Gene flow from planted common ash (Fraxinus excelsior L.) stands to old-growth forests – implications for conservation and use of genetic resources affected by ash dieback

Common ash (Fraxinus excelsior L.) is a keystone species, ecologically and economically important in European forests since it supports many associated species, and also produces valuable timber. The transformation of continuous forests to smaller forest fragments influences how ash trees exchange genes, and reduced gene flow may lead to a decrease or a loss of diversity and increased inbreeding. The patterns of genetic diversity are also influenced by plantation programs. Planted forests are often based on seeds of non-local origin, and sometimes from seed sources with narrow genetic variation. Seed and pollen from such plantings will mix with that of local trees, and the combined influence of fragmentation and planting programmes will, over generations, influence the genetic resources of ash trees and most likely their ability to respond to new challenges including climate change, new pests and pathogens. In recent years, the emerging infectious pathogen, Hymenoscyphus fraxineus, has caused severe ash dieback with high mortality throughout Europe. This may influence the genetic diversity of ash forests dramatically, because high mortality can create inbreeding and loss of diversity in the absence of effective gene flow, but also because the ability to resist the disease has been shown to be under strong genetic control. It is therefore important to understand how gene flow takes place across forests and landscapes in order to support sound management of endangered European ash trees and forests. The present thesis presents and discusses three studies that contribute new knowledge on how pollen and seeds are dispersed across forests and landscapes. The studies are based on field observations and DNA analyses using population genetics approaches.  Paper 1 addresses seed and pollen movement within – and into ‒ a small, native forest fragment surrounded by agricultural land at Rösenbeck, Germany. The ash trees in the small forest are of putative local origin, while roadside trees close to the forest were planted. Categorical and neighbourhood model approaches were applied to estimate gene flow patterns at the study site. Local meteorological data was used to evaluate the effect of wind on the dispersal patterns of seed and pollen. The results showed that the forest fragment has maintained a high level of genetic diversity. Long distance dispersal (LDD) events of both seed and pollen, and the influence of local wind conditions on the dispersal patterns, were documented. The majority of the seeds (55‒64 %) and seedlings (75‒98 %) sampled in the native forest patch had both parents among mature trees within the same patch, and pollen flow from the cultivated road side trees was very limited (2 %). However, a considerable amount of the seeds and seedlings (26‒45 %) must have been pollinated by trees located in more remote parts of the landscape, suggesting substantial pollen immigration into the study site. The results show that fragmented ash forests are likely to receive substantial gene flow through pollen, but also that trees located within a maximum of 100 m distance will most likely be able to mate. This finding is important for planning of gene conservation stands. Available wind data from weather stations in proximity can support in the planning process.  Paper 2 addresses the gene flow and reproductive success of ash forests based on a study site in Valby Hegn, Denmark. Compared to Rösenbeck, Valby Hegn represents a much larger forest complex, where single and planted compartments of ash trees are intermixed with mainly oak and beech trees. Unlike Rösenbeck, the ash trees in Valby Hegn were already severely influenced by ash dieback when the study was performed. The study site was further selected because it included a mature planting with trees from an old Danish-seed orchard (‘FP202’). The setup of the FP202 planting made it possible to estimate seed and pollen flow from a planted seed source into surrounding older stands, which is of special interest, since it may provide a model for enriching existing ash forests with planted trees that are highly resistant to ash dieback. This is an option that can be considered to improve the fitness of future generations of existing native ash forests. The neighbourhood model was applied to estimate mating system parameters and genealogies, and to understand the relationship between individual reproductive success and the phenotypic traits including fruit set, flower intensity, size, and crown dieback (i.e. ash dieback symptoms). The results showed that a large amount of seeds (39 %) and pollen (31 %) were dispersed from the planted trees into the surrounding forest, although the frequency of the gametes from the planted trees decreased rapidly with increased distance. The mean seed and pollen dispersal distances were 67 m and 347 m, respectively. The findings suggest that seed dispersal beyond 100‒140 m is rare within a forest, but LDDs through pollen over hundreds (sometimes thousands of meters) must be expected, allowing trees that survive ash dieback to mate. Planted trees with high levels of ash dieback resistance are therefore likely to spread their genes into fairly large neighbourhoods as pollen donors, while seeds from enrichment plantings probably will be able to colonise areas with 0‒100 m from the plantings. A very interesting finding was that the seedlings randomly sampled in Valby Hegn predominantly originated from healthy parents. This was the case for both male and female parents. The findings suggest that enrichment plantings with seeds from tolerant ash genotypes can potentially improve the future health of the species in European forests affected by ash dieback.     Paper 3 directly addresses the effect of ash dieback susceptibility on male and female reproductive success. This study is based on a clonal field trial in Tuse Nӕs, Denmark, where the ash dieback susceptibility of 39 clones has already been studied for several years. Gender, seed productivity and paternity success of the 39 clones were estimated in order to test the hypothesis that the clones with the lowest levels of susceptibility had the highest levels of reproductive success. The results revealed that females with low levels of susceptibility produced more seeds compared to highly susceptible females. On the male side, the pattern was the same, although less clear. The male clone with the lowest level of susceptibility (Clone no. 18) was by far the most effective pollen donor. However, highly susceptible males also produced some offspring, showing that selection for lower levels of susceptibility is slower on the male side. The overall result is that healthy ash trees are likely to contribute more to the next generation due to sexual selection in favour of less susceptible individuals. This supports the observations from Valby Hegn (Paper 2 as discussed above), that healthy trees are likely to become overrepresented when founding the next generation ash trees at generation turnover. This is good news for European ash forests, although the future of ash forests remains uncertain, and is a still major concern.