The authors have declared that no competing interests exist.
Conceived and designed the experiments: TMN TLR MVB. Performed the experiments: TMN. Analyzed the data: TMN MVB. Contributed reagents/materials/analysis tools: TMN TLR MVB. Wrote the paper: TMN.
After birth, mammals acquire a community of bacteria in their gastro-intestinal tract, which harvests energy and provides nutrients for the host. Comparative studies of numerous terrestrial mammal hosts have identified host phylogeny, diet and gut morphology as primary drivers of the gut bacterial community composition. To date, marine mammals have been excluded from these comparative studies, yet they represent distinct examples of evolutionary history, diet and lifestyle traits. To provide an updated understanding of the gut bacterial community of mammals, we compared bacterial 16S rRNA gene sequence data generated from faecal material of 151 marine and terrestrial mammal hosts. This included 42 hosts from a marine habitat. When compared to terrestrial mammals, marine mammals clustered separately and displayed a significantly greater average relative abundance of the phylum
Bacteria inhabiting the gastro-intestinal tract of mammals expand their host’s metabolic potential by harvesting energy that would otherwise be inaccessible
Mammalian hosts first acquire their gut bacterial community during transport through the birth canal and subsequently through maternal, social and environmental transmissions
In a pioneering study, Ley et al. 2008
Further insight could be gained by comparing a diverse range of extant mammals with differing life history traits. One group of mammals that have been relatively understudied are marine mammals. Their comparatively recent evolution and differing life history traits and adaptation to a marine habitat
To understand patterns in gut bacterial communities, the faecal bacterial communities of a broad range of terrestrial and marine mammals were compared. Marine mammals included two species of seals inhabiting the Antarctic
Samples collected from southern elephant seals and leopard seals were carried out in strict accordance with the recommendations in the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes. Protocols used in the study were approved by the University of New South Wales Animal Care and Ethics Committee (permit number 08/83B and 03/103B). The southern elephant seal is listed as vulnerable under the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act) and listed under the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES). Permission to export southern elephant seal biological materials was obtained from the Australian Government Department of the Environment, Water, Heritage and the Arts (permit number 2008-AU-534289).
Permission to access regions in Antarctica where seals were located was approved by the Ministry of Foreign Affairs and the Dirección Nacional del Antártico. Southern elephant seals in this study were located in Antarctic Special Protected Area 132 ‘Peninsula Potter’ and additional permissions were obtained through the Dirección Nacional del Antártico under Article 3, Annex II of the Madrid Protocol to the Antarctic Treaty (no permit number). For southern elephant seal males and leopard seals, individuals were anaesthetised using a mixture of tiletamine and zolazepam (Zoletil – Virbac Australia) at a combined dose of 1 mg/kg. Female southern elephant seals were anaesthetised with 3–6 mg/kg ketamine hydrochloride. On all occasions, procedures were performed by qualified personnel and all efforts were made to minimize suffering.
Studies for comparison were selected on the basis that analysis methods for bacterial community composition sequenced a region of the 16S rRNA gene using highly conserved bacterial or universal primer sets. The specific methods each study used to generate this data are outlined in
Individual sequence data were obtained from the National Centre for Biotechnology Information website (
Sequence taxonomy was assigned using the Ribosomal Database Project II (RDP) v.10 Classifier tool
Meta-analyses such as this may be prone to study effects, whereby similarity or differences in methodology, rather than ecology, generate the observable patterns. To address concerns over any potential study effects, we selected and analysed data from four host phylogenetic families, the
In addition, to examine the effect of data standardisation versus data rarefaction (to enable incorporation of datasets that contained fewer than 100 sequences per host), the complete dataset was rarefied to 24 sequences per host, which was the lowest number present in any study used. When analysed using the same techniques as described below (see methods) this subsampled dataset generated the same statistically significant overall patterns as those observed when using the expanded dataset (see
The assessments of study effects give provide the author’s with confidence that the results reported herein describe ecological effects rather than methodological effects.
The combined dataset was standardised prior to transformation. This involved converting abundance counts into relative percentages for each individual host. A Bray-Curtis dissimilarity matrix
Differences in the composition of the gut bacterial community between hosts were tested with the non-parametric permutation procedure ANOSIM (Analysis of Similarity) as it is more robust to heterogeneous dispersion of data
Herbivores and carnivores displayed significant differences in the composition of their gut bacterial communities (ANOSIM: R = 0.49,
nMDS ordination plot dislaying similarity of the gut bacterial community in the host mammal as grouped by diet and habitat. See Figure S4 for detailed display of this figure.
