Farm microbes show surprising stability without keystone species
A new study of aquaponic farm systems—which integrate fish and vegetable production—found that microbial communities remain resilient even when individual dominant species disappear. The finding challenges conventional wisdom and could simplify farm management by reducing the need to monitor specific microbial populations.
Originaltitel: Multi-method keystone screening reveals distributed ecological resilience in aquaponic microbiomes
Introduction: Aquaponic systems couple aquaculture and crop production with microbial communities underpinning both the waste removal needs of the fish and nutrient supply needs of the plants. Most ecological environments exhibit keystone species playing an outsized role in community structure. Understanding their population dynamics can reveal important insights into environmental stability and community robustness. Whether individual keystone taxa exist in aquaponics in a similar form as observed in natural aquatic ecosystems is unknown, yet the answer would directly guide microbiome monitoring and day-to-day farm management. We applied a multi-method consensus framework for keystone identification to 16S rRNA gene amplicon data from a research-scale aquaponic facility operating both coupled and decoupled hydraulic configurations. Results: 88 amplicon libraries from three compartments (clarifier, fish tank, grow bed) across six sampling dates were processed with DADA2, yielding 825 ASVs after prevalence filtering. Three independent keystone screens were applied: leave-one-out PERMANOVA (67 taxa with FDR < 0.05 and R² > 0.02), co-occurrence network hub analysis (four hubs), and LASSO regression on temporal turnover (no individual predictors retained). No taxon was supported by more than one method. Cetobacterium sp. (Fusobacteriota) ranked highest in the presence-impact screen (R² up to 0.11) and was strongly enriched in fish tank samples, consistent with its gut-associated ecology; network hubs were environmental biofilm-associated Proteobacteria and Planctomycetota. Conclusions: The absence of cross-method convergence is consistent with a distributed ecological architecture in which community function is spread across habitats and functional guilds rather than concentrated in any single taxon. Aquaponic microbiome monitoring should therefore target compartment-specific indicators and operating conditions rather than universal keystone candidates. Cetobacterium nonetheless stood out as a useful fish-tank community marker, but its influence does not extend system-wide.