Microbial community coalescence restructures co-occurrence architecture and generates regime-specific patterns of...

Microbial community coalescence (MCC), the mixing of distinct microbiomes, is increasingly recognized as an important process in natural and managed microbial systems. However, how MCC reorganizes species association architecture and whether this restructuring alters community stability remain poorly understood. Here, we investigated how MCC modifies microbial co-occurrence networks and whether resulting architectures deviate from expectations based on stochastic reshuffling or direct compositional mixing of donor communities. Using controlled soil microcosms, we manipulated pairwise coalescence scenarios across a gradient of biotic dilution and tracked community reassembly over 30 days. Empirical association networks were then compared to two reference frameworks: a stochastic balanced-averaging null model and a donor-preserving compositional mixing model. Across treatments, MCC produced extensive restructuring of species association architecture, with empirical networks exhibiting substantially higher node turnover and a predominance of novel correlations relative to both reference expectations. Higher-order network properties also frequently diverged from expected simulated outcomes, indicating that coalescence reshapes correlation architecture beyond stochastic or additive donor mixing. The consequences of this restructuring for network stability were strongly context-dependent. In one coalescence regime, reductions in biotic complexity led to coordinated declines in network complexity and robustness, accompanied by increased fragmentation and vulnerability, whereas other regimes remained insensitive to the biotic dilution gradient. Moreover, community diversity predicted network architecture only under specific coalescence contexts, revealing regime-specific relationships between community diversity, network complexity, and network stability. Together, these results demonstrate that MCC reorganizes microbial association networks in patterns consistent with context-dependent ecological processes that generate distinct complexity and stability patterns.