Abstract

Mechanisms selecting for the evolution of cooperative breeding are hotly debated. While kin selection theory has been the central paradigm to explain the seemingly altruistic behaviour of non-reproducing helpers, it is increasingly recognized that direct fitness benefits may be highly relevant. The group augmentation hypothesis proposes that alloparental care may evolve to enhance group size when larger groups yield increased survival and/or reproductive success. However, there is a lack of empirical tests. Here we use the cooperatively breeding cichlid fish Neolamprologus pulcher, in which group size predicts survival and group stability, to test this hypothesis experimentally by prompting two cooperative tasks: defence against an egg predator (left video) and digging out sand from the breeding shelter (right video). We controlled for alternative mechanisms such as kin selection, load lightening and coercion. As predicted by the group augmentation hypothesis, helpers increased defence against an egg predator in small compared with large groups. This difference was only evident in large helpers owing to size-specific task specialization. Furthermore, helpers showed more digging effort in the breeding chamber compared with alternative personal shelters, indicating that digging is an altruistic service to the dominant breeders.