Feeding Experiments with a Supposed Mimetic Complex in Salamanders

The supposed mimetic complex consisting of the red- cheeked salamander, Plethodon jordani, and the red-cheeked variants of the Allegheny Mountain salamander, Desmognathus ochrophaeus, was tested by subjecting members of this complex to laboratory predation by four natural predator species. Predation by the eastern garter snake, the large short-tailed shrew, the Blue Ridge spring salamander, and the black-bellied salamander failed to reveal discriminatory feeding behavior. The possibility that mimicry has not been involved in the evolution of the red cheek patch in D. ochrophaeus remains. The hypothesis that infrequent but discriminate predation by uncommon predators such as birds could confer slight selective advantage upon the D. ochrophaeus the range of red-cheeked P. jordani the incidence of the red cheek coloration among normally patterned D. ochrophaeus is low. Numerous individuals have suggested that the similarity can be best explained in terms of mimicry. According to the theory of Bate- sian mimicry, the red cheek patch of P. jordani (the model) could be considered warning or aposematic coloration, which would "ad- vertise" to, predators some distasteful or undesirable quality. Huheey (1960) suggests that this undesirable characteristic is a slimy exuda- tion from the tail, not found in D. ochrophaeus. The red-cheeked variant of D. ochrophaeus (the mimic), which exhibits pseudapose- matic or false warning coloration, would thus gain survival value from its close resemblance to the model, even though it possessed none of the undesirable qualities. In Miillerian mimicry, both P. jordani and the red-cheeked D. ochrophaeus would possess unde- sirable qualities and thus the model and the mimic are both protected by undesirable characteristics. In the study of any suspected mimicry complex, the determina- tion of a natural predator that discriminates between the supposed mimics and nonmimetics is of utmost importance. If one can ascer- tain the natural predators of the models and mimics by field obser- vations or stomach analyses, and then prove the discriminating behavior of at least one such predator, he greatly strengthens the hypothesis 147