Scour of Chinook Salmon Redds on Suction Dredge Tailings

-We measured scour of the redds of chinook salmon Oncorhynchus tshawytscha on dredge tailings and natural substrates in three tributaries of the Klamath River, California. We measured maximum scour with scour chains and net scour by surveying before and after high winter flows. Scour of chinook salmon redds located on dredge tailings exceeded scour of redds on natural substrates, although the difference varied among streams. Our results show that fisheries managers should consider the potential negative effects of dredge tailings on the spawning success of fall-spawning fishes such as chinook salmon and coho salmon O. kisutch. Suction dredging for gold is common in many streams and rivers in western North America and in gold-bearing lotic habitats worldwide (Hall 1988). Studies of the effects of dredging on fishes have focused on survival following entrainment (Griffith and Andrews 1981) or the immediate responses of fishes to changes in habitat caused by dredging (Harvey 1986). The effect of suction gold dredging on fish spawning has not been studied, in part because dredging rarely overlaps in time with spawning by species of special concern to fisheries managers. Also, in many unregulated streams, most fishes spawn in spring after dredge tailings from the previous summer and fall are redistributed by high winter flows (Thomas 1985; Harvey 1986). However, dredging during summer may affect the reproductive success of fall-spawning fishes such as chinook salmon Oncorhynchus tshawytscha. Because of low streamflow during late summer and early fall, dredge tailings may retain their original form during the spawning period of these species. Tailings often contain substrate appropriate for redd construction and may be used by fallspawning salmonids. The significance of dredge tailings to fish populations depends in part on the extent dredge tailings are used for spawning, which is itself probably affected by the availability of suitable unaltered substrates and the relative quality of dredge tailings as spawning sites. Because dredge tailings may be more unstable than * Corresponding author: [email protected] Received June 24, 1998 Accepted October 6, 1998 natural substrates, redds on tailings may be subject to greater scour than those on unaltered substrates. Greater scour of tailings would significantly decrease their quality as spawning sites because mortality of preemergent salmonids can be sensitive to small increases in scour depth (Holtby and Healey 1986; Montgomery et al. 1996). Our objective in this study was to test the null hypothesis that chinook salmon redds on dredge tailings and those on natural substrates are scoured equally. Methods Study sites.-We made scour measurements in three tributaries of the Klamath River in Siskiyou County, northwestern California: Elk Creek, the South Fork Salmon River, and the Scott River (Table 1). Regional streamflow is highly seasonal; most peak flows result from rainfall or rain-onsnow events during wet winters. Suction dredging occurs from June to September. Spawning by chinook salmon occurs most often in October and November, and storm flows capable of mobilizing streambed material typically occur from December to March. All study reaches are single-thread, slightly sinuous alluvial channels with limited floodplains bounded by valley walls. Bed surfaces are predominated by cobbles and boulders. Scour and fill of the streambeds can be expected to vary annually because, although these channels have high sediment supplies typical of the Klamath Mountains, winter streamflow is highly variable. Scour and fill also can be expected to vary spatially because patches of gravel are transported annually whereas boulders move less frequently. Measurement of scour.-We measured maximum and net scour of redds on dredge tailings and on natural substrate. We directly measured maximum scour at each redd with two scour chains positioned on either side of the redd where bed elevation approximately equaled the surrounding substrate, about midway along the longitudinal axis of the tailspill. We chose these chain locations to avoid damaging embryos and to measure local scour depths presumably equal to those at the bracketed redds. We inserted scour chains in October and November, before large increases in stream discharge obscured the locations of redds