POTTHOFF: OSTEOLOGICAL DEVELOPMENT AND VARIATION IN YOUNG TUNAS 



prezygapophyses on the centra, T. thynnus and 

 T. obesus have them near the centra but on the 

 haemal arches, and T. albacares and T. atlanticus 

 have them well ventrad of the centra on the 

 haemal arches (Gibbs and Collette, 1967). In 

 juveniles below 50 mm SL, this specific difference 

 for the adults is not apparent because the more 

 anterior haemal prezygapophyses arise at about 

 the same position on the anterior parts of the 

 haemal arches for all species. In specimens 

 larger than 50 mm SL, the differences between 

 the species can be gradually perceived with 

 increased size. 



In my collection not even the largest juvenile 

 T. alalunga (91 mm SL) had its anterior haemal 

 prezygapophyses on the centra; instead, they were 

 close to the centra on the haemal arches as in 

 T. thynnus. The largest juvenile T. atlanticus 

 (87 mm SL) had the anterior haemal prezyga- 

 pophyses well ventrad on the haemal arches. 

 Consequently, large juveniles of T. alalunga and 

 T. thynnus cannot be separated by the position 

 of their anterior haemal prezygapophyses, but 

 large juvenile T. atlanticus can be separated 

 from T. alalunga and T. thynnus because the shift 

 of the more anterior haemal prezygapophyses to 

 ventrad is accomplished at a smaller size in 

 T. atlanticus than the shift dorsad to the edge 

 of the centra in T. alalunga. Lack of larger 

 Thunnus spp. specimens prevents their separa- 

 tion by this character. I am certain that large 

 juveniles (>100 mm SL) of the Thunnus spp. 

 complex could be separated to the species on the 

 basis of the position of the anterior haemal 

 prezygapophyses since differences have been 

 observed on adult T. albacares and T. obesus as 

 mentioned above. 



Ventrolateral Foramina 



(Figures 3 to 6; Table 3) 



The ventrolateral foramina are the last gross 

 anatomical feature to develop on the vertebral 

 column. They begin to form as projecting bony 

 bridges from the anterior edges of the haemal 

 postzygapophyses to the posterior edges of the 

 haemal arches in specimens from 19 to 22 mm 

 SL. On rare occasions 17-mm-SL fish may show 

 beginning development of the structure. The very 

 first developing ventrolateral foramen can be 

 located beneath any vertebra from the 27th to 

 30th, but it is most often found beneath the 28th. 

 During growth, new structures are added beneath 



the centra, anterior and posterior to the first 

 foramen. All of the ventrolateral foramina are 

 developed at about 25 mm SL, generally from the 

 22nd to the 36th vertebra. There is considerable 

 overlap for all the species with some modal 

 separation for specimens greater than 25 mm SL 

 in the vei-tebral position of the first ventrolateral 

 foramen. The ventrolateral foramina can be found 

 posteriorly to the 36th vertebra on all 25- to 

 35-mm-SL specimens of all species. After 35 mm 

 SL, some posterior openings are gradually filled 

 in by ossification. Juvenile T. thynnus from 71 to 

 117 mm SL had the last ventrolateral foramen 

 on the 32nd and 33rd vertebra, T. alalunga from 

 54 to 91 mm SL on the 31st, and T. atlanticus 

 from 52 to 87 mm SL on the 29th or 30th. 

 My largest Thunnus spp. specimen (47 mm SL) 

 had the last ventrolateral foramen on the 34th 

 vertebra. In adult Thunnus the last foramen is 

 found on the 29th to 30th vertebra according to 

 my own examination of seven skeletons of adults 

 (400 to 700 mm SL). Gibbs and Collette (1967) 

 found it on the 29th to 33rd vertebra. 



All species of Thunnus have initially circular 

 or nearly circular shaped openings, which 

 gradually decrease in diameter posteriorly. Ju- 

 veniles larger than 60 mm SL lose the circular 

 shape on the anteriormost foramina, acquiring a 

 more triangular or oval shape. Specific dif- 

 ferences in the size and shape of the anterior 

 openings are not distinct for juveniles as they 

 are for adults (Gibbs and Collette, 1967). 



First Haemal Spine 



(Figure 7) 



The first haemal spine is located on the haemal 

 arch of the first caudal vertebra. It is a consider- 

 ably elongated process. Anterior to this process, 

 in close proximity, are the first two anal pterygio- 

 phores (Figures 3 to 6). The preceding haemal 

 processes on the haemal arches of the precaudal 

 vertebrae are shorter than the first haemal spine 

 and also have at their tips flattened parapophy- 

 ses for rib articulation. Yabe et al. (1958), 

 Matsumoto (1963), and Yoshida (1965) reported 

 on the flattening of the first haemal spine in T. 

 alalunga and stated that this character may be 

 unique to the species. I believe that this is true 

 for the adults but not for the juveniles because 

 in all Thunnus species the first haemal spine 

 goes through a variety of flattened shapes during 



571 



