SMITH: SPAWNING BIOMASS OF NORTHERN ANCHOVY 



Table 7. — A comparison of the regional census estimate (A), separately weighted 

 nearshore (B), and 1 month per quarter sampling (C), in the southern California 

 inshore region, 1951-60. 



* February 



B r = 0.85 

 s 



C r = 0.90 



to biomass estimation is solely a function of the 

 variability these biases cause with respect to 

 year-to-year differences between the capture and 

 retention of sardine and anchovy larvae. For ex- 

 ample, in Figure 5 the product moment regres- 

 sion coefficients, r for sardine %%% or larval cen- 

 sus estimate versus sardine biomass are plotted 

 by total eggs, total larvae (1951-59) , total larvae 

 with outlyer censored (1953-59) , and each larval 

 size interval through 15.75 mm. All the coeffi- 

 cients are high and positive with pronounced 

 minima at the q%z stage and at the 9.75 mm stage. 

 One might ascribe the minimum associated with 

 eggs to the effect of "patchy" distribution on 

 precision of estimate. I have no ready expla- 

 nation of the 9.75 mm minimum or the 12.75 mm 

 maximum which follows it. 



Sette and Ahlstrom (1948, p. 521) discussed 

 the concept of "area of station" relative to sar- 

 dine eggs. Trial calculations of the same kind 

 suggest that with respect to sardine larvae, as- 

 signing equal areas to stations is also close to 

 the more exact method of erecting perpendicular 

 bisectors to each nearest station and using the 

 area of the polygon so formed. For anchovy lar- 

 vae, however, there may be an important prob- 

 lem. To study this problem, the southern Cal- 

 ifornia inshore region was divided into "near- 

 shore", i.e., the standard station closest to shore 

 on each line, and "nearshore-excluded" segments. 

 In the 40 quarters of the years 1951-60, the mean 

 concentration of eggs per positive station in the 

 "nearshore" region exceeded that of the "near- 

 shore-excluded" section by 35%. Similarly the 



859 



