Long-term freshwater time series reveals recurrent and episodic microbial dynamics driven by distinct assembly mechanisms

Long-term freshwater time series reveals recurrent and episodic microbial dynamics driven by distinct assembly mechanisms | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Long-term freshwater time series reveals recurrent and episodic microbial dynamics driven by distinct assembly mechanisms Hongjae Park, Suhyun Kim, Miri. S. Park, Ilnam Kang, Jang-Cheon Cho This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7144747/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Freshwater bacteria are key contributors to ecosystem stability and resilience, yet their inherent microdiversity and rapid community turnover make interpreting their seasonal dynamics within broader ecological frameworks challenging. Furthermore, the underlying mechanisms of community assembly, driven by deterministic and stochastic processes, remain poorly understood. Results We conducted a five-year time series sampling of microbial communities, alongside key environmental parameters, in an oligo-mesotrophic lake. Seasonal community assembly in this freshwater ecosystem was primarily driven by drift, dispersal limitation, and homogeneous selection, with the relative contributions of these processes varying markedly across months and depths. Six distinct clusters of seasonally recurring microbes shaped the timeline of annual community transitions, while episodic populations sporadically occupied the remaining ecological niches. Network-based reconstruction of microbial dynamics uncovered a highly interconnected structure, where microbial and environmental components exhibited coordinated, time-dependent variation. Conclusions This study highlights the importance of long-term ecological monitoring for resolving microbial dynamics at a fine scale. Our findings establish a foundation for future approaches aiming to apply microbial metrics to freshwater ecosystem assessment and management. Freshwater ecosystem Microbial community Seasonal dynamics Ecological model Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Background Freshwater is essential for human life, supplying water for drinking, agriculture, industrial processes, and transportation [ 1 ]. In addition to serving human needs, freshwater habitats such as lakes, rivers, and ponds support diverse plant and animal species and maintain local biodiversity [ 2 ]. At the core of these ecosystems are freshwater bacteria, which play a pivotal role in nutrient cycling, breaking down organic matter, and forming the backbone of the aquatic food web [ 3 ]. Understanding their assembly mechanisms and spatiotemporal dynamics is key to assessing ecosystem health and ensuring long-term water quality management. Our understanding of microbial community assembly has been shaped by two conceptual frameworks: niche-based and neutral theories [ 4 ]. According to the niche theory, deterministic forces, including abiotic filtering and species interactions, play a dominant role in shaping community assembly by favoring taxa best suited to local environments [ 5 ]. In contrast, neutral theory emphasizes the role of stochastic processes, such as ecological drift, dispersal limitation, and homogenizing dispersal, under the assumption that all taxa are ecologically equivalent [ 6 ]. While once viewed as competing explanations, it is now widely accepted that both deterministic and stochastic processes operate in tandem [ 7 , 8 ], with the key challenge being to quantify their relative contributions across space and time. In freshwater ecosystems, the mechanisms underlying seasonal community assembly remain poorly understood [ 9 , 10 ]. The plankton ecology group (PEG) model has long been a cornerstone for interpreting the seasonal development of large planktonic organisms in lakes [ 11 ]. Recent efforts to extend this framework by integrating bacterial populations have shed light on their coordinated succession within the broader planktonic community [ 12 ]. However, despite these advances, a few blind spots remain to be resolved. For instance, winter microbial dynamics have been underexplored in many freshwater studies [ 12 , 13 ], resulting in an incomplete understanding of their annual transition. In addition, the temporal architecture of microbial community transformation might differ considerably among geographically distant lakes [ 14 ], emphasizing the need for cross-system comparisons. Alongside seasonally recurring bacterial groups, episodic populations that emerge unpredictably in response to environmental fluctuations deserve greater attention, especially in the context of long-term ecological change [ 15 ]. Addressing these gaps is essential for capturing the full spectrum of microbial dynamics and developing a generalized ecosystem model for freshwater environments. This study presents a five-year microbial time series from Lake Soyang, the largest freshwater lake in South Korea. By leveraging this long-term dataset, we examined how deterministic and stochastic processes interact to influence microbial community assembly over time. We further reconstructed a generalized timeline of microbial succession at lake epilimnion, identifying annually recurring groups with robust seasonal signatures, as well as episodic populations that emerged unpredictably. In addition, an association network uncovered a high degree of interconnectedness among microbial and environmental components, pointing to their coordinated succession as a collective outcome of complex interdependencies. Materials and Methods Study site and sample collection Over a five-year period, 300 water samples were collected monthly from the dam station of Lake Soyang (37°56'51.0"N 127°49'08.0"E). The sampling was carried out in two distinct phases. In the first phase (May 2015–May 2017), water was collected from depths of 1, 10, 20, 40, and 70 meters. In the second phase (January 2019–December 2021), the 70-meter sampling point was replaced by a 90-meter depth to capture near-bottom microbial communities. At each sampling event, 10–20 L of water was collected using a Niskin water sampler (General Oceanics, USA). To harvest microbial biomass, 1 L of water was first filtered through a 3-µm cellulose acetate membrane (Advantec, Japan) to remove coarse particles and larger organisms. The resulting filtrate was then passed through 0.22-µm polyethersulfone membranes (Pall, USA). Membrane filters containing the biomass were stored at − 80°C until further processing. Environmental measures Vertical profiles of temperature, pH, conductivity, and dissolved oxygen (DO) concentration were measured at the sampling station using a YSI 556 MPS Multiprobe system (YSI, USA). Chlorophyll a concentrations were determined using a UV-VIS spectrophotometer (Model UV-2600, Shimadzu, Japan), following extraction with acetone from a GF/F glass fiber filter (Whatman, USA). Additional physicochemical parameters, including ammonium, nitrite, nitrate, phosphate, and silicate, were measured from water filtered through a 0.45-µm mixed cellulose ester membrane (Advantec, Japan), using a QuAAtro microflow analyzer (SEAL Analytical, UK) during the first sampling phase, and a Hach reagent kit (Hach, USA) during the second phase. Genomic DNA extraction and 16S rRNA gene amplicon sequencing Genomic DNA was extracted using the PowerWater DNA isolation kit (Qiagen, Germany) for the first batch, and the DNeasy PowerSoil Pro Kit (Qiagen, Germany) for the second. DNA concentration was quantified using a Qubit fluorometer (Invitrogen, USA). PCR amplification was performed using the primer pair 515FY and 926R (targeting the V4–V5 region) for the first batch, and 515FY and 806RB (targeting the V4 region) for the second. Library preparation and sequencing were carried out at CJ Bioscience (Seoul, South Korea). Amplicon libraries were sequenced using the 2×300 bp (first batch) and 2×250 bp (second batch) paired-end chemistry on the Illumina MiSeq platform (Illumina, USA). Sequence processing Adapter and primer trimming was performed using Cutadapt [ 16 ], with 515FY and 806RB primer sequences removed from all reads to standardize downstream processing. Trimmed reads were further processed using the R package DADA2 v1.34.0 [ 17 ], following the standard pipeline available at https://benjjneb.github.io/dada2/tutorial.html . Taxonomic assignment of amplicon sequence variants (ASVs) was conducted using the SILVA database v138.2 [ 18 ]. ASVs affiliated with the hgcI clade were further resolved into specific actinobacterial lineages using the SBDI Sativa curated 16S GTDB database [ 19 ]. Subsequent data analysis was conducted using the R package Phyloseq v1.50.0 [ 20 ]. ASVs assigned to eukaryotes, chloroplasts, and mitochondria were excluded before rarefying the dataset to a uniform sequencing depth of 6,000 reads per sample. Rarefaction curves for individual samples were generated using the ‘ggrare’ function from the R package ranacapa v0.1.0 [ 21 ]. Species accumulation curves were generated using the ‘specaccum’ function from the R package vegan v2.6-8 [ 22 ]. Plots and heatmaps were generated using the R packages ggplot2 v3.5.1 [ 23 ] and pheatmap v1.0.12 ( https://github.com/raivokolde/pheatmap ). Ecological processes governing community assembly ASV sequences were aligned using MAFFT v7.453 with the default settings [ 24 ]. A maximum-likelihood phylogenetic tree was constructed using IQ-TREE2 v2.2.0 with 1000 iterations of ultrafast bootstrapping [ 25 ]. The final tree was visualized in the interactive Tree of Life (iTOL) v6 ( https://itol.embl.de/ ). Phylogenetic-bin-based null model analysis was performed using the R package iCAMP v1.5.12 to determine the relative contributions of deterministic and stochastic processes to prokaryotic community assembly [ 8 ]. The bin size limit was set to five, and samples from different depths were run separately. The 70-meter depth was excluded due to insufficient representation across the full sampling period. Identification of annually recurring populations and episodic bloomers Bacterial seasonality at the lake epilimnion was assessed using soft clustering implemented in the R package Mfuzz v2.66.0 [ 26 ], which allows the identification of temporal expression patterns with overlapping cluster membership. The cluster number was initially set to four and increased stepwise. A higher cluster number was accepted only when it yielded a new cluster with at least three ASVs with membership values above 0.8. This process led to a final cluster number of six. Episodic bloomers were identified using a Bayesian ensemble algorithm for time-series decomposition and change-point detection implemented in the R package Rbeast v.1.0.1 [ 27 ]. Since this analysis requires multi-year time series data for accurate detection of change points, only samples from depths down to 40 meters were included. The minimum probability for the change point detection was set to 0.7. Network-based analysis of microbial interactions A time-resolved microbial association network was constructed using extended local similarity analysis (eLSA) [ 28 ], with parameters: -r 1 -s 13 -d 1 -p perm -x 1000 -f linear -n percentileZ. The analysis included all measured environmental variables as well as ASVs classified as either annually recurrent or episodic populations. The resulting network was visualized using Cytoscape v3.10.3 [ 29 ]. Results Five-year overview of lake characteristics We collected 300 water samples at monthly intervals from the dam station of Lake Soyang (Fig. 1 a). These samples were taken from depths ranging between 1 and 90 meters over two periods: May 2015–May 2017 and January 2019–December 2021. For each sample, a comprehensive set of environmental and chemical parameters was measured (Figs. 1 b–c and S1– 2 ). The repeated pattern of thermal stratification in summer, followed by water mixing in winter, clearly showed that Lake Soyang is a monomictic lake (Fig. 1 b). Temperature profiles further indicated that the lake remained above 4 ℃ throughout the entire sampling period, resulting in ice-free winters (Fig. S1 ). The chlorophyll a concentrations in surface water averaged 1.1 ± 0.9 µg L⁻¹, reflecting the oligo-mesotrophic nature of the lake (Fig. 1 c) [ 30 ]. Although the temperature profiles at 70 and 90 meters were similar (5.4 ± 0.6 ℃ and 5.5 ± 0.6 ℃, respectively), variations in other environmental parameters suggested a meaningful environmental distinction between the two depths. A sharp decline in dissolved oxygen concentrations at 90 meters (3.1 ± 1.7 mg L − 1 ) indicated that this water layer is an oxygen-limited zone. Moreover, elevated chlorophyll a (Fig. 1 c) and ammonium concentrations (Fig. S2 ) at this depth supported the inference that 90 meters is near the lake's bottom [ 31 ]. Long-term microbial time series in Lake Soyang To explore spatiotemporal microbial dynamics, we analyzed prokaryotic community compositions across all collected samples. The analysis initially revealed a total of 5,745 ASVs. After filtering out rare ASVs (< 10 reads), 2,119 ASVs representing 36 distinct phyla remained (Table S1 ). The rarefaction curves plateaued in all samples (Fig. S3a), suggesting that microbial diversity was well represented at the given sequencing depth. Similarly, the species accumulation curve for the entire dataset (n = 300) showed that the sample size was sufficient to capture the majority of prokaryotic populations at the sampling location (Fig. S3b). Alpha diversity gradually increased with water depth (Fig. S4), except at 90 meters near the lake bottom. Temporally, alpha diversity declined markedly during summer in surface layers above the thermocline but increased during periods of water column mixing (Fig. S5). Non-metric multidimensional scaling (NMDS) revealed clear clustering of microbial communities by both season and water depth (Fig. 2 ). Repeated seasonal trends were evident in samples from depths down to 20 meters (Fig. 2 a). However, these trends gradually weakened in deeper waters, where temporal environmental variations were less pronounced. Samples collected near the bottom (90 meters) were clearly distinguished from those at shallower depths (Fig. 2 b). A closer examination of microbial abundance patterns between the 70- and 90-meter samples (Fig. S6) highlighted the presence of microbial population adapted to oxygen-limited conditions at 90 meters, including ASV12 from Methylotenera , ASV64 from Nitrosarchaeum , and three ASVs (ASV48, ASV49, and ASV335) from Flavobacterium . Surprisingly, ASV31 from acI-A (‘ Candidatus Planktophila ’ ), a lineage typically dominant in oxygenated water columns [ 32 ], also emerged as part of this population. This novel ASV was notably absent at other depths (Fig. S7), further suggesting its ecological specialization for the oxygen-limited hypolimnion. Ecological processes driving seasonal microbial dynamics We next quantified the ecological processes driving the seasonal dynamics of community assembly using a phylogenetic bin-based null model analysis (iCAMP) [ 8 ]. To enhance data interpretability and reduce technical noise, we focused on 243 ASVs (hereafter referred to as core ASVs) that exhibited high relative abundance (> 1% maximum abundance) and frequent occurrence (present in > 10% of the samples). These core ASVs (Fig. S7) accounted for 83.6% of total reads. Analysis of this curated dataset revealed that community assembly was influenced by both deterministic (homogeneous or heterogeneous selection) and stochastic (dispersal limitation, homogenizing dispersal, and drift) processes, with their relative importance varying markedly across months and depths (Figs. 3 a–e). Drift was the most dominant stochastic process overall, followed by homogeneous selection as the leading deterministic process. Dispersal limitation had a greater influence (35% on average) in the oxygen-limited bottom layer (Fig. 3 e) compared to shallower depths (26–33.9%). To explore fine-scale temporal dynamics, we focused on the lake’s epilimnion (1 meter), which exhibited the most pronounced temporal variations in community compositions (Fig. 2 b). The core ASVs at the epilimnion were assigned to 22 phylogenetic bins (Fig. 3 f). Since the importance of ecological processes was weighted by the relative abundance of each bin [ 8 ], a small subset (n = 6) of dominant bins (bin4–6, bin9, bin11, and bin21) accounted for over 67% of all processes (Table S2 ). During winter and early spring (Jan.–Mar.), homogeneous selection was driven primarily by bin4 and bin5, both represented by genome-streamlined actinobacteria such as acI and acIV clades. Stochastic processes were most influential in April, accounting for up to 96.6% of all inferred processes (Fig. 3 a). This exceptionally high prevalence could be attributed to the combined contribution of multiple bins (Fig. 3 f), highlighting the complex and unpredictable nature of community dynamics during this period. As water temperatures rose following this period, homogeneous selection gradually gained influence, reaching a maximum of 29% in August (Fig. 3 a). This increased relative importance of the deterministic process in mid-season was primarily contributed by bin6, represented by Fonsibacter . In the final quarter of the season (Oct.–Dec.), Sediminibacterium and Flavobacterium from bin 21 made notable contributions to the deterministic process (Fig. 3 f). In contrast, low-abundance bins were predominantly influenced by stochastic processes throughout the year, rather than by deterministic factors. Annually recurring and episodic succession of planktonic bacteria We next aimed to accurately capture the timing, intensity, and duration of microbial succession driven by the complex interplay of deterministic and stochastic processes. To systematically classify the observed patterns, we applied an unsupervised soft clustering approach to uncover annually recurring microbial succession in the lake epilimnion (1 meter). In parallel, episodic blooming events throughout the water column (down to 40 meters) were identified using a Bayesian model averaging algorithm, which decomposes time-series abundance trajectories to detect abrupt shifts. Samples from 2015 and 2017 were excluded from the analysis due to incomplete 12-month coverage. Based on these analyses, approximately one-third (n = 85) of the core ASVs were categorized as annually recurring, 28% (n = 69) as episodic, and the remaining (n = 89) as having weak or inconsistent temporal signals (Tables S3 and S4). The annually recurring microbes were further categorized into six distinct clusters (Table S3), each characterized by a unique seasonal development pattern. The first two clusters (C1 and C2) were associated with winter (Fig. 4 a), a season often considered as a ‘dormant period’, and displayed peaks of differing durations. ASVs from C1 exhibited a broad peak throughout the winter, predominantly consisting of cold-adapted taxa (Fig. 4 b) such as Methylopumilus [ 33 ] and Rhodoferax [ 12 ]. In contrast, ASVs associated with C2 showed a narrow peak (Jan.–Feb.), likely linked to water column turnover. During this period, microbial populations typically confined to the cold hypolimnion, including Methylobacter , Methylotenera , Nitrospira , and Opitutus , became enriched. As temperatures gradually increased in spring, copiotrophs such as Limnohabitans , Flavobacterium , Polynucleobacter , and the uncharacterized ‘Candidatus Planktoluna’ from C3 became dominant in surface waters. In a previous model [ 12 ], oligotrophic bacteria like Fonsibacter were strongly associated with the clear-water phase typically observed in May or June. In our study, C4 aligned with this phase and included Fonsibacter , CL500-3, Sandarakinorhabdus , Limnobacter , and Nemorincola . During peak summer temperatures, ASVs from Cyanobium and potential phytoplankton-associated bacteria such as Terrimonas , Noviherbaspirillum , and Roseomonas dominated, collectively forming C5. Lastly in C6, highly similar to the previous model [ 12 ], potential decay-associated taxa such as ‘Candidatus Nitrosotenuis’, Nitrosarchaeum , and Chthoniobacter became dominant during the fall season (Sep.–Nov.). These seasonal clusters exhibited robust interannual resilience and were consistently observed at depths down to 20 meters (Figs. S8–10). Although phylogenetic distance is often regarded as a proxy for niche separation [ 8 ], genome-streamlined actinobacteria demonstrated extensive niche diversification, even among closely related members within individual lineages. Annually recurring ASVs (n = 15) from these lineages, including acI-A, acI-B ( ‘Candidatus Nanopelagicus’), acI-C, and acIV (CL500-29), were distributed throughout the year and associated with four of the six identified clusters (Fig. 4 b). A previous high-resolution study in Lake Zurich reported that the seasonal abundance of these lineages was highly associated with algal and cyanobacterial blooms during the growing season [ 12 ]. However, nearly half of the ASVs linked to the winter cluster (C1) in our dataset indicated the existence of cold-adapted populations that remain poorly characterized. In contrast to the predictable patterns of annually recurring taxa, episodic bloomers exhibited irregular seasonal trajectories throughout the sampling period. Based on the number of change points (CPs) detected in their abundance trends, we classified these events into three categories: persistent (CP = 1), transient (CPs = 2), and intermittent (CPs = 3) (Fig. 5 a). Persistent bloomers, the most common type across depths (Fig. 5 b), underwent abrupt shifts in seasonal patterns that did not revert to their original states in subsequent years (Fig. 5 d). Transient bloomers were characterized by short-lived and non-repeating peaks, appearing in a specific year and absent in others (Fig. 5 e). At 40 meters, a few ASVs displayed intermittent blooming, appearing in the first and third years but absence in the intervening year, suggesting that such disturbances may recur under specific environmental conditions (Fig. 5 f). Interestingly, the number of episodic bloomers increased with depth, suggesting a higher degree of randomness in microbial succession in deeper water layers (Fig. 5 b). Genome-streamlined actinobacteria and fast-growing opportunists such as Limnohabitans from Gammaproteobacteria and were the most frequently identified taxa among episodic bloomers (Fig. 5 c and Table S4). Time-dependent associations among microbial and environmental components To further elucidate the complex interactions between microbial communities and environmental parameters, an association network was constructed using local similarity analysis (Fig. 6 ). The resulting network revealed 762 associations (edges) among 84 components (nodes), including 503 positive and 259 negative associations, as well as 494 synchronous and 268 delayed associations. As expected, 76% of the positive and synchronous associations occurred within the same clusters, demonstrating strong internal connectivity within the clusters (Fig. 6 a). In contrast, 78.7% of the delayed associations were observed between different clusters, highlighting a well-coordinated and sequential succession of microbial populations. Temperature emerged as the most central environmental component in the network, with the highest number of associations (n = 55). Among microbial components, ASV40 from acI-C and ASV6 from Anaerolineaceae were identified as keystone taxa, each exhibiting 41 distinct associations with other nodes. Several episodic bloomers were also integrated into the network, providing insights into underlying mechanisms potentially driving their dynamics. Notably, the blooming of ASV81 ( Cyanobium ) was found to be influenced by ASV46 ( Limnohabitans ) and ASV138 ( Beijerinckiaceae ), the latter of which is known as a potential nitrogen-fixer [ 34 ] (Fig. 6 b). Other episodic bloomers, including ASV100 ( Polynucleobacter ), ASV37 (SL56 marine group from Chloroflexota ) and ASV65 ( Verrucomicrobiia ), formed up to 25 associations with other nodes, indicating complex microbial cascades underlying their succession. Discussion Our long-term microbial time series data, combined with quantitative modeling and statistical approaches, facilitated an in-depth analysis of freshwater microbial dynamics. Investigating the ecological processes underlying bacterial community assembly highlighted a balance between stochastic and deterministic influences, with stochastic processes playing a more dominant role (Fig. 3 ). The strong influence of drift across seasons suggests that inter-annual variations, such as the timing of phytoplankton blooms [ 35 ] or the intensity of thermal stratification and mixing [ 36 ], may amplify the randomness and unpredictability of bacterial community structures. Previous studies have shown that the relative importance of ecological processes might vary considerably within a single ecosystem [ 9 , 37 , 38 ], depending on the monitoring framework, including sampling frequency (high vs. low), scale (spatial vs. temporal), and study duration (weeks to years). This emphasizes a key challenge in ecological research, which involves designing sampling strategies and analyzing datasets to detect underlying ecological processes with greater accuracy and sensitivity. By successfully incorporating samples from winter and early spring (Jan.