294 



Abstract.— We examined the relation- 

 ship between otolith size and larval 

 standard length (SLl at hatching for 

 Atlantic cod. Gadus morhua, on the 

 Scotian Shelf. We found a weak corre- 

 lation between SL and area-based and 

 radius-based measures of size of both 

 lapillar and sagittal otoliths. Correla- 

 tions of SL with the area of the lapillus 

 were strongest. However, the predictive 

 ability of all relationships was low. For 

 example, the range of predicted SL at 

 hatching of larvae with lapilli of aver- 

 age area included more than 90'^t of 

 observed SLs in newly hatched larvae 

 from the Scotian Shelf collected over 

 two spawning seasons. These results 

 suggest that otolith-based attempts to 

 backcalculate the size of cod larvae may 

 be prone to substantial error if sizes of 

 particularly young larvae are esti- 

 mated. We recommend that, where pos- 

 sible, stock- and season-specific esti- 

 mates of the relationship between the 

 area of the lapillus and larval size at 

 hatching be used in back-calculation 

 techniques. 



The relation between otolith size and larval size 

 at hatching for Atlantic cod, Gadus morhua* 



Thomas J. Miller 



Chesapeake Biological Laboratory 



University of Maryland Center for Environmental Science 



PO Box 38 



Solomons, Maryland 20688-0038 



E-mail address: millerg cbl umces edu 



Tomasz Herra 



William C. Leggett 



Department of Biology 



Queens University 



Kingston, Ontario, K7L 3N6 Canada 



Manuscript accepted 22 June 1998, 

 Fish. Bull. :294-305( 1999), 



The variation in year-class strength 

 in fish populations has profound 

 implications for our ability to man- 

 age stocks wisely. Recently, consid- 

 erable effort has been focused on 

 trying to understand why the few 

 individuals that survive do so, 

 rather than why the majority die 

 (Crowder et al., 1992), This ap- 

 proach, which relies inherently on 

 characterizing traits in individual 

 fish, has been widely applied in both 

 marine and freshwater studies 

 (Herman et al,, 1996; Rice et al., 

 1997 for example). The philosophy 

 behind the approach is that by un- 

 derstanding how survivors differ 

 from those that die, one may expose 

 the mechanisms that regulate re- 

 cruitment (Fritz et al,, 1990). 



We used this approach to explore 

 recruitment mechanisms in Atlan- 

 tic cod, Gadus morhua, on the 

 Scotian Shelf, In this area, cod 

 spawning is bimodal, beginning in 

 late October and continuing to the 

 following April, with peaks in De- 

 cember and March (Miller et al,, 

 1995), Recent studies have sug- 

 gested the importance of small-scale 

 physical oceanographic features, 

 such as gyres and fronts, for larval 

 survival (Taggart et al., 1996; 

 Lochmann et al,, 1997). Abundant 

 populations of copepods, particu- 

 larly F^seudocalanus and Paraca- 



lanus, co-occur with cod larvae 

 ( McLaren and Avendano, 1995 ) and 

 support rapid larval growth. By fol- 

 lowing birth-date cohorts through 

 time and repeatedly estimating the 

 distribution of phenotypic and ge- 

 notypic traits in the cohort, we 

 hoped to quantify whether survi- 

 vors were different functionally 

 from the majority that died, or 

 whether they were simply lucky 

 (Miller, 1997), 



Genetic evidence suggests that 

 cod larvae collected on the Scotian 

 Shelf originate from distinct spawn- 

 ing events (Ruzzante et al., 1996). 

 However, Ruzzante et al. (1996) 

 have shown that the genetic struc- 

 ture within the population remains 

 stable over time. Meekan and 

 Fortier (1996) repeatedly sampled 

 two autumn-spawned cohorts of 

 Scotian Shelf cod, following each for 

 approximately six months. The pat- 

 tern of survival for the two cohorts 

 differed. In 1991-92, the growth 

 rate distribution of survivors dif- 

 fered little from the growth distri- 

 bution of the cohort from which they 

 were drawn. In contrast, in 1992- 

 93, Meekan and Fortier (1996) 

 found evidence of strong selection 



* Contribution .3013 from the LIniversity of 

 Maryland Center for Environmental Sci- 

 ence. Solomons. MD. 



