Activity budget and gut microbiota stability and flexibility across reproductive states in wild capuchin monkeys in a seasonal tropical dry forest

Abstract

Background: Energy demands associated with pregnancy and lactation are significant forces in mammalian evolution. To mitigate increased energy costs associated with reproduction, female mammals have evolved behavioural and physiological responses. Some species alter activity to conserve energy during pregnancy and lactation, while others experience changes in metabolism and fat deposition. Restructuring of gut microbiota with shifting reproductive states may also help females increase the energy gained from foods, especially during pregnancy. The goal of this study was to examine the relationships among behaviour, gut microbiota composition, and reproductive state in a wild, non-human primate to better understand reproductive ecology. We combined life history data with > 13,000 behavioural scans and 298 fecal samples collected longitudinally across multiple years from 33 white-faced capuchin monkey (Cebus imitator) females. We sequenced the V4 region of the 16S rRNA gene and used the DADA2 pipeline to analyze microbial diversity. We used PICRUSt2 to assess putative functions.

Results: Reproductive state explained some variation in activity, but overall resting behaviours were relatively stable across pregnancy and lactation. Foraging was less frequent among females in the early stage of nursing compared to the cycling stage, though otherwise remained at comparable levels. Maximum temperature was a strong, significantly positive predictor of resting, while social dominance had a small but significantly negative effect on resting. Ecological variables such as available fruit biomass and rainfall had a small but significantly positive effects on measures of foraging time. Gut microbial community structure, including richness, alpha diversity, and beta diversity remained stable across the reproductive cycle. In pairwise comparisons, pregnant females exhibited increased relative abundances of multiple microbial ASVs, suggesting small changes in relation to reproductive state. Reproductive state was not linked to differential abundance of putative metabolic pathways.

Conclusions: Previous data suggest that activity budget and the gut microbiome shifts considerably during reproduction. The present study finds that both activity and gut microbial communities are less associated with reproduction compared to other predictors, including ecological contexts. This suggests that behavioural flexibility and gut microbial community plasticity is contrained by ecological factors in this population. These data contribute to a broader understanding of plasticity and stability in response to physiological shifts associated with mammalian reproduction.

Keywords: Animal behaviour; Gut microbiome; Non-human primates; Reproductive ecology.

Fig. 1

Visualization of estimated increases in energy requirements during the reproductive cycle of a non-human primate. Female primates face a ca. 25% increase in daily energy requirements during gestation, and up to a 50–100% increase during lactation [35]

Fig. 2

Reproductive states of study individuals observed between April 1, 2014 and June 30, 2018. Fecal samples collection took place between April 29, 2014 and September 27, 2016. Behavioural data collection took place between April 20, 2016 and June 22, 2018. One individual (SF) disappeared from the population in early 2017. Two individuals (NP and VN) reached reproductive maturity during the study, but were never observed to be pregnant. At 15 time points, females were observed to be pregnant via protrusion of the abdomen, but were subsequently observed with no protrusion. In these cases the orange (pregnancy) segments are followed by green (cycling) and not by purple (nursing) segments. Nursing segments that are shorter than 12 months represent cases where infants died. We did not include samples from individuals who experienced potential pregnancy loss in our analysis

Fig. 3

Proportions of daily scans spent in each behavioural category across the reproductive cycle. Daily scans from all individuals were summed for each reproductive state, then divided by total scans per day in that state to determine relative frequency of each behavioural category. We collected 13,721 individual scans over the course of 222 contact days from 33 adult female capuchins in four social groups

Fig. 4

Incidence rate ratios and standard error (A) for predictors from a GLMM of resting scans per day. The reference dominance category is low social rank; the reference reproductive stage is cycling (pre-conception). The grey vertical line represents “no effect”. Values to the right of the grey line represent positive effects and values to the left represent negative effects. Significant predictors (p < 0.05) are denoted with asterisks. We plotted the predicted number of resting scans per day for each level of reproductive stage variable (B)

Fig. 5

Incidence rate ratios and standard error (A) for predictors from GLMM of foraging scans per day. The reference dominance category is low social rank; the reference reproductive stage is cycling (pre-conception). The grey vertical line represents “no effect”. Values to the right of the grey line represent positive effects and values to the left represent negative effects. Significant predictors (p < 0.05) are denoted with asterisks. We plotted the predicted number of resting scans per day for each level of reproductive stage variable (B)

Fig. 6

For each sample, Bray–Curtis dissimilarity values were computed and ordinated using non-metric multidimensional scaling (NMDS) (A). Reproductive status was not a significant predictor of dissimilarity. Relative abundance of phyla were visualized across reproductive statuses (B). Phyla with relative abundances below 0.01 were grouped in the category “ < 1% abund”