Temporal Variation of Plankton Community in Typical Lake in Middle Reaches of Yangtze River: Structure, Environmental Response and Interactions

Abstract

Liangzi Lake, a typical shallow lake in the middle reaches of the Yangtze River, is important for water resource and biodiversity conservation. With the development of urbanization, anthropogenic activities have posed serious threats to the water quality and biodiversity of Liangzi Lake. To assess the aquatic ecosystem health of Liangzi Lake, the structure, the environmental response, and the interactions of plankton were investigated in 2022 and 2023. The results indicated that water temperature was a pivotal factor regulating plankton dynamics, with the assemblage patterns predominantly shaped by the phytoplankton species, which were Bacillariophyta in spring and Chlorophyta in summer. In terms of the phytoplankton, dissolved oxygen and the N:P ratio significantly affect cyanobacteria distribution. The high biomass and abundance of cyanobacteria in summer highlight the potential risk of harmful algal blooms. In contrast to the phytoplankton, the zooplankton exhibited enhanced resilience to changes in the surrounding environment. Rotifera was the dominant group in summer in terms of both abundance and biomass. Most core genera of plankton were jointly identified by eDNA metabarcoding and microscopical analysis, and eDNA metabarcoding had advantages in revealing a higher diversity. However, some taxa among rotifers such as Liliferotrocha were only identified using microscopical analysis. Therefore, a combination of both the methods is recommended to better understand the structuring mechanisms of plankton assemblages in lake ecosystems.

1. Introduction

Plankton are essential components in aquatic ecosystems, playing vital roles in energy flow and material cycling [1,2]. The high genetic diversity and adaptation ability of plankton make them widely distributed in various aquatic environment [1,3]. Among plankton, phytoplankton are crucial contributors to primary production, and zooplankton serve as the trophic links between primary producers and higher-trophic-level aquatic organisms [4,5]. The biodiversity of plankton has non-negligible effects on the stability of food webs, the functioning of trophic networks, and water environmental quality [6]. However, growing evidence reveals that the biodiversity of plankton is easily affected by environmental and climatic changes, and even anthropogenic activities [7,8]. For example, the outbreak of algal blooms in aquatic ecosystems has been widely reported [9,10,11]. Therefore, assessing the status and drivers of plankton diversity is vital to reveal the potential environmental problems of aquatic ecosystems.

In recent years, the environmental DNA (eDNA) metabarcoding technique has emerged to monitor the biodiversity of plankton [3,11,12]. Compared to the conventional microscopical method, eDNA metabarcoding can effectively identify species with similar morphological characteristics such as copepods at the larval stage [8,13]. However, eDNA metabarcoding still has limitations in many cases. For example, it is inefficient for revealing cyanobacteria and fish taxa [14]. To date, few studies have compared the biodiversity of both phytoplankton and zooplankton using these two methods in actual aquatic habitats [8].

Liangzi Lake, the second-largest freshwater lake in Hubei province, is located in the middle and lower reaches of the Yangtze River. It is a typical grass-type lake with extensive populations of aquatic macrophytes, such as Potamogeton crispus, Bythophyton indicum, Egeria densa, and Elodea nuttallii [15]. Due to its high water quality and rich biological resources, Liangzi Lake is considered a “fossil lake” and a “species gene pool” [16]. Over the last few decades, Liangzi Lake has been overburdened with tourism, and the main production modes are aquaculture and natural fisheries. Due to anthropogenic effects, a series of environmental problems has arisen. For instance, the decline in aquatic biodiversity and the degradation of aquatic macrophytes [15]. What is worse, the risk of cyanobacteria blooms has increased in recent years [17]. Thus, the risk assessment and management of plankton is necessary in Liangzi Lake. Therefore, the objectives of this study were as follows: (1) to assess the key environmental factors that shape the structure of the plankton in Liangzi Lake within spring and summer; and (2) to explore the differences between the molecular and traditional approaches in revealing the plankton community.

