Invasive and native lake fish species display different responses of their gut microbiota and metabolite compositions along a eutrophication gradient

Invasive and native lake fish species display different responses of their gut microbiota and metabolite compositions along a eutrophication gradient | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 29 September 2025 V1 Latest version Share on Invasive and native lake fish species display different responses of their gut microbiota and metabolite compositions along a eutrophication gradient Authors : Alice Navarro , Nicolas Loiseau , Marc Troussellier , Pierre Foucault , Charlotte Duval , Julie Leloup 0000-0002-2777-284X , Manon Quiquand , Emilie Lance , Midoli Goto , Benjamin Marie , and Sebastien Duperron 0000-0002-6422-6821 [email protected] Authors Info & Affiliations https://doi.org/10.22541/au.175916041.10007367/v1 255 views 143 downloads Contents Abstract Figures Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Biological invasion and eutrophication are the main components of global changes in lacustrine ecosystems. The gut microbiome plays a vital role in animal health, yet its response to anthropogenic impact in the context of invasion has largely been overlooked. Here we examine the gut microbiota of two fish species, an invasive (Lepomis gibbosus) and a native one (Perca fluviatilis), in six lakes along a gradient of eutrophication revealed by environmental chlorophyll a (Chla) concentrations. We found that eutrophication level affects gut community and metabolite compositions of both species. At high Chla concentration we observed strong changes in microbiota composition, a signal of dysbiosis. Although metabolome and microbiota community compositions were correlated, the metabolome is less affected, suggesting limited functional consequences for the holobiont. This mismatch may be due to the dominant influence of host-derived metabolites and the functional redundancy within gut microbial communities. This redundancy acts as a buffer for metabolic outputs, despite shifts in microbiota composition. However, the response is species-specific, with the invasive L. gibbosus exhibiting more stable gut microbiota composition and functions across a wider range of conditions compared to the native P. fluviatilis. Moreover, tipping point (Chla threshold) beyond which gut microbiota dysbiosis occurs is lower in P. fluviatilis than in L. gibbosus. In light of the global rise in bloom frequency and intensity, we hypothesize that the enhanced resilience of gut microbiota and associated functions may represent an underappreciated advantage that enables certain invasive species to outcompete native counterparts. Introduction Harmful algal blooms (HABs) have been identified as a major threat to the rapidly declining freshwater biodiversity (Reid et al. 2019). The proliferation of phototrophic microorganisms plays a crucial role in shaping freshwater ecosystems (Huisman et al. 2018; Paerl & Otten 2013; Pavagadhi & Balasubramanian 2013). Climate change and increased eutrophication lead to increased frequency and intensity of blooms worldwide (Erratt et al. 2023; O’Neil et al. 2012). In this context, evaluating the various consequences of blooms, and more generally high phytoplankton abundances, on fauna is of prime importance to inform conservation biology (Danaher et al. 2022). Fish occupy upper trophic levels and thus are important indicators of freshwater ecosystem health. Phytoplanktonic blooms impose stress on fish populations either directly through exposure to their toxins (e.g. microcystins produced by the cyanobacterium Microcystis aeruginosa, Pavagadhi & Balasubramanian 2013) or indirectly via hypoxic/anoxic conditions provoked by bloom decays (Duperron et al. 2019). Lethal ( e.g. , mortality events) or sub-lethal effects ( e.g., adverse effects on organs, development, reproduction and behavior) of cyanotoxins are often reported (Le Manach et al. 2016, 2018; Malbrouck & Kestemont 2006; Saraf et al. 2018), yet non-lethal impact of chronic high phytoplankton abundances on fish remain understudied. Due to their role in host health, nutrition, immunity, homeostasis, and ultimately fitness (Egerton et al. 2018; Sullam et al. 2012), fish-associated microbiota have emerged as relevant to fish biology and ecology (Evariste 2019; Llewellyn et al. 2014). Located at the interface between a host and its environment, be it on the skin, gill or gut, the microbiota interacts with contaminants including toxins, making it a key player in host resistance and resilience (Duperron et al. 2020). The link between blooms, fish, and fish-associated microbiota has received attention only recently. Microcosm experiments on the medaka and zebrafish have shown that exposure to high yet environmentally-relevant concentrations of cyanobacterial metabolites (including microcystins) as well as live Microcystis aeruginosa cells could trigger changes in gut microbiota composition within a couple of days (Duperron et al. 2019; Qian et al. 2019; Foucault et al. 2022). These changes also affect metabolites content, a relevant proxy to evaluate functional traits of the holobionts (Foucault et al. 2022). Different levels of exposure induce dose-dependent responses, and gut microbiota composition was shown to be resilient, tending to return to its pre-exposure composition within days after the end of a simulated bloom, emphasizing the dynamics of microbiota response (Gallet et al. 2023). The same study however suggested that iterative blooms have comparable effects, with no evident priming effect observed following the first exposure. However, field studies are needed to confirm these findings in natural settings, and involving other fish and phytoplankton species. Besides blooms, biological invasions pose another threat to aquatic ecosystems, sometimes even more than anthropogenic-led changes, with invasive species topping the charts as primary drivers of plant and animal extinctions (Lopez et al. 2022). Freshwater fish are among the most commonly introduced organisms, with over 500 species translocated worldwide (Muñoz‐Mas et al. 2023). Once the species are established, their population growth can be facilitated by anthropogenic pressures, often leading to habitat modification, ecosystem degradation and biodiversity loss (Jakubčinová 2018). The success of an invasive non-indigenous species (NIS) depends on various known factors (Crooks & Rilov 2009). Differential resistance of host-associated microbiota towards highly eutrophicated waters has been hypothesized as another, yet unaccounted for, asset favoring the success of some invasive species (Escalas et al. 2022). Yet, despite that the microbiota of some invasive fish has been analyzed in a few species including carps (Eichmiller et al. 2016) or gobies (Gallo et al. 2020), studies on the microbiota of fish invaders remain scarce (Zhang et al. 2023), and very few have actually tried to compare the microbiota of invasive versus native species. Moreover, to our knowledge, no