Unravelling the molecular diversity of marine cyanobacterial communities in the lagoon of Nouméa (New Caledonia): impact of a cyclonic event
Christophe Six, Mathisse Meyneng, Morgane Ratin, Priscilla Gourvil, Raffaele Siano, Hugues Lemonnier
TL;DR
This study explores how cyclones and environmental factors shape cyanobacterial communities in a tropical lagoon, revealing shifts in species diversity and distribution.
Contribution
The study identifies a 'minor cyanosphere' and links cyclonic events to changes in cyanobacterial community structure in a subtropical lagoon.
Findings
A single Synechococcus ASV dominated nutrient-rich coastal waters, while diverse subclusters and Prochlorococcus ecotypes were found in other zones.
Cyclone Uesi caused a surge of estuarine Synechococcus strains transported to offshore reefs, altering community composition.
Cyanobacterial assemblages showed spatial structuring, indicating potential as bioindicators of water masses in lagoon systems.
Abstract
In tropical marine ecosystems, cyanobacteria are key components of phytoplankton communities, yet their diversity and spatio-temporal dynamics in tropical lagoons remain poorly documented. Here, we use the Nouméa lagoon (New Caledonia) as a model system to genetically characterize cyanobacterial communities in a subtropical lagoon subjected to both natural and anthropogenic pressures. Using metabarcoding, we investigated seasonal and spatial variations along environmental gradients, spanning from estuarine zones to oligotrophic reef areas. A single Synechococcus ASV dominated most communities, particularly in nutrient-enriched coastal waters. Beyond this dominant lineage, we uncovered a diverse “minor cyanosphere,” which comprised Synechococcus subclusters 5.1, 5.2, and 5.3, several Prochlorococcus ecotypes, an as-yet-uncultivated group of picocyanobacteria, and various diazotrophic…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7- —CRESICA
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMicrobial Community Ecology and Physiology · Marine and coastal ecosystems · Marine and coastal plant biology
Introduction
Phytoplankton forms the foundation of marine ecosystems, sustaining ecological balance. Among them, cyanobacteria, particularly the picocyanobacteria Prochlorococcus and Synechococcus, dominate global oceanic phytoplankton communities and rank among the most abundant phototrophs on Earth. Over the past decade, advances in molecular tools, especially metagenomics, have greatly expanded our understanding of their genetic diversity and revealed fine-scale ecological adaptations and niche partitioning ([1–3]. In warm oligotrophic tropical oceans, Prochlorococcus typically dominates, with several genetically distinct ecotypes adapted to varying light regimes [4, 5]. Likewise, Synechococcus encompasses multiple phylogenetic lineages with differing ecological preferences [6]. Clade II of subcluster 5.1 often predominates in mesotrophic tropical waters and may account for over 80% of Synechococcus populations in non–iron-limited regions [1]. In contrast, the CRD1 clade of the same subcluster is well adapted to iron-depleted environments [7]. In tropical estuarine and coastal environments, members of subcluster 5.2, alongside certain clades from subcluster 5.1, are frequently detected.
Another major component of tropical cyanobacterial communities is nitrogen-fixing diazotrophs. Equipped with the nitrogenase enzyme, they convert atmospheric N₂ into bioavailable forms. Diazotrophic cyanobacteria display diverse morphologies and lifestyles, from filamentous Trichodesmium spp. forming buoyant colonies to free-living unicellular types such as Crocosphaera watsonii (UCYN-B) and UCYN-C [8]. Some, like UCYN-A and Richelia intracellularis, form symbioses with prymnesiophytes or diatoms, respectively [9]. In oligotrophic waters, these organisms play pivotal roles in nitrogen cycling and vertical export [10–12].
While oceanic cyanobacterial communities have received considerable attention, tropical lagoons remain largely underexplored. The Nouméa Lagoon in New Caledonia is an emblematic tropical lagoon ecosystem located in the Southwest Pacific Ocean, north of the Tropic of Capricorn. This ecosystem is renowned for its exceptional endemic biodiversity and has been designated a UNESCO World Heritage Site. Grande Terre, the main island of the archipelago, is encircled by ~1000 km of coral reefs that delineate a lagoon of 22 000 km^2^. The coral reef barrier is interrupted by natural passes, relics of ancient river mouths, which today allow the exchange between oceanic and lagoonal waters. The Nouméa Lagoon experiences a mild tropical climate with pronounced seasonality, with sea surface temperatures ranging from ~23°C in winter (May–October) to over 29°C in austral summer (November–April) [13]. Precipitation peaks typically occur in February [14], coinciding with the cyclone season. Indeed, while strongly affected by anthropogenic pressure, the Nouméa Lagoon is also particularly exposed to extreme weather events, including intense cyclones that bring heavy rainfall and strong winds [15]. These disturbances result in marked increases in freshwater inputs from rivers such as the Dumbéa, potentially altering the lagoon’s ecological dynamics [16, 17].
Previous studies have characterized phytoplankton in the Nouméa Lagoon, reporting seasonal Chl a maxima in May–June, similar to patterns observed in other tropical lagoons [13, 18]. Microphytoplankton in the bays is generally dominated by chain-forming diatoms, while outside these areas, nanoplanktonic coccolithophores, picoeukaryotes, and dense populations of Prochlorococcus and Synechococcus cyanobacteria prevail [19, 20]. Flow cytometry and spectrofluorometric analyses have revealed at least two distinct Synechococcus optical types in the lagoon: green-light specialists near the coast and blue-light specialists in offshore and barrier reef zones [21, 22]. However, no data are currently available on the genetic structure of this abundant picocyanobacterial community, hindering the identification of existing ecotypes and the potential discovery of lineages characteristic of tropical lagoon ecosystems. In addition, as nitrogen limitation has been recognized as a key factor shaping phytoplankton productivity [18], a number of studies have documented the diversity and activity of diazotrophic cyanobacteria in the lagoon [8, 23, 24].
