From assembly inference to co-occurrence organization: conservation contexts are associated with gut bacterial...

Conservation programs increasingly rely on in-situ and ex-situ interventions to prevent the loss of endangered species; however, it remains unclear whether such interventions preserve host-associated microbial organization. Understanding how conservation contexts influence gut microbiome structure and function is therefore essential for evaluating their ecological effectiveness.

We analyzed gut bacterial communities of the endangered high-altitude fish Gymnocypris przewalskii across wild, in-situ conservation, and ex-situ conservation contexts using 16S rRNA gene sequencing. Analyses included host morphometric covariate assessment, alpha- and beta-diversity statistics, co-occurrence network analysis, PICRUSt2-based functional prediction, and null-model-based community assembly inference.

Gut bacterial communities were consistently dominated by Proteobacteria, with Firmicutes, Bacteroidetes, Actinobacteria, and Chloroflexi as secondary taxa. Alpha diversity showed no significant differences among groups, whereas beta diversity revealed modest but significant compositional shifts associated with conservation context, independent of multivariate dispersion and host morphometric traits. Functional predictions indicated shifts in pathways related to carbohydrate, energy, lipid, amino acid, terpenoid, polyketide metabolism, secondary metabolite biosynthesis, and xenobiotic degradation. Co-occurrence networks exhibited context-dependent restructuring: wild communities showed highest connectivity and positive interactions, in-situ networks were most modular and sparsest, and ex-situ communities showed the highest proportion of negative interactions. Null-model analyses indicated dominance of stochastic processes across all groups, with highest drift and stochasticity in in-situ populations, while ex-situ communities showed stronger deterministic influence and homogeneous selection.

These results demonstrate that conservation contexts are associated not only with taxonomic turnover but also with functional potential shifts, altered co-occurrence topology, and changes in community assembly processes. Within the constraints of an observational and site-confounded design, our findings suggest that microbiome-informed conservation assessment should move beyond diversity metrics and incorporate compositional, functional, network, and assembly-based perspectives when evaluating conservation outcomes in endangered aquatic hosts.