GREENBLATT ASSOCIATIONS OF TL'NA WITH FLOTSAM 



Table 5. — Variability of fork length (centimeters) in single set 

 data by tuna species and set type, as indicated by average stan- 

 dard devation (Si and mean ix) and results of Kruskal-Wallis 

 one-way analysis of variance of standard deviations- (Source of 

 data: SWFC.i 



in catch rate, it was necessary to examine data 

 that could indicate which factors have most 

 influenced catch rate. Changes in attractability. 

 changes in residence time, technological ad- 

 vances, and increased knowledge of the fisherman 

 could not be rejected or confirmed with existing 

 data. It was possible, with some assumptions, to 

 determine if overall abundance changes or in- 

 creased skill of the fisherman could explain the 

 increased catch rate. 



If the overall abundance of tuna had increased 

 from 1963 to 1975, then one would expect the catch 

 rate to have increased correspondingly. If one ac- 

 cepts the supposition that flotsam-associated fish 

 were from the same population as unassociated 

 schoolfish. then one would expect the catch rate on 

 unassociated schoolfish to have likewise in- 

 creased. The catch per set on unassociated 

 schoolfish (Figure 6) showed no increase. How- 



ever, the year-to-year variations in catch per set of 

 flotsam-associated tuna and unassociated tuna 

 (Figure 6) were remarkably similar. The above 

 evidence indicates that changes in overall abun- 

 dance does not explain the long-term increase of 

 catch per set in flotsam-associated sets. The simi- 

 larity of the year-to-year variations supports the 

 hypothesized relationship between unassociated 

 schoolfish and flotsam-associated tuna. The year- 

 to-year variation may represent changes of abun- 

 dance. 



The second explanation for the changes in the 

 catch rate of flotsam-associated tuna schools, 

 changes in the skill of the fisherman, can be 

 evaluated by the percentage of successful sets. An 

 increase in the percentage of successful sets on 

 schools associated with flotsam would explain the 

 apparent increase in catch per set. Such an expla- 

 nation is untenable because the percentage of suc- 

 cessful sets would have had to more than double. 

 Greenblatt (1977) calculated the percentage of 

 successful sets associated with flotsam for 1974- 

 76, finding the average percentage of successful 

 sets to be 75'7f . To account for the change in catch/ 

 set, the percentage of successful sets in 1963 would 

 have had to have been about 35%, an unreason- 

 ably low figure. Pella and Psaropulos ( 1975, fig. 2) 

 showed that the percentage of successful sets of 

 unassociated schoolfish sets and flotsam- 

 associated sets ( considered as one category ) did not 

 increase enough in 1961-71 to account for the ob- 

 served changes in catch per set. Bayliff and 

 Orange (1967) reported percentages of successful 

 sets stratified by set type for a limited area of the 



FIGURE 7,— Distribution of catch by 5 

 short ton intervals for flotsam- 

 associated sets and unassociated 

 schoolfish from 1974 through 1976, 

 Solid boxes indicate percentage of zero 

 catch sets. The average catch per suc- 

 cessful set (CPSSi in short and metric 

 tons is given for each category. Metric 

 tons are in parentheses. 



1974 

 UNflSSOCiaTED SCHOOLFISH 



CPSS = 2071 

 llB7e) 



IrlThrrPn 



Iw 



OU- 



25 i.0 75 >iOO 



FLOTSftM ASSOCIATED 

 U-- 265 

 CPSS = 24 07 

 {21 63) 



1975 

 UNflSSOCIATEO SCHOOLFISH 



CPSS = 16 07 

 (14 5B) 



I hi n n r 



1975 



FLOTSAM ASSOCIATED 



N^20e 



CPSS = 27 30 



124 76) 



njfrh-rw. 



1976 



UNASSOCIATED SCHOOLFISH 



N = 98l 



CPSS = 21 21 



(19 24) 



25 50 75 >I00 



\m. 



_a — q 



1 



SHORT TONS 



1976 



FLOTSAM ASSOCtATED 



N= 550 



CPSS = 26 04 



(23 621 



1 1 1 nhp-ii-n r 



25 50 7l> 



SHORT TONS 



153 



