Network resilience buffers salinity-induced phase transitions in microbial communities

Authors:

Abstract
Predicting how microbial communities maintain stability under environmental stress remains a central challenge for environmental management. Here, microcosm experiments were conducted to test how community complexity regulates microbial responses to salinity disturbance. Microbial communities with contrasting complexity were generated using simple or heterogeneous carbon resources and then exposed to medium or high salinity. By time-series 16S rRNA gene sequencing, this study showed that microbial communities with higher initial complexity showed weaker phase transitions and greater resilience under salinity stress than communities assembled with lower diversity. ASV-level trajectories indicated that taxa from the same genus could display different response timings and directions depending on community composition. Dynamic co-occurrence networks further showed that network resilience, rather than alpha-diversity stability alone, was the primary determinant of community phase transition, explaining 91.0% of the relative importance in hierarchical partitioning. These findings indicate that microbial resilience under salinity stress is governed less by diversity alone than by the capacity of interaction networks to absorb compositional change, providing a network-centered framework for assessing microbial vulnerability under salinization.