study has tested for differences in host-associated microbiota resistance to stress, such as high abundances of phytoplankton, between native and invasive species. This crucial comparison represents the initial step towards gaining a deeper understanding of how characteristics of associated microbiota could ultimately influence fish invasion success. In this study, gut prokaryotic microbiota and holobiont metabolome compositions are investigated and compared in two fish species sampled from six peri-urban lakes located around Paris (France), namely the native perch Perca fluviatilis and the invasive pumpkinseed Lepomis gibbosus . Sampled lakes display overall similar background characteristics (depth, geological and peri-urban context, meteorology…) yet different levels of phytoplankton abundances at the time of sampling, correlated to their different trophic status ranging from eutrophic to hypereutrophic (Foucault et al. 2025), thus mimicking a eutrophication gradient measured through chlorophyll a (Chl a ) levels. We determine the types of gut bacteria present and associated metabolome, and differences within and between perch and pumpkinseed that occur along this gradient. Finally, changes in microbiota diversity and composition are compared with changes occurring in the metabolome composition of the gut of the fish. To our knowledge, this study is the first comparison of gut microbiota and holobiont metabolomes made between a native and an invasive fish species along a phytoplankton abundance gradient. Material and methods Water sampling and environmental parameters Water samples and specimens of Eurasian perch ( Perca fluviatilis , 6–14 cm) and pumpkinseed ( Lepomis gibbosus , 8–12 cm) were collected from six artificial lakes located within a 66 km radius of Paris (France) between July 19th and 23rd, 2021 (Table S1, Supplementary Figure 1). The sampled lakes included Cergy-small (CERS), Cergy-large (CERL), Créteil (CTL), Verneuil-sur-Seine (VSS), La Grande-Paroisse (LGP), and Champs-sur-Marne (CSM). Reference water samples (500-1000 mL filtered on 0.22- µ m mesh) were collected using a Niskin bottle from three locations in each lake. Eluates were collected in duplicate (2x12 mL) for nutrient analyses in polyethylene tubes, with a specific acidification (three droplets of 3% HNO 3 solution) for orthophosphate (PO 4 3- ions) analysis. Dissolved mineral nitrogen (NH 4 + , NO 3 - and NO 2 - ions) and PO 4 3- concentrations were determined as described by (Holmes et al. 1999). The Chl a content (a proxy of phytoplankton biomass), was measured from 500 mL raw water filtered onto 0.7-µm filters (GF/C, Whatman, UK), in triplicate, by spectrophotometry (Cary 60 UV-Vis, Agilent, USA) (Yéprémian et al. 2016). Water temperature, dissolved O 2 (DO) and pH were measured on shore upon recovery (KS-2 MultiLine© probe, WTW, USA). Fish sampling Fish were sampled by electric fishing (Hydrosphere, France). Upon recovery, 10 specimens of Perca fluviatilis and Lepomis gibbosus per lake were sacrificed in accordance with regulations, for a total of 120 fish from six lakes. Whole guts including content, in order to prevent bias associated with differential removal of gut content, were dissected immediately, flash-frozen on site in liquid nitrogen and stored at -80 °C. Metabolite extraction and mass spectrometry analysis Serial extraction of metabolites and DNA was performed following the protocol described previously (Duperron et al. 2023). Supernatants containing metabolite extracts were transferred to mass spectrometry analyses. Gut pellets were dried and kept at -80 °C for subsequent DNA extraction on the same gut tissue (see below). Metabolites extracts were analyzed by Ultra high-performance liquid chromatography (UHPLC; ELUTE, Bruker) coupled with a high-resolution mass spectrometer (ESI-Qq-TOF Compact, Bruker) at 2 Hz speed, on simple MS mode then on broad-band Collision Ion Dissociation (bbCID) or autoMS/MS mode on the 50-1500 m / z range. Feature peak lists were generated from MS spectra within a retention time window of 1-15 minutes and a filtering of 5000 counts using MetaboScape 4.0 software (Bruker). The peak lists consisted of the area-under-the-peaks of extracted analytes from the gut tissue (Marie & Gallet 2022). Metabolites were annotated using GNPS-based molecular networking approach with publicly available MS/MS databases, including GNPS, HMDB, MS-DIAL, MetaboBase, LipidBlast and CASMI (Wang et al. 2016). DNA extraction and community 16S rRNA gene amplicon sequencing DNA was extracted from gut pellets and water filters using the ZymoBIOMICS DNA Miniprep kit (Zymo Research, California). Prior to DNA extraction, mechanical lysis was achieved on a bead beater (TissueLyser II, Qiagen, 6x1 min). The V3-V4 region of the prokaryotic 16S rRNA encoding gene was amplified using primers 341F (5’-CCTACGGGNGGCWGCAG -3’) and 806R (5′-GGACTACVSGGGTATCTAAT-3′) (Parada et al. 2016) and sequenced on an Illumina MiSeq 250×2 bp platform (GenoToul, Toulouse, France). An extraction blank was sequenced as a control. Reads were deposited into the Sequence Read Archive (SRA) database (Bioproject PRJNA1086840; supplementary table S2). 16S rRNA gene amplicon data processing Sequence analysis was performed using the QIIME2-2022.8 pipeline (Bolyen et al. 2018). Forward and reverse reads were trimmed at 240 and 220 bp, respectively. Amplicon Sequence Variants (ASVs) were obtained with the DADA2 plugin (ee=2, min. length = 10, (Callahan et al. 2016)) and affiliated with the SILVA 138-99 database (Quast et al. 2012). ASVs form mock communities, as well as those that were unassigned or assigned to Eukaryota, Chloroplast and Mitochondria, were discarded. ASVs which did not represent at least 1% of the reads in at least one of the samples were discarded. Sampling depth was rarefied to 5,500 reads, with all samples reaching saturation by then. The final dataset consisted of 2,650 ASVs (60 Perca fluviatilis , 58 Lepomis gibbosus , 18 water samples). Because the number of rare ASVs is very high, we then performed a sensitivity analysis by selecting only ASVs representing at least more than 10% of the number of reads in at least one sample. Biodiversity indices The diversity of microbial communities was analyzed using Hill numbers (Chao et al. 2014). Hill numbers provide a parametric family of diversity indices that differ by a parameter ”q” that determines their sensitivity to species relative abundances, all expressed in the same unit as ”equivalent number of species” (sensu species with equal dominance and equal functional distance). Taxonomic and phylogenetic diversities were described by two indices: richness (i.e. q = 0, indicating that all species have equal weight) and entropy (i.e. q = 1, as an exponential of the Shannon index). In the same way, dissimilarity in taxonomic composition (i.e. change in composition q = 0) and structure (q= 1) were computed. The metabolite diversity and dissimilarity was analyzed using the same framework. Statistical analyses Principal coordinates