occurring in small groups or forming small schools 

 earlier than other species. 



Our catches of C. hippos and C. latus are too small 

 to allow remarks on spawning periods. Our catches of 

 C. ruber substantiate Berry's contention that spawn- 

 ing occurs from mid-February to mid-August; and 

 the lack of small individuals in our catches implies 

 most spawning occurs to the south of the Florida 

 Straits or in waters contributing to the Gulf Stream. 

 The 14.9 mm specimen we caught off South Carolina 

 indicates some spawning takes place off the 

 southeastern coast of the United States. Berry (1959) 

 reported that C. bartholomaei spawned from mid- 

 February to mid-September. Our catches (11.0 and 

 11.4 mm) on 4 February indicate a slightly earlier 

 start in 1968, probably in mid- to late-January. 

 McKenney et al. (1958) thought that C. fusus spawned 

 the year-round, mostly from January through August 

 with a peak in the summer. Berry (1959) thought most 

 spawning occurred from April through September. 

 Our catches are consistent with these observations, 

 both in year-round occurrence of juveniles and in- 

 creased summer abundance. 



The various species of the genus Caranx were about 

 equally distributed in temperatures from 25.0° to 

 30.0°C. Although they were caught in salinities rang- 

 ing from 34.0 to 38.0 %o , 74^r were taken from water 

 with salinities between 35.0 and 35.9 %o. 



Decapterus punctatus apparently spawns the year- 

 round with a peak in activity during the spring (Table 

 5). It is a creature of the middle shelf, not directly 

 associated with inlets, estuaries, or the Gulf Stream. 

 We caught this species during all seasons and over the 

 entire range of our sampling area. Judging by our 

 catches juveniles apparently rise to the surface during 

 the pre-dawn period, occupy the surface during the 



day, and descend at night (Table 5). The diel 

 difference breaks down in October due to a single 

 catch of 32 fish made at 0544 h, a time within my 

 definition of night. Our specimens are generally 

 smaller than Dooley's (1972) possibly because smaller 

 fishes are more strongly surface oriented, than larger 

 ones. 



Little is known of the early life history of the rain- 

 bow runner, Elagatis bipinnulata. Its restriction in 

 our collections to Gulf Stream waters off Florida in- 

 dicates an oceanic origin south of the Florida Straits. 

 These specimens were identified following the 

 description by Okiyama (1970). Key features are the 

 serrated preopercular spines in fishes less than about 

 20.0 mm and the pair of unpigmented lines on each 

 side of the otherwise darkly pigmented caudal pedun- 

 cle (Fig. 14). 



Because the diagnostically important body bands 

 are not yet present on small Seriola sp. (Mather 1958), 

 I was able to identify only those fish greater than 18.0 

 mm FL to species. Many larger fish (up to 35.8 mm) 

 are identified only to genus but are probably either S. 

 dumerili or S. rivoliana — all have seven spines in the 

 first dorsal fin, five body bands plus a sixth on the 

 peduncle, and intergrading numbers of rays in the sec- 

 ond dorsal fin. 



Dooley (1972) concluded that S. dumerili and S. 

 rivoliana spawn during the spring. Munro et al. (1973) 

 caught ripe iS. dumerili in August and November and 

 ripe S'. rivoliana in November on Pedro Bank in the 

 Caribbean. Our catches indicate spawning during fall, 

 winter, and spring for the former species, summer and 

 fall for the latter. 



Our catches of three species of Trachinotus confirm 

 Fields' (1962) observations: 1) spawning of T. 



Table 5. — Summary of catch data for Decapterus punctatus. 



Diel Analysis 



Figure 14. — Juvenile Elagatis bipinnulata, 27. .5 mm FL. 



23 



