Factors affecting brood cannibalism in laboratory colonies of the imported fire ant, Solenopsis invicta Buren (Hymenoptera: Formicidae)

We examined the effects of nutrition, access to brood by the workers, starvation, and use of${\rm CO}_{2}$anesthesia on brood cannibalism in Solenopsis invicta Buren. To measure these effects, worker ants were subjected to different feeding regimes and placed with larvae and queens. Larvae were either left on the nest floor or housed in a brood chamber which restricted worker access. Workers were either anesthetized with${\rm CO}_{2}$to facilitate counting of larvae, or larvae were counted in the absence of${\rm CO}_{2}$by partitioning the nest. Larvae and pupae were counted daily for 5-10 days. Food (insects, egg yolk, honey, or soybean oil) affected both the number of larvae missing and the percent that pupated. Brood cannibalism was highest in colonies fed oil and egg yolk and significantly less in colonies fed insects and honey. Pupation was highest in colonies fed insects. Brood cannibalism was greater when larvae were confined inside the brood chamber than when they were kept on the nest floor. In contrast, rates of pupation and survivorship were less when the larvae were exposed and in piles on the nest floor. Cannibalism increased and the rate of pupation decreased when the workers were starved or when${\rm CO}_{2}$was used to facilitate counting of the brood. Insect-fed colonies in which the brood was housed in a chamber and counted daily without the use of${\rm CO}_{2}$had a brood cannibalism rate of 8-13%.