In this study, we present a comprehensive genetic characterization of cyanobacterial communities in the Nouméa Lagoon using a metabarcoding approach. We analyze community structure at the ecotype level across multiple bays and along two environmental gradients extending from coastal areas to the barrier reef, over different seasons, allowing the comparison of lagoonal communities with those from other tropical coastal and open-ocean systems. We also assess the ecological response of these communities to a major disturbance caused by a cyclonic event that occurred in February 2020 and, by integrating results from laboratory-maintained cyanobacterial strains, we explore potential mechanisms underlying the subsequent community restructuring. Our findings provide new insights into the ecological functioning of tropical lagoons and highlight the potential of cyanobacterial communities as bioindicators of environmental variations.
Material and methods
Study area and sampling strategy
The study took place in the Southwest Lagoon (Fig. 1A), influenced by the Dumbéa, Coulée, and Pirogues rivers and connected to the ocean via the Dumbéa and Boulari passages. We sampled two coast–offshore transects (stations A–C, G–I) and three additional coastal stations (D–F) during two dry-season periods with minimal river input: September 2019 and December 2020 (Fig. 1B, Table S1). These samples provided spatial and temporal insights under non-disturbed “baseline” conditions. In February 2020, Cyclone Uesi (category 3, N–S trajectory) triggered major water runoff in the Dumbéa estuary [17]. Two stations, Dumbéa Bay (DB) and Grande Rade Bay (GR), were monitored daily from February 12–17. In addition, two drifting buoys (B1, B2) released from DB on days 2 and 3 tracked and sampled surface waters over two days each (Fig. 1C).
Localization of the study area (A) and sampling plan in the Nouméa lagoon. B: During two dry seasons (September 2019 and December 2020), five coastal stations were sampled (A: Dumbéa Bay; F: Grande Rade; E: Sainte-Marie Bay; D: Coulée Bay; G: Pirogues), along with two coast-offshore transects (A, B, C: Dumbéa Bay, Dumbéa Intermediate, Dumbéa passage; G, H, I: Pirogues, Pirogues intermediate, Boulari passage). C: In February 2020, the Dumbéa Bay (DB) and Grande Rade (GR) stations were monitored for six days, starting from the day after cyclone Uesi. Two drifting buoys were deployed (green lines) from the DB station on day 1 (B1) and day 2 (B2), with water samples collected twice along their trajectories. See also Table S1.
Measurements of environmental parameters
Temperature and salinity were recorded with a SBE19 CTD (Sea-Bird). Salinity was also measured in the water samples collected at 1 m depth (see below). Chemical analyses on 20 L seawater samples included NH₄^+^, Si(OH)₄, PO₄^3−^, and NO_x_ (NO₃^−^ + NO₂^−^), measured as previously described [17]. Chl a was measured fluorometrically on acetone extracts (GF/F filters, Whatman). Dissolved organic matter (<0.2 μm) optical properties allowed the calculation of the biological (BIX) and humification (HIX) indices, reflecting respectively recent autochthonous DOM and its source/maturation [17, 25, 26].
Flow cytometry analyses of environmental samples
Seawater samples collected at 1 m depth were fixed with 1% glutaraldehyde (v/v), stored in liquid N₂, and analyzed by flow cytometry (Accuri™ C6, BD-Biosciences; 488/640 nm lasers, four detectors). Prochlorococcus populations were unresolved, but Synechococcus were clearly identified and enumerated based on side scatter (SSC), phycoerythrin orange (585/40 nm) and chlorophyll red (670 nm) fluorescence, with 1 μm beads as internal standards.
Amplicon sequencing and statistical analyses
Seawater samples (4–15 L) collected at 1 m depth were sequentially filtered through 20 μm, 3 μm and 0.2 μm filters, covering 3 size fractions representing approximatively micro-, nano- and picoplankton, resulting in a total of 129 samples (Table S2). All filtration units were thoroughly rinsed with seawater from the sampling site prior to use, and new filter holders were employed for each size fraction to prevent cross-contamination. DNA was extracted using the Nucleospin Plant II Mini DNA extraction kit (Macherey–Nagel). The V4-V5 region of the 16S rRNA gene was PCR-amplified [27] in triplicate. PCR reactions were performed in a total volume of 30 μL containing 0.12 μL GoTaq Flexi G2 DNA Polymerase (5 U μL^−1^; Promega), 6 μL of 5× GoTaq buffer, 1.8 μL MgCl₂ (25 mM), 0.6 μL dNTP mix (10 mM), 0.45 μL of each primer (20 μM), 18.58 μL sterile water, and 2 μL of template DNA. The PCR thermal profile consisted of an initial denaturation at 95°C for 5 min, followed by 32 cycles of denaturation at 95°C for 30 s, annealing at 50°C for 60 s, and extension at 72°C for 60 s, with a final extension at 72°C for 10 min. Three negative controls were included in each PCR batch to monitor potential contamination. The amplicons were pooled and sequenced using Illumina technology at the GeT-Biopuces platform (INSA, France). Raw data are available on Sextant Ifremer, at doi:10.12770/26840434-8354-4856-9d5ee562c8de7252.
Raw data were processed using the Standardized and Automated MetaBarcoding Analyses (SAMBA) pipeline developed by the Ifremer’s department SeBiMer (https://github.com/ifremer-bioinformatics/samba) [17]. Rarefaction curves confirmed saturating sequencing depth, yielding 378 924 cyanobacterial reads (2937 ± 2264 per sample, Table S3) and capturing most bacterial 16S rRNA diversity, as shown by saturation curves and Good’s coverage values (Fig. S1 and S2, Table S4). ASVs were automatically assigned using SILVA v138.1. Cyanophyta sequences were extracted and manually curated with Megablast to remove heterotrophic bacteria and plastids (Table S5). Picocyanobacterial ecotypes were refined via phylogenetic analysis (Maximum Likelihood and Neighbor Joining, MEGA-X) using a reference 16S V4–V5 database of cultured cyanobacteria with publicly available genomes (https://cyanorak.sb-roscoff.fr/). ASV counts were expressed as per-sample relative abundances to mitigate major sequencing-depth effects. Interpretations focused on compositional rather than absolute abundance changes. Given that the dominant taxa Prochlorococcus and Synechococcus (subcluster 5.1) typically harbor two 16S rRNA gene copies in the known genomes, relative comparisons among them were considered more reliable.
