Continental scale metagenomics reveals Microcystis-induced synchrony between aquatic taxa is exacerbated by trophic cascades

Harmful Microcystis proliferation severely threatens aquatic ecosystems globally, but their impacts on food-web stability remain poorly understood. Here, we analyzed 1,309 metagenomic samples from freshwater ecosystems across China to investigate how Microcystis abundance affects synchrony among taxa in multi-trophic food-webs (from bacteriophages to piscivorous fish). We found that Microcystis abundance significantly positively correlated with anthropogenic factors (e.g. GDP, population density, nitrogen and phosphorus emissions), and their proliferation strongly enhanced synchrony between aquatic taxa (enhanced synchrony normally correlate low stability), whereas other aquatic taxa (e.g., fish, zooplankton and bacteria) showed no clear enhancing effects on the synchrony. The enhanced synchrony caused by Microcystis proliferation occurred more frequently between the consumers and their foods than between the aquatic taxa without consumer-food relationship. Moreover, trophic cascades more strongly drove these synchrony dynamics in middle and high trophic-levels, while synchrony in low trophic-levels was mainly affected directly by Microcystis. Overall, our results reveal ecosystem stability relies strongly on trophic linkages and provide a pioneering framework for integrating extensive metagenomic datasets to enhance our understanding of ecosystem stability across broad taxa spectrum. Continental scale metagenomics reveals Microcystis -induced synchrony between aquatic taxa is exacerbated by trophic cascades Running title: Microcystis synchronizes taxa in food webs

Harmful Microcystis proliferation severely threatens aquatic ecosystems globally, but their impacts on food-web stability remain poorly understood. Here, we analyzed 1,309 metagenomic samples from freshwater ecosystems across China to investigate how Microcystis abundance affects synchrony among taxa in multi-trophic food-webs (from bacteriophages to piscivorous fish). We found that Microcystis abundance significantly positively correlated with anthropogenic factors (e.g. GDP, population density, nitrogen and phosphorus emissions), and their proliferation strongly enhanced synchrony between aquatic taxa (enhanced synchrony normally correlate low stability), whereas other aquatic taxa (e.g., fish, zooplankton and bacteria) showed no clear enhancing effects on the synchrony. The enhanced synchrony caused by Microcystis proliferation occurred more frequently between the consumers and their foods than between the aquatic taxa without consumer-food relationship. Moreover, trophic cascades more strongly drove these synchrony dynamics in middle and high trophic-levels, while synchrony in low trophic-levels was mainly affected directly by Microcystis . Overall, our results reveal ecosystem stability relies strongly on trophic linkages and provide a pioneering framework for integrating extensive metagenomic datasets to enhance our understanding of ecosystem stability across broad taxa spectrum. Key words: Microcystis ; aquatic ecosystem stability; environmental DNA; synchrony; human activity; aquatic food web; cyanobacterial bloom Introduction Cyanobacteria are the key base of aquatic food webs, but can sometimes cause harmful effects on aquatic ecosystems when they form blooms (Ho et al., 2019). Previous studies have demonstrated that harmful cyanobacterial blooms can have negative impacts on a range of aquatic organisms (e.g., bacteria, protozoa, zooplankton, fish and hydrophytes), by producing toxic substances, increasing turbidity and depleting dissolved oxygen (Huisman et al., 2018, Liu et al., 2019). However, ecological communities are not just assemblages of different species, but contain multiple interactions between each other. These interactions form ecological networks that underpin the structure of biodiversity and ecological functions. Among these ecological networks, the food web represents one of the most important components. However, to date, how cyanobacterial proliferation affects the structure and stability of food webs is a fundamental but unresolved question (Novotny et al., 2023, Han et al., 2021). The cyanobacterial proliferation may have multiple impacts on food web stability. First, increased cyanobacterial abundance and associated aquatic environmental changes (e.g., elevated nutrient concentrations (eutrophication) and increased water turbidity) decreased environmental heterogeneity which lead to convergent adaptation of aquatic taxa in food webs, thereby synchronizing these taxa (Huisman et al., 2018, Zhang et al., 2021, Wu et al., 2021). Increased synchrony between different taxa is considered to reduce stability of ecosystems (Walter et al., 2024, Jin et al., 2025). For example, high synchrony among taxa can lead to drastic fluctuations in total community biomass (abundance) due to the lack of compensatory dynamics among taxa. High synchrony also increases extinct risk for species, because strong correlations between species make them more vulnerable to simultaneous decline (Mintrone et al., 2024). Second, the synchrony may propagate across trophic levels, leading to cascade effects. When food resources become synchronously abundant in a region, it may attract or support various consumers, thereby increasing the synchrony between these consumers. For instance, a previous study demonstrated that increased kelp detritus on coastal sandy beaches promoted spatial synchrony of shorebird abundances, as these shorebirds move between beaches to forage on invertebrates (the consumers of kelp detritus) (Walter et al., 2024). As shown in previous studies, cyanobacteria, other phytoplankton and bacteria (which utilize algal exudates) increased with the elevated nutrient concentration in water bodies (Cao et al., 2023). The increased food resources (cyanobacteria were also the import food resource for algivores (Novotny et al., 2023)) would stimulate the growth of various consumers (e.g., bacterivores, algivores), further promote the growth of various higher trophic-level predators, and ultimately increase the synchrony between taxa in food webs. Moreover, from the perspective of top-down effects, cyanobacterial proliferation is expected to impair foraging efficiency of visual predators, such as planktivorous and piscivorous fish, by increasing the turbidity and reducing light penetration within the water column (Zanghi and Ioannou 2025). A previous study found that largemouth bass ( Micropterus salmoides ), primarily a visual predator, exhibited a prey capture success rate of 15% in high turbidity environment (250 NTU), whereas success reached 100% in clear environment (0 NTU) (Huenemann et al., 2012, Zhang et al., 2025). Therefore, a reasonable inference is that the impairment in foraging efficiency compels predators to spend more energy on foraging activities. The increased energetic demands further cause predators to require higher feeding rates, thereby enhancing top-down control and further synchronizing the predators and prey. Third, with increase of cyanobacterial abundance, the extent of synchrony between species may be different at different trophic levels. Due to their more similar feeding