FISHERY BULLETIN: VOL, 71. NO. 1 



Table 1. — Sources of length data of Atlantic yellowfin tuna caught in the surface fishery 



off Africa. 



Abidjan 



Dakar 



Gulf of 

 Guinea 



Polnte- 

 Noire 



Year 



Type of 



length 



measurement 



Type of vessel sampled 



Bait- 

 boat 



Small 

 seiner^ 



Large 

 seiner- 



Source 



X 

 X 

 X 



X 

 X 



X 

 X 

 X 

 X 

 X 



O.R.S.T.O.M., 1971 



O.R.S.T.O.M., 1971 

 X O.R.S.T.O.M., 1971 



Champagno-t and Lhomme, 1970 



Champagnot and Lhomme, 197C 

 X O.R.S.T.O.M., 1971 



X Staff, Tuna Population 



Dynamics Project, 1971' 



O.R.S.T.O.M., 1971 



Lo Guen et al., 1969 



O.R.S.T.O.M., 1971 



O.R.S.T.O.M., 1971 



Unpublished data (Le Guen) 



1- Small purse seiner = less than 500 metric tons capacity. 



^ Large purse seiner = larger than 503 metric tons capacity. 



* Staff, Tuna Population Dynamics Project. 1971. Size composition of the yellowfin and skipjack tuna purse 

 seine fishery off the west coast of Africa 1968-1970. Unpublished manuscript, 28 p. Southwest Fisheries Center, 

 National Marine Fisheries Service, NOAA, La Jolla, CA 92037. 



they could not be separated as such, they were 

 treated separately from the French data. 



Two methods were employed in our analysis. 

 One approach ("age unknown") was based on 

 all samples from the four regions, for years 1965- 

 70 and with age of size groups unknown. The 

 second approach ("apparent age known") was 

 slightly different. Only fish that were caught in 

 an area from Sao Tome to southern Angola, 1967- 

 71, and with the apparent age of each size group 

 known, were employed. 



Growth was estimated with the von Berta- 

 lanffy growth function. This function is often 

 expressed as, 



Lt = L, [1 — exp — K {t - UU, 



where Lt = length at age t, L^ = asymptotic 

 length, K = growth rate, and to = theoretical 

 ■age when Lt = 0. It is fitted to growth data 

 by various procedures (e.g., Walford, 1946; 

 Abramson, 1963; Ricklefs, 1967; Gulland, 1969; 

 Knight, 1969), most of them require data on 

 size at known age. A least-squares procedure 

 that does not contain this limitation was de- 

 scribed by Fabens (1965) . He fitted a von Bert- 

 alanflfy function of the form 



Lt + A = Lt + (L„ — Lt) (1 — exp - K) 

 to tag-return data, but his procedure is equally 



applicable to length observations of untagged 

 fish made at t and again at a later date, t + A, 

 when the age of the fish is unknown. For tuna, 

 Rothschild (1967) and Joseph and Calkins 

 (1969) employed Fabens' procedure to estimate 

 growth of skipjack tuna {Katsuivomis pelamis) 

 from tagging data. We used the Fabens' pro- 

 cedure with monthly mean lengths for individ- 

 ual year classes to estimate growth of yellowfin 

 tuna of unknown age. A computer program 

 written by Tomlinson (Abramson, 1971) was 

 employed to estimate L„ in centimeters and 

 K, expressed on a monthly basis. For growth 

 estimates based on apparent known age fish, we 

 used a computer program written by Abramson 

 (1963) and modified by Psaropulos (1966) of a 

 least-squares procedure described by Tomlinson 

 and Abramson (1961). 



ANALYSIS WITH UNKNOWN 

 AGE FISH 



METHODS 



Fish landed at Abidjan, Dakar, and Pointe- 

 Noire (Figure 1) were measured for predorsal 

 length (tip of snout to anterior base of the dor- 

 sal fin) by French scientists; fish were measured 

 for fork length by lATTC scientists. In order 

 to standardize the length measurements, we em- 



176 