The statistical analyzes were performed using R v4.4.1 [28] (https://github.com/Matmey/Six_et_al_Cyanobacteria_NC). Good’s coverage index [29] and rarefaction curves were computed to assess sequencing depth and overall sequencing quality. Cumulative Sum Scaling normalization was applied before any compositional analysis [30], while presence/absence transformation was done for beta-diversity analyses. Variations in alpha diversity were assessed using ASV richness and the Shannon index using the phyloseq package [31]. PERMANOVA analyses were then conducted using the adonis() function to evaluate the influence of qualitative factors (season, cyclone occurrence, sampling campaign, and size fraction) on community composition. β-diversity was assessed via Principal Coordinates Analysis for each size fraction to evaluate community changes across sampling campaigns. Spearman correlations explored relationships among taxa and with environmental parameters. To investigate the relationships between environmental parameters and the distribution of the cyanobacterial community at the station scale, we performed a distance-based redundancy analysis (db-RDA) using presence/absence data. The db-RDA was computed using the capscale() function from vegan [32] with Jaccard distance, including temperature (T), ammonium (NH₄^+^), phosphate (PO₄^3−^), silicate (Si), oxidized nitrogen (NO_x_), total chlorophyll a (TChla), and two fluorescence-derived indices of organic matter composition (HIX and BIX) as explanatory variables. To assess the relative contribution of each group of parameters, we applied hierarchical partitioning using the rdacca.hp package [33]. Environmental variables were grouped into four categories: Temperature, Nutrients (NH₄^+^, PO₄^3−^, Si, NO_x_), Organic matter (HIX, BIX), and Chl a. The adjusted R^2^ values from the db-RDA were partitioned among these groups to quantify their independent and joint effects on cyanobacterial community structure.
Isolation and halotolerance capacities of Caledonian Synechococcus
Synechococcus sp. RCC7811 was isolated from a mangrove directly connected to the lagoon and to a shrimp pond, 80 km north of Nouméa (E 166° 5′21.22″; S 21°55′59.13″). Water was filtered (3 μm) and enriched with modified PCR-S11 medium [34]. Clonality was achieved via two serial dilutions to extinction, and the 16S rRNA gene was sequenced (Accession PQ037838).
Synechococcus sp. RCC7811, along with Synechococcus sp. NOUM97-013, which was isolated from offshore lagoon water (E 166°19′59″; S 22°19′59″), were grown in vented polycarbonate flasks at 23°C under 80 μmol photons m^−2^ s^−1^ continuous white light. Halotolerance was assessed by acclimating cultures to varying seawater concentrations while keeping nutrients constant. Cell density and cytometric parameters were monitored by flow cytometry (Advanteon, Agilent), and growth rates were calculated from the exponential growth phase slope [34].
Results and discussion
The September and December samplings reflected non-disturbed dry-season conditions. Seawater was ~3°C warmer in December, with coastal stations generally warmer than offshore ones (Fig. 2). Despite minimal river input, a coast-to-offshore gradient of suspended matter and nutrients persisted, consistent with previous studies [13, 20, 35]. Our results support these observations, showing higher organic matter indices (HIX, BIX) and nutrients (PO₄^3−^, NH₄^+^, NOx, Si(OH)₄) at coastal stations compared to the more stable offshore sites in both months (Fig. 2). Chl a, a proxy for phytoplankton biomass and nutrient status [36, 37], showed a gradient, averaging 0.72 ± 0.40 μg L^−1^ at coastal stations and 0.23 ± 0.02 μg L^−1^ offshore. These values agree with earlier studies [8, 19, 22], and resemble non-eutrophic lagoons [36, 38].
Variations in hydrological parameters at all sampled stations during the two studied dry seasons (September and December) and at Dumbéa and Grande Rade bays during the monitoring period following the cyclone event. Coastal stations: A: Dumbéa Bay; F: Grande Rade; E: Sainte-Marie Bay; D: Coulée Bay; G: Pirogues. Coast-offshore transects: A: Dumbéa Bay, B: Dumbéa intermediate, C: Dumbéa passage; G: Pirogues, H: Pirogues intermediate, I: Boulari passage. DB: Dumbéa bay; GR: Grande Rade; B: buoy.
Cyanobacterial ASV diversity in the Nouméa lagoon
We identified 51 cyanobacterial ASVs in Nouméa lagoon (Fig. 3, Table S5) and successfully assigned most sequences to genus or ecotype, including four Prochlorococcus ASVs (Fig. S3), 29 Synechococcus ASVs (Fig. S4 and S5), and 18 diazotrophic cyanobacteria ASVs (Fig. 3, S6). Three Synechococcus SC5.1 ASVs remained unresolved due to short sequence length (374 nt). These cyanobacterial ASVs represented a dominant proportion of the bacterial community (mean of 42.9 ± 9.35%, 33.0 ± 11.4%, and 33.3 ± 11.5% of reads in September 2019, December 2020, and February 2020, respectively, Fig. S7).
Phylogenetic tree and read relative abundance of cyanobacterial ASV sequences. This figure presents a maximum likelihood phylogenetic tree (Kimura 2-parameter model with gamma distribution) including 51 ASV sequences (rRNA 16S V4-V5, 374 nucleotides), using the ancestral cyanobacterium Gloeobacter violaceus as outgroup. A comparable topology was observed using the neighbor joining method. Bootstrap values greater than 80% are marked with dark grey circles. On the right, the relative read abundance of each ASV is depicted with a color scale across the 25 samplings, differentiated by three size classes. The dominant Synechococcus ASV’s read abundance is illustrated with diamonds on a yellow-red color scale, representing its proportion relative to the total cyanobacterial reads in each sample. The read abundances of the other ASVs are shown as circles on a green-violet color scale, indicating their proportion relative to the total reads of the minor cyanosphere, i.e. all reads except those of the dominant ASV. Synecho: Synechococcus, Prochloro: Prochlorococcus, Inter: intermediate, Off: offshore. Coastal stations: A: Dumbéa Bay; F: Grande Rade; E: Sainte-Marie Bay; D: Coulée Bay; G: Pirogues. Coast-offshore transects: A: Dumbéa Bay, B: Dumbéa intermediate, C: Dumbéa passage; G: Pirogues, H: Pirogues intermediate, I: Boulari passage. DB: Dumbéa bay; GR: Grande Rade; B: buoy. The complete read abundance dataset is available in Table S3. For relative read abundance of the phylogenetic groups, see Fig. S6.
