Habitat dissimilarity and GuniFrac distances between the communities were not correlated (Mantel test: ntrials = 15, ngroups = 6, r = ? 0.149, p = 0.553; late dry 2016: nsamples = 15, ngroups = 6, r = 0.008, p = 0.972; early dry 2017: nsamples = 21, ngroups = 7, r = ? 0.154, p = 0.561; late dry 2017: nsamples = 21, ngroups = 7, r = 0.064, p = 0.776; Table S8). The model examining the effects of habitat overlap and diet dissimilarities on groups’ GuniFrac distances was also not significant (LMM II: ? 2 = 3.264, df = 2, p = 0.196, R 2 m/c = 0.08/0.98) (Table S9).
The new 18S rRNA gene study of your home vegetation found in faecal samples revealed that at least within all the way down taxonomic account, i.age. up until the relatives top, diet plan don’t seem to apply at between-classification variation when you look at the microbiome constitution. Even after obvious ranging from-group type for the dining plant configurations, groups’ microbial microbiome compositions don’t echo this type of variations whenever aesthetically examining the new respective graphs (Fig. 2A, B). We receive, however, regular losing weight designs. In early inactive 12 months both in studies ages, faecal examples consisted of the vast majority off vegetation from the parents Combretaceae and you will Salicaceae, while inside the late dead 12 months Fabaceae and you may Sapindaceae had been ate from inside the greater numbers (Fig. 2B).
We examined the effects of maternal relatedness coefficients on GuniFrac distances among all individuals, i.e. between both, group members and individuals from different groups. The interaction between the relatedness coefficient and group membership (same or different) was not significant (likelihood ratio test comparing the model with and without the interaction: ? 2 = 0.105, df = 1, p = 0.746), which is why we excluded it from the model. The model without the interaction was highly significant (LMM III:? 2 = , df = 1, p < 0.001, R 2 m/c = 0.51/0.92) (Table S10). Maternal relatives had a more similar microbiome than unrelated individuals, and this effect was independent of whether these relatives lived in the same group or not (Fig. 3).
GuniFrac distances of all analysis animals when considering the maternal relatedness coefficient and you will class membership. An enthusiastic Rc away from 0.25–0.fifty means dyads whereby we can’t see whether they is actually complete- or 1 / 2 of-sisters
The model examining correlations of dyadic GuniFrac dissimilarity with seasonality, sex, age classes, and the time two group members spent affiliating was significant (LMM IV: ? 2 = , df = 10, p < 0.001, R 2 m/c = 0.70/0.91) (Tables S11). Bacterial microbiomes of group members increased in similarity across the study period; they were least similar in the early and late dry season 2016 and most similar in the late dry season 2017. Samples of adults differed most from tendermeets desktop each other, whereas samples among juveniles and infants were more similar (Fig. 4A). Neither sex nor time spent affiliating significantly affected microbiome similarity.
Differences in gut similarity and association networks within groups per age category, female reproductive state, and male dominance. A, C GuniFrac distances between group members of different or same age categories or rank categories of adult group members only. As there is only one dominant male per group, we could not compare two dominant individuals. We did not have enough adult female group members to compare their GuniFrac distances during different reproductive stages. B, D, E ASVs associated with the different age categories, adult female reproductive stages, or rank categories within groups, respectively. The association network was calculated and visualised in the same way as described in Fig. 1. The network for age categories only contains data from the late dry seasons since animals were only considered infants, when they were < 9 months of age. Hence, during the early dry seasons, there were no infants in the population