Testing assertions of dietary specialisation: a case study of the diet of Aparallactus capensis

The lack of detailed natural history information required to precisely characterise the diets of many organisms often results in the use of generalised descriptions of those diets. These descriptions can become dogmatic, and should be challenged with novel observational data when those data become available. We tested the characterisation of cape centipede eaters (Aparallactus capensis) as centipede-eating specialists that consume centipedes to the exclusion of other prey types. We dissected 62 preserved A. capensis specimens and identified stomach contents. Fifteen specimens contained a total of 16 prey items, all of which were identifiable as centipedes. Taking a simulation approach, we calculate the probability of all known wild prey items (n = 21) being centipedes under a range of different simulated diets. We show that if cape centipede eaters do eat non-centipede prey, they do so very infrequently. Although all reported prey items for A. capensis are from the order Scolopendromorpha, a survey of 453 citizen science records of South African centipedes suggests that members of this order account for two thirds of all observations. These data do not allow us to reject the hypothesis that cape centipede eaters consume different centipede prey proportionally to what the encounter.