Extinction risks and the re-introduction of plant species in a fragmented europe

Exploitable results

Sampling involved establishing the fundamental relationships between work to be undertaken by WP5 and other workpackages. Initial discussions were held with other workpackages about collection of shared data to minimise overlap of work. Protocols for long-term field and laboratory experiments were designed. These included: i) Experiments on recruitment of seedlings under field conditions, ii) Seed burial and recovery experiments, iii) Studies of the size and composition of soil seed banks under field conditions, conducted by soil core sampling, iv) Experiments on seed dispersal under field conditions, and measurement of seed terminal velocities under controlled laboratory conditions, v) Comparative studies of germination responses to temperature of seeds collected from different countries of origin including all of the countries participating in TRANSPLANT. Protocols for experimental work were finalised and distributed to other work packages. A novel design was developed for the equipment used to measure seed terminal velocities. Components were ordered, assembled and tested exhaustively prior to being used for measurement of seed dispersal characteristics and modelling of patterns of seed distribution around source plants and populations. Extensive fieldwork was undertaken to identify field sites appropriate for experiments to be conducted under WP5, and for experiments undertaken on behalf of WP2 and WP3. In addition, the field sites chosen were characterised using physical and edaphic measurements. Seeds were collected from all participating WPs for terminal velocity trials, for the UK-based seed burial experiment, and for experiments undertaken as part of WP2. The legal implications, and customs regulations for sending soil, seeds and plant material from partner countries participating in TRANSPLANT, and from other EU and non-EU countries, were examined. Appropriate protocols were designed for processing and packaging seeds for shipment from different countries without causing damage to the seeds or alteration of their physical or architectural properties.

Understanding the persistence of seeds within the soil seed bank is a fundamental requirement for understanding plant population dynamics. Knowledge about seed longevity can help determine whether species can potentially re-establish from seeds stored in the seed bank after local extinction events. Seed longevity was studied for the following six species of dry calcareous grassland: Carlina vulgaris, Daucus carota, Hypochoeris radicata, Pimpinella saxifraga, Succisa pratensis and Tragopogon pratensis. Workpackages 2-6 collected seeds from the UK, The Netherlands, Germany, Sweden and the Czech Republic. These seeds were air-dried, and sent, under standardised packing conditions, to the UK, where they were stored at room temperature until used for experiments. Prior to use in experiments, initial percentage viability levels were determined for seeds of all species collected from all countries. Seeds were subjected to one of four treatments: (i) burial in the field for 7 months, (ii) burial in the field for 19 months, (iii) storage under ideal conditions to preserve viability in the Millennium Seed Bank (MSB), Wakehurst Place, for 7 months, (iv) storage under ideal conditions at MSB for 19 months. Seeds placed under field conditions were sewn into nylon bags with a small volume of soil, and buried at a depth of 5 cm in dry calcareous grassland field sites. The turf was replaced after burial of the seeds. Seeds to be stored at MSB were first dried through a series of reduced relative humidity rooms, and then vacuum-sealed in light-proof bags. They were then stored within the MSB vault at -20 degrees Celsius. After retrieval from the burial treatments, seeds were rinsed free of soil and test-crushed for viability. Seeds stored at MSB were brought back to room temperature, and tested for germination in a controlled environment room. At the end of the period allowed for germination, seeds that had not germinated were checked for viability using the crush test. In the burial treatments all species exhibited reductions in viability over time. Viability of Daucus carota decreased during the first 7 months, but remained stable thereafter. Differences in seed viability due to country of origin were seen in all species, but the differences were not consistent across species. For most species from most countries of origin, storage under ideal conditions at the MSB resulted in little if any reduction in viability over time. Exceptions that did exhibit declining viability included Succisa pratensis seeds from the UK and the Netherlands, and Tragopogon pratensis ssp. pratensis. Results from the seed burial treatments indicated that a considerable number of seeds of all species examined would be capable of remaining viable under field conditions for at least 19 months. Daucus carota can be characterised as having a long-term seed bank. It was confirmed that the conditions of storage at the MSB are successful in maintaining viability of the majority of the species we tested. Variation in maintenance of viability of seeds of the same species collected from different countries demonstrates that checks of the viability of stored seeds may need to be done on a population or country basis rather than at the level of species, to avoid failure to detect loss through time of stored seeds.

