Microbial Species-Area Relationships on the Skins of Amphibian Hosts.

Abstract

Unlike species-area relationships (SARs) that have been widely reported for plants and animals on Earth, there is no clear understanding of the SARs for microorganisms. In this study, 358 specimens of 10 amphibian host species collected from the rural Chengdu region of southwest China were selected as island models for evaluating SAR curve shapes and assessing the skin microbiota from different amphibian species. The results showed that skin microbial diversity, measured using Hill's number, presented significant differences between hosts, but the difference was insignificant between habitat-specific classifications of hosts. As for microbial SARs, other than the classical power-law (PL) model describing an expected steady increase in microbial diversity as sampled skin area increases, two additional trends were observed: (i) microbial diversity first rises and gradually decreases after reaching a maximum accrual diversity (MaxAD) and (ii) microbial diversity decreases and starts to rise after reaching the minimum accrual diversity (MinAD). Among the four SAR statistical models compared, it was consistently found that the models that can describe MaxAD were favorably selected in the highest frequency. Models that can describe MinAD and PL model also performed reasonably well. However, PL had the poorest fitting power, implying the necessity of introducing biologically meaningful complex SAR models in microbial diversity research. In conclusion, through multihost analyses, our study provided compelling evidence that microbial SARs are complex and nonlinear. A variety of ecological mechanisms may be used for explaining these, including, but not limited to, community saturation, small-island effects, or sampling heterogeneity. IMPORTANCE In this study, we investigate species-area relationships (SARs) for skin-borne symbiotic microbes of wildlife hosts. Unlike the traditional SARs for plants and animals, symbiotic microbial SARs were complex. We found that both U-shaped and inverted U-shaped SAR models were widely favored for microbial taxa than the well-known power-law model in different host species. These favored models presented interesting statistical features, including minimal or maximal accrual diversity or inflection point. We provide intuitive derivations of these statistical properties. We showed that different habitat-specific amphibian hosts did not present distinct microbial diversity and skin-related SAR patterns. We predicted that approximately 600 to 1,400 cm2 (in two-dimensional [2D] measurement) or approximately 1,200 to 3 500 cm2 (in 3D measurement) are the skin area threshold range that can allow the emergence of minimal or maximal accrual microbial diversity with high chances. Finally, we list a variety of ecological mechanisms that may be used for explaining the observed nonlinear SAR trends.