–Apr.), our study addressed key gaps in a recent investigation [ 12 ], and provided a more continuous view of the seasonal dynamics of freshwater bacteria (Fig. 4 a). The identification of two distinct winter clusters emphasized the role of cold-adapted bacteria in driving seasonal microbial succession. This study also illustrated freshwater microbial dynamics at a finer resolution by reconstructing seasonal succession patterns at the ASV level rather than the broader taxonomic level [ 12 ]. This approach led to the discovery of previously overlooked ecological niches for genome-streamlined actinobacteria [ 32 , 39 ], including those associated with cold winter periods (Fig. 4 b) and oxygen-limited hypolimnion (Fig. S6). The remarkable consistency of microbial resilience observed over multiple years suggests that these communities are likely to serve fundamental and irreplaceable roles in maintaining ecosystem function. Tracking microbial dynamics over multiple years allowed us to distinguish annually recurring microbes from episodic bloomers (Figs. 4 and 5 ). Episodic bloomers may reflect two ecologically distinct scenarios: i) they may represent functionally redundant populations that temporarily occupy available niches without serving essential ecosystem functions, or ii) they could emerge in response to major environmental shifts, such as species invasions [ 15 ], global warming [ 40 ], or eutrophication [ 41 ]. In the latter scenario, episodic bloomers could act as sensitive indicators of ecological instability or transitional states. Understanding the cumulative impact of episodic bloomers requires continuous, high-resolution monitoring of microbial communities [ 15 ], which is particularly crucial given the accelerating pace of global environmental change and the increasing frequency and intensity of ecological disruptions [ 42 ]. By quantifying all detectable connections among microbial and environmental components, we further resolved the full spectrum of microbial dynamics into an association network (Fig. 6 ). This network-based approach offers a valuable opportunity to advance beyond conventional ecosystem assessments, which typically rely on metrics such as physical and chemical parameters [ 43 ] or biological indices based on a limited number of taxa [ 44 , 45 ]. By capturing patterns such as premature or delayed successions, missing microbial components at critical time points, or the emergence of new components forming novel connections, we can gain deeper insights into how microbial disturbances are triggered and propagate throughout the network [ 46 , 47 ], ultimately altering ecosystem functions and stability over time. Conclusions This long-term microbial time series successfully addressed critical gaps in earlier studies, providing a comprehensive view of seasonal microbial dynamics in a freshwater ecosystem. Future studies could expand the scope of this approach by integrating additional trophic levels, such as viruses or protists, incorporating a wider range of environmental parameters, and refining microbial network models for more accurate ecosystem assessments. Ultimately, these advancements hold the potential to transform freshwater ecosystem management and conservation through more effective, data-driven strategies. Abbreviations PEG Plankton ecology group DO Dissolved oxygen ASV Amplicon sequence variant NMDS Non-metric multidimensional scaling CP Change points Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. Availability of data and materials The sequencing data from this study were deposited to the NCBI Sequence Read Archive (SRA) under BioProject PRJNA1253076. All other relevant data supporting the findings of this study are available within the paper and its supplementary information files. Funding This study was supported by the Mid-Career Research Program (NRF-2022R1A2C3008502 to J.-C.C.) and the Sejong Science Fellowship (NRF-2022R1C1C2004070 to S.K.) through the National Research Foundation (NRF) funded by the Ministry of Sciences and ICT of the Korea government. Contributions J-CC and IK conceived and designed the experiment. SP and SK collected the data. HP, SK, and SP conducted data analyses. HP wrote the initial draft of the manuscript with all authors contributing to its revisions. All authors reviewed and approved the final manuscript. Acknowledgements We gratefully acknowledge captain Do Gyeom Lee and the crews of the R/V of the Korea Water Resources Corporation (K-water) for their assistance with freshwater sample collection. References Baron JS, LeRoy Poff N, Angermeier PL, Dahm CN, Gleick PH, Hairston NG, et al. Meeting ecological and societal needs for freshwater. Ecol Appl. 2002;12:1247–60. Dudgeon D, Arthington AH, Gessner MO, Kawabata ZI, Knowler DJ, Lévêque C, et al. Freshwater biodiversity: Importance, threats, status and conservation challenges. Biol Rev Camb Philos Soc. 2006;81:163–82. Arora-Williams K, Olesen SW, Scandella BP, Delwiche K, Spencer SJ, Myers EM, et al. Dynamics of microbial populations mediating biogeochemical cycling in a freshwater lake. Microbiome. 2018;6:1–16. Vellend M. Conceptual synthesis in community ecology. Q Rev Biology. 2010;85:183–206. Chase JM, Myers JA. Disentangling the importance of ecological niches from stochastic processes across scales. Philosophical Trans Royal Soc B: Biol Sci. 2011;366:2351–63. Chave J. Neutral theory and community ecology. Ecol Lett. 2004;7:241–53. Riddley M, Hepp S, Hardeep FNU, Nayak A, Liu M, Xing X, et al. Differential roles of deterministic and stochastic processes in structuring soil bacterial ecotypes across terrestrial ecosystems. Nat Commun. 2025;16:2337. Ning D, Yuan M, Wu L, Zhang Y, Guo X, Zhou X, et al. A quantitative framework reveals ecological drivers of grassland microbial community assembly in response to warming. Nat Commun. 2020;11:4717. Aguilar P, Sommaruga R. The balance between deterministic and stochastic processes in structuring lake bacterioplankton community over time. Mol Ecol. 2020;29:3117–30. Ge Y, Meng X, Heino J, García-Girón J, Liu Y, Li Z et al. Stochasticity overrides deterministic processes in structuring macroinvertebrate communities in a plateau aquatic system. Ecosphere. 2021;12. Sommer U, Adrian R, De Senerpont Domis L, Elser JJ, Gaedke U, Ibelings B, et al. Beyond the plankton ecology group (PEG) model: Mechanisms driving plankton succession. Annu Rev Ecol Evol Syst. 2012;43:429–48. Park H, Shabarova T, Salcher MM, Kosová L, Rychtecký P, Mukherjee I, et al. In the right place, at the right time: the integration of bacteria into the Plankton Ecology Group model. Microbiome. 2023;11:112. Paver SF, Newton RJ, Coleman ML. Microbial communities of the Laurentian Great Lakes reflect connectivity and local biogeochemistry. Environ Microbiol. 2020;22:433–46. Kent AD, Yannarell AC, Rusak JA, Triplett EW, McMahon KD. Synchrony in aquatic microbial community dynamics. ISME J. 2007;1:38–47. Robin R, Rohwer, Riley J, Hale, Zanden MJ, Vander TR, Miller, Katherine D, McMahon. Species invasions shift microbial phenology in a two-decade freshwater time series. Proceedings of the National Academy of Sciences. 2023;120:e2211796120. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011;17:10–2. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13:581–3. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Res. 2013;41:590–6. Swedish Biodiversity Infrastructure. (SBDI; 2021). SBDI Sativa curated 16S GTDB database. https://doi.org/10.17044/scilifelab.14869077 McMurdie PJ, Holmes S. Phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data. PLoS ONE. 2013;8. Kandlikar GS, Gold ZJ, Cowen MC, Meyer RS, Freise AC, Kraft NJB, et al. ranacapa: An R package and Shiny web app to explore environmental DNA data with exploratory statistics and interactive visualizations. F1000Res. 2018;7:1734. Dixon P. Computer program review VEGAN, a package of R functions for community ecology. J Veg Sci. 2003;14:927–30. H W. ggplot2: elegant graphics for data analysis. New York: Springer-Verlag. 2016. p. ISBN 978 3. Katoh K, Standley DM. MAFFT multiple sequence alignment software version 7: Improvements in performance and usability. Mol Biol Evol. 2013;30:772–80. Minh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, Von Haeseler A, et al. IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era. Mol Biol Evol. 2020;37:1530–4. Kumar L, Futschik ME, Mfuzz. A software package for soft clustering of microarray data. Bioinformation. 2007;2:5–7. Zhao K, Wulder MA, Hu T, Bright R, Wu Q, Qin H, et al. Detecting change-point, trend, and seasonality in satellite time series data to track abrupt changes and nonlinear dynamics: A Bayesian ensemble algorithm. Remote Sens Environ. 2019;232:111181. Xia LC, Ai D, Cram J, Fuhrman JA, Sun F. Efficient statistical significance approximation for local similarity analysis of high-throughput time series data. Bioinformatics. 2013;29:230–7. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13:2498–504. Kong D, Kim B. Suggestion for Trophic State Classification of Korean Lakes. J Korean Soc Water Environ. 2019;35:248–56. Bhadury P, Sen A. Understanding Impact of Seasonal Nutrient Influx on Sedimentary Organic Carbon and Its Relationship With Ammonia spp. in a Coastal Lagoon. Front Mar Sci. 2020;7. Okazaki Y, Fujinaga S, Tanaka A, Kohzu A, Oyagi H, Nakano SI. Ubiquity and quantitative significance of bacterioplankton lineages inhabiting the oxygenated hypolimnion of deep freshwater lakes. ISME J. 2017;11:2279–93. Salcher MM, Schaefle D, Kaspar M, Neuenschwander SM, Ghai R. Evolution in action: habitat transition from sediment to the pelagial leads to genome streamlining in Methylophilaceae. ISME J. 2019;13:2764–77. Marín Irma and Arahal DR. The Family Beijerinckiaceae. Rosenberg Eugene and DeLong EF and LS and SE and TF, editor. The Prokaryotes: Alphaproteobacteria and Betaproteobacteria [Internet]. Berlin, Heidelberg: Springer Berlin Heidelberg; 2014. pp. 115–33. Yuan J, Cao Z, Ma J, Li Y, Qiu Y, Duan H. Influence of climate extremes on long-term changes in cyanobacterial blooms in a eutrophic and shallow lake. Sci Total Environ. 2024;939. Woolway RI, Sharma S, Weyhenmeyer GA, Debolskiy A, Golub M, Mercado-Bettín D et al. Phenological shifts in lake stratification under climate change. Nat Commun. 2021;12. Wang J, Shen J, Wu Y, Tu C, Soininen J, Stegen JC, et al. Phylogenetic beta diversity in bacterial assemblages across ecosystems: Deterministic versus stochastic processes. ISME J. 2013;7:1310–21. Roguet A, Laigle GS, Therial C, Bressy A, Soulignac F, Catherine A et al. Neutral community model explains the bacterial community assembly in freshwater lakes. FEMS Microbiol Ecol. 2015;91. Neuenschwander SM, Ghai R, Pernthaler J, Salcher MM. Microdiversification in genome-streamlined ubiquitous freshwater Actinobacteria. ISME J. 2018;12:185–98. Ouyang J, Wu H, Yang H, Wang J, Liu J, Tong Y et al. Global warming induces the succession of photosynthetic microbial communities in a glacial lake on the Tibetan Plateau. Water Res. 2023;242. Xie G, Zhang Y, Gong Y, Luo W, Tang X. Extreme trophic tales: deciphering bacterial diversity and potential functions in oligotrophic and hypereutrophic lakes. BMC Microbiol. 