Viability of seeds of different weights, belonging to five species, was analysed by examining seeds from the two extremities, and from the centres, of the individual seed weight distributions for each species. For Hypochoeris and Pimpinella, all seed weight categories exhibited >80% germinability. For Carlina and Tragopogon, heavy and medium seeds displayed >70% germinability whereas small seeds exhibited over 50% germinability. Succisa germinability fell as seed weight declined, to a value of 30-50% for the lightest seeds. Many of the lightest seeds of these calcareous grassland species are capable of germination. Initial viability of seeds of six species collected from the five countries in the Transplant consortium has also been recorded to enable comparisons of seed germination and viability properties, and loss of viability with the passage of time. Analysis of initial germinability, and germination after 7 and 19 months, under ideal storage conditions at the Millennium Seed Bank (MSB), Wakehurst Place, has allowed calculation of rates of decay in viability. These results will be shared with Workpackage 4. An experiment was carried out to determine whether recruitment into populations of species of dry calcareous grassland is dispersal limited (i.e. limited by the ability of seeds to reach nearby but isolated sites) or access limited (limited by inability of species to establish in previously unoccupied sites). Partners from Workpackages 2-6 conducted this experiment in the UK, the Netherlands, Germany, Sweden and the Czech Republic. Four species (Tragopogon pratensis, Hypochoeris radicata, Carlina vulgaris and Succisa pratensis) were chosen for study. Seeds were collected from each of the participating countries, and used for analysis of recruitment within the country of collection. The reason for this was to ensure maximisation of recruitment potential. Failure of recruitment could otherwise have been entirely attributable to incompatibilities between local species adaptations and environmental conditions in each country. Within each country, five permanent quadrants were established in one occupied and three unoccupied sites, and seeds were sown into these sites. The unoccupied sites were chosen as suitable sites from which each of the species was absent prior to the experiment. All unoccupied sites were grassland areas selected for increasing distance from the occupied sites. Recruitment was monitored soon after seeds were sown and also during the following spring and summer. Studies were conducted to compare germination responses to temperature for seeds of three selected dry calcareous grassland species (Daucus carota, Succisa pratensis, Pimpinella saxifraga) collected from eight European countries (the UK, the Netherlands, Germany, Sweden, Czech Republic, Estonia, Crete and Italy). Seeds from each were sown separately onto agar plates in petri dishes at 13 temperature values, on a thermogradient plate with a constant, unidirectional temperature gradient from 9 degrees Celsius to 34 degrees Celsius. The light regime was set at 12hours light/12 hours dark. Differences have been established in speed of germination, optimum and range of germination temperatures for seeds from different provenances. Further runs of the experiment are almost analysed. Work packages 2-6 supplied WP5 with seeds for this experiment. Differences in growth of plants grown from seeds of different provenances, under standard conditions, have been studied, using seeds of six species (Carlina vulgaris, Daucus carota, Hypochoeris radicata, Succisa pratensis, Pimpinella saxifraga and Tragopogon pratensis) from each of five European countries (the UK, the Netherlands, Germany, Sweden, the Czech Republic). Seedlings were sown into peat and grown in a greenhouse at 20±10 degrees Celsius under a 16-hour light/8 hour dark light regime. Records were made of plant height, length of longest leaf and number of leaves, at weekly intervals from week 3 until week 8 of growth, and on three occasions afterwards at five-weekly intervals.

Due to habitat destruction and fragmentation for many plant species in Europe more populations are currently becoming extinct than are founded. A useful tool to re-establish a balance between extinctions and colonisations could be the re-introduction of species into sites where the habitat is suitable, but the species do not occur due to insufficient dispersal of seeds. However, little is known about the importance of the source of seeds for the success of reintroductions. Large populations are assumed to be less affected by genetic drift and inbreeding than small populations and may therefore be better sources of seeds. However, frequently, large local populations from which seeds could be collected are not available, whereas in other areas large populations still exist. If plants were adapted to regional or even local environmental conditions, choosing the right source for re-introductions would be very important for the success of these measures. We carried out reciprocal transplant experiments with several plant species both at a large scale (among several European regions) and a small scale (among several sites within European regions) and studied the survival and performance of plants of different origin. The likelihood of gene flow among populations is larger in species with good dispersal ability and the generation time longer in long-lived species. Therefore, in long-lived and well-dispersed species the differentiation among populations and their adaptation to local conditions may be less pronounced than in short-lived species of poor dispersal ability. We used three species with different life-history traits for reciprocal transplant experiments at different geographical scales: Carlina vulgaris (short-lived, monocarpic, poor dispersal ability), Hypochoeris radicata (polycarpic, high dispersal ability) and Pimpinella saxifraga (long-lived, low dispersal ability). In collaboration with the project partners (WP 3, WP4 and WP6), seedlings of C. vulgaris were reciprocally transplanted among five different European regions (Czechia, Germany, Sweden, Switzerland, Luxembourg) and among four populations of different size within these different European regions (local scale); seedlings of H. radicata were reciprocally transplanted among three regions (Czechia, Germany, The Netherlands) and within the regions among 3-4 populations each; and seedlings of P. saxifraga were reciprocally transplanted among three regions (Czechia, Germany, The Netherlands) and within the regions among 4 populations each. Survival and performance of the transplants were recorded during at least two growing seasons. In Pimpinella the survival rate was very low in all regions independent of the origin of the seedlings. Thus, reintroduction of this species using seedlings would require large numbers. In Carlina the region of origin of the plants influenced their performance. Overall, plants from Sweden performed much worse than plants from the other regions. In Hypochoeris the effects of population origin were much weaker than in Carlina, which supports the hypothesis that differentiation among populations is stronger in the short-lived species with low dispersal ability (Carlina). The most important result was that in both species there were clear indications of adaptation at the regional scale. Survival and overall performance of Carlina transplants were higher in their home region and decreased with increasing geographical distance to their site of origin and transplants of Hypochoeris survived best in their region of origin. At the local scale, plants from their home site performed better than plants from other populations of origin in Hypochoeris, but not in Carlina. Transplants of Carlina originating from large populations had a higher fitness than those from small populations. In Hypochoeris population size did not have an effect, suggesting that this better dispersed species was suffering less from the effects of inbreeding in small populations. The conclusion from the studies is that plant species that occur in different European countries consist of different ecotypes that are adapted to the specific conditions in these regions. It is therefore important to conserve large viable populations in the different regions to preserve plant genetic resources. Moreover, red data lists at the regional level and management measures at that level to conserve populations of rare plants are important. Because of the strong regional genetic differentiation, a strong decline of a species in one or more regions is a reason for concern, even if at the European scale the species is not threatened. For reintroductions material from large populations within the same region should be used, in particular, if a species is not well dispersed. If no large populations are left, plants from small populations in the same region should be propagated and used.

