Taylor and Hoenig: Growth per molt of Chionoecetes opitio from Newfoundland 



757 



Figure 5 



Recapture data for Chionoecetis tipilm from Conception Bay. Solid 

 symbols represent animals presumed to have molted once: open sym- 

 bols represent animals which molted twice. Solid lines show the 95% 

 confidence band obtained from the linear regi-ession fitted to the open 

 circle data. Dashed lines give an approximate 95% confidence band 

 for the size after two molts based on the regression for the presumed 

 single-molters (solid symbols). (See Appendix A for method of con- 

 structing the confidence band.) The dashed lines enclose the solid 

 lines, indicating that the two predictions of size after two molts are 

 not significantly different. This supports the idea that the two clouds 

 in Figure 3 re|;)resent single- and doulile-molters. 



Then the growth parameters a and h can be estimated 

 by regressing Y on X and Z in the regression model 



Y 



abZ + hX + b{h - 1)ZX 



-t- e 



(3) 



where e is a random error term. Equation (3) reduces 

 to (1) when Z = and to (2) when Z = 1. The model is 

 generalized in Appendix B to account for a quadratic 

 relationship between post- and pre-molt size. The pre- 

 dicted sizes from the fitted equation (3) (Table 2) are 

 very similar to the results obtained using equations (1) 

 and (2) separately. 



Table 2 



Estimates of the parameters a and b in the nonlinear regres- 

 sion model relating size at recapture (Y) of Conception Bay 

 Chionoecetes opilio to size at tagging (X) for animals molting 

 once or twice. The model is: Y = a -t- abZ + bX + b{b - 1)ZX 

 + e. where e is the error term and Z = if the animal molted 

 once, and 1 if the animal molted twice. 



Parameter Estimate Standard error 



7.398 

 1.038 



2.074 

 0.021 



correlation 

 -0.998 



Figure 6 



Plot of size at recapture vs. size at tagging for recaptures of Chio- 

 noecetes opilio from Conception Bay (circles) and Bonavista Bay 

 (triangles and stars). Solid circles and stars represent animals be- 

 lieved to have molted once: open triangles and circles represent 

 animals which molted twice. Diamond represents an animal from 

 Bonavista Bay at liberty for ~1000 days which was presumed to have 

 molted three or more times. Large circle represents an animal which 

 normally would be assumed to have molted twice: however, it was 

 reportedly at liberty for only 29 days and consequently was excluded 

 from analysis. Regression lines were determined using equation (3) 

 from all data on single and double molters from both bays. 



Bonavista Bay 



Molt information was obtained from 18 animals re- 

 captured in Bonavista Bay. Of these, four animals ap- 

 peared to have molted once; 13 animals appeared to 

 have molted twice; and one animal at liberty for ap- 

 proximately 1000 days appeared to have molted three 

 or more times (Fig. (3). This interpretation is supported 

 by the fact that the animals presumed to have molted 

 once were at liberty for an average of 198 days, while 

 the animals presumed to have molted twice were at 

 liberty for an average of 698 days. 



The molt increments appeared very similar to those 

 from Conception Bay. Since there is not much infor- 



mation from Bonavista Bay, we computed a regression 

 to predict molt increments for Bonavista Bay crabs 

 using the combined recapture data on single and dou- 

 ble molters from Bonavista and Conception Bays. The 

 predicted size after molting is given by 



Predicted size = 9.21 + 1.02 (pre-molt size). 



Other populations 



The predicted molt increments from equation (3) for 

 Conception Bay crabs (Table 2) were very close to a 

 constant (10.5 - 1 1.7 mm) over the size range of ani- 

 mals we studied (82-113 mm CW). These estimates are 



