46 



Fishery Bulletin 93(1), 1995 



weights (F 554 =0.02, P>0.99) of the crabs did not dif- 

 fer significantly between treatments. Change in ju- 

 venile crab body weight was estimated from initial 

 and final measurements (32 d). 



Female crabs were randomly placed in 20 groups 

 (including controls) in a complete 4 (temperature) 

 by 5 (length of exposure) design, with 7 to 8 crabs 

 per group. Treatment temperatures ranged from 

 -3.1 to -20.3°C and exposure duration ranged from 

 (controls) to 32 minutes (Table 1). Two additional 

 groups were tested at 5°C for 8 and 32 minutes (Table 

 1). The crabs did not differ significantly in length 

 (F 21U9 =1.13, P=0.324) or weight (P. 21 149 =1.36, 

 P=0.149) between treatments. Exposure took place 

 16 and 17 February (about six days after capture). 

 Observation continued through 22 June. 



Mortality and limb autotomy were monitored daily. 

 Crabs were judged dead when scaphognathite move- 



ment stopped. Generally, dead crabs were rechecked 

 the following day before they were removed from test 

 tanks. The number of legs missing on each crab was 

 counted and autotomized legs were removed from the 

 tanks. 



Righting response (the time it took a crab to right 

 itself when placed on its back underwater), which 

 we considered to be a measure of vigor, was timed to 

 the nearest 0.01 second immediately after aerial ex- 

 posure and 1, 2, 4, 8, 16, 24, and 32 days thereafter. 

 Crabs that could not right themselves after 2 min- 

 utes were recorded as "not righting" and were placed 

 upright in the tank. 



A subset of 40 female crabs randomly selected from 

 the entire exposure range was used for reproductive 

 observations. The crabs were isolated 32 days after 

 exposure in covered 70-L tanks that overflowed into 

 19-L buckets containing conical 363-u mesh nets de- 

 signed to trap zoeae. Flow rates were approximately 

 1.5 L/minute; 95% turnover time was 2.3 hours and 

 water temperatures ranged from 5.2 to 5.9°C during 

 this period (23 March-11 May). 



Feeding rates were measured before and after the 

 zoeal hatch while the 40 ovigerous females were in- 

 dividually isolated. Mussels, Mytilus trossulus, were 

 fed ad libitum to crabs during each feeding period. 

 Live mussels were cut in half and drained tissue- 

 side down on paper towels for five minutes, weighed, 

 then placed in the tanks. Twenty-four hours later 

 the remaining food was removed, drained, and 

 weighed as before. At each feeding, four food portions 

 were placed as controls in tanks without crabs. Con- 

 sumption was corrected for the mean weight changes 

 in the control portions. Feeding observations were 

 repeated every 1 to 3 days, from 41 to 60 and from 

 85 to 98 days after exposure. 



Zoeae were collected daily, rinsed from the nets, 

 concentrated in a known volume, and subsampled 

 with a 5- or 10-mL Hensen-Stemple pipette (Carls 

 and O'Clair, 1990). Subsamples, which contained a 

 minimum of 200 zoeae, were preserved in 5% forma- 

 lin and counted later; the occasional large subsample 

 was divided with a Folsom plankton splitter before 

 being counted. After zoeal hatching, all debris from 

 each tank bottom was preserved to determine the 

 number of dead eggs and zoeae. 



Responses of the crabs were related to aerial ex- 

 posure, expressed as the product of air temperature 

 (°C) and length of time in air (hours), i.e. degree-hours 

 (°h). In a similar experiment, Carls and O'Clair ( 1990) 

 demonstrated the usefulness of this technique for 

 interpreting responses to aerial exposure in adult 

 king crabs, Paralithodes camtschaticus. Because the 

 responses of the Tanner crabs to exposure (in °h) were 

 similar in form to those of the king crabs over identi- 



