Dishonest signalling in a ddler crab

Animal communication theory predicts that low-frequency cheating should be common in generally honest signalling systems. However, perhaps because cheats are designed to go undetected, there are few examples of dishonest signals in natural populations. Here we present what we believe is the ¢rst example of a dishonest signal which is used commonly by males to attract mates and ¢ght sexual rivals. After losing their large claw, male ¢ddler crabs (Uca annulipes) grow a new one which has less mass, is a less e¡ective weapon and costs less to use in signalling than an equivalent-length claw of the original form. Males with original claws do not di¡erentially ¢ght males with regenerated claws even though they are likely to win. Regenerated claws e¡ectively blu¡ ¢ghting ability and deter potential opponents before they ¢ght. During mate searching, females do not discriminate against males with low-mass, regenerated claws, indicating that they are deceived as to the true costs males pay to produce sexual signals. Up to 44% of males in natural populations have regenerated claws, a level unanticipated by current signalling theory. The apparent rarity of cheating may be an artefact of the usual di¤culty of detecting cheats and dishonesty may be quite common.