872 



Fishery Bulletin 100(4) 



Discussion 



Skeletal anatomy and morphometries 



We had to use skeletal anatomy, including the chondro- 

 neurocranium, palatoquadrate, and pectoral girdle for 

 unambiguous identification of the Sanzo (1912) embryo 

 after capture data and the vertebral count suggested that 

 the embryo might be a shortfin mako rather than a white 

 shark. Our attempts to use morphometries, dentition, and 

 DNA analysis were not successful. 



Sanzo (19121 correctly placed the embryo in the family 

 Lamnidae using only morphometric criteria. Only three 

 species of the family Lamnidae normally occur in the 

 Strait of Messina of the Mediterranean Sea: porbeagle, 

 white, and shortfin mako sharks (Compagno, 1984; Fergus- 

 son, 1996). However, the morphometric arguments used by 

 Sanzo (1912) for identification to species were not charac- 

 teristic, leading him to incorrectly eliminate the genus Isu- 

 rus. The upturned snout was probably caused by distortion 

 during preservation, and we have shown that the mouth 

 width-to-length ratio, eye shape, and relative positions of 

 the origins of second dorsal and anal fins are not suitable 

 criteria for distinguishing between white and shortfin 

 mako sharks. Other promising morphometries also failed 

 to distinguish between the two species. These conclusions 

 are tentative — confirmation will depend on obtaining mea- 

 surements of these characters from small embryonic white 

 sharks, which were missing from our database. 



Dentition 



Tooth shape is species-specific in postnatal Lamna, Isurus, 

 and Carcharodott (Compagno, 1984) but not in embryos. 

 Lamnid embryos have specialized "embryonic" teeth that 

 are adapted for grasping and tearing the membrane of 

 the eggcases on which they feed (Gilmore, 1993; Francis 

 and Stevens, 2000). We observed fanglike embryonic teeth 

 in porbeagle embryos, which lacked the characteristic 

 cusplets of adult specimens. We observed fanglike embry- 

 onic white shark teeth lacking serration in the intestine of 

 a nearterm embryo similar to the embryonic teeth of the 

 Sanzo embryo (Francis. 1996; Francis and Stevens. 2000). 

 Embryonic teeth are similar in all lamnid embryos and do 

 not appear to be suitable for identification. Shortfin mako 

 embryos shed their embryonic dentition at about 45-50 

 cm TL and nearterm embryos have emerging adultlike 

 teeth (Gilmore, 1993; Mollet et al.. 2000). 



It is difficult to describe a dentition completely without 

 the benefit of prepared jaws, particularly in a relatively 

 small embryo. In addition, the tooth formula of a small 

 embryo may be different from that of a postnatal speci- 

 men. Sanzo (1912) reported an upper jaw tooth formula of 

 10-0-10; we observed 14-0-14. the full adult complement of 

 replacement tooth files. That suggests that Sanzo (1912) 

 had not observed the four replacement teeth in files 11-14. 

 He reported two replacement teeth behind the functional 

 teeth in files 7-8 and 8-9. whereas we observed replace- 

 ment teeth in all position of two rows by pulling back the 

 dental lamina (Table 2). 