2024;24. Minière A, von Schuckmann K, Sallée JB, Vogt L. Robust acceleration of Earth system heating observed over the past six decades. Sci Rep. 2023;13. Suresh K, Tang T, Van Vliet MTH, Bierkens MFP, Strokal M, Sorger-Domenigg F et al. Recent advancement in water quality indicators for eutrophication in global freshwater lakes. Environ Res Lett. 2023;18. Fortunato CS, Eiler A, Herfort L, Needoba JA, Peterson TD, Crump BC. Determining indicator taxa across spatial and seasonal gradients in the Columbia River coastal margin. ISME J. 2013;7:1899–911. Aylagas E, Borja Á, Tangherlini M, Dell’Anno A, Corinaldesi C, Michell CT, et al. A bacterial community-based index to assess the ecological status of estuarine and coastal environments. Mar Pollut Bull. 2017;114:679–88. Röttjers L, Faust K. From hairballs to hypotheses–biological insights from microbial networks. FEMS Microbiol Rev. 2018;42:761–80. Carlström CI, Field CM, Bortfeld-Miller M, Müller B, Sunagawa S, Vorholt JA. Synthetic microbiota reveal priority effects and keystone strains in the Arabidopsis phyllosphere. Nat Ecol Evol. 2019;3:1445–54. Additional Declarations No competing interests reported. Supplementary Files supplementarytablev2.xlsx Supplementaryfigures.pdf Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7144747","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":545726052,"identity":"3db3c360-1125-47af-a74c-5e9a8998ff0d","order_by":0,"name":"Hongjae Park","email":"","orcid":"","institution":"Inha University","correspondingAuthor":false,"prefix":"","firstName":"Hongjae","middleName":"","lastName":"Park","suffix":""},{"id":545726053,"identity":"873c7120-b39b-48b9-9497-725d62a74d19","order_by":1,"name":"Suhyun Kim","email":"","orcid":"","institution":"Inha University","correspondingAuthor":false,"prefix":"","firstName":"Suhyun","middleName":"","lastName":"Kim","suffix":""},{"id":545726054,"identity":"a540f5d6-c5c6-4090-b15c-6d73a1255f74","order_by":2,"name":"Miri. S. Park","email":"","orcid":"","institution":"Inha University","correspondingAuthor":false,"prefix":"","firstName":"Miri.","middleName":"S.","lastName":"Park","suffix":""},{"id":545726055,"identity":"69448449-f843-44f0-a11c-ff4df93c8c7a","order_by":3,"name":"Ilnam Kang","email":"","orcid":"","institution":"Inha University","correspondingAuthor":false,"prefix":"","firstName":"Ilnam","middleName":"","lastName":"Kang","suffix":""},{"id":545726056,"identity":"3051c2fc-cc5e-4797-a523-446dfb35ea8b","order_by":4,"name":"Jang-Cheon Cho","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuUlEQVRIiWNgGAWjYDACCQYGA4YKCSABBgnEajlDqhYGxjYGErTwz24+UPBznkWeOfsBxg8/GNLyCVty51iCYe82iWLLngRmyR6GHMsGQloMJHIMjBm3SSRuOJDAIM3AUGFA0BaIljlALecfMP8mQUsDUMuNBDagLTmEtUjcSEsw7DkG0vKwzbLHII2wFv4ZyccMftTUAR2WfPjGj4pkwlqAgA2qirEBHjuEAPMD4tSNglEwCkbBiAUAfJc3QAz1COwAAAAASUVORK5CYII=","orcid":"","institution":"Inha University","correspondingAuthor":true,"prefix":"","firstName":"Jang-Cheon","middleName":"","lastName":"Cho","suffix":""}],"badges":[],"createdAt":"2025-07-17 04:23:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7144747/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7144747/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96113849,"identity":"2c23f5d0-3fca-44a2-a039-7a5dbe964d18","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":306708,"visible":true,"origin":"","legend":"","description":"","filename":"microbiomesoyang16sparketal2025v19.docx","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/d26a1400618be0fb4e23d705.docx"},{"id":96250495,"identity":"97d02351-cd97-4a73-96ee-7870e884cd02","added_by":"auto","created_at":"2025-11-19 07:38:27","extension":"json","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6872,"visible":true,"origin":"","legend":"","description":"","filename":"917ec93a5178413a902c4b7971bc4fd7.json","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/af17b4e3c6a8bcd4628a6937.json"},{"id":96250529,"identity":"7c50d817-8178-4fea-ab46-c8141a065141","added_by":"auto","created_at":"2025-11-19 07:38:35","extension":"pdf","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11803240,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryfigures.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/053275871ffcb09e28c43ec0.pdf"},{"id":96113850,"identity":"480dc006-4c5b-4696-af80-865ad782a623","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"xlsx","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2330376,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarytablev2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/18be6e0811d0ae101639951a.xlsx"},{"id":96113852,"identity":"ca71d07e-613e-4b09-ae82-cc4152702b7f","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"xml","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":109114,"visible":true,"origin":"","legend":"","description":"","filename":"917ec93a5178413a902c4b7971bc4fd71enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/8e3008585704fe468bf976b2.xml"},{"id":96248727,"identity":"a01d99fc-7342-4520-ac12-c9acc4dc9a53","added_by":"auto","created_at":"2025-11-19 07:29:04","extension":"pdf","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2561263,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/bc9af102f3b8681b09ce429e.pdf"},{"id":96113859,"identity":"d7fe9ff0-88b7-40bc-af0d-c0ecb74c92c5","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"pdf","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":488831,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/3f188128ac7a4d8f34708c3d.pdf"},{"id":96249033,"identity":"c659b006-13bc-492e-bc9f-ae4d2d0127ed","added_by":"auto","created_at":"2025-11-19 07:29:57","extension":"pdf","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":725632,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/3f80098490064398b0f09f3d.pdf"},{"id":96113853,"identity":"446adf97-58ec-49e5-8324-9de5fc92ce62","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"pdf","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":585301,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/757fa4eb72300a907fb7b82a.pdf"},{"id":96113858,"identity":"1976a05e-bbff-40af-b620-38ba54b5cc55","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"pdf","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":519175,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/d4c42417b4912d95eeeac32c.pdf"},{"id":96113860,"identity":"6f9985bb-138f-437e-9eef-627e4f130284","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"pdf","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1813992,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/081737f7ec453e5822acbf7c.pdf"},{"id":96113856,"identity":"10c57cc7-8ba9-4cbd-b730-225313fc41e6","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"xml","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":107690,"visible":true,"origin":"","legend":"","description":"","filename":"917ec93a5178413a902c4b7971bc4fd71structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/9068099dd0e9a57bf41bae1a.xml"},{"id":96247641,"identity":"793a496b-2cad-4d75-979c-537871e37f10","added_by":"auto","created_at":"2025-11-19 07:27:39","extension":"html","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":117980,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/835d40494ca79ea3f66a606f.html"},{"id":96113842,"identity":"dc0e3abd-afac-427b-b3f0-d522580133f3","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":161874,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFive-year microbial time series in an oligo-mesotrophic lake.\u003c/strong\u003e \u003cstrong\u003ea\u003c/strong\u003e, Timeline of monthly sampling at the dam station of Lake Soyang. Filled circles indicate months when samples were collected, while open circles indicate months with missing data. Months were categorized into five seasonal phases (Spring, Clear-water, Summer, Fall, and Winter) based on a previously established framework [12]. \u003cstrong\u003eb\u003c/strong\u003e, Two-dimensional contour plot of temperature at the sampling station. \u003cstrong\u003ec\u003c/strong\u003e, Box plots of key environmental parameters measured across all depths and sampling periods. The lower and upper edges of the boxplots correspond to the first and third quartiles, the whiskers extend to the largest or smallest value at 1.5 times the interquartile, and the black bars across the box represent median values. Additional plots for environmental parameters are shown in Figs. S1–S2.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/1305d0bfcacbba0445890875.png"},{"id":96113841,"identity":"7b5b1dd6-80cf-4ef3-82b4-08baed07846a","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":45756,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNon-metric multidimensional scaling (NMDS) of bacterial community compositions.\u003c/strong\u003e Each circle represents an individual sample, colored by season (a) and by water depth (b). In panel (a), the NMDS1 (x-axis) is scaled independently for each depth to enhance resolution.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/1d277954ef7aaf98524fb9bc.png"},{"id":96113846,"identity":"0b2f1444-8ffd-48e1-ab5b-caa250c31e84","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":120929,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCommunity assembly mechanisms in Lake Soyang. a-e\u003c/strong\u003e, Temporal variation in the relative contributions of five ecological processes, heterogeneous selection (HeS), homogeneous selection (HoS), dispersal limitation (DL), homogenizing dispersal (HD), and drift (DR), at different water depths (1, 10, 20, 40, and 90 meters). f, Community assembly mechanisms of 22 phylogenetic bins in the lake epilimnion (1 meter). The phylogenetic tree on the left shows relationships among 243 core ASVs. Bootstrap values \u0026gt;95 are marked on the tree and the scale bar represents 1 substitution per nucleotide position. On the right, bin contribution (%) indicates the proportional importance of individual bins on each process (Table S2). The three most dominant processes (homogeneous selection, dispersal limitation, and drift) are shown. Abundance change (%) represents the monthly average change in the total abundance of ASVs within each bin. Error bars represent the standard deviation of abundances among different years.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/3c803290a3676c1082eb9a95.png"},{"id":96113844,"identity":"1f4d40eb-a496-4546-8102-6928d06ec847","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":101988,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAnnually recurring ASVs at the lake epilimnion (1 meter). a\u003c/strong\u003e, Standardized abundance changes (Z-scores) of annually recurring ASVs, grouped into six distinct clusters (C1–C6) using soft clustering. Only ASVs with membership value \u0026gt; 0.8 are shown. Vertical green lines indicate January of each year\u003cstrong\u003e b,\u003c/strong\u003e Heatmap showing the seasonal abundance patterns (log\u003csub\u003e2\u003c/sub\u003e read counts) of the ASVs. Columns and rows represent different time points and individual ASVs, respectively. Taxonomic affiliations of selected ASVs are marked. \u0026nbsp;C1-C6, Cluster1-Cluster6.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/f2127a37c6610374953145b4.png"},{"id":96113845,"identity":"5d8a2e40-b6a7-4808-ba01-be90936a4499","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":39135,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eIdentification of episodic bloomers throughout the water column (down to 40 meters). a\u003c/strong\u003e, Categorization of blooming events based on the number of change points. \u003cstrong\u003eb\u003c/strong\u003e, Distribution of episodic bloomers by water depth, showing the number of ASVs observed in each category. \u003cstrong\u003ec\u003c/strong\u003e, Taxonomic affiliation of 69 ASVs identified as episodic bloomers. \u003cstrong\u003ed–f,\u003c/strong\u003e Representative examples of episodic bloomers categorized as persistent (CL500-3), transient (acIV), and intermittent (acI-A). For each panel, trend probability (Trend Pr) and slope sign probability (slpSign) are shown. In the slpSign panel, the upper red section represents the probability that the trend slope is positive, the middle green section indicates the probability that the slope is zero, and the lower blue section reflects the probability that the slope is negative. CP, change point; Pr, probability of change point occurring over time; slpSign, slope sign.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/06e3d224eec227d1cc418d27.png"},{"id":96113847,"identity":"9617925f-9f2b-4a61-b3ef-71cddab72902","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":236733,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTime-dependent association network depicting interactions among microbial and environmental components in the lake epilimnion (1 meter). \u003c/strong\u003eCircles represent ASVs and diamonds represent environmental factors. ASVs are colored according to seasonality (a) and taxonomy (b). Node labels on the panel b indicate the lowest taxonomic levels for each ASV. Edges connecting nodes represent strong (|local similarity|\u0026gt;0.6) and significant (P-value\u0026lt;0.001 and Q-value\u0026lt;0.01) correlations. Edge thickness is proportional to the absolute local similarity score. Line arrows indicate a 1-month shift/delay in the correlation. Asterisks denote keystone ASVs.