Armament under direct sexual selection does not exhibit positive allometry in an earwig

The allometric scaling relationships of armaments and ornaments have been subject to extensive debate. A large body of empirical evidence suggests that sexually selected traits typically exhibit positive static allometry, where the large individuals express proportionally larger traits. Recent theory suggests that this need not be the case. We confirm this prediction using the earwig Euborellia brunneri as a model species—unusually, the male armament in this species does not exhibit positive allometry. We experimentally assessed the strength of direct and indirect selection on armament length and morphology and on body size and weight. In a 3-stage experiment, we first permitted females to choose between 2 males and assigned 1 male as the preferred and the other the nonpreferred male. We then allowed the same pair of males to establish a dominance hierarchy in fighting trials. Last, to evaluate the implications of female choice, we conducted mating trials where half the females were mated to their preferred male and half with their nonpreferred male. We found that male armament length and body weight were under direct sexual selection through intrasexual competition. In contrast, female mate preferences did not relate to any measured male trait. Finally, mating behavior was not related to male preference status or armament length. Thus, armament size is sexually selected through intrasexual competition although it does not exhibit positive allometry. Our finding suggests that researchers should be cautious when inferring the absence of sexual selection in traits that do not exhibit positive allometry.