habits and lifestyles, the top predators (e.g., piscivorous fish) may exhibit higher synchrony in response to synchronous food resources. In contrast, non-piscivorous fish and bottom trophic-level taxa (e.g., microbes, plankton, benthic invertebrates) exhibit greater diversity in feeding habits, lifestyles, and habitats. For example, non-piscivorous fish feed on diverse food resources (e.g., herbivorous fish, omnivorous fish, filter-feeding fish). In general, although current studies have demonstrated that cyanobacterial proliferation has strong influences on aquatic organisms, most of them have focused on species composition, diversity and abundance. A comprehensive understanding of how cyanobacterial proliferation affects aquatic food webs remains largely unknown (Srednick and Swearer 2024). In addition, current understanding on the effects of cyanobacterial proliferation on structure and stability of aquatic food webs have mostly been based on short-term controlled experiments or limited range of spatiotemporal scales that consider a narrow range of taxa spectrum (Bandara et al., 2025, Novotny et al., 2023). Whether these effects and mechanisms show consistency in natural aquatic food webs across multiple trophic levels (from microorganisms to macroorganisms) at large spatial scales remains unclear. Recently, environmental DNA (eDNA) applications have been significantly expanded by metagenomic sequencing and public taxonomic reference database (e.g. Non-Redundant Protein Sequence (NR) Database and Nucleotide Sequence (NT) Database), which can rapidly assess almost entire biomes in a single sequencing effort. By using this approach, researchers assessed biodiversity across whole biomes in airborne eDNA (Nousias et al., 2025), and revealed marine ecosystem shifts based on a broad taxa spectrum (ranging from cyanobacteria to marine mammals) (Zimmermann et al., 2023). In this study, 1,309 metagenomic samples were collected from previous studies, covering most provinces across China and spanning more than 6,000 kilometers. The sample types included freshwater lakes, reservoirs, rivers and wetlands. After subsampling to achieve uniform sequencing depth, the retained sequences were quality-filtered and taxonomically classified using Kraken2 with core-NT database. In recent decades, eutrophication and cyanobacterial blooms have become one of the most critical environmental challenges in China due to rapid economic growth and increasing human populations. Microcystis is one of the most pervasive bloom-forming cyanobacterial genera in global freshwater ecosystems (Harke et al., 2016). In China, a previous study showed that over 80% of cyanobacterial blooms in 115 representative lakes and large reservoirs were caused by Microcystis alone or mixture of Microcystis and other harmful cyanobacterial species (Huo et al., 2020). Although numerous studies have investigated the impacts of Microcystis proliferation on species turnover and diversity of aquatic communities, to our knowledge, no study has examined its effects on aquatic food webs on continental scale. Therefore, based on experimentally verified trophic interactions in previous studies, we constructed the China aquatic food webs that span multiple trophic levels from bacteriophages to piscivorous fish. Our aim is to obtain the direct evidence that effects of Microcystis proliferation on stability of aquatic food webs across a broad taxa spectrum on continental scale.

Data collection The metagenomic sequence data were obtained from previous studies. Specifically, we conducted a comprehensive literature search in Web of Science (http://www.webofknowledge.com) to identify relevant peer-reviewed journal articles from January 2015 to December 2024. Although the high-throughput sequencing for metagenome has been utilized for decades, it has undergone rapid development in the past ten years, with significant improvements in sample size, sequencing depth and quality. We used various keyword combinations, including China + lake + metagenome, China + river + metagenome, China + reservoir + metagenome, China + wetland + metagenome, China + freshwater + metagenome and China + aquatic + metagenome. Moreover, the samples from water bodies with salinity > 10‰ were excluded from subsequent analyses. According to the accession number, we then downloaded the metagenomic sequence data from the National Center for Biotechnology Information (NCBI) database (https://www.ncbi.nlm.nih.gov) and the Genome Sequence Archive (GSA) database (https://ngdc.cncb.ac.cn/gsa). In total, we obtained 1,309 metagenomic samples (data), representing most provinces across China. 853 sample were obtained from rivers, 274 from lakes, 135 from wetlands and 47 from reservoirs. The detailed information about these samples was listed in Table S1. Metagenomic data processing and annotation To normalize sequencing efforts, we randomly subsampled 37,500,000 read pairs from each sample by using seqtk (v1.4). After normalization, reads were filtered using Trimmomatic (v0.39) to generate high-quality data (LEADING: 3, TRAILING: 3, SLIDINGWINDOW: 5:20, MINLEN: 50) (Bolger et al., 2014). The remaining reads were then processed with FastUniq (v1.1) to remove duplicate reads. For the taxonomic classification of reads, we used the k-mer based classifier Kraken2 (v2.1.3) in the paired mode for paired-end reads (Lu et al., 2022, Zimmermann et al., 2023, Wood et al., 2019). We downloaded the National Center for Biotechnology Information (NCBI) core non-redundant nucleotide database and the NCBI taxonomy (obtained using kraken2-build in September 4, 2024. https://benlangmead.github.io/aws-indexes/k2) for taxonomic classification and built the Kraken2 database. The outputs from Kraken2 were further estimated for abundance at the species or higher-level taxa by using Bracken (v2.9). It estimates species (or higher-level taxa) abundances in metagenomics samples by probabilistically re-distributing reads in the taxonomic tree (Lu et al., 2017). To eliminate the false positive taxa, the annotation results for fish and hydrophytes were further verified against the Catalogue of Life China 2024 (http://www.sp2000.org.cn/download), Chinese Fish Database (https://www.cnfishbase.cn/), China Wetland Plant Database (https://zgsdzw.com/open/habitat/zszw), China Hydrophyte Species Database (http://nsii.org.cn/2017/minglu/shuisheng.html) and previous relevant studies (Kang et al., 2014). Moreover, we searched the recent literatures on Web of Science to identify the major harmful cyanobacteria (HC) in China, such as species in genera Microcystis, Aphanizomenon, Cylindrospermopsis, Dolichospermum, Anabaena, Nostoc, Planktothrix and Pseudanabaena (Table S2). Food web construction According to the food web construction approach in previous studies (Gauzens et al., 2020, D’Alessandro and Mariani 2021, Boyse et al., 2025), we systematically searched for peer-reviewed journal articles in Web of Science and China National Knowledge Infrastructure (https://www.cnki.net) to obtain the demonstrated trophic interactions (i.e. binary links in the food webs). In total, 73 scientific articles based on experimental verification of trophic interactions were identified for our food web construction. For simplification, we eliminated cannibalistic, parasitic (except for bacteriophages and bacteria) or symbiotic interactions. We first selected six piscivorous fish species. Due to the extremely low abundance and presence frequency in the metagenomic data of this study, other piscivorous fish were eliminated from further analyses. We then identified the prey for the six piscivorous fish species from the literatures. If the prey were also detected in the metagenomic data, we included them in the food web diagram. Then, we further identified the foods for the detected prey species from the literatures. The other non-piscivorous fish were classified into other omnivorous fish according to their feeding habits. The benthic invertebrates, zooplankton, harmful cyanobacteria (HC), phytoplankton (excluding HC), protozoa, bacteria and bacteriophages were not classified into lower taxonomic levels because of their high diversity and the non-selective feeding behavior of their consumers. To simplify the food web, hydrophytes were also not classified into lower taxonomic levels. The demonstrated binary links in the food webs and their corresponding references were shown in Table S3. Source of anthropogenic data According to the sampling year and location (province), we downloaded the anthropogenic data (i.e., annual gross domestic product (GDP), population density, annual total nitrogen emission and annual total phosphorus emission) in different provinces of China from the National Bureau of Statistics (https://www.stats.gov.cn). All these data were shown in Table S1. Synchrony calculation We used a moving window of 100 samples (step width = 10 samples) to compute the synchrony between any two taxa in the food webs. All the metagenomic samples were initially sorted in ascending order of organism read number (e.g the read number of Microcystis ), and then calculated the synchrony between any two taxa in the food webs along the gradient of increased organism read number (e.g the read number of Microcystis ) by using moving window method. For bacteriophages, bacteria, HC, phytoplankton (excluding HC), zooplankton, protozoa, benthic invertebrates, Decapoda and other omnivorous fish, we directly used its total abundance (e.g. zooplankton abundance) to calculate synchrony. The synchrony φ was quantified as (Loreau and de Mazancourt 2008) \begin{equation} \varphi=\sigma^{2}/(\sigma_{1}+\sigma_{2})^{2}\nonumber \\ \end{equation} σ 2 is the variance of total abundance of the two taxa. σ 1 and σ 2 are the standard deviation of abundance for taxon 1 and 2, respectively. The synchrony ranges from 0 (perfectly asynchronous) to 1 (perfectly synchronous). To assess the robustness of our synchrony estimates, we further calculated the pairwise synchrony with window width 100 samples and step width 5 samples, and with window width 50 samples and step width 10 samples. All these statistical analyses were performed in R (v4.3.1) (2020). Structural equation model (SEM) construction The relationships among anthropogenic factors, Microcystis abundance and pairwise synchrony were assessed in a structural equation model (SEM). We started with an initial model that included all plausible pathways between the factors (Fig. S5). Subsequently, the significance of each path coefficient was tested by its critical ratio, and nonsignificant paths were removed stepwise until all remaining paths were significant ( P < 0.05) (Kline 2005). The overall fit of the final model was evaluated by the goodness-of-fit index (GFI), the Bentler comparative fit index (CFI) and Chi-squared test (Kline 2005). The SEM analysis was performed using the software package AMOS version 28. Abundance of harmful cyanobacteria (HC) and effects of anthropogenic factors on abundance of aquatic taxa The genus Microcystis was the dominant harmful cyanobacteria (HC) in China’s freshwater ecosystems (Fig. 1a-1c). It exhibited significantly higher abundance (read counts) and relative abundance (percentage of Microcystis reads to total phytoplankton reads) in the east of the Hu line than in the west (Fig. 1a and 1b) (over 90% of Chinese people live in the east of the Hu line). Among the HC, the abundance and relative abundance of Microcystis were 1.63 ×10 8 and 37.4%, followed by Cylindrospermopsis (2.81 ×10 7 and 6.5%), Planktothricoides (1.72 ×10 7 and 3.9%) and Pseudanabaena (1.49 ×10 7 and 3.4%) (Fig. 1c). The Microcystis abundance showed strongly positive correlation with the annual gross domestic product (GDP) (r s = 0.383, P < 0.01), population density (r s = 0.380, P < 0.01), annual total nitrogen (TN) emission (r s = 0.428, P < 0.01) and annual total phosphorus (TP) emission (r s = 0.260, P < 0.01) across provinces (Fig. 1d and S1). Moreover, all these anthropogenic factors showed strong positive correlations with bacteriophage abundance, whereas most of these anthropogenic factors showed weak positive correlations with abundance of protozoa, phytoplankton (excluding HC) and other HC (excluding Microcystis ). However, most of these anthropogenic factors showed significantly negative correlations with abundance of fish, hydrophytes, benthic invertebrates, bacteria and zooplankton (Fig. 1d and S1). Correlations between abundance of Microcystis and abundance of other aquatic taxa According to the generalized additive model (GAM), the Microcystis abundance showed hump-shaped responses to the abundance of hydrophytes, protozoa, zooplankton, benthic invertebrates and fish (Fig. 2). Moreover, the Microcystis abundance showed clearly positive correlation with abundance of bacteriophages, phytoplankton (excluding HC) and Other HC (excluding Microcystis ), while showed clearly negative correlations with bacterial abundance (Fig. 2). Effects of elevated abundance of aquatic taxa on synchrony between taxa in the food webs We used sliding window analysis (window width 100 samples and step width 10 samples) to calculate the synchrony between any two taxa in the food webs (Fig. 3 and 4) (see Fig. 3g for the food webs construction). We found that with increase of Microcystis abundance, 63.1% of taxa pairs showed significantly enhanced synchrony (spearman correlation, P < 0.05), followed by increase of phytoplankton abundance (excluding HC, 48.3%), hydrophyte abundance (40.0%), benthic invertebrate abundance (36.0%), fish abundance (29.8%), zooplankton abundance (29.5%), protozoa abundance (28.6%), other HC abundance (26.5%), bacteriophage abundance (25.8%) and bacterial abundance (17.5%) (Fig. 3a). We further extracted the consumer-food synchrony from Fig. 3a (i.e., synchrony between any one consumer and one of its foods). We found that with increase of Microcystis abundance, 74.5% of consumer-food pairs showed significantly enhanced synchrony, followed by increase of phytoplankton abundance (excluding HC, 49.1%), hydrophyte abundance (40.9%), benthic invertebrate abundance (39.1%), protozoa abundance (33.6%), fish abundance (31.8%), bacteriophage abundance (28.2%), zooplankton abundance (27.3%), other HC abundance (21.8%) and bacterial abundance (13.6%) (Fig. 3b). Similarly, we extracted the synchrony between any two taxa at same trophic levels from Fig. 3a (i.e. trophic-level 1 to 3, see Fig. 3g). We found that with increase of Microcystis abundance, 63.2% of taxa pairs showed significantly enhanced synchrony, followed by increase