European and national guidelines currently encourage the planting of local tree seed, over seed from non-native sources. One objective of FRAXIGEN was to provide scientific evidence to inform this policy. We carried out a range of field and laboratory studies to investigate patterns of genetic variation and local adaptation in three species (Fraxinus excelsior, F. angustifolia, F. ornus) and six countries (U.K., Sweden, Spain, Greece, Romania and Slovakia). OBSERVATIONS OF PHENOLOGY: Our studies of flowering, fruiting and leaf flushing showed that different stands flush at different times, and this is related to latitude, longitude and elevation. For example, Romanian material planted in provenance trials of F. excelsior in southern England flushes earlier than local material and can be damaged by late frosts, causing forking. Late frosts can also kill flowers and reduce fruit set, even in local material. F. angustifolia is particularly at risk because it flowers early in the year (January-March). KEY RECOMMENDATION: Avoid moving material over large latitudinal or longitudinal ranges since it is likely to be poorly adapted to local climate, causing forking and poor seed set. STUDIES OF GENETIC VARIATION: (i) Reciprocal transplant experiments: We established field trials in UK (F. excelsior) and Greece (F. angustifolia, F. ornus) to assess the geographic scale of adaptive variation. Local material was tested against material collected at sites increasingly further away, and the trial was replicated at all the sites. We measured germination percentage of direct sown seed, and survival and growth of one-year-old seedlings. Preliminary results showed no correlation of height growth with patterns of ecological or geographic variation. No local adaptation was observed at the scale of the experiments, covering distances up to about 250km. KEY CONCLUSIONS & RECOMMENDATIONS: 1. Provenance description for ash can operate at the scale of at least 200-300km without compromising local adaptation. 2. Good provenances can be moved within this range without compromising performance. (ii) Molecular markers: We assessed levels of neutral genetic variation in stands of ash across Europe using both chloroplast and nuclear markers. (a) Chloroplast markers Chloroplast markers, being maternally inherited, can be used to investigate the movement of species via seed dispersal over long time periods. They provide a good picture of broad scale population structure over large distances. Fraxinus angustifolia and F. ornus both have southerly distributions in Europe, with postglacial migration predominantly from east to west, originating from glacial refugia in southeastern Europe and western Asia. Fraxinus excelsior is today widespread across central and northern Europe. Glacial refugia were probably located in Iberia, Italy, the eastern Alps and the Balkan Peninsula. Migration was predominantly northerly and from more than one direction from these refugia, after the retreat of the ice. KEY CONCLUSIONS & RECOMMENDATIONS: 1. Where conservation of genetic diversity is the highest priority it should aim to be guided at the largest scale by the distribution of haplotypes 2. In countries where only a single ash haplotype is present, chloroplast markers can be used to detect some non-native material. 3. Most haplotypes have widespread distributions, giving little support to a strategy of dividing large regions of provenance into small seed zones. (b) Nuclear markers We used nuclear microsatellite markers to assess the overall level of genetic variation in ash species, the differences within stands and between stands, and how this variation is distributed across Europe. We found high levels of genetic diversity within stands in all three Fraxinus species. Fraxinus excelsior and F. angustifolia showed few genetic differences between stands, suggesting high levels of gene flow via pollen movement (as expected for species with wind-dispersed pollen). At the northern edge of the range of F. excelsior, stands are more isolated and have lower levels of diversity. They are more differentiated because there is less pollen flow between them. Fraxinus ornus pollen is dispersed by insects. The genetic differences between stands in any one locality are few, but they increase with geographic distance. The foraging behaviour of the insect pollinators tends to move pollen between neighbouring trees. Closer individuals are therefore more likely to be related than more distant ones. KEY CONCLUSIONS & RECOMMENDATIONS: 1. High levels of gene flow between stands argue against the strategy of dividing large regions of provenance into small seed zones. 2. Geographically distant stands show more genetic differences in insect-pollinated Fraxinus ornus than wind-pollinated F. excelsior and F. angustifolia.
An important element of FRAXIGEN has been the interpretation of our research findings into a language and format that would make them accessible to our end-user community, including policy-makers, conservation bodies, forestry practitioners and seed collectors. Early in the project, a brochure was produced in six languages (English, Swedish, Spanish, Greek, Romanian, Slovak) to give an overview of the purpose and approach of FRAXIGEN. The content and layout was modified in each version to take account of each country's priorities and species: Fraxinus excelsior in northern and Central Europe, F. angustifolia and F. ornus in Central/Eastern Europe and the Mediterranean. The brochures were intended for dissemination of general information about the project to potential end-users of the research as well as interested members of the public. They were widely distributed in the respective project countries and are still available in downloadable format on the project website. The website (http://www.fraxigen.net) is a key dissemination medium for the project. It gives current information about FRAXIGEN including contact details for all project partners, downloadable versions of the project brochures, and reports of user group activities. The book summarising the project results (see below "Ash species in Europe: Biological Characteristics and Practical Guidelines for Sustainable Use") is also downloadable from the website. As more public results become available they will also continue to be posted. We aim to maintain the website until 2010. The most important non-academic output of FRAXIGEN is the book "Ash species in Europe: Biological characteristics and practical guidelines for sustainable use", in which we have summarised our research methodology, principal findings, and most importantly, the conclusions and recommendations arising from our research which are of direct relevance to both policy-makers and practitioners in the field of forest genetic resource conservation. This is being distributed through the end-user networks, which we have developed in all the project countries.
