We analyzed intraspecific mitochondrial DNA variation in chinook salmon (Oncorhynchus tshawytscha) from drainages in the Yukon River (Alaska and Yukon Territory), the Kenai River (Alaska), and Oregon and California rivers; and chum salmon (O. keta) from the Yukon River and Vancouver Island, and Washington rivers. For each species, three different portions of the mtDNA molecule were amplified separately using the polymerase chain reaction and then digested with at least 19 restriction enzymes. Intraspecific sequence divergences between haplotypes were less than 0.01 base substitution per nucleotide. Nine chum salmon haplotypes were identified. Yukon River chum salmon stocks displayed more haplotypes (eight) than the stocks of Vancouver Island and Washington (two). The most common chum salmon haplotype occurred in all areas. Seven chinook salmon haplotypes were identified. Four haplotypes occurred in the Yukon and Kenai rivers and four occurred in Oregon/California, with only one haplotype shared between the regions. Sample sizes were too small to quantify the degree of stock separation among drainages, but the patterns of variation that we observed suggest utility of the technique in genetic stock identification.

From 1979 to 1982,188 chinook salmon (Oncorhynchus tshawytscha) were tagged with radio transmitters to locate spawning areas in the glacial Kenai River, southcentral Alaska. Results confirmed that an early run entered the river in May and June and spawned in tributaries, and a late run entered the river from late June through August and spawned in the main stem. Spawning peaked during August in tributaries influenced by lakes, but during July in other tributaries. Lakes may have increased fall and winter temperatures of downstream waters, enabling successful reproduction for later spawning fish within these tributaries. This hypothesis assumes that hatching and emergence can be completed in a shorter time in lake-influenced waters. The time of upstream migration and spawning (mid- to late August) of the late run is unique among chinook stocks in Cook Inlet. This behavior may have developed only because two large lakes (Kenai and Skilak) directly influence the main-stem Kenai River. If run timing is genetically controlled, and if the various components of the two runs are isolated stocks that have adapted to predictable stream temperatures, there are implications for stock transplantation programs and for any activities of man that alter stream temperatures.

No abstract

Genetic differentiation among subpopulations of sockeye salmon (Oncorhynchus nerka) was investigated within nine intensively sampled lake systems located throughout the species' range using allozyme allelic frequency data collected by researchers in Canada, Russia, and the United States. Allelic frequencies at up to nine highly polymorphic loci were used to examine genetic diversity among 163 samples collected from 68 distinct spawning sites and to identify subpopulation structure within lakes. Significant heterogeneity was detected among sites within all lakes. The greatest differentiation was evident among subpopulations exhibiting different run timing (earlier vs. later) or utilizing different spawning habitat (tributary vs. littoral). These findings indicate that sockeye home precisely to natal streams, not just to lake systems, and underscore the importance of conserving individual spawning sites within sockeye populations.

Substantial genetic divergence was found among chum salmon (Oncorhynchus keta) populations collected from North America and Russia. Five major groups of populations can be identified by geographic region: (1) lower Yukon River summer run; (2) upper Yukon River fall run; (3) Bristol Bay area; (4) Alaska Peninsula; and (5) Russia. Mean heterozygosities were 0.064, 0.062, 0.065, 0.064, and 0.063, respectively; and the percent polymorphic loci values at the 0.99 level were 33.7, 31.3, 32.6, 30.6, and 30.9%, respectively. The hierarchical gene diversity analysis showed that 95.42% of the diversity can be explained by heterogeneity within sites, 1.36% among sites, 0.49% between Yukon River run timing, 1.69% among areas, and 1.04% among countries. The Alaska Peninsula populations are genetically more similar to populations from Russia than to those from western Alaska, and two populations from the upper Yukon River are distinct from other nearby populations. Possible reasons for these findings concern the complex glacial histories of the watersheds.