Estimating genetic diversity and the extent of gene flow between populations in fragmented landscapes was the other main goal of our work package. Two different molecular markers were developed to estimate genetic diversity and gene flow, respectively. Genetic diversity was analysed using chloroplast DNA whereas gene flow will be assessed using microsatellite Chloroplast DNA (cpDNA hereafter) is commonly used in evolutionary and phylogenetic studies given its particular characteristics. In particular, cpDNA evolves slowly and most cpDNA polymorphisms are thought to be caused by structural rearrangement or mutation. However, several other studies have used cpDNA to assess the large-scale genetic relationship among and between populations of plant species across a considerable part of their distribution area. Genetic variation of several TRANSPLANT study species (Succisa pratensis, Hypochaeris radicata, Tragopogon pratensis, Scabiosa columbaria, Pimpinella saxifraga, Ranunculus bulbosus and Carlina vulgaris) was analysed using all cpDNA markers available (around 15 cpDNA markers) and the method of polymerase chain reaction - restriction fragment length polymorphism (PCR¿RFLP). To our knowledge, it must be noted that the analysis of cpDNA variation in these species represents a novel result. Microsatellites are nuclear markers based on the repetition of bases that can show high levels of polymorphism. It still remains controversial whether the variation shown by microsatellites is neutral or not. Nevertheless, microsatellites have been proved to be highly effective for population genetics studies, such as parental analysis, population genetic structuring, or gene flow among fragmented populations. Between 5-10 polymorphic microsatellites were successfully developed for Hypochaeris radicata and are being used to analyse the gene flow among fragmented populations in a Dutch landscape.

Regional distributions of animals and plants in landscapes are the result of local population dynamics within habitat patches and dispersal among patches. The dynamics and extinction risks of local populations are strongly influenced by local habitat conditions. Dispersal resulting in establishment of new populations or augmentation of existing populations critically depends on landscape configuration. Metapopulation theory explains distributions as determined by the balance between extinctions and colonisations, and it has generated an increased interest in large-scale spatial dynamics and landscape processes, such as habitat fragmentation. Hence, assessments of population viability within a larger landscape fragment must consider both these aspects. To model the importance of landscape configuration for the long-term survival probabilities of four target plant species in an agricultural landscape of 7 square kilometres, we collected information about landscape configuration, in terms of categorization of habitat into suitable or unsuitable and presence or absence of each of the target species. This information formed part of the GIS-model (see "Building GIS mode"). We continued by building more complex models that included also information on habitat quality and the land use history of each habitat patch. Lastly, we incorporated the local population dynamics of plant species assigned a particular type of local demography to a particular habitat quality type by using the demographic data collected (see "Sampling of demographic data"). Together with WP 4 we then constructed a dynamic model that simulated the dynamics of species in the landscape. The model was a realistic, spatially explicit, and dynamic metapopulation model and by using this model we were able to examine how the distribution of target species among patches was affected by local population dynamics, and by short-distance and long-distance dispersal. As a baseline in simulations we used the current land use conditions. We were also able to investigate how land use history influences the current distributions. The effect of landscape development on future distributions and the possibilities of long-term persistence in the landscape were examined through different land-use scenarios. Simulation results showed that the colonization and extinction dynamics in this agricultural landscape may be slow and the time frame for the population system to attain equilibrium in a constant landscape is very long. Hence, landscape history is important to interpret current distribution patterns. They also suggested that species only very slowly colonizes new habitats that are made available through changes in management. Sensitivity analyses demonstrated that the very small fraction of seeds dispersing over long distances have a large influence on the regional dynamics of species, whereas regular short-distance dispersal has almost no effect. The model also allowed us to assess how demographic processes affect not only local population growth but also regional dynamics. Our analyses of the importance of the configuration of an agricultural landscape on the survival possibilities of plant species are central to biological monitoring and risk assessment. One insight from the models is that conservation efforts should focus as much on dispersal processes and landscape configuration as on local habitat conditions. The results are also important in the sense that they illustrate that the dynamics of many perennial plant species is slow and that therefore time lags are important. The response of the landscape to changes in land use may therefore be slow and the changes that have already occurred are only partly expressed in the current-day distribution. These findings have important implications for assessing the effects of different management practices in agriculture, and show that it is not only important what management that is practiced a particular site but that also the landscape configuration of different management types play a role. From a research point of view our results are important because they suggest a way to study metapopulation dynamics also in systems with a very slow turn-over rates, thus being in non-equilibrium. The study system also provides unique opportunities to integrate, cultural geography, landscape history and ecology in education. Lastly, landscapes like our study system have high recreational values. These values are associated both with properties of local areas and properties of the landscape as a whole. To maintain these values we must therefore ensure that both local conditions and landscape configuration are such that the continued survival of key species is guaranteed.

