FIG 2.

Carbohydrate-degrading capabilities in the gut microbiome of gorillas across two seasons of variable dietary intake and in humans under two different subsistence strategies. (a) Principal-coordinate analysis using Bray-Curtis distances generated from the CAZyome shows distinctions in carbohydrate-degrading capabilities in the microbiome of gorillas across dry and wet seasons and between BaAka hunter-gatherers and Bantu agriculturalists (PERMANOVA: R² = 0.20, P = 0.01**). The amplified ordination panel on the left specifically shows functional distinctions between gorillas across seasons (PERMANOVA: R² = 0.13, P = 0.01**). (b) PCo2 ordination score reflects functional similarities of BaAka hunter-gatherers with gorillas in both dry and wet seasons. (c) Higher interindividual variability in CAZyme content was observed in Bantu agriculturalists and gorillas during the wet season. (d) Heatmap of CAZy families and subfamilies shows discriminant patterns among groups and subgroups. Two-sided Wilcoxon rank sum test was applied for each pair (dry versus wet season and BaAka versus Bantu) (false-discovery rate [FDR]-corrected P < 0.05). Color code of each CAZyme shows the broad carbohydrate utilization capabilities. The heatmap is color coded based on the row normalized z-scores. (e) Distinctions in broad CAZyme categories among human and gorilla groups and subgroups, plotted by their respective distributions. The color key in panel a applies to all panels. A nonparametric two-sided Wilcoxon rank sum test was used for testing the box plot distributions. The center values indicate the medians, and error bars depict the SD. ns, not significant; *, P < 0.05; **, P < 0.01.