Collective behavior in slave-making ants: how ecology and social structure shape raiding strategies

Abstract

In this dissertation, I studied the raiding behavior of a slave-making ant, Temnothorax americanus, a social parasite whose workforce consists of different ant species. ‘Slave’ workers are acquired in coordinated raids during which slave-maker workers attack a host colony and steal the immatures (brood). I focused on the mechanisms colonies use to reach collective decisions over where raid and on the attack once a raiding site was chosen. In Chapter One, I first established which nest characteristics make an optimal raiding choice by measuring the benefit of raids (stolen brood), and counterbalancing costs (mortality), at target colonies that differed in the number of brood, workers, or their ratios. These experiments showed that slave-makers are more successful when attacking colonies where brood outnumber workers and where there is an intermediate number of workers. In Chapter Two, I tested whether slave-maker colonies demonstrate a preference for such nests and characterize their decision-making strategy. Choice trials showed that slave-maker colonies exhibit no preference over host colony features. This result led to the question of why not be more selective? I then tested for ecological conditions that could favor their low acceptance threshold and show that slave-maker colonies encounter host colonies at a very low rate relative to the time when brood is available to raid. Slave-maker colonies therefore ought to raid every nest they find rather than pay the opportunity cost of waiting to raid only the best host colonies. In Chapter Three, I investigated the mechanisms of attack and show that successful raiding parties effectively evacuate workers while guarding the door to keep the brood inside. I also tested how conflict, especially prominent in T. americanus colonies, affects this collective behavior. I found that workers from colonies with higher levels of worker reproduction initiate attacks alone rather than as a group. In Chapter Four, I explored how cooperation and collective behaviors co-evolve by modeling these interactions analytically. I adapted mathematical models of cooperative binding from biochemistry to characterize collective behaviors that arise in a non-linear fashion.