In 2001 and 2002, Workpackages 2-6 collected seeds of five species from five countries (UK, The Netherlands, Germany, Sweden, the Czech Republic). The species were chosen from those assigned to different cells of a colonisation/longevity matrix previously established by TRANSPLANT partners. The species were Tragopogon pratensis (ssp. pratensis and orientalis), Hypochoeris radicata, Carlina vulgaris, Succisa pratensis and Pimpinella saxifraga. A standard sampling scheme was used to sample seeds from each of three small and three large populations in each country, for all species for which this was possible. Seeds were transported to the UK under controlled conditions, to ensure that they suffered no damage or change to their natural characteristics, and stored under standard conditions. Measurements were carried out on seed weights, lengths, pappus widths (in the cases where a pappus was present), and terminal velocity, using a machine designed specifically for this part of the project. Data were analysed for within-species differences in the measured characteristics that could be referred to country of origin, population size and year of collection. The assumptions that seed size is correlated with dispersal distance, and that lighter seeds are more prone to be incapable of germination, were also tested. For the latter test, germination was compared for seeds from the two extremities, and from the centres, of the individual seed weight distributions for each species. For Hypochoeris and Pimpinella, all seed weight categories exhibited >80% germinability. For Carlina and Tragopogon, heavy and medium seeds displayed >70% germinability, whereas small seeds exhibited just over 50% germinability. The germinability of Succisa seeds fell as seed weight declined, to a value of 30-50% for the lightest seeds. Nevertheless, many of the lightest seeds dispersed by all species are clearly capable of germination, and thus these seeds must be considered when dispersal and colonisation potentials of dry calcareous grassland species are investigated. These light seeds have the potential to establish in suitable habitat at the most extreme distances to which seeds can disperse. Deterministic models have been utilised, using measured individual seed and plant characteristics, to establish the seed dispersal kernels for the five species. Particular attention was paid to the long distance dispersal tail of the distributions. Characteristics measured at the species' field sites, including wind speeds and their frequency distributions, were used to create realistic scenarios of the conditions under which seeds disperse at each of the sites of collection. In addition, extreme wind conditions and standard wind profiles were used to model dispersal kernels. Using standard wind profiles, the characteristics of the dispersal kernels for the five species were found to be affected by different characteristics. For Pimpinella, average dispersal distances differed significantly between seeds obtained from different countries and between seeds from small and large populations (seeds from small populations had larger average values). Dispersal of Succisa seeds also differed significantly between countries, and seeds from small populations showed smaller dispersal distances. Carlina dispersal differed for seeds collected in different years. Tragopogon dispersal did not differ significantly between country of origin and size of population, although for seeds collected from different countries, significantly different proportions were uplifted by the wind profiles used in the models. Dispersal kernels for Hypochoeris seed showed significant variation due to country of origin, size of source population and year of collection. When extreme wind profiles were used in the model, the overall dispersal distances increased for all species. Ultimately, the dispersal kernels demonstrate that the majority of populations in fragmented patches of calcareous grassland habitats in Europe are effectively isolated, since seeds cannot disperse across the intervening countryside, even under the most favourable conditions. It can therefore be confirmed that many populations in fragmented calcareous grassland habitats are already effectively ecologically and evolutionarily isolated. Detailed examination has also been undertaken by WP5 of seed dispersal under field conditions using sticky trapping methods, study of soil seed banks by soil coring methods, and investigations of seed dispersal via dog-walking activities. Results indicate that 80% of seeds collected by dogs' coats are deposited within 0.25km, although 7% can be dispersed at least 3km. Seeds carried for such long distances would be capable of being moved between isolated fragments of calcareous grassland. Although most of the more widely dispersed seeds possessed structures obviously promoting adhesive dispersal, this was not always the case.

