Spatial variability of stream bed scour and fill: a comparison of scour depth in chum salmon (Oncorhynchus keta) redds and adjacent bed

Scour depth in egg pockets of chum salmon (Oncorhynchus keta) egg nests (redds) in a short gravel-bed spawning reach (45 × 20 m) of Kanaka Creek, British Columbia, was not significantly different from that in the adjacent bed during 1997-1998 winter flood events, whereas the scour depth in tailspills of redds was greater. Over the course of the incubation period, none of the egg pocket locations (zero of four), all of the tailspills (three of three), and 17% of the immediately adjacent bed locations (three of 18) scoured to the assumed egg burial depth of 15 cm below the initial postspawning surface elevation. Egg pocket scour depth has not previously been monitored, and the reliance of earlier studies on tailspill monitoring as an indication of redd scour may have led to faulty assessment of embryo loss. Only one flood event, which exceeded bankfull, caused widespread and deep scour and fill. Despite implementation of the most spatially intensive array of wiffle-ball scour monitors to date, scour was so variable that there was no spatial autocorrelation of scour depths.