METHOT and KRAMER GROWTH OF NORTHERN ANCHOVY LARVAE 



Suppose variation in growth is such that indi- 

 viduals of age 25 days range in size from 11.4 to 

 16.6 mm and that the exponential growth rate 

 parameter which gives rise to this variation has 

 an uniform statistical distribution. If mortality 

 during this 25-day period is random with respect 

 to size, the mean size of individuals which survive 

 to age 25 days will be 13.86 mm. However, if the 

 instantaneous daily mortality rate is related to 

 size by Af = 3.5L " (this is consistent with current 

 estimates of egg, young larvae, and adult mortal- 

 ity rates for anchovy (MacCall 1974; Smith and 

 Lasker 1978)), the mean size of individuals which 

 survive to age 25 days will increase only slightly to 

 14.08 mm. We conclude that an overestimate of 

 growth because of size selective mortality is un- 

 likely to occur. 



Another possibility is that mortality is related 

 to growth rate rather than size, the mortal fraction 

 being that portion of the population which is slow 

 growing and weak and therefore more susceptible 

 to predators (Isaacs 1964). If we make the extreme 

 assumption that the daily mortal fraction is the 

 slowest growing lO^f ofthe cohort, the survivors at 

 age 25 days will be the fastest growing !"< ofthe 

 original cohort, but this extreme model unrealisti- 

 cally predicts that variations in size at age would 

 decline as the slow growing larvae die. Some in- 

 termediate degree of growth rate selective mortal- 

 ity could affect estimates of growth rate in the sea, 

 especially if mean growth rate is slow and the 

 slower growing individuals are near starvation. 



Net Avoidance 



Avoidance ofthe ring net by anchovy larvae in 

 daylight begins at a length of about 5 mm and 

 increases with size iLenarz 1973; Murphy and 

 Clutter 1972). The Bongo net catches larvae more 

 effectively but avoidance still occurs in the larger 

 size classes. The degree to which we underesti- 

 mated mean size at age depends upon how rapidly 

 avoidance increases with increasing size. We at- 

 tempted to minimize this bias by only considering 

 larvae with fewer than 25 increments ( lengths less 

 than about 15 mm) and samples taken at twilight 

 or dark. There was no difference in size at age 

 between the twilight samples (Dl. D5) and the 

 night samples (D2-D4) along the transect. Al- 

 though fast growth was observed only in samples 

 taken with the Bongo or neuston net (Bl, CI, C2), 

 we do not believe this was an artifact caused by 

 size-selective avoidance of the ring net. If this 



were an artifact, then ranges in size at age in 

 samples taken by the less selective gear would 

 have been broader and overlapped the distribution 

 of size at age in samples taken with gear which 

 allowed larger larvae to escape. This was not ob- 

 served. 



DISCUSSION 



Growth rates of larval northern anchovy < 1 mo 

 old were determined with size at age data. In 9 of 

 12 samples, growth rates at 8 mm were very simi- 

 lar, ranging from 0.34 to 0.40 mm/day. The growth 

 rate estimated from the combined data of these 

 samples was 0.37 mm/day. Growth in the three 

 other samples was faster. 0.47-0.55 mm/day. Var- 

 iation in size at age between individuals was 

 small. A typical 95% confidence interval for larvae 

 with 12 daily growth increments was 6.5-9.5 mm. 



When the trend of growth rate on temperature, 

 obtained from several rearing experiments, was 

 compared with the field results (Figure 3) it was 

 obvious that growth in the sea was similar to 

 growth in the laboratory. The range in growth rate 

 between field samples at the same temperature 

 was similar to the range in growth rate between 

 laboratory rearing experiments conducted at the 

 same temperature (Kramer and Zweifel 1970). 

 Variation of size at age was also similar in the sea 

 and the laboratory. In no case were sea-caught 

 larvae growing as slowly as larvae reared in the 

 laboratory on inadequate rations (LV, LD, F in 

 Figure 3). At 17.5° C (Lasker etal. 1970) anchovy 

 leirvae fed only a dinoflagellate grew at about 0. 15 

 mm/day (LD) and larvae fed dinoflagellate and 

 veligers ( LV) grew still slower than larvae fed wild 

 plankton (K) (Kramer and Zweifel 1970). Al- 

 though the availability of suitable prey may limit 

 the feeding success rate of first feeding anchovy 

 larvae (Lasker 1975), larvae which get enough 

 food to survive apparently get enough food to grow 

 rapidly. 



There was no obvious relationship between 

 growth rate and temperature. This is not surpris- 

 ing considering the narrow temperature range 

 considered, the variation in growth rate between 

 laboratory experiments at the same temperature, 

 and the uncertainty in our measurement of the 

 temperature experienced by the larvae in the sea. 

 The samples we examined came from near the 

 center of distribution of northern anchovy larvae 

 with respect to space, time, and temperature. As 

 samples from the periphery of this species range 



421 



