Lenarz and Shaw: Estimates of tag loss from double-tagged Anoplopoma fimbria 
297 
Table 1 
Double-tag releases and recoveries of sablefish, Anoplo- 
poma fimbria, during first six years at liberty. Number of 
releases are shown in parentheses. 
Recoveries 
Single tag 
Years at liberty 
(Midpoint) Both tags Anterior 
Posterior 
Total 
1986 releases (2,652) 
0.5 
116 
21 
12 
49 
1.5 
77 
10 
13 
100 
2.5 
29 
8 
6 
43 
3.5 
37 
11 
5 
53 
4.5 
16 
18 
3 
37 
5.5 
31 
17 
8 
56 
Total 
306 
85 
47 
438 
1987 releases (1,872) 
0.5 
74 
3 
5 
82 
1.5 
16 
4 
1 
21 
2.5 
19 
7 
2 
28 
3.5 
19 
3 
4 
26 
4.5 
11 
5 
2 
18 
5.5 
11 
6 
1 
18 
Total 
150 
28 
15 
193 
1988 releases (6,021) 
0.5 
272 
16 
11 
299 
1.5 
159 
34 
14 
207 
2.5 
98 
23 
12 
133 
3.5 
86 
26 
14 
126 
4.5 
37 
4 
11 
52 
5.5 
26 
16 
4 
46 
Total 
678 
119 
66 
863 
Total releases (10,545) 
0.5 
462 
40 
28 
530 
1.5 
252 
48 
28 
328 
2.5 
146 
38 
20 
204 
3.5 
142 
40 
23 
205 
4.5 
64 
27 
16 
107 
5.5 
68 
39 
13 
120 
Total 1,134 
232 
128 
1,494 
Table 2 
Maximum-likelihood estimates of rates of immediate tag 
retention ( p ) and tag-shedding rates for anterior tags ( L p 
and posterior tags ( L 2 )for sablefish. Also shown are esti- 
mates of the averages, medians, standard deviations, and 
ranges of the rates from 2,000 bootstrap replicates. 
Parameter 
P 
^2 
Maximum-likelihood estimate 
0.9516 
0.0304 
0.0694 
Bootstrap average 
0.9517 
0.0304 
0.0693 
Median 
0.9519 
0.0302 
0.0694 
Standard deviation 
0.0098 
0.0062 
0.0075 
Minimum 
0.9176 
0.0108 
0.0457 
Maximum 
0.9855 
0.0515 
0.0968 
ging is necessary to estimate tag-loss rates. Thus we 
recommend that double tagging be considered, when 
feasible, for at least a portion of any tagging study. 
The number of fish released in our study was not 
affected by double tagging. It is possible, however, 
that in some situations double tagging could increase 
the time required to process fish so as to decrease the 
number of fish released. The tradeoff between the po- 
tential reduction in number of fish released and the 
potential increase in number of fish recovered should 
be considered when designing a tagging program. 
In summary, analysis of returns from double-tag re- 
leases indicates that initial shedding of tags was 0.048. 
The long-term instantaneous rates of shedding were 
0.030 and 0.069 for the anterior and posterior positions, 
respectively. Because there was a difference in the long- 
term instantaneous rates and because fish released 
with single tags are only tagged in the anterior posi- 
tion, corrections made for single-tagging experiments 
should be done only with the anterior tag loss rates. 
ging the fish, the total recoveries appeared to be in- 
creased by 9%. The cost of the double tagging was 
low compared to the cost that would have been in- 
curred by increasing time at sea by 9%. 
The parameter estimates of this study indicated 
that by the middle of the sixth recovery period, 19% 
of the anterior tags ( J 1>6 ) and 35% of the posterior 
tags ( J 2 6 ) had been shed, and 7% of the fish had 
lost both tags (( J 1>6 ) ( J 2 6 )). Thus, even though shed- 
ding rates are low for sablefish, these rates are suffi- 
ciently high to affect analysis of tag-return data from 
this long-lived species. 
Tag-shedding rates were high enough in many of 
the reviewed studies to warrant incorporation of tag- 
loss rates in analysis of tag-return data. Double tag- 
Acknowledgments 
We owe considerable thanks to the many people who 
tagged and recovered sablefish. In particular, we are 
indebted to Norm Parks for the role he played in or- 
ganizing and implementing the trap surveys. We thank 
Steve Ralston for helping us with SAS and Mark 
Wilkins, Steve Ralston, and two anonymous reviewers 
for their constructive reviews of the paper. 
Literature cited 
Baglin, R. E., Jr., M. I. Farber, W. H. Lenarz, and 
J. M. Mason Jr. 
1980. Shedding rates of plastic and metal darttags from 
