regression technique described by Stevens (1951), a 

 curve of the form y = A + BR ' was fitted* to the data. In 

 this model, y is the average arcsine and x is the duration 

 of time in days. The asymptote (A) of the curve, when re- 

 transformed to the original scale, is the estimated value 

 to which the cumulative percentage escapement tends. 

 Estimated values (in the arcsine scale) of the param- 

 eters follow: 



Parameters 



Size 



B 



Standard 

 deviation 



of A 



Legal-size crab 



(120-139 mm) -42.098 0.8421 63.715 5.451 

 Undersize crab 



(150-169 mm) -51.711 0.7604 73.758 2.278 



mated value was 92.2%, with lower and upper confl- 



dence limits of 87.4% and 95.9%. 



Effect of Bait or Dead Crab 



Bait in the form of chopped herring or dead crab did 

 not "hold" tagged king crab in the test pots. Crab left 

 baited pots more rapidly than unbaited pots (Fig. 4). The 

 occurrence of new untagged king crab at time of retrieval 

 in the test pwts baited with herring was many times 

 greater than pots containing dead crab or no bait (Fig. 5). 

 After 7 days, the new entries in herring-baited pots ap- 

 proached the rates for the unbaited and dead-crab- 

 baited pots, indicating that the herring became less 

 effective as bait with time. Dead crab in test pots did not 

 attract live king crab to the pots. 



Using these values and retransforming to the original 

 scale, we estimated the cumulative percentage escape- 

 ment with time (Fig. 3). Also shown in Figure 3 are 

 observed data points and a 95% confidence interval for 

 the asymptote. For legal-size crab, the estimated asymp- 

 tote was 80.4% with lower and upper 95% confidence 

 limits of 63.5% and 93.0%. For undersize crab the esti- 



^The asymptotic regression analysis was completed using a computer 

 program written by George Hirschhom of the Northwest and Alaska Fish- 

 eries Center, National Marine Fisheries Service, NOAA, Seattle, WA 

 98112. 



Mortality in Pots 



Crab remains found in each pot were examined for tag 

 numbers or scars. No unmarked crab were found dead. It 

 is imlikeiy that any dead crab were unaccounted for as 

 the carapaces remained intact even during a 15-day soak, 

 and the pxit mesh-size was small enough to prevent shells 

 from falling out during the lift. Mortality did not in- 

 crease with time of soak (Table 4). However, the 15-day 

 soak was a relatively short period to observe mortality 

 caused by confinement. Natural mortality, tagging in- 

 jury, and handling probably all contributed to the 



LEGAL CBABS 



Figure 3. — Average escapement retransformed to 

 percent for small and large king crab from standard 

 pots (data from 1974 and 1975). 



e e 10 12 



DAYS SOAKED 



e 8 10 12 

 DAYS SOAKED 



Figure 4. — King crab escapement f^om baited and 

 unbaited pots after various soak Intervals. 



DAYS SOAKED 



