Allozyme and RAPD variation in the eastern Pacific yellowfin tuna (Thunnus albacares).

Abstract--Stock structure of eastern Pacific yellowfin tuna was investigated by analyzing allozymes and random amplified polymorphic DNAs (RAPDs) from 10 samples of 20-30 individuals each, collected between 1994 and 1996 from fishing vessels operating in the Inter-American Tropical Tuna Commission (IATTC) yellowfin regulatory area (CYRA). Allozyme analysis resolved 28 loci, eight of which were polymorphic under the 0.95 criterion: Aat-[S.sup.*], Glud, Gpi-[F.sup.*], Opi-[S.sup.*], La, Lgg, Pap-[F.sup.*], and 6-Pgd, resulting in a mean heterozygosity over all allozyme loci of H = 0.052. Four polymorphic RAPD loci were selected for analysis, resulting in a mean heterozygosity of H = 0.43. Eight of 45 pairwise comparisons of allozyme allele frequencies among the ten samples showed significant differences after correction for multiple testing (P<0.0001), all of which involved comparisons with the Gulf of California sample. Confirmation of this signal of population structure would have management implications. No significant divergence in RAPD allele frequencies was observed among samples. Weir and Cockerham [theta] estimated for allozyme loci ([theta]=0.048; P<0.05) and RAPD loci ([theta]=0.030; P>0.05) revealed little population structure among samples. Mantel tests demonstrated that the genetic relationships among samples did not correspond to an isolation-by-distance model for either class of marker. Four of eight comparisons of coastal and offshore samples revealed differences of allele frequencies at the Gpi-[F.sup.*] locus (P<0.05), although none of these differences was significant after correction for multiple testing (P>0.001). Results are consistent with the hypothesis that the CYRA yellowfin tuna samples comprise a single genetic stock, although gene flow appears to be greater among coastal samples than between coastal and offshore samples.

**********

Yellowfin tuna (Thunnus albacares) is a cosmopolitan species inhabiting tropical and subtropical waters in the Atlantic, Pacific, and Indian oceans. This species has accounted for more than a third of the world's tuna production since 1970. The eastern Pacific has contributed from 21% to 26% of the global catch from 1993 through 1997, representing 273,329 metric tons (t) in 1990 to 264,426 t in 1998 (IATTC, 1999).

Yellowfin tuna is a large pelagic fish with a common size of 150 cm (Collette and Nauen, 1983). Spawning occurs throughout the year in the tropical oceans, preferably near islands and coasts (Leis et al., 1991). Growth is rapid and individuals reach maturity by the end of the second year (Suzuki et al., 1978). Schooling of individuals of similar size is observed near surface waters and is often associated with floating objects (Wild, 1994).

Yellowfin tuna is currently considered to comprise a single species (Gibbs and Collete, 1967), although significant morphometric and meristic differences, limited fish movements, and differences among catch data, have been reported for the different regions of the Pacific Ocean (Godsil and Greenhood, 1951; Schaefer, 1955; Joseph et al., 1964; Suzuki et al., 1978; Schaefer, 1991). Population structure in yellowfin tuna has been addressed in the Pacific Ocean by using several independent methods. Morphometric and meristic based studies have shown significant differences (Godsil and Greenhood, 1951; Schaefer, 1955; Kurogane and Hiyama, 1957), and at least three stocks or discrete units (western, central, and eastern Pacific) have been proposed. More recent studies using morphometric multivariate analysis suggest the presence of different stocks between north and south regions in the eastern Pacific (Schaefer, 1991), as well as across the Pacific Ocean (Schaefer, 1991). Additionally, differences in larval distribution, catch rates, and size composition data of yellowfin tuna caught along the equatorial Pacific by longline and purse-seine have been used by Suzuki et al. (1978) to distinguish between western, central, and eastern Pacific groups.

Tagging experiments have shown limited movement of yellowfin tuna between western and eastern Pacific waters (Joseph et al., 1964; Fink and Bayliff, 1970). In the eastern Pacific, the presence of two groups has been suggested: a northern group off Baja California coast and the Revillagigedo Islands and a southern group from the Maria Islands through Chile. Some mixing occurs between them (Fink and Bayliff, 1970). There seem to be marked movements between north and south groups along the coast with limited westward movements (Joseph et al., 1964).

Some studies of population structure using genetic analyses have not revealed the presence of discrete stocks along the Pacific Ocean. Barret and Tsuyuki (1967) used transferrin analysis and did not find differences in allele frequencies between samples from Hawaii and eastern Pacific samples, although heterogeneity was detected within the eastern Pacific samples (IATTC, 1975). Allozyme variation studies in the esterase locus (Fujino, 1970) did not show enough evidence of genetic differentiation between eastern Pacific and Hawaii samples. Furthermore, Scoles and Graves (1993) used restriction fragments length polymorphisms (RFLP) and analysis of mitochondrial (mt) DNA to examine five samples collected across the Pacific Ocean and one from the Atlantic Ocean. Although they detected 34 haplotypes and considerable genetic variation, no evidence of genetic differentiation among samples was found.

However, more recent genetic studies have provided limited evidence of genetic heterogeneity. Ward et al. (1994) analyzed four polymorphic allozyme loci and 18 mtDNA haplotypes in yellowfin tuna from the Pacific Ocean. Although no unique haplotypes were found in the analyzed populations through RFLP analysis, the eastern Pacific samples were found to be different from the central and western Pacific samples in frequency differences at a single locus GPI-[F.sup.*], suggesting that the signal of population structure exhibited is due to selective factors contributing to the divergence. Eastern Pacific samples (n=41) were collected in the northeast Pacific off California and at an unspecified site off Mexico (n=40). Comparisons of GPI-[F.sup.*] allele frequencies from eastern Pacific also included two samples previously analyzed by Sharp (1978) from Roca Partida (Central America) and Ecuador. Their results showed population homogeneity at the GPI-[F.sup.*] locus for this region.

