260 
Fishery Bulletin 11 6(3-4) 
Table 1 
For the simulated data scenario produced by 4 competing models, total yearly abundance estimates <[/ Tot ) 
for juvenile Chinook salmon (Oncorhynchus tshawytscha) in Idaho, bias from true L/r 0 t *23,527 individuals), 
95% credible intervals (CIs), credible interval width in relation to the total yearly abundance estimate as a 
percentage (% Cl width), and the percentages of temporal strata with 95% Cl bounds that covered the true 
value of unmarked abundance (Strata coverage %). The M PS model structure includes pooled capture prob¬ 
ability and simple abundance parameters, the Mhw model has a hierarchical within-year parameter struc¬ 
ture, and the Mspline model uses a P-spline hierarchical prior for unmarked abundance. The hierarchical 
multiyear M HB model allows information about capture probability and abundance parameters during the 
same temporal period to be shared among years. Data scenarios may have excluded data for 4 or 8 temporal 
strata and may have reduced the number of marked and recaptured individuals by 60% across all strata. 
Because of predetermined P-spline knot characteristics, the M SPLINE model was not able to produce estimates 
for data missing 8 strata. 
Models 
Data scenario 
E/ T ot 
Bias 
95% Cl 
% Cl 
width 
% coverage 
of strata 
M PS 
Full 
25,944 
2467 
(25,214, 
26,685) 
6 
34.3 
4 strata' 
23,059 
-418 
(22,455, 
23,682) 
5 
77.1 
4 strata 60% reduction 
24,205 
728 
(23,351, 
25,163) 
7 
85.7 
8 strata 60% reduction 1 
26,640 
3163 
(25,528, 
27,841) 
9 
31.4 
M hw 
Full 
24,020 
543 
(22,484, 
25,717) 
13 
0.8 
4 strata 
24,492 
1015 
(22,412, 
35,035) 
52 
88.6 
4 strata 60% reduction 
23,934 
457 
(21,771, 
34,895) 
55 
88.6 
8 strata 60% reduction 
30,541 
7064 
(25,780, 
45,167) 
63 
82.9 
Mspline 
Full 
23,993 
516 
(21,939, 
26,298) 
18 
82.9 
4 strata 
22,659 
-818 
(20,955, 
24,586) 
16 
77.1 
4 strata 60% reduction 
22,697 
-780 
(20,401, 
25,606) 
23 
77.1 
8 strata 60% reduction 
- 
- 
- 
- 
- 
Mhb 
Full 
23,245 
-232 
(22,538, 
23,976) 
6 
100 
4 strata 
23,216 
-261 
(22,517, 
23,968) 
6 
100 
4 strata 60% reduction 
22,509 
-968 
(21,536, 
23,594) 
9 
100 
8 strata 60% reduction 
22,552 
-925 
(21,534, 
23,698) 
10 
100 
'Posterior parameter distributions for strata that were missing data were removed from the analysis to simu¬ 
late a naive Lincoln-Peterson estimator. 
Table 2 
Total abundance estimate (C/ Tot ), 95% credible intervals (CIs), absolute Cl width, and percent Cl width 
for rotary screw trap mark-recapture data collected for juvenile Chinook salmon (Oncorhynchus tshawyts¬ 
cha) at Marsh Creek and Big Creek, Idaho, in 2014 from the pooled-simple (M PS ), hierarchical within-year 
(M hw ), hierarchical penalized-spline (M SPLINE ), and hierarchical multiyear (M HB ) models. 
Model 
t>Tot 
95% Cl 
Cl width 
% Cl width 
Marsh Creek 
M PS 7 
104,594 
(103,149, 106,064) 
2915 
3 
M hw 
121,718 
(110,295, 208,526) 
98,231 
81 
Mspline 
112,299 
(105,910, 120,212) 
14,302 
13 
M hb 
114,035 
(106,991, 123,920) 
16,929 
15 
Big Creek 
Mps 1 
106,141 
(101,856, 104,729) 
2873 
3 
M hw 
216,292 
(154,931, 584,029) 
425,498 
197 
Mspline 
NA 
NA 
NA 
NA 
Mhb 
148,110 
(119,758, 239,889) 
120,131 
81 
' Posterior parameter distributions for temporal strata that were missing data were removed from the analysis. 
