Origin, behaviour, and genetics of reproductive workers in an invasive ant

Background Worker reproduction has an important influence on the social cohesion and efficiency of social insect colonies, but its role in the success of invasive ants has been neglected. We used observations of 233 captive colonies, laboratory experiments, and genetic analyses to investigate the conditions for worker reproduction in the invasive Anoplolepis gracilipes (yellow crazy ant) and its potential cost on interspecific defence. We determined the prevalence of worker production of males and whether it is triggered by queen absence; whether physogastric workers with enlarged abdomens are more likely to be reproductive, how normal workers and physogastric workers compare in their contributions to foraging and defence; and whether worker-produced males and males that could have been queen- or worker-produced differ in their size and heterozygosity. Results Sixty-six of our 233 captive colonies produced males, and in 25 of these, some males could only have been produced by workers. Colonies with more workers were more likely to produce males, especially for queenless colonies. The average number of days between the first appearance of eggs and adult males in our colonies was 54.1 ± 10.2 (mean ± SD, n = 20). In our laboratory experiment, queen removal triggered an increase in the proportion of physogastric workers. Physogastric workers were more likely to have yolky oocytes (37–54.9%) than normal workers (2–25.6%), which is an indicator of fertile or trophic egg production. Physogastric workers were less aggressive during interspecific aggression tests and foraged less than normal workers. The head width and wing length of worker-produced males were on average 4.0 and 4.3% greater respectively than those of males of undetermined source. Our microsatellite DNA analyses indicate that 5.5% of worker-produced males and 14.3% of males of undetermined source were heterozygous, which suggests the presence of diploid males and/or genetic mosaics in A. gracilipes. Conclusions Our experimental work provides crucial information on worker reproduction in A. gracilipes and its potential cost to colony defence. The ability of A. gracilipes workers to produce males in the absence of queens may also contribute to its success as an invasive species if intranidal mating can take place between virgin queens and worker-produced males.