Pielou's evenness [47], Good's coverage [48], beta diversity metrics (weighted UniFrac) [49], unweighted UniFrac, Jaccard distance and Bray-Curtis dissimilarity were estimated using the diversity plugin, with samples rarefied to sequences [50]. We used NCBI Organelle Genome Resources for the BLAST search. The information from the databases was merged to build a new database by using the Blastn tool. We obtained the specific composition table of the fish community in each sample at each classification level by using the statistics of the flattened OTU table, and we then used the R package (v3.3.2) script to construct the data from the table as a histogram (Figure 2). In the following process, we showed the composition of fish, and the related analyses were based on the level of species classification. Finally, we took into account the conditions of the Yangtze River to exclude the species that should not have appeared; we used Fishbase (https://www.fishbase.se/search.php, 9 October 2021) and also consulted books about fishery resources in the Yangtze River for localization certification [51].

Seasonal Difference Comparison
We compared the results of eDNA surveys in spring and autumn and analyzed the variation in fish resources in the mainstream of the Yangtze River. Nonmetric multidimensional scaling (NMDS) was used to reveal the seasonal differences [52]. NMDS analysis was performed on each distance matrix separately using the Vegan package under the R script, and the structure distribution of community samples was described by a two-dimensional sorting diagram. A thermal image was used to show the difference in species composition based on average abundance ranking or the degree of similarity between the two seasons. In addition, canonical correspondence analysis (CCA) was applied to analyze the correlation between seven environmental factors (Chl, TP, TN, SiO 3 , NH 4 , NO 3 , NO 2 ) and the distribution pattern of fish assemblages between the two seasons; the fish composition was based on the analysis of the relative abundance of the obtained sequence [38]. Seven environmental factor parameters are presented in Supplementary Material Table S1.

Fish Resource Results by eDNA
The results of the three negative controls proved that there was no cross-contamination during eDNA extraction. After the quality filtering procedure, the total readings of high-quality sequences detected in spring were 1227540, while there were 1317686 in autumn, from 21 samples (Supplementary Material Table S2). Samples were collected from 10 sampling sites in the spring and 11 sampling sites in autumn (Table 1). In spring, the domain species were a number of small and economic species known to inhabit shallow waters, such as Sinibotia superciliaris, Asian swamp eel Monopterus albus and the Chinese lizard gudgeon Saurogobio dabryi. In addition, some deep-water fish, such as the flathead grey mullet Mugil cephalus and the yellow catfish Tachysurus fulvidraco, were found, which displayed significantly greater relative sequence abundances in shoreline samples compared to interior samples ( Figure 2, indicated as "SE"). In autumn, several medium-sized and large-sized fish species were typically found, such as the grass carp Ctenopharyngodon idella, Common carp Cyprinus carpio, yellow catfish Tachysurus fulvidraco and Rainbow trout Oncorhynchus mykiss (Figure 2, indicated as "AE").

Diversity of Fish Communities
The spring and autumn samples were divided into two groups and were represented by different colors. In the figures, each panel corresponds to an alpha diversity index, marked in gray at the top. The box plot area of the "spring" Simpson and Shannon index plate was larger than that for "autumn", which indicated that the community diversity in spring was greater than in autumn (Figure 3a). In addition, the box plot area of observed species index analysis for "autumn" was larger than that in the "spring", indicating that the abundance in spring was higher than that in autumn (Figure 3a). In the two-dimensional sorting graph of the NMDS analysis, the points of the "spring" and "autumn" groups were closer, indicating that the difference in fish communities in spring and autumn was smaller, i.e., the composition of fish species in the Yangtze River mainstream in 2019 was similar in spring and autumn (Figure 3b).

Seasonal Variation Analysis
The heatmap of the relative sequence abundances of individual fish taxa revealed disparities in spatial distributions among species in each season (Figure 4). A total of 13 fish species were identified from the two seasons in the mainstream of the Yangtze River in 2019; they were common carp Cyprinus carpio, the Chinese lizard gudgeon Saurogobio dabryi, Saurogobio dumerili, yellow catfish Tachysurus fulvidraco, Acanthogobius hasta, Sinibotia superciliaris, amur sturgeon Acipenser schrenckii, grass carp Ctenopharyngodon idella, Odontamblyopus rubicundus, flathead grey mullet Mugil cephalus, stone flounder Kareius bicoloratus, the Asian swamp eel Monopterus albus and rainbow trout Oncorhynchus mykiss.

Relationship with Environmental Factors
The relationships between the environmental factors and the detected species were illustrated in a CCA analysis diagram using data from the 13 species detected in the two seasons and the set of seven environmental factors ( Figure 5). As shown in the CCA plot, Chlorophyll-a was the key environmental factor affecting the fish assemblages, as the first axis was strongly correlated with TN, NO 2 and Chlorophyll-a. The correlation between environmental factors and the distributions of the different species was inconsistent, and Chlorophyll-a had the strongest effect. The fish detected in spring and autumn, which are marked in purple in Figure 5, were not significantly affected by environmental factors.

