TROPHIC DIFFERENTIATION IN ILYODON , A GENUS OF STREAM‐DWELLING GOODEID FISHES: SPECIATION VERSUS ECOLOGICAL POLYMORPHISM

An impressive diversity of trophic adaptations is a hallmark of teleostean fishes, and has been a characteristic of their evolution that dates from their first appearance in the fossil record. Students of this diversity have usually regarded a particular suite of functionally related trophic specializations as attributes of a given species that adapt it to a particular ecological niche and tend to exclude it from other niches. Generally, though allowance is made for developmental variation or sexual dimorphism, a species has been presumed to have only one set (or a narrow range) of trophic adaptations. It has been thought extremely unlikely for contemporaneous adult members of the same Mendelian population to possess radically different (and mutually exclusive) arrays of trophic features. Thus, especially at or near the species level, ichthyologists have tended to equate trophic differentiation with systematic divergence. The generality of this relationship has recently been seriously challenged. Sage and Selander (1975) have demonstrated that three trophically specialized sympatric cichlids (Cichlasoma) in the Cuatro Cienagas basin of Coahuila, Mexico, a molluskivore (with molariform pharyngeal teeth and a short gut), an algal detritus feeder (papilliform pharyngeal teeth and a long gut) and a piscivore all previously thought to be distinct species (Taylor and Minckley, 1966)-exhibited coordinate geographic variation at several allozyme loci, and were therefore most likely conspecific. Vrijenhoek (1978) detected two trophically divergent, sympatric clones of the triploid unisexual fish species Poeciliopsis 2-monacha-lucida. The clones, though obviously closely related, were differentiated in dental morphology and feeding behavior. The implications of the discovery of discontinuous trophic polymorphisms in single populations are far-reaching. From the perspective of ichthyology, the existence of such polymorphisms (involving well defined "taxonomic" characters) suggests that the number of biological species involved in several extreme radiations of trophic diversity, most notably those of cichlids in the African Rift lakes (Fryer and Iles, 1972) and of cyprinids in Lake Lanao (Myers, 1960), may have been seriously overestimated by morphological inferences (see also Kornfield, 1978). The nature of the radiations themselves, as well as the systematic utility of trophic characters in general, may have to be reevaluated. From the perspective of ecology, trophic polymorphism implies that "a single species occupies niches that are as distinctly different as is usually the case among validly different species" (Hutchinson, 1978, p. 179), and leads to a host of questions concerning the ecological relationships of the trophic morphs. From the perspective of evolutionary genetics, the discontinous trophic variation may well be the type of niche-specific polymorphism that Maynard Smith (1966) and others have regarded as a fundamental requisite for sympatric speciation. The only other known case of trophic polymorphism in teleosts is that postulated in species of the characoid genus Saccodon by Roberts (1974): up to four dental mornhs occur in what qnnear to he con1 Present address: Department of Biology, Virginia Polytechnic Institute & State University, Blacksburg. Virginia 24061.