Back in the 90s the Adams River sockeye run arrived 3 to 4 weeks early and then held in the lower Fraser for quite some time. Many of those fish died of disease. Same thing happened to the Harrison/Weaver run around the same time. Those fish moved into the Harrison. Water temperatures were high and a large % died. I remember fishing the river in Sept for trout and the banks and river floor was covered with dead rotted salmon. In both cases this behaviour lasted 3 or 4 years then the runs shifted back to their historical timing.
Interesting to note the Harrison has been opened for sockeye for the first time since then. Prior to that event they were caught on 'krill', just the small cocktail shrimp you can buy at Safeway etc. No need to floss.
Below is a link to a paper on the phenomena I mentioned above. It lasted at least 7 years ('95 to 2002) and lead to very high mortalities of over 90% in some cases. At least at the time the catastrophic decline of Cultus Lake sockeye was attributed to this delay:
Abstract:
Late-run stocks are unique among Fraser River sockeye salmon populations in that they typically delay in Georgia Strait4 to 6 weeks prior to migrating upstream to spawn. Beginning in 1995, and continuing through 2002, this delay period to 6 weeks earlier than normal. Each year, the early river entry has been associated with high rates of en-route and prespawning mortality (>90% in 2000 and 2001). The abnormal behavior and mortality was less extreme in 2002 than the prior two years and results from a radio tagging program indicated that individual late-run sockeye exhibited a range of behaviors from little or no delay to near normal delay. The abnormal behavior and subsequent mortality is already threatening the viability of small populations (e.g. Cultus Lake sockeye). The Pacific Salmon Commission has held workshops and funded several studies to begin investigating the causes of this behavior. Many competing hypotheses have been proposed (e.g. involving physiology, environmental conditions, contaminants, parasites, and predators) regarding the reason(s) for early migration and high mortality. However, the cause of the behavior and mortality remain unknown. Precautionary management actions have resulted in large foregone catches of late-run and summer-run sockeye populations that co-migrate with late-run stocks in the fishing areas (e.g. Adams and Horsefly River sockeye). Similar behavior has been observed in other Fraser River salmon species, but the high mortality rates have not yet been observed
https://www.watershedwatch.ca/wp-content/uploads/2011/04/Exh-72-NonRT.pdf