Abstract
Three alternative hypotheses have been advanced to explain the dynamics of individually produced nestmate-recognition cues in colonies of social insects: ( i ) that there is no effective transfer of cues among nestmates (individual hypothesis); ( ii ) that cues are shared reciprocally among nestmates (collective hypothesis); and ( iii ) that cues derived from the queen are transferred to all colony members and dominate all other cues in nestmate recognition (queen hypothesis). In the present study, a bioassay based on aggression by laboratory colonies toward workers introduced into their nests was used in conjunction with isolation and interspecific-adoption experiments to test these hypotheses for colonies of two closely related, polygynous, and polydomous ant species, Leptothorax ambiguus and Leptothorax longispinosus . The results provide strong evidence for the collective hypothesis. A collective system has long been postulated as one of the primary modes of nestmate discrimination among social insects but to my knowledge has never before been clearly demonstrated.