For sampling purposes c. 120 sites in Germany within an area of 30 x 25 km were visited and studied. About 80 sites containing populations of different size and degree of isolation of the target species (Carlina vulgaris, Hypochoeris radicata, Pimpinella saxifraga, Tragopogon pratensis, and Succisa pratensis and Daucus carota for other WPs) were then selected. Seeds were sampled in each population for the common garden studies (results: 9382) and the reciprocal transplant experiments (results: 9380), and also for studies performed by WP1 (crossing experiments), WP3 (demography studies) and WP5 (recruitment experiments, seed burial experiments, terminal velocity studies). Leaf samples were collected for WP1 (genetic variation). Biotic (e.g. species composition) and abiotic variables were recorded to characterise the sites, and soil samples were collected to carry out a bioassay in order to assess soil fertility. Seeds for the common garden study and the transplants experiments were collected and other seeds received from WP3, WP4, WP5 and WP6. The seeds were counted and weighed and batches of seeds were prepared for the transplant experiments. In spring 2001 batches of seeds containing random samples of equal size from each population were sent back to the project partners who carried out the reciprocal transplant experiments (WP3, WP4; WP6). Seeds were germinated, raised in nutrient-poor gardening soil in each study region for a few weeks and then the seedlings were transplanted into natural field populations either at their site of origin ("home site") or at one of the other regions ("away sites"). To study the effect of population size on reproduction and performance at the field sites, the number of seeds per plant and plant size was determined in a number of German populations of Carlina, Hypochoeris, Pimpinella and Tragopogon. These four species differ in the two traits dispersability and longevity: Hypochoeris and Tragopogon are well and Carlina and Pimpinella poorly dispersed, Carlina and Tragopogon are short-lived (monocarpic), whereas Hypochoeris and Pimpinella are perennial. Because of their short-generation time, short-lived species are thought to suffer faster from the negative genetic effects of small population size, and poorly dispersed plants are presumably more strongly isolated. However, we found that the number of seeds per plant increased with population size in the two obligate outcrossing species Carlina and Hypochoeris, but not in Pimpinella and Tragopogon. Thus, there were no consistent differences between well and poorly dispersed, and between short- and long-lived species. Instead, the results suggest that the breeding system is a better predictor of the effects of small population size on plants than longevity or dispersability: Reproduction of obligate outcrossers (Carlina and Hypochoeris) was more strongly affected by population size than that of self-compatible plants (Pimpinella and Tragopogon), presumably because of pollination limitation. In contrast to reproduction, plant performance in the field was not influenced by population size, the effects of habitat type were far more important. In conclusion, fragmentation has already negative effects on the reproduction of plants in natural populations, which could present a particular problem for monocarpic plants like Carlina that have to re-establish frequently from seeds.

The objective of this study was to investigate whether plants of the monocarpic perennial composites Tragopogon pratensis and Carlina vulgaris show local adaptations at a small scale within populations. The two species differ in their dispersability: Dispersal of the seeds of Carlina is relatively poor in comparison to the dispersal of the seeds of Tragopogon. This difference in dispersability may affect the degree of local adaptation in the two species. It was expected that Carlina, which is relatively poorly dispersed, would be locally adapted to a higher degree than the more better dispersed Tragopogon. This study complements the larger scale adaptation studies (see 9380). To study the occurrence of small-scale (within-population) local adaptation, reciprocal transplant experiments were carried in one German population of each species. Ten mother plants of each species that were growing within 300m of each other were selected, their position marked, and the seeds of these plants were collected and germinated in the greenhouse. Several seedlings from each mother plant were then transplanted into the vicinity of each mother plant. A number of fitness-related traits (survival, growth, flowering) was examined in the transplants over two growing seasons. No evidence for small-scale local adaptation was found in the species. There was no home site advantage and no effect of distance to the mother plant. Instead, some target sites were consistently favourable and others unfavourable irrespective of the origin of the transplants. These differences in performance could not be related to habitat characteristics recorded at the target microsites. The results indicate that even in populations that are not obviously heterogeneous, there is strong spatial heterogeneity in the quality of microsites for plants at a small scale. This is surprising, because the successful reproduction of the mother plants indicates that all microsites had been favourable for the species. Local adaptation appears not to be important at this scale; selection may be too weak and genes flow too strong for local adaptation to develop. In conclusion, the results suggest that in restoration projects seeds should be broadcast over the whole site to increase their probability of reaching at least some favourable microsites.