analysis was performed using the vegan package (2.6-6.1; (Oksanen et al. 2013) in R (4.3.2) to characterize lakes based on their environment. Collinearity between the environmental (explanatory) variables was also checked using a correlation matrix, followed by a calculation of the variance inflation factor (VIF) (Sohil et al. 2022). The smallest possible value for the VIF, 1, indicates the absence of collinearity, while value greater than 5 indicates very high collinearity (Zuur et al. 2010). Differences in ASVs relative abundances and presence/absence between species and among lakes were tested using the Kruskal-Wallis (KW) test, followed by Wilcoxon-Mann-Whitney (WMW) tests (pairwise lake-to-lake comparisons), with p-values adjusted for multiple comparisons (Bonferroni correction). Determinants of ASVs diversity We used generalized linear mixed models (GLMMs) to investigate the effect of environmental drivers on the four diversity. Selected variables included Chl a , pH, [dissolved O 2 ], temperature and salinity. Lakes and species were used as random effects. GLMMs were implemented in R (glmmTMB package v1.1.9, Brooks et al. 2017). The models were specified with diversity indices as the dependent variable, environmental variables as fixed effects, and lakes and species as random effects, following the general formula : Diversity_index ~Temperature + Chla + Salinity + pH + Oxygen + (1|lake) + (1|species). To directly compare diversity between the invasive and native species, we used non-parametric statistical tests. A Kruskal–Wallis test was performed to assess overall differences in diversity indices between species. When significant, pairwise comparisons were conducted using Wilcoxon–Mann–Whitney tests (p-values were adjusted using the Bonferroni correction). Determinants of ASVs composition To test the influence of environmental drivers on ASV composition (taxonomic and phylogenetic diversity), we applied generalized dissimilarity modeling (GDM, Ferrier et al. 2007), using the gdm package v1.5.0.9.1 in R. GDM examines the dissimilarities among pairs of host-associated microbiota as a nonlinear multivariate function of environmental variables of those host-associated microbiota, overcoming nonlinearity in dissimilarity between assemblages and environmental dissimilarities, and uneven rates of dissimilarities along environmental gradients (Ferrier et al. 2007). We specified the GDM using formula gdm(data = gdmTab.dis, geo = TRUE), after formatting the input data with formatsitepair() using a dissimilarity matrix. We selected the three I-spline basis function options. We accounted for spatial autocorrelation by including the geographic distance between pairs of lakes as a predictor variable. Percentage contribution of each variable was determined by comparing deviance of the full model to deviance of models without the variable of interest. Integration of bacterial community and metabolite datasets A correlative analysis was performed with DIABLO MixOmics ( v6.26.0) (Singh et al. Rohart et al. 2017), to analyze putative correlated dynamics between gut microbiota ASVs and gut metabolites. Briefly, the block.plsda() function integrates several datasets (named “blocks”) and performs a Pattern Latent Structure Discriminant Analysis. A correlation score for the given blocks was computed with the plotDiablo() function. Dose-response of metabolites Dose-response of metabolites identified in the metabolome analysis were evaluated against the Chl a concentrations measured in the different lakes for both species using DROMICS v2.6.2 (Larras et al. 2018). Plots displaying Empirical Cumulative Distribution Function (ECDF), representing the cumulative proportion of gut metabolites for which the dose-response threshold is reached versus normalized Chl a concentration (benchmark dose, BMD) were generated for both fish species. Physico-chemical parameters of sampled lakes At the time of sampling, the six lakes displayed temperatures between 23.5 and 26.3°C, oxygen saturation levels ranging from 115 to 165%, and pH values between 8.6 and 9.2 (Table S1). VSS lake displayed high orthophosphate levels compared to other lakes (3.42 vs . NH 4 + values compared to other lakes (4.75 and 1.12 µM, respectively, vs 0.48 to 0.81 µM for other lakes). Chl a concentrations were highly variable among lakes, ranging from 6.37 to 96.53 µg.L -1 in CERL and CSM lakes, respectively, with a range of intermediate values in between. Diversity of fish gut-associated communities Overall, 1370 and 1303 ASVs were identified in P. fluviatilis (each specimen displaying 14 to 174 ASVs) and L. gibbosus gut samples, (each specimen displaying 22 to 220 ASVs) respectively (average 172.1 ± 72.7 ASVs). No difference was found in ASV richness between gut communities from the two species (KW; p >0.05). Similar dominant genera and phyla were recovered from both species and from the different lakes, including members of the Firmicutes, Proteobacteria, and Cyanobacteriota (Figure 1, Supplementary Figure 2). Within a given lake, Firmicutes tended to display higher relative abundances in L. gibbosus compared to P. fluviatilis . Specifically, the relative abundance of Clostridium sensu stricto 1 was higher in L. gibbosus (0.0949 ± 0.0673) than in P. fluviatilis (0.0556 ± 0.0315). Romboutsia showed the same pattern (0.1101 ± 0.0852 vs. 0.0371 ± 0.0259). Phylum Verrucomicrobiota (including genus-level taxon LD29, Figure 1) was abundant in both species collected at CSM, the lake displaying highest Chl a levels (93.8 µg.l -1 ), and in P. fluviatilis specimens from LGP, the lake with the second highest Chl a level (41.9 µg.l -1 ); this phylum was otherwise rare in guts of fish from other lakes (supplementary Figure 2). Among the genera found abundant in the gut, genus LD29 (Verrucomicrobiota) was also abundant in the water at CSM, rare in LGP, and almost absent in other lakes (not shown). Influence of lake and fish species on gut bacterial community compositions The intra- and inter-lake pairwise dissimilarity levels in taxonomic composition (q0) and structure (q1) were high overall for both P. fluviatilis and L. gibbosus (Supplementary table S3), with intra-lake q0, q1 and inter-lake q0, q1 values for P. fluviatilis of 0.70±0.14, 0.64±0.23 and 0.89 ± 0.05, 0.91 ± 0.10 and for L. gibbosus of 0.71 ± 0.14, 0.58 ± 0.23 and 0.86 ± 0.07, 0.87 ± 0.18. Taxonomic q1 dissimilarities were lower than q0 dissimilarities in intra-lakes comparisons, for example 0.64 ± 0.23 vs . 