Source of variation | Pair-wise comparisons | ||
Diet | 0.39* | <0.01 | |
Herbivores, omnivores | 0.33 | <0.01 | |
Herbivores, carnivores | 0.49* | <0.01 | |
Omnivores, carnivores | 0.26 | <0.01 | |
Diet and habitat | 0.52** | <0.01 | |
Terrestrial herbivores,terrestrial omnivores | 0.33* | <0.01 | |
Terrestrial herbivores,terrestrial carnivores | 0.62** | <0.01 | |
Terrestrial herbivores,marine carnivores | 0.65** | <0.01 | |
Terrestrial omnivores,terrestrial carnivore | 0.32 | <0.01 | |
Terrestrial omnivores,marine carnivores | 0.50** | <0.01 | |
Terrestrial carnivores,marine carnivores | 0.69** | <0.01 | |
Phylogenetic order | 0.19 | <0.01 | |
Phylogenetic family | 0.50** | <0.01 | |
Gut morphology | 0.11 | <0.01 | |
Hindgut fermenters,simple guts | 0.14 | <0.01 | |
Hindgut fermenters,foregut fermenters | 0.17 | <0.01 | |
Simple guts,foregut fermenters | 0.15 | <0.01 |
ANOSIM of gut bacterial abundance data was used to generate a permutated Global R statistic (R) and permutated p-value (
Marine carnivores possessed a significantly lower average relative abundance of the phylum
Average relative abundance of each major phyla in the gut bacterial community of host mammals grouped by habitat and diet. Error bars represent standard errors (SE). The lack of replication in the marine herbivore grouping does not allow for estimation of SE or significance testing. Student’s paired t-test were conducted between groups as displayed in
Comparison | ||||
Terrestrial herbivores,terrestrial omnivores | 0.473 | 0.996 | 0.423 | 0.001** |
Terrestrial herbivores,terrestrial carnivores | 0.033 | 0.001** | 0.828 | 0.002** |
Terrestrial herbivores,marine carnivores | <0.001*** | 0.502 | 0.018* | <0.001*** |
Terrestrial omnivores,terrestrial carnivores | 0.165 | 0.001** | 0.266 | 0.735 |
Terrestrial omnivores,marine carnivores | <0.001*** | 0.536 | 0.004** | <0.001*** |
Terrestrial carnivores,marine carnivores | <0.001*** | 0.002** | 0.151 | 0.005** |
Terrestrial mammals,marine mammals(all diet types) | <0.001*** | 0.991 | 0.002** | <0.001*** |
Student’s paired t-test of gut bacterial abundance data to generate a p-value (
The phylum
nMDS plot displays gut bacterial community of all host mammals in the order
Members of the family
Herbivores possessed a faecal bacterial community significantly richer than that of carnivores or omnivores (
Richness of the gut bacterial community was measured using Chao 1 mean. Error bars represent standard error (SE). Student’s paired t-test were conducted between groups with significance level:
The composition of the gut bacterial community of the marine mammals available at the time of this study is clearly distinct from that of the available terrestrial mammals. Differences in the gut bacterial community of carnivorous marine mammals appears to be due, in part, to their considerably reduced abundance of
The occurrence of greater than average abundance of
Composition of the leopard seal gut bacterial community was previously shown to differ significantly in captivity compared with wild hosts, as a result of specific dietary items and local habitat differences
Several challenges are faced when consuming plant material as a primary food source due to the indigestible cell walls
(TIF)
(TIF)
(TIF)
(TIF)
(TIF)
Overview of main methods employed by included studies.
(DOCX)
Characteristics of mammalian hosts used in the study. Abbreviated table data is as follows: number of sequences used (No. of seq.); gut morphology (Gut morph.); hindgut fermenter (HG); foregut fermenter (FG); simple gut (S); marine (M); terrestrial (T); carnivore (C); herbivore (H); and omnivore (O).
(DOCX)
Comparison of ANOSIM results between the genera rarefied to the minimum of 24 and those subsampled to 100. Results display those reported in
(DOCX)
Characteristic genera in the gut bacterial community of mammal hosts grouped by diet and habitat. The foremost ten characteristic genera in the gut bacterial community of hosts identified using SIMPER analysis. Hosts are grouped based on diet and habitat.
(DOCX)
The authors are grateful for the field assistance and sample collection provided by Instituto Antártico Argentino, the Australian Marine Mammal Research Centre and Taronga Zoo. Three independent reviewers are thanked for their assessment of the study.