Demography data have been sampled from the study area, Nynäs, during three years for four plant species, Succisa pratensis, Tragopogon pratense, Ranunculus bulbosus and Carlina vulgaris. In total 13 populations were sampled, over a spectrum of habitats. Studies were made in the field using permanent plots at selected target populations, but also using experimental plots subjected to seed addition. The data concern the whole plant life cycles, from flowering and seed production over recruitment to development and survival of established plants. The data have been collected on a yearly basis and used as a basis for further analyses for the other deliverables in the same work package, but also exported to other workpackages in order to perform different tasks relating to comparative studies across Europe. Transition matrix models have been constructed based on the data, and used for simulation models and for analyses of historical and habitat configuration effects on species demography. Thus, the demographic data are not published separately. Demographic data are used for biological monitoring and risk assessment for endangered species. More specifically they can be used to determine trends in population size and probability of population extinction over a certain time interval, related to ongoing changes in land use. Different species were selected as representatives of existing variation in life cycles, i.e. mainly life span and dispersal ability. These traits were selected because they are fundamental for species response to habitat change and fragmentation. Moreover, data were collected in several habitat types, representing different management. By assessing how population viability of the target species is related to land use, we are able to predict the consequences of changes in management practices on the possibilities of long-term survival in the landscape. Knowledge of demography and population viability is relevant for agricultural planning, global change in land cover as well as for general ecology and plant biology. Biodiversity assessment is an important task for agricultural planning because a majority of plant species in Europe inhabit agricultural landscapes, particularly those with traditional management. Global change in land cover is therefore basically dependent on species response to land use change. To assess how species respond to such changes knowledge of variation in demographic performance related to habitat quality is essential. One central aspect is the time-scale of response, relevant for assessing the need for artificial transplantation of endangered species, and the time frames for restoration programmes. Semi-natural grasslands have high biodiversity values and a high recreational value and the preservation of these grasslands is therefore generally important. Since the recreational values largely depend on floristic composition, viability assessments of endangered species are an important tool for conservation. The sampled demographic data can also be useful for further research in population dynamics, risk assessment and landscape ecology. The produced knowledge will also be useful in education, both basic natural science and advanced teaching in ecology at universities.

Population demographic field data provide a wealthy source of information on the status of the population. Specific data give immediate insight in aspects of the life cycle, such as the structure of individual plant sizes telling about the survival and growth opportunities, and seed germination giving the opportunities for sexual recruitment. However, only by combining the population data the different aspects of the life cycle can be evaluated together and provide an integrated measure of the viability of the population. Models are powerful tools for integrating complex datasets. For the purpose of population viability studies, matrix projection models in particular have proven their usefulness. Matrix models logically sort the demographic parameters obtained in the field, categorized in age or size classes comprising groups of individuals with similar demographic fates. The major output variable that the model computes is the population growth rate, an appropriate measure of the viability of the population. In addition, sensitivity analyses may be carried out by which one may quantify what aspects of the life cycle contribute most to the growth rate of the population. The measures that the analyses generate (so called elasticity) have proven their usefulness in conservation ecology. Parameters with high elasticity are usually the most important targets for conservation management, as their improvement will boost the growth rate of the population most. A final tool within the context of the matrix modelling technique is Life Table Response Experiment (LTRE) analysis. This recently developed analysis gives insight in the factors that determine differences in the population growth rates between species, populations, types of management, and alike. It is a very powerful tool for unravelling what aspects of the life cycle are critically affected by certain habitat conditions and specific management actions. Elasticities and LTRE together thus comprise the matrix analysis toolbox that is indispensable for a proper evaluation of the viability of populations. We have applied these techniques to the populations monitored, as described under the result ¿Demography¿. The design of our European-wide monitoring programme on field demography allows systematic comparisons to be made between species, differing in elemental traits such as longevity and dispersal ability, between regions within the distribution range of species, in particular central vs. marginal populations, and population within regions, differing in size and in the degree of isolation. All individuals were classified on the basis of size criteria (number of leaves or leaf size, depending on the species), and flowering individuals were placed in a separate class from vegetative non-flowering individuals. To allow comparisons, the criteria were identical for a given species across its entire range. The results of the analyses show large differences between the populations, with a major part of the variation expressed at the level between the regions. Even in situations in which the population growth rates are similar (i.e. all populations are slightly increasing), the contributions of different life cycle stages can markedly differ. This indicates that the populations in different regions survive in radically different ways, and may respond to management actions in equally different ways. Our analyses identified the short-lived species Carlina vulgaris as the most variable within our species group. This small thistle can either grow fast and flower rapidly or extend its flowering and grow bigger but only in habitats and climates in which the survival chances are good. Carlina indicates that populations of the same species in different regions can behave almost as different species, underscoring the importance of conservation and a diversification of management actions at the regional level.

