Bacterial systematics currently lacks a theory-based approach to identify the fundamental units of ecology and evolution. Consequently, a single bacterial species typically contains multiple phylogenetically and ecologically distinct lineages. Thus, the species of bacterial systematics are too broadly conceived to help the microbial ecologist who seeks to behold the full ecological diversity within a community, and to determine the interactions and functions of its ecologically distinct populations. Here, we review some of our recent work aiming to discover the ecologically distinct populations of a clade through analysis of DNA sequence diversity. We have developed an algorithm to compare simulations of bacterial sequence evolution with observed sequence diversity patterns within a clade. By finding the set of parameters that give a maximum likelihood fit between evolutionary simulations and actual sequence data, we are able to estimate lineage-specific rates of evolution and to make predictions about sequence clusters that correspond to ecologically distinct populations (ecotypes). We applied this approach to bacterial isolates of "Evolution Canyon" III in the southern Negev Desert. Within the Bacillus subtilis-B. licheniformis clade, our simulation identified numerous putative ecotypes, some of which were shown to be strongly associated with different microhabitats, confirming their ecological distinctness. Also, several confirmed ecotypes were found to be grouped within a single named species, demonstrating the power of the algorithm to discern ecologically significant variation that is beyond the current focus of bacterial systematics. These findings highlight the promise of a theory-driven approach to identify fundamental units of bacterial diversity.
