The ISME Journal

Understanding how intra- and interspecific differentiation arises in natural microbial populations is central to explaining the processes that drive bacterial evolution. Motivated by the co-occurrence of several genospecies closely related to Shewanella baltica in Baltic Sea sediments, we investigated the genomic structure of this species complex across fine spatial scales. We analyzed 112 genome sequences from strains collected across multiple sediment cores and depths (0-6 cm) at Vaxon (Stockholm archipelago, Sweden), including sympatric isolates from this site as well as earlier isolates and allopatric strains from other locations in the Stockholm region obtained from both sediments and the water column. Using a reverse-ecology population genomics approach, we found that these strains form a species complex that resolves into three cohesive evolutionary lineages (G1, G2, and G3). Each lineage is characterized by extensive gene turnover, driven largely by horizontal gene transfer (HGT), and displays distinct genomic signatures of metabolic specialization. While G1 consists predominantly of a single species (S. baltica), G2 and G3 comprise a diverse set of divergent genospecies, many of which are repeatedly recovered from sediment samples. Patterns of homologous recombination indicate that speciation within G2 and G3 is primarily recombination-driven ( sexual), and that both groups derive from a common ancestor. Together, these results capture a snapshot of early-stage speciation within a shared ecosystem and provide insight into the mechanisms that diversify sympatric, recombining bacterial populations, with a sediment-associated lifestyle likely promoting this process.

1.Structured Abstract1.1 AbstractSoybean red crown rot, caused by the soil-borne fungus Calonectria ilicicola, causes substantial yield losses, but the response of the root-associated bacterial microbiome remains poorly understood. Here, we combined 16S rRNA gene sequencing, shotgun metagenomics, and single-cell genomics to characterize bacterial communities in soybean root-associated soils. 16S rRNA gene sequencing showed that diseased plants had rhizosphere and, more strikingly, rhizoplane microbiomes distinct from those of healthy plants, often with increased Enterobacterales. Shotgun metagenomics further revealed enrichment of genes associated with antibiotic resistance, particularly cationic antimicrobial peptide resistance, in diseased rhizoplane samples. Single-cell genomics recovered seven nonredundant Enterobacterales genomes and showed that plant pathogenicity-related genes were broadly distributed across these lineages. In contrast, dlt genes, which are associated with cationic antimicrobial peptide resistance, were detected only in the Enterobacterales lineages enriched in diseased rhizoplane soils. These results support a model in which soybean red crown rot is accompanied by microbiome restructuring and opportunistic enrichment of specific Enterobacterales lineages carrying putative cationic antimicrobial peptide resistance genes. More broadly, this study highlights the value of strain-resolved single-cell genomics for linking disease-associated community shifts to specific bacterial traits. 1.2 ImportanceUnderstanding crop disease requires resolving not only the primary pathogen but also the root-associated bacteria that respond to infection. Here, we used 16S rRNA gene sequencing, shotgun metagenomics, and single-cell genomics to examine the soybean rhizoplane microbiome under red crown rot. Diseased plants showed reproducible shifts in bacterial composition, including frequent enrichment of Enterobacterales and antimicrobial resistance-related functions. Strain-resolved genomes further revealed that the Enterobacterales lineages enriched in diseased rhizoplane soils specifically carried putative dlt-mediated resistance to cationic antimicrobial peptides, whereas general pathogenicity-related genes were broadly shared. These findings suggest that host defense-associated selection, rather than pathogenicity genes alone, may help shape disease-associated root microbiomes. This study demonstrates how single-cell genomics can uncover strain-level traits hidden within bulk community data and thereby clarify plant-pathogen-microbiome interactions.

