Suttonia

Abstract: Relationships among epinepheline genera are investigated based on cladistic analysis of larval and adult morphology. Five monophyletic tribes are delineated, and relationships among tribes and among genera of the tribe Grammistini are hypothesized. Generic composition of tribes differs from Johnson's (1983) classification only in the allocation of Jeboehlkia to the tribe Grammistini rather than the Liopropomini. Despite the presence of the skin toxin grammistin in the Diploprionini and Grammistini, we consider the latter to be the sister group of the Liopropomini. This hypothesis is based, in part, on previously unrecognized larval features. Larval morphology also provides evidence of monophyly of the subfamily Epinephelinae, the clade comprising all epinepheline tribes except Niphonini, and the tribe Grammistini. Larval features provide the only evidence of a monophyletic Epinephelini and a monophyletic clade comprising the Diploprionini, Liopropomini and Grammistini; identification of larvae of more epinephelines is needed to test those hypotheses. Within the tribe Grammistini, we propose that Jeboehlkia gladifer is the sister group of a natural assemblage comprising the former pseudogrammid genera (Aporops, Pseudogramma and Suttonia). The "soapfishes" (Grammistes, Grammistops, Pogonoperca and Rypticus) are not monophyletic, but form a series of sequential sister groups to Jeboehlkia, Aporops, Pseudogramma and Suttonia (the closest of these being Grammistops, followed by Rypticus, then Grammistes plus Pogonoperca). The absence in adult Jeboehlkia of several derived features shared by Grammistops, Aporops, Pseudogramma and Suttonia is incongruous with our hypothesis but may be attributable to paedomorphosis. The generic phytogeny of the Grammistini proposed herein emerges as the single most parsimonious hypothesis largely because of the method chosen for analyzing multistate characters. Few percoid families have received as extensive systematic treatment as the Serranidae. Since its recognition almost 200 years ago, the family has been expanded, restricted and subdivided, and often has comprised a bewildering array of genera whose affinities to one another were unclear. By restricting it to the subfamilies Anthiinae, Epinephelinae and Serraninae (sensu Jordan and Eigenman, 1890, with some modifications), Gosline (1966) made an important step toward defining the Serranidae as a natural assemblage. Johnson (1983) largely corroborated Gosline's hypothesis, but cladistically refined it, citing four derived features (presence of three opercular spines and absence of the procurrent spur, third preural radial cartilages and posterior uroneural) as evidence of the monophyly of the family. Furthermore, Johnson (1983) diagnosed a monophyletic subfamily Epinephel