It should be emphaaized that catches made by longlLne fishing do not represent cross 

 sections of the population. Rather these catches, particularly in the instance of the tunas, repre- 

 sent the larger, older individuals in the population, and of course only that portion of the latter 

 that occupy subsurface levels. This topic is treated in some detail in the section immediately 

 following the discussion of numerical abundance. 



The analysis of the expeditions' catch records was facilitated by utilizing IBM punch 

 cards, requiring the adoption of a coding system. Preliminary inspection indicated that the 

 variation in catch rate was more closely related to latitude than to longitude, and for this reason 

 all catches were assigned to blocks of 1 of latitude and 5 of longitude. Temporally, the 

 catches were coded by month and by year. This grouping has the disadvantage of obscuring in- 

 dividual catches, but it would have been impracticable to handle the approximately 6, 000 in- 

 dividual daily catch records without it. 



Variation in the catch with longitude and time 



We shall first consider the variation in the longline catch rates of yellowfin, bigeye, 

 and black marlin as they are related to longitude and time, deferring consideration of variation 

 with latitude. Working statistical control of latitude was achieved in the following manner. 

 Trial latitudinal plots of individual longitudes indicated that the catch rates were relatively con- 

 stant between 1 and 5 N. latitude, in that there was relatively little variation within these 

 bounds as compared to the variation between this zone and the area to the north. Confining the 

 consideration of longitudinal and temporal variation in catch rates to the relatively restricted 

 latitudes bounded by 1 and 5 provided partial control of latitude. Additional statistical control 

 was effected by taking advantage of the fact that at nearly all longitudes there was some fishing 

 at each latitude between the limits of 1 and 5 N. This was done by computing the average 

 catch rate for each of the four degrees of latitude and computing the unweighted mean catch rate 



of the four samples. The end result was a catch rate representative of an area of the ocean 



o o 



surface that is 5 of longitude in length and 4 of latitude in height for 1 month of 1 year. 



The catch rates derived by this method are shown in table 4 arranged by species, 

 year, month, and longitude. Each of the three species --yellowfin, bigeye, and black marlin-- 

 will be considered separately because, as will be shown later, there is little or no relationship 

 between the presence of bigeye and the presence of yellowfin, and there is no reason to expect 

 the presence or absence of marlin to be in anyway related to the presence of the tunas. 



The catch rates of the yellowfin (table 4) show that during some months, e.g., July 

 of 1950, they were more abundant at the eastern end of the area than at the western end. During 

 other months, e.g., September 1951, the gradient appecirs to have been reversed. Of the three 

 months for which good records were obtained in both 1950 and 1951 (June, August, and Septem- 

 ber), two show a reversal of the gradients between the two years. For these reasons it seems 

 likely that the longitudinal gradients, even if real, are not caused by regularly recurring phenome- 

 na. 



Averaging all the unweighted monthly catches for each longitude appears to eliminate 

 these apparently transient differences associated with longitude. If this is done, the average 

 catches for the four zones of longitude esist of 150 E. , for which sampling is reasonably well 

 stratified, have a range of only 2.20-2.50 yellowfin per 100 hooks. If the area under considera- 

 tion is extended west to 140 E. (the samples to the west are not as well stratified as those to the 

 east of 150 E.), the range of variation is increased to only 1.84-2.50, or a 36-percent difference 

 between the highest and the lowest values. This is a remarkably constant abundance for any 

 species of animal to maintain over a 2, 000-mile range (140 -170 E. ), and probably means that 

 the environment is highly uniform over that range. 



