High phenotypic diversity correlated with genomic variation across the European Batrachochytrium salamandrivorans epizootic
Fig 5
Correlation between genomic and phenotypic variation.
A) Principal component analyses (PCA), performed using prcomp() in R, with first four principle components (PCp) explaining 66.5% of variation of phenotypic data (spore counts, sporangia counts and mean sporangia size) across all temperatures and time points. Colour of isolate label and ellipses depict cluster identity according to K-mean clustering analysis. B) Estimates with 95% confidence intervals of coefficients from linear model (model structure PCp1*PCp2*PCp3* PCp4 ~ PCg1*PCg2*PCg3, R2 = 0.827) where PCg = genomic principle component and PCp = phenotypic principal component. C) Genomic PCA output. Pairwise scatterplots of first three principal components with isolates colour indicating isolate group as shown in legend positioned in D.