FISHER\ BULLETIN: VOL. 72. NO. 



its ontogeny. In addition, there is great indi- 

 vidual variability in the shape of the first haemal 

 spine within each species. Nevertheless, T. 

 alalunga exhibits the greatest flattening of the 

 first haemal spine during its ontogeny and Thun- 

 nus spp. the least. Thunnus thynnus and T. 

 atlanticus generally show less flattening than T. 

 alalunga but more than Thunnus spp. 



In T. alalunga the first haemal spine begins to 

 flatten a little at 21 mm SL. Flattening continues 

 to increase to about 35 mm SL. From this size on 

 changes in shape occur, but the degree of flatten- 

 ing remains essentially the same. Thunnus 

 atlanticus does not show any flattening before 

 60 mm SL and T. thynnus not before 30 mm SL. 

 Thunnus spp. showed slight flattening at 47 mm 

 SL. 



As the shape of the first haemal spine is a char- 

 acter of degree and cannot be accurately assessed, 

 I suggest that only persons familiar with the 

 Atlantic species of Thunnus juveniles in all 

 sizes use this character. 



Table 4. — Variability of spine and ray counts of the dorsal 

 and anal fins in the various species for juveniles of Thunnus. 



FINS AND FIN SUPPORTS 



First Dorsal Fin 



(Table 4) 



All species develop the full complement of 

 spines in the first dorsal fin before 8 mm SL. Four- 

 teen spines were regularly counted in the first 

 dorsal, even in the smallest specimens (8 mm SL). 

 The count of 14 spines is remarkably constant for 

 juveniles with a variability of 0% to 3% for the 

 various species. This remarkably constant count 

 of 14 can serve as a generic character to separate 

 juveniles of the genus Thunnus from other 

 scombrid genera in the Atlantic Ocean such as 

 Euthynnus (15-16), Katsuwonus (15-16), and 

 Auxis (10-12) (Potthoff and Richards, 1970); 

 Scomberomorus (15-19) (W. J. Richards, pers. 

 comm.)3; Scomber (9-13) (Matsui, 1967); Acan- 

 thocybium (24-26) (Rivas, 1951); Sarda (20-22); 

 and Orcynopsis (13) (Collette and Chao, 1973).^ 

 There is conflict between the consistency of first 

 dorsal fin counts in juveniles and greater varia- 

 tion of counts in adults (Frade, 1931; Rivas, 1951; 



^Southeast Fisheries Center, National Marine Fisheries 

 Service, NOAA, Miami, FL 33149. 



"Collette, B. B., and L. N, Chao. 1973. Systematics and 

 anatomy of the bonitos (Sarda and their relatives). Unpublished 

 manuscript. 



BuUis and Mather, 1956; Gibbs and Collette, 

 1967). In some adults the posteriormost spines 

 become embedded in the dorsal groove and sur- 

 rounding tissue and are consequently overlooked. 

 The first dorsal fin can be easily separated 

 from the second dorsal fin because the last spine 

 of the first dorsal is always shorter than the first 

 element in the second dorsal (Figures 3 to 6), and 

 the spacing between spines of the first dorsal fin 

 is greater than that between rays of the second 

 dorsal. The space between the last spine of the 

 first dorsal and the first element of the second 

 dorsal is wider than the following spaces between 

 the rays of the second dorsal. This diff'erence in 

 spacing is due to the shape, structure, and 

 spacing of pterygiophores, which support the 

 visible elements of the fins. 



Second Dorsal Fin and Finlets 



(Table 4) 



Eight-mm SL larvae of all species have already 

 acquired the full complement of rays in the second 

 dorsal fin, but lack two or three of the posterior- 

 most finlets. By 11 to 13 mm SL all finlets 

 are developed. Thunnus atlanticus develops its 

 second dorsal rays and finlets to a full comple- 

 ment at a slightly smaller size, usually by 10 



572 



