DE SYLVA: POLYNEMOIDEI 



541 



Table 129. Comparison of Meristic Characters of Mwg;/ and. 4go- 



nostomus (Mugilidae), Polydaclylus (Polynemidae), and Sphyraena 



(Sphyraenidae) from the Western Atlantic Ocean (Data from 



Miller AND JoRGENsoN, 1973). 



Fig. 29 1 . Larvae of the polynemoid. Galeoides polydaclylus. (A) 2.75 

 mm; (B) 3.13 mm; (C) 4.3 mm; (D) 4.4 mm; (F) 7.6 mm. From Abous- 

 souan(1966cl). 



upon adult specimens (Marathe and Bal. 1958). No studies of 

 the external or internal anatomy have been undertaken on any 

 polynemid. 



The suborder Percesoces was established by Starks (1900) to 

 show the close relationships among the families Atherinidae. 

 Mugilidae, and Sphyraenidae. To this group Tate Regan (1929) 

 added the Polynemidae, based upon the well-developed cranial 

 crests, the position of the exoccipitals and basioccipitals, the 

 alisphenoid juncture, the poor development of the parapoph- 

 yses, and the 24 vertebrae shared with the Sphyraenidae. Based 

 upon extensive osteological evidence, Gosline ( 1 962) concluded 

 that the Polynemidae, Sphyraenidae, Atherinidae, and Phallo- 

 stethoidei are more closely related to one another than to other 

 fish groups, and placed them in a separate order, Mugiliformes. 

 He did. however, show that the Polynemidae. Sphyraenidae, 

 and Mugilidae were more closely related to each other based on 

 the similar number of vertebrae, the postcleithral strut, and the 

 possession of nonadhesive eggs, than to the Alhennidae and 

 Phallostethoidei. The pelvic morphology of the polynemids and 

 sphyraenids is so primitive as to suggest that these groups could 

 not have arisen from any advanced percoid groups, and that 

 they must be derived from a very low level of percoid. 



Gosline (1971) removed the Phallostethoidei from the sub- 

 order Mugiloidei (the old Mugiliformes), but concluded that the 

 superfamily Atherinoidae belonged in this suborder together 

 with the superfamilies Polynemoidae, Sphyraenoidae, and Mu- 

 giloidae. However, Rosen (1964) had removed the atherinoids 

 from the percesocine group and had established them as part 

 of a new, separate order, the Atheriniformes, a practice followed 

 widely today. Thus, the Polynemoidei, Sphyraenoidei, and Mu- 

 giloidei have no relatives closer to them than they are to each 

 other. These are presently recognized as separate suborders within 

 the order Perciformes. 



There is no salient feature in the early life history which relates 

 the Polynemoidei to other taxa. The development of the eggs 

 and larvae of Polynemoidei, Mugiloidei. and Sphyraenoidei seem 

 to follow approximately the same pattern, and all have pelagic 

 eggs. However, a major departure of the Polynemoidei from the 

 other two is that one species, Polydaclylus sexfilis. is a protan- 

 drous hermaphrodite. It matures first as a male at a fork length 

 of about 20 to 29 cm, and then transforms into the female at 

 between 30 and 40 cm following a hermaphroditic stage (San- 

 terreandMay, 1977). As far as is known, none of the Mugiloidei 

 or the Sphyraenoidei is ever hemaphroditic. 



Comparisons of meristic characters offer some noteworthy 

 data (Table 129). The vertebral count of the Polynemoidei is 

 10 + 14 = 24; the other two suborders have a count of 12 -I- 

 12 = 24. The number of dorsal and anal elements of the Mu- 

 giloidei and Sphyraenoidei resemble each other more closely 

 than they do the Polynemoidei. The vertebral formula, as well 

 as the number of dorsal and anal elements, are more closely 

 related to the Gerreidae. In fact, the habits of the Polynemoidei 

 closely resemble those of the Gerreidae. To my knowledge, there 

 is nothing published on the eariy life history of the Gerreidae 

 which might disclose any similarities to the Polynemidae. 



The Polynemoidei (i.e.. Polydaclylus) have a higher number 

 of first dorsal, second dorsal, anal, and caudal elements than 

 the other groups (Table 129). However, a companson on a 

 worldwide basis is required before such an analysis can reveal 

 phyletic relationships. 



RosENSTiEL School of Marine and Atmospheric Science, 

 University of Miami, 4600 Rickenbacker Causeway, 

 Miami, Florida 33149. 