Picocyanobacteria
Picocyanobacteria numerically dominate phytoplankton in the New Caledonian lagoon and adjacent oceanic waters [19]. In Nouméa lagoon, our results show that the Synechococcus community is dominated by a single SC5.1 ASV from the clades II/III/EnvC/UCA radiation (ASV9; Fig. 3, S5). Although sequence length prevents precise placement, it likely belongs to clade II, consistent with reports that such Synechococcus often comprise >80% of tropical, iron-replete open-ocean communities [1–3]. In iron-limited regions, CRD1 Synechococcus are usually dominant [7, 39] (Table S6), but they were not detected in Nouméa lagoon, where iron concentrations can be substantial (3.5–38 nM, Dec 2020; unpublished). The dominant Synechococcus ASV9 co-occurred with eight closely related ASVs (SC5.1, clades II/III EnvC/UCA) across the lagoon (Fig. 3, S5, Table S5), some correlating with coastal parameters like silicates and organic matter (Fig. 4). While ASV2 and ASV5 were rare, ASV7 persisted year-round. Other ASVs showed seasonal patterns: ASV1 in December, ASV3 and ASV6 from February to September. Notably, one or more of these ASVs were always present in the lagoon, regardless of the season. Some of these ASVs may belong to clade III, common in warm waters (Table S6), but close phylogenetic similarity with clade II prevents definitive assignment [3, 40, 41].
Spearman correlation heatmap of the read abundances between each cyanobacterial ASV and to the environmental parameters, during the dry seasons (September and December). Only significative correlation coefficients are shown (P < .05), following the color scale. The dominant Synechococcus ASV is shown in bold. T: seawater temperature.
The dominant Synechococcus ASV9, mainly retrieved in smaller size fractions (Fig. 3), comprised ~90% of cyanobacterial reads in summer, slightly less in September. Its contribution declined in central/offshore waters, notably near the Boulari passage (Fig. 3), reflecting coastal-to-offshore nutrient gradients influenced by terrestrial inputs versus oceanic waters. In the lagoon’s oligotrophic offshore waters, Prochlorococcus showed higher relative abundance (Fig. 3) and negative correlations with coastal parameters like silicate, organic matter and Chl a (Fig. 4). It dominated the smallest size fractions at intermediate and offshore stations, and at coastal Pirogues. The Pirogues–Boulari transect (GHI) is known to be more oligotrophic than Dumbéa (ABC) due to trade-wind-driven nutrient advection, leading to lower Chl a [22]. At Boulari passage, Prochlorococcus accounted for ~80% of cyanobacterial reads (Fig. S6, S8), consistent with previous divinyl-Chl a based observations in Nouméa lagoon and similar sites [42, 43]. Prochlorococcus ASV diversity included three HLII clade sequences: one from the globally common surface HLII subclade and two related to HLII-DG, recently described as inhabiting mid-depth waters [44]. Additionally, one ASV belonging to the Prochlorococcus LLI clade was detected, a lineage commonly found in mixed water environments [45]. Therefore, this Prochlorococcus community appears well-adapted to the lagoon’s variable conditions and reef-upwelling influence, and may constitute an indicator of the lagoon’s oligotrophic status.
Besides the single dominant Synechococcus ASV, our study unveiled a developed assemblage of other cyanobacteria showing interesting occurrence patterns. We thereafter collectively refer to these communities as the “minor cyanosphere” of the Nouméa lagoon. The diversity of this minor cyanosphere included other marine Synechococcus clades from the highly diversified SC5.1 lineage, along with SC5.2 and SC5.3 Synechococcus, Prochlorococcus ecotypes, and different groups of nitrogen fixers.
The SC5.1 Synechococcus included three ASVs of the WPC1 clade, previously detected in the eastern Mediterranean Sea [1, 46] and along the Chinese coasts [47–49], and whose ecological preferences are poorly known (Table S6). In Nouméa Lagoon, they appeared as coastal cyanobacteria occurring mostly in September and February in our study. Synechococcus clade WPC1 often co-occurred with clade V Synechococcus (Fig. 4), a clade usually detected in coastal waters [50]. In the Pearl river estuary, the presence of Synechococcus clade WPC1 is highly correlated with Clade VI [51], another typically coastal group of Synechococcus closely related to clade V [3, 6]. Clade VI Synechococcus was not detected, but unassigned ASV3 was potentially related to clade VI (Figs. 3, S5). WPC1’s absence in December and presence in September and February suggests a seasonal pattern independent of temperature.
SC5.1 clade IX Synechococcus, initially described in the Gulf of Aqaba, remain ecologically poorly characterized [52, 53] (Table S6). Though rare in the global ocean [1, 3], they are ubiquitous in coastal Chinese waters, often co-occurring with halotolerant SC5.2 Synechococcus[48–51] and some strains can physiologically acclimate to a wide salinity range [50]. Our study revealed that clade IX Synechococcus are a prominent component of the coastal minor cyanophere in the Nouméa lagoon (Fig. 3, Table S3), represented by two frequently co-occurring ASVs. They were most abundant in December and February when temperatures were elevated (Fig. 3 and 4), and exhibited a clear affinity for coastal stations, especially in the estuary of the Dumbéa River. Notably, although particle adhesion or aggregation cannot be excluded, their frequent detection in the largest size-fraction samples (Figs 3, S6, S8) suggests that these Synechococcus may form colonies or engage in symbiotic associations.
SC5.2 Synechococcus, which markedly co-occur with clade IX (Fig. 4), are estuarine picocyanobacteria (Table S6) and include at least two clades, CB4 and CB5, sometimes classified as Cyanobium [6, 54]. We detected a single SC5.2 ASV (CB5), found exclusively at coastal stations, particularly in the Dumbéa estuary, where it can represent ~40% of the minor cyanophere (Figs 3, S4, S8). Its prevalence in total cyanobacterial reads markedly increases in December and February, suggesting a preference for warm waters. Given its distinct niche in the Nouméa lagoon, this Synechococcus appears as an interesting indicator of warm waters characterized by variable salinity.
