<p>Mosquito larvae develop in aquatic habitats that harbor highly variable communities of bacteria and other microorganisms, which have been well demonstrated to shape individual fitness outcomes in laboratory settings. However, relatively little is known about how this microbial variation contributes to or is influenced by mosquito population dynamics in the field. To investigate potential associations between mosquito population dynamics and microbial community assembly, we characterized bacterial communities in naturally occurring larval habitats with variable historical mosquito productivity using amplicon sequencing. We then applied a null model approach to quantify the relative importance of selection, dispersal, and drift processes in bacterial community assembly. Habitat microbiota clustered into two distinct biotypes: Biotype 1 communities were dominated by Proteobacteria, while Biotype 2 communities were dominated by Cyanobacteria. Both biotypes were shaped by a combination of selection and neutral (i.e., dispersal and drift) processes. However, selection played a more prominent role in habitats with Biotype 1 communities, whereas drift was more influential in Biotype 2 habitats. Variation partitioning further identified historical mosquito productivity and the spatial aggregation of sites with similar productivity histories as key drivers of selection. These findings suggest that mosquito population dynamics are associated with differences in microbial community structure, potentially through feedbacks between mosquito activity and habitat conditions. This study lays the foundation for future work to disentangle causal relationships and to integrate patterns of microbiota diversity and mosquito occurrence into vectorial capacity models for improved prediction of mosquito-borne disease transmission dynamics in the field.</p>