138 
Fishery Bulletin 11 6(2) 
ors do not impair the reflexes). In analyses, in absence 
of a control to evaluate score-0 crab and with low num¬ 
bers of reflex-impaired animals, crab with scores great¬ 
er than zero were combined. Therefore, crab with low 
mortality probabilities (e.g., score-1) were mixed with 
moribund animals (score-6), which obfuscated a clear 
comparison of crab with different levels of impairment 
and prevented the detection of statistical differences 
in relative short- and long-term survival rates between 
the laboratory and tagging studies. 
These limitations highlight the importance of be¬ 
ing able to tag and recapture a sufficient number of 
animals for each reflex impairment score when exe¬ 
cuting a tag-return study using the RAMP approach. 
Hueter et al. (2006), in describing the methods used 
here to look at relative short-term survival, tagged a 
minimum of 365 animals per condition. For the com¬ 
mercial ocean fishery, which had the highest num¬ 
ber of crab with scores greater than 0, of the 5594 
crab assessed overall, only 202 had reflex impair¬ 
ment scores greater than 1 (129 score-2; 46 score-3; 
14 score-4; 5 score-5; and 8 score-6; Yochum et al., 
2017). These sample sizes are even smaller when di¬ 
vided into subcategories for sex and shell hardness, 
and when the lower-impact fisheries are considered. 
Moreover, the requirement of tagging a sufficient 
number of soft-shell crab becomes logistically imprac¬ 
tical because the highest return rates are linked with 
release events early in the fishing season, a time 
when there are few to no soft-shell animals. 
The efficacy of applying a tag-return approach for 
discard mortality research is also linked with the ex¬ 
tensive list of factors that determine whether or not 
a tag will be returned. For this study, these included 
natural mortality, fishing-induced mortality (both re¬ 
tention and handling-induced mortality if a tagged ani¬ 
mal is recaptured and released without the tag being 
observed), catchability, molt failure attributed to the 
tag, tag loss, tag-induced mortality, and reporting rate. 
Each of these factors is a potential source of bias in 
tag-return rates. In the case of Dungeness crab, natural 
mortality, for example, differs for crab that are and are 
not molting (Zhang et al., 2004). In addition, fishing- 
induced mortality must be considered when evaluating 
soft-shell crab that, after a period of time, become hard 
and recruit back into the fishery, or when evaluating 
sublegal crab that become legal size after molting dur¬ 
ing the study period. Moreover, different catchabilities 
could apply to recently molted male crab (Taggart et 
al., 2004), female crab based on whether or not they 
are egg-bearing (Swiney et al., 2003), and crab with 
different reflex impairment scores if a reduction in vi¬ 
tality affects a crab’s inclination to eat and therefore 
enter a pot. Also, for the commercial fishery, there are 
temporal and spatial patterns in fishing practices (Got- 
shall, 1978; Barry 10 ) that could affect tag return rates 
differently. The majority of effort and landings for the 
commercial Dungeness crab fishery occurs shortly af¬ 
ter the season opens in December and during the first 
2 months, after which fishermen often switch to an 
alternative fishery (Didier, 2002; Goblirsch and The- 
berge 11 ). Because of this temporal trend, crab tagged 
and released near the beginning of the season (or crab 
at large during multiple season openings) have a high¬ 
er chance of being observed than those tagged toward 
the end of the season when fishing effort is lower. The 
physical presence of the tag may also affect return 
rates given its potential to impede molting, to be lost, 
or contribute to mortality. In addition, tag reporting 
is potentially dependent on successful and extensive 
outreach, and the willingness and ability of fishermen 
to participate (Pollock et al., 2001). Finally, conducting 
a tag-return study on discarded animals can be chal¬ 
lenging because they are less likely to be inspected 
than retained animals. Furthermore, some fishermen 
reported that, because female and sublegal-size male 
crab are not allowed to be retained, they were not sure 
if they were supposed to remove the tag or leave it in 
place. However, if sufficient tag returns permit rela¬ 
tive analysis by score for a single release event, then 
many of the aforementioned variables of concern be¬ 
come irrelevant. 
Reliability of estimating mortality under captive holding 
conditions 
To address the goal of the field validation in ascertain¬ 
ing the accuracy of the laboratory-based mortality rates 
estimated by Yochum et al. (2017) for Dungeness crab, 
we conclude that, although direct rate comparisons 
were not possible, similar patterns in mortality- and 
tag return- rates from the 2 studies lend support to the 
reliability of the laboratory-generated rates. However, 
results from the tagging study suggest that laboratory- 
based mortality rates for discarded females should be 
evaluated to determine whether there is a discrepancy 
between hard- and soft-shell crab. In addition, mortal¬ 
ity would be underestimated if the laboratory-based 
commercial rates were applied to a shoreside fishery 
where the discarded crab fell more than 6 m (and po¬ 
tentially less). With respect to chronic mortality, for all 
fisheries, there was no conclusive evidence that sug¬ 
gests mortality rates were underestimated in the labo¬ 
ratory study because of short-term holding. 
In general, trade-offs between laboratory holding 
and tag-return for estimating discard mortality rates 
depend on the overall objectives of the study, logistic 
constraints, and the level of impairment caused by 
the stressor(s) being studied. For our 2 studies, the 
laboratory-based RAMP approach was superior in its 
ability to quantify mortality rates and to detect signifi¬ 
cant differences among several variables. Regardless, 
the tag-return study identified important influences 
on mortality and allowed increased collaboration and 
outreach in the fishing community. In addition, advan¬ 
tages of the tag-return approach over captive holding 
include the fact that the conditions with the former 
u Goblirsch, G., and S. Theberge. 2008. Traps. Sea Grant 
Oregon ORESU-G-08-002, 2 p. Oregon State Univ., Corval¬ 
lis, OR. [Available from website.] 
