Fig. 3.

Scatter plot showing distribution of hormone content in relation to age. There is a significant positive association between JH content and age in age-typical colonies (A), but no such correlation exists for single-cohort colonies (B). Similarly, a significant negative correlation is observed between ecdysteroid content and age in age-typical colonies (C), but not in single-cohort colonies (D). This relationship illustrates that the association between hormone content and behavior is not necessarily age associated, as the relationship disappears in single-cohort colonies. Asterisks denote significant correlations, and Spearman coefficients (ρ) are indicated (Spearman rank correlation, P<0.05).