RICHARDSON: SPAWNING BIOMASS AND EARLY LIFE OF NORTHERN ANCHOVY 



beginning or end of the season, the biomass 

 estimates would be high. 



Differences in biomass estimates between the 2 

 yr based on eggs reflects the fact that more than 

 twice as many eggs were collected in 1975 than in 

 1976. This could be the result of a decrease in 

 spawning biomass between the years although 

 small sample size is probably just as important. 

 Biomass estimates based on eggs are likely to be 

 more variable than those based on larvae. North- 

 ern anchovy is a schooling fish; therefore eggs 

 released from spawning schools are clumped, re- 

 sulting in a sample of high variance (Pacific 

 Marine Fishery Management Council footnote 2). 



Simpson Egg Method 



Estimates of daily egg production, Equation (7), 

 44.68 X 10^1 in 1975 and 22.59 x 10^ in 1976 using 

 the Sette and Ahlstrom Census ( Table 3 ) and 35.02 

 X 10" and 17.83 x 10" using the Smith Census, 

 were plotted against the cruise middate. The area 

 under the resulting triangle was then equated 

 with egg production for the entire spawning sea- 

 son, 1,316.27 X 10" in 1975 and 708.87 x 10" in 

 1976 with the Sette and Ahlstrom Census and 

 1,061.76 X 10" and 614.50 x 10" with the Smith 

 Census. Biomass estimates were then obtained 

 with Equation (4). 



Spawning biomass estimates were 365,631 t in 

 1975 and 196,909 t in 1976 using the Sette and 

 Ahlstrom Census and slightly smaller with the 

 Smith Census (Table 3). These biomass values are 

 nearly one-half those obtained by the Sette and 

 Ahlstrom Egg Method. This reflects the different 

 assumption of this method regarding egg produc- 

 tion where it is high in mid season and low at both 

 ends. 



Saville Egg Method 



Egg production is assumed to follow a normal 

 curve throughout the 62 -d spawning season from 

 15 June to 15 August. Each cruise within that 

 period represents a proportion of the area under 

 that normal curve. In this study, each cruise was of 

 9 d duration and was made near the peak of the 

 spawTiing period. In 1975 the cruise represented 

 33.35% of the curve and in 1976, 30.81%. Seasonal 

 egg production was then estimated by Equation 

 (8) and biomass by Equation (4). 



Estimates of biomass using this method were 

 smaller, 334,951 t in 1975 and 183,289 t in 1976 



using the Sette and Ahlstrom Census, than in the 

 two previous methods (Table 3). If egg production 

 is skewed from a normal distribution, large errors 

 could be introduced into the biomass estimate. 



Smith Larva Method 



This method of estimating spawning biomass 

 assumes that a similar linear relationship exists 

 between numbers of larvae and adult spawning 

 biomass in both the central and northern subpopu- 

 lations. Although this assumption seems reason- 

 able, the recent fecundity estimate for northern 

 anchovy off California, 340 eggs/g total female 

 weight (Hunter and Goldberg 1980 1, is consider- 

 ably less than the fecundity estimate obtained for 

 northern anchovy off Oregon, 720 eggs/g total 

 female weight (Laroche and Richardson 1981). 

 These data indicate it would take more planktonic 

 eggs to represent 1 g offish in the northern than in 

 the central subpopulation. It follows then that it 

 would also take more pelagic larvae to represent 1 

 g offish in the northern subpopulation given that 

 spawning frequency and larval growth and mor- 

 tality conditions are similar. Thus, biomass esti- 

 mates obtained by this method may be too high. 



Other assumptions of the method seem reason- 

 able (see Methods section). The census estimate 

 obtained from one cruise during the shortened 

 spawning season in the north should be equiva- 

 lent to a quarterly census estimate obtained dur- 

 ing a peak spawning period in the south. Some- 

 times a quarterly estimate for CalCOFI is based 

 on only one cruise, other times a mean of several 

 cruises. However, my cruise dates were purpose- 

 fully selected at the time of peak spawning. Some- 

 times CalCOFI quarterly cruises may be con- 

 ducted near the beginning or end of a quarter and 

 not necessarily at the peak of spawning. Thus 

 larger numbers of smaller larvae may have been 

 collected in our study, and the relationship be- 

 tween numbers of larvae and biomass may be 

 biased to give a higher biomass estimate in the 

 north. 



In general, water temperatures at the time of 

 peak spawning appear to be similar off Oregon 

 and California (Ahlstrom 1959; Baxter 1967) so 

 that growth rates and length of time in the water 

 column are assumed to be similar. Methot (in 

 press) demonstrated that growth rates of larvae in 

 the two subpopulations are similar at similar 

 temperatures. 



Sampling techniques used in this study were 



869 