0.70 ± 0.14 for P. fluviatilis , indicating that shared ASVs tended to be abundant. On the other hand, taxonomic q0 and q1 dissimilarity levels were comparable in inter-lake comparisons, for example 0.86 ± 0.07 and 0.87 ± 0.18 for L. gibbosus. Phylogenetic q1 dissimilarities, which account for phylogenetic distances between ASVs in pairwise comparisons, were markedly lower than taxonomic q1 ( e.g. 0.28 ± 0.14 versus 0.64 ± 0.23 in intra-lake comparisons of P. fluviatilis specimens, Supplementary table S 3). This indicates that abundant non-shared ASVs tended to be phylogenetically closely related. For all these metrics, within-species intra-lake comparisons yielded slightly lower dissimilarity values compared to inter-lake, indicating lake-related differentiation of gut communities in each species. All intra-lake dissimilarities between the gut communities of P. fluviatilis and L. gibbosus were higher than those found within a given species, indicating species-related differentiation of gut community compositions. These patterns were confirmed by PERMANOVA analyses, which showed that lake had a significant effect on gut bacterial community compositions of both species whatever the diversity metrics used (p < 0.001; Supplementary table S4), and that the explained variance (R²) was similar for both species for any given metrics. For taxonomic q0 dissimilarity for example, lakes explained 28.8% ( L. gibbosus ) and 30.8% ( P. fluviatilis ) of the variance. When accounting for abundance (q1), explained variance was even higher (R² = 35.7% for P. fluviatilis and 36.4% for L. gibbosus for taxonomic q0), indicating that lake of origin affects both the identity as well as abundances of ASVs. Ordination of community compositions according to the taxonomic and phylogenetic dissimilarities revealed several clusters that correspond to species and lakes (Figure 2). Regarding the taxonomic q0, the first axis separated fish from CSM (highest Chl a concentration) from all other lakes, while the second axis clearly separated communities associated with P. fluviatilis and L. gibbosus (Figure 2A). When accounting for abundances (q1), the PCoA plot still emphasized the difference between CSM and other lakes for both species, indicating that this difference is driven by identity as well as abundance variations of ASVs, while the distinction between species was less evident (figure 2B). Separation according to fish species was evident in the phylogenetic q0 dissimilarity plot, while separation according to lake was not (Figure 2C). This suggests that ASVs associated with the two fish species are more phylogenetically different than ASVs found in the different lakes within a given species. Finally, the phylogenetic q1 dissimilarity also separated species, although less markedly (Figure 2D). Levels of intra-lake taxonomic structure (q1) dissimilarity tend to decrease in P. fluviatilis with increasing Chl a , dropping sharply at VSS (Figure 3A), indicating more homogeneous communities in the gut of fish from the most Chl a -rich lakes. In L. gibbosus , lower intra-lake taxonomic structure (q1) dissimilarities are observed for both the least (CERL) and the most Chl a -rich lake (CSM, Figure 3B), peaking at intermediate concentrations. This suggests species-specific differences in the dose response. These trends are clear in both taxonomic and structure dissimilarities, but not in phylogenetic dissimilarities (not shown), suggesting replacement of ASVs by closely related ASVs in most cases. When comparing inter-species taxonomic q1 levels within a given lake, the dissimilarity drops at CSM, indicating that communities from both species become more similar in this lake which displays the highest Chl a level (Figure 3C). Factors structuring community compositions and consequences of high Chl a levels The GLMM models applied to alpha diversity shows low values for marginal R² and conditional R², whatever the index used. None of the relationships tested reached statistical significance (p > 0.05), suggesting a weak explanatory capacity of the variables included in the models to describe the observed variation in alpha diversity. The GDMs models were best at explaining the change in taxonomic composition q0, accounting for 45.8% and 32.1% of the total variance in P. fluviatilis and L. gibbosus , respectively (Figure 4). Their explanatory power was slightly lower when relative abundance was taken into account (q1 taxonomic dissimilarities, respectively 36.1 and 28.1%). Explanatory power was much lower on phylogenetic dissimilarities (11.3 for q0 and 19.0% for q1). The explanatory power of GDM models on taxonomic q0 and q1 dissimilarities were higher in P. fluviatilis compared to L. gibbosus, indicating an overall higher variation of microbiota compositions in relation to tested environmental variables in the former (Figure 4). For all models except phylogenetic q0 in L. gibbosus (the latter dominated by an effect of pH), Chl a concentration was the best predictor for change in ASV composition, contributing to 65% and 47% of the explained deviance. The spatial effect was also more important in all models for P. fluviatilis than for L. gibbosus. For both fish species, the proportion of gut-associated 16S rRNA reads corresponding to ASVs that also occurred in surrounding water was higher in lakes LGP and CSM compared to lakes with lower Chl a levels (Figure 5). For any lake, this proportion was higher in P. fluviatilis compared to L. gibbosus . The difference was maximal at LGP, suggesting a marked threshold effect in P. fluviatilis . The high abundance of Verrucomicrobiota in the gut of both species at CSM, and to a lesser extent in the gut of P. fluviatili s at LGP, illustrates this trend (Figure 1). Indeed, most of the reads corresponded to genus LD29 (supplementary figure S2), a genus also found abundant in the water at CSM. Variability of gut metabolome composition and correlation with community compositions and Chl a levels The gut metabolome composition was analyzed from the gut of all 120 fish specimens used for DNA extraction. A total of 2,061 metabolites were detected, of which 277 were successfully annotated. This approach allowed the detection and the annotation of various lipids (LipoPhosphoCholines, PhosphoCholine, Phosphatidylethanolamine, Cholic acid, palmitic acid), several nucleic acids and various small peptides, that are classically retrieved from fish gut metabolome by similar approach (Marie 2022) (Supplementary Figure S3). The PERMANOVA analysis performed on metabolite dissimilarity indices shows an effect of the lake factor on the composition of metabolites associated with Perca fluviatilis and Lepomis gibbosus . For the taxonomic dissimilarity (q0), R² values (0.149–0.180) indicate that the lake influences the presence/absence of metabolites, with a stronger effect for P. fluviatilis (R² = 0.180, F = 2.34, p = 0.0009) than for L. gibbosus (R² = 0.149, F = 1.825, p = 0.001). The lake effect is even more pronounced when considering the relative abundances of metabolites (q1), with R² reaching 0.285 and 0.251 for P. fluviatilis and L. gibbosus , respectively (Supplementary Table S5). Gut metabolome compositions of P. fluviatilis and L. gibbosus were well-differentiated in the PCoA plot (Figure 2E). In P. fluviatilis , compositions dissimilarity (q0) displayed strong structuration along the two first axes, with lakes CSM and LGP on the left and the four other lakes on the right (Supplementary Figure S4 A-B). Gut metabolome composition of L. gibbosus specimens from the different lakes were much more overlapping