Abstract
Despite their prevalence in nature, the evolution of sex-specific female ornaments is still not well understood. Although in some cases (often carotenoid-based ornaments) they appear to honestly signal quality, such as fecundity, it has been suggested that some female ornaments have evolved to deceptively to obtain matings. We address these two hypotheses in the long-tailed dance fly ( Rhamphomyia longicauda ), where females possess two sex-specific ornaments: pinnate scales on the hind femur and tibia and abdominal sacs that are inflated in female-biased display swarms. Although several studies have suggested that female ornaments in this species are deceptive, evidence is mixed and requires further investigation. Here, we use static allometry (with body size as a proxy for condition) of both ornamental and non-ornamental traits in females (and homologous non-ornamental traits in males) in order to determine whether they are honest or deceptive signals of quality. Most male traits scaled isometrically with body size, however male leg hairs showed positive static allometry, probably because they are involved in nuptial-prey capture or in grasping mates. Ornamental traits in females (abdomen area and tibia scale length) showed significant positive allometry and had steep slopes relative to non-ornamental traits. As larger females invest more in ornamentation relative to smaller females, this suggests that these traits are likely honest, condition-dependent signals of quality. We note that honesty and deception are not mutually exclusive hypotheses. Individuals may vary in their signalling strategy, resulting in, for example, deception from some low condition individuals but honesty overall. Although our finding of positive allometry makes it unlikely this occurs in long tailed dance flies, simultaneous honesty and deception should be considered in future studies of female ornamentation.