3)4 



Fishery Bulletin 101(2) 



B 



1999/2000 WDS 

 (71=2060) 



20 40 60 80 100 120 140 160 180 200 



20 40 60 80 100 120 140 160 180 200 



Sub- sample of 

 998/ 1999 WDS 

 (/;=769) 



D 



Sub-sample of 

 1999/2000 WDS 



(/!=367) 



^ 



¥^ 



20 40 60 80 100 120 140 160 180 200 



20 40 60 80 100 120 140 160 180 200 



Carapace width (mm) 



Figure 1 



Size-frequency distributions for blue crabs collected in 1998-99 (A) and 1999-2000 (B ) winter dredge surveys and size-frequency 

 distributions (C and D) of subsamples used in this study. Hatched bars in C and D indicate samples excluded from lipofuscin 

 analysis. 



Results 



Size frequencies of blue crabs collected in the winter dredge 

 survey differed significantly between the two sampling 

 years (chi-square test; a=0.05; Fig. 1, A and B). Although 

 the total number of crabs collected from 1998-99 and 

 1999-2000 winter surveys were equivalent, a greater 

 proportion of crabs >80 mm CW occurred in the second 

 year. For both years, frequency distributions of CW data 

 supported the fit of two modes (Fig. 1) and juveniles were 

 clearly distinguishable from adult crabs as individuals 

 <80 mm CW (or 90 mm CW for 1998-99). For 1998-99 

 WDS, males and females subsets of crabs had mean CW 

 of 56.4 ±2.0 mm (mean ±SD; n = 165) and 52.2 ±2.0 mm 

 (n = 152), respectively. For 1999-2000 WDS, CW for males 

 and females was 91.5 ±2.7 mm (n = 162) and 72.0 ±3.0 mm 

 (n=97), respectively. The number of crabs sampled for 

 lipofuscin analysis was halved (from n=769 to 367) in the 

 second year because of a much more restricted analysis of 

 juveniles <80 mm CW that were assumed to be <1 year of 

 age. In both years, the subsamples examined for lipofuscin 

 were skewed towards crabs >60 mm CW iFig. 1, C and D), 

 reflecting our analytical criteria and objective to include all 

 adults in the analysis of age structure. 



Lipofuscin index varied positively with CW (Table 1, 

 Fig. 2) but there was high variability in lipofuscin index 

 for a given size. All subcategories (sex and yearl supported 



significant regressions of lipofuscin on CW, but coefficients 

 of determination were quite low, ranging as low as 0.04 

 for females in the first sampling year. ANCOVA analysis 

 showed that lipofuscin adjusted for CW effects was signifi- 

 cantly different only between sexes for the first sampling 

 year (Table 2 ). We had expected that the lipofuscin index in 

 mature females (CW >120 mm) would be higher at a given 

 CW than in males because females discontinue molting 

 following their first mating (Millikin and Williams, 1984). 

 As a result, females living beyond their final molt would 

 be predicted to accumulate lipofuscin independently of size. 

 Results were scattered, however, and there was no statisti- 

 cal evidence for higher lipofuscin index in mature females. 

 The lack of compelling evidence for differences in overall 

 lipofuscin levels among sexes allowed males and females to 

 be combined in the subsequent age structure analysis. 



Frequency distributions for the lipofuscin index were fit- 

 ted by multiple modes in both years (Fig. 3, A and B). Inter- 

 vals between modal means were greater than two standard 

 deviations for the first three modes in each sample (Table 3 ) 

 and lipofuscin index-based modal distributions were not 

 significantly different from expected normal distributions 

 (chi-square test; a=0.05). Modes were consistent with the 

 presence of 0, 1, and 2 age classes in each collection. Ages 

 derived from lipofuscin index values (Eq. 1) for the first 

 mode in each year were higher (ages 0.8-1.0 yr) than those 

 expected for juveniles produced during the previous spawn- 