To date, the methods and logistics used to study divergence in the Pacific yellowfin tuna have been focused on a global rather than a local scale, and sampling has been focused on the wide areas of the west and central Pacific. Local structure in the eastern Pacific yellowfin tuna has not been addressed through a more intense sampling strategy to examine genetic homogeneity in this region. Because tagging studies have shown restricted longitudinal movements by yellowfin tuna, population structure and isolation by distance hypotheses can be tested. To evaluate the stock structure of yellowfin tuna in eastern Pacific, we employed analyses of allozymes and of randomly amplified polymorphic DNA (RAPDs).

RAPDs have proven to be useful genetic markers because of their high levels of polymorphisms (Williams et al., 1990; Welsh et. al., 1991). They have been used to estimate population structure in fishes, including the cod (Kenji, 1998), red mullet (Mamuris et al., 1998), and striped bass (Bielawski and Pumo, 1997). The use of RAPDs, considered as neutral markers, and the simultaneous use of allozyme analyses with intense sampling in a more local area, might provide evidence about the relationship between gene flow and spatial distribution of the eastern Pacific yellowfin tuna, as well as evidence of the presence of local selective factors responsible for the divergence suggested by Ward et al. (1994).

Materials and methods

Sampling

A total of 327 tissue samples from specimens of ten locations were obtained from commercial tuna boats fishing in the tropical eastern Pacific from 1994 to 1996 (Fig. 1). Muscle tissue samples were dissected from specimens at the time of landing and were transported in liquid nitrogen or on dry ice to the Laboratorio de Genetica de Organismos Acuaticos of the Instituto de Ciencias del Mar y Limnologia in Mexico City. Samples were maintained at -70[degrees]C until processing.

[FIGURE 1 OMITTED]

Allozyme analysis

For allozyme analysis, 1 [cm.sup.3] (about one gram) of tissue sample was ground with a manual homogenizer in 1.5 mL of extraction buffer (0.01M Tris-0.001M EDTA, pH 6.8, and 1% NADP) and centrifuged at 2500 g at 4[degrees]C. Electrophoretic runs were performed in 12% (w/v) starch gels (Sigma Chemicals, St. Louis, MO). Four buffer systems were used to analyze nineteen enzymes that resolved 28 loci, eight of which showed polymorphism: Aat-[S.sup.*] (aspartate aminotransferase), Glud (glutamate dehydrogenase), Gpi-[F.sup.*] and Gpi-[S.sup.*] (glucose phosphate isomerase), La (leucil-L-alanine), Lgg (L-leucil-glycil-glycine), Pap-[F.sup.*] (L-leucil-L-proline) and 6-Pgd (phosphogluconate dehydrogenase). Enzymes AK (adenilate kinase), CK (creatinine kinase), GAPDH (glyceraldehyde-3-phosphate dehydrogenase), LDH (lactate dehydrogenase), MDH (mMalate dehydrogenase), ME (malic enzyme) and SOD (superoxide dismutase), displayed twenty more loci that were presumably monomorphic. Buffer systems for enzyme analysis were 1) amino-citrate: 0.04M citric acid, 15mL/L of N-3-aminopropyl-morpholine, pH 6.5 (AAT, GPI, and LA); 2) 0.008M Tris, 0.003 M citric acid, pH 6.7 (GLUD and 6-PGD); 3) 0.025 M Tris, 0.192 glycine, pH 8.5 (GPI and LGG); 4) 0.0.076 M Tris, 0.005 M citric acid, pH 8.7 (PAP). Enzyme assays were performed following Harris and Hopkinson (1976). Enzymes showing polymorphism were analyzed for all samples and subjected to population genetic analysis.

RAPD analysis

For RAPD analyses, genomic DNA was extracted from muscle tissue by using standard phenol-chloroform protocols (Sambrook et al., 1989), resuspended in TE buffer (10mM Tris-0.1mM EDTA pH 8.0), and quantified with a Hoefer DyNA quant 200 fluorometer. DNA was amplified with primer F-10 (Operon[R] Alameda, CA; 5'-GGAAGCTTGG-3'). Amplifying reactions were performed in a final volume of 22 [micro]L consisting of 0.7 to 1 ng/[micro]L of DNA in amplification buffer, 10 mM Tris-HCl, 50 mM KCl, 1.5 mM Mg[Cl.sub.2], 33 ng of primer, 10 mM dNTPs, and 1 U of Taq polymerase. Amplification of genomic DNA was performed in a Perkin Elmer[R], Foster City, CA (mod. 480), thermal cycler. The program was set for 1 cycle of 1 min. at 36[degrees]C, followed by 44 cycles of 1 min. at 36[degrees]C; 1 min. at 94[degrees]C; 2 min. at 72[degrees]C, and a final cycle of 15 min. at 72[degrees]C. Optimal DNA concentrations for amplification were determined by testing several dilutions, one of which was taken as the standard for every subsequent amplification.

Amplified fragments were resolved by electrophoresis in 1.5% agarose gels (Sigma Chemicals) for 3 to 4 h. at 90 mA (100 V). A 100bp DNA Ladder (GibcoBRL, Gaithersburg, MD, 15628-019) was used as size standard. After electrophoresis, gels were stained with ethidium bromide and photographed in a UV light transilluminator.

