BAYLIFF: INTEGRITY OF SKIPJACK TUNA SCHOOLS 



between release and recapture. A description of how 

 this was done is given below. The data used in the 

 analyses are summarized in Table 1, and a more 

 detailed summary of the data for tagging cruise 

 1079, which will be discussed in more detail than the 

 other cruises, is given in Table 2. 



First, the units of time to be employed were 

 selected. Tagged skipjack tuna have been recaptured 

 after as long as 3 years at liberty, but the great 

 majority of recaptures have been made within 6 

 months after release. It was decided to examine the 

 data by monthly intervals because, for the experi- 

 ments with sufficient total numbers of returns, that 

 would produce several intervals vdth sufficient num- 

 bers of returns of tagged fish for each interval. Also, 

 the statistical and size-frequency data of the Inter- 

 American Tropical Tuna Commission (lATTC) are 

 routinely calculated by month (and quarter and 



Table 2— Data from cruise 1079 used for analysis of integrity of 

 schools of skipjack tuna. 



year), so no special calculations are required. 



Second, the areas of study were selected. If the 

 tagged fish are released at a particular location, the 

 vessels fishing near that location would catch more 

 tagged fish, at least during the first few months 

 after release, than would vessels fishing several hun- 

 dred or more miles away. Thus only data for the 

 vessels fishing near the location of release should 

 be considered. The areas of study were selected by 

 examining charts of the distributions of fishing ef- 

 fort and recaptures of tagged fish, by 1 -degree areas 

 and by months, and arbitrarily excluding those with 

 lower recaptures per unit of fishing effort. This was 

 quite simple, as during the periods in question prac- 

 tically no tagged fish were recaptured south of lat. 

 20°N, and there was little fishing effort north of lat. 

 20 °N outside the area- time strata selected. The 

 areas of study for the data for tagging cruise 1079 

 are shown in Figure 1. 



Third, a list of sets for each area-time stratum was 

 prepared. This included the weights of skipjack tuna 

 caught (Table 3, column 2) and the numbers of 

 tagged fish returned (Table 3, column 4). (Through- 

 out this report the weights are expressed in short 

 tons [0.907 metric tons] and pounds [0.454 kg]. The 

 lATTC uses this system because the fishermen 

 estimate the weights of the fish caught in individual 

 sets in short tons, and these estimates are an im- 

 portant component of its data base.) In a few cases 

 the catches were recorded only as weights of mixed 

 skipjack tuna and some other species, and in those 

 cases the weights were divided by 2, assuming that 

 they consisted of equal weights of skipjack tuna and 

 the other species. (The rest of this table will be dis- 

 cussed later.) 



It can be seen in Table 1 that ranges instead of 

 individual values are given for "Sets" for cases when 

 the month of recapture is the same as the month 

 of release. This is because sets made before the date 

 of release were not considered for the analyses, and 

 it was also decided not to use data for tagged fish 

 recaptured on the date of release because these 

 could not have mixed with the rest of the fish in the 

 area to any appreciable extent. Therefore for cmise 

 1079, for example, there were 297 June sets after 

 9 June, 294 after 10 June, 252 after 17 June, 240 

 after 18 June, and 232 after 19 June (Table 2). 



Fourth, an average weight for each month of re- 

 capture for each experiment was selected. Monthly 

 average weight data for purse seine- and baitboat- 

 caught fish from area 1 in Peterson (1982: fig. 30) 

 were used for this purpose because they closely cor- 

 respond to the strata selected for study. The aver- 



633 



