Discussion
We found clear localization of the microbiome in the S. virgatusgut and reproductive tracts. While localization has been seen in other reptiles (Colston et al., 2015; Costello et al., 2010; Keenan et al., 2013; Kohl et al., 2017), the cloaca in previously sampled reptiles was found to have relatively high diversity and to host representatives from the upper regions of the gut. This pattern may be expected as the cloaca is the terminus of the GI and reproductive systems, and is often assumed to be both inoculated by feces that pass through and influenced by sexual transmission during copulation (Videvall et al., 2018; Wen et al., 2021; White et al., 2011). Similarly, the reproductive tract of chickens shows increasing levels of diversity from oviducts to cloaca, which has the highest diversity (Wen et al., 2021). In contrast to these previously described patterns, the cloacal microbiome of S. virgatus females is low diversity, especially at the Family and Phylum level at which it was dominated byEnterobacteriaceae and Protobacteria, and is distinct from upper intestine, oviductal, and fecal microbiomes. This pattern may indicate strong selection for the specific cohort of bacteria found in theS. virgatus cloaca. S. virgatus females deposit beneficial microbes on their eggs during oviposition that facilitate egg survival (Bunker et al., in review). This function could be the driving force behind the winnowing of the microbiome in the cloaca, especially considering that many members of the Enterobacteriaceae family are known to have antifungal properties (Dhar Purkayastha et al., 2018; Gutiérrez-Román et al., 2015; Kalbe et al., 1996).
Cloacal swabs taken from a given individual lizard resulted in communities similar to the lower intestine and the cloacal tissue, and distinct from higher regions of the gut and reproductive tract. The cloacal swab community was predominantly Enterobacteriaceae but, like the lower intestine, had several families that made small but noticeable contributions to the community and was dispersed slightly differently than the cloacal tissue microbiome. The upper intestine and oviduct have different microbial communities than the lower sections of the intestinal tract but also have differences in major families from one another. Recent evidence has contradicted the long-accepted idea that embryonic development is sterile (Funkhouser & Bordenstein, 2013). Microbial communities similar to gut microbes have been found in chicken oviducts (Shterzer et al., 2020), and Trevaline et al. (2018) found that microbes may colonize bird and lizard eggs from within the oviduct before the shell develops. The unique community in the oviduct here could indicate an internal egg microbiome which is seeded during egg development; the functional significance of this microbiome should be investigated further.
In contrast to the adequate sampling of the cloacal and lower intestinal microbial communities by cloacal swabs, fecal pellets contain strikingly distinct communities from all gut regions, with greatest similarity to the upper intestine. Pre-defecation cloacal swabs and the fecal samples taken from the same individuals were found to have different microbial communities from one another by every metric we examined. As described above, the swabs were more similar to the cloaca and lower intestine, and largely dominated by Enterobacteriaceae . The fecal sample community was most similar to that found in the upper intestine, with a relatively high abundance of Ruminococcaceae and Bacteroidaceae,but was dominated by Lachnospiraceae which was largely absent from all regions of the gut, and lacked the high abundance ofEnterobacteriaceae that was found across the gut. Overall, these patterns caution against the common practice of using fecal samples as a proxy for the gut microbiome without first validating this approach.
While Study 2 found some changes between pre- and post-defecation, those changes were largely non-significant, and likely due to increased variation within the post-defecation swabs, as indicated by differences in dispersion, more so than composition, between time points. The variation is due to a distinct bifurcation in the recovered swab communities from post-defecation swabs, with most of these swabs mirroring cloacal tissue, with its high abundance of Proteobacteria, and a smaller portion displaying a more “feces-like” community. This variation in post-defecation cloacal swabs was also present in the field swabs of lizards from Study 1; 5 of the 8 cloacal swab communities from Study 1 were made up of >95% Enterobacteriaceae andHelicobacteraceae , while other samples were dominated byLachnospiraceae and Bacteroidaceae , two feces-associated taxa. Because recovered communities from feces and pre-defecation cloacal swabs were distinct, it seems unlikely that the cloaca is being inoculated with fecal microbes. Rather, a more likely explanation is that small amounts of fecal material may attach to swabs if they are taken shortly after defecation, which then masks the much less diverse cloacal community when it is sequenced. The idea of temporary contamination is further supported by the fact that the same lizards, when re-swabbed for a different study, often had comparatively lower diversity and a greater relative abundance ofEnterobacteriaceae -associated ASVs (Online Resource 3). Some animals continued to vary between the feces-like and cloacal tissue phenotype when they were resampled, likely due to repeated defecation events. Cloacal swabs often have been found to be inconsistent and unreliable (Videvall et al., 2018; Williams & Athrey, 2020), and potential contamination by fecal material could account for some of this variation. Indeed, cloacal swabs have been used to collect feces in the past (Stanley et al., 2015). The fact that these fecal microbes do not colonize and grow in the S. virgatus cloaca supports the hypothesis that the low microbial diversity of the cloaca is due to selection, and that there is a mechanism to maintain stability of the microbiome in that region. This type of selection has been seen in other species (Nyholm & McFall-Ngai, 2004; Zhang et al., 2016), and supports a model in which function is the driving force of microbiome diversity (Reese & Dunn, 2018).
At the phylum level, Bacteroidetes and Firmicutes are often considered to represent the core gut microbiome across vertebrates (Colston & Jackson, 2016; Ley et al., 2008). However, this dogma is based largely on studies that rely on fecal sampling in mammals and so may not accurately represent endogenous microbial communities of diverse vertebrate taxa. For instance, Bacteroidetes have been found to be relatively rare in some gut regions of wild reptiles and birds (Colston et al., 2015; Hird, 2017; Keenan et al., 2013; Kreisinger et al., 2015), and were relatively uncommon across all S. virgatus gut regions and even in fecal pellets, which were dominated by Firmicutes. Firmicutes and Proteobacteria were more equally represented in the S. virgatusupper intestine, though Proteobacteria were more common, and Proteobacteria increased in dominance down the GI tract to the cloaca, which was nearly entirely composed of this phylum. Proteobacteria have been found to be highly abundant in the GI tract of other non-mammalian vertebrates (Colston & Jackson, 2016), particularly in studies that sampled directly from the GI tract rather than relying on fecal sampling alone, but its dominance in the S. virgatus cloaca appears to be extreme even in comparison to those systems. While other studies have found a majority of the microbiome was composed of Proteobacteria, the lowest relative abundance of this cloacal community member was 96.9% with greater than 99% on average, and the majority of those reads belonged to a single family. The cloacal swabs were less consistent, as discussed above, although still 85% Proteobacteria on average, and seemed to land between the cloacal tissue community and that of the lower intestine.
Variation in the endogenous microbiome across taxa has been considered in relation to diet and other aspects of animal life history and ecology (Colston & Jackson, 2016; Ley et al., 2008). As descriptions of vertebrate microbiomes continue to accumulate, it will be interesting to examine the potential influence of reproductive mode (i.e., viviparity vs. oviparity), as this separates mammals from most non-mammalian vertebrates. We propose that oviparous species, and especially those without egg-tending, may have unique selective pressures on the cloacal microbiome to transfer antifungal or otherwise egg-protective bacteria to eggshells during oviposition. Future research will compare the cloacal microbiome of oviparous and viviparous Sceloporus lizards.