Antibiotic resistance genes (ARGs) are ubiquitous in marine environments, yet whether their distribution primarily reflects anthropogenic pollution or intrinsic ecological functions remains unresolved. We used genome-resolved metagenomics to characterize resistomes in 371 genomic operational taxonomic units (gOTUs) across a gradient of human impact: the heavily impacted Baltic Sea, the moderately impacted North Sea, and the minimally impacted West Greenland shelf. ARG density was distinctly elevated in the Baltic Sea (3.20 ARGs Mbp-1) relative to the North Sea (1.90) and West Greenland (1.67), which did not differ significantly from each other, suggesting a relatively uniform oceanic baseline. Variance partitioning revealed that taxonomic identity explained 20.1% of ARG density variation, with environment contributing 11.4%; critically, Baltic gOTUs carried 35.1% more ARGs than predicted from taxonomy alone, indicating environment-driven enrichment beyond baseline taxonomic carriage. Lifestyle-dependent ARG partitioning between particle-attached and free-living prokaryotes emerged only under anthropogenic pressure: free-living bacteria were enriched in multiple resistance classes in the Baltic Sea but showed no differentiation in West Greenland. Only 0.85% of detected ARGs showed [≥]70% amino acid identity to clinically characterized sequences in the CARD database, showing that marine ARGs are highly divergent from clinical resistance determinants. Virulence factor annotations were widespread but weakly coupled with ARG abundance, suggesting independent ecological selection. Our results suggest that marine resistomes integrate an intrinsic baseline of ecological functions with selective enrichment of specific resistance mechanisms under anthropogenic pressure, and that genome-resolved approaches are able to quantify the relative contributions of each.

Microbial community composition is determined by numerous factors, including environmental filtering, biotic interactions, and stochasticity. Disentangling the relative contribution of these distinct processes is a fundamental problem for microbial ecology. Towards addressing it, here we used controlled experiments on replicate freshwater communities that vary individual factors, and used contig-based metagenomics methods to capture functional composition independently from uneven genome recovery across taxa. After repeated sub-culturing in minimal media lacking organic carbon but containing nitrate and vitamins, communities with distinct initial compositions consistently converged toward similar taxonomic structures and metabolic functions. These included oxygenic photosynthesis and nitrate metabolism, consistent with the imposed growth regime, and also showed reproducible enrichment of anoxygenic photosynthesis, vitamin biosynthesis and degradation pathways. These patterns indicate strong environmental filtering during assembly while also revealing a consistent role for emergent environments arising from microbial activity and metabolic interactions. To more directly test specific environmental drivers, we initiated replicate cultures from a single community and propagated them in the original medium and in variants lacking vitamins or both vitamins and nitrates. Only nitrate removal produced a distinct and statistically significant shift in both composition and function, with nitrate-free communities enriched for cyanobacterial nitrogen fixation and supporting metabolic functions in heterotrophs. Together, these results support a hierarchy of environmental filters determining community outcomes and provide a quantitative framework for predicting and steering community function through rational environmental design.

Abstract-shortLand-locked meromictic lakes are characterized by long-term stratification and steep redox gradients that sustain vertically structured microbial communities and tightly coupled biogeochemical processes. Because complete overturns of the lake are rare, the dynamics of microbial reassembly after redox gradients collapse and subsequently recover remain poorly resolved. We investigated a mixing-restratification transition in Lake Shira (Siberia) with depth-stratified sampling of oxic, chemocline, anoxic, and water-sediment interface layers across four stages: intermittent holomictic (IH), complete holomictic (CH), developing meromictic (DM), and stable meromictic (M). 16S rRNA gene amplicons showed that CH homogenized the water column, dominated by Gamma- and Alphaproteobacteria, Campylobacteria, and Cyanobacteriia. As stratification re-formed (DM-M), communities became strongly depth-partitioned, with Desulfobacterota and other anaerobes re-established in sulfidic deep waters and assemblages concentrating near the redox transition. Nanopore metagenomics reconstructed 401 MAGs, revealing stage- and depth-specific functional repertoires consistent with redox zonation. Core MAGs, including Yoonia spp., persisted across phases, suggesting functional continuity underpinning rapid ecosystem recovery. These data provide a system-wide view of biogeochemical reassembly during collapse and restoration of stratification meromictic lake.

1Synthetic microbial communities (SynComs) could help plants withstand biotic stress and reduce the need for pesticides. With this in mind, we created two SynComs, comprising bacterial strains isolated from the rhizospheres of barley and wheat. We then studied their potential to trigger induced systemic resistance against the barley pathogen Blumeria graminis f. sp. hordei (Bgh). To investigate the plant-microbial interactions from the perspective of both plants and microbes, we performed DAF staining to quantify Bgh propagation in plant leaves, analysed leaf transcriptomes and conducted rhizosphere 16S rRNA gene metabarcoding and rhizosphere metatranscriptome analysis. Our results demonstrate that the SynComs elicit defence responses in barley against Bgh in a manner similar to that of the positive control strain Pseudomonas simiae WCS417r. The SynComs act without triggering a strong gene response prior to inoculation with the plant pathogen or affecting plant-associated prokaryote communities; they only mildly influence bacterial gene expression in the rhizosphere. Instead, they act as priming agents, preparing the plant for further pathogen attack. These findings suggest that protective SynComs can be applied in the field without causing signficant disruption to native microbial communities.

