Abstract
Well-adapted, high quality reproductive material is key to the success of forest plantations. Consequently in many countries the collection and trade of forest reproductive material is regulated. Paper documents are usually the only evidence for the origin of forest reproductive material. Certification schemes already established in Germany use genetic inventories to compare reference samples collected at different steps of the chain-ofcustody. A new approach using DNA-fingerprints efficiently controls the origin of seed sources without these multiple reference samples. Only a sample of adult trees within the seed stands is needed. The control is directly made for each suspicious plant or a group of suspicious plants by use of multilocus genotype assignment. We made a field test with samples of adults and seedling from 5 registered seed stands of Quercus robur in Western Germany. Eight highly variable nuclear microsatellites were used to genotype each individual. We found in total 255 different alleles at all loci in the adult populations. The observed levels of genetic variation (Ae= 9.18), genetic differentiation (delta = 0.187) and population fixation (FST = 0.01) were slightly higher than results of similar studies. Individual and group assignment tests were performed with the Bayesian multi-locus approach. The proportion of correctly assigned seedlings was 65% for individuals with completely scored genotypes. In all 5 cases the groups of seedlings were assigned to the correct seed stand and an additional sample of seedlings from another stand could be successfully excluded with a probability test. The conclusion of the field study is that a large scale application of this new approach to control of the origin of forest reproductive material is feasible.