Figure 1.Mitochondrial DNA haplotype frequencies in lake sturgeon from the nothernpartof their range See Table 1 for sample sizes.
Haplotype 1=black. haplotype 2=clear.much migration isrequired toerodemtDNAhaplo- HincII)or nucleotidevariation atposition 54 in the
type divergence than nuclear gene divergence. 275 base pairssequenced(details inFerguson etal.
Thus,the basicpremise isthatfishoriginatingfrom 1993).Both analyses provided equivalent haplotyp-
differentrefugiawilldiffer genetically inmtDNA icdesignations.Threepopulationsfrom the Hud-
haplotypefrequencies(Billington&Hebert1991). son-James Baywatershed innorthernManitoba.
Ward et al.(1989)found thatmtDNAhaplotypesof Ontario, and Quebec shared the same twocommon
walley, Stizostedionvitreum, cluster intothree haplotypes. However, the vast majority ofmore
main groups, each reflecting recolonization ofCan- southerlyfishfrom the Great Lakes/St.Lawrence
adianwatersfrom each of three glacial refugia.A and Mississippi watershedswereofa single haplo-
similar approach has been used for lake whitefish, type.This included three specimens housed atthe
Coregonus clupeaformis (Bernatchez & Dodson British MuseumofNatural History and collected in
1990). lake charrSalvelinusnamaycusharctic charr thelate1800’s.The only apparent disparityinthis
S.alpinus(Wilson1995),and brook charrS.fontina- patternwasthe collectionofa very small numberof
lis(Danzmann&Ihssen1995). haplotype 2 fish near the confluence of the Ottawa
We analyzed mtDNA variationoflake sturgeon andSt.Lawrence Rivers by Guenetteetal. (1993).
fromthe northernpartoftheirrange toassess fac- Thedistributionof mtDNA haplotypes (Figure
tors which may influence the distribution of 1) can explained by two alternative scenarios
mtDNA haplotypelineages(Ferguson et al.1993, corresponding toasingle refugium hypothesis ver-
Guenette et al. 1993).Most of the 396 lake sturgeon susatworelugiumhypothesis. The resolutionof
analyzed were characterized byonlytwomtDNA the controversy dependsonthe ability of eachsce-
haplotypes based on a restrictionfragmentlength nariotoexplainthecommonalityof haplotype 2
polymorphism (RFLP)analysis ordirectsequene- fish in northern samples and rarity in southern sam-
ing of 275 nucleotidesinthemtDNAcontrolregion ples.Inthe firstscenario. thedisjunctdistribution of
(Table1.Figure1).Fish could becategorizedinto haplotype 2 isbecauselakesturgeonusedtwo dis-
two majorhaplotypesusingeitheroftwo restriction tinctroutesofcolonizationfrom tworefugia (pos-
enzymeswhichdetectedpolymorphisms(AvaII, siblyMissourianandMississippian)toreinvade the