ANOVA-style results of linear mixed effects models testing for differences in the proportion of vacant sites with colonizations per year.

Random effects were: Year (Intercept) variance = 0.004, standard deviation = 0.066, from 227 observations in 6 groups (years). Effect size is given as the proportion of variance explained by explanatory variables, partial eta-squared (h2) = (SSeffect)/(SSeffect+SSerror).