Goetz and Quinn: Behavioral thermoregulation by adult Oncorhynchus tshawytscha prior to spawning 
267 
Temperature (°C) 
Figure 6 
Frequency distribution of hourly temperature records from tags implanted in Chinook salmon ( Onco¬ 
rhynchus tshawytscha ) in 2005 and 2006, by migration segment from Puget Sound, in western Wash¬ 
ington, to entry into a spawning tributary. In the order that a Chinook salmon would move through 
them, the migration segments are (A) Puget Sound, (B) the locks and (C) ship canal in the Lake 
Washington watershed, (D) Lake Washington, (E) Sammamish River, and (F) Lake Sammamish. 
2010), although Chinook salmon in Puget Sound did not 
display diel vertical migrations (Smith et al., 2015). Even 
with behavioral adaptation, fish residing in the upper 
estuary were continuously exposed to potentially stress¬ 
ful temperatures (18-21°C) while experiencing pulses 
of brackish water (1-18) and variable periods of low DO 
concentration (<6 mg/L). The available alternatives to this 
refuge were to revisit marine waters or swim up 10.8 km 
through even warmer waters (21-22°C) to Lake Washing¬ 
ton. Lake Union provided access to cooler water (14—15°C 
at a depth of 11 m) between the upper estuary and Lake 
Washington, but the salmon did not occupy this habitat. In 
that lake, DO levels below the thermocline were <5 mg/L 
in August and early September (USACE 2 ), a level causing 
2 USACE (U.S. Army Corp of Engineers). 2005-2006. Unpubl. 
data. P.O. Box 3755, Seattle, WA 98124-3755. 
avoidance or delay in migrating Pacific salmon (Fujioka, 
1970; Hallock et al., 1970; Alabaster, 1989). In Lake Wash¬ 
ington, Chinook salmon had access to cooler water below 
the thermocline without restriction by low DO concentra¬ 
tion, yet they did not occupy such depths, in contrast to 
sockeye salmon (Newell and Quinn, 2005). 
Comparisons to salmonids in other systems reveal a 
variety of responses to adverse conditions along the migra¬ 
tory corridor. Most Chinook salmon tagged in the Klam¬ 
ath River’s estuarine lagoon returned to the ocean and 
remained for about 10 d near the river plume before mov¬ 
ing upriver (Strange, 2013), as did Chinook salmon in our 
study although for a shorter period. Similarly, some sock- 
eye salmon populations hold in the Fraser River plume 
prior to upriver migration (Cooke et al., 2008). The migra¬ 
tory behavior of Pacific salmon returning to estuaries 
and rivers includes several species (e.g., Chinook salmon, 
