UC Irvine

Microbial communities play an essential role in sustaining ecosystem processes and life on Earth. As such, the field of microbial ecology has gradually shifted from establishing taxonomic patterns of distribution to identifying the traits that dictate where different microorganisms live to isolating the eco-evolutionary processes that create and maintain these patterns. Yet our understanding of these matters still largely reflects broad taxonomic levels. For soil bacteria in particular, most studies rely on highly conserved markers (e.g., the 16S rRNA gene). However, genomic variation at these scales overlooks substantial trait divergence that is potentially ecologically relevant. As a result, our understanding of how closely related lineages, akin to bacterial species, partition the soil environment and assemble into communities remains limited. In this dissertation I aimed to address this gap by using the Curtobacterium genus, an abundant bacterial taxon in soil leaf litter, as a model organism and ask: 1) How does the distribution of different lineages of Curtobacterium change along environmental gradients? 2) Does the phenotypic characterization of representative isolates of these lineages align with the biogeographic patterns observed? and 3) Are the biogeographic and phenotypic patterns observed a result of local adaptation to combinations of environmental conditions?In the first section of this dissertation, I addressed the first two questions by conducting a large-scale survey on 24 locations across California to capture a wide environmental gradient in terms climate, ecosystems and vegetation types. I collected both grass and dominant-vegetation litter at each site to decouple substrate effects from climate, and isolated Curtobacterium strains from these samples for phenotypic characterization. Using metagenomic sequencing I characterized microbial communities, and applied random forest modeling, distance-based ordination, and lab-based thermal and pH performance curves to link lineages distributions to environmental drivers. Overall Curtobacterium abundance was primarily predicted by both litter chemistry (cellulose content) and climate. The four most abundant ecotypes (lineages) exhibited distinct biogeographic patterns: ecotype IIIA was more abundant in cooler, wetter sites with lignin-rich litter, while ecotypes IC and IVB were associated with warmer, drier conditions, and ecotype IIG showed broader thermal tolerance. These field patterns were corroborated by phenotypic data, with ecotype IC showing significantly higher temperature and pH optima than IIIA. These results demonstrate that Curtobacterium ecotypes undergo niche partitioning shaped by both climate and litter chemistry and suggest that accounting for heterogeneity in the soil substrate can reveal underlying climate signal.To test whether these biogeographic and phenotypic patterns reflect adaptation, in the second section of this dissertation I addressed the third question by conducting a replicated reciprocal transplant anchored at a focal site in southern California and replicated across a subset of eight sites from the initial survey. I manipulated site (as a proxy of climate), leaf litter substrate and inoculum community origin, to tease apart the effect of these three factors on compositional responses and asses the adaptive response of individual ecotypes to the experimental manipulations. Transplanted communities shifted in composition toward that of the native away community, with both climate and litter substrate chemistry independently driving convergence. These patterns were consistent across two phylogenetic scales: the whole bacterial community and the genus Curtobacterium, for which independent biogeographic and phenotypic evidence supports ecotypic differentiation along the same environmental axes. Specifically, the four most abundant Curtobacterium ecotypes shifted in relative abundance across transplant paths in directions predicted by their thermal preferences and biogeographic distributions, suggesting that community-level convergence reflects the sorting of evolutionarily differentiated lineages rather than stochastic assembly. Together, these results demonstrate that leaf litter bacterial communities are locally adapted to both climate and litter substrate chemistry and establish community-level convergence as a metric for detecting local adaptation in microbial systems where individual-level approaches remain infeasible. As microbial communities underpin critical ecosystem processes including decomposition and nutrient cycling, understanding the eco-evolutionary mechanisms that structure their diversity across environmental gradients has important implications for predicting how these processes will respond to ongoing environmental change.