on the PCoA plot, with no self-evident outlier lake (Supplementary Figure S4 C-D). High correlation values were obtained when comparing the metabolites and ASVs datasets in both P. fluviatilis (0.92) and L. gibbosus (0.84), also when considering only the 277 annotated metabolites (0.89 and 0.87, respectively), supporting a strong correlation between differences observed in gut community compositions and differences in holobionts gut metabolites profile. Significant correlations were observed between variations in abundances of numerous annotated (and unannotated) metabolites and ASVs in both species, most of these being positive (Supplementary Figure S5, supplementary table S6). In order to explore dose-response of metabolites composition versus Chl a , a DROMICS analysis was conducted. Abundances of 683 metabolites from P. fluviatilis fitted one of the four types of dose-response curves used by DROMICS: namely bell, decrease, increase and U-shaped (41, 108, 476 and 58 metabolites, respectively). In L. gibbosus , abundances of 245 metabolites fitted one of the four types of dose-response curves: Bell, decrease, increase and U-shaped (35, 67, 107, 36 metabolites, respectively). P. fluviatilis thus displayed 2.78 times more dose-responsive metabolites compared to L. gibbosus . However, the cumulative distribution functions of metabolites in relation to Chl a concentrations mostly overlapped between the two species (Figure 6). Most changes occurred below 50% benchmark dose, indicating that most dose-responsive metabolites respond to intermediate Chl a levels, while far less metabolites respond to highest doses (>50% BMD). Discussion High Chla levels lead to gut microbiota dysbiosis Lake of origin and host species, the native Perca fluviatilis and the invasive Lepomis gibbosus , both affect gut community and metabolite compositions. The six lakes have similar contexts, that of shallow peri-urban artificial lakes resulting from quarrying activities, and their main use is for recreational activities. They were sampled on the same week under highly similar meteorological conditions, limiting the influence of confounding factors. Chl a concentration is commonly used as a proxy of water body eutrophication levels. While potential confounding variables cannot completely be ruled out, it remains a reasonable primary variable for our analysis. Our results confirm that Chl a is a primary driver of fish gut community and metabolome composition. For both species, gut bacterial microbiota includes representatives of genera commonly reported as fish resident gut symbionts, including Cetobacterium , Aeromonas , Romboutsia , and Clostridium sensu stricto 1 (Llewellyn et al. 2014; Tsuchiya et al. 2007). Despite that many genera are shared between P. fluviatilis and L. gibbosus , gut community compositions differ significantly, in particular at the ASVs level and in terms of their abundances. Within each species, specimens from a single lake display more similar gut communities than those from different lakes, highlighting the influence of localized environmental factors, even among lakes located in close geographical vicinity (Sullam et al. 2012, 2015). Gut community compositions from specimens sampled at CSM differed most from those found in other lakes, emphasizing the influence of the highest Chl a level. In particular, the aforementioned typical gut symbiont genera are less abundant in this lake. Decrease of an important gut symbiont in the context of high phytoplankton abundance was documented in recent lab-based studies on the medaka fish, in which relative abundances of resident Firmicutes were shown to decrease after exposure to simulated Microcystis aeruginosa blooms (Foucault et al. 2022; Gallet et al. 2023). On the other hand, the observed high abundance of Verrucomicrobiota (genus LD29) in guts of both P. fluviatilis and L. gibbosus at CSM in this study mirrors its high abundance in the water (Foucault et al. 2025). LD29 is reportedly abundant in the eutrophicated Baltic Sea and in mesotrophic to eutrophic lakes, where it has been documented to live attached to the phytoplankton, likely degrading polymers (Bergen et al. 2014). This taxon was also previously identified in the gut of fish living in the hypereutrophic Aghien lagoon (Gallet et al. 2022). The decrease of typical gut-associated genera and increase in bacteria normally found in the water are indicative of a major shift in gut microbiota composition. Indeed, the fact that ASVs identical to some occurring in water samples also occur in gut communities in both species at CSM suggests colonization of the gut by opportunistic bacteria, and a limited resistance of resident gut communities (Allison & Martiny 2008). As a result, increased homogenization of gut community compositions is observed among specimens in both species, as well as between the two species at CSM. This points to high Chl a levels as a likely external forcing, and we hypothesize that the observed changes are signs of dysbiosis induced by the consequences of high eutrophication and phytoplankton levels (Iebba et al. 2016). Further support for the dysbiosis hypothesis arises from metabolite compositions. Often used as a proxy to fingerprint functional processes (Bundy et al. 2009), the metabolome compositions of fish gut and their associated microbiota display dose-response to Chl a contents. However, the fact that most of this dose-response occurs at low-to-intermediate Chl a levels indicates that the response becomes less effective at highest concentrations, suggesting that the holobiont struggles to acclimatize and enters a stress state (Creusot et al. 2022). Despite that they are correlated, metabolome compositions are more marginally affected than prokaryotic community compositions. This is congruent with results from a previous study on the medaka fish exposed to a simulated bloom of M. aeruginosa showing that changes in microbiota composition occur faster (and possibly before) changes in holobiont’s metabolome (Foucault et al. 2022). The more limited response of the metabolome indicates limited functional consequences for holobionts. The difference between microbiota and metabolites response might be related to the important weight of host-produced metabolites in the metabolome signal, and to the fact that changes in microbiota composition do not necessarily translate directly into changes in metabolites due to extensive functional redundancy in gut communities (Louca et al. 2018). Either way, these results emphasize the importance of addressing gut microbiota composition in combination with a more functional approach such as metabolomics. Strong influence of experimentally simulated blooms of M. aeruginosa on the gut microbiota and metabolome compositions of Oryzias latipes and Danio rerio was recently reported (Gallet et al. 2023; Qian et al. 2019). Results from this study confirm that high phytoplankton abundances are also correlated with changes in gut microbiota and metabolome compositions in natural fish populations, in