Environmental Biology of Fishes 69: 37–50, 2004. © 2004 Kluwer Academic Publishers. Printed in the Netherlands. Genetic population structure of chum salmon in the Pacific Rim inferred from mitochondrial DNA sequence variation Shunpei Sato a , Hiroyuki Kojima b , Junko Ando a , Hironori Ando a , Richard L. Wilmot c , Lisa W. Seeb d , Vladimir Efremov e , Larry LeClair f , Wally Buchholz g , Deuk-Hee Jin h , Shigehiko Urawa i , Masahide Kaeriyama b , Akihisa Urano a,j & Syuiti Abe k,l a Division of Biological Science, Graduate School of Science, Hokkaido University, Sapporo 060-0810, Japan b Graduate School of Science and Engineering, Hokkaido Tokai University, Sapporo 005-8601, Japan c Auke Bay Laboratory, Alaska Fisheries Science Center, NOAA, Juneau, U.S.A. d Alaska Department of Fish and Game, Anchorage, U.S.A. e Russian Academy of Science, Vladivostok, Russia f Washington Department of Fish and Wildlife, Olympia, Washington, U.S.A. g U.S. Fish and Wildlife Service, Anchorage, AK, U.S.A. h Kangnung National University, Kangnung, Korea i Salmon Resources Center, Sapporo 062-0922, Japan j Field Science Center, Hokkaido University, Sapporo 060-0811, Japan k Laboratory of Animal Cytogenetics, Center for Advanced Science and Technology, Hokkaido University, Sapporo 060-0810, Japan (e-mail: sabe@ees.hokudai.ac.jp) l Laboratory of Breeding Science, Graduate School of Fisheries Sciences, Hokkaido University, Hakodate 041-8611, Japan Received 17 April 2003 Accepted 27 April 2003 Key words: mtDNA control region, Pacific salmon, haplotype genealogy, genetic divergence Synopsis We examined the genetic population structure of chum salmon, Oncorhynchus keta, in the Pacific Rim using mitochondrial (mt) DNA analysis. Nucleotide sequence analysis of about 500 bp in the variable portion of the 5 ′ end of the mtDNA control region revealed 20 variable nucleotide sites, which defined 30 haplotypes of three genealogical clades (A, B, and C), in more than 2,100 individuals of 48 populations from Japan (16), Korea (1), Russia (10), and North America (21 from Alaska, British Columbia, and Washington). The observed haplotypes were mostly associated with geographic regions, in that clade A and C haplotypes characterized Asian populations and clade B haplotypes distinguished North American populations. The haplotype diversity was highest in the Japanese populations, suggesting a greater genetic variation in the populations of Japan than those of Russia and North America. The analysis of molecular variance and contingency χ 2 tests demonstrated strong structuring among the three geographic groups of populations and weak to moderate structuring within Japanese and North American populations. These results suggest that the observed geographic pattern might be influenced primarily by historic expansions or colonizations and secondarily by low or restricted gene flow between local groups within regions. In addition to the analysis of population structure, mtDNA data may be useful for constructing a baseline for stock identification of mixed populations of high seas chum salmon. Introduction Chum salmon, Oncorhynchus keta, has the widest nat- ural geographical distribution among all Pacific salmon species in the North Pacific Rim, ranging from Korea and Japan northward to the Arctic coasts of Russia and North America and then southward to Oregon (Salo 1991). Spawning adults, like other Pacific salmons, are