Association of maternal genetics with the gut microbiome and eucalypt diet selection in captive koalas

Abstract

Background: Koalas, an Australian arboreal marsupial, depend on eucalypt tree leaves for their diet. They selectively consume only a few of the hundreds of available eucalypt species. Since the koala gut microbiome is essential for the digestion and detoxification of eucalypts, their individual differences in the gut microbiome may lead to variations in their eucalypt selection and eucalypt metabolic capacity. However, research focusing on the relationship between the gut microbiome and differences in food preferences is very limited. We aimed to determine whether individual and regional differences exist in the gut microbiome of koalas as well as the mechanism by which these differences influence eucalypt selection.

Methods: Foraging data were collected from six koalas and a total of 62 feces were collected from 15 koalas of two zoos in Japan. The mitochondrial phylogenetic analysis was conducted to estimate the mitochondrial maternal origin of each koala. In addition, the 16S-based gut microbiome of 15 koalas was analyzed to determine the composition and diversity of each koala's gut microbiome. We used these data to investigate the relationship among mitochondrial maternal origin, gut microbiome and eucalypt diet selection.

Results and discussion: This research revealed that diversity and composition of the gut microbiome and that eucalypt diet selection of koalas differs among regions. We also revealed that the gut microbiome alpha diversity was correlated with foraging diversity in koalas. These individual and regional differences would result from vertical (maternal) transmission of the gut microbiome and represent an intraspecific variation in koala foraging strategies. Further, we demonstrated that certain gut bacteria were strongly correlated with both mitochondrial maternal origin and eucalypt foraging patterns. Bacteria found to be associated with mitochondrial maternal origin included bacteria involved in fiber digestion and degradation of secondary metabolites, such as the families Rikenellaceae and Synergistaceae. These bacteria may cause differences in metabolic capacity between individual and regional koalas and influence their eucalypt selection.

Conclusion: We showed that the characteristics (composition and diversity) of the gut microbiome and eucalypt diet selection of koalas differ by individuals and regional origins as we expected. In addition, some gut bacteria that could influence eucalypt foraging of koalas showed the relationships with both mitochondrial maternal origin and eucalypt foraging pattern. These differences in the gut microbiome between regional origins may make a difference in eucalypt selection. Given the importance of the gut microbiome to koalas foraging on eucalypts and their strong symbiotic relationship, future studies should focus on the symbiotic relationship and coevolution between koalas and the gut microbiome to understand individual and regional differences in eucalypt diet selection by koalas.

Keywords: Captive animal; Diet selection; Diet specialist; Eucalyptus; Gut microbiome; Hindgut fermentor; Koala; Marsupial; Maternal transmission; Mitochondrial DNA.

Figure 1. The ML tree of the D-loop in koalas.

This is inferred using the Tamura three-parameter model with a discrete Gamma distribution (five categories), which has the lowest Bayesian Information Criterion scores (unrooted). The tree with the highest logarithmic likelihood (−990.84) is shown. The percentage (≥ 50%) of trees based on 1,000 bootstrap replications is shown below the branches. This tree shows four clades (Northern 1, Northen 2, Central, and Southern), which corresponding to the clades found by Neaves et al. (2016).

Figure 2. Alpha diversity.

The differences between management groups in (A) Shannon index and in (B) Chao1 and mitochondrial lineages in (C) Shannon index and (D) Chao1. Statistical tests are conducted by the pairwise Kruskal–Wallis test with Benjamini–Hochberg correction.

Figure 3. Management groups and mitochondrial lineages significantly influenced the gut bacterial community.

Clustered by the management groups in (A) unweighted UniFrac (pairwise PERMANOVA tests: pseudo- F = 16.698158; P = 0.001; number of permutations = 999) and (B) weighted UniFrac (pairwise PERMANOVA tests: pseudo- F = 27.864924; P = 0.001; number of permutations = 999). Clustered by mitochondrial lineages in (C) unweighted UniFrac (pairwise PERMANOVA tests, Center vs North2: pseudo- F = 8.025514; q = 0.001; Center vs South: pseudo- F = 8.133812; q = 0.001; North2 vs South: pseudo- F = 7.289387; q = 0.001; number of permutations = 999) and (D) weighted UniFrac (pairwise PERMANOVA tests, Center vs North2: pseudo- F = 5.784462; q = 0.0015; Center vs South: pseudo- F = 11.216916; q = 0.0015; North2 vs South: pseudo- F = 5.085741; q = 0.002; number of permutations = 999). (E) Volcano plot of the results of the analysis of the composition of microbiomes (ANCOM) between management groups or (F) mitochondrial lineages at the genus level. Each circle represents a taxon. Those with statistically significant differences based on the W statistics between mitochondrial lineages are colored red.

Figure 4. Relationships of diet and gut microbiome.

(A) Diet proportion of each individual. (B) Results of NMDS using foraging data. Mitochondrial lineages significantly influenced the foraging pattern (F = 5.88; R² = 0.677; P = 0.014; number of permutations = 719). (C) Correlation between the foraging diversity and gut microbiome diversity (Shannon index). A significant correlation was observed between foraging diversity and gut microbiome diversity (Spearman’s rank correlation; ρ = 0.886, P = 0.009). (D) The heat map shows changes in the relative abundance of gut bacteria because of changes in the proportion of eucalypt species in the diet (the color bar shows the strength of correlation). Spearman’s rank correlation was conducted with Benjamini–Hochberg correction, and bacteria with significant correlations are indicated with an asterisk (q < 0.05; see Table S7 for individual values). The genera with the correlation coefficient ∥ρ ∥ of > 0.7 are shown. T, E. tereticornis; M, E. microcorys; CR, E. camaldulensis; R, E. robusta; P, E. punctata.