Eukaryotes originated from the symbiosis of an Asgard archaeon, the alphaproteobacterial ancestor of mitochondria, and possibly additional bacterial contributions. This transition occurred in redox-transition environments such as microbial mats or shallow sediments [~]2 billion years ago, when atmospheric oxygen was far lower than today. We investigated Asgard-enriched microbial mats from the low-oxygen, sulfidic Catherine volcano lake (Afar region, Ethiopia), mimicking early Proterozoic conditions. 16S rRNA gene metabarcoding, metagenomics, and metagenome-assembled genome analyses across redox-stratified layers of in situ and mesocosm-maintained mats revealed that Asgardarchaeota thrived in the sulfate-reduction zone, mainly co-occurring with Desulfurobacterota-Myxococcota, among others. Lokiarchaeia and Thorarchaeia preferred anoxic layers. Within Heimdallarchaeia, Heimdallarchaeales were enriched in upper layers, correlating with oxygen-tolerant hydrogenase and sulfate-reduction genes, and Hodarchaeales, in anoxic layers, correlating with methanogenesis. Although reactive-oxygen-species defense mechanisms were widespread, Asgardarchaeota lacked aerobic respiration. These results support the idea that Asgard archaea engaged primarily in syntrophic interactions with sulfate-reducers under early-Earth-like conditions.

Even though complex microbial communities are ubiquitous and provide essential services for natural and human-associated ecosystems, our knowledge about their assembly and dynamics is incomplete. There is an ongoing debate whether the behavior of complex communities can be predicted from the outcome of pairwise competition of species, and whether communities reach alternative stable states depending on the level and complexity of resource provided for growth. To estimate the effect of two resource gradients, total carbon availability and resource complexity, on the compositional dynamics of a complex microbial community, we conducted a 16-day serial passage experiment, transferring a 16-species synthetic community in 96 different resource environments. We observed that although both resource dimensions influenced community composition, total carbon exerted a considerably larger effect. Additionally, we saw the emergence of a tristable pattern along the total carbon gradient, a feature not observed for the resource complexity gradient. Using monoculture assays, we identified lag phase duration as the dominant predictor of competitive success at carbon extremes, with maximum growth rate increasing in importance as lag times converged. Total carbon availability thus structured community state transitions and regulated which growth trait governed competitive sorting. These results suggest the importance of total carbon level over resource complexity and identifying dominant species for the quest to successfully manage, maintain and manipulate complex microbial communities.

Many of the important biogeochemical implications of marine viruses focus on the activity of lytic viruses. It is generally thought that lytic viral activity contributes to maintaining bioavailable nutrient pools and controlling host community composition through specific infection and lysis processes. Among lytic viruses, genome replication is under stringent selection pressure as advantages in replication speed and fidelity can influence viral fitness. Deoxyribonucleotides (dNTPs) are the precursor substrate for DNA synthesis and ribonucleotide reductase (RNR) is the only known enzyme capable of reducing ribonucleotides into dNTPs. Thus, for a virus, encoding an RNR gene provides control over dNTP production. Overall, distributional patterns of the RNR-encoding virioplankton community mirrored that seen for total virioplankton throughout the global ocean. A majority of RNR-encoding virioplankton ([~]66%) carried the Class II enzyme and most of these carried the monomeric (NrdJm) gene. This was significant as NrdJm utilizes triphosphate ribonucleotides as opposed to the diphosphate ribonucleotides used by Class I and Class II dimeric RNRs. The distribution and frequency of Class I RNRs followed the concentration of required iron and manganese co-factors. Oceanic virioplankton encompassed a high diversity of Class I and II RNRs with the discovery of new phylogenetic clades not previously observed among viruses or within cellular life. Representing [~]10% of all virioplankton populations, RNR-encoding viral populations demonstrated surprising fidelity with the major biogeographical features defining oceanic ecosystems.

