FISHERY BULLETIN: VOL. 70, NO. 3 



PC 



1.000 



0.100 



0.010 



0.001 



.2 .3 .4 



Figure 10. — Alternative analysis of skipjack data from Figure 8, assuming constant 

 population density. Left panel: Three possible choices of fit between catch data and 

 theoretical curves of P^. at constant Xq/R. Curve A shows fit for best sampled class 

 intervals. Curve B shows fit for midrange class intervals. Curve C shows fit for regions 

 of maximum slope. Right panel : P^ curves from left panel, plotted on linear coordinates. 



speed assumptions, so the analysis yields usable 

 estimates of population abundance of skipjack 

 larvae and juveniles between 17.5 and 43.5 mm 

 in length, if population structure is not import- 

 ant. From curve B of Figure 10, No is 40 9^^ 

 (of 499) or 200 individuals per class interval, 

 as can be seen by placing the "9^ of catch" scale 

 of Figure 9 in the proper position on Figure 10. 

 Since abundance apparently does not seriously 

 affect the length-frequency curve, each of the 19 

 class, intervals sampled must also represent a 

 population of about 200 individuals. Thus the 

 sample of 510 skipjack was taken from a pop- 

 ulation of about 3,800, an overall catch efficiency 

 of 13%. The total catch resulted from 41 ef- 

 fective tows (out of 74), an average catch per 

 effective tow of 12.3. Accordingly, the popula- 



tion sampled by the Cobb trawl must have num- 

 bered roughly 94 skipjack per effective tow, to 

 yield the observed catches at 13% efficiency. By 

 far the greatest number of skipjack were caught 

 by the cod end of the trawl, so the volume ef- 

 fectively sampled was not more than 230,000 m"" 

 (100% filtering efficiency). This leads to an 

 estimated mean population density for larval and 

 juvenile skipjack of one fish per 2,500 m^ in 

 areas where skipjack were present during the 

 three Toivnsend Cromwell equatorial cruises. 



Since Pc is by definition a minimum probabil- 

 ity of capture, reference to equation (3) shows 

 that the above estimate of mean population den- 

 sity (during a tow) represents a maximum val- 

 ue. On the other hand, of course, the skipjack 

 probably were not randomly distributed, so that 



816 



