758 



Fishery Bulletin 88(4), 1990 



consistent with the Hmited information from a tagging 

 study in the Gulf of Alaska. McBride (1982) reported 

 the mean growth increment of six tagged snow crabs 

 at liberty for up to 1 year was 14.7 mm. McBride's 

 animals were somewhat larger than those in our study 

 (x 113.8 mm, range 108-124 mm CW at tagging) and 

 would, on the basis of our regression, be predicted to 

 have slightly larger molt increments. 



Laboratory studies of snow crab growth are also con- 

 sistent with our findings. O'Halloran (1985) reported 

 an average molt increment of 1 1 .6 mm for eight ani- 

 mals. However, five of the eight animals had their 

 eyestalks ablated bilaterally and these animals died dur- 

 ing ecdysis. Consequently, these estimates are not very 

 reliable. Miller and Watson (1976) and R.J. Miller (Dep. 

 Fish. Oceans, Halifax, Nova Scotia, Canada 83J 2S7, 

 pers. commun., June 1989) observed a mean molt in- 

 crement of 15.5 mm for 18 crabs ranging in pre-molt 

 size from 59.3 to 101.1 mm CW (.r 80.5 mm). G. Hur- 

 ley (Hurley Fish. Consulting, 45 Alderney Dr., Dart- 

 mouth, Nova Scotia, Canada B2Y 2N6, pers. commun., 

 June 1989) observed a mean increment of 13.9 mm for 

 47 crabs, size range 60.7-83.4 mm CW (.7 66.1 mm). 

 In these studies, crabs were fed ad libitum. In another 

 lal)oratory study, Moriyasu et al. (1987) reported a 

 regression line that indicates that animals in the size 

 range we studied (82-113 mm CW) have predicted molt 

 increments of 14-16 mm CW. 



Molt increments have also been estimated l)y size- 

 frequency analysis for Japanese populations of snow 

 crabs. Kon et al. (1968) reported estimates for small 

 individuals. Ito (1970) suggestetl that 81 -mm CW in- 

 dividuals molt to 97 mm, then to 111 mm and 121 mm 

 (i.e., molt increments of 16, 14, and 10 mm). Kon (1980) 

 suggested that 80-mm CW individuals molt to 93.4 mm, 

 then to 105.6 and 1 16.7 mm (molt increments of 13.4, 

 12.2, and 11.1 mm). These estimates are similar to the 

 ones we derived from the tagging data. Robichaud et 

 al. (1989) analyzed size-frequency distributions of snow 

 crabs from the Gulf of St. Lawrence. Their samples 

 consisted of small animals, so their results are not com- 

 pai-able to the results presented here. 



Effect of possible terminal molt 



Some workers believe that male snow crabs undergo 

 a "terminal" or final molt which is associated with a 

 change in allometry. Although this idea is controver- 

 sial (see Jamieson and McKone 1988 for reviews), it 

 is worth examining whether the two groups evident in 

 Figure 4 might reasonably be interpreted as those 

 molting to the terminal state (e.g., lower group) and 

 those molting but not to the terminal state (upper 

 group). To accept this hypothesis implies that all of the 

 following are accepted: 



(1) males do, in fact, undergo a terminal molt; 



(2) the size increment at the terminal molt is con- 

 siderably different (presumably smaller) than the pen- 

 ultimate molt increment; 



(3) it is a coincidence that the animals in the upper 

 group were at liberty twice as long as those in the lower 

 group (828 vs. 468 days on average), and that the 

 animals which did not molt were at liberty for less than 

 a year on average {x 322 days); 



(4) it is a coincidence that the size predicted after 

 two molts by the regression fitted to the lower cloud 

 of points is in close agreement to the size after two 

 molts predicted by the regression fitted to the upper 

 cloud of points; 



(5) there must be an as yet unidentified reason why 

 a third cloud of points, corresponding to animals which 

 molted twice, is not evident in Figure 4. 



Even if males undergo a terminal molt, there is no 

 evidence to suggest that the final molt increment is 

 distinctly smaller than the penultimate molt increment. 

 In an aquarium study, Moriyasu et al. (1987) found, for 

 male snow crabs in the size range 60-70 mm CW, that 

 animals molting to the large-clawed state have molt 

 increments about 3 mm smaller than those molting to 

 a small-clawed state. This small difference in molt in- 

 crements is not sufficient to account for the two clouds 

 we identified in the field observations. Ennis et al. 

 (1988) have shown that males can become functionally 

 mature before attaining the morphometry associated 

 with the terminal molt. Hence, the terminal molt is not 

 necessarily associated with a diversion of energy from 

 growth to reproductive processes, and the terminal 

 molt increment need not be small. We conclude that 

 if animals undergo a terminal molt, the molt increment 

 at this time is similar to the penultimate molt inci'e- 

 ment. We do not rule out the possibility that some of 

 the scatter about the regressions through the two 

 clouds of points is due to mixing terminal and non- 

 terminal molt data. 



In summary, our interpretation of the tagging data 

 is supported by two lines of evidence. Animals pre- 

 sumed to have molted twice were at liberty twice as 

 long on average as animals presumed to have molted 

 once. Also, the predicted size after two molts, as esti- 

 mated from data on animals presumed to have molted 

 once, agrees closely with the predicted size estimated 

 from data on animals presumed to have molted twice. 

 We conclude that male snow crabs in the size range 

 75-115 mm have molt increments of around 11 mm. 

 Our results are consistent with the limited information 

 ai)OUt snow crab growth in the literature. 



