and Indian oceans. Populations of T. ionggol in the 

 Red Sea and western Indian Ocean appear to have 

 more gill rakers than those to the eastward. 



MORPHOMETRIC CHARACTERS 



Relative lengths of body parts have limited value 

 in species identification of tunas, although they have 

 been widely used by many investigators. Allometric 

 growth has been demonstrated for many, if not most, 

 body parts and has been responsible for manj^ mis- 

 conceptions. A classic example of allometry in- 

 volves the dorsal and anal fins of T. albacares which 

 become relatively much longer in large specimens. 

 Furthermore, the length attained by these fins varies 

 geographicall.y in a complex fashion (Royce, 1965). 

 Lack of consideration of these factors has resulted 

 in the description of numerous nominal species. 

 Many limited analyses have shown statistically sig- 

 nificant differences between populations of widely 

 distributed species (cf . Kurogane and Hiyama, 1957b, 

 1958a, 1959 for T. alahinga). 



Of the many measurements that can be made, 

 only the following appear to be of importance in 

 distinguishing tuna species: dorsal and anal fin 

 heights, pectoral fin length, placement of second 

 dorsal fin, greatest body depth, and diameter of eye 

 (or orbit). 



Our conclusions are based, as often as possible, on 

 data fiom many parts of the range of each species. 

 Where these data have shown no noteworthy geo- 

 graphic differences, we consider them together as a 

 single vmit. Otherwise, the differences are men- 

 tioned. In atklition to our original data, we have 

 leaned heavily upon Dung and Royce (195.3) for raw 

 data, and have used many other sources. Our in- 

 formation has significant gaps that can be filled only 

 by future work or by use of unpublished data from 

 other workers. All species are more or less deficient 

 in morphometric data for specimens below 400 mm. 

 For the Atlantic, such data are few or lacking for 

 T. alalunga less than 900 mm.; for T. atlanticus 

 larger than 650 mm.; and for T. obcsiis less than 

 1,000 mm. For the Pacific and Indian Oceans, T. 

 maccoijii is represented by data on only four speci- 

 mens outside the 6.50-750 mm. range, and our sparse 

 Indian Ocean data for all species are almost entirely 

 from specimens sent to us as a result of cruises of the 

 R/V Anton Bruun during the International Indian 

 Ocean Expedition. Our interpretations must be 

 judged with these deficiencies in mind. 



Heights of the second dorsal and anal fins have 



received much attention, especially in T. albacares, 

 in which the positive allometry of both these fins 

 relative to fork length, apparently characteristic of 

 all species, is most pronounced. This allometry has 

 led to the description of long-finned nominal species 

 in both the Atlantic and Pacific and even to the 

 establishment of a new genus {Semathunnus). The 

 species are compared in figs. 24 and 25, in which the 

 large T. albacares are western Atlantic specimens. 

 The few data for T. tnaccoyii fall in the range of T. 

 thijnniis and are not discussed separately. 



As many workers have shown and as Royce (1965) 

 has most recently demonstrated, different popula- 



FORK lENCTH 



Figure 24. — Relative height of second dorsal fin in Thunniis 

 species. Data, in addition to our own, include Dung and 

 Royce (1953: tables 27, 28, 42). 1— T. atlanticus; 2—T. 

 Ionggol; 3 — T. albacares; 4 — T. obesus; 5 — T. alalunga; 

 6 — T. thynnus. 



nil 



iiiiii « 



FORK LENGTH 



Figure 25. — Relative height of anal fin in Thuntius species. 

 Data, in addition to our own, include Dung and Royce 

 (1953: tables 27, 28, 42). 1— T. atlanticus; 2—T. tonggol; 

 3 — T. albacares; 4 — T. obesus; 5 — T. alalunga; 6 — T. thynnus. 



ANATOMY AND SYSTEMATICS OF TUNAS 



93 



