Problems Identifying Tuna Larvae Species 

 (Pisces: Scombrldae: Thunnus) 

 from the Gulf of Mexico 



William J. Richards 



Southeast Fisheries Science Center, National Marine Fisheries Service, NOAA 

 75 Virginia Beach Drive, Miami, Florida 33149 



Thomas Potthoff 



Narragansett Laboratory, Northeast Fisheries Center 



National Mariane Fisheries Service. NOAA 



South Ferry Road, Narragansett, Rhode Island 02882 



Jong-man Kim 



Korea Ocean Research and Development Institute 

 Ansan, P O Box 29, Seoul 425-600, Korea 



The identification of tuna larvae of 

 the genus Thunnus is very difficult. 

 Five species of the genus Thunnus 

 may spawn in the Gulf of Mexico: 

 albacares, obesus, alalunga, thi/nnus, 

 and atlanticus. Criteria for their 

 identification have been established 

 by Matsumoto et al. (1972), Potthoff 

 (1974), and by Richards and Pott- 

 hoff (1974). A combination of pig- 

 ment patterns and osteological chai-- 

 acters is the criteria used. Pigment 

 patterns are used for larvae in the 

 3mm-NL to lOmm-SL size range, 

 and osteological characters are used 

 in larvae larger than 5 mm NL just 

 during and after notochord flexion. 

 Larvae larger than 7 mm SL are 

 rare and pigment characters are not 

 as useful since juvenile pigmenta- 

 tion begins to obscure the larval 

 pigmentation, and larvae less than 

 5 mm NL usually do not have suffi- 

 cient cartilage or bone development 

 to determine vertebral counts or 

 first closed hemal arch location. The 

 combination of both character sets 

 does allow for reliable identification 

 in most cases. Clearing and staining 

 larvae and the subsecjuent osteo- 

 logical examination are very time- 

 consuming, whereas pigment pat- 

 tern analysis is quite fast during 

 normal larval ichthyoplankton iden- 



tification. We conduct extensive ich- 

 thyoplankton surveys of the Gulf of 

 Mexico annually (Richards et al. 

 1984, Kelley et al. 1986), and it is 

 impractical to clear, stain, and per- 

 form osteological examination on all 

 Thunnus obtained. In this study we 

 cleared, stained, and osteologically 

 examined all Thunnus larvae col- 

 lected in 1982 (Richards et al. 1984) 

 to determine a baseline value for 

 species occurrence. 



Methods 



A total of 362 Thunnus (4.5 mm 

 NL-9.1 mm SL) larvae were ex- 

 amined for the presence of black 

 pigment on the upper and lower jaw 

 tips, along the ventral midline of the 

 tail, along the dorsal midline, along 

 the lateral midline, and on the cau- 

 dal fin. Since these specimens had 

 been fixed in formalin and pre- 

 served in ethanol, red pigments 

 were no longer present. All Thun- 

 nus larvae taken during 1982 and a 

 few from 1983 were utilized, except 

 for those which were typical T. 

 thynnus (possessing pigment on 

 both the dorsal and ventral margins 

 of the tail) or those specimens less 

 than 4.4 mm NL. Following pig- 

 mentation examination, specimens 



were cleared and stained following 

 Potthoff (1984), and vertebral counts 

 and position of the first closed 

 hemal arch determined. Species 

 identification followed the criteria 

 established by Matsumoto et al. 

 (1972), Potthoff (1974), and by 

 Richards and Potthoff (1974) for 

 North Atlantic Thunnus species 

 (Table 1). 



Results 



Of the 362 larvae, 95 larvae will not 

 be considered further because 4 

 were typical thynnus. 11 had no tail 

 pigment and were too small for 

 osteological data, 61 had ventral tail 

 pigment and were too small for 

 osteological data, 3 were not Thun- 

 nus. and 16 were lost or badly dam- 

 aged during the clearing and stain- 

 ing process. Of the remaining 267 

 larvae, 241 could be reliably iden- 

 tified (Table 2). Most of the" larvae 

 consisted of the atlanticus morph 

 with ventral tail pigment (72.7%), 

 although a few (6.0%) lacked this. 

 Those with ventral tail pigment are 

 inseparable from obesus except by 

 osteological criteria, and those with- 

 out it are inseparable from alba- 

 cares. Only 4.9% were obesus, only 

 6.0% were albacares, and 2 larvae 

 (0.7%) were alalunga. The remain- 

 ing 26 larvae (9.7%) cannot be posi- 

 tively identified (Table 3). These 

 data demonstrate the variability 

 which can be found in these species. 



Discussion 



When Richards and Potthoff (1974) 

 made a similar study, their speci- 

 mens were mainly from the Carib- 

 bean Sea and Straits of Florida. 

 They also found that the abundant 

 species was atlanticus, with few 

 albacares and obesus larvae con- 

 firmed in the western Atlantic. 

 These results confirm this same 



Manuscript accepted 30 January 1990. 

 Fishery Bulletin. U.S. 88:607-609. 



607 