of phytoplankton abundance (excluding HC, 48.1%), hydrophyte abundance (36.8%), zooplankton abundance (36.8%), benthic invertebrate abundance (35.8%), other HC abundance (31.1%), fish abundance (21.7%), protozoa abundance (18.9%), bacterial abundance (18.9%) and bacteriophage abundance (17.9%) (Fig. 3c). Moreover, when we specifically examined the consumer-food pair between any one piscivorous fish (trophic-level 3) and one of its preys, with increase of Microcystis abundance, 76.4% of consumer-food pairs showed significantly enhanced synchrony (Fig. 3e). Similarly, when we focused on the consumer-food pair between any one non-piscivorous fish (trophic-level 2) and one of its foods, with the increase of Microcystis abundance, 72.7% of consumer-food pairs showed significantly enhanced synchrony (Fig. 3e). Additionally, at the highest trophic level (trophic-level 3), with the increase of Microcystis abundance, 93.3% of taxa pairs exhibited significantly enhanced synchrony. At the middle trophic level (trophic-level 2), with the increase of Microcystis abundance, 60.0% of taxa pairs exhibited significantly enhanced synchrony. At the bottom trophic level (trophic-level 1), with the increase of Microcystis abundance, 55.6% of taxa pairs exhibited significantly enhanced synchrony (Fig. 3f). In addition, with increase of Microcystis abundance, the enhanced synchrony phenomenon was more commonly observed in the consumer-food pairs (74.5% of the pairs showed enhanced synchrony, n = 110 pairs) and taxa pairs in same trophic level (63.2% of the pairs showed enhanced synchrony, n = 106 pairs) (Fig. 3b and 3c). In contrast, with increase of Microcystis abundance, the enhanced synchrony phenomenon was less commonly observed in other taxa pairs ( the two taxa belong to different trophic levels and have no consumer-food relationship; 55.6% of these pairs showed enhanced synchrony, n = 133 pairs) (Fig. 3d). We further used food web graphic to show the synchrony trends (between consumers and their foods) along the gradient of increased Microcystis abundance (Fig. 3g), and used network graphic to show the synchrony trends (between any two taxa at same trophic levels) along the gradient of increased Microcystis abundance (Fig. 3h: trophic-level 3, Fig. 3i: trophic-level 2, Fig. 3j: trophic-level 1). To assess the robustness of our synchrony estimates, we further calculated the pairwise synchrony with window width 100 samples and step width 5 samples (Fig. S2), and with window width 50 samples and step width 10 samples (Fig. S3). Moreover, to assess the influences of different sampling seasons on our results, we further calculated the pairwise synchrony using samples only collected in warm months (May to September; window width 100 samples and step width 10 samples; n = 681) (Fig. S4). All these calculations consistently demonstrated that compared to other taxa, the increased Microcystis abundance strongly enhanced the pairwise taxa synchrony in the food webs. Factors that influenced the taxa synchrony along the gradient of increased Microcystis abundance. We used structural equation models (SEM) to further explore the factors that influenced the taxa synchrony along the gradient of increased Microcystis abundance. All significantly enhanced synchrony along the gradient of increased Microcystis abundance were extracted for this analysis. The consumer-food synchrony (between trophic-level 3 and their foods) was more strongly influenced by consumer-food synchrony (between trophic-level 2 and their foods) (correlation coefficients = 0.753) than by Microcystis abundance (0.221) (Fig. 5a). In contrast, the consumer-food synchrony (between trophic-level 2 and their foods) was more strongly influenced by Microcystis abundance (0.718) than by consumer-food synchrony (between trophic-level 3 and their foods) (0.404) (Fig. 5b). Synchrony between trophic-level 3 taxa was more strongly influenced by synchrony between trophic-level 2 taxa (0.909) than by Microcystis abundance (0.096). Similarly, synchrony between trophic-level 2 taxa was more strongly influenced by synchrony between trophic-level 1 taxa (0.701) and synchrony between trophic-level 3 taxa (0.949) than by Microcystis abundance (0.233) (Fig. 5c and 5d). In contrast, synchrony between trophic-level 1 taxa was more strongly influenced by Microcystis abundance (0.794) than by synchrony between trophic-level 2 taxa (0.468) (Fig. 5d).

Among the water bodies investigated, the dominated harmful cyanobacterial taxa were Microcystis (Fig. 1a-1c). As the most common bloom-forming cyanobacteria, Microcystis exhibits remarkable phenotypic plasticity (Harke et al., 2016, Dick et al., 2021). Morphological advantages like colony formation and gas vesicles inside cells contribute to its widespread distribution and ecological dominance (Dick et al., 2021). Moreover, compared to other aquatic taxa in the food webs, we found Microcystis exhibited strongly positive correlations with anthropogenic factors (Fig. 1d and S1), indicating its close association with human activities. We further found that with increase of Microcystis abundance, most taxa in the food webs became more synchronous with other taxa (Fig. 3). Elevated synchrony between taxa has negative effects on ecosystem stability. The increased synchrony tends to cause drastic fluctuations of total community biomass (abundance) and increases the risk of species extinct on spatial and temporal scales, due to the lack of compensatory effects (a decline in one species is offset by an increase in another, thereby maintaining ecosystem stability) (Wilcox et al., 2017). Generally, along the gradient of elevated algal abundance (e.g. Microcystis ), the environmental factors in waters would change correspondingly to an identical direction, such as increased nutrient concentration (eutrophication), elevated phytoplankton biomass and elevated water turbidity. Previous studies have indicated that these changes lead to environmental and biotic homogenization (Geng et al., 2022, Liu et al., 2024, Liu et al., 2025), consequently inducing the convergent (synchronous) responses of aquatic taxa in the food webs to the identical environmental conditions. However, we found that the significantly enhanced synchrony along the gradient of increased Microcystis abundance was more commonly observed in the consumer-food pairs, and these synchrony in high trophic-levels were more strongly driven by their adjacent trophic-levels than by Microcystis abundance (Fig. 3 and 5). This suggests potential propagation of synchrony across trophic levels, leading to cascade effects. For example, a case study found that increased resources (i.e., detrital kelp biomass) on beaches synchronized local abundance of shorebirds, because these birds moved between beaches to forage for invertebrates that fed on these kelp subsidies (Walter et al., 2024). With increase of food resources, the consumer-resource interactions may enhance, and thus lead to the increased synchrony between them. According to the Holling type III functional response, when prey density is below a security threshold, predator feeding rates stay low because prey are more likely to hidden or was not fully detected by predators. However, when prey density exceeds this threshold, the feed rates increase