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/a75fe6bc003c942e7eb31aa1.png"},{"id":104993715,"identity":"c29095d7-60eb-45ce-bc5d-7087ff989fde","added_by":"auto","created_at":"2026-03-19 15:56:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1674068,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/2aa164df-c062-477e-bf6d-f4ca78d9f4f0.pdf"},{"id":96247235,"identity":"50853470-aa57-4fa1-bf1a-86953c17275b","added_by":"auto","created_at":"2025-11-19 07:27:16","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2330376,"visible":true,"origin":"","legend":"","description":"","filename":"supplementarytablev2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/832d552a8d89d287c3438a66.xlsx"},{"id":96113862,"identity":"18c8c80f-41f9-4470-b914-3bdb445ce501","added_by":"auto","created_at":"2025-11-17 17:55:58","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":11803240,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementaryfigures.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7144747/v1/cd23b3f9e0f8368dcb5dbace.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Long-term freshwater time series reveals recurrent and episodic microbial dynamics driven by distinct assembly mechanisms","fulltext":[{"header":"Background","content":"\u003cp\u003eFreshwater is essential for human life, supplying water for drinking, agriculture, industrial processes, and transportation [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. In addition to serving human needs, freshwater habitats such as lakes, rivers, and ponds support diverse plant and animal species and maintain local biodiversity [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. At the core of these ecosystems are freshwater bacteria, which play a pivotal role in nutrient cycling, breaking down organic matter, and forming the backbone of the aquatic food web [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Understanding their assembly mechanisms and spatiotemporal dynamics is key to assessing ecosystem health and ensuring long-term water quality management.\u003c/p\u003e\u003cp\u003eOur understanding of microbial community assembly has been shaped by two conceptual frameworks: niche-based and neutral theories [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. According to the niche theory, deterministic forces, including abiotic filtering and species interactions, play a dominant role in shaping community assembly by favoring taxa best suited to local environments [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. In contrast, neutral theory emphasizes the role of stochastic processes, such as ecological drift, dispersal limitation, and homogenizing dispersal, under the assumption that all taxa are ecologically equivalent [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. While once viewed as competing explanations, it is now widely accepted that both deterministic and stochastic processes operate in tandem [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], with the key challenge being to quantify their relative contributions across space and time. In freshwater ecosystems, the mechanisms underlying seasonal community assembly remain poorly understood [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe plankton ecology group (PEG) model has long been a cornerstone for interpreting the seasonal development of large planktonic organisms in lakes [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Recent efforts to extend this framework by integrating bacterial populations have shed light on their coordinated succession within the broader planktonic community [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. However, despite these advances, a few blind spots remain to be resolved. For instance, winter microbial dynamics have been underexplored in many freshwater studies [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e], resulting in an incomplete understanding of their annual transition. In addition, the temporal architecture of microbial community transformation might differ considerably among geographically distant lakes [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], emphasizing the need for cross-system comparisons. Alongside seasonally recurring bacterial groups, episodic populations that emerge unpredictably in response to environmental fluctuations deserve greater attention, especially in the context of long-term ecological change [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Addressing these gaps is essential for capturing the full spectrum of microbial dynamics and developing a generalized ecosystem model for freshwater environments.\u003c/p\u003e\u003cp\u003eThis study presents a five-year microbial time series from Lake Soyang, the largest freshwater lake in South Korea. By leveraging this long-term dataset, we examined how deterministic and stochastic processes interact to influence microbial community assembly over time. We further reconstructed a generalized timeline of microbial succession at lake epilimnion, identifying annually recurring groups with robust seasonal signatures, as well as episodic populations that emerged unpredictably. In addition, an association network uncovered a high degree of interconnectedness among microbial and environmental components, pointing to their coordinated succession as a collective outcome of complex interdependencies.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cb\u003eStudy site and sample collection\u003c/b\u003e\u003c/p\u003e\u003cp\u003eOver a five-year period, 300 water samples were collected monthly from the dam station of Lake Soyang (37\u0026deg;56'51.0\"N 127\u0026deg;49'08.0\"E). The sampling was carried out in two distinct phases. In the first phase (May 2015\u0026ndash;May 2017), water was collected from depths of 1, 10, 20, 40, and 70 meters. In the second phase (January 2019\u0026ndash;December 2021), the 70-meter sampling point was replaced by a 90-meter depth to capture near-bottom microbial communities. At each sampling event, 10\u0026ndash;20 L of water was collected using a Niskin water sampler (General Oceanics, USA). To harvest microbial biomass, 1 L of water was first filtered through a 3-\u0026micro;m cellulose acetate membrane (Advantec, Japan) to remove coarse particles and larger organisms. The resulting filtrate was then passed through 0.22-\u0026micro;m polyethersulfone membranes (Pall, USA). Membrane filters containing the biomass were stored at \u0026minus;\u0026thinsp;80\u0026deg;C until further processing.\u003c/p\u003e\u003cp\u003e\u003cb\u003eEnvironmental measures\u003c/b\u003e\u003c/p\u003e\u003cp\u003eVertical profiles of temperature, pH, conductivity, and dissolved oxygen (DO) concentration were measured at the sampling station using a YSI 556 MPS Multiprobe system (YSI, USA). Chlorophyll \u003cem\u003ea\u003c/em\u003e concentrations were determined using a UV-VIS spectrophotometer (Model UV-2600, Shimadzu, Japan), following extraction with acetone from a GF/F glass fiber filter (Whatman, USA). Additional physicochemical parameters, including ammonium, nitrite, nitrate, phosphate, and silicate, were measured from water filtered through a 0.45-\u0026micro;m mixed cellulose ester membrane (Advantec, Japan), using a QuAAtro microflow analyzer (SEAL Analytical, UK) during the first sampling phase, and a Hach reagent kit (Hach, USA) during the second phase.\u003c/p\u003e\u003cp\u003e\u003cb\u003eGenomic DNA extraction and 16S rRNA gene amplicon sequencing\u003c/b\u003e\u003c/p\u003e\u003cp\u003eGenomic DNA was extracted using the PowerWater DNA isolation kit (Qiagen, Germany) for the first batch, and the DNeasy PowerSoil Pro Kit (Qiagen, Germany) for the second. DNA concentration was quantified using a Qubit fluorometer (Invitrogen, USA). PCR amplification was performed using the primer pair 515FY and 926R (targeting the V4\u0026ndash;V5 region) for the first batch, and 515FY and 806RB (targeting the V4 region) for the second. Library preparation and sequencing were carried out at CJ Bioscience (Seoul, South Korea). Amplicon libraries were sequenced using the 2\u0026times;300 bp (first batch) and 2\u0026times;250 bp (second batch) paired-end chemistry on the Illumina MiSeq platform (Illumina, USA).\u003c/p\u003e\u003cp\u003e\u003cb\u003eSequence processing\u003c/b\u003e\u003c/p\u003e\u003cp\u003eAdapter and primer trimming was performed using Cutadapt [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e], with 515FY and 806RB primer sequences removed from all reads to standardize downstream processing. Trimmed reads were further processed using the R package DADA2 v1.34.0 [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], following the standard pipeline available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://benjjneb.github.io/dada2/tutorial.html\u003c/span\u003e\u003cspan address=\"https://benjjneb.github.io/dada2/tutorial.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Taxonomic assignment of amplicon sequence variants (ASVs) was conducted using the SILVA database v138.2 [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. ASVs affiliated with the hgcI clade were further resolved into specific actinobacterial lineages using the SBDI Sativa curated 16S GTDB database [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Subsequent data analysis was conducted using the R package Phyloseq v1.50.0 [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. ASVs assigned to eukaryotes, chloroplasts, and mitochondria were excluded before rarefying the dataset to a uniform sequencing depth of 6,000 reads per sample. Rarefaction curves for individual samples were generated using the \u0026lsquo;ggrare\u0026rsquo; function from the R package ranacapa v0.1.0 [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Species accumulation curves were generated using the \u0026lsquo;specaccum\u0026rsquo; function from the R package vegan v2.6-8 [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Plots and heatmaps were generated using the R packages ggplot2 v3.5.1 [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] and pheatmap v1.0.12 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/raivokolde/pheatmap\u003c/span\u003e\u003cspan address=\"https://github.com/raivokolde/pheatmap\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cb\u003eEcological processes governing community assembly\u003c/b\u003e\u003c/p\u003e\u003cp\u003eASV sequences were aligned using MAFFT v7.453 with the default settings [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. A maximum-likelihood phylogenetic tree was constructed using IQ-TREE2 v2.2.0 with 1000 iterations of ultrafast bootstrapping [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. The final tree was visualized in the interactive Tree of Life (iTOL) v6 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://itol.embl.de/\u003c/span\u003e\u003cspan address=\"https://itol.embl.de/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Phylogenetic-bin-based null model analysis was performed using the R package iCAMP v1.5.12 to determine the relative contributions of deterministic and stochastic processes to prokaryotic community assembly [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. The bin size limit was set to five, and samples from different depths were run separately. The 70-meter depth was excluded due to insufficient representation across the full sampling period.\u003c/p\u003e\u003cp\u003e\u003cb\u003eIdentification of annually recurring populations and episodic bloomers\u003c/b\u003e\u003c/p\u003e\u003cp\u003eBacterial seasonality at the lake epilimnion was assessed using soft clustering implemented in the R package Mfuzz v2.66.0 [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e], which allows the identification of temporal expression patterns with overlapping cluster membership. The cluster number was initially set to four and increased stepwise. A higher cluster number was accepted only when it yielded a new cluster with at least three ASVs with membership values above 0.8. This process led to a final cluster number of six.\u003c/p\u003e\u003cp\u003eEpisodic bloomers were identified using a Bayesian ensemble algorithm for time-series decomposition and change-point detection implemented in the R package Rbeast v.1.0.1 [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Since this analysis requires multi-year time series data for accurate detection of change points, only samples from depths down to 40 meters were included. The minimum probability for the change point detection was set to 0.7.