522 



ONTOGENY AND SYSTEMATICS OF HSHES-AHLSTROM SYMPOSIUM 



stage (along with the caudal fin rays) in most species, closely 

 followed by neural and haemal spines. Vertebrae, neural, and 

 haemal spines ossify sequentially, anteroposteriorly. Centra os- 

 sify from their anterior margin posteriorly. Neural spines of the 

 abdominal vertebrae, and neural and haemal spines of caudal 

 vertebrae begin to ossify before their respective centra. Ribs 

 ossify at about the same time and also develop anteroposte- 

 riorly. Pleural ribs ossify before the epipleural ribs. The urostyle 

 begins to ossify before the posteriormost two or three vertebrae 

 during the flexion stage. Ossification proceeds from its anterior 

 base towards its distal tip as it also does in the hypurals. 



Pigmentation. — Details concerning the development and vari- 

 ety of pigmentation characters are discussed by Laroche et al. 

 (MS) and are summarized for genera in Table 126. Although 

 many species have not been observed and this table is tentative, 

 it reflects the potential utility of pigmentation characters. 



It is not possible to describe a generalized pigmentation pat- 

 tern that is unique to and diagnostic for all carangid larvae. By 

 the end of the preflexion stage, most species have rows of me- 

 lanophores along the dorsal and ventral margins of the tail. 

 Melanophores appear on the head over the brain and eventually 

 form a cap of pigmentation. Dorso- and ventrolateral pigmen- 

 tation may be present or absent depending on the species (Fig. 

 270A, B). A row of small melanophores develops along the 

 lateral midline at midbody during the preflexion stage and per- 

 sists into the juvenile stage (Figs. 270-275). When these me- 

 lanophores are expanded, they appear as a line of pigmentation. 

 This pigmentation along the lateral midline has been referred 

 to as the "lateral line streak" by Ahlstrom and Ball (1954) and 

 Miller and Sumida (1974). The amount and pattern of mela- 

 nistic pigmentation on the head, body, and fins of carangid 

 larvae is otherwise quite diverse, grading from very light to very 

 dark pigmentaton. However, larvae can usually be categorized 

 as either lightly or darkly pigmented (Table 126, Figs. 270-275). 

 Darkly pigmented forms usually have a lightly pigmented caudal 

 peduncle (Figs. 272, 273). 



Systematic considerations 



Although considerable taxonomic confusion still exists re- 

 garding carangids, and developmental stages for most species 

 remain unknown, similarities among larvae of species assigned 

 to the same genus suggest a congruence between adult and larval 

 similarities which may reflect the naturalness of some generic 

 groups. For example, all species of the genus Selene for which 

 larvae are known share precocious development of the spinous 

 dorsal, pelvic, and caudal fins, while all species of Decapterus 

 for which larvae have been described begin development of a 

 finlet at the posterior of the dorsal and anal fins before more 

 anterior elements begin to develop. Interestingly, Selar cni- 

 menophthalmus (which lack finlets as adults) larvae also begin 

 development of a fin element at the posterior of the dorsal and 

 anal fins before more anterior elements begin to develop (Fig. 

 270C). This character may reflect a relationship between De- 

 capterus and Selar. This type of information is encouraging and 

 may tend to raise confidence in the naturalness of taxonomic 

 groups and in the potential utility of developmental characters 

 for use in systematic studies of carangids. 



Developmental information is available for too few species 

 to allow interpretation of character patterns which might reflect 

 phylogenetic relationships within the Carangidae. Of course, 

 investigation of Carangidae's relationship to other groups within 

 Perciformes is a much larger problem and will require that 

 similar information be gathered for other taxa. Careful, com- 

 parative developmental studies are needed to supply this critical 

 information and provide the most direct route towards a better 

 understanding of relationships. 



(W.A.L.) School of Natural Resources, Department of 

 Fisheries, Humboldt State University, Arcata, Cal- 

 ifornia 95521; (W.F.S.-V.) Department of Ichthyology, 

 The Academy of Natural Sciences, 19th and The 

 Parkway, Logan Circle, Philadelphia, Pennsylvania, 

 19103; (S.L.R.) Gulf Coast Research Laboratory, East 

 Beach Drive, Ocean Springs, Mississippi 39564. 



Carangidae: Relationships 

 W. F. Smith- Vaniz 



DESPITE the great economic importance and broad geo- 

 graphic distribution of the Carangidae, knowledge of their 

 systematics is very inadequate. The few attempts to determine 

 their phylogenetic relationships have been both limited in scope 

 and methodologically flawed. These classifications largely reflect 

 the distribution of characters shared between taxa rather than 

 being based on evolutionarily derived characters. Lack of knowl- 

 edge of an appropriate out-group for comparison has also lim- 

 ited progress in this area. 



In his pioneering study of carangid osteology and relation- 

 ships, Starks (1911) recognized four subfamilies but stressed the 

 difliiculty of establishing intrafamilial relationships. Suzuki ( 1 962) 

 described and illustrated the osteology of 1 8 genera of carangids. 



Unfortunately only Japanese species were considered and, al- 

 though much useful descriptive information was presented, little 

 progress was made towards attaining a better understanding of 

 carangid phylogeny. Vergara (1972) described the osteology of 

 the Cuban species assigned to Caran.x and presented a phyletic 

 analysis of their relationships. In a subsequent paper Vergara 

 (1974) expanded his analysis to include all Cuban genera of 

 Carangidae and evaluated the phenetic relationships of Cuban 

 Caran.x. Smith-Vaniz and Staiger (1973) concentrated their ef- 

 forts on the Scomberoidini and presented evidence suggesting 

 a sister-group relationship between Parana and Scomheroides + 

 Oligopliles. The detailed comparison and osteological descrip- 

 tion of Nematistius by Rosenblatt and Bell (1976) provided 