2. Materials and Methods

2.1. Sampling and Laboratory Analysis

Liangzi Lake (114°32′~114°43′ E, 30°01′~30°16′ N) is situated south of Ezhou City, adjacent to Wuhan City. With an average water depth of 4 m, and the lake spans approximately 339 km² within a 2085 km² basin characterized by southern highlands and northern lowlands [18]. Using the Liangzi island town, located in the center of Liangzi Lake, as a boundary, Liangzi Lake can be divided into East Lake, the Qianjiang Lakes, West Lake, and Niushan Lake. Notably, West Lake primarily supports aquaculture while East Lake maintains natural fisheries. Four sampling campaigns were conducted during March and June of 2022 and 2023 across key locations (Figure 1). Among these sampling sites, L7 is located at the only outlet of Liangzi Lake, and L10 is located at the estuary of Gaoqiao River, the largest inflowing river of Liangzi Lake. Filed measurements of water temperature (WT), pH, and dissolved oxygen (DO) were obtained using a multiparameter controller (Milti3430, WTW, Munich, Germany), and transparency (SD) was measured using a Secchi disk. Surface water samples (0~0.5 m depth) were collected with an organic glass hydrophore and temporarily stored in polyethylene plastic bottles with different volumes for the further analysis of water nutrients, chlorophyll a (Chl.a), and plankton. A total of 20 L of surface water was filtered through a NO. 13 zooplankton net and concentrated in 100 mL polyethylene bottles for the further analysis of Cladocera, Copepoda, nauplii, and larger Rotifera, and then 4 mL of 40% formaldehyde solution was added.

After transporting the water samples to the laboratory, the samples for chl.a analysis were filtered through 0.7 µm GF/F filters and stored at −20 °C until analysis. A total of 1 L of water samples for phytoplankton, protozoa, and smaller rotifers analyses were fixed with 1.5% (v/v) Lugol’s iodine. After sedimentation for 48 h, the supernatant was carefully siphoned off with a 2 mm diameter pipette. The remaining sample was then concentrated to 30 mL. Subsequently, all the samples were transported to the Analysis and Testing Center, the Institute of Hydrobiology, the Chinese Academy of Sciences, for microscopic identification. The concentrated samples were shaken well, and 0.1 mL was transferred to the 0.1 mL counter box for species-level cell counts using perioptometry. Protozoa and smaller rotifers were counted from the phytoplankton samples, with counts performed using 0.1 mL and 1 mL whole pieces from the concentrated 30 mL sample, respectively [19]. The currently accepted names of phytoplankton species were based on AlgaeBase [20]. Cladocera and Copepoda were counted from the zooplankton samples, with counts performed using a 1 mL counter box. The average values of a triplicate experiment were taken to calculate the density and biomass of the plankton samples. The taxonomic identification of qualitative samples was conducted at the species level based on the “Atlas of Freshwater Microorganisms” and the “Atlas of Freshwater Organisms in China” [19].

The measurements of total nitrogen (TN), total phosphorus (TP), and the permanganate index (COD_Mn) were conducted based on the Chinese national standard methods of China as described in our previous study [21,22].

In June 2023, surface water samples were also filtered through 0.22 μm filter membranes to collect the eDNA for the analysis of the compositions of plankton in Liangzi Lake. All the filter membranes were temporarily stored at −80 °C before further analysis.

2.2. DNA Extraction, and Sequencing Analysis

The total DNA of the plankton samples was extracted using the E.Z.N.A. soil DNA kit (OMEGA) based on the manufacturer’s protocol. After the quality assessment, the qualified DNA was used for further sequencing analysis. According to the previous studies, the primers A23SrVF2 (CARAAAGACCCTATGMAGCT) and A23SrVR2 (TCAGCCTGTTATCCCTAG) were selected to amplify the 23S rRNA gene region of the phytoplankton [23]. The primers V8F (ATAACAGGTCTGTGATGCCCT) and 1510R (CCTTCYGCAGGTTCA-CCTAC) were used to amplify the 18S rRNA gene region of zooplankton. Subsequently, the PCR products were purified, quantified, pooled, and sequenced on an Illumina MiSeq platform (Illumina, San Diego, CA, USA) at Majorbio Bio-Pharm Technology Co. Ltd., Shanghai, China. All the progress was conducted following our previous study [24]. The raw data were first annotated against the Nucleotide Sequence Database (NT, v20210917) and then some irrelevant sequences (e.g., bacteria, plant, and fungus sequences) were removed. The filtered sequences were further used to analyze the planktonic community. The sequence raw datasets have been deposited in the NCBI Sequence Read Archive with the Bio-Project accession number PRJNA1114416.