Synechococcus SC5.3 constitute a monophyletic radiation encompassing at least six clades [1, 55, 56]. In marine environments, they have been detected in specific locations of the world ocean, including the Sargasso Sea, the Mediterranean Sea, the Red Sea, Bermuda, the Mozambique Channel and along the Chinese coasts [1, 46, 47, 49, 57]. Despite limited knowledge on their biology and ecology (Table S6), they appear to prefer relatively warm waters, with certain lineages possibly displaying depth niche partitioning, an uncommon feature among marine Synechococcus [46, 55]. Four SC5.3 ASVs were detected in the lagoon, with the two major ones belonging to clade I ^55^ (Fig. S4). These ASVs show distinct spatio-temporal patterns: ASV1 predominates offshore in September, while ASV2 is widespread during December and February (Fig. 3), indicating strong seasonality that may explain their patchy detection in global metagenomic surveys.
We identified a cluster of five related ASVs forming a distinct clade with strong bootstrap support (Fig. S5), unrelated to any known cultivated Synechococcus. The closest BLAST hits came from SC5.1 Clade I sequences from Massachusetts coasts (e.g. Synechococcus sp. MV1307) [58, 59], which are adapted to cold temperate/subpolar waters [34, 60–62] and absent from subtropical Nouméa Lagoon. The group of ASVs we identified in New Caledonia exhibits a rather low sequence identity with Clade I, ranging from 93% to 95% (Table S5) and occupy a basal position within marine Synechococcus cluster 5 (Figs. 3, S5). They occur mainly in February, correlating with coastal parameters (Fig. 3 and 4). Related environmental sequences (max 95.7% identity) have been reported from California, Red Sea, China Sea, Fiji, and Tasmania. These findings suggest a distinct, potentially novel marine Synechococcus lineage, for which longer sequences, additional markers, and culture isolation are needed for confirmation.
Three SC5.1 sequences remained unassigned, limiting discussion of their distribution and seasonality. Unassigned ASV1 occurred mainly in September, while ASV2 was widespread (Fig. 3). Phylogenetic placement, though weakly supported, suggests some may belong to the understudied clade VII. Notably, a model strain from clade VII, Synechococcus sp. RCC2433 (NOUM97013, Cyanorak database), with a complete genome, was isolated from offshore Nouméa Lagoon.
Diazotrophic cyanobacteria
Nitrogen has been identified as one of the major limiting factors for phytoplankton in the Nouméa Lagoon [8, 18, 19]. In addition to picocyanobacteria, we detected filamentous and/or endosymbiotic diazotrophic cyanobacteria mainly in larger size fractions. While picocyanobacterial richness was uniform across stations, diazotroph molecular richness increased at offshore stations compared to coastal ones (Fig. S9 and S10), consistent with previous observations of rising nifH gene abundance from coast to barrier reef [8].
We detected one ASV of the filamentous diazotroph Trichodesmium erythraeum (>100–1000 μm), a major contributor to tropical N₂ fixation [63]. Its presence and activity in New Caledonian waters are well documented [8, 22, 23, 64] was sporadically detected (Fig. 3), consistent with its tendency to form large patchy blooms in coastal and offshore stations of the lagoon, during all seasons. In our dataset, T. erythraeum was sporadic (Fig. 3), reflecting its patchy distribution in all the lagoon throughout the year, consistent with previous observations, though peak abundances usually occur in summer [8].
The UCYN-A group (Candidatus Atelocyanobacterium thalassa) comprises nitrogen-fixing cyanobacteria living as obligate endosymbionts of unicellular eukaryotes [9, 65, 66], now often considered “nitroplasts” [67, 68]. These are emerging as significant diazotrophic systems, being reported as dominant and active in various coastal and offshore environments [69–71]. We identified a single UCYN-A ASV, found only in large size fractions. Comparison with 18S rRNA data from the same samples [17] showed that 92% of UCYN-A-positive samples also contained the Prymnesiophyte Braarudosphaera, its known host [9]. The detected UCYN-A ASV exhibits 100% sequence identity with the 16S V4-V5 region of the CPSB-1 isolate genome (Table S5). Despite the less conventional use of the 16S rRNA gene, as opposed to the nifH gene in most studies, this strongly suggests the affiliation of this cyanobacterial endosymbiont with the UCYN-A2 clade. UCYN-A2 is considered a more coastal ecotype than UCYN-A1, which aligns with its occurrence in a lagoon environment [71]. However, unlike T. erythraeum, this ASV was absent from coastal stations (Fig. 3 and 4), suggesting specialization for ocean-influenced waters within the lagoon [72]. Previous studies in Nouméa Lagoon have examined UCYN-A ecotypes, showing higher UCYN-A2:UCYN-A1 ratios inside the lagoon and elevated UCYN-A1 offshore, with temperature influencing their distribution [11, 24, 73–75]. Our data, despite not detecting UCYN-A1, suggest that UCYN-A2 associates with lower temperatures, with higher read abundances in September, consistent with findings in other cooler environments [71, 76].
Other diazotrophic cyanobacteria, often heterocyst-forming, are part of diatom-diazotroph associations (DDAs), including Richelia strains with Rhizosolenia (Het-1) and Hemiaulus (Het-2) [77], and Calothrix (Het-3), epiphytic on Chaetoceros but sometimes free-living [78]. Due to their ballasted nature, DDAs contribute to the export of particulate matter and organic carbon to the deep sea, especially in subtropical gyres [12, 79]. Among DDAs, we identified three ASVs matching Calothrix (98% identity; Table S5), found only in the largest size fractions. They were consistently detected across the lagoon in September and December, often at relatively high abundance in the minor cyanosphere (Fig. 3, S6). An exception was the Dumbéa estuary, where Calothrix was absent, possibly due to sensitivity to fluctuating salinity. These symbiotic cyanobacteria were the most prevalent diazotrophs by ASV reads, though genome 16S copy number is unknown, contrasting with a previous report of low nifH abundances for Calothrix [24].