A seed collection with a broad genetic base will avoid inbreeding depression; ensure that planted stands remain viable and produce seed; and maintain adaptability in future environments. Decisions are needed on which trees within a stand to collect from; whether to collect from isolated trees; and whether or not to collect in poor seed years. We have studied reproductive biology in Fraxinus to determine the factors determining genetic diversity in seed. Each European ash species has a different sexual system. F. excelsior has male, female and hermaphrodite trees, with no clear divisions between genders. F. ornus has only males and hermaphrodites, whilst F. angustifolia is purely hermaphrodite. Controlled pollinations and molecular markers showed that selfing is possible on hermaphrodite trees, but naturally Fraxinus is highly outcrossing with negligible selfing in both mast and non-mast years. Mating between closely related trees rarely occurs, because most pollen is moved over tens of meters by insects (F. ornus) or hundreds of meters by wind (F. angustifolia, F. excelsior). Therefore there is little risk of inbreeding in natural populations. Pollen production in F. angustifolia overlaps with the opening of female flowers in F. excelsior. This means that there is potential for hybridisation in mixed stands, and the hybrid seed is most likely to occur on F. excelsior trees. GUIDELINES FOR SEED COLLECTION Adequate sampling depends on the number of seed collected per tree and the species' mating/pollination system. Wind pollination (F. excelsior, F. angustifolia) is likely to result in trees producing seed from many fathers, whereas insect pollination (F. ornus) is likely to produce seed of fewer fathers per tree. To make a seed collection with similar levels of genetic diversity requires collection from roughly 50% more trees in F. ornus than in F. excelsior or F. angustifolia. In F. excelsior or F. angustifolia, a collection from only ten mother trees, of at least 500 seeds per tree, will capture 95% of all the available genetic variation in the collection area, apart from rare alleles (gene variants) which comprise less than 5% of the total. In F. ornus, fifteen trees must be included in the collection to ensure the same genetic variation in the seed. Spacing between seed trees is also important for diverse collections. In F. excelsior seed trees should be at least 150m apart to ensure collection from both unrelated females and different pollen pools. With F. ornus most seed trees mate with different fathers, so it is sufficient to collect from trees separated by the seed dispersal distance (30m). KEY CONCLUSIONS AND RECOMMENDATIONS 1. Seed can be collected from females and hermaphrodites, isolated trees (as part of larger collections), and in non-mast years, without risk of inbreeding. Contrary to popular belief, F. excelsior trees do not change sex. 2. In F. excelsior and F. angustifolia, seed collections should include at least ten trees separated by at least 150 m; in F. ornus, the trees need be only 30m apart but at least 15 trees should be included in the collection. 3. It is better to spread a collection over a wider area, keeping the minimum distance between seed trees, rather than restrict the number of trees or reduce the distance between trees. 4. In mixed F. excelsior/F. angustifolia stands the seed from F. excelsior trees may be hybrid. LESSONS FOR NATURAL REGENERATION Sustainable management of native woods or their restoration through natural regeneration requires genetically diverse seed. In F. angustifolia and F. excelsior high levels of pollen flow mean natural regeneration maintains high levels of genetic diversity, even in small fragmented woodland remnants. Even where isolated trees form the basis for regeneration, the diverse pollen pool fathering seed means regeneration will be genetically diverse. Conversely, the high level of gene flow means native woods are not 'isolated, uncontaminated, local gene pools'. Natural regeneration in a reserve also contains 'foreign' alleles from external pollen sources. KEY CONCLUSIONS AND RECOMMENDATIONS 1. In F. ornus small isolated remnants (<15 seed bearing trees) should be avoided as the only source of natural regeneration, since insect pollination patterns means regeneration is likely to be lower in genetic diversity than in larger remnants. Native populations of all three European Fraxinus species maintain high levels of genetic variation. Neither the species, nor their genetic resources are seriously threatened and in need of specific conservation programmes. Nevertheless activities such as seed collection, planting and restoration through natural regeneration need to take note of FRAXIGEN's recommendations if genetic diversity is to be maintained locally, with future productivity and adaptability unaffected.