2.3. Calculation Methods

Integrated Nutritional Index (TLI)

The TLI was used to assess the nutritional state of the lake. The calculation of TLI is based on the principle of “Methods for the assessment of surface water environmental quality of China” China” [25].

$$TLI={\sum }_{i=1}^n{W}_{i}TLI\left(i\right)$$

(1)

where TLI(i) is the trophic status index of parameter i, and W_i is the relative weight of the nutritional status index of parameter i.

Taking chlorophyll a (chl.a) as the reference parameter, the relevant weights for the normalization of the parameter i are calculated as follows:

$$W_i=r_{ij}^2/{\sum }_{i=1}^n{r}_{ij}^2$$

(2)

where r_ij is the correlation coefficient between parameter i and the reference parameter, and n is the number of evaluation parameters. The r_ij of some parameters is described as follows: r_ij² (chla) = 1, r_ij² (TP) = 0.7056, r_ij² (TN) = 0.6724, r_ij² (SD) = 0.6889, and r_ij² (COD_Mn) = 0.6889.

The formula for calculating the nutritional status index of each item is:

TLI(Chl.a) =10(2.5 + 1.086lnChl.a)

(3)

TLI(TP) = 10(9.436 + 1.624lnTP)

(4)

TLI(TN) = 10(5.453 + 1.694lnTN)

(5)

TLI(SD) = 10(5.118 − 1.94lnSD)

(6)

TLI(COD_Mn) = 10(0.109 + 2.66lnCOD_Mn)

(7)

If the TLI < 30, the lake is in a state of oligotrophy; 30 ≤ TLI ≤ 50, mesotrophy; 50 < TLI ≤ 60, light eutrophication; 60 < TLI ≤ 70, moderate eutrophication; and 70 < TLI, severe eutrophication.

2.4. Statistical Analysis

The sequencing data were analyzed and visualized using the online Majorbio Cloud Platform [26]. After testing the normality of the data in our study using the Shapiro–Wilk test, the non-parametric data, including water quality and diversity indices, were analyzed using one-way Kruskal–Wallis ANOVA tests, followed by Duncan’s multiple comparison tests [27]. The data are presented as mean ± standard deviations (SDs), and the letters mean a significant difference in the index between the different groups. Redundancy analysis (RDA) and correlation analysis were used to analyze the relationship between the environmental factors and the planktonic groups using the R studio 4.03 software with R packages “Vegan” and “Corrplot”, respectively. Figures were visible by GraphPad Prism 8, and Origin2022. The sampling map in this study was generated by ArcGIS 10.3. Partial least squares path modeling (PLS-PM) was performed in SmartPLS version 4 software using the PLS-SEM algorithm. In addition, the taxa (relative abundance ≥ 1%) were considered as core taxa, and only the core taxa were subjected to correlation analysis and visualized in network analysis by Gephi version 0.9.2 (Spearman’s |r| > 0.6 and BH-adjusted p < 0.05) [27,28].

3. Results

3.1. Environmental Characteristics

Nutrients such as TP showed no noticeable variations across the four sampling campaigns (Figure 2), while TN exhibited significant seasonal differences. The water temperature ranged from 14.4 to 24.9 °C in spring and from 26.8 to 31.7 °C in summer. The concentration of Chl.a was significantly higher in summer compared to that in spring (K W test, χ² = 19.14, p < 0.001). The TLI of the water sample was approximately 50, a criterion for light eutrophication, with no significant seasonal difference (K W test, χ² = 3.95, p > 0.05). During the four sampling campaigns, COD_Mn, TN, and the N:P ratio were significantly higher in summer 2023, while DO was significantly lower at the same time.

