The mystery of the missing toes: extreme levels of natural mutilation in island lizard populations

P>1. Ecological interactions that involve aggressive confrontations between animals are important in shaping the evolution of morphology, behaviour and life history. However, as such confrontations are rarely witnessed, direct quantification of the intensity of these processes in natural populations is notoriously difficult. While the utilization of the frequency of non-lethal injuries is fraught with difficulties, it may provide information concerning different types of interaction, such as predation, intraspecific aggression and interspecific interference competition. 2. In this paper, we report on an exceptionally large difference in toe loss incidence between two populations of Podarcis sicula lizards living on small, neighbouring islands in the Adriatic Sea. We caught 900 lizards and recorded the number and location of missing toes. Subsequently, we investigated five non-mutually exclusive hypotheses concerning differences in bite force capacity, bone strength, predation intensity, average age and intraspecific aggression that may provide proximate explanations for the observed differences in injury frequencies. 3. Bite force differences differed considerably between the populations, but bone strength was found to be stronger in the populations with a higher frequency of natural scars. Predation pressure clearly differed between the populations, but we found higher injury rates under predation relaxation. 4. Our results indicate that density and consequently an increased intraspecific competition is the most likely explanation for the observed high frequencies of injuries. We suggest that the intensity of toe amputation between lizard populations may be a useable indirect indication for the intensity of intraspecific competition. 5. This study shows how a combination of morphological, physiological, behavioural and ecological measurements can be used to test assumptions implicit to alternative explanations of an observed phenomenon. Such tests can reveal how likely each of these explanations is, even if the processes leading to the phenomenon are difficult to observe directly.