Marine bivalve mortality events cause substantial economic losses in aquaculture and threaten global food security. While pathogenic Vibrio species are frequently implicated, growing evidence suggests that loss of beneficial microbes can increase host susceptibility to disease. We previously observed that Vibrio mediterranei was consistently isolated from healthy oysters but systematically disappeared prior to mortality events, coinciding with proliferation of pathogenic Vibrio species. Here, we test whether this pattern reflects a protective functional role. Pre-colonization with V. mediterranei strain Vm02 increased Crassostrea virginica oyster larval survival from 10-19% (pathogen-only controls) to 94-97% when challenged with V. harveyi or V. coralliilyticus, representing near-complete protection from pathogen-induced mortality. Protection was maintained at both ambient (28{degrees}C) and thermal stress (32{degrees}C) temperatures where pathogen virulence is enhanced, rapid (effective from co-inoculation), and durable (maintained for >96 hours). Fluorescence microscopy confirmed stable colonization of larval digestive tissues by fluorescently-tagged Vm02. Larval colonization by eleven V. mediterranei strains revealed three distinct phenotypes - protective, pathogenic, and intermediate - corresponding to monophyletic clades with 97.1-97.8% average nucleotide identity between protective and pathogenic lineages. Pangenome analysis identified 230 protective-specific versus 80 pathogenic-specific orthogroups. Protective strains encode unique regulatory systems, stress tolerance mechanisms, and metabolic versatility while lacking Type I and Type VI secretion system variants associated with pathogenicity. Together, these findings demonstrate that beneficial versus pathogenic phenotypes are phylogenetically constrained within distinct V. mediterranei lineages. This supports reports that V. mediterranei acts as both a pathogen and potential symbiont in marine hosts and reveals a clade that provides robust protection against oyster pathogens. ImportanceAquaculture disease management has traditionally emphasized either prophylactic treatment using antibiotics to avoid disease and dysbiosis or has focused entirely on pathogen detection. Both of these approaches have overlooked the potential contributions of beneficial microbes to host defense and grow-out performance. Developing beneficial probiotic tools for disease prevention represents an emerging opportunity for sustainable aquaculture management. This study demonstrates that specific lineages of Vibrio mediterranei function as protective symbionts capable of rescuing oyster larvae from near-complete pathogen-induced mortality. By integrating field observations of microbial succession during mortality events with experimental validation and comparative genomics, we show that protective versus pathogenic phenotypes are phylogenetically constrained within V. mediterranei clades separated by 97.1-97.8% average nucleotide identity. This resolution of strain-level functional variation provides fundamental insights into how host-microbe mutualisms evolve within species complexes that also harbor pathogens. The unique genomic markers identified here enable reliable screening for protective symbionts, while the temperature-stable and durable protection demonstrated in this study highlights the potential for biological control strategies in shellfish hatcheries increasingly affected by warming oceans and Vibrio-driven mortality events.

Members of the UBA868 group within the order Arenicellales are globally distributed marine Gammaproteobacteria predicted to participate in sulfur and carbon cycling, yet their physiology and ecological roles remain unknown due to the absence of cultured representatives. Here, we report the isolation and characterization of the first heterotrophic representative of the previously uncultured UBA868 group. Using dilution-to-extinction cultivation, we obtained four isolates from the Yellow Sea whose high-quality genomes represent a single UBA868 species. One strain, IMCC57338, maintained in axenic culture, exhibited small coccoid morphology and slow growth (doubling time [~]2.9 days), consistent with an oligotrophic lifestyle. Genome analysis revealed a predominantly aerobic chemoorganoheterotrophic lifestyle with a streamlined central carbon metabolism, including a complete glyoxylate shunt and limited carbohydrate utilization capacity, suggesting adaptation to low-nutrient conditions. The genome also encodes pathways for methylated amine oxidation coupled to formaldehyde assimilation via the serine cycle, indicating a capacity for methylotrophy. Genes encoding sulfur oxidation (Sox) and reverse dissimilatory sulfite reductase (rDsr) pathways further suggest a capacity for sulfur-based chemolithoheterotrophy. Global metagenomic and metatranscriptomic read recruitment showed that the species represented by IMCC57338 is widely distributed across ocean basins and pelagic depth layers, with higher abundance and transcriptional activity in mesopelagic waters. Our findings provide the first physiological and genomic insights into the UBA868 group and suggest that members of this lineage contribute to the cycling of organic carbon, C1 compounds, and sulfur in marine ecosystems.

