Geographical variation in signals and responses: individual identity signals linked with capacity for individual face learning across Polistes fuscatus wasp populations

Communication requires both signals and receiver responses, yet we know little about how signals and responses covary when recognition capacity differs across a species' geographical range. Previous work has shown that Polistes fuscatus wasps from Michigan, U.S.A. are capable of individual face recognition, while P. fuscatus from central Pennsylvania, U.S.A. are not. Here, we provide a broader assessment of intraspecific variation in signals and responses across the range of P. fuscatus. We quantify the amount of perceivable facial pattern variation in each population as a measure of individual identity signalling. We also measure how accurately wasps from each population learn to discriminate between a standard set of conspecific faces as a measure of individual face recognition. We show that wasps are equally adept at learning to discriminate faces of wasps from their own population and faces of wasps from different populations, confirming that discriminating a standard set of faces provides a comparable measure of individual face learning capacity. We find that there is striking variation in signals and responses across wasp populations and that individual identity signals and receiver responses are linked. Wasps from populations with more variable individual identity signals learn to discriminate between a standard set of conspecific faces more accurately than wasps from populations with less variable individual identity signals. Overall, we find surprisingly high levels of intraspecific geographical variation in individual identity signals and responses. Work in additional taxa will be important to assess whether signal phenotype and receiver responses are commonly linked in geographically variable communication systems.