3.2. Biomass of Planktonic Communities

In spring 2022 and 2023, Bacillariophyta dominated the phytoplankton biomass, followed by Chlorophyta (Figure 3). The highest phytoplankton biomass was found in the sampling sites (L6~L9) located in East Lake. In the summers of 2022 and 2023, cyanobacteria became the predominant group, accounting for 49.8% and 61.2% of the total biomass, respectively. Compared to that in spring, the phytoplankton biomass was significantly higher in summer. For the zooplankton (Figure 3b), no obvious variations were observed between spring and summer. Cladocera and Copepoda, belonging to Crustacea, were the dominant groups and accounted for 23.2~96.6% of the total biomass. In spring 2023, Rotifera contributed the highest proportion of biomass.

3.3. Abundance of Planktonic Communities

In the four sampling campaigns, Bacillariophyta and cyanobacteria were the predominant groups in spring (Figure S1), accounting for 32.0~92.2% of the total abundance, and the relative abundance of these two groups was higher at the sampling sites located in East Lake and around the Liangzi island town (L5~L9, with an average of 78.83%). The absolute abundance of phytoplankton in summer exceeded 108 cells/L, from one to two orders of magnitude higher than that in spring (Figure S3a). Cyanobacteria was the only predominant group, accounting for 88.8% and 88.2% of the total abundance in summer 2022 and 2023, respectively (Figure S3b). In contrast to the total biomass, Ciliata and Rotifera were always the predominant groups in Liangzi Lake (Figures S2 and S3c,d), and in spring 2023, they accounted for 98.5% of the total abundance.

3.4. Environmental Drivers of Planktonic Communities

At the species level, the phytoplankton were significantly more diverse (Figure 4a) and evenly distributed (Figure S4) in spring. In terms of the zooplankton, they were significantly less diverse and evenly distributed in summer 2022 and spring 2023 (Figure 4b and Figure S4). At the genus level, Aulacoseira was the dominant genus in the spring 2022 and 2023, but the rest dominant genera in these two sampling times varied greatly (Figure 4c). In summer, Raphidiopsis, Leptolyngbya, and Pseudanabena were the top three genera and accounted for more than 62.0% of the phytoplankton. The dominant genera of zooplankton in spring 2022 were Tintinnopsis, Strobilidium, and Conochilus, and in spring 2023, they were Epistylis and Strobilidium (Figure 4d). In summer, the dominant genera were Tintinnopsis, Liliferotrocha, Difflugia, and Trichocerca. RDA analysis further confirmed that the core genera of the plankton in summer showed high similarity (Figure 4e,f), while in spring, an obvious distinction was revealed. Among these measured environmental factors, WT and DO significantly affect the distribution of core genera of the plankton in Liangzi Lake.

Correlation analysis revealed that WT significantly positively correlated with the abundance of cyanobacteria (p < 0.001), the phytoplankton (p < 0.001), Copepoda (p < 0.001), Rotifera (p < 0.001), and the zooplankton (p < 0.01) (Figure 5a). In addition, TP significantly negatively correlated with the abundance of phytoplankton (p < 0.05) and cyanobacteria (p < 0.01), whereas a high N:P ratio is beneficial for the growth of phytoplankton and cyanobacteria in that lake ecosystem (p < 0.05). As for these planktonic groups, the phytoplanktonic groups have significant positive effects on the growth of Rotifera (e.g., cyanobacteria and Rotifera, p < 0.001). PLS path modeling further demonstrated that WT, DO, and COD_Mn exhibited positive effects on the abundances of the phytoplankton, cyanobacteria, the zooplankton, and Rotifera (Figure 5b). The effects of environmental factors on them based on these standardized total effects followed this order: Rotifera (λ = 0.719) > cyanobacteria (λ = 0.697) ≈ phytoplankton (λ = 0.682) > chl.a (λ = 0.607) > zooplankton (λ = 0.486).

