Abstract
Experimental field studies in Costa Rica reveal that A. fasciatus density in a pool connected to a subterranean cavity is negatively correlated with fishing bat activity. Since fish tend to remain within the cavity when bats are actively foraging, the vulnerability of A. fasciatus to bat predation may be reduced. Fish from the study pool and those living in nearby rivers do not differ in gross morphology. Taking refuge in the cavity may be a first step in cave colonization; this behavioral change can occur prior to gross morphological changes. This result is consistent with Mayr's hypothesis that behavioral changes may precede morphological ones during the invasion of a new niche, and that behavior may act as the pacemaker of evolutionary change.All A. fasciatus collected from caves in Mexico exhibit scotophilia. Larger (older) fish are more strongly scotophilic than smaller (younger) ones. Scotophilia varies inversely with the degree of reduction in pigmentation and development of the eyes. There is thus a negative correlation between behavior that accompanies cave colonization and the morphology typical of cave-dwelling animals.Phenotypic intermediacy between surface and cave forms found in one Mexican population is suggested to be the result of introgression that has taken place in less than forty-three years. The contention that cave colonization necessarily requires drastic genotypic and phenotypic changes is thus rejected for A. fasciatus.The concept of "regressive evolution" is evaluated from historical, semantic and conceptual perspectives and is found inadequate on several grounds.