Previous studies have reported other photosynthetic diazotrophs in New Caledonian waters, including UCYN-A1, UCYN-B/C (Crocosphaera, Cyanothece) [80] and Richelia DDAs [23, 81]. We did not detect the endosymbiotic, heterocyst-forming Richelia, whereas it was reported as abundant in mesocosm experiments in the Nouméa lagoon [24].
Sponges (Porifera) are a major component of Nouméa lagoon’s benthic fauna. Among the ~50 bacterial phyla associated with sponges, cyanobacteria are typically the most abundant [82, 83]. We identified an assemblage of 12 cyanobacterial ASVs related to Leptolyngbya, a key component of sponge microbiomes [84]. Over the past decade, recognition that many traditional genera, including Leptolyngbya, Lyngbya, and Pseudanabaena, are polyphyletic has led to extensive taxonomic revisions [85, 86]. In particular, Leptolyngbya has been split into many new genera, including Nodosilinea, Haloleptolyngbya, Oculatella, Pantanalinema, Alkalinema, Leptothoe, Salileptolyngbya, Pegethrix, Drouetiella, Rhodoplaca, Cymatolege, and Metis [82, 84, 87–92]. In this study, we collectively refer to these ASVs as the “Leptolyngbya group” (Fig. 3, S6). The blast best hits of each ASVs are available in Table S5.
Studies across various sponge species show that up to half of symbiotic diazotrophs may be cyanobacteria actively expressing nifH, such as Leptolyngbya minuta [93, 94]. These cyanobacteria are typically intracellular, can be transmitted vertically or horizontally and their presence in the water column is expected [92]. While this may include some non-diazotrophs, Leptolyngbya has been detected globally in marine, coastal, and terrestrial ecosystems, often as a significant component of cyanobacterial communities [89, 92, 95–97]. In this study, Leptolyngbya ASVs were absent in December and largely missing from coastal stations, except at the more oligotrophic Pirogues site (Fig. 3, S6). Prevalence patterns divided them into two groups: nine ASVs clustered tightly together, suggesting similar distribution and abundance, while three grouped with major diazotrophs (T. erythraeum, UCYN-A, Oscillatoria, Calothrix), indicating closer ecological affinities (Fig. 4, S11). Leptolyngbya has been largely overlooked in Nouméa lagoon diazotroph studies, yet this sponge-associated community*—*also found as coral symbionts [98, 99]- may substantially contribute to nitrogen cycling across reef benthic and pelagic compartments.
Dynamics of the cyanobacterial communities after a cyclonic event
In February 2020, cyclone Uesi caused a 40-fold increase in Dumbéa River flow within one day (Fig. S12), while the Coulée and Pirogues rivers were less affected. Modeling showed floodwaters first spreading toward the coral barrier, then moving southeast [17]. Grande Rade was slightly impacted on day 2. The plume peaked on day 3 before diluting into the lagoon. Environmental parameters were monitored in Dumbéa Bay and Grande Rade for six days after the event.
Immediately after the cyclone, water temperature in Dumbéa Bay dropped to 24°C (Fig. 2), unusually low for February but typical of freshwater inflow from the Dumbéa River. It then rose over three days to the seasonal norm of ~29°C, while Grande Rade showed a smaller, delayed increase. Salinity measurements (Fig. S13) and modeling [17] revealed a freshwater layer ~150 cm in Dumbéa Bay on day 1, which thinned to ~30–50 cm while propagating into the lagoon, indicating that metabarcoding samples from day 1 may have been influenced by the surface freshwater layer. The humification index (HIX), a proxy for terrestrial humic inputs, was high on day 1 in Dumbéa Bay but quickly declined to normal values, while remaining stable at Grande Rade (Fig. 2). Silicate showed the same trend. NO_X_ concentrations spiked immediately after the cyclone before returning to baseline, whereas ammonium rose progressively, peaking on day 4 and then decreasing (Fig. 2), suggesting active organic matter degradation. Overall, the cyclone introduced a pulse of cooler freshwater enriched in dissolved nitrogen, silicate, and humic substances, which subsequently degraded over time.
Only a few studies have examined phytoplankton community responses to cyclones in tropical and subtropical regions. Most reported increases in seawater Chl a concentrations following such events [100–103]. For instance, in New Caledonia, cyclone ERICA in 2003, though shorter than Uesi, caused a 5- to 6-fold rise in surface Chl a lasting several days, associated with the development of Chl (c_1_ + c_2_)-containing taxa, most likely diatoms [22]. In a similar way, our study revealed a 5-fold increase in Chl a concentration at the Dumbéa Bay station over the six-day monitoring period following cyclone Uesi (Fig. 2). This rise, attributed to a diatom bloom [17], was likely triggered by nutrient inputs from cyclone-driven river discharge. In contrast, phytoplankton biomass appeared unaffected at the Grande Rade station.
We investigated the cyclone’s effect on cyanobacteria first by analyzing changes in their contribution to phytoplankton, calculating the number of detected cyanobacterial reads normalized to the Chl a concentration in seawater (Fig. 5). During the dry seasons, this parameter remained fairly stable across all lagoon stations in both September and December, suggesting a consistent overall contribution of cyanobacteria to the phytoplankton communities, despite variations in their composition (Figs 3, S14, S15, Table S7). In Dumbéa Bay, however, the cyclone induced a sharp decrease in the cyanobacterial contribution to phytoplankton, especially in the larger size fractions, while picocyanobacteria remained unchanged, though slightly lower than in dry seasons. Over the next three days, the contribution of larger fractions gradually recovered to dry-season levels (Fig. 5). No significant change was observed at Grande Rade. Overall, this aligns with a previous study [17], which also reported a post-cyclone decline of cyanobacterial reads relative to total bacterial reads in large fractions.
Variations of the contribution of cyanobacteria to phytoplankton, expressed in decimal logarithm of cyanobacterial read abundance per chl a seawater concentration, during the dry seasons and following the Uesi cyclone. Inter: intermediate, Off: offshore. Coastal stations: A: Dumbéa Bay; F: Grande Rade; E: Sainte-Marie Bay; D: Coulée Bay; G: Pirogues. Coast-offshore transects: A: Dumbéa Bay, B: Dumbéa intermediate, C: Dumbéa passage; G: Pirogues, H: Pirogues intermediate, I: Boulari passage. DB: Dumbéa bay; GR: Grande Rade; B: buoy.
