We examined the distribution and ancestral relationships of genetic caste determination (GCD) in 46 populations of the seed-harvester ants Pogonomyrmex barbatus and P. rugosus across the east-to-west range of their distributions. Using a mtDNA sequence and two nuclear markers diagnostic for GCD, we distinguished three classes of population phenotypes: those with GCD, no evidence of GCD, and mixed (both GCD and non-GCD colonies present). The GCD phenotype was geographically widespread across the range of both morphospecies, occurring in 20 of 46 sampled populations. Molecular data suggest three reproductively isolated and cryptic lineages within each morphospecies, and no present hybridization. Mapping the GCD phenotype onto a mtDNA phylogeny indicates that GCD in P. rugosus was acquired from P. barbatus, suggesting that interspecific hybridization may not be the causal agent of GCD, but may simply provide an avenue for GCD to spread from one species (or subspecies) to another. We hypothesize that the origin of GCD involved a genetic mutation with a major effect on caste determination. This mutation generates genetic conflict and results in the partitioning and maintenance of distinct allele (or gene set) combinations that confer differences in developmental caste fate. The outcome is two dependent lineages within each population; inter-lineage matings produce workers, while intra-lineage matings produce reproductives. Both lineages are needed to produce a caste-functional colony, resulting in two reproductively isolated yet interdependent lineages. Pogonomyrmex populations composed of dependent lineages provide a unique opportunity to investigate genetic variation underlying phenotypic plasticity and its impact on the evolution of social structure.

The genus Pogonomyrmex is one of three ant genera with an effective mating frequency (me) > 2.0. We developed microsatellites to determine me for P. rugosus because mating frequency of P. rugosus was known only from observational data which do not allow an estimate of me. We genotyped 474 workers from 20 colonies for two microsatellite loci. Observed mating frequencies ranged from 3 to 12 and me for P. rugosus was 4.71. Observed patriline frequencies were significantly different from the expected patriline frequencies generated with a simulated data set under the assumption of equal patriline representation. The available mating frequency data and phylogenetic information of the genus Pogonomyrmex suggest that multiple mating is the ancestral state in the North American Pogonomyrmex sensu stricto. Established P. rugosus colonies raid and destroy smaller conspecific colonies. During these raids ant workers were observed carrying pupae and larvae from the raided colony into the nest of the raiding colony. However, it was not clear whether raided brood emerged in the raiding colony and were subsequently recruited into the work force (intraspecific slavery) or were used as food (predation). Our analyses indicate 6 of 14 field colonies contained foreign P. rugosus workers (43%). The range of the intracolonial frequency of foreign workers collected directly from the nest entrance was between 4 and 28%.

The interplay of genetic and environmental factors in the determination of social insect castes has long intrigued biologists. Though an overwhelming majority of studies establish that factors such as nutrition, pheromones and temperature determine the developmental fate of worker larvae, genetic components have recently been shown to play a role in the determination of morphological worker castes in leaf-cutting ants. Here we demonstrate that the determination of worker castes in the strongly polyandrous Florida harvester ant, Pogonomyrmex badius, has a genetic component. The overall distribution of caste members among patrilines in our study colonies is significantly different from the intracolonial caste ratio. Though this effect was not apparent in all colonies, our results suggest that workers of different patrilines in P. badius differ significantly in their propensities to develop into a certain worker caste. This genetic basis of worker polymorphism may go unnoticed in many social hymenopterans because of their low intracolonial genetic diversity due to monogamous colony structure. The worker polymorphism of P. badius is a taxonomic isolate and presumably a young trait in the genus. Therefore, a common genetic component of the determination of morphological and behavioral worker castes in social insects might be farranging taxonomically and may even be based on a genetic machinery inherent to all hymenopterans, but dormant in most.

No abstract

The North American seed-harvester ant Pogonomyrmex (Ephebomyrmex) pima displays a dimorphism that consists of winged (alate) and wingless (intermorph) queens; both types of queens are fully reproductive. Microsatellite allele frequencies and a mitochondrial phylogeny demonstrate (1) alate and intermorph queens represent an intraspecific wing polymorphism, and (2) an absence of assortative mating and inbreeding by males. Surveys at our field site in southcentral Arizona, USA, demonstrated that only one type of queen (intermorph or dealate) occurred in each colony, including those excavated during the season in which reproductive sexuals were present. Colony structure appeared to vary by queen type as most intermorph colonies contained multiple mated queens. Alternatively, dealate queen colonies rarely contained a mated queen. Our inability to find mated dealate queens in these colonies probably resulted from difficulty in excavating the entire colony and reproductive queen, especially given that these colonies were only excavated over one day.A morphometric analysis demonstrated that intermorph queens are intermediate in size to that of workers and alate queens, but that intermorph queens retain all of the specialized anatomical features of alate queens (except for wings). Some colonies had queens that foraged and performed nest maintenance activities, and these queens sometimes accounted for a significant portion of colony foraging trips. Dissections revealed that these queens were uninseminated; some of these queens produced males in the laboratory.