42 
Fishery Bulletin 108(1) 
between surveys; 122 and 128 individuals captured 
in the first and second surveys, respectively. Of the 
total 250 individuals captured, the numbers caught 
among the pot types was as follows: 107 in lair pots, 
35 in Korean hair pots, 102 in sablefish pots, and 6 in 
shrimp pots. 
Sablefish pots captured significantly heavier indi- 
viduals than all other pot types (Fig. 3), but the weight 
of these giant octopuses captured in the other three 
pot types did not differ significantly from one another 
(Welch ANOVA, df=3, 32.241, F=28.115, PcO.OOl Tukey 
HSD). The depth at which pots were set, however, was 
not determined at random. Larger giant octopuses may 
be found in deeper water resulting in biased estimates. 
Although a regression of weight by depth showed a very 
weak correlation (coefficient of correlation (r 2 )=0.0019, 
P=0.49), in order to mitigate the potential confounding 
effects of depth, we re-analyzed weight by pot type us- 
ing only giant octopuses caught within Eldred Passage 
where depth did not differ significantly by pot type. 
Giant octopuses caught in sablefish pots remained sig- 
nificantly heavier than those caught in both lair and 
shrimp pots, but there was not a significant difference 
in mean weight between sablefish and Korean hair 
crab pots (ANOVA, df=3, F=10.599, PcO.OOl Fisher’s 
LSD). Among the 112 males and 126 females captured, 
males were heavier than females (/-test, df=l, F=7.166, 
P=0.008). However, this finding did not seem to be 
driving the observed difference in weight by pot type 
because for each pot type both sexes were equally likely 
to be captured (lair: % 2 = 0.154, P=0.695, Korean hair 
crab: % 2 =0.030, P=0.862, sablefish: x 2 =0.853, P=0.356, 
and shrimp: x 2 =0, P= 11- 
Capture efficiency varied both temporally and with 
the CPUE index (kg/days soaked vs. kg/pot set). Be- 
tween the two surveys the efficiency of both lair and 
Korean hair crab pots increased and the efficiency of 
sablefish pots decreased. The average CPUE for all pot 
types between surveys ranged from 5.5 kg/pot (±1.7 SD 
[standard deviation]) for the lair pots to 0.4 kg/pot for 
shrimp pots, and 17.6 kg/days soaked (±2.7 SD) for the 
sablefish pots to 2.1 kg/days soaked for shrimp pots 
(Fig. 4). Lair pots that caught giant octopuses were not 
soaked longer than pots that did not capture octopuses 
(/-test df=l, F=0.214, P=0.644). 
Both sablefish and Korean hair crab pots caught large 
numbers of commercially important crab and fish spe- 
cies as bycatch. Korean hair crab pots caught Tan- 
ner crabs ( Chionoecetes bairdi) and Pacific cod ( Gadus 
macrocephalus; sizes ranging from 26 to 100 cm total 
length). Sablefish pots caught C. bairdi, G. macrocepha- 
lus, and halibut ( Hippoglossus stenolepis ; sizes rang- 
ing from 29 to 112 cm total length). Lair and shrimp 
pots did not contain bycatch of commercially important 
species. Other bycatch species, caught in all four pot 
types were lyre crabs (Hyas lyratus), decorator crabs 
( Oregonia gracilis), sunflower sea stars ( Pycnopodia 
helianthoides), and Oregon hairy tritons (Fusitriton 
oregonensis ). 
Discussion 
The internal volume of each pot may be a better indica- 
tor than the size of the pot entrance for the potential 
size of giant octopuses that will be caught. Although lair, 
Korean hair crab, and sablefish pots had similar size 
openings, the success of the sablefish pots may have been 
due to their volume being three times greater than that 
of the Korean hair crab pots and nearly twenty times 
that of the lair pots. The ability of 
giant octopuses to enter a trap is 
most likely determined by the size 
of its beak, the only hard part of 
its body. 
Variation in the size of North 
Pacific giant octopuses caught by 
the different pot types may be in- 
fluenced by differences in depth or 
substrate rather than by pot type 
alone. However, when pots laid at 
similar depths in Eldred Passage 
were analyzed, we observed similar 
trends in the data, indicating that 
sablefish pots have the potential to 
capture larger giant octopuses than 
the other pots. 
Pot efficiency changed with dif- 
ferent measurements of effort. Lair 
pots, because of their long soak 
times, may be more useful in a fish- 
ery where operators are fishing for 
giant octopus and another species 
concurrently. Anecdotal evidence 
from the use of lair pots indicates 
ie 
14 
12 
10 • 
8 - 
6 - 
4 - 
2 - 
0 
— 5 — 
' 
_ 
Hh 
iftsa 
Lair Korean hair crab Sablefish 
(n=107) (n=35) (n=102) 
Pot type 
Shrimp 
(n=6) 
Figure 3 
Mean weight (±standard error [SE]) in kilograms and numbers (in paren- 
theses) of North Pacific giant octopus ( Enteroctopus dofleini ) caught in each 
of the four pot types, during two surveys in Kachemak Bay, Alaska, in 2006, 
evaluated for size selectivity for use in a directed North Pacific giant octopus 
fishery. The asterisk indicates a significant difference at PcO.OOl. 