Fish Diversity in the Mainstream of the Yangtze River Based on eDNA
Our study investigated the composition of fish diversity in the spring and autumn in the Yangtze River mainstream, showing that eDNA technology can be used to investigate fish diversity. However, some common Yangtze fish, such as silver carp, bighead carp, grass carp and herring, were not detected in this investigation [51]. This may be attributed to the following aspects.
(1) Insufficient sample collection: The fish activities in the Yangtze River mainstream are relatively large; in the present study, we only set 11 sampling points in the Yangtze River mainstream, which may have led to insufficient sampling. Ficetola et al. (2015) pointed out that replication, i.e., the number of water samples collected as opposed to the number of technical replicates run from a single sample, can be adjusted to compensate for false negatives [53]. Therefore, adding sampling points according to the study area would increase the possibility of detecting fish. (2) Less eDNA preservation: The amount of eDNA secreted by fish is related to itself; studies have shown that different species secrete DNA at different rates and quantities [54]. Moreover, the preservation of eDNA is also very important, as the long transportation after sampling, temperature, time and other factors will affect the preservation effect of the eDNA experiment. In addition, delayed extraction from the Sterivex filter may also result in less DNA being extracted. Thus, special attention should be paid to the preservation of samples, and timely filtration after sampling will reduce the degradation rate of eDNA. (3) Lack of local reference database information: This study did not establish a more accurate local reference database for fish in the Yangtze River mainstream, which may have reduced the number of native species identified. The effectiveness of eDNA depends on the quality of the reference sequence database and the classification parameters employed [55]. In future investigations and studies, we will consider establishing a local database. At the same time, the results of this survey will also serve as a reference for future surveys of fish in the Yangtze River mainstream. (4) Inapplicability of 12S primer to freshwater fish: The primers that we used may not be suitable for the identification of fish in the mainstream of the Yangtze River, which could explain the small number of identification results. Many studies have used other primers, such as the mitochondrial cytochrome c oxidase subunit I (COI) sequence and the 16S primer for the fish amplification [56]. In future studies, different primers can be designed to explore primers that are more suitable for Yangtze River fish. In the non-closed fishing season, fish body muscle tissue can be used for primer design, which will be our next research direction.
Shen et al. (2019) gathered a DNA barcode reference library of 2830 sequences from 238 species, with the objective of re-examining Yangtze River ichthyofauna diversity through a DNA-based method of species delimitation [1]. In order to obtain as many individuals per species as possible, various gears were used to capture specimens at different locations [1]. This will cause harm to the monitored species itself, and the eDNA technology only needs to collect water samples to detect the species. Therefore, it is very important to use the eDNA technique to monitor species and implement ecological protection in the future [38]. In addition, we detected some bottom-dwelling fish such as the Asian swamp eel Monopterus albus, which is difficult to be observed during normal activities due to their hidden living environment; this also demonstrates the sensitivity of our eDNA monitoring.

Seasonal Variation between Spring and Autumn
In the present study, the thermal imaging, through NMDS, showed that the composition of fish communities detected by eDNA technology was similar in spring and autumn. Nine species of fish were detected in both seasons; the research sampling was conducted in April and October in the Yangtze River. This result might be related to the similar temperature of the two seasons, as well as the living habits of the species themselves [57]. In addition, studies have shown that some water systems lack seasonal patterns regarding the fish species composition [38]. Alpha diversity analysis showed higher diversity in autumn, which may be related to the higher content of eDNA during the breeding period, which is consistent with the results of other previous investigations [38,57]. The current results suggest that continuous investigation would be of great help to fully understand the biodiversity of the mainstream of the Yangtze River.

Relationship between Fish Assemblage and Environmental Factors
Environmental DNA has been investigated in many waters in China since its inception [3]. Recently, there has been research on the construction of a database of fish in the Yangtze River mainstream through DNA barcoding technology, which is the key to the successful application of eDNA technology [1]. Chlorophyll-a had the greatest effect on fish species distribution, which is consistent with the findings of fish communities in other studies [38,55,57]. Our survey results also provide a reference for future surveys on the Yangtze River mainstream, and the eDNA technology used in species detection in this area is worthy of popularization and application. In addition, the impact of environmental factors on the detection of eDNA technology is also worthy of in-depth exploration, and more investigation and verification are required here [27].

Conclusions
In this study, we investigated the fish composition in the spring and autumn in the Yangtze River mainstream in 2019 through eDNA technology, which provides a reference for future surveys on the Yangtze River mainstream. Currently, there are few studies on the use of eDNA technology to investigate the seasonal composition of fish in the mainstream of the Yangtze River. It is worth noting that eDNA can represent a tool to assess composition in fish resource surveys, especially in the context of the ten-year ban on fishing in the Yangtze River. Although there are still some shortcomings caused by technical operations, pollution and insufficient databases, etc., this area needs to be studied in depth in the future. Therefore, we suggest that relevant research should continue to be carried out in the future to monitor the fish resources in the Yangtze River mainstream, so as to provide a scientific basis for the development and utilization of fishery resources in the region.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/fishes7010001/s1, Table S1: Seven environmental factors of two seasons, Table S2: Sequencing statistics for each sample in two seasons.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The original data file for the article can be found online at https: //www.ncbi.nlm.nih.gov/sra/PRJNA783738, accessed on 11 December 2021.