Multi-level selection for hygienic behaviour in honeybees.

Disease is one of the main factors driving both natural and artificial selection. It is a particularly important and increasing threat to the managed honeybee colonies, which are vital in crop pollination. Artificial selection for disease-resistant honeybee genotypes has previously only been carried out at the colony-level, that is, by using queens or males reared from colonies that show resistance. However, honeybee queens mate with many males and so each colony consists of multiple patrilines that will vary in heritable traits, such as disease resistance. Here, we investigate whether response to artificial selection for a key resistance mechanism, hygienic behaviour, can be improved using multi-level selection, that is, by selecting not only among colonies as normal but also among patrilines within colonies. Highly hygienic colonies were identified (between-colony selection), and the specific patrilines within them responsible for most hygienic behaviour were determined using observation hives. Queens reared from these hygienic patrilines (within-colony selection) were identified using DNA microsatellite analysis of a wing-tip tissue sample and then mated to drones from a third highly hygienic colony. The resulting colonies headed by queens from hygienic patrilines showed approximately double the level of hygienic behaviour of colonies headed by sister queens from non-hygienic patrilines. The results show that multi-level selection can significantly improve the success of honeybee breeding programs.

Nontransitivity of paternity in a bird.

In many animals reproductive success is determined after insemination by the interaction of male and female processes. While sperm competition is reasonably well understood in some taxa, other processes, such as cryptic female choice and differential early embryo mortality resulting from genetic incompatibilities, are less well understood. The relative importance of these different factors contributing to reproductive success is difficult to assess. Here we control for male-mediated effects (which are often considerable and can mask more subtle processes) through the artificial insemination of known numbers of sperm in the domestic fowl to reveal that male reproductive success is nontransitive across females: the success of a particular male depends on the background against which his sperm compete for fertilization. Two potential processes could account for this effect: cryptic female choice (sperm choice) or differential early embryo mortality. Regardless of the mechanism, nontransitivity of male reproductive success has important evolutionary consequences, including the maintenance of variation in male fitness.

Anarchy in the UK: Detailed genetic analysis of worker reproduction in a naturally occurring British anarchistic honeybee, Apis mellifera, colony using DNA microsatellites.

Anarchistic behaviour is a very rare phenotype of honeybee colonies. In an anarchistic colony, many workers' sons are reared in the presence of the queen. Anarchy has previously been described in only two Australian colonies. Here we report on a first detailed genetic analysis of a British anarchistic colony. Male pupae were present in great abundance above the queen excluder, which was clearly indicative of extensive worker reproduction and is the hallmark of anarchy. Seventeen microsatellite loci were used to analyse these male pupae, allowing us to address whether all the males were indeed workers' sons, and how many worker patrilines and individual workers produced them. In the sample, 95 of 96 of the males were definitely workers' sons. Given that approximately 1% of workers' sons were genetically indistinguishable from queen's sons, this suggests that workers do not move any queen-laid eggs between the part of the colony where the queen is present to the area above the queen excluder which the queen cannot enter. The colony had 16 patrilines, with an effective number of patrilines of 9.85. The 75 males that could be assigned with certainty to a patriline came from 7 patrilines, with an effective number of 4.21. They were the offspring of at least 19 workers. This is in contrast to the two previously studied Australian naturally occurring anarchist colonies, in which most of the workers' sons were offspring of one patriline. The high number of patrilines producing males leads to a low mean relatedness between laying workers and males of the colony. We discuss the importance of studying such colonies in the understanding of worker policing and its evolution.