particular at highest levels. Besides this, a novel finding is that dysbiosis involves colonization of the gut by environmental, non-resident bacteria of which the short- and long-term effect on host health and fitness need to be further investigated. Invasive L. gibbosus displays a more resistant gut microbiota compared to autochtonous P. fluviatilis L. gibbosus is a North American species introduced in Europe in the 1880s (Pascal et al. 2003) and is considered invasive in France and other European regions. In contrast, P. fluviatilis is a native species that may be experiencing population decline, partly due to competition with invasive species including L. gibbosus (Erratt et al. 2023; Huisman et al. 2018). Both species show signs of dysbiosis in lake CSM, characterized by its very high Chl a content. Yet, apart from this lake, differences occur between the two species. The microbiota compositions in P. fluviatilis are slightly more variable from lake to lake, a higher proportion of the gut community corresponds to bacteria also present in the water, and more metabolites display dose-response to Chl a levels. P. fluviatilis thus displays more lake-specific community compositions, meanwhile L. gibbosus communities and metabolome composition are more homogeneous among lakes. The lower explanatory power of tested factors, including Chl a , in L. gibbosus community compositions, as well as the lower number of metabolites that show dose-response indicate less sensitive communities and lower influence on holobionts functions. Overall, results suggest that the gut microbiota composition and associated metabolites of invasive L. gibbosus are more stable, over a broader range of Chl a values compared to that of autochthonous P. fluviatilis , and thus likely more resistant to variations. Besides, P. fluviatilis shows signs of dysbiosis at lower Chl a levels compared to L. gibbosus , including a drop in intra-lake dissimilarity levels indicative of community homogenization due to colonization by bacteria from the water, higher proportion of these non-resident bacteria, and higher abundances of the typical eutrophic water genus LD29 in LGP. This points to the existence of tipping points beyond which gut microbiota dysbiosis occurs, and that the corresponding Chl a threshold value is lower in P. fluviatilis than in L. gibbosus . However, the fact that both species display signs of dysbiosis at the most Chl a -rich lake, CSM, indicates possible limits to acclimatization even in the invasive species. In the context of the rapidly declining biodiversity of freshwater ecosystems and increasing harmful algal blooms (Reid et al. 2019), differential tolerance to high abundances of phytoplankton could be an important yet overlooked selective advantage enhancing the success of invasive, such as the pumpkinseed, versus native species such as the perch. Invasive species are among the main factors contributing to fish diversity decline worldwide (Gallardo et al. 2016). With almost 40% of European and North American freshwater fish species at risk of extinction, increase of blooms intensity and frequency could thus contribute to further accelerate their decline. Conclusion This study confirms that high Chl a levels have a major impact on fish gut community and holobiont metabolites compositions in natural populations. Response is species-specific, with invasive L. gibbosus microbiota and functions being more stable over a broader range of conditions compared to native P. fluviatilis . In the context of increasing bloom frequencies and intensities worldwide, we hypothesize that increased resistance of gut microbiota and functions could be one underestimated asset allowing some invasive species to outcompete native ones. To test this hypothesis, long-term monitoring of fish holobionts’ response is necessary in order to evaluate the impact of increased Chl a levels on microbiota-influenced fitness relevant parameters (feeding efficiency, reproductive success…). Ethics Fish sampling was performed by Hydrosphere with all due accreditation for animal welfare. Acknowledgments We thank the COM2LIFE consortia for their participation in the fieldwork and especially Haeitz Aloui for the administrative support. We thank the leisure’s centers directors for their monthly access approval, Hydrosphere for the fishing, and the GENOTOUL sequencing facility. We thank the PtRMN (MNHN) and PARI platform (Plateforme d’Analyse Haute Résolution, IPGP, UMR 7154, CNRS, Paris, France) for nutrient measurements. Authors contribution SD, NL and BM conceived the study; SD, MQ, MG, EL, BM, PF and CD performed the sampling and DNA analyses; AN, NL, MT, MQ, BM and SD analyzed and interpreted the data; SD, NL and AN drafted the ms, all authors edited the manuscript and agreed to contents. Funding This work, as well as PF, MG and AN grants were funded by the Agence Nationale de la Recherche project COM2LIFE (ANR-20-CE32-0006), MNHN and Sorbonne Université. Competing interest statement The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data Availability: Sequence data was deposited into the Sequence Read Archive database (Bioproject PRJNA1086840); All data, code, figures are available from github https://github.com/LoiseauN/Invasive_native_microbiome.git Reference Allison, S.D. & Martiny, J.B. (2008). Resistance, resilience, and redundancy in microbial communities. Proc. Natl. Acad. Sci. , 105, 11512–11519.Bergen, B., Herlemann, D.P., Labrenz, M. & Jürgens, K. (2014). Distribution of the verrucomicrobial clade S partobacteria along a salinity gradient in the B altic Sea. Environ. Microbiol. Rep. , 6, 625–630.Bolyen, E., Rideout, J.R., Dillon, M.R., Bokulich, N.A., Abnet, C., Al-Ghalith, G.A., et al. (2018). QIIME 2: Reproducible, interactive, scalable, and extensible microbiome data science.Brooks, M.E., Kristensen, K., van Benthem, K.J., Magnusson, A., Berg, C.W., Nielsen, A., et al. (2017). glmmTMB Balances Speed and Flexibility Among Packages for Zero-inflated Generalized Linear Mixed Modeling, 9.Bundy, J.G., Davey, M.P. & Viant, M.R. (2009). Environmental metabolomics: a critical review and future perspectives. Metabolomics , 5, 3–21.Callahan, B.J., McMurdie, P.J., Rosen, M.J., Han, A.W., Johnson, A.J.A. & Holmes, S.P. (2016). DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods , 13, 581–583.Chao, A., Chiu, C.-H. & Jost, L. (2014). Unifying Species Diversity, Phylogenetic Diversity, Functional Diversity, and Related Similarity and Differentiation Measures Through Hill Numbers. Annu. Rev. Ecol. Evol. Syst. , 45, 297–324.Creusot, N., Chaumet, B., Eon, M., Mazzella, N., Moreira, A. & Morin, S. (2022). Metabolomics insight into the influence of environmental factors in responses of freshwater biofilms to the model herbicide diuron. Environ. Sci. Pollut. Res. , 29, 29332–29347.Crooks, J.A. & Rilov, G. (2009). The Establishment of Invasive Species. In: Biological Invasions in Marine Ecosystems , Ecological Studies (eds. Rilov, G. & Crooks, J.A.). Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 173–175.Danaher, C., Newbold, T., Cardille, J. & Chapman, A.S. (2022). Prioritizing conservation in sub‐Saharan African lakes based on freshwater biodiversity and algal bloom metrics. Conserv. Biol. , 36, e13914.Duperron, S., Foucault, P., Duval, C., Goto, M., Gallet, A., Colas, S., et al. (2023). Multi-omics analyses from a single sample: prior metabolite extraction does not alter the 16S rRNA-based characterization of prokaryotic community in a diversity of sample types. FEMS Microbiol. Lett. , 370, fnad125.Duperron, S., Halary, S., Gallet, A. & Marie, B. (2020). Microbiome-aware ecotoxicology of organisms: relevance, pitfalls, and challenges. Front. Public Health , 8, 407.Duperron, S., Halary, S., Habiballah, M., Gallet, A., Huet, H., Duval, C., et al. (2019). Response of Fish Gut Microbiota to Toxin-Containing Cyanobacterial Extracts: A Microcosm Study on the Medaka ( Oryzias latipes ). Environ. Sci. Technol. Lett. , 6, 341–347.Egerton, S., Culloty, S., Whooley, J., Stanton, C. & Ross, R.P. (2018). The gut microbiota of marine fish. Front. Microbiol. , 9, 873.Eichmiller, J.J., Hamilton, M.J., Staley, C., Sadowsky, M.J. & Sorensen, P.W. (2016). Environment shapes the fecal microbiome of invasive carp species. Microbiome , 4, 44.Erratt, K.J., Creed, I.F., Lobb, D.A., Smol, J.P. & Trick, C.G. (2023). Climate change amplifies the risk of potentially toxigenic cyanobacteria. Glob. Change Biol. , 29, 5240–5249.Escalas, A., Auguet, J.-C., Avouac, A., Belmaker, J., Dailianis, T., Kiflawi, M., et al. (2022). Shift and homogenization of gut microbiome during invasion in marine fishes. Anim. Microbiome , 4, 37.Evariste, L. (2019). Gut microbiota of aquatic organisms: A key endpoint for ecotoxicological studies. Environ. Pollut. Ferrier, S., Manion, G., Elith, J. & Richardson, K. (2007). Using generalized dissimilarity modelling to analyse and predict patterns of beta diversity in regional biodiversity assessment. Divers. Distrib. , 13, 252–264.Foucault, P., Gallet, A., Duval, C., Marie, B. & Duperron, S. (2022). Gut microbiota and holobiont metabolome composition of the medaka fish (Oryzias latipes) are affected by a short exposure to the cyanobacterium Microcystis aeruginosa. Aquat. Toxicol. , 253, 106329.Foucault, P., Halary, S., Duval, C., Goto, M., Marie, B., Hamlaoui, S., et al. (2025). A summer in the greater Paris: trophic status of peri-urban lakes shapes prokaryotic community structure and functional potential. Environ. Microbiome , 20, 24.Gallardo, B., Clavero, M., Sánchez, M.I. & Vilà, M. (2016). Global ecological impacts of invasive species in aquatic ecosystems. Glob. Change Biol. , 22, 151–163.Gallet, A., Halary, S., Duval, C., Huet, H., Duperron, S. & Marie, B. (2023). Disruption of fish gut microbiota composition and holobiont’s metabolome during a simulated Microcystis aeruginosa (Cyanobacteria) bloom. Microbiome , 11, 108.Gallet, A., Yao, E.K., Foucault, P., Bernard, C., Quiblier, C., Humbert, J.-F., et al. (2022). Fish gut-associated bacterial communities in a tropical lagoon (Aghien lagoon, Ivory Coast). Front. Microbiol. , 13, 963456.Gallo, B.D., Farrell, J.M. & Leydet, B. (2020). Use of next generation sequencing to compare simple habitat and species level differences in the gut microbiota of an invasive and native freshwater fish species. PeerJ , 8, e10237.Holmes, R.M., Aminot, A., Kérouel, R., Hooker, B.A. & Peterson, B.J. (1999). A simple and precise method for measuring ammonium in marine and freshwater ecosystems. Can. J. Fish. Aquat. Sci. , 56, 1801–1808.Huisman, J., Codd, G.A., Paerl, H.W., Ibelings, B.W., Verspagen, J.M.H. & Visser, P.M. (2018). Cyanobacterial blooms. Nat. Rev. Microbiol. , 16, 471–483.Iebba, V., Totino, V., Gagliardi, A., Santangelo, F., Cacciotti, F., Trancassini, M., et al. (2016). Eubiosis and dysbiosis: the two sides of the microbiota. New Microbiol. , 39, 1–12.Jakubčinová, K. (2018). Distribution patterns and potential for further spread of three invasive fish species (Neogobius melanostomus, Lepomis gibbosus and Pseudorasbora parva) in Slovakia. Aquat. Invasions , 13, 513–524.Larras, F., Billoir, E., Baillard, V., Siberchicot, A., Scholz, S., Wubet, T., et al. (2018). DRomics: a turnkey tool to support the use of the dose–response framework for omics data in ecological risk assessment. Environ. Sci. Technol. , 52, 14461–14468.Le Manach, S., Khenfech, N., Huet, H., Qiao, Q., Duval, C., Marie, A., et al. (2016). Gender-specific toxicological effects of chronic exposure to pure microcystin-LR or complex Microcystis aeruginosa extracts on adult medaka fish. Environ. Sci. Technol. , 50, 8324–8334.Le Manach, S., Sotton, B., Huet, H., Duval, C., Paris, A., Marie, A., et al. (2018). Physiological effects caused by microcystin-producing and non-microcystin producing Microcystis aeruginosa on medaka fish: A proteomic and metabolomic study on liver. Environ. Pollut. , 234, 523–537.Llewellyn, M.S., Boutin, S., Hoseinifar, S.H. & Derome, N. (2014). Teleost microbiomes: the state of the art in their characterization, manipulation and importance in aquaculture and fisheries. Front. Microbiol. , 5, 207.Lopez, B.E., Allen, J.M., Dukes, J.S., Lenoir, J., Vilà, M., Blumenthal, D.M., et al. (2022). Global environmental changes more frequently offset than intensify detrimental effects of biological invasions. Proc. Natl. Acad. Sci. , 119, e2117389119.Louca, S., Polz, M.F., Mazel, F., Albright, M.B., Huber, J.A., O’Connor, M.I., et al. (2018). Function and functional redundancy in microbial systems. Nat. Ecol. Evol. , 2, 936–943.Malbrouck, C. & Kestemont, P. (2006). Effects of microcystins on fish. Environ. Toxicol. Chem. , 25, 72–86.Marie, B. (2022). Fish metabolome from sub-urban lakes of the Paris area (France) and potential influence of noxious metabolites produced by cyanobacteria.Marie, B. & Gallet, A. (2022). Fish metabolome from sub-urban lakes of the Paris area (France) and potential influence of noxious metabolites produced by cyanobacteria. Chemosphere , 296, 134035.Muñoz‐Mas, R., Essl, F., Van Kleunen, M., Seebens, H., Dawson, W., Casal, C.M.V., et al. (2023). Two centuries of spatial and temporal dynamics of freshwater fish introductions. Glob. Ecol. Biogeogr. , 32, 1632–1644.Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O’hara, R., et al. (2013). Package ‘vegan.’ Community Ecol. Package Version , 2, 1–295.O’Neil, J.M., Davis, T.W., Burford, M.A. & Gobler, C.J. (2012). The rise of harmful cyanobacteria blooms: the potential roles of eutrophication and climate change. Harmful Algae , 14, 313–334.Paerl, H.W. & Otten, T.G. (2013). Harmful cyanobacterial blooms: causes, consequences, and controls. Microb. Ecol. , 65, 995–1010.Parada, A.E., Needham, D.M. & Fuhrman, J.A. (2016). Every base matters: assessing small subunit rRNA primers for marine microbiomes with mock communities, time series and global field samples. Environ. Microbiol. , 18, 1403–1414.Pascal, M., Lorvelec, O., Vigne, J.