In Dumbéa Bay, the contribution of dominant Synechococcus after the cyclone remained stable and comparable to dry-season levels (Fig. 3, S16). However, Principal Coordinates Analysis revealed a shift in ASV composition relative to dry seasons, particularly in the picoplanktonic fraction (Fig. S14). Although PCR biases or template competition cannot be entirely excluded, Prochlorococcus reads -robustly detected in dry-season samples—were not observed in Dumbéa Bay after the cyclone, indicating a strong decline in abundance. During dry seasons, this taxon contributed moderately in Dumbéa Bay and dominated the minor cyanosphere in Grande Rade. In the six days following the cyclone, no Prochlorococcus reads were detected in Dumbéa Bay, whereas they reappeared in Grande Rade on the fifth day. After the cyclone, the minor cyanosphere in Dumbéa Bay included markedly more reads from the Synechococcus 5.2 ASV than during dry seasons (Fig. 3). These halotolerant, brackish-water cyanobacteria were likely transported from inland waters and probably also grew in situ, as their contribution continued increasing until the final day in Dumbéa Bay. They were also detected at Grande Rade (station F) until day three, despite being nearly absent there during dry seasons (Fig. 3). Overall, while the dominant Synechococcus maintained the global picocyanobacterial contribution, these shifts reveal notable community changes in response to the cyclone.
Flow cytometry analyses further supported these observations. Although Prochlorococcus could not be reliably enumerated, two Synechococcus cytometry populations were clearly distinguished by their fluorescence properties (Fig. S17). Marine Synechococcus display diverse optical types based on variations in their light-harvesting system, the phycobilisome [104–106]. Upon excitation by a cytometer laser at 488 nm, it has been shown that estuarine/coastal Synechococcus, whose phycobilisome is often rich in phycocyanobilin and phycoerythrobilin, form dim fluorescing cytometry populations, while offshore Synechococcus contain much phycourobilin which leads to higher fluorescence levels. To examine relative variations between these two optical types, we calculated the cell density ratio of dim to bright populations (Fig. 6). As expected, during the dry seasons, dim Synechococcus were generally more abundant than bright ones at coastal stations.
Variations of the ratio of the dim (Syn1) to bright (Syn2) Synechococcus cells, as determined by flow cytometry, during the dry seasons and following the Uesi cyclone. Inter: intermediate, Off: offshore. Coastal stations: A: Dumbéa Bay; F: Grande Rade; E: Sainte-Marie Bay; D: Coulée Bay; G: Pirogues. Coast-offshore transects: A: Dumbéa Bay, B: Dumbéa intermediate, C: Dumbéa passage; G: Pirogues, H: Pirogues intermediate, I: Boulari passage. DB: Dumbéa bay; GR: Grande Rade; B: buoy.
At the end of the cyclone, the dim/bright Synechococcus ratio was abnormally high in Dumbéa Bay (station A), reflecting a surge of dim cells that persisted until day five before returning to December-like values (Fig. 6). These dim cells, likely transported by the Dumbéa River, were also detected in Grande Rade, where the ratio remained elevated until day four. Previous works showed that, among the phycoerythrin-containing Synechococcus, the communities of the lagoon are dominated by phycoerythrobilin-rich cells [22], i.e. optical types 2 and 3a [105, 106]. We hypothesize that the bright Synechococcus population corresponds to phycoerythrobilin-rich cells, while the dim population is phycocyanobilin-rich (optical type 1). This would indicate that cyclone-induced river flow introduced a substantial input of inland and estuarine-type Synechococcus into the lagoon. Notably, most 5.2 subcluster Synechococcus are phycocyanin-rich, consistent with the metabarcoding results.
Two drifting buoys were released from Dumbéa Bay on days 2 and 3 to track river water dispersal (Fig. 1C). Buoy B1 circumnavigated the Nouméa peninsula toward the Coulée-Pirogues estuaries, while B2 drifted southwest toward the Dumbéa reef passage. Metabarcoding samples collected along their trajectories enabled assessment of cyclone-induced changes in offshore cyanobacterial communities. As expected, the four offshore samples contained offshore cyanobacteria such as Prochlorococcus, Synechococcus SC5.3 ASV1, and diazotrophs including Leptolyngbya, Trichodesmium, and UCYN-A (Fig. 3, S8). More unexpectedly, Synechococcus from subcluster 5.2 and SC5.1 clades IX and V, normally associated with coastal or estuarine habitats (Fig. 3), were also present. This indicates that the cyclone transported coastal cyanobacteria into offshore lagoon waters, generating atypical communities that persisted up to the coral barrier. Flow cytometry further supported this, revealing elevated abundances of dim coastal Synechococcus in offshore waters compared to dry-season conditions (Fig. 6).
It is noteworthy that we did not observe a massive influx of typical freshwater cyanobacterial ASVs following the river flow increase. However, in the buoy samples, we identified an ASV belonging to Synechococcus SC5.1 clade VIII (Fig. 3). Members of this clade are euryhaline picocyanobacteria [52], with most isolates being phycocyanobilin-rich (optical pigment type 1). These organisms are scarce in open-ocean environments [1] and are generally considered restricted to estuaries and inland waters, where salinity is highly variable. A previous study reported several ribotypes of clade VIII Synechococcus in shrimp ponds of the Nouméa region [107], which are connected to the lagoon via coastal mangroves. In our survey, this ASV was detected along both buoy trajectories, in the lagoon’s central area and near the Dumbéa passage (Fig. 3). Given that these organisms usually thrive in brackish environments, their occurrence in offshore waters is highly unexpected and strongly suggests transport from coastal or inland sources, such as mangroves and shrimp ponds, extending to the coral barrier.
