Amphibian Skin Microbiota Response to Variable Housing Conditions and Experimental Treatment across Space and Time
Host-associated microbial ecology research is becoming a popular tool in conservation. For amphibians, there is recognition that environmental factors and anthropogenic activities can alter the composition and function of skin microbiotas. Despite growing studies on this topic, the environmental conditions and experimental methods that amphibians are exposed to vary among studies, potentially influencing our ability to develop generalizations. Using Northern Leopard Frogs (Lithobates pipiens), we investigated how the movement of individuals between housing conditions affected the amphibian skin microbial communities. In addition, we evaluated whether variation in experimental venue (e.g., culture pools, mesocosms, or laboratory) and time influenced the effect of a common antibacterial (sulfadimethoxine; SDM) on the skin microbiota. We found that the microbial community diversity decreased when tadpoles were transferred from culture pools to mesocosms and subsequently increased between mesocosms and the laboratory. When comparing the effect of SDM exposure on the skin microbiota of tadpoles across experimental venues (mesocosm and lab), there was no effect of SDM on alpha diversity. However, we noted opposing patterns between the control and SDM-treated individuals within the mesocosm and laboratory groups. In the laboratory, there were differences in the abundances of operational taxonomic units (OTU) while in the mesocosm there were differences in OTU turnover. Finally, we found that SDM treatment on amphibian microbial communities was consistent across time in the laboratory. Because researchers are integrating microbial assessments into our understanding of conservation biology, our results underscore the importance of standard housing conditions and taking into consideration that experimental design may yield variable results.Abstract

Alpha diversity values measured from the skin of Northern Leopard Frog tadpoles in three different environments (culture pools, mesocosms, and laboratory). Long horizontal lines representing the mean and standard error lines are presented.

Nonmetric multidimensional scaling (NMDS) plot of Northern Leopard Frogs raised in three separate environments. Analyses depict coordinates derived using (A) unweighted UniFrac, (B) weighted UniFrac, and (C) Bray-Curtis distance matrices.

Average pairwise distances between control and SDM exposed individuals in mesocosm and laboratory environments. Average distances are presented for three different distance metrics; unweighted UniFrac (presence/absence-based), weighted UniFrac (abundance-based), and Bray-Curtis (abundance-based). Long horizontal lines representing the mean and standard error lines are presented.

Photos of the housing conditions utilized in this study. Individuals were moved from (A) common garden culture pools to (B) seminatural mesocosms and finally into (C) laboratory conditions (Table 1).

Relative abundance of most common OTUs (relative abundance >1.0%) characterized on the skin of Northern Leopard Frogs inhabiting three different captive environments.

Bipartite network displaying the distribution of OTUs across the core microbiotas of Northern Leopard Frogs in three rearing environments (core microbiota corresponds to OTUs present in 80% of individuals). Lines connect OTUs to rearing environment and are weighted by relative abundance.
Contributor Notes