Coevolution between brood-parasitic honeyguides and their hosts

Abstract Obligate brood parasites lay their eggs in the nests of other species, foisting the costs of parental care onto the host parents. The success of the parasite and host are then at odds, with both parties evolving defences and countermeasures in an evolutionary arms race. This reciprocal influence of acting upon both species’ evolution — a process known as coevolution — has forged the natural world around us. Avian brood parasites and their hosts are now model systems for studying such coevolutionary interactions between species, providing striking examples of the adaptations that arise when the life histories of two species become entangled. In this thesis I highlight the adaptations that have arisen in response to coevolutionary selection pressures in a group of understudied avian brood parasites, the honeyguides (Indicatoridae), and their hosts. This study focuses on the greater (Indicator indicator) and lesser (I. minor) honeyguides, and their respective primary hosts, the little bee-eater (Merops pusillus) and the black-collared barbet (Lybius torquatus). The interactions between honeyguides and their cavity-nesting hosts of the Old World tropics are evolutionarily ancient, contrasting with the majority of studies of avian brood parasitism which have predominantly focused on temperate brood-parasitic species targeting open cup-nesting hosts. Therefore, honeyguides and their hosts are an ideal study system in which look for novel adaptations that have not evolved in other systems. Using field observations and experimental manipulations at each stage of the parasitic life cycle — before parasitism, during egg-laying, during incubation, and during chickrearing — I examine how honeyguides and their hosts have evolved in response to the selection pressures they exert on each other. First, in chapter two, I consider whether the nest structure of the little bee-eater – host to the greater honeyguide – can act as a defence against brood parasitism. Experimental manipulation of the size of bee-eater nest tunnels demonstrates that bee-eaters with narrower nest tunnels are less likely to be parasitized by greater honeyguides than those bee-eaters whose nests have wider nest tunnel entrances. This study provides the first experimental evidence of a host nest functioning as a frontline defence against brood parasitism. In chapter three, I take a comparative approach and use a phylogenetic framework to investigate, across multiple avian brood parasite species, the evolutionary drivers of rapid egglaying. This trait is shared by most brood-parasitic birds, but not by non-parasitic birds. I find strong evidence that the egg-laying speed of avian brood parasites is ecologically and physiologically constrained, but find no evidence that variation in the costs incurred during ii parasitism events have driven variation in the rapidity of egg-laying among brood-parasitic species. In chapter four, I examine whether there are costs associated with the virulent egg puncturing behaviour of greater honeyguides, and whether honeyguides can adjust their level of virulence in accordance with these costs. I find strong support for the idea that virulence is costly to honeyguides, as bee-eater hosts are more likely to reject clutches that contained eggs punctured by honeyguides. Such punctured clutches are also more likely to be predated. Honeyguides appear to adjust how much they puncture host eggs in accordance with the severity of these costs, providing the first evidence of an avian brood parasite moderating its virulence in response to the associated costs. In chapter five, I examine egg rejection behaviour in the black-collared barbet, a common host of the lesser honeyguide. I consider whether the (smaller) size of a parasitic egg could be used as a cue for egg rejection inside the dark environment of a cavity nest. Through observations of natural parasitism events, and experimental parasitism of host nests using different sized eggs, I demonstrate that barbets are more likely to reject a clutch of eggs when they detect a small egg within the nest. This seems to be achieved through a process of true recognition, a mechanism that involves a specific innate or learnt template of what size eggs a host should reject. Barbets do not appear to rely on discordancy – comparing all eggs within their clutch in order to reject the odd one out – in order to make rejection decisions. Finally, in chapter six I explore whether honeyguides elicit additional provisioning from their foster parents by using vocal mimicry, and investigate why such extra food would be required. I demonstrate that both greater and lesser honeyguides mimic the sound of a brood of chicks of their respective hosts in order to receive higher levels of provisioning from their foster parents. I establish that greater and lesser honeyguides do this for contrasting reasons. Greater honeyguides require higher levels of provisioning to support their fast growth rate to a size much larger than their host siblings, whereas lesser honeyguides require more food in order to offset a sub-optimal diet provided to them by their foster parents.