-D., Keith, P. & Clergeau, P. (2003). Évolution holocène de la faune de Vertébrés de France : invasions et extinctions Juillet 2003.Pavagadhi, S. & Balasubramanian, R. (2013). Toxicological evaluation of microcystins in aquatic fish species: Current knowledge and future directions. Aquat. Toxicol. , 142, 1–16.Qian, H., Zhang, M., Liu, G., Lu, T., Sun, L. & Pan, X. (2019). Effects of different concentrations of Microcystis aeruginosa on the intestinal microbiota and immunity of zebrafish (Danio rerio). Chemosphere , 214, 579–586.Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., et al. (2012). The SILVA ribosomal RNA gene database project: improved data processing and web-based tools. Nucleic Acids Res. , 41, D590–D596.Reid, A.J., Carlson, A.K., Creed, I.F., Eliason, E.J., Gell, P.A., Johnson, P.T.J., et al. (2019). Emerging threats and persistent conservation challenges for freshwater biodiversity. Biol. Rev. , 94, 849–873.Rohart, F., Gautier, B., Singh, A. & Lê Cao, K.-A. (2017). mixOmics: An R package for ‘omics feature selection and multiple data integration. PLoS Comput. Biol. , 13, e1005752.Saraf, S.R., Frenkel, A., Harke, M.J., Jankowiak, J.G., Gobler, C.J. & McElroy, A.E. (2018). Effects of Microcystis on development of early life stage Japanese medaka (Oryzias latipes): comparative toxicity of natural blooms, cultured Microcystis and microcystin-LR. Aquat. Toxicol. , 194, 18–26.Singh, A., Shannon, C.P., Gautier, B., Rohart, F., Vacher, M., Tebbutt, S.J., et al. (2019). DIABLO: an integrative approach for identifying key molecular drivers from multi-omics assays. Bioinformatics , 35, 3055–3062.Sohil, F., Sohali, M.U. & Shabbir, J. (2022). An introduction to statistical learning with applications in R: by Gareth James, Daniela Witten, Trevor Hastie, and Robert Tibshirani, New York, Springer Science and Business Media, 2013, $41.98, eISBN: 978-1-4614-7137-7. Stat. Theory Relat. Fields , 6, 87–87.Sullam, K.E., Essinger, S.D., Lozupone, C.A., O’CONNOR, M.P., Rosen, G.L., Knight, R., et al. (2012). Environmental and ecological factors that shape the gut bacterial communities of fish: a meta‐analysis. Mol. Ecol. , 21, 3363–3378.Sullam, K.E., Rubin, B.E., Dalton, C.M., Kilham, S.S., Flecker, A.S. & Russell, J.A. (2015). Divergence across diet, time and populations rules out parallel evolution in the gut microbiomes of Trinidadian guppies. ISME J. , 9, 1508–1522.Tsuchiya, C., Sakata, T. & Sugita, H. (2007). Novel ecological niche of Cetobacterium somerae, an anaerobic bacterium in the intestinal tracts of freshwater fish. Lett. Appl. Microbiol. Wang, M., Carver, J.J., Phelan, V.V., Sanchez, L.M., Garg, N., Peng, Y., et al. (2016). Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. Nat. Biotechnol. , 34, 828–837.Yéprémian, C., Catherine, A., Bernard, C., Congestri, R., Elersek, T. & Pilkaityte, R. (2016). Chlorophyll a Extraction and Determination. In: Handbook of Cyanobacterial Monitoring and Cyanotoxin Analysis (eds. Meriluoto, J., Spoof, L. & Codd, G.A.). Wiley, pp. 331–334.Zhang, L., Yang, Z., Yang, F., Wang, G., Zeng, M., Zhang, Z., et al. (2023). Gut microbiota of two invasive fishes respond differently to temperature. Front. Microbiol. , 14, 1087777.Zuur, A.F., Ieno, E.N. & Elphick, C.S. (2010). A protocol for data exploration to avoid common statistical problems: Data exploration . Methods Ecol. Evol. , 1, 3–14. Figures Figure 1 : Composition of gut-associated communities in Perca fluviatilis and Lepomis gibbosus , phylum level, based on median values from 10 replicate specimens. Phyla that were below 1% of reads are grouped under “other”. Lakes are ranked from left to right with increasing Chl a content, and abbreviated as follows : Cergy Large (CERL), Cergy Small (CERS), Créteil (CRE),Verneuil-sur-Seine (VSS), La Grande-Paroisse (LGP), Champs-sur-Marne (CSS). Figure 2: Principal Coordinates Analysis (PCoA) of prokaryotic gut communities of fish based on the framework of Hill numbers. A: Taxonomic dissimilarity (q0), B: Taxonomic dissimilarity (q1), C: Phylogenetic dissimilarity (q0), D: Phylogenetic dissimilarity (q1). E: Metabolite dissimilarity based on the analysis of 2,061 metabolites (annotated and unannotated) was performed. q0. Solid points represent Perca fluviatilis , and hollow points represent Lepomis gibbosus . Figure 3: Boxplot representing the values of intra-specific q1 taxonomic dissimilarity within lakes for P. fluviatilis ( A ) and L. gibbosus ( B ); and inter-specific comparison ( C ). Lakes are ranked according to increasing Chl a level. Figure 4 : Comparison of coefficient percentages, representing relative importance, by predictor of each GDM model (taxo_q0, taxo_q1, phylo_q0, phylo_q1) based on beta diversity results for Perca fluviatilis (A) and Lepomis gibbosus (B). Numbers above each stacked bar represent the percentage of total variance explained by each model. Figure 5: Total percentage of fish gut microbiota reads that correspond to ASVs also present in the water for the two species along the Chl a gradient. The red curve represents P. fluviatilis and the yellow curve represents L. gibbosus . Figure 6: Curves displaying the Empirical Cumulative Distribution Function (ECDF) of metabolites showing a dose-response versus benchmark dose (BMD) representing the normalized Chl a concentration in P. fluviatilis (red, 683 metabolites) and L. gibbosus (yellow, 245 metabolites). Information & Authors Information Version history V1 Version 1 29 September 2025 Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords aquatic microbial ecology eutrophication fish microbiome lake resilience Authors Affiliations Alice Navarro Montpellier University View all articles by this author Nicolas Loiseau MARBEC View all articles by this author Marc Troussellier Montpellier University View all articles by this author Pierre Foucault Museum National d'Histoire Naturelle View all articles by this author Charlotte Duval Museum National d'Histoire Naturelle View all articles by this author Julie Leloup 0000-0002-2777-284X Sorbonne University View all articles by this author Manon Quiquand Museum National d'Histoire Naturelle View all articles by this author Emilie Lance Reims Champagne-Ardenne University Faculty of Natural Sciences View all articles by this author Midoli Goto Museum National d'Histoire Naturelle View all articles by this author Benjamin Marie Museum National d'Histoire Naturelle View all articles by this author Sebastien Duperron 0000-0002-6422-6821 [email protected] Museum National d'Histoire Naturelle View all articles by this author Metrics & Citations Metrics Article Usage 255 views 143 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Alice Navarro, Nicolas Loiseau, Marc Troussellier, et al. Invasive and native lake fish species display different responses of their gut microbiota and metabolite compositions along a eutrophication gradient. Authorea . 29 September 2025. DOI: https://doi.org/10.22541/au.175916041.10007367/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175916041.10007367/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a003524cee94e2c5',t:'MTc3OTUzMTU1Nw=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();