ABSTRACT
Little is known about the relative contributions of selective and neutral forces on human-associated microbiota assembly. Here, we characterize microbial community assembly in 52 Tsimane infant-mother pairs, using longitudinally collected stool and tongue swab samples profiled with 16S rRNA gene amplicon sequencing. The Tsimane are an indigenous Bolivian population who practice infant care associated behaviors expected to increase mother-infant dispersal. Infant consumption of dairy products, vegetables, and chicha (a fermented drink inoculated with oral microbes) was significantly associated with gut microbiota composition. At both body sites, maternal microbes at higher relative abundance were more likely to be shared. Shared microbes were also higher in abundance in infants at both body sites, but decreased in average relative abundance with age and were not significantly higher by 12 months of age. Infant microbiotas were modeled using a neutral community model of assembly, which showed that the prevalence of more than two thirds of infant-colonizing microbes could be explained using neutral processes alone. The same method was applied to datasets from Finnish and Bangladeshi infants, confirming that the majority of microbes colonizing infants from different countries were neutrally distributed. Among the Tsimane infant and adult gut microbiota samples, neutral processes were less prominent in villages with more market access. These results underscore the importance of neutral processes during infant microbiota assembly, and suggest that cultural changes associated with market integration may be affecting traditional modes of microbiota assembly by decreasing the role of these neutral processes, perhaps through changes in diet, sanitation, or access to medical care.