3.5. Planktonic Co-Existence Patterns

As shown in Figure 5a, the abundance of zooplankton, particularly Rotifera and Copepoda, exhibited significant positive correlations with phytoplankton abundance. To better understand the temporal dynamics of the plankton in Liangzi Lake, network analysis was performed with the core taxa at the module level because taxa within the same module are more likely to co-exist [5]. The phytoplankton and zooplankton networks exhibit distinct patterns during the different sampling campaigns (Figure 6a). In detail, the phytoplankton networks are divided into four dominant modules, and the proportion of dominant module was higher in summer (98.8% and 95.3%), and the relationships among the taxa were tighter in that season, while the proportion of dominant modules in summer 2022 was relatively lower. The further analysis of the planktonic co-existence patterns revealed that complex interactions occurred between the phytoplankton and the zooplankton, and the distribution patterns of dominant modules are similar to those that occurred in the phytoplankton. In spring, the taxa belonging to Bacillariophyta (phytoplankton), Protozoa, and Rotifera (zooplankton) predominated the dominant modules with highest number of nodes (Figure 6b), while in summer, the predominant taxa in the dominant modules were Chlorophyta and Rotifera.

3.6. Comparison of eDNA Metabarcoding and Microscopical Analysis

In this study, eight phyla, 89 genera, and 132 species of phytoplankton were identified by microscopical analysis, and six phyla, 95 genera, and 152 species were identified by eDNA metabarcoding (Figure 7a–c). Phyla cyanobacteria, Bacillariophyta, Chlorophyta, Euglenophyta, and Cryptophyta were jointly identified by these two methods. As for the zooplankton, more taxa at the genera and species levels were identified using eDNA metabarcoding. However, the number of identified core genera was higher using microscopical analysis (Figure 7d,e). In terms of the phytoplankton, most of the jointly identified genera belonged to cyanobacteria, and Raphidiopsis was the predominant genus (40.1% and 46.5%). For the zooplankton, the unique core genera identified using eDNA metabarcoding mainly belonged to Ciliata, whereas the unique core genera identified by microscopical analysis mainly belonged to Rotifera.

4. Discussion

The biomass of the phytoplankton in Liangzi Lake was significantly higher in summer than that in spring, while the zooplankton biomass exhibited no significant difference between both the seasons (Figure S3), highlighting the different lifecycles of phytoplankton and zooplankton. The reason for this difference is that Liangzi Lake is a typical lightly eutrophic lake (TLI ≈ 50) (Figure 2). According to the modified plankton ecology group (PEG) model described by Sommer, Adrian [29], the biomass of phytoplankton in eutrophic waters will reach a peak in summer due to sufficient nutrients. In contrast, the proliferous of inedible phytoplankton [30] and the predation of filtering-feeding fish such as Culter alburnus [31] may further limit the growth of zooplankton.

4.1. Temporal Dynamics and Environmental Response of Planktonic Communities

As a result of adaptation to environmental variations, the composition of the phytoplankton in Liangzi Lake changed dramatically from spring to summer. The seasonal dynamics of plankton have been also observed in other freshwater systems [19,32]. In terms of the phytoplankton, they were significantly most abundant in summer; this is because the WT of Liangzi Lake in summer ranged from 26.8 to 31.7 °C, which fell within the optimal temperature ranges for the growth of most organisms [33]. In Liangzi Lake, Bacillariophyta and cyanobacteria were dominated in spring and summer, respectively (Figure S3). Compared with the other species, Bacillariophyta is competent to suffer harsh environmental conditions, and thus dominant in low-temperature waters [34]. When it comes to summer, the rising temperature gives cyanobacteria competitive advantages over other species [35], leading to the bloom of cyanobacteria (Figure 3) [36,37]. Similar temporal dynamics of the phytoplankton were also reported in Yongli Lake [38]. Particularly, the biomass and proportion (approximately 88% of the total abundance) of the cyanobacteria in Liangzi Lake are significantly higher than those in other eutrophic lakes, such as Erhai Lake [39] and Hongze Lake [40], indicating the potential risk of cyanobacteria bloom in Liangzi Lake. In addition to WT, P limitation, and nutrient stoichiometry (N:P) are other important determinants that regulate the lifecycle of phytoplankton, especially cyanobacteria [41]. Our result showed that P limitation and high N:P ratios enhanced the community competition of cyanobacteria (Figure 5a); a similar phenomenon was also observed in a gravel-bed urban river [42]. Due to the low concentration of nutrients in summer 2022 (Figure 2), the proportions of nitrogen-fixing cyanobacteria such as Leptolyngbya and Pseudanabena [43] were at high levels. In addition, phytoplankton with unicellular or filamentous shapes are efficient in nutrient use, which also contributed to the dominance of these species in nutrient-limited conditions [44].

