Fig. 2

Parity affects the gut microbiota trajectory during gestation. Clustering each of the 34 individuals by parity (zero = no previous pregnancies, low = 1–3 previous pregnancies, high = 4–7 previous pregnancies) reveals that parity explains a significant amount of the variation between gut microbiota trajectories during gestation. A Non-parametric microbial interdependence test (NMIT), which calculates correlations between each pair of taxa for each individual over time, was performed for each individual. A principal coordinate analysis (PCoA) plot of the NMIT data shows how the gut microbiota trajectory differs between animals of different parities across the first two axes. B Non-metric multidimensional scaling (NMDS) plot of the NMIT data depicts the differences between the gut microbiota trajectories, with 95% confidence intervals displayed as ellipses around each parity bin. C Boxplots depict the differences in NMIT between each individual and each high parity individual. Boxplots show the median and the first and third quartiles, with whiskers that extend to outliers up to 1.5 times the interquartile range. The adjusted P values and R² are shown for each comparison. D Bray-Curtis beta diversity was calculated between each sample and the day 37 sample from the same individual. E Bray-Curtis beta diversity was calculated between each sample and the previous week’s sample from the same individual. F Dirichlet multinomial mixtures (DMM) showing only samples collected from high parity animals. G DMM showing only samples collected from zero parity animals