markedly until saturation (Holling 1959). Compared to low prey density, predators eat more and grow faster at higher prey density (Murdoch 1971). For example, the predatory mosquito larva ( Toxorhynchites towadensis ) grew slowly under low prey density conditions, whereas they grew fast at high prey density (Yasuda and Hagimori 1998). Likewise, a previous study showed that intensity of fish ( Cheilodactylus nigripes ) attacking prey (measured by attack rate and duration) primarily depended on prey density, with higher attack intensity observed at higher prey densities (Wellenreuther and Connell 2002). In this study, we found the Microcystis abundance showed strong positive correlation with the phytoplankton abundance (excluding HC) (Fig. 2). Therefore, the increased abundance of Microcystis and other phytoplankton (primary production) likely gradually stimulated the consumer growth (e.g. algivorous zooplankton, protozoa and fish), ultimately supporting the growth of higher trophic levels. Therefore, increased Microcystis abundance may cause food resources to become synchronously abundant in a region. The abundant food resources attract or support various consumers, ultimately increasing the synchrony between taxa. In contrast to bottom-up control, the top-down control may also propagate the synchrony downward along the trophic levels. Normally, with increase of algal abundance, the water turbidity increased correspondingly. Algae-induced turbidity can decrease the foraging efficiency of visual predators (e.g., planktivorous and piscivorous fish) by hindering prey detection (Engström-Öst et al., 2006, Zanghi and Ioannou 2025) As a result, in environments with high turbidity, visual predators may expend more energy on foraging activities. This could intensify trophic interactions between visual predators and their prey, as the predators require higher feeding rates to meet their increased energetic demands. In fact, Zanghi and Ioannou collected 281 studies from peer-reviewed literatures and found that increased algae biomass is one of the key factors causing increase of water turbidity. Elevated turbidity levels had significantly negative impacts on predators due to impaired visibility (especially, the adult piscivorous fish, e.g., Perca fluviatilis ) (Zanghi and Ioannou 2025). Therefore, under high algae-induced turbidity, the interaction strength between visual predators and prey increases, which may lead to increased synchrony between these predators and prey. In addition, the Microcystis proliferation had different impacts on different trophic levels. For instance, 93.3% of pairwise synchrony between piscivorous fish (trophic-level 3) increased with rising Microcystis abundance, whereas this percentage was 60.0% for non-piscivorous fish (trophic-level 2) and 55.6% for bottom trophic levels (trophic-level 1: e.g., hydrophyte, zooplankton, phytoplankton, benthic invertebrates, bacteria and bacteriophages) (Fig. 3f). Communities with more diverse species (e.g., more species differ in their traits) generally exhibit lower synchrony, because species with different ecological strategies tend to fluctuate less synchronously than more similar species (Zhao et al., 2023, van Klink et al., 2019). In this study, the lifestyles, feeding habits and habitats of non-piscivorous fish and bottom trophic-level taxa are more diverse than those of piscivorous fish. For example, we found that with increase of Microcystis abundance, the significantly decreased and non-significant synchrony between non-piscivorous fish were primarily associated with other omnivory fish, filter feeding fish ( Hypophthalmichthys molitrix + Hypophthalmichthys nobilis ) and Misgurnus anguillicaudatus (they have more diverse feeding habits than piscivorous fish) (Fig. 3i). Likewise, the bottom trophic levels included diverse taxa groups, such as hydrophyte, zooplankton, phytoplankton, benthic invertebrates, bacteria and bacteriophages. In contrast, the six piscivorous fish species had similar feeding habits. As piscivorous top predators, due to their similar feeding habits and lifestyles, they were more vulnerable to the impacts of increased Microcystis abundance compared to taxa at other trophic levels. Top predators play a critical role in driving ecosystem stability. Their decline may favor small-sized fish, which is often accompanied by deterioration of water quality and ecosystem function. A experimental study demonstrated that small-sized fish (zooplanktivorous and omnivorous fish) increased biomass of phytoplankton and relative biomass of cyanobacteria through their predation on large-sized zooplankton (Guo et al., 2023). Our metagenomic data span across space and time, as the samples were collected in different years and sites. However, in this study, we used the abundance of biological taxa (e.g abundance of Microcystis ) as the driving factors that shaped the food web stability. Compared to a single environmental factor, the abundance of biological taxa reflects the integrated effects of multiple factors. For example, when investigating food web dynamics along a nutrient gradient, samples should be collected under similar water temperature, because temperature strongly influences the algal growth. In contrast, the abundance of biological taxa reflects the integrated effects of both temperature and nutrient centration. Therefore, although two samples with high Microcystis abundance were collected from different sites or in different years, they may have suffered similar impacts from the Microcystis proliferation.

In conclusion, by employing large-scale metagenomic data, we constructed a multiple trophic-level food web spanning a broad taxa spectrum. Our results demonstrated that elevated Microcystis abundance strongly enhanced synchrony among taxa in the food webs. Moreover, the potential cascade effects (potential propagation of synchrony across trophic levels) and vulnerability of top predators in food webs, exacerbated synchronous extent of taxa in food webs. This suggests that ecosystem stability relies more on indirect trophic linkages than direct environmental responses. Therefore, overlooking interaction among taxa may lead to an underestimation of anthropogenic disturbance effects on aquatic ecosystems (Yin et al., 2024, Srednick and Swearer 2024). Although food webs constitute the base of ecosystems, research on them has lagged behind species turnover studies in the molecular ecology era. Our findings highlight the necessity to further evaluate ecosystem responses to environmental disturbances through coupling eDNA metagenomic data and food-web construction. It should be noted that the food webs constructed in this study could not reflect the realistic predation dynamics due to the samples being collected from different times and water bodies. However, our results revealed that the anthropogenic disturbances strongly influenced the relationships between consumers and available food resources across spatiotemporal scales. Given the increasing amount of available metagenomic data from aquatic ecosystems, our work provides an idea for integration of large-scale eDNA metagenomic datasets to investigate the effects of anthropogenic disturbances on stability of ecological networks and functions across multiple trophic levels.

