Allozyme variation in Pinus cembra and P. sibirica: differentiation between populations and species

Two closely related Eurasian species of 5-needle pines, Swiss stone pine(Pinus cembra L.) and Siberian stone pine (P. sibirica Du Tour) occupy two disjunctive parts of the formerly common range in Europe andSiberia, respectively. These forms show so close morphological and genetic similarity that in some classifications they are treated as subspecies. Using a set of 29 allozyme loci (Adh-1,-2, Fdh, Fest-2, Gdh, Got-1,-2,-3, Idh, Lap- 2,-3, Mdh-1,-2,-3,-4, Mnr-1, Pepca, 6-Pgd-1,-2,-3, Pgi-1,-2, Pgm-1,-2, Skdh-1,-2, Sod-2,-3,- 4) we analyzed genetic differentiation within the Alpine-Carpathian part of the range (P.cembra) and found relatively low genetic diversity for conifers (HE=0,08) and moderate level of differentiation (FST=7,4%). For thesame loci set within Siberian populations (P. sibirica) genic diversity was higher (0,14), while differentiation was lower (3%). The factthat differentiation within the highly fragmented range of Swiss stone pine is just 2.5 times higher than in widespread closely related P.sibirica makes us consider factors other than unlimited gene flow responsible for uniformity of allelic frequencies. Among these factors the leading role belongs to balancing selection.Heterozygote superiority leads to both (i) increasing of heterozygosity in course of stand development and (ii) through balancing selectionto stable equilibrium state. Under this equilibrium, virtually the same genetic structure is maintained, even when remote and isolatedparts of the species' ranges are compared. For many studied loci, Swiss and Siberian stone pines have the same allelic profiles despite the fact that gene flow among them ceased a long time ago. According to one point of view fragmentation of the formerly united range might take place in the Atlantic time of the Holocene (about 5000 years BC), however, an alternative hypothesis refers to a much earlierPleistocene glacial time. Analysis of a combined data set (P. cembra + P. sibirica) including samples of both species showed that a subdivision of genetic variation is about 10 times higher than within each species. A problem of distinction between selective and nonselectivedifferentiation for allozymes and other genetic markers is discussed.