Status, Relationships, and Characters of the Percid Fish, Poecilichthys sagitta Jordan and Swain

Abstract

Listen

Translate article

N one of his famous walking trips, David Starr Jordan and his party stopped briefly to collect fishes in Wolf Creek and other tributaries of the Clear Fork of the Cumberland River near Pleasant View, Whitley County, Kentucky. From the fishes taken, Jordan and Swain (1883) described Poecilichthys sagitta on the basis of a single specimen which has remained the only known example until the present time. In the original description they noted that there was but a single anal spine, an unusual condition, since all species of Poecilichthys normally have two anal spines. They expressed the opinion that the unique individual was probably abnormal in this respect. Subsequently, however, Jordan and Evermann (1896: 1066, 1068, and 1080) apparently regarded the presence of a single spine as of marked significance since they erected the subgenus Torrentaria for the reception of sagitta and a Mexican fish, Etheostoma australe Jordan, which also has one anal spine. Jordan, Evermann, and Clark (1930: 293) elevated Torrentaria to generic rank, and included the same two species. Recently Hubbs (1936) noted that Torrentaria Jordan and Evermann is preoccupied and proposed the substitute name Austroperca for the Mexican species only. He regarded as uncertain the status and relationships of sagitta. Recently the Museum of Zoology has secured four lots (comprising 24 specimens) of sagitta including some taken near the type locality. Despite the reported occurrence of a single anal spine in the type, sagitta does not differ from the other species of Poecilichthys in this character since each of the 22 specimens examined has two anal spines; if counted correctly the type is atypical. No intimately related species are revealed by a comparison of sagitta with the species referred to Poecilichthys (sensu latu) by current authors, including the genera Poecilichthys, Nanostoma, Nothonotus, Rafinesquiellzs, Oligocephalus, Nivicola, Claricola, Catonotus, Torrentaria (in part), and Boleichthys (in part) of Jordan, Evermann, and Clark, 1930. However, two species, nianguae and spilotus, heretofore referred to the genus Hypohomus, agree with sagitta in many characters. Those forms are regarded as incorrectly assigned to Hypohomus (Table I), and their inclusion in Poecilichthys seems proper. Within the genus Poecilichthys these species plus sagitta almost certainly constitute a compact natural group, here regarded as a subgenus. To erect a genus for these species would only add to the already overburdened complexities of generic classification in the Etheostomatinae. The three species (nianguae, spilotus, and sagitta) agree with Poecilichthys in that the premaxillae are nonprotractile; the belly is covered with unspecialized scales of normal size; there are no enlarged modified scales on the breast or between the pelvic fin bases; there is no semicircular flangelike ridge on the lower side of the caudal peduncle; the anal fin, though rather long, has a smaller area than the soft dorsal; the pelvic fins are separated by a space not greater than two-thirds the fin base; the lateral line is