Fig. 2

Genome size correlations with repetitive DNA content, ecological variables and genome assembly quality metrics. a. Repetitive DNA as a function of genome size. Top left: The number of TEs ("# TEs"). Top right: The percentage of the genome being TEs ("% TEs"). Bottom left: The number of STRs ("# STRs"). Bottom right: The percentage of the genome being STRs ("% STRs"). The blue line and the shaded area indicate a linear regression with the 95% confidence interval, which does not take the correlation between genetic relatedness and the residuals into account. The gray line indicates the local regression. The box above each plot indicates the adjusted R² and the P-value of the fit in PGLS regression. b. PGLS regression model with genome assembly size as the response, showing the estimates of the explanatory power of repetitive DNA content, ecological variables and genome assembly quality metrics. The variables were standardized by subtracting the means and divided by the standard deviation. Asterisks indicate that the explanatory variables had significant contributions at the 1% alpha level (**) or 0.1% alpha level (***). See Supplementary Table 2 for model estimates without normalization