Scoles and Graves: Genetic analysis of the population structure of Thunnus albacares 



691 



tances and between regions (Fink and Bayliff, 1970; 

 Bayliff, 1984; Itano and Williams, 1992). 



Population structure was indicated by investigations 

 of both meristic and morphometric characters which 

 revealed significant differentiation among yellowfin 

 tuna from the eastern, central, and western Pacific 

 regions (Schaefer, 1955; Kurogane and Hiyama, 1957), 

 as well as clinal character variation across the equato- 

 rial Pacific (Royce, 1964). Further investigation using 

 discriminant function analysis of morphometric vari- 

 ables suggested mixing occurs between morphologically 

 differentiated northern and southern yellowfin tuna of 

 the eastern Pacific (Schaefer, 1989), as well as across 

 the Pacific (Schaefer, 1991, 1992). 



Analysis of fishery data also suggests population 

 structuring of yellowfin tuna within the Pacific. Kami- 

 mura and Honma ( 1963) provided evidence for two or 

 more semi-independent subpopulations based on size 

 composition and catch data of equatorial Pacific yel- 

 lowfin tuna from longline landings. Suzuki et al. (1978) 

 examined longline and purse-seine length composition 

 data and suggested the existence of semi-independent 

 eastern, central, and western Pacific subpopulations. 

 Additionally, homogeneity within the western Pacific 

 was indicated by the widespread distribution of 

 fish contaminated by radioactivity resulting from the 

 1954 U.S. nuclear tests at Bikini Atoll (Suzuki et al., 

 1978). 



While the results of several analyses suggest yel- 

 lowfin tuna exhibit population structure within the 

 Pacific Ocean, genetic analyses have revealed no sig- 

 nificant genetic differentiation. Suzuki ( 1962) reported 

 that the blood agglutinogen Tg2 occurs in similar fre- 

 quencies in samples from the Indian Ocean and the 

 eastern Pacific. Additionally, allozyme analysis did not 

 reveal genetic differentiation between samples of yel- 

 lowfin tuna collected off Hawaii (n=529) and Baja Cali- 

 fornia (a; =207) at the polymorphic serum esterase lo- 

 cus, although overall variation was low (Fujino, 1970). 

 Preliminary evidence for frequency differences occur- 

 ring at two other loci (phosphoglucose isomerase and 

 transferrin A) was reported for both within and be- 

 tween samples of Atlantic and Pacific yellowfin tuna 2 - 3 . 

 However these loci have not been used to examine 

 population structure. 



Our understanding of the population structure of 

 yellowfin tuna in the Pacific Ocean remains problem- 

 atic. Much evidence is available which suggests that 

 population structure exists, yet genetic analyses did 

 not reveal differentiation among samples collected from 



-Anonymous. 1977. Inter-American Tropical Tuna Commission, An- 

 nual Report, 1976. 



'Anonymous. 1978. Inter-American Tropical Tuna Commission, An- 

 nual Report, 1977. 



distant locations. To further examine the genetic basis 

 of the population structure of Pacific yellowfin tuna 

 we employed restriction fragment length polymorphism 

 (RFLP) analysis of mitochondrial DNA (mtDNA). Be- 

 cause mtDNA evolves rapidly (Moritz et al., 1987; 

 Brown et al., 1979) and displays considerable polymor- 

 phism within animal populations (Avise and Lansman, 

 1983), mtDNA analyses have been useful in revealing 

 population structure within marine fishes (Ovenden, 

 1990). Using this technique, we demonstrated consid- 

 erable genetic variability within yellowfin tuna, but 

 we could not reject the null hypothesis that samples 

 share a common gene pool. 



Materials and methods 



Hearts were taken from 50 yellowfin tuna at each of 

 five Pacific locations and one Atlantic location (Fig. 1); 

 however, only 20 specimens per location were analyzed. 

 Samples from the Pacific were collected during 1990 

 at Manta, Ecuador (ECU); Revillagigedo Islands, 

 Mexico (MEX); Oahu, Hawaii (HAW); Manus Island, 

 Papua New Guinea (PNG); and New South Wales, Aus- 

 tralia (AUS). The sample from the Atlantic was col- 

 lected during 1991 at Hatteras, North Carolina (ATL). 

 Hearts were dissected within 12 hours of capture and 

 placed on crushed ice. Hearts from fish collected in the 

 Pacific were frozen at -20°C and shipped to the Inter- 

 American Tropical Tuna Commission, La Jolla, CA, 

 where they were stored at -20°C for more than one 

 year before shipment to our laboratory. Hearts from 

 fish collected in the Atlantic were transported on 

 wet ice and frozen at -70°C within four hours of 

 dissection. 



MtDNA was purified from 3 g of heart tissue from 

 Atlantic specimens following the CsCl-ethidium bro- 

 mide gradient centrifugation protocol of Lansman et 

 al. (1981). MtDNA yields averaged about 350 ng of 

 supercoiled mtDNA per g of heart tissue. Aliquots of 

 mtDNA were digested with the following 12 informa- 

 tive restriction endonucleases (Stratagene and BRL) 

 according to the manufacturers' instructions: Apal, 

 Aval, Banl, Bell, Bgll, Dral, EcoRl, Hindlll, Neil, PstI, 

 Pvull and Xhol. Restriction fragments were end-la- 

 beled with a mixture of all four a- 35 S-dNTP's by using 

 the Klenow fragment, electrophoresed at 2 volts/cm 

 overnight in YJc agarose gels, and visualized by auto- 

 radiography (Sambrook et al, 1989). 



Yields of supercoiled mtDNA from Pacific specimens 

 were low, possibly due to sub-optimal storage condi- 

 tions. For these specimens, mtDNA-enriched genomic 

 DNA was isolated from 4 to 6 g of heart tissue follow- 

 ing the protocols of Chapman and Powers ( 1984), modi- 

 fied by the omission of sucrose step gradients and the 



