were present at the dorsal fin base of 9.5- to 11.0- 

 mm larvae. They numbered from 3 to 8 at about 

 14.0 mm and then as many as 15 at 18.0 mm. Two or 

 three stellate melanophores were present at the 

 anal fin base on 11.0- to 12.0-mm larvae; these 

 numbers increased rapidly to 11 or 12 at 18.0 mm. 

 Numbers of melanophores at the dorsal and anal 

 fin bases were variable among individuals of the 

 same length. A single stellate melanophore 

 developed at the pelvic fin base on specimens as 

 small as 12.3 mm. Some tiny melanophores began 

 to occur in the pectoral, dorsal, and anal fins at 16.2 

 mm. 



Caudal Region 



Newly hatched larvae have melanophores on 

 both the dorsal and ventral sides of the notochord 

 tip. Numbers on the dorsal tip range from one to 

 two, while those on the ventral tip range from one 

 to three. In addition, one or two melanophores are 

 located along the ventral midline posterior to the 

 anus. Pigment along the ventral tip of the no- 

 tochord began to migrate into the caudal finfold at 

 7.0 to 7.2 mm. This pigment was associated with 

 developing caudal rays in larger larvae. Pigment 

 in the caudal fin increased rapidly when larvae 

 exceeded 10.0 mm. Internal melanophores were 

 first present in the hypural plate region on larvae 

 10.3 mm and longer. Larvae longer than 16.5 mm 

 invariably had many melanophores among the 

 rays of the caudal fin. 



Transformation 



Transformation of larvae to juveniles ap- 

 parently was complete between 20 and 23 mm. 

 Unfortunately, we preserved no specimens that 

 were between 18.0 and 22.7 mm. However, our 

 specimens of 17.8 and 18.0 mm were not complete- 

 ly transformed while the 22.7-mm specimen had 

 acquired nearly all of the juvenile characteristics. 

 Larvae began transforming at about 14.0 mm. At 

 that time, proportional measurements relating 

 preanal length, predorsal length, body depth, head 

 length, snout length, and eye diameter to standard 

 length (Table 4) began to change rapidly. Except 

 for body depth, which continued to increase, the 

 proportional measurements became nearly con- 

 stant at 22.7 mm. The distribution of myomeres 

 relative to other body parts became stable for 

 specimens 22.7 to 36.2 mm (Table 3). Fin rays were 

 completely ossified at 18.0 mm (Table 5), but the 



epural bones of the caudal skeleton were still 

 unossified at that size. Although some scales were 

 present on our 17.8- and 18.0-mm specimens, the 

 22.7-mm specimen was the smallest that was fully 

 scaled. The slender, rodlike shape of the larva was 

 replaced by the deeper bodied, laterally 

 compressed form of the juvenile between 17.8 and 

 22.7 mm. The silvery coloration of juveniles was 

 present on our specimens from 22.7 to 36.2 mm. 

 Transformation included the following features: 

 forward movement of the dorsal fin; shortening of 

 the gut; forward movement of the anal fin; and 

 relative increases in head length, snout length, eye 

 diameter, and body depth. The same features were 

 noted for transforming larvae of Atlantic thread 

 herring (Richards et al. 1974) and scaled sardine 

 (Houde et al. 1974). 



COMPARISONS 



Eggs and larvae of the genus Brevoortia almost 

 always can be distinguished from those of other 

 clupeid genera that spawn in marine waters of the 

 south Florida and Gulf of Mexico region. Members 

 of the genera Alosa and Dorosoma have demersal 

 eggs, unlike those of Brevoortia which are pelagic. 

 Dorosoma spawns in fresh waters and Alosa in 

 fresh or nearly fresh waters, so that occurrence of 

 their larvae with those of Brevoortia is unlikely. 

 Larvae of Alosa spp. have more total myomeres 

 than any species of Brevoortia, and the genera can 

 be easily distinguished. Dwarf herrings (Jenkin- 

 sia spp.) might occur with B. smithi at the 

 southern extreme of their range in Florida. 

 Neither spawning habits nor eggs and larvae of 

 Jenkinsia have been described, but the total 

 myomeres of Jenkinsia did not exceed 42 (Miller 

 and Jorgenson 1973), making it unlikely that lar- 

 vae could be confused with Brevoortia which have 

 higher myomere numbers {Brevoortia, 44-48). 

 Since B. smithi may occur with either B. tyrannus 

 or B. patronus and because hybrids are known 

 (Dahlberg 1970), the specific identification of 

 menhaden eggs and larvae from plankton collec- 

 tions is still in doubt where the species' ranges 

 overlap. 



Eggs of B. smithi are smaller than those of 

 Harengula jaguana which range from 1.55 to 1.78 

 mm in diameter (Houde et al. 1974), and they can- 

 not be confused with those of Etrumeus teres 

 because eggs of that species have no oil globule. 

 Our B. smithi eggs were similar to those of 

 Opisthonema oglinum (Richards et al. 1974) and to 



671 



