HUNTER: BEHAVIOR OF LARVAL ANCHOVY 



Table 3. — Comparison of visual and photographic es- 

 timates of speed of intermittent swimming for anchovy 

 larvae of three sizes. 



Larval Tail beat 

 length amplitude^ 

 (cm) (cm) 



Intermittent 



speed estimates 



(cm/sec) 



Distance traveled* 

 (m/hr) 



Visual- Photographic' Visual Photographic 



0.5 

 1.0 



1.5 



0.197 

 0.282 

 0.367 



0.414 

 0.592 

 0.771 



0.304 

 0.823 

 1.342 



12.3 

 17.6 

 22.9 



9.0 

 24.5 

 39.9 



1 Estimated from J = 0.112 + 0.170Z.. 



a Estimated from JV^ = 0.047 + 1.308f, where F = 1.57. 



3 Estimated from y = —0.215 + 1.038i. 



* When 82.6% time spent swimming. 



did not change with age over the first 30 days 

 of larval life. That tail beat frequency did not 

 change implied that the interval of rest between 

 beats, the principal determinant of frequency, 

 was dependent on variables unrelated to size or 

 development. The same conclusion was obtained 

 from analysis of the structure of intermittent 

 swimming in the preceding section. Thus, the 

 change in speed associated with increased length 

 or age may be a function of only the increase 

 in tail beat amplitude with length. 



The mean tail beat frequency for intermittent 

 swimming, 1.57 ± 0.03 beats/sec, was substi- 

 tuted into the speed equation for intermittent 

 swimming. Speed of intermittent swimming 

 could then be determined for larvae of any length 

 by substitution of the appropriate tail beat am- 

 plitude into the equation. Speed estimates in 

 which the above procedure was used and ones 

 based on photographic analysis are compared in 

 Table 3. The two sets of estimates are reason- 

 ably close for 0.5-cm larvae but they diverge 

 for larger ones. No reason exists to disregard 

 either set of estimates. It seems reasonable to 

 assume that the true values lie somewhere be- 

 tween them. 



The proportion of daylight hours devoted to 

 intermittent swimming must be considered to 

 estimate the distance traveled per hour. Visual 

 observations are preferable for this purpose be- 

 cause of the greater number of observations 

 (318) and because visual observations were 

 systematically taken at different times of day. 

 Between ages 4 and 30 days, no trend with age 

 existed in the proportion of time devoted to 

 swimming although time spent swimming de- 

 creased slightly on days of intensive feeding. 



The mean proportion of time devoted to inter- 

 mittent swimming was 82.6 ± 1.2%. To arrive 

 at this estimate I considered periods of inactivity 

 longer than 5 sec as rest and periods equal to or 

 less than 5 sec as a part of intermittent swim- 

 ming bouts. Estimates of the distance traveled 

 per hour, assuming 82.6% of the time is spent 

 swimming, are shown in Table 3. These values 

 will be combined with others to estimate rate 

 of food search in a later section. 



FEEDING BEHAVIOR 

 DESCRIPTION OF FEEDING BEHAVIOR 



After a larva sighted a prey, the head turned 

 toward it so that the prey was perpendicular to 

 the tip of the snout and thus in about the center 

 of the binocular field of the larva. Then, while 

 keeping the prey in the center of the binocular 

 field, the larva swam slowly toward the prey by 

 executing one or more tail beats. After swim- 

 ming ceased, the larva contracted its body into 

 an S-shaped striking posture typical of the lar- 

 val clupeoid fishes (Figures). During contrac- 

 tion of the body, the prey was maintained di- 

 rectly in front of the snout and small movements 

 by the prey were compensated for by slight ad- 

 justments in the orientation of the head and 

 larger movements by rotating the entire body 

 with the pectoral fins. The larva continuously 

 moved toward the prey while forming the strike 

 posture by high frequency (50 to 60 beats/sec), 

 low amplitude vibration of the finfold or caudal 

 fin. 



The order and rate at which portions of the 

 body were contracted to form the S-strike pos- 

 ture were not fixed. The order appeared to be re- 

 lated to the initial orientation of the head and 

 trunk. Frequently swimming movements were 

 integrated into the beginning of the strike pos- 

 ture. The larva while approaching a prey often 

 ceased swimming with the body partially bent, 

 and the contractions to form the strike posture 

 were carried onward from that point. Varia- 

 tions in the rate of contraction were related to 

 movements of the prey. If the larva did not keep 

 up with a moving prey, contraction of the body 

 was often interrupted and the incomplete pos- 

 ture held for an extended period. The ampli- 



827 



