Increased interest in native woodlands has exposed gaps in our knowledge of our native species. Aspen (Populus tremula L.), one such species, is of considerable conservation value though has been subject to very little previous research. It is widely distributed in Scotland at low density occupying a marginalised niche. Aspen is dioecious but reproduces mainly by asexual means and recruitment is reputed to occur only very rarely. Low population density , a marginal niche and an inability to colonise new sites renders the Scottish aspen resource potentially vulnerable to genetic erosion from natural and anthropogenic disturbance potentially compromising its long term future. Intervention is clearly required. Successful conservation management requires a thorough understanding of genetic structure at the national, regional and stand scale. The genetic structure of the Scottish native aspen resource was therefore investigated at the national, regional and stand scales using selectively neutral molecular markers. Isozyme analysis was used to estimate genetic variation and investigate population structure and genetic differentiation in six Scottish aspen populations. Eight systems, yielding eleven putative loci, were employed in screening 275 aspen samples. Genetic variation was shown to be greater than mean values obtained for species sharing similar characteristics and comparable with the homologous species P. tremuloides (Pp=54.5, Hep=0.174, Ap=2.00. Aes=0.121) suggesting that no significant loss of genetic variation has occurred. Some inbreeding was detected (f=0.153) and attributed to assortive mating between a small subset of clones retaining sexual function. A small but significant amount of structuring was detected (q 0.014) although UPMGA clustering of genetic identity (I) indicated little differentiation amongst populations providing little support of r a multiple origin hypothesis. It is concluded that the genetic structure in the present Scottish native aspen resource (SNAR) is little changed from the genetic structure in the SNAR when aspen ceased widespread flowering. The genetic structure of Tomnagowhan Wood, Strathspey, Scotland's largest aspen dominated woodland, was investigated using the same suite of markers. 186 aspen were samples, mapped and genotyped in an area of 4.6 ha. Twenty one clones were identified using a combination of isozyme analysis and probability theory. Clonal diversity in large aspen stands would appear much greater than is often asserted. A wide variety of clone sizes and strategies were also uncovered and leaf flushing date was shown to be an efficient phenotypic clonal marker. Analysis of the distribution of DBH suggests that aspen ramets flow a 'window of opportunity' mode of regeneration with the last burst of regeneration occurring around World War II in Tomnagowhan. Genotypic diversity in the small relict island apsen population of Orkney was estimated using the same suite of markers. Fifteen aspen locations were sampled. Multi-locus genotyping, probability theory and local knowledge were used to identify 12 independent clones. A review of the literature and legislation on the "safe" minimum number of clones to use in plantations suggests that the twelve Orkney aspen clones are sufficient to be safely used in the expansion of native woodland in Orkney thus enabling only locally derived stock to be used. The findings of the three surveys are discussed within an ecological and historical context providing a basis for recommendations for the future management and genetic conservation of the Scottish native aspen resource including guidelines on the conservation of the existing resources, expansion of the resource, the production and deployment of aspen planting stock and strategic initiatives.