FISHERY BULLETIN: VOL. 72, NO. 2 



except T. atlanticus, have 22 anal fin elements 

 (15-7) but only 20 pterygiophores. Thunnus 

 atlanticus has 21 fin elements (14-7) but only 

 19 pterygiophores. As in the second dorsal fin, 

 all the anal fin rays and finlets have a bifurcate 

 base that contains a small distal radial (Figure 

 9B, C). The last anal finlet, as well as the last 

 dorsal finlet, is serially associated with a pterygio- 

 phore but has a secondary association with a small 

 bone that I assume to be a greatly reduced 

 proximal radial of a lost finlet. 



Pterygiophore development is more or less 

 synchronous with fin spine and ray development. 

 My smallest 8-mm-SL Thunnus specimens lacked 

 about two to four posteriormost pterygiophores 

 dorsally and ventrally (see earlier fin sections). At 

 13 mm SL, all specimens had acquired a full count 

 of pterygiophores. Thunnus atlanticus has a full 

 count at the slightly smaller size of 11 to 12 

 mm SL. In the very small sizes of 8 to 11 mm 

 SL, the pterygiophores cannot be differentiated 

 into proximal, middle, and distal radials. At about 

 11 mm SL, differentiation first begins under the 

 anteriormost section of the first dorsal fin. A few 

 of the distal radials begin to separate from the 

 main mass of the pterygiophores in tiny nonalate 

 blocks. The separation and development sequence 

 is in a posterior direction. At 20 to 35 mm SL, all 

 the distal radials are separated under the first 

 dorsal fin, are well ossified, and gradually assume 

 an alate shape. 



The middle radials under the finlets (usually 

 eight dorsal and seven ventral) begin to separate 

 and ossify over a great size range. Some speci- 

 mens showed no separation at 30 mm SL, whereas 

 others showed some separation and ossification as 

 small as 22 mm SL. The middle radials of the 

 finlet pterygiophores separate and develop in an 

 anterior direction. First to develop are the 

 posterior middle radials of the last finlets. Most 

 specimens had all their ventral middle radials 

 developed by 40 mm SL, but a few still lacked 

 the first (e.g., anteriormost) dorsal middle radial. 

 By 60 mm SL, all specimens had their eight 

 dorsal middle radials developed. 



In juvenile Thunnus it is difficult to determine, 

 externally, the exact number of finlets because of 

 the gradual intergradation from fin ray to finlet. 

 A finlet can now be precisely defined as having 

 a middle radial serially associated with it. I think 

 that at 50 mm SL one can, with some certainty, 

 count middle radials to determine finlet number. 

 Caution is warranted, however, because a few 



o 



t=::\ 



Figure 9. — Schematic representation of the relationship 

 between fin elements and distal radials (anterior view) in 

 juveniles of Thunnus. A. First dorsal fin spine and distal 

 radial near base; B. Second dorsal fin ray and distal radial 

 within bifurcate base; C. Finlet and distal radial within 

 bifurcate base. 



specimens may still lack the first (e.g., anterior- 

 most) dorsal middle radial. Table 7 has been 

 compiled from my specimens above 40 mm SL. 

 There was not enough material available in the 

 larger sizes to assess variability and specific 

 differences in the number of middle finlet radials. 

 The most common combination of middle dorsal 

 and ventral radials is 8/7. Variability from this 

 combination seems to be low in T. atlanticus and 

 T. alalunga and high in T. thynnus. 



Most of the interneural and interhaemal spaces 

 bounded by the neural and haemal spines are 

 occupied by one or more pterygiophores in 

 Thunnus (Figures 3 to 6). The association of 

 pterygiophores with the interneural and inter- 

 haemal spaces is limited to smaller juveniles. By 

 100 mm SL, the anteriormost pterygiophores 

 under the first dorsal fin are already situated 

 above the neural spines and do not insert into 

 any interneural spaces. As the juveniles grow, 

 more pterygiophores lose their association with 

 their respective interneural spaces. I did not 

 examine specimens larger than 117 mm SL and 

 cannot say if all pterygiophores, dorsally and 

 ventrally, pull away by the time the fish become 

 adults. For the identification of juveniles, how- 

 ever, the serially associated pterygiophores of the 

 second dorsal fin can be used successfully in 



576 



