Represents the results from bivariate regression analysis.

Significant correlations between 14 variables and the top 100 OTUs are depicted after correcting for the correlation between season and OUT composition. The x-axis depicts the respective OTUs and the y–axis the 15 variables tested. The heatmap shows significant correlations (p-value less than 0.05) from univariate analysis. Blue squares show positive changes in relative abundance, whereas red squares show negative correlations. The intensity of colour correlates with the magnitude of the (log) fold change value (see colour key).