Abstract/Summary

Despite theoretical predictions, dishonest signalling has rarely been observed in aggressive interactions. We present evidence of such signalling in the nonpollinating. g wasp Philotrypesis sp. A ex Ficus rubiginosa. First, morphometric data indicated that an alternative 'atypical' male morph (17.8% of individuals) exists that tends to be larger in body size and has longer mandibles for a given body size than other 'typical' males. Second, behavioural observations suggested that males use mandible gape width (which depends on mandible length) as a cue to assess opponents before fights and retreat without escalating if they are unlikely to win, and, probably because their greater mandible gape width causes more opponents to retreat without escalating, that atypical males engaged in fewer fights than typical males for a given body size but had higher mating success. Third, atypical males were less likely to win fights than typical males of similar mandible length relative to opponents. In addition, we found that atypical males incur more injuries (greater receiver-dependent signal costs) than typical males of similar body size relative to rivals. We discuss the implications of our findings for future work on dishonest signalling. (C) 2009 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.