An ideal species to use for examination of the evolution of cooperation among unrelated individuals has recently been found - a social insect that occurs in discrete populations, one of which contains cooperative and the other non-cooperative, aggressive individuals. The Principal Investigator plans to exploit the existence of these two social morphs to perform the first experiments directly testing many of the fundamental precepts of a group-selection hypothesis of the evolution od cooperation. The behavioral experiments outlined in this proposal also provide the first empirical investigation of whether "group level punishment" of cheaters exist. In addition, the P.I. will attempt a molecular genetic analysis of cooperative and non-cooperative morphs to examine whether one morph is "invading the other. Given the crucial role social insects have played in the development of kin selection theory and its eventual transfer to studies of primate and other social systems, insights generated from this study testing group selection in the evolution of social insect behavior can be expected to be transferable to other social systems.