Ornamentation is associated with social costs in male red-backed fairywrens ( <i>Malurus melanocephalus</i> )

Abstract Sexually-selected ornaments typically confer reproductive benefits, but ornamented and unornamented male phenotypes often co-occur in the same population. Reduced social costs are one potential compensatory benefit of unornamented male phenotypes, but few studies have tested this hypothesis in a natural context. Here we present a field-based experiment testing whether ornamented plumage carries social costs in the red-backed fairywren (Malurus melanocephalus), a small Australian songbird that exhibits delayed but flexible plumage maturation. Only one-quarter of 1-yr-old male red-backed fairywrens acquire ornamented plumage, but young ornamented males obtain greater reproductive success than males that retain brown, female-like plumage. We experimentally tested whether ornamented plumage is associated with social costs by implanting brown 1-yr-old males with testosterone to induce molt into ornamented plumage, then recorded the resulting social interactions. Testosterone-implanted males developed ornamentation and were chased and attacked more often by conspecific males than were control-implanted males without ornamentation, supporting the hypothesis that molt into ornamented plumage is socially costly. However, despite these apparent costs, some 1-yr-old males naturally acquired ornamented plumage after pairing with a female. Paired 1-yr-olds who naturally acquired ornamented plumage exhibited smaller home ranges than the unpaired, implanted males in their natal groups. Post-hoc simulations suggested that smaller home ranges could result in fewer interactions with neighbors, meaning smaller home ranges may reduce the social costs associated with acquiring ornamentation. We discuss the aspects of manipulating ornamentation with testosterone and suggest that our results add to growing evidence that social costs can enforce the honesty of sexually-selected ornaments.