136 
Fishery Bulletin 116(2) 
Figure 5 
Cumulative proportion of Dungeness crab (Cancer magister) that were tagged 
and released during commercial fishing trips off Oregon between October 
2012 and April 2014 and whose tags were returned over the duration of the 
study, by release event (year and month released). Darker symbols indicate 
release events earlier in the fishing season, fading in color as the season 
advances (from December to July). The gray background indicates when the 
fishing season was open. 
dropped from 3 m, 18 from 6 m, and 22 from 8 m. Re¬ 
spectively, 83%, 89%, and 82% were male; and only 1 
soft-shell male and 3 soft-shell females were dropped 
for all heights combined. By drop height, 0% (3 m), 0% 
(6 m), and 14% (8 m), respectively, died within 5 d of 
holding. From 8 m, only males died (3 of 18 males com¬ 
pared to 0 of 4 females), and all dead crab had major 
carapace cracks. 
Results from logistic regression indicated that mor¬ 
tality from the return to water is primarily influenced 
by whether or not the carapace cracks as a result of 
the drop. Whether a carapace cracked was closely 
linked with drop height and whether or not the crab 
was in water before being dropped. Both experiments 
combined, by height, revealed that 0% (1 m), 0% (3 m), 
6% (6 m), and 33% (8 m) of crab had cracked cara¬ 
paces as a result of the drop, indicating an increased 
probability of cracking as drop height increases. Con¬ 
sidering whether or not the crab was left out of water 
before it was dropped, for just the 8 m drop, 55% of 
those left out of water (experiment 1) and 14% of those 
kept in water (experiment 2) had cracked carapaces. 
These results suggest that mortality could be linked 
with whether or not a carapace cracks, which is po¬ 
tentially influenced by both drop height and whether 
or not a crab is left out of water before being returned 
to water. Further investigation is needed to determine 
the role of shell hardness in induc¬ 
ing mortality. 
Discussion 
Patterns in delayed discard mortality 
based on tag-returns 
Commercial ocean fishery Probabil¬ 
ity of tag return for the commer¬ 
cial ocean fishery was influenced by 
reflex impairment score, sex, shell 
hardness, carapace width, and time 
of release in relation to the open¬ 
ing of the fishing season. Males 
had higher tag-return rates than 
females and, for both sexes, hard¬ 
shell crab had higher rates than 
soft-shell crab. Similar percentag¬ 
es of score-0 crab were tagged for 
each category, which suggests that 
differences in returns were not in¬ 
fluenced by discrepancies in compo¬ 
sition of tagged crab by score, but 
rather by differences in survival. 
The few returns of tagged crab with 
scores greater than 0 (36) limited 
an analysis of return trends by in¬ 
dividual score. However, decreasing 
returns as scores increased for fe¬ 
males and soft-shell males and low¬ 
er scores, on average, for returned 
crab are consistent with the idea that crab with higher 
reflex impairment scores had higher rates of mortal¬ 
ity. Moreover, lower average scores for noninjured crab 
compared with injured crab supports the assumption 
that injury is reflected in the score. These tag-return 
patterns mirror results from the laboratory study in 
which delayed discard mortality rates were dependent 
upon level of reflex impairment (increasing mortality 
with higher scores), sex (higher mortality for females), 
and shell hardness (soft-shell males had higher mortal¬ 
ity than hard-shell males; few soft-shell females were 
captured requiring soft- and hard-shell females to be 
combined for the laboratory study). 
Patterns in tag-return rates indicated potential bias 
due to carapace width and month of release. For both 
sexes, the frequency of returns was higher for crab 
with larger carapace widths than for smaller tagged 
crab, possibly indicating that size affected the likeli¬ 
hood of return. This could be due to the fact that fish¬ 
ermen immediately discard crab with carapaces that 
are clearly smaller than the legal-size requirement 
without evaluation. In addition, a tag was more likely 
to be returned if tagged and released closer to the be¬ 
ginning of the season, likely because effort is highest 
at this time (i.e., there are more vessels on the water 
and, therefore, more pots pulled) and decreases over 
the fishing season. 
