A Test of Optimal Caste Ratio Theory Using the Ant Camponotus (Colobopsis) Impressus

Oster and Wilson developed theoretical models predicting optimal caste ratios for social insects. We have tested some of these predictions in the field using an ant species that fits the relevant assumptions of their models. The ant Camponotus (Colobopsis) impressus has two worker castes: major workers (=soliders), which guard the nests and may also store food, and minor workers (=workers), which are responsible for foraging. For such a species, Oster and Wilson predicted a concave relationship between the number of soldiers and the number of workers in the breeding season, and a convex relationship in the nonbreeding season. These predictions were supported by field censuses of C. impressus colonies. Also, increased proportions of soldiers (number of soldiers per worker) were found in the breeding versus the nonbreeding season, as predicted by the models. However, although the models predicted that breeding caste ratios should be optimal from the point of view of reproductive success, in C. impressus the production of alates was positively correlated with the number of soldiers, implying that the population had not reached the optimal caste ratio functions predicted by theory. Data on colony growth patterns and ecological variables suggested that the lack of a complete fit between theory and results was due to several oversimplified assumptions about the proximal mechanisms of caste ratio regulation. C. impressus actively altered caste ratios on a seasonal basis and on an ontogenetic basis, but there was no evidence of any adjustments to annual or spatial variations in resources, competitors or predators. Moreover, laboratory experiments in which different proportions of soldiers were removed revealed that the ability of C. impressus to regulate caste ratios was severely constrained by a low, constant production ratio of new soldiers to new workers. The existence of physiological and temporal constraints to caste ratio adjustments should be taken into account in studies of caste ratios for social insects.