http://R-project.org/ R Core Team. R: a language and environment for statistical computing. .Bandara, T., S. Brugel, D. C. P. Lau, I. Jurgensone, E. Griniene, A. Garbaras, L. Butkus, L. Bučinskas, E. Fridolfsson, J. Carstensen & A. Andersson (2025) Blooms of filamentous cyanobacteria reduce zooplankton quality and pelagic food web efficiency. Limnol. Oceanogr. , 1–15.Bolger, A. M., M. Lohse & B. Usadel (2014) Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics, 30 , 2114–2120.Boyse, E., K. P. Robinson, I. M. Carr, E. Valsecchi, M. Beger & S. J. Goodman (2025) Inferring species interactions from co-occurrence networks with environmental DNA metabarcoding data in a coastal marine food web. Mol. Ecol., 34 , e17701.Cao, J., Z. Y. Hou, Z. K. Li, B. H. Zheng & Z. S. Chu (2023) Spatiotemporal dynamics of phytoplankton biomass and community succession for driving factors in a meso-eutrophic lake. J. Environ. Manage., 345 , 118693.D’Alessandro, S. & S. Mariani (2021) Sifting environmental DNA metabarcoding data sets for rapid reconstruction of marine food webs. Fish Fish., 22 , 822–833.Dick, G. J., M. B. Duhaime, J. T. Evans, R. M. Errera, C. M. Godwin, J. J. Kharbush, H. S. Nitschky, M. A. Powers, H. A. Vanderploeg, K. C. Schmidt, D. J. Smith, C. E. Yancey, C. C. Zwiers & V. J. Denef (2021) The genetic and ecophysiological diversity of Microcystis . Environ. Microb., 23 , 7278–7313.Engström-Öst, J., M. Karjalainen & M. Viitasalo (2006) Feeding and refuge use by small fish in the presence of cyanobacteria blooms. Environ. Biol. Fishes, 76 , 109–117.Gauzens, B., B. C. Rall, V. Mendonça, C. Vinagre & U. Brose (2020) Biodiversity of intertidal food webs in response to warming across latitudes. Nat. Clim. Chang., 10 , 264–269.Geng, M., W. Zhang, T. Hu, R. Wang, X. Cheng & J. Wang (2022) Eutrophication causes microbial community homogenization via modulating generalist species. Water Res., 210 , 118003.Guo, C., S. Li, J. Ke, C. Liao, A. G. Hansen, E. Jeppesen, T. Zhang, W. Li & J. Liu (2023) The feeding habits of small-bodied fishes mediate the strength of top-down effects on plankton and water quality in shallow subtropical lakes. Water Res., 233 , 119705.Han, M., C. Dong, S. Ma, C. Feng, C. Lei, Z. Gu & X. Liu (2021) Food web responses to a cyanobacterial bloom in a freshwater eutrophic lake. Water, 13 , 1296.Harke, M. J., M. M. Steffen, C. J. Gobler, T. G. Otten, S. W. Wilhelm, S. A. Wood & H. W. Paerl (2016) A review of the global ecology, genomics, and biogeography of the toxic cyanobacterium, Microcystis spp. Harmful Algae, 54 , 4–20.Ho, J. C., A. M. Michalak & N. Pahlevan (2019) Widespread global increase in intense lake phytoplankton blooms since the 1980s. Nature, 574 , 667–670.Holling, C. S. (1959) Some characteristics of simple types of predation and parasitism. Canadian Entomologist, 91 , 385–398.Huenemann, T. W., E. D. Dibble & J. P. Fleming (2012) Influence of turbidity on the foraging of largemouth bass. Trans. Am. Fish. Soc., 141 , 107–111.Huisman, J., G. A. Codd, H. W. Paerl, B. W. Ibelings, J. M. H. Verspagen & P. M. Visser (2018) Cyanobacterial blooms. Nat. Rev. Microbiol., 16 , 471–483.Huo, D., N. Gan, R. Geng, Q. Cao, L. Song, G. Yu & R. J. H. A. Li (2020) Cyanobacterial blooms in China: diversity, distribution, and cyanotoxins. Harmful Algae, 109 , 102106.Jin, Y., K. Zhang, Q. Lin, S. Huang, X. Zhou, J. Ren & M. E. Meadows (2025) Aquatic community synchrony dynamics in response to multitrophic regime shifts: Paleoecological insights from a shallow lake. Water Res., 287 , 124402.Kang, B., J. Deng, Y. Wu, L. Chen, J. Zhang, H. Qiu, Y. Lu & D. He (2014) Mapping China’s freshwater fishes: diversity and biogeography. Fish Fish., 15 , 209-230.Kline, R. B. 2005. Principles and Practice of Structural Equation Modelling . New York, USA: Guilford Press.Liu, L., H. Chen, M. Liu, J. R. Yang, P. Xiao, D. M. Wilkinson & J. Yang (2019) Response of the eukaryotic plankton community to the cyanobacterial biomass cycle over 6 years in two subtropical reservoirs. ISME J., 13 , 2196–2208.Liu, L., H. Wang, Y. Guo, Q. Yan & J. Chen (2024) Human-induced homogenization of microbial taxa and function in a subtropical river and its impacts on community stability. Water Res., 252 , 121198.Liu, L., S. Zhu, Y. Gu, S. Tao, B. Krock, C. Wang, X. Shi, Q. Yan, X. Pan, J. Chen, S. Lin & Z. Luo (2025) Holocene and contemporary marine dinoflagellate community patterns predict expansion of generalist dinoflagellate blooms in warming oceans. ISME J., 19 , wraf095.Loreau, M. & C. de Mazancourt (2008) Species synchrony and Its drivers: neutral and nonneutral community dynamics in fluctuating environments. Am. Nat., 172 , E48–E66.Lu, J., F. P. Breitwieser, P. Thielen & S. L. Salzberg (2017) Bracken: estimating species abundance in metagenomics data. Peer J Comput. Sci., 3 , e104.Lu, J., N. Rincon, D. E. Wood, F. P. Breitwieser, C. Pockrandt, B. Langmead, S. L. Salzberg & M. Steinegger (2022) Metagenome analysis using the Kraken software suite. Nat. Protoc., 17 , 2815–2839.Mintrone, C., L. Rindi & L. Benedetti-Cecchi (2024) Stabilizing effects of spatially heterogeneous disturbance via reduced spatial synchrony on a rocky shore community. Ecology, 105 , e4246.Murdoch, W. W. (1971) The developmental response of predators to changes in prey density. Ecology, 52 , 132–137.Nousias, O., M. McCauley, M. R. Stammnitz, J. A. Farrell, S. A. Koda, V. Summers, C. B. Eastman, F. G. Duffy, I. J. Duffy, J. Whilde & D. J. Duffy (2025) Shotgun sequencing of airborne eDNA achieves rapid assessment of whole biomes, population genetics and genomic variation. Nat. Ecol. Evol., 9 , 1043–1060.Novotny, A., B. Serandour, S. Kortsch, B. Gauzens, K. M. G. Jan & M. Winder (2023) DNA metabarcoding highlights cyanobacteria as the main source of primary production in a pelagic food web model. Sci. Adv., 9 , eadg1096.Srednick, G. & S. E. Swearer (2024) Understanding diversity–synchrony–stability relationships in multitrophic communities. Nat. Ecol. Evol., 8 , 1259–1269.van Klink, R., J. Lepš, R. Vermeulen & F. de Bello (2019) Functional differences stabilize beetle communities by weakening interspecific temporal synchrony. Ecology, 100 , e02748.Walter, J. A., K. A. Emery, J. E. Dugan, D. M. Hubbard, T. W. Bell, L. W. Sheppard, V. A. Karatayev, K. C. Cavanaugh, D. C. Reuman & M. C. N. Castorani (2024) Spatial synchrony cascades across ecosystem boundaries and up food webs via resource subsidies. Proc. Natl. Acad. Sci. USA, 121 , e2310052120.Wellenreuther, M. & S. D. Connell (2002) Response of predators to prey abundance: separating the effects of prey density and patch size. J. Exp. Mar. Biol. Ecol., 273 , 61–71.Wilcox, K. R., A. T. Tredennick, S. E. Koerner, E. Grman, L. M. Hallett, M. L. Avolio, K. J. La Pierre, G. R. Houseman, F. Isbell, D. S. Johnson, J. M. Alatalo, A. H. Baldwin, E. W. Bork, E. H. Boughton, W. D. Bowman, A. J. Britton, J. F. Cahill Jr., S. L. Collins, G. Du, A. Eskelinen, L. Gough, A. Jentsch, C. Kern, K. Klanderud, A. K. Knapp, J. Kreyling, Y. Luo, J. R. McLaren, P. Megonigal, V. Onipchenko, J. Prevéy, J. N. Price, C. H. Robinson, O. E. Sala, M. D. Smith, N. A. Soudzilovskaia, L. Souza, D. Tilman, S. R. White, Z. Xu, L. Yahdjian, Q. Yu, P. Zhang & Y. Zhang (2017) Asynchrony among local communities stabilises ecosystem function of metacommunities. Ecol. Lett., 20 , 1534–1545.Wood, D. E., J. Lu & B. Langmead (2019) Improved metagenomic analysis with Kraken 2. Genome Biol., 20 , 257.Wu, Z., X. Lai & K. Li (2021) Water quality assessment of rivers in Lake Chaohu Basin (China) using water quality index. Ecol. Indic., 121 , 107021.Yasuda, H. & K. Hagimori (1998) Effects of prey density, prey instar, and patch size on the development of the predatory mosquito, Toxorhynchites towadensis . Entomol. Exp. Appl., 86 , 189–195.Yin, J., C. Duan, F. Zhou, L. Gong, M. D. K. L. Gunathilaka, X. Liu, D. Liu, A. Shen & Y. Pan (2024) Microplastics affect interspecific interactions between cladoceran species in the absence and presence of predators by triggering asymmetric individual responses. Water Res., 248 , 120877.Zanghi, C. & C. C. Ioannou (2025) The impact of increasing turbidity on the predator–prey interactions of freshwater fishes. Freshwater Biol., 70 , e14354.Zhang, H., S. Huo, X. Cao, C. Ma, J. Zhang & F. Wu (2021) Homogenization of reservoir eukaryotic algal and cyanobacterial communities is accelerated by dam construction and eutrophication. J. Hydrol., 603 , 126842.Zhang, K., Z. Ye, M. Qi, W. Cai, J. L. Saraiva, Y. Wen, G. Liu, Z. Zhu, S. Zhu & J. Zhao (2025) Water quality impact on fish behavior: A review from an aquaculture perspective. Rev. Aquacult., 17 , e12985.Zhao, Q., P. J. Van den Brink, C. Xu, S. Wang, A. T. Clark, C. Karakoç, G. Sugihara, C. E. Widdicombe, A. Atkinson, S.-i. S. Matsuzaki, R. Shinohara, S. He, Y. X. G. Wang & F. De Laender (2023) Relationships of temperature and biodiversity with stability of natural aquatic food webs. Nat. Commu., 14 , 3507.Zimmermann, H. H., K. R. Stoof-Leichsenring, V. Dinkel, L. Harms, L. Schulte, M.-T. Hütt, D. Nürnberg, R. Tiedemann & U. Herzschuh (2023) Marine ecosystem shifts with deglacial sea-ice loss inferred from ancient DNA shotgun sequencing. Nat. Commu., 14 , 1650. Supplementary materials Fig. S1 Spearman’ correlations show the relationships between abundance of aquatic taxa and anthropogenic factors. Fig. S2 a) Sliding window analysis (window width = 100 samples, step width = 5 samples) shows the proportion of three trends of synchrony: with increase of aquatic taxa abundance (e.g. Microcystis ), synchrony between any two taxa significantly increased (trend 1), significantly decreased (trend 2) and non-significant (trend 3) (Fig. S2a = Fig. S2b + Fig. S2c + Fig. S2d). b) extracted from Fig. S2a. Proportion of three trends of synchrony (between consumers and their foods). c) extracted from Fig. S2a. Proportion of three trends of synchrony (between taxa at same trophic-levels. See Fig. 3g for the three trophic-levels). d) extracted from Fig. S2a. The remaining synchrony. e) Proportion of three trends of synchrony (between trophic-level 3 taxa and their foods, and between trophic-level 2 taxa and their foods), along the gradient of increased Microcystis abundance. f) Proportion of three trends of synchrony (between taxa at trophic-level 3, trophic-level 2 and trophic-level 1), along the gradient of increased Microcystis abundance. Fig. S3 a) Sliding window analysis (window width = 50 samples, step width = 10 samples) shows the proportion of three trends of synchrony: with increase of aquatic taxa abundance (e.g. Microcystis ), synchrony between any two taxa significantly increased (trend 1), significantly decreased (trend 2) and non-significant (trend 3) (Fig. S3a = Fig. S3b + Fig. S3c + Fig. S3d). b) extracted from Fig. S3a. Proportion of three trends of synchrony (between consumers and their foods). c) extracted from Fig. S3a. Proportion of three trends of synchrony (between taxa at same trophic-levels. See Fig. 3g for the three trophic-levels). d) extracted from Fig. S3a. The remaining synchrony. e) Proportion of three trends of synchrony (between trophic-level 3 taxa and their foods, and between trophic-level 2 taxa and their foods) and their foods), along the gradient of increased Microcystis abundance. f) Proportion of three trends of synchrony (between taxa at trophic-level 3, trophic-level 2 and trophic-level 1), along the gradient of increased Microcystis abundance. Fig. S4 a) Sliding window analysis (only samples in warm months (May to September) were selected; window width = 100 samples, step width = 10 samples) shows the proportion of three trends of synchrony: with increase