Introduction
Male ants are ephemeral in the annual cycle of a colony. They have been likened to ‘sperm missiles’ (a statement attributed to E. O. Wilson) – a transient, but essential function – delivering the male gametes of the superorganism to gynes so that the gynes can become queens and found new colonies, beginning the colonial life cycle anew (Hölldobler et al. 1990). Because of this males are often unseen and unsampled. Males are also haploid, typically developing from unfertilized eggs (males can be fathers, but have no father). Males are thus genetically and behaviorally very very different from the standard (female) ant, and their morphology is also clearly distinct as they typically have small heads with big eyes, a large muscly thorax with wings, and an abdomen tipped with an intromittent organ rather than a stinger. They are very clearly different than females; so different to look a different species (Fig. 1). And so it is clear why males are largely ignored in the empirical and theoretical literature on caste and polyphenism in the social insects – they are different.
Recent studies on gene expression during development have revealed that queen-worker caste determination shares many genes in common with male-female sex determination (Klein et al. 2016, Johnson and Jasper 2016, McAfee et al. 2019, Warner et al. 2019, Roth et al. 2019). In fact, in both an ant and honeybee there was a large suite of shared gene expression patterns between males and workers to the exclusion of gynes (Warner et al. 2019). Furthermore, in a CRISPR gene knockout of early sex determination genes in honeybee it was shown that the Fem(feminizer) gene appears to control gonad size regardless of sex (though it also interacts with sex determination)(Roth et al. 2019). The emerging picture is that worker caste differentiation (from queens) arose via the co-option of male genes. That is, workers might be a gene-expression developmental mosaic (Molet et al. 2012) of male and female.
The view that male-ness contributed (and contributes) to worker evolution is logical. Male and female ants, despite behavioral, genetic, and morphological dissimilarity, arise from the same genome, whether haploid or diploid. Thus the raw material for natural selection to work from is the combination of phenotype expressed by both males and females; the colonial lifestyle of social insects may also buffer novel developmental mosaics from strong selection (Molet et al. 2012). Sexual dimorphism is present in most clades of Hymenoptera at least partly because of specialization by females for parasitism/predation and brood care (in some lineages)(Stubblefield and Seger 1994). Due to sexual dimorphism there was morphological variation for selection to act on – possibly enough to recombine developmental programs and produce a novel phenotype/caste that changed ecological and evolutionary history (Smith and Szathmary 1997), the worker. In this study I examine whether there is evidence for sexual dimorphism explaining variation in worker caste polymorphism. In extant species, does the degree of sexual dimorphism correlate with how variable the worker caste is – in other words, does male-ness have a hand in the evolution and elaboration of the worker caste?
Two species of Pogonomyrmex harvester ant, P. barbatus andP. badius , were used to examine the relationship between ant size and shape across all adult castes. These species were chosen because sufficient samples were available across all castes for detailed measurement and because they differ in regard to both sexual and worker dimorphism. P. barbatus has very little sexual dimorphism while in P. badius , gynes are more than three times the size of males by dry mass and have an energetic cost to the colony of 4.5X of a male (Smith and Tschinkel 2006). Similarly, workers of P. barbatus are characterized as monomorphic while those of P. badius are dimorphic, with clearly distinct major and minor worker sub-castes (Tschinkel 1998).