The hindgut of lower termites is generally coinhabited by multiple morphologically identifiable protist species. However, it is unclear how many protist species truly coexist in this miniaturized environment, and moreover, it is difficult to define the fundamental unit of protist diversity. Species delineation of termite gut protists has therefore been guided without a theory-based concept of species. Here, we focused on the hindgut of the termite Reticulitermes speratus, where 10 or 11 morphologically distinct oxymonad cell types, that is, morphospecies, coexist. We elucidated the phylogenetic structure of all co-occurring oxymonads and addressed whether their diversity can be explained by the “ecotype” hypothesis. Oxymonad-specific 18S rRNA gene amplicon sequencing analyses of whole-gut samples, combined with single-cell 18S rRNA sequencing of the oxymonad morphospecies, identified 210 one-nucleotide-level variants. The phylogenetic analysis of these variants revealed the presence of microdiverse clusters typically within 1% sequence divergence. Each known oxymonad morphospecies comprised one to several monophyletic or paraphyletic microdiverse clusters. Using these sequence data sets, we conducted computational simulation to predict the rates of ecotype formation and periodic selection, and to demarcate putative ecotypes. Our simulations suggested that the oxymonad genetic divergence is constrained primarily by strong selection, in spite of limited population size and possible bottlenecks during intergenerational transmission. A total of 33 oxymonad ecotypes were predicted, and most of the putative ecotypes were consistently detected among different colonies and host individuals. These findings provide a possible theoretical basis for species diversity and underlying mechanisms of coexistence of termite gut protists.