We identified 64 bacterial OTUs in the cloacae of adult and chick black-legged kittiwakes, a majority of which were identified as Firmicutes and Actinobacteria, while a smaller proportion belonged to the Proteobacteria. A greater number of total OTUs were identified in chick cloacae than in adults. However, the number of OTUs hosted per individual increased with age. Older chicks hosted more OTUs per individual than younger chicks and adults hosted more OTUs per individual than chicks. Surprisingly very little overlap existed in the bacterial assemblages between chicks and adults. Only seven of 64 OTUs were shared between the two age groups (e.g. OTU 3: order Lactobacillales, OTU 9: genus: Lactobacillus and OTU 37: genus Escherichia), while the vast majority were found exclusively in only chicks or adults. For example, Corynebacterium was the most common genus of bacteria identified in adults, but it was virtually absent from chicks. This pronounced difference between adults and chicks resulted in strong statistical clustering of bacterial assemblages according to host age.
ARISA has been widely used to provide important insights in a wide range of fields within microbiology e.g.[34–38]. Coupled with a clone library for species identification, it represents a highly relevant research tool to allow the rapid and inexpensive characterisation of environmental bacterial assemblages. Despite the advantages of ARISA, this technique has some limitations (which are not all necessarily restricted to ARISA). For example, biases inherent during DNA extraction and PCR are also known to affect the apparent composition of bacterial assemblages[39, 40]. ARISA can potentially underestimate species richness, as eight percent of bacterial species are known not to have the 23S and 16S rRNA organised in an operon (i.e. they have no intergenic spacer region) or have very large IGS lengths that cannot be detected by ARISA. In addition, divergent bacteria may share the same IGS length and therefore be associated with the same ARISA peak [this study,. In contrast, some species have several operons in their genomes resulting in multiple ARISA peaks for single species and potentially leading to overestimates in diversity. In our dataset, we conservatively assumed that any two clones with the same 23S sequence derived from the same OTU. We consequently identified many OTUs with multiple operons (1–9 operons were identified per OTU). We also only included OTUs in our analyses for which we could confidently assign ARISA peaks, resulting in several genuine OTUs being excluded. This conservative approach means that our estimates of OTU richness are likely to represent a minimum. However, these biases apply equally for both chicks and adults and we were still able to identify a relatively large number of OTUs suitable for community-level comparisons.
The establishment of bacterial communities in the gastrointestinal tract of young animals is characterised by a high turnover of many transient species and large changes in community structure over short periods of time. For example, González-Braojos et al. 2012 found that, in faecal sacs of nestling pied flycatchers (Ficedula hypoleuca), Enterobacteriaceae loads decreased when the nestlings aged from 7 to 13 days, while Enterococci loads concurrently increased. Age-related changes in the composition of other important gastrointestinal microbes, such as fungi, are also known to occur e.g.. The rapid changes in bacterial community structure in young animals may arise due to a number of reasons including resource competition between bacterial species, shifts in host diet or age-related variation in the chemical and physiological state of the gastrointestinal tract[8, 10, 23]. For example, in humans, the early colonisation of the gut by facultative anaerobes (e.g. Enterococci and Enterobacteria) reduces gut oxygen levels which allows anaerobic bacteria to become established. Eventually, gastrointestinal bacterial communities are known to transition to a stable adult state. Adult bacterial communities may differ from those of young individuals due to the more developed immune system of adults[41–43], the low mobility of young animals resulting in a restricted environment from which to obtain bacteria or contrasting chemical and anatomical cloacal environments that are differentially hospitable or hostile to various bacteria[7, 8].
Our data support these findings. For example, the fact that the chick rarefaction curve failed to plateau and that more OTUs were unique to individual chicks than adults suggests that chicks host more rare, and presumably transient, bacterial species than adults. Second, our finding that adults host a greater number of OTUs per individual than chicks, and that the number of OTUs hosted by chicks increases with age, is in accordance with previous studies that have shown that species richness in bacterial assemblages increases as animals reach adulthood e.g.[19, 20, 26]. Last, we identified substantial variation in bacterial assemblage composition between chicks and adults, highlighting the dynamic nature of bacterial communities within the gut. It is, unfortunately, not possible to deduce the fitness consequences on hosts of age-related changes in bacterial microbiota from our genetic data, especially given the great intrageneric diversity in ecological roles and pathogenicity of bacteria. It, for example, remains unknown why bacteria of the genus Corynebacterium are so prevalent in adults, but almost absent in chicks. However, some inferences can still be made. The seven OTUs shared between chicks and adults may be beneficial or commensal and therefore retained in the gastrointestinal microbiota as the hosts age. For example, two of the shared OTUs were Lactobacillales species, many of which are known to competitively exclude pathogenic bacteria and increase antibody levels, thus increasing immunity to pathogens[44, 45]. Similarly, another shared OTU belonged to the genus Escherichia. Escherichia bacteria are common commensals in the gastrointestinal tract, which are known to rapidly colonise the gut of young birds[18, 19, 24].
The data generated in this study will provide new opportunities to investigate the causes and consequences of variation in bacterial assemblages in a wild bird species. Although much is known in domesticated bird species, relatively little is known about the acquisition of gastrointestinal bacteria assemblages in wild birds. For example, assemblages are known to vary with both external factors, such as nutrition e.g. and environment e.g., and host-related traits such as genotype e.g., body condition e.g., immune system e.g., and sex and mating behaviour e.g.. Our data will allow us to build on these studies with large experimental data sets to explore how specific bacterial species are acquired (e.g. via cross-fostering experiments to ascertain how environment and genotype affect microbial assemblages in chicks) and how they impact on host condition and fitness (e.g. whether individuals that host different bacterial assemblages cf. enterotypes: differ in condition or reproductive success).