378 



Fishery Bulletin 98(2) 



Abundance (ind/m-) 



Belle Isle slrait 



Geographic distribution and abundance of redfish larvae at stations sampled in June 

 1992 in the Gulf of St. Lawrence. 



merated. The number of larvae frozen on board and 

 those later sorted from the plankton samples were 

 summed to provide estimates of the total abundance 

 of redfish larvae (individuals/m^) at each station. 



Electrophoretic analyses 



All electrophoretic analyses were performed within 

 two to three months after specimen collection. The 

 plexiglass plates supporting the larvae were always 

 kept on ice during laboratory examination and analy- 

 sis. The tissue in the liver area of individual lai-va was 

 dissected under a stereomicroscope, placed directly 

 into an individual sample well, and homogenized in 

 an extraction solution containing 10 ]xh of a 0.0 IM 

 Tris-HCl (pH 8.0) composed of 307^ sucrose, 0.005M 

 dithiothreitoKDTT), 0.5% polyvinylpyrrolidone (PVP) 

 and 0.00 IM phenylmethylsulfonyi fluoride (PMSF). 

 The tissue was homogenized with a fine glass rod and 

 an aliquot was applied on the cellulose acetate gels. 

 Cellulose acetate gel electrophoresis and gel staining 

 were carried out as described by Hebert and Beaton 

 ( 1989). A specimen of genotype MDWA1A2 was used 

 as a standard on every gel to assess both the quality 

 of the electrophoretic separation and to ensure allele 



identification. In 1991, a total of 735 larvae was frozen 

 and analyzed for genotypic variation. Interpretable 

 results were obtained for 697 specimens. In 1992, 

 1041 larvae were analyzed and scored. 



Two alleles segregated at the MDH* locus, resul- 

 ting in the presence of three phenotypes ( Rubec et al. 

 1991 and references therein). Homozygotes for the 

 slow allele {MDH-A2) were assigned to S. fasciatus, 

 and the homozygotes for the fast allele iMDH^Al) 

 were classified as S. mentella. Heterozygotes at the 

 MDH* locus were left unclassified (see "Results" and 

 "Discussion" sections). 



Statistical analyses 



The adequacy of genotypic proportions to Hardy- 

 Weinberg expectations was tested for each sampling 

 station where genotypic variation was observed by 

 using the G-test of goodness-of-fit (Tables 1 and 2). 



Standard length data were either not homoscedas- 

 tic or not normally distributed. Therefore, compa- 

 risons of larval standard length between 1991 and 

 1992, between the different sampled areas of the 

 Gulf, were carried out with a Mann Whitney U-test. 

 Heterogeneity in the distribution of standard lengths 