Succession is an ecosystem building process in which a habitat and its community interact predictably by increasing diversity, habitat engineering, and ultimately reaching a climax community, where other ecological processes influence its dynamic. Key to succession is the establishment of primary producing habitat forming species, which drives niche differentiation leading to increasing diversity. Here, we use the primary colonizing and habitat forming seagrass, Halophila ovalis, to demonstrate that it drives bacterial succession in a meadow ecosystem, and its microbiome, both rhizoplane and phylloplane, are under host selection. Many of the characteristics attributed to plants for habitat modification are microbial processes such as nitrogen fixation and sulfide detoxification and succession is often extrapolated to such processes. To determine if succession (increasing diversity) or selection (reducing diversity) drives changes in diversity (16S rRNA gene) or habitat modifying processes (nifH, soxB, aprA, dsrA), molecular analysis was performed along chronosequences (as a proxy for succession) of seagrass patches. Bacterial communities were sampled within the meadow ecosystem and the microbiomes of H. ovalis (both rhizoplane and phylloplane). Genes involved in biogeochemical cycling are differentially impacted within the microbiome and meadow sediments, with only nifH under succession. All genes from all niches sampled for community analysis are under directional community trajectories, despite being subjected to distinct ecological processes, signifying that many ecological processes, including succession and host association, drive community assemblage.

Bacterivorous protists are central to aquatic food webs, mediating the transfer of carbon and nutrients to higher trophic levels through the microbial loop. In natural communities, a major challenge remains in linking protist grazing activity to environmental sequences and identifying which taxa are actively feeding at the community level. Here, we present the first application of quantitative stable isotope probing (qSIP) in a grazing experiment. By combining qSIP with 18S rRNA gene amplicon sequencing, we linked prey assimilation to the identity of active protist predators at the operational taxonomic unit (OTU) level. In a replicated 36-h bottle-experiment, live 13C, 15N-labeled Limnohabitans planktonicus cells were added to natural samples from a lake pelagic site and its main inlet stream. Although hydrologically connected, protist richness was higher in the inlet than in the lake, yet a similar number of taxa incorporated prey biomass, comprising 108 OTUs in the inlet and 107 OTUs in the lake, including both rare and abundant taxa. Of these, 26 OTUs were labeled at both sites. The most strongly labeled protist in the inlet was a putative phago-mixotrophic prasinophyte, whereas in the lake it was an uncultured chrysophyte. Across sites, prey incorporation occurred in a broad range of taxa, including heterotrophs (e.g., choanoflagellates, cercozoans, ciliates, centrohelids), putative mixotrophs (e.g., cryptophytes, chrysophytes, dictyochophytes), parasitic protists and fungi. These results demonstrate the potential of qSIP to resolve trophic interactions at fine taxonomic resolution in natural communities and highlight new opportunities to study complex microbial food webs across environmental systems.

Predicting microbiome function remains challenging as microbial interactions scale from pairwise encounters to emergent community properties. This is particularly true of disease protective microbial consortia, where pathogen invasion has typically been studied either in terms of single biocontrol agents or in terms of microbiome diversity at the full community level, but rarely in between. Focusing on a 16-member synthetic tomato phyllosphere bacterial community, we combined reciprocal spent-media growth assays of over 600 pairwise and community-level combinations with comparative genomics to dissect the ecological and metabolic drivers of community interactions. Across the interaction network, negative interactions dominated, with community-derived spent media consistently exerting stronger inhibitory effects on bacterial growth across the community than any single-species filtrate. While two isolates (Exiguobacterium sibiricum and Bacillus thuringiensis) exhibited strong inhibitory effects in monoculture assays, community spent media analyses revealed that no single strain was responsible for the pathogen-suppressive phenotype observed in community, indicating that protection against Pseudomonas syringae is an emergent property of the particular community composition. Furthermore, using correlations and cross-validated multivariate models, inhibition strengths were poorly predicted by either genomic annotations or phenotypic strategies. Instead, community context strongly constrained environmental modification and buffered strain-specific effects observed in isolation. Together, these results demonstrate that microbial community function cannot easily be inferred from pairwise interactions or individual strain properties alone, and that both direct and indirect interactions shape phyllosphere community structure and function, with emergent properties such as pathogen suppression arising from collective properties rather than the presence/absence or dominance of individual keystone taxa.