Fragmentation in many European landscapes is accompanied by a slow deterioration of many of the remaining habitats. This deterioration is largely due to change in the landscape management, such as cessation of grazing or mowing, increased load of nutrients etc. which typically lead to increased intensity of competition from grasses and/or all tall forbs. While the rare plants are often able to survive in these deteriorated habitats for some time, their response to the changed habitat conditions affects their ability to produce seeds that otherwise might serve as means for colonization of other habitats. The major part of the research in the WP4 concentrated on quantification of this response by manipulating the competitive stress (as a proxy for habitat deterioration) and observing changes in plant demography and allocation patterns. These experiments were conducted both in the experimental garden (using seeds collected in the sampling phase) and in field habitats. The experiments showed that increased competition of neighbours leads to a change in allocation pattern of experimental plants. In both common garden experiments and in the field sites, increased competition negatively influenced fecundity; survival was not affected. Decrease in investment to sexual reproduction was a function of smaller plant size of individuals growing with competitor (Hypochoeris radicata, Succisa pratensis); there was little evidence of shift to reproductive functions in worse conditions. In Hypochoeris radicata, size-independent effects of competition on allocation pattern were detected. Namely, plants grown without competitor produced more seeds than plants of the same size cultivated with competitor. However, increased allocation to generative reproduction did not result in lower investment to survival (to root biomass). This should be interpreted as the absence of resource-based trade-off between these two essential life functions. In one target plant (Succisa pratensis), the allocation shifts were compared over the European scale (the Czech Republic, the Netherlands and Sweden) to identify site-independence of the effects and the possibility to generalize the results. This comparison of the species response over the European scale showed strong site-dependent effects in the allocation pattern and its response to competition. While the Succisa populations from the Netherlands showed increased flowering at the expense of further survival at deteriorating habitats, no similar response was found for populations coming from other sites. This means that there is no general patterns of the species response to the habitat deterioration; always site or country specific information has to be sought to assess the species response to deterioration and assessing its change in dispersal capacities in response to that.

The objective of these experiments was to study offspring performance and quantitative genetic variation in Pimpinella saxifraga, Carlina vulgaris, and Hypochoeris radicata in relation to population size and isolation, and plasticity in Carlina. In a first set of experiments, plants were raised from seeds that had been collected in populations of a wide range of size and degree of isolation in different European regions (see 9379, 9380). Fragmentation reduces the size of populations and increases their isolation, and this may result in genetic erosion in the remaining populations. Poorly dispersed species like Carlina and Pimpinella would be expected to suffer more strongly from the negative effects of reduced population size and increased isolation than well-dispersed species like Hypochoeris. We examined several traits, both morphological and fitness-related, in the plants grown in a common garden over two years and analysed the distribution of quantitative genetic variation among regions, populations within regions, and families. There was strong variation in most of the investigated characters among regions, among populations within regions, and among seed families within populations in all three species, although regional differences were not expressed in all characters in Pimpinella and Hypochoeris. A partitioning of variation revealed higher differentiation among regions and among populations within regions in Carlina than in Hypochoeris, which was expected because of the higher dispersal ability and thus higher gene flow among populations in Hypochoeris. Population size and isolation did not influence the variability of morphological traits within populations in any of the species, but affected fitness-related traits in the two poorly dispersed species. In Pimpinella many reproductive traits and thus total fitness increased strongly with population size, but population isolation had little effect. In Carlina the size of flowering plants and multiplicative fitness increased with the size of the population of origin and decreased with increasing isolation. In contrast, there were almost no effects on the traits measured in Hypochoeris. In a second study, we studied phenotypic plasticity in Carlina to investigate whether plants of Carlina vulgaris originating from different European regions and from populations of different size and degree of isolation differed in their degree of plasticity. Seeds were collected in populations in Czechia, Germany, Switzerland, Luxembourg, The Netherlands, and Sweden that differed in size and isolation. From the seeds plants were raised in a common garden at the University of Marburg, Germany. Overall, more than 1200 plants from more than 50 populations were grown. During two growing seasons, each plant was subjected to one of four treatments: (1) control, (2) drought, (3) fertilizer, (4) drought and fertilizer. Drought is a common feature of the habitats of the species and eutrophication a problem that is affecting many populations. Fertilizing had a positive effect on vegetative growth and on reproduction, but reduced survivorship until the flowering period as well as flowering probability in surviving plants. Drought had weaker but overall favourable effects on plant performance, suggesting that this treatment was closer to optimal conditions for this species than were control conditions. For some traits the response to fertilizing differed significantly among regions, but the geographical pattern differed among traits. Fertilizing influenced the probabilities of survival and flowering of plants from the various populations within regions differently, but population size or isolation had virtually no effect on the reaction norms of plants. The results of the first study indicate that fragmentation is a serious problem for plants and that efforts should be made to avoid further reductions of population size. As expected, the negative genetic effects of fragmentation were stronger in poorly dispersed species like Carlina and Pimpinella than in well dispersed like Hypochoeris. Thus, knowledge of the dispersability of species may be useful for the prediction of the effects of fragmentation on different plant species. The strong genetic differentiation found among different European regions is consistent with the results of the transplant studies (see 9380) and indicates that it is important to conserve large populations of plants in different European regions to preserve the genetic variability of a species. The strong genetic variation among populations indicates very limited gene flow among the remaining fragmented populations within the regions. From the second study we conclude that for Carlina there was no indication that habitat fragmentation already had a negative effect on the phenotypic plasticity of this species. However, populations differed in their plasticity, which is consistent with the overall pattern of genetic isolation among the populations.