Data analysis

Allelic frequencies, test of conformity of genotype distributions with Hardy-Weinberg, and heterozygous deficit were determined by using Genepop version 3.3 (Raymond and Rousset (1)). Homogeneity of allozyme and RAPD allele frequencies was evaluated by using the exact probability test (Raymond and Rousset, 1995) consisting of a contingency analysis for every polymorphic locus and an estimation of their probability values by the combined probability of Fisher (Sokal and Rohlf, 1995) as implemented in the TFPGA program (Miller (2)). Pairwise comparisons were conducted to determine allele frequency differences among samples in order to define sources of variation. Based on the longitudinal differentiation pattern observed by Ward et al. (1994) and the morphological latitudinal differences within eastern Pacific samples reported by Schaefer (1991) at north-south of the 15-20[degrees]N range, spatial homogeneity was tested at the following levels: overall samples (O), among longitudinal regions (L; coastal, intermediate, and offshore localities), and latitudinal regions (N; north-south of the 15-20[degrees] range). The Gulf of California sample was excluded from this analysis because the large variation found in this sample (reflected in the significant differences shown in allele frequency homogeneity from pairwise comparisons) would not allow an accurate assessment of whether longitudinal or latitudinal differentiation exists or not. Significance levels were adjusted for multiple testing through the Bonferroni sequential method (Rice, 1989).

Population subdivision was estimated by using the Weir and Cockerham (1984) method through the TFPGA program. Standard error and confidence intervals were obtained through jackknife and bootstrapping procedures, respectively, with [F.sub.ST] Pro 1.0 (Weir, 1990). Estimates of population subdivision were partitioned into the following levels: O, over all the samples; L, longitudinal regions (coastal, intermediate and offshore); and N, latitudinal regions.

We used the [theta] statistic to estimate gene values between sample pairs (Slatkin, 1993) that are defined as [M.sub.[theta]]. An "isolation by distance" model was evaluated from the correlation between the distance between localities measured as geographic separation in nautical miles (nmi), and the [M.sub.[theta]] values by means of the Mantel test (Hellberg, 1994) in both allozymes and RAPDs.

The patterns of the amplification products resulting from the RAPD analysis were subjected to the same analyses as allozymes with the procedures described in Lynch and Milligan (1994). RAPD fragments were interpreted under the following assumptions: 1) fragments were considered to behave as dominant genes (Williams et al., 1990); 2) every polymorphic fragment was considered derived from a two-allele locus; 3) the equilibrium of Hardy-Weinberg was assumed for all genotypes, and 4) each fragment was considered to be an independent locus.

Only those fragments clearly defined and having consistent intensity were recorded. Because of this, the Michoacan sample, with poor consistency in the banding patterns, was excluded from the RAPD analysis. The allele frequency of every fragment was calculated on the basis of the inferred homozygous recessive genotypes. Because of the dominant nature of the alleles, and in order to correct the bias originated by calculating the recessive allelic frequencies, we chose the estimation based on the Taylor expansion (Kendall and Stuart, 1977, cited in Le Corre et al., 1997) as implemented in TFPGA program. This reduction on the bias is based on the equation resulting from the second order expansion of Taylor (see details in Lynch and Milligan, 1994).

Results

Allozymes

Of the 28 analyzed allozyme loci, eight (28.5%) showed polymorphism under the 0.95 criterion. The observed heterozygosities per sample over all allozyme loci ranged from 0.027 to 0.083 (mean 0.052). Allozyme frequencies for the eight polymorphic loci detected are shown in Table 1. After adjusting levels of significance by the Bonferroni procedure, significant deviation of genotypic frequencies from those expected under Hardy-Weinberg was found in the loci Lgg and Pap-[F.sup.*] for the Gulf of California sample and in the Aat-[S.sup.*] locus for two localities--west of Revillagigedo Islands, and the Gulf of California (P<0.0006, Table 1). Deviations displayed for both locations corresponded to a heterozygous deficit (P<0.0006, after Bonferroni correction).

Comparison of allozyme allele frequencies among all collections (overall) by the exact probability test revealed significant heterogeneity at loci Glud, La, and Lgg, after Bonferroni adjustment (P<0.006). Pairwise comparisons among samples to test allele homogeneity gave significant differences for nine of 45 comparisons after correction for multiple tests (P<0.001), seven of which involved comparisons with the Gulf of California (GC) sample. The remaining significant differences were between Guerrero-Nayarit, and Cape San Lucas-Nayarit comparisons (Table 2), resulting from significant heterogeneity at Glud, La, and Lgg loci.

In general, allozyme analysis displayed low levels of differentiation. The [theta] value over all loci was different from zero (P<0.05) and showed that 4.8% of the variance was attributable to differences among samples (Table 3).

Individual loci showed [theta] values ranging between 0.0037 and 0.27. Highly significant values at loci La (0=0.13 [+ or -] 0.089) and Lgg ([theta]=0.27 [+ or -] 0.253) evidently resulted from their weak polymorphism in some samples (Table 3).

Allele frequency homogeneity was tested among coastal, intermediate, and offshore regions (Table 4). Significant heterogeneity was found by exact test between coastal and offshore comparisons (P=0.0043) but was found to be not significant between coastal and intermediate regions (P=0.0632). Subdivision as measured by [theta] among coastal, intermediate, and offshore localities was not different from zero. However, for the Gpi-[F.sup.*] locus the population subdivision among regions (0.0058) was twice as large as that noted among samples ([theta]=0.003), but neither value was significant (Table 3). No latitudinal differentiation by the exact test or population subdivision estimated by the [theta] index was found between north and south regions (data not shown).

