| The venom glands of fishes represent a widespread, putatively adaptive antipredatory trait, which have received little attention relative to those of other venomous organisms. I review the literature for ecologically and evolutionarily relevant information regarding venomous fishes and find poor support for an adaptive hypothesis, due to the lack of empirical evidence for an antipredatory function of venom glands in most groups of venomous fishes. Additional information, including phylogenies for venomous groups, the phylogenetic distribution of venom glands within these groups, intra- and interpopulational variation in the presence and/or structure of venom glands, and the functional design of venom constituents is also found to be lacking. Using histological and toxicological assays, I demonstrate that 1250-1625 catfish species should be presumed to be venomous, nearly doubling estimates of venomous fish species diversity. Interfamilial comparisons of venom toxicity and composition indicate that a broad range of venom activities and constituents exist in catfish venoms, which character optimization analyses suggest has resulted from a limited number of independent evolutionary origins of venom glands in the order Siluriformes. Behavioral experiments demonstrate that the venom glands of the tadpole madtom (Noturus gyrinus) provide a significant antipredatory advantage relative to individuals lacking venom glands, supporting an adaptive hypothesis for venom glands in catfishes. The broader examination of venom toxicity and composition in 22 ictalurid species reveals significant variation in levels of venom toxicity, which appears to demonstrate correlated evolution with several aspects of ictalurid life history. Additional experiments demonstrate that the presence of venom glands can also influence the evolution of other aspects of a catfish species' biology. A model predatory species ( Micropterus salmoides) quickly became conditioned to avoid attacking catfishes possessing the distinctive color pattern displayed by Lake Tanganyikan Synodontis species, and were unable to distinguish between different Tanganyikan Synodontis species, supporting a hypothesis of Müllerian mimicry in this group. Statistical examinations of Synodontis phylogenetic history and modern species' color patterns indicate that the color pattern conservatism seen in Lake Tanganyikan Synodontis is likely due to the protection provided by this color pattern, which is predicated on the defensive capabilities afforded by these species' venoms. |
