Dispersal of seed and effective pollen in small stands of European beech (Fagus sylvatica L.)

Two isolated small beech stands were used to study gene flow. Seeds were sampled on the ground under each adult tree. Ten polymorphic enzyme coding gene loci were utilized for multilocus-genotyping trees and seeds. The modified most-likely method was used to assess effective pollen dispersal by inferring paternity of offspring. The difference in LOD scores between the most likely candidate parent and the next-most likely candidate parent (i.e. the statistic ∆ according to Marshall et al. 1998) was used to evaluate confidence in parentage of the most likely parent. The most likely parent pair was inferred and seed dispersal was estimated. Preliminary results showed that the migration distance of seeds within the isolated beech stand 10A2 (70 trees on 0.5 ha) is limited to 50m and 75% of the seeds was dispersed within 20m. Most pollen originated from within 50m of the seed tree, although in stand 34B1 (24 trees on 0.2 ha) a low proportion of pollen movement was from as far as 300m. In the stand 10A2, 6.1% of the seeds were fertilized by pollen parents occurring at least 140m outside the stand. Two of the trees in stand 10A2 had unique allozyme alleles (LAP-A1 and IDH-A4, respectively), which were used to directly measure pollen movement away from those trees. The frequency of the unique LAP and IDH alleles in seeds declines as the distance from the source tree increases. The mathematical models of pollen and seed dispersal were fitted and the probability distributions of pollen and seed dispersal were derived.