of aquatic taxa abundance (e.g. Microcystis ), synchrony between any two taxa significantly increased (trend 1), significantly decreased (trend 2) and non-significant (trend 3) (Fig. S4a = Fig. S4b + Fig. S4c + Fig. S4d). b) extracted from Fig. S4a. Proportion of three trends of synchrony (between consumers and their foods). c) extracted from Fig. S4a. Proportion of three trends of synchrony (between taxa at same trophic-levels. See Fig. 3g for the three trophic-levels). d) extracted from Fig. S4a. The remaining synchrony. e) Proportion of three trends of synchrony (between trophic-level 3 taxa and their foods, and between trophic-level 2 taxa and their foods), along the gradient of increased Microcystis abundance. f) Proportion of three trends of synchrony (between taxa at trophic-level 3, trophic-level 2 and trophic-level 1), along the gradient of increased Microcystis abundance. Fig. S5 Initial structural equation model (SEM) that included all plausible pathways between the factors. a) factors that influenced synchrony between consumers and foods. b) factors that influenced synchrony between taxa at same trophic levels (See Fig. 3g for the three trophic-levels). GDP: annual gross domestic product (10 4 yuan·km -2 ·a -1 ), population density (person·km -2 ), TN: annual total nitrogen emission (ton·km -2 ·a -1 ), TP: annual total phosphorus emission (ton·km -2 ·a -1 )). Table S1 General information of metagenomic data and anthropogenic factors. Table S2 List of harmful cyanobacterial species reported in previous studies. Table S3 Demonstrated binary links in the food webs and their corresponding references. Figures Fig. 1 Sampling map and characteristics of harmful cyanobacteria. a) Map of sampling sites (dot size indicates Microcystis abundance (read counts)). b) Microcystis abundance and relative abundance (percentage of Microcystis reads to total phytoplankton reads) in the east and west of the Hu line (over 90% of Chinese people live in the east of the Hu line). c) Abundance and relative abundance of different harmful cyanobacteria (HC). d) Spearman’ correlations show the relationships between aquatic taxa abundance and anthropogenic factors. GDP: annual gross domestic product (10 4 yuan·km -2 ·a -1 ), population density (person·km -2 ), TN: annual total nitrogen emission (ton·km -2 ·a -1 ), TP: annual total phosphorus emission (ton·km -2 ·a -1 ). * P < 0.05, ** P < 0.01 (Mann-Whitney U test). Phytoplankton did not include HC. Fig. 2 Generalized additive models show the relationships between Microcystis abundance and abundance of other taxa in the aquatic ecosystems. Phytoplankton did not include HC. Fig. 3 a) Sliding window analysis (window width = 100 samples, step width = 10 samples) shows the proportion of three trends of synchrony: with increase of aquatic taxa abundance (e.g. Microcystis ), synchrony between any two taxa significantly increased (trend 1), significantly decreased (trend 2) and non-significant (trend 3) (Fig. 3a = Fig. 3b + Fig. 3c + Fig. 3d). b) extracted from Fig. 3a. Proportion of three trends of synchrony (between consumers and their foods). c) extracted from Fig. 3a. Proportion of three trends of synchrony (between taxa at same trophic-levels. See Fig. 3g for the three trophic-levels). d) extracted from Fig. 3a. The remaining synchrony; e) Proportion of three trends of synchrony (between trophic-level 3 taxa and their foods, and between trophic-level 2 taxa and their foods), along the gradient of increased Microcystis abundance. f) Proportion of three trends of synchrony (between taxa at trophic-level 3, trophic-level 2 and trophic-level 1), along the gradient of increased Microcystis abundance. g) Food web constructed in this study shows the three trends of synchrony (between consumers their foods), along the gradient of increased Microcystis abundance. h-j) Network shows the three trends of synchrony, along the gradient of increased Microcystis abundance: h) between taxa at trophic-level 3, i) at trophic-level 2 and j) at trophic-level 1. Filter feeding: Hypophthalmichthys molitrix + Hypophthalmichthys nobilis . Decapoda: Procambarus clarkii + Macrobrachium rosenbergii . Fig. 4 Synchrony values between any two taxa in each sliding window (window width = 100 samples, step width = 10 samples). Synchrony was calculated along the gradient of elevated a) Microcystis abundance, b) bacteriophage abundance; c) bacterial abundance, d) phytoplankton (excluding HC) abundance, e) other HC abundance, f) hydrophyte abundance, g) protozoa abundance, h) zooplankton abundance, i) invertebrate abundance and j) fish abundance. Significantly increased, significantly decreased and non-significant synchrony were shown, respectively. * P < 0.05, ** P < 0.01 (Spearman’s correlation). Fig. 5 Structural equation models (SEM) show the factors that influenced the taxa synchrony along the gradient of increased Microcystis abundance. a and b) factors that influenced synchrony between consumers and foods (a: bottom-up influence, b: top-down influence). c and d) factors that influenced synchrony between taxa at same trophic levels (a: bottom-up influence, b: top-down influence). GDP: annul gross domestic product, TP: annul total phosphorus emission. * P < 0.05, ** P < 0.01. Information & Authors Information Version history Copyright This work is licensed under a Non Exclusive No Reuse License.

Authors Metrics & Citations Metrics Article Usage 202views 175downloads Citations Download citation lemian Liu, Siqi Zhu, Xiaoqin Huang, et al. Continental scale metagenomics reveals Microcystis-induced synchrony between aquatic taxa is exacerbated by trophic cascades. Authorea. 22 December 2025. DOI: https://doi.org/10.22541/au.176643739.98425458/v1 DOI: https://doi.org/10.22541/au.176643739.98425458/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.