FISHERY BULLETIN: VOL. 78, NO. 1 



015 



0.10 



O05 



S(0. 



Figure 3. — Fraction of time spent actively swimming, ^, versus 

 oxygen concentration in percentage of saturation. S[02], for 

 newly hatched (day 0) and 24-h-old (day 1) northern anchovy 

 larvae. Each point on the full curves is an average of between 20 

 and 25 individual 5-min observations. Bounds are standard er- 

 rors. Dashed lines indicate idealized model of constant fraction, 

 I,, at high S[02l and linearly increasing i at low SLOj]. 



The duration of bursts increased monotonically 

 as the oxygen levels decreased, while the number 

 of bursts dropped significantly to a minimum of 

 1/min at 60-80%, increasing sharply after that 

 (Figure 4). No satisfactory explanation has been 

 found for the drop in the number of bursts at 80% 

 of saturation. The main result illustrated by Fig- 

 ure 4 is that both the frequency and duration of 

 bursts increase markedly at low oxygen concen- 

 trations, both contributing to the increase in time 

 spent swimming. 



The center of gravity of anchovy larvae is in the 

 vicinity of the head and therefore they tend to be 

 oriented in an oblique head-dowm configuration 

 after swimming ceases. More mature larvae, 

 which have converted significant amounts of yolk 

 into denser tissue are negatively buoyant (Hunter 

 and Sanchez 1976) and tend to sink head down- 

 ward at rates of approximately 1-2 mm/s. To check 

 the vertical station-keeping hypothesis, the direc- 

 tion of swimming was recorded, as well as the body 



40% 60% 



sLo^] 



Figure 4. — Average duration (t) of swimming bouts (full lines) 

 and number of bouts (n) in 5 min (dashed lines) for day (solid 

 triangles) and day 1 (solid circles) larvae, versus oxygen in per- 

 centage of saturation concentration (SIOjl). 



orientation, when swimming started. The results 

 of over 1 ,400 recorded swimming periods appear in 

 Table 1, which lists average values of the body 

 angle at the onset of swimming and the direction 

 of swimming. 



No significant variation in swimming direction 

 with oxygen concentration was found for either 

 day or day 1 larvae (Table 1). The spread in 

 results was large, as is noticeable from the stan- 

 dard errors. The total possible spread of data is 

 ±90°, which suggests that swimming direction is 

 actually a random phenomenon for day larvae. 

 At age 1 day, a positive bias was observed in the 

 swimming direction, still with large variation. 

 The body inclination at the beginning of the 

 swimming periods was consistent, at around -65° 

 with the exception of the day 0, 20% oxygen data, 

 which is influenced by additional factors, dis- 

 cussed below. 



DISCUSSION 



The analytical model predicted that a motion- 

 less larva would be able to pick up oxygen at a 

 decreasing rate at any given spot (Equation 12). 



Table l. — Initial orientation of the body and duration of swimming during bouts of continuous sv«mming by newly hatched and 

 1-d-old northern anchovy larvae. Error bounds are standard error. Angles are measured from the horizontal. Positive values indicate 

 upward motion. Averages of day do not include 20% O^ values as these (indicated by question marks) include different phenomena. 



114 



