Julie Miller: University of California, Los Angeles

The slave-making ants are social parasites that steal their workers in coordinated raids on other colonies. Their parasitic lifestyle has made these ants a model of host-parasite co-evolution, however their charismatic brood-raiding behavior is virtually unstudied. Here I explore how colonies make collective decisions when selecting a raiding site and how colonies coordinate their attack. Slave-making has independently evolved multiple times in ants, but I focused this study on one species of North American slave-maker, Temnothorax americanus. I staged raids in table-top arenas in the lab to standardize conditions and made detailed observations of individual and colony-level behaviors. I found that the success of group raids depends on the ability of slave-makers to coordinate the timing of two roles: herding and guarding the entrance. Having established that raiding is a coordination problem, I then investigated how the colony selects a raiding site in the first place. I first measured colony preferences using choice experiments, but colonies demonstrated no preference for any size-related host features. This apparent lack of preference led to a separate question: why do colonies disregard fitness-relevant host variation? Theoretical work has offered suggestions about which conditions ought to favor low-choosiness, so I empirically tested whether they are met by T. americanus colonies, specifically testing the hypotheses that slave-makers experience (1) low host encounter rates, (2) high time constrains, or (3) low variability in host quality. To test these hypotheses, observations of lab raids were combined with spatial field data on host distribution and brood phenology. These data support that raiding is constrained by both the brief window of host brood availability, particularly of the highly valuable pupae, combined with low encounter rates of host colonies in the field. Variation in host nest quality was relatively high, and thus unlikely to favor low acceptance thresholds of slave-maker colonies. The implications of these results are that slave-maker colony raiding decisions are selected to maximize the number of raids per season, and not to selectively exploit the few most profitable ones.