For statistical analysis, we used a generalized linear mixed model using conditioning and time as main effects and colony as a random effect. We estimated least-squares means and compared between treatments for each timepoint using bonferroni adjusted p-values