Gene marker analysis concerning hybrid formation and natural regeneration of Sorbus latifolia agg. in Thuringia

Abstract

In Thuringia 13 small stands of Sorbus latifolia s.l. from different locations and two reference groups of Sorbus aria and Sorbus torminalis with in total 94 samples were analyzed (Table 1), to check the hybrid status of S. latifolia s.l. individuals with genetic markers. In one of the S. latifolia s.l. stands all (19) adult trees, 10 young plants and 21 seeds were harvested from the adult trees and investigated to check the apomixes hypothesis and the dissemination of genes during natural regeneration. Nuclear DNA (biparental inherited) and chloroplast DNA (maternal inherited) was analyzed using microsatellites (Table 2). With the exception of one sample, genotypes gave no evidence for polyploidy. Genetic differentiation (Phi(PT)) is high with 65% and 37% for cpDNA and nuclear DNA respectively. Principal component analysis (Fig. 1) revealed at least two groups which are mainly differentiated concerning their cpDNA haplotypes. S. aria with haplotype S9 and S. torminalis with haplotypes S14 and S22 are completely differentiated. Some groups of S. latifolia s.l. show either haplotype S9 or S14, in other groups mixtures of these haplotypes occur sometimes with additional haplotypes (Tab. 3). Analysis of nuclear DNA corroborates the hybrid status of the S. latifolia hybrids (Fig. 2 and Fig. 4). At MSS16 S. aria and S. torminalis are completely differentiated while S. aria shows only three alleles of fragment size 142, 144 and 146 (bp), longer fragments occur in S. torminalis only. With two exceptions, all potential hybrids show exact hybrid genotypes with one allele of S. aria and S. torminalis each. In sum, these results indicate that the analyzed S. latifolia s.l. individuals are primary hybrids. The comparison of adult trees, seeds and natural regeneration in a single stand shows one and the same genotype for all adult trees and all plants of the natural regeneration. Low seed quality (i.e. viability of <20%) indicate effects of incompatibility due to clonal inbreeding. The seeds show high genetic variation (Tab. 4). Thus, for most of the analyzed seeds (81%) the apomixis hypothesis has been rejected. At least 57% of pollen gene flow originates from external sources. Despite the fact that genetically variable seeds were produced over years, it seems that genetic structures of adult trees and their natural regeneration are the result of clonal propagation only.