GENE FLOW PATTERNS OF THE EUPHAUSIID, MEGANYCTIPHANES-NORVEGICA, IN THE NW ATLANTIC BASED ON MTDNA SEQUENCES FOR CYTOCHROME-B AND CYTOCHROME-OXIDASE-I

Population genetic structure and patterns of gene flow are described f or the euphausiid, Meganyctiphanes norvegica, in the NW Atlantic Ocean based on DNA sequence variation of two regions of mitochondrial DNA ( mtDNA). DNA sequences were determined for portions of cytochrome oxida se I (COI; 400 base pairs; 76 individuals) and cytochrome b (CYB; 300 base pairs; 101 individuals) for euphausiids collected from the Gulf o f Maine in 1991, the Gulf of St Lawrence in 1993 and 1994, the Scotian Shelf in 1994, and Georges Bank in 1994. For comparison, M.norvegica collected from the fjords of western Norway in 1992 were sequenced for COI (20 individuals) and CYB (18 individuals). COI was less variable than CYB, based on both haplotype (h = 0.6847 for COI; 0.9077 for CYB) and nucleotide diversities (pi = 0.0038 for COI; 0.0182 for CYB). Hap lotype frequencies among all samples of M.norvegica were significantly heterogeneous for CYB (P < 0.001 by chi(2)), COI/CYB (P < 0.034) and for the 37 sites in the COI/CYB sequence at which there were multiple substitutions (MSUB; P < 0.0001). Haplotype frequencies among regional populations (i.e. pooled samples collected within each region) in the NW Atlantic were different for MSUB haplotypes (P < 0.021), but not f or any other gene portion. Comparison of NW Atlantic versus Norwegian fjord populations (based on pooled samples) revealed significant diffe rences in haplotype frequencies for CYB (P < 0.004) and MSUB (P < 0.00 01), and by analysis of molecular variance (AMOVA) for CYB (P < 0.01) and COI (P < 0.01). All evidence considered together indicated that: ( i) genetic heterogeneity among samples reflected high levels of molecu lar genetic diversity and, perhaps, undersampling of the population ge netic structure; (ii) gene flow of M.norvegica within the NW Atlantic was sufficient to prevent the formation of distinctive geographical po pulations, except in the case of the Gulf of St Lawrence sample collec ted in 1994; (iii) genetic differentiation of NW Atlantic and Norwegia n fjord populations indicated highly restricted gene how across the N Atlantic Ocean. We conclude that mtDNA genes may vary significantly in the amount and pattern of variation, and that this variation affects the results and conclusions from molecular population genetic analyses . Consideration of genetic traits with small numbers of moderately fre quent variants, such as the MSUB haplotypes, markedly increased the po wer of statistical analyses. Resolution of the population genetic stru cture of marine zooplankton at spatial scales smaller than ocean basin s will require the analysis of numerous individuals (i.e. thousands), samples collected at appropriate temporal and spatial resolutions, and molecular markers with appropriate levels and patterns of variation.