As for the zooplankton, though Cladocera and Copepoda account for a high proportion of the total biomass in spring and summer (Figure 3), the dominant group in spring was Ciliata (Protozoa) in terms of abundance (Figure S3). This is because the body sizes of most zooplankton belonging to Ciliata are smaller than those of Cladocera and Copepoda. In addition, the results also indicate that the zooplankton in the eutrophic lake was usually dominated by small- and medium-sized species. Similar phenomena were also revealed in previous studies, indicating that small zooplankton taxa are dominated in aquatic environments [45,46]. Similarly to the phytoplankton, WT was the most important determinant that drives the succession of the zooplankton (Figure 4f and Figure 5). With the increase in WT in summer, Rotifera became the dominant group due to their high reproductive rate and environmental adaptability at high-temperature conditions [47,48,49]. The results of RDA and correlation analyses also showed that chl.a and DO have significant effects on the distribution of zooplankton. As mentioned in a previous study, the low levels of dissolved oxygen in summer synergically enhanced the advantages of rotifers [50]. In addition, rotifers are less affected by planktivorous fishes, which may also further increase their abundance in Liangzi Lake [51].

4.2. Interactions of Planktonic Communities

The results of network analysis implied that the phytoplankton were more stable and complex in summer (Figure 6a), indicating that the increased WT amplified their biological interactions [5,52]. The high proportion of dominant modules in summer further suggested a higher niche differentiation degree, which indicated that co-occurring phytoplankton could use different strategies to utilize the available resources from the surrounding environment to avoid competition and maintain their advantages in eutrophic lakes [3,53]. Therefore, it is no wonder that the phytoplankton in summer was only dominated by a few species such as Raphidiopsis, leading to the low diversity and uneven distribution of the phytoplankton (Figure 4c and Figure S4). The high abundance of cyanobacteria and low phytoplanktonic diversity have also been previously found in eutrophic shallow lakes [35,54]. Although a high WT led to the increased abundance of zooplankton in summer, compared to phytoplankton, the biological interaction among zooplankton was inactive (Figure 6a). As revealed in previous studies, phytoplankton are more sensitive to a rise of WT, and this results in rapid growth rates, while zooplankton cannot track the changes in primary production under increased WT [55].

As for the relationship between the phytoplankton and the zooplankton, the core taxa within the dominant modules in spring belonged to various groups, such as Bacillariophyta, Euglenophyta, Protozoa, Rotifera, and Cladocera (Figure 6b), while in summer, the core taxa within the dominant modules mainly belonged to Chlorophyta and Rotifera. It seems that the main relationship maintained among the phytoplankton and the zooplankton was changed from competitive to predatory. As described in a previous study, niche differentiation played a more important role in the community assembly of the plankton, especially in spring [19]. Therefore, there is no doubt that temperature bottom-up shapes phytoplankton to construct phytoplankton–zooplankton interactions [52]. In addition, the amplified interactions among the phytoplankton and the zooplankton in summer indicated the higher resistance of the planktonic communities to environmental disturbances.

4.3. Difference Between Morphological Analysis and eDNA Metabarcoding

In terms of the plankton that exist in aquatic ecosystems, some species, especially small-sized ones, are difficult to distinguish through morphological characteristics [56,57], and misidentification via microscopical analysis would result in us believing the plankton are not diverse [14]. In this study, more species were identified using eDNA metabarcoding compared to microscopical analysis (Figure 7a–c), implying that eDNA metabarcoding has advantages in revealing a higher diversity of plankton [3,8,12]. The relative abundance of most unique species identified by eDNA metabarcoding is lower than 1%, which highlights its high detection sensitivity [58]. Therefore, eDNA metabarcoding is useful as a useful tool to assess plankton biodiversity.

