FISHERY BULLETIN: VOL. 86, NO. 3 



at any of the stations, and Lough suggested that a 

 mass mortahty of C. magister larvae might have oc- 

 curred. In contrast, C. oregonensis larvae were in 

 greater abundance during 1971 than 1970. 



Results from the current study and Booth et al. 

 (1985) suggest that the sampling gear and proce- 

 dures used by Reilly (1983) and Lough (1976) gave 

 poor abundance estimates of megalopae. The num- 

 ber of megalopae caught during daylight is relatively 

 low, and the presence or absence of only a few 

 megalopae can greatly influence the estimated 

 megalopal density per unit of sea surface. This sen- 

 sitivity to bias can only be compensated for by many 

 replicate samplings, and it is logistically easier to 

 estimate absolute megalopal abundance by sampling 

 at night when megalopae are concentrated in a 

 relatively easily sampled, narrow depth range, 

 namely the neuston (Booth et al. 1985). Reilly (1983) 

 sampled once both day and night at 12 stations, and 

 noted no significant differences in megalopal den- 

 sity between day and night at the surface and in 

 oblique tows. However, his surface tows were not 

 neuston tows, and consisted of a 0.5 m diameter net 

 towed about 0.25-0.5 m below the water surface, 

 or below the wave troughs if the surface was rough 

 (P. Reilly fn. 5). This qualifies his observations, and 

 may explain the suggested low abundance of mega- 

 lopae in his study area (maximum density was 43 

 100 m"^ at the surface). 



On occasion, megalopae have been observed to be 

 in association with Velella velella at the surface dur- 

 ing daylight (Wickham 1979; Reilly 1983; G. Jamie- 

 son unpub. data), but we have generally observed 

 relatively few megalopae in surface waters during 

 the day in comparison to the number observed 

 present at night. Accepting that relative megalopal 

 abundance is best determined at night from neuston 

 tows, megalopae off British Columbia were shown 

 to be abundant in specific areas offshore and at these 

 locations, early and mid-stage megalopae predomi- 

 nated. 



All three studies indicate that 1) later stage C. 

 magister larvae are found offshore and, since they 

 are hatched at inshore locations, some mechanism 

 must be transporting them away from shore, 2) 

 megalopae are found inshore, with Hatfield (1983) 

 and this study showing that earlier stage megalopae 

 predominate furthest offshore, and 3) while long- 

 shore drift of larvae may occur, it is difficult to 

 establish its significance in the absence of simul- 

 taneous current and larval distribution data, both 

 geographically and vertically in the water column. 

 None of these studies demonstrates conclusively 

 that larvae which ultimately settle inshore were ever 



very far offshore, and while it is shown that larvae 

 can be carried offshore in abundance, this may well 

 represent larval wastage, in that relatively few of 

 them, if any, may ultimately return inshore to sur- 

 vive at settlement. The same is assumed to apply 

 for C. oregonensis. 



The California data on larval distribution, which 

 is the most extensive spatially, are somewhat am- 

 bigious in that they may be interpreted either as 

 onshore-offshore-onshore movement, or as a general 

 offshore dispersal coupled with northwards-south- 

 wards drift of those larvae remaining inshore. To 

 account for an absence of middle stage larvae in the 

 nearshore waters of Reilly's (1983) study area, the 

 latter scenario assumes that larval crab production 

 south of the Gulf of the Farallones is relatively low, 

 and that the water mass entering the sample area 

 from the south as part of the Davidson Current con- 

 tains few larvae. Little data appear to exist to refute 

 this possibility; the commercial fishery for Dunge- 

 ness crab only exists at a few locations south of San 

 Francisco, and then only sporadically (Warner 

 1985). Our data and Lough's (1976) data are inade- 

 quate to investigate longshore drift because bio- 

 logical samples in each study were only collected 

 along one offshore transect and no data on currents 

 was simultaneously obtained. 



Off the west coast of Vancouver Island, we found 

 that megalopae may be concentrated in areas 

 between surface currents flowing in opposite direc- 

 tions, specifically in the shear between the Van- 

 couver Island Coastal Current and the outer Shelf- 

 Break Current. Megalopae were abundant 40-70 km 

 from shore, with C. magister megalopae particularly 

 concentrated about 50 km from shore. In 1985, there 

 was no major crab settlement observed on the outer 

 coast near Tofino, suggesting that the Coastal Cur- 

 rent may sometimes be an effective barrier to on- 

 shore movement and successful settlement. Under- 

 standing the horizontal and vertical distributions of 

 crab larvae in the water column now seems essen- 

 tial if transport mechanisms of larvae and their 

 movement from hatching to settlement are to be 

 understood. 



ACKNOWLEDGMENTS 



We wish to thank the officers and crew of the 

 Canadian Research Vessel G. B. Reed for their en- 

 thusiastic assistance in collecting data; D. Hartley, 

 skipper of the charter vessel Beatrice, for his skill 

 and persistence in meeting the objectives of the 

 charter; and the students and term employees who 

 participated in collecting and analyzing data. Glen 



540 



