Graves et al: Stock structure of Pomatomus ssltatnx along the mid-Atlantic coast 



707 



hybridization protocols of Maniatis et al. (1982) were 

 followed after digestion and electrophoresis. Highly 

 purified bluefish mtDNA, nick translated with biotin- 

 7-dATP, was used as a probe for mtDNA fragments. 

 Hybridization filters were visualized after strigency 

 washes using the BRL BlueGene Nonradioadtive 

 Nucleic Acid Detection System. 



All mtDNA samples were digested with the follow- 

 ing nine restriction endonucleases used according to 

 the manufacturers' instructions: Aval, Dral, EcoRV, 

 HindUl, Neil, Pstl, Pvull, Sstl, and Sstll. The dif- 

 ferent restriction-fragment patterns produced by each 

 restriction endonuclease were assigned a letter, and a 

 composite mtDNA genotype, consisting of nine letters 

 representing the fragment patterns generated by each 

 of the restriction endonucleases, was constructed for 

 each individual. The nucleon diversity (Nei 1987) was 

 calculated for each sample and for the pooled samples. 

 The nucleotide sequence divergence among mtDNA 

 genotypes was estimated by the site approach of Nei 

 and Li (1979). The mean nucleotide sequence diversity 

 within samples and mean nucleotide sequence diver- 

 gence between samples were calculated following the 

 method of Nei (1987), with the latter value being cor- 

 rected for within-group diversity (Nei 1987). The dis- 

 tribution of genotypes was evaluated for homogeneity 

 among collections using the G-test (Sokal and Rohlf 

 1981); however, as several of the genotypes were 

 represented by one individual, we employed the Roff 

 and Bentzen (1989) Monte Carlo approach to estimate 

 the significance of heterogeneity 

 X" values determined from the 

 raw data. 



Results 



The analysis of 472 mid- Atlantic 

 bluefish with 9 restriction endo- 

 nucleases revealed 40 mtDNA 

 genotypes, and 2 mtDNA geno- 

 types were encountered among 

 19 Australian bluefish. A total of 

 77 restriction fragments was vis- 

 ualized, and the average individ- 

 ual was scored for 34 fragments, 

 accounting for approximately 

 1.4% of the mtDNA genome. 

 The restriction endonucleases, 

 HindlU and Pstl, revealed no 

 variant fragment patterns, while 

 the remaining seven enzymes 

 revealed from two (Sstl and 

 Sstll) to eight (Neil) different 

 fragment patterns. Restriction- 

 site gains or losses were inferred 



from completely additive changes in fragment patterns. 



Considerable RFLP variation was detected within 

 Atlantic bluefish samples (Table 2). The most common 

 mtDNA genotype, AAAAAAAAA, ranged in frequen- 

 cy from 0.43 (NC 1990 YOY) to 0.75 (NJ 1990 YOY). 

 The large number of variant genotypes resulted in 

 nucleon diversities ranging from 0.416 to 0.798 (Table 

 3). Because many of the variant genotypes differed 

 from the common genotype by several site changes, the 

 within-sample mean nucleotide sequence diversities 

 were also relatively high, varying from 0.63% to 1.49%. 

 In contrast to the mid-Atlantic bluefish, the Australian 

 sample was quite depauperate of variation. Of the 19 

 fish in the sample, 18 shared a common mtDNA geno- 

 type (AAAEEAAAD), and one fish had a genotype dif- 

 fering from the common type by a single site change 

 (Table 2). The lack of variation in the Australian sample 

 was reflected in a low nucleon diversity (0.105) and a 

 within-sample mean nucleotide sequence diversity of 

 0.07%. 



Significant genetic differentiation was not found 

 between the samples of spring- and summer-spawned 

 YOY bluefish collected in New Jersey during the sum- 

 mer of 1990. The corrected mean nucleotide sequence 

 divergence between the two samples was extremely 

 small (0.02%), indicating that average sequence diver- 

 gence between two individuals randomly drawn from 

 either the spring- or summer-spawned sample was the 

 same as the divergence between two individuals ran- 

 domly drawn from each group. 



