Luring cannibal: dishonest sexual signalling in the springbok mantis

Sexual signals that attract the opposite sex are typically honest indicators of quality. However, sexually antagonistic selection can favour dishonest signals if poor quality members of the signalling sex can improve their own fitness at a cost to the opposite sex by conveying deceptive information in their signals. Here, using the sexually cannibalistic mantis Miomantis caffra, we found evidence that females in low body condition use sexual deception to exploit males as prey. Females of high condition produced 97% heavier oothecae than low-condition females, indicating that high-condition females were the optimal choice for males. However, in a simultaneous mate choice experiment using a T-maze olfactometer, males chose the pheromonal signals of low-condition females twice as often as those of high-condition females, indicating that female pheromonal attractiveness did not match female quality. In subsequent mating experiments, low-condition females attacked males 3 times more frequently and cannibalised them 4 times more often than high-condition females did. Cannibalism provided significant material benefits for females of low condition, since oothecae produced by these females were 52% heavier when they cannibalised a male compared to when they did not. These results strongly suggest that low-quality females attract males dishonestly with their pheromones to boost fecundity through sexual cannibalism. Such deception could be widespread in sexually cannibalistic systems, especially if pheromone production is cheap, food limitation is common, and males represent a substantial meal for hungry females.

Methods
We conducted a series of laboratory experiments to investigate whether hungry (low-condition) females of the sexually cannibalistic springbok mantis, Miomantis caffra, deceive males with their pheromonal signals to reap the material benefits of cannibalising them. We used a T-maze olfactometer to test whether males preferred the pheromones of high- or low-condition females. We also performed mating trials to assess whether females of high or low condition were more or less aggressive and cannibalistic toward males. Finally, we collected measurements on the difference in fecundity between high- and low-condition females, as well as the effect of cannibalism on the fecundity of low-condition females.