296 
Fishery Bulletin 95(2), 1 997 
L 
Figure 2 
Estimated distribution functions of instantaneous tag-shedding rates, 
L ; , of anterior and posterior tags for sablefish from a 2,000 replicate 
bootstrap. Intersections of the vertical lines with the distribution func- 
tions mark the estimated 90% confidence bands. 
Figure 3 
Observed and expected double- and single-tagged (anterior and poste- 
rior tags separately) recoveries of sablefish. Expected values were cal- 
culated from maximum-likelihood parameter estimates. 
another time. Although we believe that most, if not all, 
reports of single tag recaptures were accurate, 
misreporting may have caused underestimation of p. 
Tag-loss rates in this study are similar to those of 
Beamish and McFarlane (1988) for sablefish. They 
used two types of tags (anchor and suture) and did 
not find a significant difference in the rate of loss by 
tag type. From a line fitted by eye through the data, 
they found a loss rate of approximately 10% during the 
first year and 2% per year thereafter. Examination of 
Figure 2 of their paper indicated that p was about 0.95. 
We present tag-loss rates from sablefish and other 
species in Table 3. Values were taken from the lit- 
erature and standardized, as much as was feasible 
within limitations, owing to the variety of models 
used and plethora of reporting styles. The median es- 
timate of L was 0.15, and the range was 0.00 to 3.93. 
Estimates of L for most species were higher than that 
for sablefish. The distribution ofL estimates had a rela- 
tively long upper tail. Only a few of the other studies 
provided estimates of p, and the estimates for sablefish 
were in the middle of the range of the other estimates. 
Although tag-shedding rates for sablefish were low, 
it still appears worthwhile to double tag. During the 
six-year recovery period, 128 sablefish were recov- 
ered with only a posterior tag. Thus, by double tag- 
