Reconciling patterns of inter-ocean molecular variance from four classes of molecular markers in blue marlin (Makaira nigricans)

• Published in: Molecular ecology 2001-05, Vol.10 (5), p.1179-1196
• Main Authors: Buonaccorsi, V. P., Mcdowell, J. R., Graves, J. E.
• Format: Article
• Language: English
• Subjects: Animals; Atlantic Ocean; Biological Evolution; Computer Simulation; DNA, Mitochondrial; Fishes - genetics; Genetic Markers; Genetic Variation; Haplotypes; Heterozygote; Linkage Disequilibrium; Makaira nigricans; Marine; microsatellite DNA; Microsatellite Repeats; Models, Genetic; mtDNA; Pacific Ocean; Polymorphism, Restriction Fragment Length; population genetics; population structure; power
• ISSN: 0962-1083; 1365-294X
• DOI: 10.1046/j.1365-294X.2001.01270.x

Different classes of molecular markers occasionally yield discordant views of population structure within a species. Here, we examine the distribution of molecular variance from 14 polymorphic loci comprising four classes of molecular markers within ≈ 400 blue marlin individuals (Makaira nigricans). Samples were collected from the Atlantic and Pacific Oceans over 5 years. Data from five hypervariable tetranucleotide microsatellite loci and restriction fragment length polymorphism (RFLP) analysis of whole molecule mitochondrial DNA (mtDNA) were reported and compared with previous analyses of allozyme and single‐copy nuclear DNA (scnDNA) loci. Temporal variance in allele frequencies was nonsignificant in nearly all cases. Mitochondrial and microsatellite loci revealed striking phylogeographic partitioning among Atlantic and Pacific Ocean samples. A large cluster of alleles was present almost exclusively in Atlantic individuals at one microsatellite locus and for mtDNA, suggesting that, if gene flow occurs, it is likely to be unidirectional from Pacific to Atlantic oceans. Mitochondrial DNA inter‐ocean divergence (FST) was almost four times greater than microsatellite or combined nuclear divergences including allozyme and scnDNA markers. Estimates of Neu varied by five orders of magnitude among marker classes. Using mathematical and computer simulation approaches, we show that substantially different distributions of FST are expected from marker classes that differ in mode of inheritance and rate of mutation, without influence of natural selection or sex‐biased dispersal. Furthermore, divergent FST values can be reconciled by quantifying the balance between genetic drift, mutation and migration. These results illustrate the usefulness of a mitochondrial analysis of population history, and relative precision of nuclear estimates of gene flow based on a mean of several loci.