The gene flow values ([M.sub.[theta]]) were high (mean 24.8 migrants per generation). A lack of correlation between gene flow estimations and geographic distance by means of the Mantel test was observed ([r.sup.2]=0.144; P=0.22), resulting in a rejection of the isolation by distance model.

For the Gpi-[F.sup.*] locus, paired tests of significance (data not shown) showed discrepancies in the Gpi-[F.sup.*]/75 allele frequencies among localities from the coast with those located at the CYRA limits (Fig. 1, Table 5). Four of eight comparisons of coastal and offshore samples revealed differences of allele frequencies at this locus (P<0.05), although none of these differences was significant after correction for multiple testing (P>0.001).

RAPDs

The primer OPF-10 produced 11 amplified fragments, with sizes from 200 to 600 bp (base pairs). Four of the fragments were polymorphic for all samples (Table 1). No significant heterogeneity of RAPD allele frequencies was found for any locus between any paired sample comparison, among all collections, nor among latitudinal or longitudinal regions (P=0.4806).

The mean [theta] value for all fragments and samples (overall), as well as regional estimations derived from RAPDs (0.0302), were not significantly different from zero and displayed some negative values. Estimations of gene flow between sample pairs ([M.sub.[theta]]) from RAPD data aver aged 29.2 migrants per generation. The evaluation of the relationships between geographic distances and the gene flow estimations in pairwise collections ([M.sub.[theta]]), through the Mantel test, showed a nonsignificant correlation ([r.sup.2]=0.413, P=0.984).

Discussion

The test of conformance to the Hardy-Weinberg frequencies showed significant differences in Lgg and Pap-[F.sup.*] loci only in the Gulf of California sample, where polymorphism at those loci was also consistently found. Similar results were obtained, with smaller differences in locus Aat-[S.sup.*] from the west of the Revillagigedo Islands and the Gulf of California samples. Considering the fact that our samples were provided by the commercial fleet, they could have included representatives of different schools with differences in genotypic distributions originated by differences in age classes or sexual ratios (or for both) among schools because recruitment of individuals into new schools has been reported to be mainly by aggregating individuals of similar sizes (Collette and Nauen, 1983). Because recruitment to the original tuna schools has been reported as well (Kimley and Holloway, 1999), random processes could also induce differences in genotypic frequencies that favor aggregation of some genotypes, while segregating some others, causing a kind of Wahlund effect that is reflected by a heterozygous deficit as shown by the homozygous excess for loci and locations having HW deviations, especially as shown in the Gulf of California sample.

The estimations of population structure based on allozymes showed a small but significant value different from zero ([theta]=0.048; P<0.01). The Gulf of California sample contributed to the significant subdivision value as shown when that collection was excluded from the regional subdivision analysis, as well as to significant heterogeneity of its allele frequencies when paired comparisons were made.

The small value of [theta] for overall estimations on RAPD data is probably due to the small sample size. The negative values of [theta] from overall and regional estimations resulted from subtracting the large value of the correction derived from the variation expected of the sample sizes from the small value of variation due to fluctuations in allele frequencies. The fact that RAPD data are considered dominant could reduce information about the true allele distributions by subestimation of null allele frequencies notwithstanding the correction applied to recessive genotypes, which is dependent on the sample sizes (Lynch and Milligan, 1994). Other assumptions for RAPD data limit the value of this marker, especially when estimations are derived from a small number of loci and sample sizes. Additional constraints are related to the limited number of alleles (two) to estimate dominant markers, which tend to subestimate the polymorphism and thus reduce the significance of relatively small discrepancies in allele distributions.

No differentiation between coastal and offshore samples was found in our study because of the slight, nonsignificant differences in the estimation of the subdivision by regions. Although the overall estimation was not different from zero, the allele homogeneity analysis showed allele-frequency heterogeneity between coastal and offshore samples, and nonheterogeneity between coastal and intermediate samples.

These results are consistent with the migration reports through tagging studies; evidence exists for the presence of two main yellowfin tuna groups in the eastern Pacific that mix to some extent (Fink and Bayliff, 1970) and that migrate longshore from around the 20[degrees]N to the mouth of the Gulf of California and to the zone between the Revillagigedo and the Clipperton islands, and back again (Joseph et al., 1964; Fink and Bayliff, 1970), although longitudinal movements are restricted to the limits of yellowfin regulatory area (CYRA). Similarly, important northward movements along the coasts to the mouth of the Gulf of California, and subsequently to the western coasts of Baja California, have been reported. Although the estimation of [theta] for allozymes showed a significant value, it was notably influenced by the heterogeneity found between the Gulf of California sample and all other samples.

Discarding the variation displayed by loci La, Lgg, and Pap-[F.sup.*], originating mainly from Gulf of California sample, the estimation of subdivision was still marginally significant after Bonferroni correction, which should be considered as evidence that the Gulf of California sample may represent a partially isolated population with different allele frequencies. Oceanographic conditions inside the Gulf are somewhat different from those of the Pacific Ocean where there are warmer waters at the end of the year, especially during yellowfin tuna spawning seasons. There is also high productivity characterized by the presence of significant biomass abundance of sardine or anchovy schools (Cisneros-Mata et al., 1995), which represents opportunities to establish the feeding and consequently the spawning grounds for eastern Pacific yellowfin tuna. Likewise, there is a trend of migratory movements through the Gulf of California by different groups of yellowfin tunas (Fink and Bayliff, 1970). These movements promote stock mixing and help to explain the wide polymorphism displayed in this sample, in contrast to the weak variation found in other samples from the coast and offshore regions. Further genetic research, including sequential temporal sampling of young fishes in order to ensure the presence of individuals that originated in discrete spawning grounds, should be undertaken to prove the presence of an independent unit inside the Gulf of California, which, if confirmed, might necessitate new stock management strategies.

