pension of copepods was added to the bowl, 

 the larvae scattered, but within 1 or 2 minutes 

 they reformed in a tighter school and swam 

 around the container at a faster rate than pre- 

 viously. Feeding began when an individual 

 larva swam away from the school, slowed 

 down, oriented slightly below and about 10 mm. 

 away from a suspended copepod, flexed its body 

 in an S-shaped position (fig. 14), and drifted 

 within 2 to 5 mm. of the prey; then, with a 

 sudden flip of its tail, the larva straightened 

 out as it lunged forward and upward to engulf 

 the copepod. After the "strike," momentum 

 carried the larva a few millimeters beyond the 

 point of capture, from whence it swam vigor- 

 ously downward, completing an arc, and then 

 returned to the original swimming depth be- 

 fore finally rejoining the school. The time re- 

 quired for an individual feeding act and re- 

 turn to the school was 3 to 5 seconds. If the 

 copepod were missed, or only partly engulfed 

 on the first try, the larva made several violent 

 lunges in an attempt to seize it. If these at- 

 tempts were unsuccessful, the larva swam 



larvae had stopped feeding (usually within 20 

 minutes), they were siphoned from the bowls, 

 rinsed in fresh, nonradioactive sea water, and 

 placed in separate 100-ml. beakers of fresh sea 

 water. The radioactivities of each lot of cope- 

 pods, of each larva (after rinsing) , and of the 

 sea water in each 100-ml. beaker were meas- 

 ured with a gamma-ray scintillation detector 

 and a well-type 76-mm. sodium iodide crystal. 

 The number of copepods ingested by an indi- 

 vidual larva was estimated by dividing its net 

 activity gain (counts per minute) after feed- 

 ing by the mean count per copepod for the lot 

 fed (attempts to feed an individual larva a 

 single, radioactive copepod were unsuccessful). 

 The presence of copepods in the alimentary 

 tract was also verified visually at the time a 

 larva was transferred to a 100-ml. beaker. Up- 

 take and initial distribution of Zn''^ in the body 

 of larvae were followed over a 3V2-day period. 

 When a larva was killed (or died), the entire 

 alimentary tract was removed, and its activity 

 level and that of the remaining body tissues 

 were determined. 



Capture and preservation of larvae. — 

 Three lots, each with eight larvae, were trans- 

 ferred to separate 500-ml. bowls and fed Ar- 

 temia nauplii. After the larvae had stopped 

 feeding, four of them were individually re- 

 moved from one bowl with a 50-mm.-diameter 

 nylon mesh dip net and placed in solutions of 

 0.4, 10, and 50 percent Formalin and saturated 

 chloral hydrate respectively. The remaining 

 four were first anesthetized by adding a solu- 

 tion of tricaine methanesulfonate (0.25 g./l.) 

 to the bowl and then transferred in a similar 

 manner. Larvae in the second and third bowls 

 were divided and transferred to the same four 

 solutions from bowl two by siphoning with a 

 25-mm.-bore plastic tube and from bowl three 

 by dipping with a 100-ml. beaker. 



FEEDING BEHAVIOR AND RESPONSES OF 

 LARVAE TO SELECTED STIMULI 



Feeding 



Atlantic menhaden larvae are selective car- 

 nivores and voracious feeders. On transfer to a 

 bowl, they schooled and began to swim slowly 

 around the container. When an aqueous sus- 



FOOD OF YOUNG ATLANTIC MENHADEN IN RELATION TO METAMORPHOSIS 



Figlike 14. — Atlantic menhaden larva in the act of cap- 

 turing a copepod. The body is flexed in an S-shaped 

 attitude about 10 mm. from the prey and ingestion 

 accomplished by a forward darting movement. 



507 



