Microbiome structure varied with geography and host immunogenetics
Geography was a significant factor in microbiome structure (beta diversity) for both bacterial and eukaryotic microbes. However, microbiome structure was not associated with genetic structure of populations at either neutral genetic markers or the MHC IIB immunogenetic locus. These results differ from a previous study on the frog Amietia hymenopus, which showed opposite patterns: there were no geographic effects on amphibian skin microbiome structure, but there was a significant association with population genetic structure (29). One possible explanation for the discrepancy between our results and the results from A. hymenopus (barring host identity factors) is that geographic structure in the host-associated microbial community is scale-dependent: our study spans a larger geographic area (~100 km compared with ~4 km in Griffiths et al. 2018). In addition, our study populations represent a set of connected mainland populations contrasted with a set of island populations that have been isolated for 12,000-20,000 years. The lack of association with genetic differentiation in our populations may be due to this relatively long period of divergence, or to isolation between island sites resulting in different environmental availability of microbes.
Microeukaryote diversity was associated with host genetic diversity, with genetically impoverished island populations possessing lower microeukaryotic diversity relative to coastal mainland populations (85.5 average OTUs on islands vs. 110.5 in coastal sites). This difference in microbiome diversity could be due to a number of factors, including less favorable environments or lower rates of host contact (i.e. , microbial transmission) on islands compared with coastal sites. However, MHC IIB heterozygosity was positively associated with microeukaryotic diversity even when only coastal populations were analyzed (19.2 average OTUs on coastal homozygotes vs. 39.1 on coastal heterozygotes). Taken together, these results imply that genetic diversity and/or MHC IIB genotype plays a significant role in determining microbiome diversity.
Microbiome structure also varied across site types and MHC IIB genotypes. Variation in microbiome structure among site types could be parsimoniously explained by variation in environmental filtering in coastal vs. island sites. However, these differences may also be driven by island isolation favoring longer-dispersing microbes, or alternatively by host genetic factors. The variation in microbiome structure across MHC IIB genotypes, although weaker than the variation due to site type, may be a clearer example of associations between endogenous host factors and the microbiome. Although MHC genes are thought to be primarily involved in pathogen resistance, results from laboratory and field studies suggest that MHC genotype and allelic composition can impact amphibian host-associated microbial assemblages (30,63). Together with our finding that microbiome structure and diversity are influenced by MHC genotype, this suggests that immune mechanisms conferred by MHC genes may influence the assembly of the overall microbiome.