Environmental selection constrains metabolic network architecture despite taxonomic turnover in anaerobic digestion communities.

Microbial ecosystems often sustain stable metabolic functions despite pronounced taxonomic turnover, yet the mechanisms underlying such reproducible functional states remain poorly understood. Here, we investigated how physicochemical constraints shape functional convergence in anaerobic digestion (AD) communities using replicated serial enrichments seeded from four distinct inocula. Across three pH levels and six substrate regimes, replicate communities from different inocula consistently converged toward reproducible metabolite profiles, with pH emerging as the dominant organizing factor. Community composition became progressively environment-driven over time, and after 30 generations, pH explained the largest fraction of compositional variance (permutational multivariate analysis of variance R2 = 0.21, P = .001), followed by substrate. Genome-resolved metagenomics revealed that convergence was accompanied by strong pH-dependent structuring of redox-balancing and terminal electron-sink pathways, whereas upstream carbohydrate-entry pathways were conserved. Taxonomic convergence was incomplete and scale-dependent: the ability to correctly assign communities to their inoculum declined from 75% at the genus level to 53% at the phylum level, indicating increasing similarity across inocula at coarser taxonomic resolution despite persistent fine-scale variability. Despite this taxonomic flexibility, communities assembled under identical conditions consistently recruited similar sets of metabolic pathways organized into comparable network architectures. Functional redundancy analyses showed high redundancy and flexible taxonomic implementation for upstream fermentative processes, contrasted with lower redundancy and stronger convergence for terminal methanogenic functions. Together, these results demonstrate that reproducible metabolic function in ad emerges from environmentally constrained assembly of shared metabolic network architectures, rather than deterministic fixation of species composition, highlighting environmental control of metabolic organization as a central principle governing microbiome function.