We used data from 14 microsatellite DNA loci in five recent collections of chinook salmon in the Puyallup River system to characterize White River spring-run chinook salmon. Patterns of allele frequency variation and direct tests of population differentiation revealed that the two hatchery populations of White River chinook (at the White River Hatchery and at the Hupp Springs Hatchery) were genetically similar to each other but were very different from the Voights Creek Hatchery fall-run stock. In contrast, a collection of unmarked and untagged adults sampled at the Buckley trap on the White River in 1998 and a collection of naturally produced smolts collected in the lower White River in 2000 were each somewhat intermediate in their genetic characteristics (allelic composition and allele frequencies) to the spring-run and fall-run hatchery populations. Assignment of individual adults and of individual outmigrating smolts to their most likely stock-of-origin using the microsatellite DNA data indicated that both groups consisted of mixtures of fish of spring-run and fall-run origin (and/or contained fish of inter-stock hybrid origin). Estimates for each group were approximately 65-70% spring-run ancestry (and 30-35% fall-run ancestry).
We reviewed the historical patterns of arrival-timing exhibited by chinook in the White River using data from Buckley trap interception records from 1949 through 2000. This process revealed that the predominant pattern of chinook arrivals at the Buckley Diversion Dam (@ RM 24.3) from 1949 through 1970 was that of a spring-timed chinook run (majority of arrivals prior to July 1st) with a small percentage of fall-timed individuals (arrivals after September 30th). In contrast, substantial runs of later-timed chinook (average of 88% of the run arrived at Buckley after the second week in August) dominated the arrivals in the nine years from 1987 through 1995. Because the Buckley trap arrival patterns from 1971 through 1986 were much more variable and the run sizes from 1977 through 1986 were much smaller due to the intensive recovery program operations in these years, it was difficult to generalize about the nature of the chinook runs during this time period. Recent arrival patterns (1996 to present) showed a slightly increasing percentage of spring-timed fish and a concomitantly decreasing proportion of falltimed fish at the Buckley trap. The increased numbers (and proportion) of spring-timed chinook and decreased proportion of summer/fall-timed chinook seen in the upper White River in the past five years may be attributable to successful spring-run chinook production by the multi-agency, hatchery-based White River Spring Chinook Recovery Program.
We also examined the patterns of estimated historical gillnet fishery harvests in the lower White and Puyallup Rivers in five selected years for information regarding patterns of run-timing. The temporal patterns of estimated catches in the years examined suggested that both spring-timed and fall-timed chinook runs were present in the lower White and/or Puyallup Rivers in 1954, 1955, 1957, 1966, and 1977. The timing of Buckley trap arrivals and the temporal patterns of these estimated fishery harvests suggested to us that the fall-timed chinook were confined to the lower White River (below the Buckley trap) from 1954 through 1977 (except in 1973). Whether or not a native fall-timed chinook run occurred in the upper White River prior to 1954 is unclear, but Buckley trap counts for 1949 - 1953 suggest this possibility.
Although the basis for the appearance of the fall-timed run in the years since 1977 is unknown, increased minimum in-stream flows since 1987 and/or changes in hatchery operations may well have contributed to this change.
Given that there is little evidence that the fall-timed chinook run observed in the upper White River in the last 20 or so years is native to this system, we are concerned about the potential for interbreeding and/or competition between the presumed native White River spring-run stock and these fall-timed fish.