526 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



Table 127. Continued. 



definitive evidence supporting its removal from the Carangidae. 

 Their work is an important contribution towards elucidating 

 carangoid relationships. 



Much more effort needs to be focused on obtaining basic data 

 on the biology, ontogeny and systematics of carangids. The data 

 presented in Table 127 and the following discussion are a first 

 step in that direction. 



Relationships 



Evidence supporting the monophyly of carangoids and several 

 major groups of carangids is discussed below. Refer to G. D. 

 Johnson (this volume) for discussion of interfamilial relation- 

 ships of the three echeneoid families. The oldest available name 

 for each of the four carangid tribes herein recognized has not 

 been determined but none is proposed as new. No synapomor- 

 phies were found to support the inclusion of Lichia in the tribe 

 Trachinotini, and its placement is one of convenience in accord 

 with the practice of previous authors and reflects my own sub- 

 jective bias. Autapomorphies that define Trachinolus. Lichia 

 and the naucratine genera are not included in the cladogram 

 (Fig. 276) because they are not informative about relationships. 

 These taxa are recognized individually in the figure to make it 

 easier for the reader to determine the character state distribu- 

 tions and the number of species comprising each genus. The 

 Carangini includes approximately 20 genera (Table 127), many 

 not well established, and their osteology poorly known. Until 

 this presumably monophyletic assemblage has been studied in 

 much greater detail no meaningful discussion of relationships 

 will be possible. Several recent authors have considered Para- 

 stromateus to constitute either a monotypic family or carangid 

 subfamily. Available evidence suggests, however, that it should 

 be assigned to the Carangini. 



The following character states are the basis for the hypotheses 

 of carangoid interrelationships inferred in Fig. 276. The pre- 

 sumed derived character state is listed first, followed by dis- 

 cussion of the character in out-groups when necessary. 



(I) Freihofer (1978) made the important observation that in 

 the Nematistiidae, Carangidae, Coryphaenidae, Rachycentridae 



and Echeneididae there are one or two tubular ossifications 

 (prenasals) around the anterior extension of the nasal canal. This 

 presumed specialization of the lateralis system is very rare in 

 percoids (also present in the unrelated and highly specialized 

 Toxotidae) and is considered to be a synapomorphy suggesting 

 that these five families constitute a monophyletic group. 



(2) The possession of small adherent cycloid scales is a derived 

 character shared by carangoids in contrast to the typically cte- 

 noid scales of most other percoids. Berry (1969) reported that 

 the carangid Elagatis has "ctenoid" scales and Zheng (1981) 

 also described the highly modified caudal peduncle scales of 

 Naucrates as ctenoid. These scales are not typically "ctenoid" 

 and appear to represent modifications of the carangoid scale- 

 type. 



(3) Two separate prenasal canal units, one membranous and 

 one bony (Carangidae) or both bony (echeneoids). In contrast, 

 Nematistiits has only a single prenasal canal unit. 



(4) Loss of the bony stay (Fig. 277) posterior to ultimate dorsal 

 and anal pterygiophores that is present in most other percoids 

 (see Table 127, G. D. Johnson, this volume). 



(5) On shoulder girdle, middle part of coracoid with its an- 

 terior margin consisting of a lamella of bone broadly extending 

 towards the median cleithral wing (Suzuki, 1962: figs. 36-44). 

 In Nematistius the middle and lower parts of the coracoid are 

 rodlike with lamellar bone restricted to its posterior margin 

 (Rosenblatt and Bell, 1976; fig. 8). 



(6) Basioccipital with a pair of foramina (Rosenblatt and Bell, 

 1976: fig. 3) into which anterior processes of the gas bladder 

 extend forward to the region of the inner ear. 



(7) Anterior shift of second predorsal bone to the first inter- 

 neural space and first pterygiophore greatly expanded and plate- 

 like. In carangids, as in most other percoids, the second pre- 

 dorsal bone occupies the second iniemeural space (predorsals 

 absent in echeneoids), and in both echeneoids and carangids the 

 first dorsal pterygiophore is not greatly expanded. 



(8) Spines of first dorsal fin very long and filamentous and 

 only basally connected by interradial membrane. 



(9) Tubular ossifications surrounding both prenasal canal units; 