However, some limitations still exist for this method. Indeed, the biggest barrier to the application of eDNA metabarcoding is that the annotated databases do not contain reference sequences for all of plankton, resulting in the questionable identification of some species [59]. The substantial intraspecific variability of plankton also affects the accurate annotation of some species. In this study, 54 and 78 OTUs of the phytoplankton and the zooplankton failed to be annotated, so were “unclassified”. This may explain why some species identified by microscopical analysis were missed via eDNA metabarcoding (Figure 7d and e). For example, Liliferotrocha subtilis was detected in the water samples with a high detected frequency (100%) and abundance (21.4%) via microscopical analysis; however, eDNA metabarcoding failed to identify this species due to the lack of a corresponding barcode sequence in the NCBI GenBank [48]. A previous study conducted by Liu et al. (2022) revealed that Asteroplanus karianus and Coscinodiscus asteromphalus, both belonging to Bacillariophyta, were not detected in Jiaozhou Bay due to the deficiency of reference sequences [60]. In addition, some decaying biological matter could be also identified by eDNA metabarcoding, which may further affect the understanding of the “true” plankton community [61]. Song and Liang (2023) previously used 18S and COI primer pairs to identify the composition of zooplankton and demonstrated that the choice of primer pairs has significant effects on the identification of zooplankton [8].

In the process of microscopical analysis, larvae, small-sized species, other uncountable species, and similar morphological species cannot be accurately identified or assigned to appropriate taxonomic levels [60]. In addition, the body sizes of zooplankton vary greatly among different groups, commonly from a few micrometers to a few millimeters. According to previous studies [8,48], the calculation of species abundance in eDNA metabarcoding is based on DNA contents, but in microscopical analysis, the calculation of species abundance is based on cell numbers. Therefore, the increased differences in body sizes would decrease the community similarity between the two methods due to algorithmic differences. As a result, different compositions of the zooplankton were identified by the two methods (Figure 4d and Figure 7e). In our study, small-sized copepods such as Pseudodiaptomus (accounting for 34.8% of the total abundance in eDNA metabarcoding) were identified with low abundances (<1%) by microscopical analysis. Despite it having some non-negligible disadvantages, microscopical analysis was more suitable for evaluating the response of plankton to environmental stress. As described in a previous study, compared with eDNA metabarcoding, morphological analysis can identify more significant variations across treatments in terms of the number of species and relative abundance [48]. Therefore, though a low matching degree still exists between the two methods, and it is difficult to unify the results of the two methods, the combination of the two methods is suggested to better understand the structuring mechanisms of plankton assemblages.

4.4. Limitations

Due to the objectives of our study and logistical constraints in sampling campaigns, we only focused on the response of plankton to the variation in environmental changes in Liangzi Lake within spring and summer; data on the plankton from autumn and winter are absent. The absence of winter and autumn data limits insights into dormancy, nutrient cycling, the transition between cold and warm periods, and the community resilience of plankton across annual cycles. For instance, winter conditions (e.g., low temperatures and light availability) likely shape the dormant stages of plankton and influence spring recruitment, while autumn mixing events may reset the nutrient regimes. These gaps preclude a full understanding of inter-seasonal linkages. In future studies, annual sampling (including four seasons) is needed to provide a more complete picture of the plankton dynamics in Liangzi Lake.

5. Conclusions

Our study analyzed the structure, the environmental response, and the interactions of the plankton in an typical lake in the middle reaches of the Yangtze River and assessed the difference between eDNA metabarcoding and microscopical analysis in identifying the plankton. The plankton composition exhibited strong temporal variations, with water temperature emerging as a critical driver. The phytoplankton, especially cyanobacteria, in summer had a high biomass and were abundant, resulting in a high potential risk of cyanobacteria blooms. Zooplankton are insensitive to changes in the surrounding environments when compared with phytoplankton. Rotifera was the dominant group in terms of both abundance and biomass in summer. Network analysis further revealed that the temperature bottom-up shapes the phytoplankton to construct phytoplankton–zooplankton interactions. Though eDNA metabarcoding has more advantages in identifying the composition of plankton compared to microscopical analysis, the supplementary taxa detected by microscopical analysis highlight the benefits of combining both the approaches for a comprehensive understanding of plankton occurrence and distribution in water ecosystems.