Allele frequencies for Gpi-[F.sup.*] locus found in the present study, apparently, correspond to those reported by Ward et al. (1994 and 1997). These authors reported a higher proportion of the allele Gpi-[F.sup.*]75 (0.571) in the eastern Pacific region and a gradual decrement of the same allele toward the central (0.423) and western Pacific regions (0.330), where allele Gpi-[F.sup.*]100 (0.650) had the higher proportion. In the present study, the highest frequencies for the allele Gpi-[F.sup.*]75 corresponded to the region of the eastern Pacific, situated in the limits of the yellowfin tuna regulatory area (offshore region), and there was a slight decrease in frequencies towards the coastal area (Table 5). Furthermore, allele frequencies for the Gpi-[F.sup.*]75 allele from the coastal locality, Colima (0.444), and the intermediate locality southeast, Clipperton Islands (0.409), have coincidences with those reported by Ward et al. (1994) for the collection Hawaii 92 (0.423) in the central Pacific region.

The similarities in the Gpi-[F.sup.*] allelic frequencies between eastern (Colima and Clipperton) and central Pacific samples (Hawaii 92) might possibly be attributed to the extended migrations of yellowfin tuna in the eastern Pacific brought about by the strong influence of warm waters on tuna movements because of the increased depth of the thermocline layer in that area, which was reflected by a decrease in catches (Joseph and Miller, 1988; Wild, 1994) and which possibly led to the mixing of the eastern and central Pacific stocks.

The low number of RAPD loci analyzed and the uncertainty of fulfilling some assumptions, such as the genetic identity of each band needed for qualitative and quantitative interpretation of data in terms of allelic frequencies, do not allow us to consider our estimations of subdivision reliable with the RAPD method. Additionally, the lack of reliability of estimations associated with high sampling variances by using randomly collected fishery samples highlights the need to design more efficient spatial and temporal sampling strategies in local and wide areas, as well as the need for alternative hypervariable markers to assess the divergence patterns observed in highly migratory species.

Table 1

Allele frequencies for allozymes and RAPDs, samples size (n) and
agreement to the Hardy-Weinberg equilibrium (HW) for every loci and
sample of Thunnus albacares. Significance values for HW tests were
adjusted for multiple comparisons with an initial [alpha] level of
0.0006 [(0.05/(8 loci x 10 samples)]. P = probability of significance
for allele frequency heterogeneity per locus, * = significant at
[alpha] = 0.006. - = no data. OCY = out of CYRA area; WRE = west of the
Revillagigedo Islands; NCL = North Clipperton islands; SCL = South
Clipperton Islands; MCH = Michoacan; GUE = Guerrero; COL = Colima;
CSL = Cape San Lucas; NAY = Nayarit; GC = Gulf of California.

                               Collection

Locus     Allele      OCY      WRE      NCL      SCL      GUE

Aat-S*       -90    0.171    0.266    0.250    0.231    0.154
            -100    0.829    0.734    0.750    0.769    0.846
               n       41       32        6       13       39
              HW      yes       no      yes      yes      yes
Glud         100    0.700    0.634    0.667    0.857    0.663
              85    0.300    0.366    0.333    0.143    0.337
               n       40       41        6       14       49
              HW      yes      yes      yes      yes      yes
Gpi-F*       135    0.071    0.134    0.250    0.091    0.190
             100    0.262    0.280    0.250    0.500    0.230
              75    0.667    0.586    0.500    0.409    0.580
               n       42       41        6       11       50
              HW      yes      yes      yes      yes      yes
Gpi-S*       -60    0.402    0.536    0.417    0.571    0.360
            -100    0.598    0.464    0.583    0.429    0.640
               n       41       42        6       14       50
              HW      yes      yes      yes      yes      yes
La           120      0.0    0.024      0.0      0.0      0.0
             100      1.0    0.976      1.0      1.0      1.0
               n       42       42        6       14       50
              HW      yes       --       --       --       --
Lgg          115     0.00    0.024      0.0      0.0      0.0
             100     1.00    0.976      1.0      1.0      1.0
               n       42       42        6       14       50
              HW      yes       --       --       --       --
Pap-F*       110      0.0      0.0      0.0      0.0      0.02
             100      1.0      1.0      1.0      1.0      0.98
               n       42       42        6       14       50
              HW       --       --       --       --      yes
6-Pgd        100    0.902    0.903    0.667    0.821    0.939
              90    0.098    0.097    0.333    0.179    0.061
               n       41       31        6       14       49
              HW      yes      yes      yes      yes      yes
F10-1          a    0.406    0.282    0.234    0.322       --
               b    0.594    0.718    0.766    0.678       --
               n        9        8        7        9       --
              HW       --       --       --       --       --
F10-2          a    0.406    0.603    0.441    0.322       --
               b    0.594    0.397    0.559    0.678       --
               n        9        8        7        9       --
              HW       --       --       --       --       --
F10-3          a    0.406    0.475    0.329    0.625       --
               b    0.594    0.525    0.671    0.375       --
               n        9        8        7        9       --
              HW       --       --       --       --       --
F10-4          a    0.115    0.282    0.234    0.178       --
               b    0.885    0.718    0.766    0.822       --
               n        9        8        7        9       --
              HW       --       --       --       --       --