\u003c/p\u003e\u003cp\u003e\u003cb\u003eNetwork-based analysis of microbial interactions\u003c/b\u003e\u003c/p\u003e\u003cp\u003eA time-resolved microbial association network was constructed using extended local similarity analysis (eLSA) [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e], with parameters: -r 1 -s 13 -d 1 -p perm -x 1000 -f linear -n percentileZ. The analysis included all measured environmental variables as well as ASVs classified as either annually recurrent or episodic populations. The resulting network was visualized using Cytoscape v3.10.3 [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cb\u003eFive-year overview of lake characteristics\u003c/b\u003e\u003c/p\u003e\u003cp\u003eWe collected 300 water samples at monthly intervals from the dam station of Lake Soyang (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). These samples were taken from depths ranging between 1 and 90 meters over two periods: May 2015\u0026ndash;May 2017 and January 2019\u0026ndash;December 2021. For each sample, a comprehensive set of environmental and chemical parameters was measured (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb\u0026ndash;c and S1\u0026ndash;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The repeated pattern of thermal stratification in summer, followed by water mixing in winter, clearly showed that Lake Soyang is a monomictic lake (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb). Temperature profiles further indicated that the lake remained above 4 ℃ throughout the entire sampling period, resulting in ice-free winters (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The chlorophyll \u003cem\u003ea\u003c/em\u003e concentrations in surface water averaged 1.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.9 \u0026micro;g L⁻\u0026sup1;, reflecting the oligo-mesotrophic nature of the lake (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec) [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Although the temperature profiles at 70 and 90 meters were similar (5.4\u0026thinsp;\u0026plusmn;\u0026thinsp;0.6 ℃ and 5.5\u0026thinsp;\u0026plusmn;\u0026thinsp;0.6 ℃, respectively), variations in other environmental parameters suggested a meaningful environmental distinction between the two depths. A sharp decline in dissolved oxygen concentrations at 90 meters (3.1\u0026thinsp;\u0026plusmn;\u0026thinsp;1.7 mg L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) indicated that this water layer is an oxygen-limited zone. Moreover, elevated chlorophyll \u003cem\u003ea\u003c/em\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec) and ammonium concentrations (Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e) at this depth supported the inference that 90 meters is near the lake's bottom [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eLong-term microbial time series in Lake Soyang\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo explore spatiotemporal microbial dynamics, we analyzed prokaryotic community compositions across all collected samples. The analysis initially revealed a total of 5,745 ASVs. After filtering out rare ASVs (\u0026lt;\u0026thinsp;10 reads), 2,119 ASVs representing 36 distinct phyla remained (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The rarefaction curves plateaued in all samples (Fig. S3a), suggesting that microbial diversity was well represented at the given sequencing depth. Similarly, the species accumulation curve for the entire dataset (n\u0026thinsp;=\u0026thinsp;300) showed that the sample size was sufficient to capture the majority of prokaryotic populations at the sampling location (Fig. S3b). Alpha diversity gradually increased with water depth (Fig. S4), except at 90 meters near the lake bottom. Temporally, alpha diversity declined markedly during summer in surface layers above the thermocline but increased during periods of water column mixing (Fig. S5).\u003c/p\u003e\u003cp\u003eNon-metric multidimensional scaling (NMDS) revealed clear clustering of microbial communities by both season and water depth (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Repeated seasonal trends were evident in samples from depths down to 20 meters (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). However, these trends gradually weakened in deeper waters, where temporal environmental variations were less pronounced. Samples collected near the bottom (90 meters) were clearly distinguished from those at shallower depths (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). A closer examination of microbial abundance patterns between the 70- and 90-meter samples (Fig. S6) highlighted the presence of microbial population adapted to oxygen-limited conditions at 90 meters, including ASV12 from \u003cem\u003eMethylotenera\u003c/em\u003e, ASV64 from \u003cem\u003eNitrosarchaeum\u003c/em\u003e, and three ASVs (ASV48, ASV49, and ASV335) from \u003cem\u003eFlavobacterium\u003c/em\u003e. Surprisingly, ASV31 from acI-A (\u0026lsquo;\u003cem\u003eCandidatus\u003c/em\u003e Planktophila\u003cem\u003e\u0026rsquo;\u003c/em\u003e), a lineage typically dominant in oxygenated water columns [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e], also emerged as part of this population. This novel ASV was notably absent at other depths (Fig. S7), further suggesting its ecological specialization for the oxygen-limited hypolimnion.\u003c/p\u003e\u003cp\u003e\u003cb\u003eEcological processes driving seasonal microbial dynamics\u003c/b\u003e\u003c/p\u003e\u003cp\u003eWe next quantified the ecological processes driving the seasonal dynamics of community assembly using a phylogenetic bin-based null model analysis (iCAMP) [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. To enhance data interpretability and reduce technical noise, we focused on 243 ASVs (hereafter referred to as core ASVs) that exhibited high relative abundance (\u0026gt;\u0026thinsp;1% maximum abundance) and frequent occurrence (present in \u0026gt;\u0026thinsp;10% of the samples). These core ASVs (Fig. S7) accounted for 83.6% of total reads. Analysis of this curated dataset revealed that community assembly was influenced by both deterministic (homogeneous or heterogeneous selection) and stochastic (dispersal limitation, homogenizing dispersal, and drift) processes, with their relative importance varying markedly across months and depths (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea\u0026ndash;e). Drift was the most dominant stochastic process overall, followed by homogeneous selection as the leading deterministic process. Dispersal limitation had a greater influence (35% on average) in the oxygen-limited bottom layer (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee) compared to shallower depths (26\u0026ndash;33.9%).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo explore fine-scale temporal dynamics, we focused on the lake\u0026rsquo;s epilimnion (1 meter), which exhibited the most pronounced temporal variations in community compositions (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). The core ASVs at the epilimnion were assigned to 22 phylogenetic bins (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef). Since the importance of ecological processes was weighted by the relative abundance of each bin [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], a small subset (n\u0026thinsp;=\u0026thinsp;6) of dominant bins (bin4\u0026ndash;6, bin9, bin11, and bin21) accounted for over 67% of all processes (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). During winter and early spring (Jan.\u0026ndash;Mar.), homogeneous selection was driven primarily by bin4 and bin5, both represented by genome-streamlined actinobacteria such as acI and acIV clades. Stochastic processes were most influential in April, accounting for up to 96.6% of all inferred processes (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). This exceptionally high prevalence could be attributed to the combined contribution of multiple bins (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef), highlighting the complex and unpredictable nature of community dynamics during this period. As water temperatures rose following this period, homogeneous selection gradually gained influence, reaching a maximum of 29% in August (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). This increased relative importance of the deterministic process in mid-season was primarily contributed by bin6, represented by \u003cem\u003eFonsibacter\u003c/em\u003e. In the final quarter of the season (Oct.\u0026ndash;Dec.), \u003cem\u003eSediminibacterium\u003c/em\u003e and \u003cem\u003eFlavobacterium\u003c/em\u003e from bin 21 made notable contributions to the deterministic process (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef). In contrast, low-abundance bins were predominantly influenced by stochastic processes throughout the year, rather than by deterministic factors.\u003c/p\u003e\u003cp\u003e\u003cb\u003eAnnually recurring and episodic succession of planktonic bacteria\u003c/b\u003e\u003c/p\u003e\u003cp\u003eWe next aimed to accurately capture the timing, intensity, and duration of microbial succession driven by the complex interplay of deterministic and stochastic processes. To systematically classify the observed patterns, we applied an unsupervised soft clustering approach to uncover annually recurring microbial succession in the lake epilimnion (1 meter). In parallel, episodic blooming events throughout the water column (down to 40 meters) were identified using a Bayesian model averaging algorithm, which decomposes time-series abundance trajectories to detect abrupt shifts. Samples from 2015 and 2017 were excluded from the analysis due to incomplete 12-month coverage. Based on these analyses, approximately one-third (n\u0026thinsp;=\u0026thinsp;85) of the core ASVs were categorized as annually recurring, 28% (n\u0026thinsp;=\u0026thinsp;69) as episodic, and the remaining (n\u0026thinsp;=\u0026thinsp;89) as having weak or inconsistent temporal signals (Tables S3 and S4).\u003c/p\u003e\u003cp\u003eThe annually recurring microbes were further categorized into six distinct clusters (Table S3), each characterized by a unique seasonal development pattern. The first two clusters (C1 and C2) were associated with winter (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea), a season often considered as a \u0026lsquo;dormant period\u0026rsquo;, and displayed peaks of differing durations. ASVs from C1 exhibited a broad peak throughout the winter, predominantly consisting of cold-adapted taxa (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb) such as \u003cem\u003eMethylopumilus\u003c/em\u003e [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e] and \u003cem\u003eRhodoferax\u003c/em\u003e [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. In contrast, ASVs associated with C2 showed a narrow peak (Jan.\u0026ndash;Feb.), likely linked to water column turnover. During this period, microbial populations typically confined to the cold hypolimnion, including \u003cem\u003eMethylobacter\u003c/em\u003e, \u003cem\u003eMethylotenera\u003c/em\u003e, \u003cem\u003eNitrospira\u003c/em\u003e, and \u003cem\u003eOpitutus\u003c/em\u003e, became enriched. As temperatures gradually increased in spring, copiotrophs such as \u003cem\u003eLimnohabitans\u003c/em\u003e, \u003cem\u003eFlavobacterium\u003c/em\u003e, \u003cem\u003ePolynucleobacter\u003c/em\u003e, and the uncharacterized \u003cem\u003e\u0026lsquo;Candidatus\u003c/em\u003e Planktoluna\u0026rsquo; from C3 became dominant in surface waters. In a previous model [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], oligotrophic bacteria like \u003cem\u003eFonsibacter\u003c/em\u003e were strongly associated with the clear-water phase typically observed in May or June. In our study, C4 aligned with this phase and included \u003cem\u003eFonsibacter\u003c/em\u003e, CL500-3, \u003cem\u003eSandarakinorhabdus\u003c/em\u003e, \u003cem\u003eLimnobacter\u003c/em\u003e, and \u003cem\u003eNemorincola\u003c/em\u003e. During peak summer temperatures, ASVs from \u003cem\u003eCyanobium\u003c/em\u003e and potential phytoplankton-associated bacteria such as \u003cem\u003eTerrimonas\u003c/em\u003e, \u003cem\u003eNoviherbaspirillum\u003c/em\u003e, and \u003cem\u003eRoseomonas\u003c/em\u003e dominated, collectively forming C5. Lastly in C6, highly similar to the previous model [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], potential decay-associated taxa such as \u003cem\u003e\u0026lsquo;Candidatus\u003c/em\u003e Nitrosotenuis\u0026rsquo;, \u003cem\u003eNitrosarchaeum\u003c/em\u003e, and \u003cem\u003eChthoniobacter\u003c/em\u003e became dominant during the fall season (Sep.\u0026ndash;Nov.). These seasonal clusters exhibited robust interannual resilience and were consistently observed at depths down to 20 meters (Figs. S8\u0026ndash;10).