Comparing the gut microbiota of Sichuan golden monkeys across multiple captive and wild settings: roles of anthropogenic activities and host factors

Abstract

Background: Captivity and artificial food provision are common conservation strategies for the endangered golden snub-nosed monkey (Rhinopithecus roxellana). Anthropogenic activities have been reported to impact the fitness of R. roxellana by altering their gut microbiota, a crucial indicator of animal health. Nevertheless, the degree of divergence in gut microbiota between different anthropogenically-disturbed (AD) R. roxellana and their counterparts in the wild has yet to be elucidated. Here, we conducted a comparative analysis of the gut microbiota across nine populations of R. roxellana spanning China, which included seven captive populations, one wild population, and another wild population subject to artificial food provision.

Results: Both captivity and food provision significantly altered the gut microbiota. AD populations exhibited common variations, such as increased Bacteroidetes and decreased Firmicutes (e.g., Ruminococcus), Actinobacteria (e.g., Parvibacter), Verrucomicrobia (e.g., Akkermansia), and Tenericutes. Additionally, a reduced Firmicutes/Bacteroidetes ratiosuggested diminished capacity for complex carbohydrate degradation in captive individuals. The results of microbial functional prediction suggested that AD populations displayed heightened microbial genes linked to vitamin and amino acid metabolism, alongside decreased genes associated antibiotics biosynthesis (e.g., penicillin, cephalosporin, macrolides, and clavulanic acid) and secondary metabolite degradation (e.g., naphthalene and atrazine). These microbial alterations implied potential disparities in the health status between AD and wild individuals. AD populations exhibited varying degrees of microbial changes compared to the wild group, implying that the extent of these variations might serve as a metric for assessing the health status of AD populations. Furthermore, utilizing the individual information of captive individuals, we identified associations between variations in the gut microbiota of R. roxellana and host age, as well as pedigree. Older individuals exhibited higher microbial diversity, while a closer genetic relatedness reflected a more similar gut microbiota.

Conclusions: Our aim was to assess how anthropogenic activities and host factors influence the gut microbiota of R. roxellana. Anthropogenic activities led to consistent changes in gut microbial diversity and function, while host age and genetic relatedness contributed to interindividual variations in the gut microbiota. These findings may contribute to the establishment of health assessment standards and the optimization of breeding conditions for captive R. roxellana populations.

Keywords: Antibiotics; Captivity; Firmicutes/Bacteroidetes; Microbiome; Non-human primate; Pedigree.

Fig. 1

Sampling information. (a) Geographic distribution and sample sizes of the nine R. roxellana populations. The eight anthropogenically-disturbed (AD) populations included seven captive populations (NJZ, SHZ, SHWP, HZZ, CDZ, BJZ, and BJWP) and one wild population with artificial food provision (wild-fed). One wild population without anthropogenic activities (wild) is used as a control. (b) Age and sex structure of the hosts. No differences in ages were detected between animal sources or genders (p > 0.05, Scheirer–Ray–Hare test). (c) Pedigree relationship network of the hosts

Fig. 2

Comparison of the microbial diversity between different populations. (a − b) Bacterial composition at the phylum (a) and genus (b) levels. (c) Variation in microbial alpha-diversity (Shannon index) between groups. Different letters denote significant differences at a threshold of p < 0.05 (one-way ANOVA and S.N.K post-hoc test). (d) Heatmap illustrating the weighted UniFrac distances between samples from different populations. The color denotes the results of pairwise PERMANOVA on the beta-diversity (BH correction). The greater the intensity of red color, the more pronounced the statistical significance of the difference. (e − f) PCoA scatter plot (e) and area plot (f) showing the similarity in microbial composition (based on weighted UniFrac distances) between samples from different populations

Fig. 3

Differential analyses of gut microbial composition between captive and wild individuals. (a) Schematic map illustrating the workflow of differential analyses. Initial comparisons on gut microbiota were made between each AD population and the wild one population. This resulted in eight differential pools, with each differential microbe met the threshold of p < 0.05 (Mann-Whitney U test). Subsequently, differential microbes shared by at least six pools were considered consistently different in abundance between AD and wild individuals. (b) Upset plot displaying the numbers of microbes in the eight differential pools. (c − e) Humanized gut microbiota of captive populations at the phylum level. (c) Bar plot showing the proportions of the screened differential bacterial phyla. (d) Quantitative relationship between Firmicutes and Bacteroidetes across groups. (e) Ratios of Firmicutes to Bacteroidetes. Different letters denote significant difference between groups (p < 0.05, one-way ANOVA and S.N.K post-hoc test). (f) Heatmap depicting the variations in significant differential bacterial genera across groups. The average abundances of each bacterial genus were scaled to 0 − 1, where black and red colors represent 0 and 1, respectively. The colors of the row names denote the phyla

Fig. 4

Differential analyses of gut microbial function between captive and wild individuals. The analysis flow is the same as that used for microbial compositional differences. (a − b) Major differential KEGG items between captive and wild gut microbiota at hierarchical levels 1 (a) and 2 (b). Different letters denote significant difference between groups (p < 0.05, one-way ANOVA and S.N.K post-hoc test). (c) Heatmap presenting the main differential KEGG metabolic pathways at hierarchical level 3. The average abundances of each bacterial genus were scaled to 0 − 1, where black and red colors represent 0 and 1 respectively. Red and green colors of the row names denote higher and lower, respectively, in the wild gut microbiota. (d − e) Main differential pathways other than metabolism at level 3. Different letters denote significant difference between groups (p < 0.05, one-way ANOVA and S.N.K post-hoc test)

Fig. 5

Associations of host factors (gender, age, and pedigree relationship) with gut microbiota in captive populations. (a − b) PCoA scatter plots showing the effects of host gender on gut microbial beta-diversity. Both the single-factor (gender) and two-factor (gender & population) PERMANOVA models were constructed to assess the significance of host gender in shaping the gut microbiota. (c − d) PCoA scatter plots showing the effects of host age on gut microbial beta-diversity. Both the single-factor (age) and two-factor (age & population) PERMANOVA models were constructed to evaluate the significance of host age in shaping the gut microbiota. (e − f) Associations of genetic (e) and maternal (f) relatedness with gut microbiota in captive populations. The associations of genetic and maternal relatedness with gut microbial similarity were analyzed using Spearman correlations. Pairwise relationships with a genetic or maternal relatedness of 0 were excluded from the correlation analyses to avoid spurious associations caused by large inter-population dissimilarity in gut microbiota

Fig. 6

A schematic diagram summarizing the effects of anthropogenic activities and host factors on the gut microbiota of R. roxellana