548 
Fishery Bulletin 11 5(4) 
250 - 
E 200 
CL 
0 
"2 150 H 
> 
0 
o 
0 
cc 100 - 
50 
# 
♦v 
♦ ♦ 
% 
♦ 
© o CO 
o o 
^ o 
a O O O 
sD o 
o onshore-shallow 
♦ onshore-deep 
• offshore-deep 
—i-1-r 
500 1000 1500 2000 2500 
Distance to shore (m) 
Figure 2 
Scatterplot showing by depth (m) and distance to shore 
(m) the distribution of receivers' used to track Chinook 
salmon (Oncorhynchus tshawytscha ) in Puget Sound, 
Washington. The 3 categories used to group receivers 
were onshore-shallow (white circles; <1 km from shore 
and <95 m deep), onshore-deep (black diamonds; <1 km 
from shore and >95 m deep), and offshore-deep (black 
circles; >1 km from shore and >95 m deep). 
categories differed by receiver type across all basins by 
using a chi-square test. Spearman’s rank correlation 
analysis was used to determine whether the number of 
individual Chinook salmon detected correlated with the 
number of individual coho salmon detected at receivers 
that detected both species. Similarly, Spearman’s rank 
correlation analysis was used to compare the number 
of days Chinook and coho salmon were detected. All 
data analyses were performed in R, vers. 3.2.1 (R Core 
Team, 2015). 
Results 
Of the 87 fish tagged, 50 were excluded, having final 
detections less than 6 d after tagging (the minimum 
time it took fish to reach the Strait of Juan de Fuca), 
a period insufficient to detect departure. Given the ex¬ 
pansive area not covered by receivers we do not pre¬ 
sume to provide natural or tag-related mortality for 
these fish, but the lack of a prolonged detection his¬ 
tory precluded us from categorizing them as transient 
or resident. The 37 fish included in the analysis (for 
which we had detections more than 6 days after tag¬ 
ging) had longer fork lengths (FLs) than the excluded 
fish (mean: 273 vs 245 mm FL, f=-3.60, P<0.01). The 
proportions of natural and hatchery origin individu¬ 
als did not differ for included fish (11 of 37 wild; 30%) 
and excluded fish (16 of 50 wild, 32%) (x 2 <0.01, df=l, 
P=1.00). 
Movement from Puget Sound 
The 37 fish analyzed were considered resident at the 
time of tagging, but 11 (30%) subsequently left Puget 
Sound and were reclassified as transients, and 26 re¬ 
mained as residents. Eleven of the 37 were of natural 
origin and 26 were of hatchery origin. The origins of 
the transients (27% natural) and residents (31% natu¬ 
ral) did not differ (x 2 <0.01, df=l, P=1.00). Transients 
were detected in Puget Sound for as little as 1 d to as 
much as 124 d before exiting. There were too few indi¬ 
vidual departures for rigorous statistical analyses, but 
departures occurred throughout the year (January=l, 
April=2, June=3, July=2, November=2, December=l). 
The time of year when fish were tagged did not af¬ 
fect the likelihood of a fish staying as a resident. Of 
the 37 fish included in analyses, 16 of 20 (80%) fish 
tagged in June (summer) remained resident in Puget 
Sound, and 10 of 17 (59%) tagged in the fall through 
November-December (x 2 = 1-09, df=l, P=0.30). Fish that 
remained resident and those that left Puget Sound 
did not differ in fork length (mean: 278 vs. 262 mm, 
t=0.99, P- 0.34) or condition factor (mean: 1.21 vs. 1.23, 
/= -0.57, P=0.58). The total duration of detection for 
fish that stayed as residents within Puget Sound was 
not significantly influenced by fish length (coefficient of 
determination [r 2 ]=0.12, P=0.08). 
Movement among basins 
Detection data indicated that fish remaining as resi¬ 
dents seldom moved between basins. Twenty-six fish 
remained residents: 10 individuals (fish 1-10, Fig. 3) 
were never detected outside central Puget Sound; 9 
individuals moved only between Admiralty Inlet and 
central Puget Sound (fish 11-19, Fig. 3); 6 individuals 
moved between central Puget Sound, Admiralty Inlet, 
and the Whidbey Basin (fish 20-25, Fig. 3); and 1 fish 
moved between central Puget Sound, Admiralty Inlet, 
and Hood Canal (fish 26, Fig. 3). The 11 fish that left 
Puget Sound had different movement patterns (fish 
27-37, Fig. 3). The last known locations varied for wa¬ 
ters within Lime Kiln State Park on San Jun Island 
(fish 27-28, Fig. 3; Arostegui et al. 2017), the Strait of 
Juan de Fuca (fish 29-31, 33-35, 37, Fig. 3), waters off 
northwestern Vancouver Island near Lippy Point (Re- 
chisky et ah, 2013; fish 32, Fig. 3), and the Washington 
State coast at Willapa Bay (fish 36, Fig. 3). Interest¬ 
ingly, these fish were not detected on the Canadian 
Northern Strait of Georgia or Queen Charlotte subar¬ 
rays. Hood Canal had 119 receivers active during at 
least part of the study period but only 2 individuals 
(fish 26 and 30, Fig. 3) were detected there. Similarly, 
south Puget Sound had 50 active receivers and only 1 
fish was detected there (fish 27, Fig. 3). 
Movement within Puget Sound 
The site-use rank sum metric was used to measure fish 
use within Puget Sound. Fish that remained within 