                               Collection

Locus       MCH      COL      CSL      NAY       GC        P

Aat-S*    0.275    0.077    0.092      0.0    0.234
          0.725    0.923    0.908      1.0    0.766
             20       26       38       25       47    0.011
            yes      yes      yes       --       no
Glud      0.775    0.914    0.788    0.680    0.395
          0.225    0.086    0.212    0.320    0.605        *
             20       29       40       25       43
            yes      yes      yes      yes      yes
Gpi-F*    0.079    0.185    0.125    0.160    0.200
          0.316    0.370    0.363    0.260    0.314
          0.605    0.444    0.512    0.580    0.486    0.526
             19       27       40       25       35
             no      yes      yes      yes      yes
Gpi-S*    0.425    0.414    0.449    0.479    0.338
          0.575    0.586    0.551    0.521    0.663    0.015
             20       29       39       24       40
            yes      yes      yes      yes      yes
La          0.0      0.0      0.0      0.1      0.2
            1.0      1.0      1.0      0.9      0.8        *
             20       24       40       20       50
             --       --       --      yes      yes
Lgg         0.0      0.0      0.0      0.0    0.349
            1.0      1.0      1.0      1.0    0.651        *
             20       24       24       24       43
             --       --       --       --       no
Pap-F*     0.05      0.0      0.0      0.0    0.083
           0.95      1.0      1.0      1.0    0.917    0.055
             20       29       40       25       48
            yes       --       --       --       no
6-Pgd       1.0    0.966      1.0    0.840    0.938
            0.0    0.034      0.0    0.160    0.062    0.007
             20       29       40       25       48
             --      yes       --      yes      yes
F10-1     0.083    0.505    0.355    0.282    0.246
          0.917    0.495    0.645    0.718    0.754    0.572
              6        9       10        8        9
             --       --       --       --       --
F10-2     0.278    0.322    0.355      1.0    0.322
          0.722    0.678    0.645      0.0    0.678    0.369
              6        9       10        8        9
             --       --       --       --       --
F10-3     0.397    0.406    0.529    0.371    0.406
          0.603    0.594    0.471    0.629    0.594    0.912
              6        9       10        8        9
             --       --       --       --       --
F10-4     0.083    0.246    0.219    0.138    0.246
          0.917    0.754    0.781    0.862    0.754    0.929
              6        9       10        8        9
             --       --       --       --       --

Table 2

Pairwise-sample comparisons of allele frequency homogeneity for Thunnus
albacares. Probability values in allozymes are above the diagonal (-)
and RAPDs are below the diagonal (-). * = significant values after
Bonferroni correction for multiple tests (initial [alpha] was 0.05). na
= data not available. OCY = out of CYRA area; WRE = west of the
Revillagigedo Islands; NCL = North Clipperton islands; SCL = South
Clipperton Islands; MCH = Michoacan; GUE = Guerrero; COL = Colima;
CSL = Cape San Lucas; NAY = Nayarit; GC = Gulf of California.

Sample      OCY        WRE        NCL        SCL        GUE

OCY         --       0.5807     0.7391     0.3414     0.4386
WRE       0.9316       --       0.9482     0.6047     0.4330
NCL       0.9797     0.9770       --       0.9232     0.3494
SCL       0.9798     0.8088     0.8915       --       0.3989
GUE       0.5935     0.4732     0.9095     0.7757       --
MCH         na         na         na         na         na
COL       0.9385     0.7864     0.9155     0.9794     0.4569
CSL       0.9475     0.9703     0.9824     1.000      0.6576
NAY       0.9656     0.9984     0.8418     0.4965     0.4101
GC        0.9458     0.9637     0.9987     0.9977     0.9743

Sample      MCH        COL        CSL        NAY        GC

OCY       0.8254     0.0295     0.1525      0.0194    *<0.001
WRE       0.1180     0.0012     0.0084      0.0192    *<0.001
NCL       0.7307     0.1567     0.0859      0.4779      0.0174
SCL       0.0717     0.3414     0.1093      0.0287    *<0.001
GUE       0.4137     0.1723     0.5932    *<0.001     *<0.001
MCH         --       0.0500     0.1532      0.0096    *<0.001
COL         na         --       0.8237      0.0082    *<0.001
CSL         na       0.9849       --        0.0012    *<0.001
NAY         na       0.4206     0.5707        --      *<0.001
GC          na       0.9715     0.9874      0.6518      --

Table 3

Estimates of population subdivision [theta] (Weir and Cockerham,
1984) for allozymes of Thunnus albacares partitioned
into longitudinal regions [[theta].sub.L] (regions) (i.e.
coastal-intermediate-offshore) and samples [[theta].sub.0] (overall).
n.v. = negative values. P = probability of significance of subdivision
estimations. Significance of single-locus values was corrected
with an initial level of 0.006 (0.05/8 loci). Means and standard
error were obtained by the jackknife method. Confidence
intervals obtained by 1000 resamplings through bootstrapping are also
shown. * = significant values after Bonferroni correction.

Locus      [[theta].sub.L (regions)]       P

Aat-2 *             0.0152               0.072
Glud                0.0009               0.227
Gpi-1 *             0.0058               0.086
Gpi-2 *             0.0014               0.780
La                   n.v.                0.012
Lgg                 0.0073               0.004
Pap-1 *              n.v.                0.068
6-Pgd                0.044              >0.006 *
Mean        0.0067 [+ or -] 0.0042
CI 95%           0.0003-0.0199

Locus      [[theta].sub.0(overall)]        P

Aat-2 *              0.029               0.007
Glud                 0.086              >0.006 *
Gpi-1 *              0.003               0.210
Gpi-2 *              0.0037              0.007
La                   0.13               >0.006 *
Lgg                  0.27                0.006
Pap-1 *              0.024               0.001
6-Pgd                0.02               >0.006 *
Mean         0.048 [+ or -] 0.022
CI 95%            0.019-0.101

Table 4

Pairwise-regions comparison of allele frequencies for
Thunnus albacares. Probabilities of nonheterogeneity for
allozymes (based on exact tests) are above the diagonal
(-) and RAPDs are below the diagonal (-). * = significant
after corrected for multiple tests (Rice, 1989).

