The relation between caste ratios and division of labor in the ant genus Pheidole (Hymenoptera: Formicidae)

Summary Ten species of Pheidole, representing as many species groups from various localities in North and South America, Asia, and Africa, were analyzed to probe for possible relationships between caste ratios and division of labor.Minor workers are behaviorally almost uniform among the species, but major workers vary in repertory from 4 to 19 behavioral acts (Table 1, Fig. 2). The major repertory size increases significantly across the species with the percentage of majors in the worker force (Fig. 3). This trend is consistent with the basic prediction of ergonomic optimization models under an assumption of colony-level selection. There is also a trend toward reduction of behavioral repertory with increase of size in the major relative to the minor, a second relation expected from theory, but the data are not sufficient to reach statistical significance.When the minor:major ratio was lowered to below 1:1 (from the usual 3:1 to 20:1, according to species), in three widely different species (guilelmimuelleri, megacephala, pubiventris), the repertory size increased by 1.4–4.5X and the rate of activity by 15–30X (Table 1, Figs. 4–6). The change occurred within 1 h of the ratio change and was reversed in comparably short time when the original ratio was restored.This abrupt and important shift in behavior permitted the major workers to serve as an emergency stand-by caste, available to be summoned to a nearly full repertory when the minor worker caste was depleted. The majors also restored 75% or more of the missing minor workers' activity rate under laboratory conditions. Their transformation allowed continued oviposition by the queen and the rearing of larvae to the adult stage.In line with these findings, a distinction is made between programmed elasticity in the repertory of individual workers and castes and the resiliency of the colony as a whole, which depends upon the pattern of caste-specific elasticity.