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAlthough phylogenetic distance is often regarded as a proxy for niche separation [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], genome-streamlined actinobacteria demonstrated extensive niche diversification, even among closely related members within individual lineages. Annually recurring ASVs (n\u0026thinsp;=\u0026thinsp;15) from these lineages, including acI-A, acI-B (\u003cem\u003e\u0026lsquo;Candidatus\u003c/em\u003e Nanopelagicus\u0026rsquo;), acI-C, and acIV (CL500-29), were distributed throughout the year and associated with four of the six identified clusters (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb). A previous high-resolution study in Lake Zurich reported that the seasonal abundance of these lineages was highly associated with algal and cyanobacterial blooms during the growing season [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. However, nearly half of the ASVs linked to the winter cluster (C1) in our dataset indicated the existence of cold-adapted populations that remain poorly characterized.\u003c/p\u003e\u003cp\u003eIn contrast to the predictable patterns of annually recurring taxa, episodic bloomers exhibited irregular seasonal trajectories throughout the sampling period. Based on the number of change points (CPs) detected in their abundance trends, we classified these events into three categories: persistent (CP\u0026thinsp;=\u0026thinsp;1), transient (CPs\u0026thinsp;=\u0026thinsp;2), and intermittent (CPs\u0026thinsp;=\u0026thinsp;3) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea). Persistent bloomers, the most common type across depths (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb), underwent abrupt shifts in seasonal patterns that did not revert to their original states in subsequent years (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ed). Transient bloomers were characterized by short-lived and non-repeating peaks, appearing in a specific year and absent in others (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ee). At 40 meters, a few ASVs displayed intermittent blooming, appearing in the first and third years but absence in the intervening year, suggesting that such disturbances may recur under specific environmental conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ef). Interestingly, the number of episodic bloomers increased with depth, suggesting a higher degree of randomness in microbial succession in deeper water layers (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb). Genome-streamlined actinobacteria and fast-growing opportunists such as \u003cem\u003eLimnohabitans\u003c/em\u003e from \u003cem\u003eGammaproteobacteria\u003c/em\u003e and were the most frequently identified taxa among episodic bloomers (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec and Table S4).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cb\u003eTime-dependent associations among microbial and environmental components\u003c/b\u003e\u003c/p\u003e\u003cp\u003eTo further elucidate the complex interactions between microbial communities and environmental parameters, an association network was constructed using local similarity analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). The resulting network revealed 762 associations (edges) among 84 components (nodes), including 503 positive and 259 negative associations, as well as 494 synchronous and 268 delayed associations. As expected, 76% of the positive and synchronous associations occurred within the same clusters, demonstrating strong internal connectivity within the clusters (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). In contrast, 78.7% of the delayed associations were observed between different clusters, highlighting a well-coordinated and sequential succession of microbial populations. Temperature emerged as the most central environmental component in the network, with the highest number of associations (n\u0026thinsp;=\u0026thinsp;55). Among microbial components, ASV40 from acI-C and ASV6 from \u003cem\u003eAnaerolineaceae\u003c/em\u003e were identified as keystone taxa, each exhibiting 41 distinct associations with other nodes.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSeveral episodic bloomers were also integrated into the network, providing insights into underlying mechanisms potentially driving their dynamics. Notably, the blooming of ASV81 (\u003cem\u003eCyanobium\u003c/em\u003e) was found to be influenced by ASV46 (\u003cem\u003eLimnohabitans\u003c/em\u003e) and ASV138 (\u003cem\u003eBeijerinckiaceae\u003c/em\u003e), the latter of which is known as a potential nitrogen-fixer [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e] (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb). Other episodic bloomers, including ASV100 (\u003cem\u003ePolynucleobacter\u003c/em\u003e), ASV37 (SL56 marine group from \u003cem\u003eChloroflexota\u003c/em\u003e) and ASV65 (\u003cem\u003eVerrucomicrobiia\u003c/em\u003e), formed up to 25 associations with other nodes, indicating complex microbial cascades underlying their succession.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur long-term microbial time series data, combined with quantitative modeling and statistical approaches, facilitated an in-depth analysis of freshwater microbial dynamics. Investigating the ecological processes underlying bacterial community assembly highlighted a balance between stochastic and deterministic influences, with stochastic processes playing a more dominant role (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The strong influence of drift across seasons suggests that inter-annual variations, such as the timing of phytoplankton blooms [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e] or the intensity of thermal stratification and mixing [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e], may amplify the randomness and unpredictability of bacterial community structures. Previous studies have shown that the relative importance of ecological processes might vary considerably within a single ecosystem [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e], depending on the monitoring framework, including sampling frequency (high vs. low), scale (spatial vs. temporal), and study duration (weeks to years). This emphasizes a key challenge in ecological research, which involves designing sampling strategies and analyzing datasets to detect underlying ecological processes with greater accuracy and sensitivity.\u003c/p\u003e\u003cp\u003eBy successfully incorporating samples from winter and early spring (Jan.\u0026ndash;Apr.), our study addressed key gaps in a recent investigation [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], and provided a more continuous view of the seasonal dynamics of freshwater bacteria (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). The identification of two distinct winter clusters emphasized the role of cold-adapted bacteria in driving seasonal microbial succession. This study also illustrated freshwater microbial dynamics at a finer resolution by reconstructing seasonal succession patterns at the ASV level rather than the broader taxonomic level [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. This approach led to the discovery of previously overlooked ecological niches for genome-streamlined actinobacteria [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e], including those associated with cold winter periods (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb) and oxygen-limited hypolimnion (Fig. S6). The remarkable consistency of microbial resilience observed over multiple years suggests that these communities are likely to serve fundamental and irreplaceable roles in maintaining ecosystem function.\u003c/p\u003e\u003cp\u003eTracking microbial dynamics over multiple years allowed us to distinguish annually recurring microbes from episodic bloomers (Figs.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Episodic bloomers may reflect two ecologically distinct scenarios: i) they may represent functionally redundant populations that temporarily occupy available niches without serving essential ecosystem functions, or ii) they could emerge in response to major environmental shifts, such as species invasions [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], global warming [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e], or eutrophication [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. In the latter scenario, episodic bloomers could act as sensitive indicators of ecological instability or transitional states. Understanding the cumulative impact of episodic bloomers requires continuous, high-resolution monitoring of microbial communities [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e], which is particularly crucial given the accelerating pace of global environmental change and the increasing frequency and intensity of ecological disruptions [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eBy quantifying all detectable connections among microbial and environmental components, we further resolved the full spectrum of microbial dynamics into an association network (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). This network-based approach offers a valuable opportunity to advance beyond conventional ecosystem assessments, which typically rely on metrics such as physical and chemical parameters [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e] or biological indices based on a limited number of taxa [\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. By capturing patterns such as premature or delayed successions, missing microbial components at critical time points, or the emergence of new components forming novel connections, we can gain deeper insights into how microbial disturbances are triggered and propagate throughout the network [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e], ultimately altering ecosystem functions and stability over time.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThis long-term microbial time series successfully addressed critical gaps in earlier studies, providing a comprehensive view of seasonal microbial dynamics in a freshwater ecosystem. Future studies could expand the scope of this approach by integrating additional trophic levels, such as viruses or protists, incorporating a wider range of environmental parameters, and refining microbial network models for more accurate ecosystem assessments. Ultimately, these advancements hold the potential to transform freshwater ecosystem management and conservation through more effective, data-driven strategies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003ePEG Plankton ecology group\u003c/p\u003e\u003cp\u003eDO Dissolved oxygen\u003c/p\u003e\u003cp\u003eASV Amplicon sequence variant\u003c/p\u003e\u003cp\u003eNMDS Non-metric multidimensional scaling\u003c/p\u003e\u003cp\u003eCP Change points\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;to participate\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe sequencing data from this study were deposited to the NCBI Sequence Read Archive (SRA) under BioProject PRJNA1253076. All other relevant data supporting the findings of this study are available within the paper and its supplementary information files.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by\u0026nbsp;the Mid-Career Research Program (NRF-2022R1A2C3008502 to J.-C.C.) and the Sejong Science Fellowship (NRF-2022R1C1C2004070 to S.K.) through the National Research Foundation (NRF) funded by the Ministry of Sciences and ICT of the Korea government.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJ-CC and IK conceived and designed the experiment. SP and SK collected the data. HP, SK, and SP conducted data analyses. HP wrote the initial draft of the manuscript with all authors contributing to its revisions. All authors reviewed and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe gratefully acknowledge captain Do Gyeom Lee and the crews of the R/V of the Korea Water Resources Corporation (K-water) for their assistance with freshwater sample collection.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBaron JS, LeRoy Poff N, Angermeier PL, Dahm CN, Gleick PH, Hairston NG, et al. Meeting ecological and societal needs for freshwater. Ecol Appl. 2002;12:1247\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDudgeon D, Arthington AH, Gessner MO, Kawabata ZI, Knowler DJ, L\u0026eacute;v\u0026ecirc;que C, et al. Freshwater biodiversity: Importance, threats, status and conservation challenges. Biol Rev Camb Philos Soc. 2006;81:163\u0026ndash;82.