Fitting microbial community adaptation of respiration and growth to warming using a genomics-informed agent-based model

 

Soil microbial communities control the fate of the largest terrestrial organic carbon pool, and their decomposition and respiration dynamics are pivotal for predicting future climate feedbacks. Community diversity, functional complexity, and adaptive responses may substantially reshape projections of the global carbon cycle.

Yet, most microbial-explicit soil biogeochemical models rely on simplified communities with static traits (e.g. growth and respiration). Approaches that incorporate microbial diversity and evolutionary processes remain largely theoretical and poorly constrained by empirical diversity and geochemical measurements, limiting their applicability in Earth system model predictions.

Here, we bridge this gap by fitting microbial community adaptation to warming using a genomics-informed, agent-based microbial model (DEMENT). We develop a framework to parameterize realistic microbial communities from metagenome-assembled genomes (MAGs), capturing taxon-specific traits related to enzyme production, resource uptake, and carbon allocation. Using long-term soil warming experiments at the Harvard Forest LTER site as a case study, we explicitly simulate the temporal dynamics of microbial community composition, respiration, and organic matter degradation under warming. We evaluate alternative evolutionary scenarios of microbial adaptation; targeting resource acquisition, growth yield, and stress responses; and identify the scenario that best reproduces observed diversity patterns as well as post-adaptation growth and respiration responses across temperature gradients.

This approach enables the identification of evolutionary pathways underlying microbial community responses to warming and provides a critical foundation for integrating adaptive microbial processes into next-generation Earth system models.