Figure 1.
Correlation of microbiome composition between replicates, individuals, and over time.
Boxplots of pairwise Spearman correlation (y-axis) of bacterial abundance is shown for data at each taxonomic level (x-axis) across different classification levels. ‘Replicates’ refer to technical replicates derived from separate DNA extractions and library preparations from the same sample.
Figure 2.
Consistent temporal shifts observed in microbiome composition.
Principal components analysis (PCA) of data obtained from samples collected in the summer (red triangle) and winter (blue circle), at the phylum (A), class (B), order (C), family (D), and genus (E) level across samples. F) Individual shifts in microbiome composition along PC1 (y-axis). Data from the same individual across seasons is connected by a line (gray lines represent individuals showing seasonal trend, black lines highlight individuals showing the opposite trend). G) The average pairwise distance along the PC1/PC2 plane is shorter where points are from the same season than points between seasons (t-test P = 3.38×10−78), supporting seasonal clustering along the first two principle components.
Figure 3.
Phylum level taxa abundances differ by season.
A) Abundance (y-axis) of the 25 bacterial phyla that were detected (only the most common bacteria are indicated in the legend) by individual (x-axis). Individual's order along the x-axis is identical in both panels. Phyla that are significantly differentially abundant between seasons (FDR <0.05) are indicated by an asterisk. B) Examples of three common phyla whose abundances (y-axis) are significantly different across seasons (x-axis): Firmicutes (q-value<0.002), Bacteroidetes (q-value<0.002), and Actinobacteria (q-value<0.002).
Figure 4.
Diversity between seasons, sexes, and age.
A) A significant difference (paired t-test P<0.002) in Shannon diversity (H′, y-axis), between season (x-axis). B) H′ between sexes is not significantly different in either winter (t-test P = 0.46) or summer (P = 0.38). C) Diversity significantly decreases with age based on the data collected from winter samples (r2 = 0.1267, P<0.006), but not based on the data collected from the summer samples (r2 = 0.01653, P = 0.33).
Figure 5.
Produce consumption varies by season.
In summer, when fruits and vegetables are grown in colony gardens, consumption of fresh produce is higher than in winter. Conversely, when fresh produce is not as available, a higher proportion of canned or frozen fruits and vegetables are eaten. Bars indicate standard error measurements. *** P<0.001 (t-test, adjusted for multiple tests using Benjamin Hochberg correction). See Table S6 for a list of all tests performed.