FISHERY BULLETIN: VOL. 70, NO. 1 



for describing larval development. No larvae 

 survived beyond 22 days after hatching, prob- 

 ably because of the inadequate diet. 



TRANSFORMATION 



Transformation from the larval to the early 

 juvenile stage occurred at about 13.5 mm SL 

 for S. boreal'is. At that size, fin ray development 

 was complete and there were no remnants of 

 the larval finfold. Additional evidence that this 

 size marked the end of the larval period was 

 the occurrence of keeled scales on specimens at 

 14.5 mm SL and the appearance of juvenile pig- 

 ment patterns between 12.5 and 14.5 mm SL. 



COMPARISONS 



Among the four early life history series of 

 barracudas that have been described in the lit- 

 erature, newly hatched larvae of S. borealis most 

 closely resemble those of S. p'mguis from Jap- 

 anese waters (Shojima et al., 1957; Uchida et al, 

 1958) , except for size at hatching. In both spe- 

 cies, the oil globule, at hatching, is located at the 

 anterior end of the yolk mass; neither develop 

 melanophores in the finfold. In contrast, the oil 

 globule is located at the posterior end of the yolk 

 mass in newly hatched larvae of S. sphyraena 

 (Vialli, 1956) and S. argentea (Orton, 1955) 

 and both have melanophores in the finfold. 



Eggs of S. pinguis (0.69 to 0.82 mm diameter) 

 were considerably smaller than those of S. sphy- 

 raena (1.11 to 1.15 mm), S. borealis (1.22 to 

 1.24 mm), and S. argentea (about 1.5 mm). It 

 is not surprising that at hatching, larvae of S. 

 pinguis were smaller than those of S. borealis, 

 measuring 1.75 mm as compared to 2.6 mm. 

 Both species were similarly pigmented at hatch- 

 ing, but differed significantly by 45 and 66 hr. 

 Melanophores were larger and more concen- 

 trated in S. pinguis larvae than in S. borealis at 

 the same stage of development. 



Older larvae of S. borealis and S. sphyraena 

 had similar pigment patterns, but those of S. 

 sphyraena had a more developed lower jaw tip. 

 Three-day-old larvae of S. argentea from Cali- 

 fornia waters had a distinctive band of melano- 

 phores on the tail portion of the body just pos- 



terior to the anus that was lacking in S. borealis. 

 Postlarvae of S. barracuda were described and 

 illustrated by de Sylva (1963) . They diflfer sub- 

 stantially from S. borealis in being deeper bodied 

 and having a relatively longer snout. Pigmenta- 

 tion of S. barracuda and S. borealis larvae be- 

 tween 5.5 and 11.9 mm SL also differs somewhat 

 in its detail. 



BEHAVIOR 



Newly hatched larvae of S. borealis drifted 

 about the rearing tank making only occasional 

 feeble swimming attempts when disturbed. 

 At 2 days after hatching, larvae maintained a 

 horizontal position and began swimming actively 

 with short darting motions. Feeding activity 

 was first observed 3 days after hatching. Sennet 

 larvae usually fed using the S-flex behavior pre- 

 viously described for many species of clupeiform 

 larvae (e.g., Rosenthal, 1969; Schumann, 1965). 

 Occasionally, however, a sennet larva would 

 strike at a food organism without first flexing 

 its body and examining the item. Zooplankton 

 organisms less than 100 yu, in body width were 

 the initial food of larvae. No stomach analyses 

 were carried out on sennet larvae, but most of 

 the food which was placed in the tank were cope- 

 pod nauplii and copepodites. 



Sennet larvae continued to feed on small zoo- 

 plankton organisms until 10 days ^fter hatching, 

 even though larger plankton, including fish 

 larvae, and nauplii of brine shrimp were present 

 in the tank beginning 7 days after larvae hatched. 

 At 10 days, large plankton (about 300 to 400 [x 

 body width) was accepted as food, as were some 

 unidentified, newly hatched fish larvae about 2 

 mm in length. A sennet larva would approach a 

 tiny fish larva, assume an <S-flex position and dart 

 out at the larva, usually seizing it crosswise. Be- 

 fore swallowing it, a sennet would shake the 

 larva and turn it so that its long axis was par- 

 allel to the sennet's alimentary tract. Sennets 

 would swallow fish larvae either head or tail 

 first. Fish larvae were the preferred food of 

 sennets longer than 9 mm SL, but this food 

 was not provided in sufficient quantities to satis- 

 fy the sennets. Cannibalism was not observed, 

 but sennets became aggressive toward one an- 



194 