Nitrogen fixing legume nodules are typically viewed as the product of a bilateral mutualism between host plants and nitrogen-fixing rhizobia, yet nodules also harbor diverse non-rhizobial endophytes whose functional significance remains poorly understood, especially in wild legumes and uncultivated soil. Here, using the wild Mediterranean shrub Calicotome villosa, we performed a soil inoculation experiment to test whether plant performance is linked to the functional composition of the nodule microbiome. Soil inocula from different natural sites produced strong differences in nodulation success, plant biomass, leaf nitrogen concentration, nitrogen fixation rate, and nodule allocation under otherwise uniform conditions. Although Bradyrhizobium dominated all nodules, species composition varied among inoculation sources, and non-rhizobial endophytes reached substantial abundance in some treatments. Functional profiles of the nodule microbiome were significantly associated with plant phenotype, with the strongest coupling observed for traits related to nodule investment. Targeted and genome-wide analyses identified trait-associated genes in both symbionts and endophytes, including genes involved in nitrogen cycling, ammonium transport, denitrification, pyrimidine degradation, sulfur assimilation, and type VI secretion systems. Several of these functions were not part of the canonical symbiosis machinery, yet were strongly associated with plant nitrogen status, biomass accumulation, or nodule mass fraction. Together, our results show that legume performance is better predicted by the collective functional composition of the nodule microbiome than by the primary symbiont alone. These findings support a broader view of nodules as multipartite microbial communities.

Plant leaves harbor diverse microbial communities influenced by environmental inputs and host traits, yet it remains unclear whether leaves act as passive substrates or active ecological filters that reorganize microbial functional capacity. Phylloplane pH regulation is one hostplant trait that has been traditionally underexplored. We used metatranscriptomics to examine microbial gene expression on the phylloplane and within whole leaves of five plant species spanning the extremes of baseline phylloplane pH, including hyperalkalinizing Gossypium species, weakly buffering Beta vulgaris, and hyperacidifying Nepenthes species. Young leaves were inoculated with a common soil-derived microbial community to quantify host-associated restructuring of taxonomic and functional profiles, and short-term pH perturbations were applied to test the effect of transient abiotic stress. Across both phylloplane and whole-leaf datasets, host species identity was the primary axis structuring microbial taxonomic composition and expressed functional repertoires. Leaf-associated communities diverged from the source inoculum, but retained a substantial shared functional backbone enriched for central biosynthetic and core metabolic pathways. Host-associated differentiation reflected selective retention and redistribution of reactions within this shared environmental pool rather than acquisition of novel metabolic capacity. Enriched pathway subsets were metabolically coherent and taxonomically distributed across multiple bacterial orders, consistent with functional redundancy and trait-based assembly. Among hosts, Gossypium exhibited the strongest restructuring relative to inoculum, suggesting comparatively stronger host-associated filtering. In contrast, short-term pH manipulation did not induce consistent community-wide functional reorganization. Microbial physiological responses to the phylloplane environment and external pH were observed at the organismal level. Together, these results position leaves as active ecological filters that reorganize microbial functional landscapes through host-specific trait regimes. This work begins to implicate some role of phylloplane pH regulation in microbial assembly and function.

Seagrasses are the only plants that flower and produce seeds in the marine environment, where they form vast meadows that fulfill important ecosystem functions. Yet, seagrass cover has declined in many coastal areas around the world and re-colonization is slow. Despite clonal growth, seed recruitment is essential for seagrass dispersal and regional genetic diversity. While the seed microbiome of several terrestrial plants has been shown to influence germination and seedling survival, the role of the seagrass seed microbiome is still unclear. We investigated the microbiome of eelgrass (Zostera marina) seeds, leaves and roots along the natural salinity gradient of the German North and Baltic Sea coasts. Despite this strong variability, Z. marina seeds harbor distinct prokaryotic and eukaryotic microbial communities compared to those on leaves and roots. Predicted microbial functions suggest roles in nutrient cycling and germination, which may be critical for recovery and restoration of seagrass ecosystems. Scientific significance statementSeagrass meadows are important but declining coastal ecosystems. Around the world, efforts are being made to restore seagrass meadows in order to halt the loss of biodiversity and maintain ecosystem function. With an evolutionary history on land, seagrasses have introduced unique features to the marine environment, such as seeds. In terrestrial plants, the seed microbiome has been shown to be important for seed germination and seedling health. Here we show that the northern hemisphere seagrass Zostera marina has a distinct seed microbiome, containing core bacterial and eukaryotic taxa and predicted functions that may play a role in seed germination and vertical microbiome transmission. Our results lay the foundation for future seagrass restoration efforts that seek to manipulate seed microbiomes to improve seedling survival. Data availability statementRaw DNA sequence data has been deposited in the European Nucleotide Archive (ENA) at EMBL-EBI under accession number PRJEB72211. Processed data and associated metadata will be made available on PANGEA via the German Federation for Biological data (GFBio.org).