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eArora-Williams K, Olesen SW, Scandella BP, Delwiche K, Spencer SJ, Myers EM, et al. Dynamics of microbial populations mediating biogeochemical cycling in a freshwater lake. Microbiome. 2018;6:1\u0026ndash;16.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVellend M. Conceptual synthesis in community ecology. Q Rev Biology. 2010;85:183\u0026ndash;206.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChase JM, Myers JA. Disentangling the importance of ecological niches from stochastic processes across scales. Philosophical Trans Royal Soc B: Biol Sci. 2011;366:2351\u0026ndash;63.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChave J. Neutral theory and community ecology. Ecol Lett. 2004;7:241\u0026ndash;53.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRiddley M, Hepp S, Hardeep FNU, Nayak A, Liu M, Xing X, et al. Differential roles of deterministic and stochastic processes in structuring soil bacterial ecotypes across terrestrial ecosystems. Nat Commun. 2025;16:2337.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNing D, Yuan M, Wu L, Zhang Y, Guo X, Zhou X, et al. A quantitative framework reveals ecological drivers of grassland microbial community assembly in response to warming. Nat Commun. 2020;11:4717.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAguilar P, Sommaruga R. The balance between deterministic and stochastic processes in structuring lake bacterioplankton community over time. Mol Ecol. 2020;29:3117\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGe Y, Meng X, Heino J, Garc\u0026iacute;a-Gir\u0026oacute;n J, Liu Y, Li Z et al. Stochasticity overrides deterministic processes in structuring macroinvertebrate communities in a plateau aquatic system. Ecosphere. 2021;12.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSommer U, Adrian R, De Senerpont Domis L, Elser JJ, Gaedke U, Ibelings B, et al. Beyond the plankton ecology group (PEG) model: Mechanisms driving plankton succession. Annu Rev Ecol Evol Syst. 2012;43:429\u0026ndash;48.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePark H, Shabarova T, Salcher MM, Kosov\u0026aacute; L, Rychteck\u0026yacute; P, Mukherjee I, et al. In the right place, at the right time: the integration of bacteria into the Plankton Ecology Group model. Microbiome. 2023;11:112.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePaver SF, Newton RJ, Coleman ML. Microbial communities of the Laurentian Great Lakes reflect connectivity and local biogeochemistry. Environ Microbiol. 2020;22:433\u0026ndash;46.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKent AD, Yannarell AC, Rusak JA, Triplett EW, McMahon KD. Synchrony in aquatic microbial community dynamics. ISME J. 2007;1:38\u0026ndash;47.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRobin R, Rohwer, Riley J, Hale, Zanden MJ, Vander TR, Miller, Katherine D, McMahon. Species invasions shift microbial phenology in a two-decade freshwater time series. Proceedings of the National Academy of Sciences. 2023;120:e2211796120.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMartin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011;17:10\u0026ndash;2.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCallahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13:581\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQuast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Res. 2013;41:590\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSwedish Biodiversity Infrastructure. (SBDI; 2021). SBDI Sativa curated 16S GTDB database. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.17044/scilifelab.14869077\u003c/span\u003e\u003cspan address=\"10.17044/scilifelab.14869077\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcMurdie PJ, Holmes S. Phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data. PLoS ONE. 2013;8.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKandlikar GS, Gold ZJ, Cowen MC, Meyer RS, Freise AC, Kraft NJB, et al. ranacapa: An R package and Shiny web app to explore environmental DNA data with exploratory statistics and interactive visualizations. F1000Res. 2018;7:1734.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDixon P. Computer program review VEGAN, a package of R functions for community ecology. J Veg Sci. 2003;14:927\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eH W. ggplot2: elegant graphics for data analysis. New York: Springer-Verlag. 2016. p. ISBN 978 3.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKatoh K, Standley DM. MAFFT multiple sequence alignment software version 7: Improvements in performance and usability. Mol Biol Evol. 2013;30:772\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMinh BQ, Schmidt HA, Chernomor O, Schrempf D, Woodhams MD, Von Haeseler A, et al. IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era. Mol Biol Evol. 2020;37:1530\u0026ndash;4.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKumar L, Futschik ME, Mfuzz. A software package for soft clustering of microarray data. Bioinformation. 2007;2:5\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhao K, Wulder MA, Hu T, Bright R, Wu Q, Qin H, et al. Detecting change-point, trend, and seasonality in satellite time series data to track abrupt changes and nonlinear dynamics: A Bayesian ensemble algorithm. Remote Sens Environ. 2019;232:111181.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXia LC, Ai D, Cram J, Fuhrman JA, Sun F. Efficient statistical significance approximation for local similarity analysis of high-throughput time series data. Bioinformatics. 2013;29:230\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13:2498\u0026ndash;504.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKong D, Kim B. Suggestion for Trophic State Classification of Korean Lakes. J Korean Soc Water Environ. 2019;35:248\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBhadury P, Sen A. Understanding Impact of Seasonal Nutrient Influx on Sedimentary Organic Carbon and Its Relationship With Ammonia spp. in a Coastal Lagoon. Front Mar Sci. 2020;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOkazaki Y, Fujinaga S, Tanaka A, Kohzu A, Oyagi H, Nakano SI. Ubiquity and quantitative significance of bacterioplankton lineages inhabiting the oxygenated hypolimnion of deep freshwater lakes. ISME J. 2017;11:2279\u0026ndash;93.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSalcher MM, Schaefle D, Kaspar M, Neuenschwander SM, Ghai R. Evolution in action: habitat transition from sediment to the pelagial leads to genome streamlining in Methylophilaceae. ISME J. 2019;13:2764\u0026ndash;77.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMar\u0026iacute;n Irma and Arahal DR. The Family Beijerinckiaceae. Rosenberg Eugene and DeLong EF and LS and SE and TF, editor. The Prokaryotes: Alphaproteobacteria and Betaproteobacteria [Internet]. Berlin, Heidelberg: Springer Berlin Heidelberg; 2014. pp. 115\u0026ndash;33.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYuan J, Cao Z, Ma J, Li Y, Qiu Y, Duan H. Influence of climate extremes on long-term changes in cyanobacterial blooms in a eutrophic and shallow lake. Sci Total Environ. 2024;939.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWoolway RI, Sharma S, Weyhenmeyer GA, Debolskiy A, Golub M, Mercado-Bett\u0026iacute;n D et al. Phenological shifts in lake stratification under climate change. Nat Commun. 2021;12.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang J, Shen J, Wu Y, Tu C, Soininen J, Stegen JC, et al. Phylogenetic beta diversity in bacterial assemblages across ecosystems: Deterministic versus stochastic processes. ISME J. 2013;7:1310\u0026ndash;21.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRoguet A, Laigle GS, Therial C, Bressy A, Soulignac F, Catherine A et al. Neutral community model explains the bacterial community assembly in freshwater lakes. FEMS Microbiol Ecol. 2015;91.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNeuenschwander SM, Ghai R, Pernthaler J, Salcher MM. Microdiversification in genome-streamlined ubiquitous freshwater Actinobacteria. ISME J. 2018;12:185\u0026ndash;98.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOuyang J, Wu H, Yang H, Wang J, Liu J, Tong Y et al. Global warming induces the succession of photosynthetic microbial communities in a glacial lake on the Tibetan Plateau. Water Res. 2023;242.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXie G, Zhang Y, Gong Y, Luo W, Tang X. Extreme trophic tales: deciphering bacterial diversity and potential functions in oligotrophic and hypereutrophic lakes. BMC Microbiol. 2024;24.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMini\u0026egrave;re A, von Schuckmann K, Sall\u0026eacute;e JB, Vogt L. Robust acceleration of Earth system heating observed over the past six decades. Sci Rep. 2023;13.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSuresh K, Tang T, Van Vliet MTH, Bierkens MFP, Strokal M, Sorger-Domenigg F et al. Recent advancement in water quality indicators for eutrophication in global freshwater lakes. Environ Res Lett. 2023;18.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFortunato CS, Eiler A, Herfort L, Needoba JA, Peterson TD, Crump BC. Determining indicator taxa across spatial and seasonal gradients in the Columbia River coastal margin. ISME J. 2013;7:1899\u0026ndash;911.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAylagas E, Borja \u0026Aacute;, Tangherlini M, Dell\u0026rsquo;Anno A, Corinaldesi C, Michell CT, et al. A bacterial community-based index to assess the ecological status of estuarine and coastal environments. Mar Pollut Bull. 2017;114:679\u0026ndash;88.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eR\u0026ouml;ttjers L, Faust K. From hairballs to hypotheses\u0026ndash;biological insights from microbial networks. FEMS Microbiol Rev. 2018;42:761\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCarlstr\u0026ouml;m CI, Field CM, Bortfeld-Miller M, M\u0026uuml;ller B, Sunagawa S, Vorholt JA. Synthetic microbiota reveal priority effects and keystone strains in the Arabidopsis phyllosphere. Nat Ecol Evol. 2019;3:1445\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Freshwater ecosystem, Microbial community, Seasonal dynamics, Ecological model","lastPublishedDoi":"10.21203/rs.3.rs-7144747/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7144747/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eFreshwater bacteria are key contributors to ecosystem stability and resilience, yet their inherent microdiversity and rapid community turnover make interpreting their seasonal dynamics within broader ecological frameworks challenging. Furthermore, the underlying mechanisms of community assembly, driven by deterministic and stochastic processes, remain poorly understood.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eWe conducted a five-year time series sampling of microbial communities, alongside key environmental parameters, in an oligo-mesotrophic lake. Seasonal community assembly in this freshwater ecosystem was primarily driven by drift, dispersal limitation, and homogeneous selection, with the relative contributions of these processes varying markedly across months and depths. Six distinct clusters of seasonally recurring microbes shaped the timeline of annual community transitions, while episodic populations sporadically occupied the remaining ecological niches. Network-based reconstruction of microbial dynamics uncovered a highly interconnected structure, where microbial and environmental components exhibited coordinated, time-dependent variation.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eThis study highlights the importance of long-term ecological monitoring for resolving microbial dynamics at a fine scale. Our findings establish a foundation for future approaches aiming to apply microbial metrics to freshwater ecosystem assessment and management.\u003c/p\u003e","manuscriptTitle":"Long-term freshwater time series reveals recurrent and episodic microbial dynamics driven by distinct assembly mechanisms","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-17 17:55:53","doi":"10.21203/rs.3.rs-7144747/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0c774416-f214-4b4f-b1f3-feab2a084671","owner":[],"postedDate":"November 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-19T15:56:10+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-17 17:55:53","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7144747","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7144747","identity":"rs-7144747","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}