LENARZ ET AL.: YIELD PER RECRUIT OF ATLANTIC YELLOWFIN TUNA 



of the methodology (Tomlinson, 1970). Calcula- 

 tions of yield per reciniit using initial values 

 of F of 0.2 and 0.8 are shown in Figures 8 and 

 9 as functions of initial values of F, effort, and 

 size at recioiitment. The values of yield per 

 recruit do not vary significantly (<10%) with 

 changes in the initial values of F, and the rela- 

 tive values are quite similar. Values of size speci- 

 fic F are shown for each gear in Figure 10 when 

 initial values of F are 0.2 and 0.8. When the 

 initial value of F is 0.8. values of F for small 

 purse seiners increase sharply with size from 

 170 to 180 cm. This does not occur when the 

 initial value of F is 0.2. Intuitively we do not 

 expect an increase in F with size past 170 cm 

 and thus choose to use the results when the 

 initial value of F is 0.2 in the remainder of the 



0.5- 



7i- 



O INITIAL F = 0.8 

 • INITIAL F = 0,2 



I I I I 



J I 



10 15 20 2 5 



MULTIPLIER OF EFFORT 



30 



35 



Figure 8. — Yield-per-recruil (kg) of Atlantic yellowfin 

 tuna, when size at recruitment is 32.5 cm, as a function of 

 the multipUer of fishing effort. 



32.5 52.5 72.5 92.5 112.5 1325 



FORK LENGTH AT RECRUITMENT (cm.) 



Figure 9. — Yield-per-recruit (kg) of Atlantic yellowfin 

 tuna, with the current level of fishing effort, as a function 

 of length at recruitment. 



>- 



_l 

 < 



I- 



cr 

 o 



to 



3 



o 



< 



< 



O BAITBOAT 



• SMALL PURSE SEINE 

 A LARGE PURSE SEINE 

 A LONGLINE 



80 100 120 140 



FORK LENGTH (cm) 



Figure 10. — Estimates of size-specific instantaneous fish- 

 ing mortality coefficients (F) by gear type when initial 

 values of F are ( A) F = 0.2. ( B) F = 0.8. 



paper. Validity of the estimates of F depends on 

 the validity of the assumption that recruitment 

 has been fairly constant for the cohorts included 

 in the analysis. The special ICCAT working 

 group noted that the cohort which entered the 

 surface fisheries in 1969 appears to be weaker 

 than the following two cohorts (ICCAT, 1972. 

 see footnote 2). Although inclusion of 5 yr of 

 data in the analysis may minimize the source 

 of error, future studies should examine the sensi- 

 tivity of the results to errors of this type. 



Estimates of Yield Per Recruit 



Results of the yield-per-recruit calculations 

 using the estimates of size-specific F when the 

 initial value of F is 0.2 and with M = 0.8 are 

 shown by gear in Table 3. Yield-per-recruit 

 isopleths and the line of eumetric fishing (size 

 at recruitment, / ^, which maximizes yield per 

 recruit at a given effort) for the entire fishery 



47 



