Fish Out Of Water: New Species of Climbing Fish from Remote Venezuela Shakes the Catfish Family Tree

The discovery of both of these characteristics in Lithogenes suggests to ichthyologists Scott Schaefer of the American Museum of Natural History and Francisco Provenzano of the Universidad Central de Venezuela that the common ancestor of the Loricariidae and Astroblepidae probably could grasp and climb rocks with its tail and mouth.

The unusual catfish caught the team’s attention twenty years ago in Caracas. An anthropologist working in the remote state of Amazonas collected samples of local foods and brought them to the Instituto de Zoologíca for identification.

"The fish was so strange in morphology that it did not fit into any taxonomic category that we were aware of," recalls Schaefer, a curator in the Division of Vertebrate Zoology at the Museum. "But it looked like it was run over by a truck. We needed better specimens."

It took years to pin down where the fish was found, but the team collected L. wahari after several trips further and further into the headwaters of the Río Cuao, a tributary of the Río Orinoco. They literally picked 84 specimens off of rocks.

The new samples of L. wahari confirmed that the species is a member of a group that bridges two catfish families. Bony plates on both its head and tail, plus other features, link the species to the Loricariidae, a widespread and successful family of fully armored catfishes. But L. wahari also has a specialized pelvic fin that decouples from its body and moves backward and forward independently. This feature—used in combination with a grasping mouth to move like an inchworm up rocks—is otherwise found only in a family of climbing catfish restricted to the Andes, the Astroblepidae.

Schaefer and Provenzano propose that L. wahari is the third known species in the subfamily Lithogeninae, and that the specialized features shared among the three species confirms their placement within the family Loricariidae at the base of this large radiation of catfishes. This phylogenetic arrangement suggests that the common ancestor to both families probably inhabited upland, rather than lowland, streams of the Amazon and Orinoco river basins, where most of the family diversity is currently found.

"We see new fish species all the time, but when you also get new information about the biological history of a group, it’s the most fun," says Schaefer. "The question is whether the grasping pelvis and climbing behavior evolved once or if it was independently acquired in these groups. I don’t think it evolved twice, although there are slight anatomical differences—so the jury is still out."

The paper is published in American Museum Novitates. Research was supported by the Constantine S. Niarchos Scientific Expedition Fund and the National Science Foundation.