To assess the potential consequences of coastal Synechococcus transport to coral barrier reefs, we examined two laboratory strains. The first, Synechococcus sp. strain RCC7811, was isolated from a mangrove intertidal zone connected to shrimp ponds. Although not an exact match to the ASV detected by metabarcoding, it belongs to SC5.2 (Fig. S4) and is phycocyanobilin-rich (Fig. 7A). We compared its fitness and salinity tolerance with Synechococcus sp. RCC2433 (SC5.1, clade VII, Fig. S5), a strain isolated from offshore lagoon waters near Maître Island. Our results showed that the coastal RCC7811 is highly halotolerant, with stable growth rates across a wide salinity range from seawater to freshwater (Fig. 7B). In contrast, the offshore RCC2433 displayed reduced growth below 25 g/L and could not grow at salinities under 12–15 g/L. Moreover, the coastal strain grew ~40% faster than the offshore strain, highlighting a clear competitive advantage under these conditions. Cytometric analyses revealed contrasting physiological responses. In RCC7811, forward light scatter (FSC, often related to cell biovolume) decreased to 50% of its seawater value at 5 g/L salinity (Fig. 7C). By contrast, RCC2433 maintained constant FSC until 12 g/L, where it suddenly increased, suggesting abnormal cell swelling. Fluorescence signals also diverged: pigment fluorescence remained stable in RCC7811, whereas RCC2433 showed elevated fluorescence at low salinity, likely reflecting chronic stress Fig. 7D and E). These results underscore the distinct salinity responses of these two Synechococcus strains and emphasize the competitive edge of coastal SC5.2 Synechococcus in these conditions.
Variations in growth rates and cytometric parameters of the coastal strain Synechococcus sp. RCC7811 and the offshore strain Synechococcus sp. RCC2433 (strain NOUM97013) grown across a range of salinities (g/L). A: Photo of batch cultures of both strains grown at 35 g/L salinity, evidencing the difference in pigmentation. B: Growth rates of cultures long-term acclimated to different salinities; C: Forward scatter (FSC), often related to the cell biovolume. D: Chlorophyll a cellular fluorescence (emission at 695 nm) and E: Phycobiliprotein (PBP) cellular fluorescence: phycocyanin for the coastal strain (emission at 667 nm) and phycoerythrin for the offshore strain (emission at 572 nm).
These observations may suggest that cyclones can transport fast-growing coastal Synechococcus from mangroves to the coral barrier, potentially allowing them to outcompete offshore populations and temporarily alter local planktonic community structure. Although RCC7811 exhibited much higher growth rates than the offshore strain, this advantage depends on the environment. Its strong growth performance is likely driven by nutrient-rich conditions, both in the laboratory and in estuarine or inland waters where resources are abundant. However, this competitive edge does not extend to oligotrophic reef waters, where nutrient scarcity probably strongly limits estuarine/coastal Synechococcus growth. Testing this hypothesis under controlled oligotrophic regimes will require dedicated future culture experiments. This highlights the crucial role of reef passages, which bring oligotrophic Pacific Ocean waters into the lagoon and help maintain the resilience of the coastal-offshore gradient in Nouméa.
Conclusion
This study provides a molecular characterization of cyanobacterial communities in the subtropical Nouméa lagoon based on 16S rRNA (V4–V5) metabarcoding, highlighting key similarities with the Pacific open ocean, particularly the dominance of clade II Synechococcus. The minor cyanosphere, however, shows distinct composition, reflecting the unique features of subtropical lagoons. These communities are well structured along environmental gradients, and may represent valuable indicators of ecosystem organization.
The lagoon’s structure and functioning rely on the interplay of multiple ecosystems, including rivers, estuaries, coastal zones, the central lagoon, and adjacent oceanic waters. Our findings show that cyclones disrupt this interconnected system, causing marked shifts in cyanobacterial communities and transporting inland species to the coral barrier. Seawater exchanges through reef passages, which maintain oligotrophic conditions in the outer lagoon, appear crucial for preserving the structured distribution of these communities following cyclonic events.
Supplementary Material
Six_et_al_-_Supplementary_material_updatedtitle_ycag032
Six_et_al_-_Supplementary_dataset_1_ycag032
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Farrant GK, Doré H, Cornejo-Castillo FM. et al. Delineating ecologically significant taxonomic units from global patterns of marine picocyanobacteria. Proc Natl Acad Sci USA 2016;113:E 3365–74. 10.1073/pnas.152486511327302952 PMC 4914166 · doi ↗ · pubmed ↗
- 2Kent AG, Baer SE, Mouginot C. et al. Parallel phylogeography of Prochlorococcus and Synechococcus. ISME J 2019;13:430–41. 10.1038/s 41396-018-0287-630283146 PMC 6331572 · doi ↗ · pubmed ↗
- 3Sohm JA, Ahlgren NA, Thomson ZJ. et al. Co-occurring Synechococcus ecotypes occupy four major oceanic regimes defined by temperature, macronutrients and iron. ISME J 2016;10:333–45. 10.1038/ismej.2015.11526208139 PMC 4737926 · doi ↗ · pubmed ↗
- 4Moore LR, Rocap G, Chisholm SW. Physiology and molecular phylogeny of coexisting Prochlorococcus ecotypes. Nature 1998;393:464–7. 10.1038/309659624000 · doi ↗ · pubmed ↗
- 5Ustick LJ, Larkin AA, Martiny AC. Global scale phylogeography of functional traits and microdiversity in Prochlorococcus. ISME J 2023;17:1671–9. 10.1038/s 41396-023-01469-y 37454234 PMC 10504305 · doi ↗ · pubmed ↗
- 6Doré H, Farrant GK, Guyet U. et al. Evolutionary mechanisms of long-term genome diversification associated with niche partitioning in marine Picocyanobacteria. Front Microbiol 2020;11:567431. 10.3389/fmicb.2020.56743133042072 PMC 7522525 · doi ↗ · pubmed ↗
- 7Ahlgren NA, Belisle BS, Lee MD. Genomic mosaicism underlies the adaptation of marine Synechococcus ecotypes to distinct oceanic iron niches. Environ Microbiol 2020;22:1801–15. 10.1111/1462-2920.1489331840403 · doi ↗ · pubmed ↗
- 8Saulia E, Benavides M, Henke B. et al. Seasonal shifts in Diazotrophs players: patterns observed over a two-year time series in the new Caledonian lagoon (Western tropical South Pacific Ocean). Front Mar Sci 2020;7:581755. 10.3389/fmars.2020.581755 · doi ↗