Using shotgun metagenomics of largetooth sawfish (Pristis pristis) skin and paired pool water samples, we provide the first taxonomic and functional characterization of a sawfish skin microbiome and assess its ecological distinctiveness, assembly, and resilience relative to the surrounding water column. The skin microbiome was strongly host-associated and compositionally distinct, with lower taxonomic richness than the water column but dominated by Bacillota, particularly spore-forming Bacilli. Functional profiling revealed enrichment of genes associated with sporulation, dormancy, anaerobic metabolism, and peptide transport, consistent with adaptation to low-oxygen, low-flow pool conditions. In contrast, water-column microbiomes were more taxonomically diverse and enriched in phototrophic and polysaccharide-utilization pathways. Despite reduced taxonomic diversity, sawfish skin communities exhibited higher functional redundancy, with gene functions accumulating more rapidly per taxon. This pattern supports a host-filtered, lottery-like assembly process that produces taxonomically variable yet functionally conserved communities. The enrichment of dormancy and anaerobic pathways suggests the skin microbiome persists through periods of host quiescence and environmental stress while maintaining metabolic potential. Together, these results demonstrate that the sawfish skin supports a resilient, functionally robust microbiome distinct from ambient aquatic communities, highlighting the potential for integrating microbial functional data into conservation strategies for Critically Endangered vertebrates.

Cross-feeding interactions are pervasive in microbial communities and profoundly shape community structure, stability, and function. While previous studies have explored how cross-feeding affects evolvability, this work has predominantly focused on bidirectional mutualistic interactions in engineered auxotrophic systems where both partners reciprocally exchange essential metabolites. However, most metabolic interactions in natural microbial communities are unidirectional, with organisms feeding on the metabolic waste products of other species. Our study addresses this gap by examining how a unidirectional cross-feeding interaction affects the evolutionary dynamics of both the producer (Acinetobacter johnsonii) and consumer (Pseudomonas putida) over 800 generations of experimental evolution. We found that co-culture constrained adaptive evolution in both species. Co-cultures exhibited lower {pi}N/{pi}S ratios (0.75 for P. putida; 1.04 for A. johnsonii) than monocultures (1.44 and 2.02, respectively) indicating stronger purifying selection against nonsynonymous mutations in the community context. Lineage tracking through whole genome sequencing of populations and clones revealed greater lineage diversity and complexity in monocultures, with more mutations showing significant parallelism across replicate populations. Additionally, P. putida evolved increased dependence on its partner; co-culture-evolved P. putida grew significantly worse than its ancestor when A. johnsonii was removed. These findings demonstrate that ecological interactions fundamentally reshape fitness landscapes and constrain adaptive evolution even when fitness benefits are unidirectional, with implications for understanding microbial community stability and predicting evolutionary dynamics in complex communities.

Viruses regulate microbial mortality and biogeochemical cycling in marine sediments; however, the ecological drivers of sediment viral communities remain unclear. Infauna, including sediment-dwelling meiofauna and macrofauna, are major ecosystem engineers that reshape sediment structures and microbial processes, but their influence on viruses is unknown. We combined infaunal gradient incubations with metagenomic and metatranscriptomic analyses to assess viral DNA and RNA responses. DNA viruses showed increased abundance (3-fold), diversity, richness, and transcriptional activity under higher infauna abundance conditions, whereas RNA viruses remained unaffected, revealing striking selectivity. This selectivity reflects an infauna-dependent component mediated by bacterial activity that cannot be explained by host abundance alone. Infection profiling revealed increased transcription of viral replication and structural genes, and lytic viruses under high infauna conditions. These findings establish infauna as a previously overlooked regulator of DNA virus dynamics, integrating viral ecology into faunal-microbial frameworks in benthic ecosystems and suggesting potential influences on geochemical cycles. TeaserInfauna selectively shape viral communities in marine sediments, revealing an overlooked effect on DNA viruses.