Region          Coastal    Intermediate    Offshore

Coastal           --          0.0632       0.0043 *
Intermediate    0.9998          --         0.5384
Offshore        0.9039        0.9126          --

Table 5

Comparison of Gpi-F* allele frequencies for Thunnus albacares among
data from the present study and those reported in Ward et al. (1994).
- = data absent.

                                                    Eastern (present
                                                         data)
Locus      Western/
Allele    Central (1)    Eastern (1)    Offshore      Intermediate

Gpi-F*
  135         0.026          0.100        0.103           0.147
  100         0.640          0.269        0.301           0.412
   75         0.332          0.631        0.596           0.441
   40         0.002           --            --              --
  n         346            178           83              17

          Eastern (present data)
Locus
Allele      Coastal    Pooled

Gpi-F*
  135         0.163      0.145
  100         0.301      0.307
   75         0.566      0.548
   40         --         --
  n         196        296

(1) Allele frequencies for Gpi-F* reported in Ward et al. (1994).

Acknowledgments

We are grateful to Ernesto Escobar from Pescados Industrializados S.A. PINSA for allowing the sampling, and Robert Olson from IATTC for providing the Gulf of California samples. We thank Monica Dominguez-Lopez, Yolanda Hornelas-Orozco, Evangelina Castillo, and Alma Hernandez-Perez, for collection and processing of samples, Luis Eguiarte and Valeria Souza for the facilities provided in their laboratory and three anonymous reviewers for their valuable comments. This manuscript benefited from the critical reading of John Graves, Jan McDowell, and Barbara Rutan. Funding for this project was provided by the Programa de Apoyo a Estudiantes de Posgrado (PADEP) and by the project IN20598 of the Programa de Apoyo a Proyectos de Investigacion e Innovacion Tecnologica. Direccion General de Asuntos del Personal Academico, Universidad Nacional Autonoma de Mexico.

(1) Raymond, M. L., and F. Rousset. 1995a. GENEPOP (version 1.2): population genetics software for exact tests and ecumenicist. J. Heredity 86:248-249.

(2) Miller, M.P. 1997. Tools for genetic populations analyses (TFPGA) 1.3: a windows program for the analysis of allozyme and molecular population genetic data, 29 p. Computer software distributed by the author at http://bioweb.usu.edu/ mpmbio.

Literature cited

Barret, I., and H. Tsuyuki.

1967. Serum transferrin polymorphism in some scombroid fishes. Copeia 1967:551-557.

Bielawski, J. P. and D. E. Pumo.

1997. Randomly amplified polymorphic DNA (RAPD) analysis of Atlantic coast striped bass. Heredity 78:32-40.

Cisneros-Mata, M. A., Nevarez-Martinez, M. O., and Hammann, M. G.

1995. The rise and fall of the Pacific sardine, Sardinops sagax caeruleus Girard, in the Gulf of California, Mexico. CALCOFI Rep. 36:136-143.

Collette, B. B., and C. E. Nauen.

1983. Scombrids of the world: an annotated and illustrated catalogue of tunas, mackerels, bonitos and related species known to date. In FAO species catalogue, vol. 2, 137 p. FAO Fish. Synop. 1325. FAO, Rome.

Fink, B. D., and W. H. Bayliff.

1970. Migrations of yellowfin and skipjack tuna in the eastern Pacific Ocean as determined by tagging experiments, 1952-1964. Inter-Am. Trop. Tuna Comm. Bull. 15:1-227.

Fujino, K.

1970. Immunological and biochemical genetics of tunas. Trans. Am. Fish. Soc. 99:152-178.

Gibbs, R. H., and B. B. Collette.

1967. Comparative anatomy and systematics of the tunas, genus Thunnus. Fish. Bull. 86:835-838.

Godsil, H. C., and E. C. Greenhood.

1951. A comparison of the populations of yellowfin tuna (Neothunnus macropterus) from the eastern and central Pacific, 33 p. Calif. Dep. Fish Game Fish. Bull. 82.

Harris, H., and D. A. Hopkinson.

1976. Handbook of enzyme electrophoresis in human genetics, 273 p. Am. Elsevier Publishing Co., New York, NY.

Hellberg, M. E.

1994. Relationships between inferred levels of gene flow and geographic distance in a philopatric coral, Ballanophyllia elegans. Evolution 48:1829-1854.

IATTC (Inter-American Tropical Tuna Commission).

1975. Annual report of the Inter-American Tropical Tuna Commission, 169 p. IATTC, LaJolla, CA.

1999. Annual report of the Inter-American Tropical Tuna Commission, 357 p. IATTC, LaJolla, CA.

Joseph, J., F. G. Alverson, B. D. Fink, and E. B. Davidoff.

1964. A review of the population structure of yellowfin tuna, Thunnus albacares, in the eastern Pacific Ocean. InterAm. Trop. Tuna Comm. Bull. 9(2):53-112.

Joseph, J., and F. R. Miller.

