FISHERY BULLETIN: VOL. 70, NO. 3 



At intermittent swimming speeds tail move- 

 ment was slow enough that the frequency of the 

 beat could be counted by eye. If the length of 

 the larva were known, intermittent swimming 

 speed could be estimated from the general speed 

 equation given above, the amplitude equation for 

 intermittent speeds, and the tail beat frequency. 

 Instead of the general speed equation it may be 

 preferable to use one based on intermittent swim- 

 ming data alone, which is V/A = 0.0466 + 

 1.308 F where Si = 1.160. 



BURST SPEEDS 



The speeds obtained by larval anchovy during 

 bursts of fast continuous swimming are of in- 

 terest because they may be a measure of the 

 larva's ability to avoid predators or possibly 

 plankton nets. Bursts of fast swimming were 

 stimulated in larvae of mean length 4.1 ± 0.1 

 mm (Table 2, continuous swimming, 0.34 to 0.44 

 cm larvae) and mean length 12.1 ± 1.2 mm 

 (Table 2, continuous swimming, 1.00 to 1.42 cm) . 

 Bursts were stimulated by moving a pin near a 

 larva and photographing the larva at the film 

 speed of 200 fps. Speeds were averaged for the 

 entire duration of the burst, which lasted about 

 0.1 to 0.2 sec. The stimulated burst speed for 

 larvae of mean length 4.1 mm was 63 ± 19 mm/ 

 sec (mean ± 2 X SE) or about 15 body lengths/ 

 sec and that for the 12.1-mm larvae was 198 ± 

 mm/sec or about 16 body lengths/sec. The 

 maximum speed obtained by larvae in the 

 4.1 mm length class was swum by a 4.2-mm 

 larva which swam 116 mm/sec (28 body 

 lengths/sec) during a burst of 0.12 sec. The 

 maximum speed for larvae in the 12.1 mm class 

 was 313 mm/sec (25 body lengths/sec) swum 

 by a 12.4-mm larva during a burst lasting 0.10 

 sec. In summary, larvae had a maximum speed 

 capability of 25 body lengths/sec or faster for 

 bursts lasting 0.1 to 0.2 sec, but the typical or 

 average burst speeds were close to 15 body 

 lengths/sec. 



ESTIMATES OF DAILY RATE OF SWIMMING 



I shall consider only intermittent swimming 

 in the estimates of daily swimming rate because 

 only intermittent swimming was associated with 

 food search, and because once feeding began. 



continuous swimming was an insignificant pro- 

 portion of daily activity (Figure 1 and Table 1). 

 Two independent methods were used to estimate 

 the rate of intermittent swimming. In the first 

 method, larvae were photographed at 2 fps in 

 a cylindrical chamber of 25 cm diameter, and 

 swimming speed calculated from frame-by-frame 

 analysis of the photographs. Five minutes of 

 swimming were analyzed for each larva; swim- 

 ming was divided into intermittent and contin- 

 uous types; and periods of inactivity equal to or 

 greater than 5 sec were tabulated as rest. In 

 the second method, visual observations of larvae 

 in the 500-liter rearing tanks were used to re- 

 cord the tail beat frequency of larvae during 

 intermittent swimming and the frequency and 

 duration of continuous swimming and rest. 

 Larvae were observed three or more times a day. 

 The speed of larvae during intermittent swim- 

 ming was calculated from the tail beat frequency 

 by use of the relationships between age, length, 

 amplitude, tail beat frequency, and speed pre- 

 viously described. 



The speed of intermittent swimming mea- 

 sured from photographs increased with larval 

 length (Figure 4), and the regression of speed 

 on length gave the relationship V = — 0.215 + 

 1.038 L where Si = 0.280. Visual measurements 

 of tail beat frequency of intermittent swimming 



3.0r 



0.0 



0.5 



1.0 1.5 



LENGTH (cm) 



2.0 



Figure 4. — Rate of movement during intermittent swim- 

 ming (cm/sec) as a function of anchovy larval length 

 (cm) . Measurements taken from analysis of cine photo- 

 graphs. Equation for line is V = —0.215 + 1.038L. 



826 