1988. El Nino and the surface fishery for tunas in the eastern Pacific. Proceedings of the Tuna Fisheries Research Conference, Japan Fish. Far Seas Fish. Res. Lab. Maguro Gyogyo Kyogikai Gijiroku, Suisancho-Enyo Suisan Kenkyusho: 199-207

Kenji, S.

1998. Genetic variation and local differentiation in the Pacific cod Gadus macrocephalus around Japan revealed by mtDNA and RAPD markers. Fish. Sci. 64(5):673-679.

Kimley, A. P., and C. F. Holloway.

1999. School fidelity and homing synchronicity of yellowfin tuna, Thunnus albacares. Mar. Biol. 133:307-317.

Kurogane, K., and Y. Hiyama.

1957. Morphometric comparison of the yellowfin tuna taken from the Equatorial Pacific. Jpn. Soc. Sci. Fish. Bull. 23: 388-293.

Leis, J. M., 21 Trnski, M. Harmelin-Vivien, J.-P. Renon, V. Dufour, M. K. El Moundi, and R. Galzin.

1991. High concentrations of tuna larvae (Pisces: Scombridae) in near-reef waters of french polynesia (Society and Taumotu islands). Bull. Mar. Sci. 48(1):150-158.

Le Corre, S., S. Dumolin-Lapegue, and A. Kremer.

1997. Genetic variation at allozyme and RAPD loci in sessile oak Quercus petraea (Matt.) Liebl.: the role of history and geography. Mol. Ecol. 6:519-529.

Lynch, M., and B. G. Milligan.

1994. Analysis of population genetic structure with RAPD markers. Mol. Ecol. 3:91-99.

Mamuris, Z., A. P. Apostodilis, A. J. Theodoru, and C. Triantaphyllidis.

1998, Application of random amplified polymorphic DNA (RAPD) markers to evaluate intraspecific genetic variation in red mullet (Mullus barbatus). Mar. Biol. 132: 171-178.

Raymond, M. L., and F. Rousset.

1995. An exact test for population differentiation. Evolution 49:1280-1283.

Rice, W. R.

1989. Analyzing tables of statistical tests. Evolution 43: 223-225.

Sambrook, J., E. F. Fritsch, and T. Maniatis.

1989. Molecular cloning: a laboratory manual, 2nd ed., 999 p. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

Scoles, D. R., and J. E. Graves.

1993. Genetic analysis of the population structure of yellowfin tuna, Thunnus albacares, from the Pacific Ocean. Fish. Bull. 91:690-698.

Schaefer, K. M.

1991. Geographical variation in morphometric characters and gill-raker counts of yellowfin tuna Thunnus albacares, from the Pacific Ocean. Fish. Bull. 9:690-698.

Schaefer, M. B.

1955. Morphometric comparison of yellowfin tuna from southeast Polynesia, Central America, and Hawaii. InterAm. Trop. Tuna Comm. Bull. 1:89-136.

Sharp, G. D.

1978. Behavioral and physiological properties of tunas and their effects on vulnerability to fishing gear. In The physiological ecology of tunas (G. D. Sharp and A. E. Dizon, eds.), p. 397-449. Academic Press, New York, NY.

Slatkin, M.

1993. Isolation by distance in equilibrium and non-equilibrium populations. Evolution 47:264-279.

Sokal, R., and F.J. Rohlf.

1995. Biometry, 3rd ed., 776 p. W.H. Freeman and Co., New York, NY.

Suzuki, Z., P. K. Tomlinson, and M. Honwa.

1978. Population structure of Pacific yellowfin tuna. InterAm. Trop. Tuna Comm. Bull. 17:273-441.

Ward, R. D., N. G. Elliott, P. M. Grewe, and A. Smolenski.

1994. Allozyme and mitochondrial DNA variation in yellowfin tuna (Thunnus albaeares) from the Pacific Ocean. Mar. Biol. 118:531-539.

Ward, R. D., N. G. Elliott, B. H. Innes, A. J. Smolenski, and P. M. Grewe.

1997. Global population structure of yellowfin tuna, Thunnus albacares, inferred from allozyme and mitochondrial DNA variation. Fish. Bull. 95:566-575.

Weir, B. S.

1990. Genetic data analysis, 377 p. Sinauer, Sunderland, MA.

Weir, B. S., and C. C. Cockerham.

1984. Estimating F statistics for the analysis of population structure. Evolution 38:1358-1370.

Welsh, J., C. Petersen, and M. McCelland.

1991. Polymorphisms generated by arbitrary primed PCR in the mouse: application to strain identification and genetic mapping. Nucl. Acids Res. 19:303-306.

Wild, A.

1994. A review of the biology and fisheries for yellowfin tuna, T. albacares, in the eastern Pacific Ocean. FAO Fish. Tech. Pap. 336(2):52-107.

Williams, J. G, A. R. Kubelik, J. Livak, J. A. Rafalaski, and S. V. Tingey.

1990. DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucl. Acids Res. 18: 6531-6535.

Pindaro Diaz-Jaimes

Instituto de Ciencias del Mar y Limnologia

Universidad Nacional Autonoma de Mexico

Circuito exterior de Ciudad Universitaria

Apdo. Postal 70-305

Mexico, D.F. 04510

E-mail address: [e-mail]

Manuel Uribe-Alcocer

Instituto de Ciencias del Mar y Limnologia

Universidad Nacional Autonoma de Mexico

Circuito exterior de Ciudad Universitaria

Apdo. Postal 70-305

Mexico, D.F. 04510

Manuscript approved for publication 19 June 2003 by Scientific Editor.

Manuscript recieved 26 June 2003 at NMFS Scientific Publications Office.