The crossing experiments aimed at assessing the effects of inbreeding depression on fitness traits in different plant species occurring in fragmented landscapes in Europe. In general, inbreeding depression reduces fitness traits (e.g. seed production, seed germination, plant growth and survival) and therefore weakens plant populations, fact that may increase the extinction risks of plant populations. Fragmentation reduces population size and increases isolation between populations, increasing the probability of inbreeding, reducing gene flow, and therefore increasing inbreeding depression. Inbreeding depression is mostly caused by the effect of recessive deleterious mutations that are expressed when individual plants self-fertilize or mate with close relatives (biparental inbreeding). The plant species used in the crossing experiments differ in mating system (self-compatible, mixed-mating system and self-incompatible), longevity (short- and long-lived), and dispersal ability (good and poor dispersers). These three life-history traits may have an important effect on the response of plants to inbreeding depression and subsequently on the viability of populations. Populations of self-compatible, long-lived plants with dispersal ability could cope with high levels of inbreeding whereas populations of self-incompatible, short-lived plants without dispersal ability could dramatically be affected. The study plant species were: Succisa pratensis (mixed-mating system, long-lived, poor disperser), Scabiosa columbaria (mixed-mating system, short-lived, poor disperser), Tragopogon pratensis (self-compatible, short-lived, good disperser), Hypochaeris radicata (self-incompatible, long-lived, good disperser), and Leontodon autumnalis (self-incompatible, long-lived, good disperser). The effects of inbreeding depression were analyzed on all life-cycle traits, from seed set (i.e. the proportion of ovules that set seed) to flowering of resulting plants. In general, results show that seed production is greatly affected by inbreeding depression for all species except Tragopogon pratensis. Seed production determines recruitment and represents a very important trait for the maintenance of populations. In fact, populations are no viable in the long term if recruitment fails or is strongly limited. The fact that seed production in Tragopogon pratensis was not affected by inbreeding is explained by the species self-compatibility system. As a result, populations of plant species with self-incompatible or mixed-mating systems might significantly reduce recruitment if inbreeding increases as a result of decreasing population size. The second important result was that dispersal was not affected by inbreeding depression. Hence, the dispersal potential of plant species of study was not reduced by genetic factors. Migration could buffer the effects of inbreeding depression on recruitment if dispersal is effective in connecting plant populations in a fragmented landscape. This mostly applies to good dispersers (Tragopogon pratensis, Hypochaeris radicata and Leontodon autumnalis), whereas poor dispersers (Succisa pratensis and Scabiosa columbaria) would have less chance to renovate the genetic load of the populations due to migration events. It must be noted, however, that poor dispersers are long-lived species. Thus, populations can persist for long periods of time during which the opportunities for long-distance dispersal events would increase. Finally, inbreeding depression effects on life-cycle traits are being included into demographic models to test the demographic implications of inbreeding depression. The study species used is Succisa pratensis. Results indicate that the species can cope with low levels of inbreeding, but under high levels of inbreeding depression, fragmented isolated populations can go extinct in 40 years. Conservation plans devoted to preserve plant populations in fragmented landscapes should take into account measures to prevent the genetic deterioration of populations and to avoid high levels of inbreeding depression. Re-introductions of plants from other populations seem to be an effective measure to restore the genetic variability of fragmented populations.

The ultimate goal of conservation efforts is to preserve species in their natural settings. Because many species survive regionally through a balance of colonisations and extinctions we need not only to preserve local habitat conditions but also the possibilities to re-colonise suitable habitat from which the species has gone extinct. However, because of continued habitat destruction the remaining habitat for many species have today become so fragmented that they are not able to re-colonise suitable habitat. Central issues are therefore to determine the extent to which the current distribution of species is limited by the dispersal capacity and if re-introduction of plant species through seed sowing experiments constitutes a realistic management option. A series of seed sowing experiments was carried out for the four target species, Succisa pratensis, Tragopogon pratense, Ranunculus bulbosus and Carlina vulgaris. The experiments included disturbance treatments and were performed both at occupied and un-occupied sites. Positive results, i.e. recruitment after seed sowing, at un-occupied sites indicate dispersal limitation, and provide support for the expected success of re-introduction of species. For all four species results suggested dispersal limitation, and, as a corollary, that re-introductions are expected to be a useful tool in restoration of species richness in semi-natural grasslands fragments in the present-day landscape. However, there were also differences among species in the degree of dispersal limitation. In addition to the results of dispersal limitation, our studies indicated that the success of the seed sowing depends on the floristic composition at the sites. This implies that local habitat conditions are also important for recruitment success. Furthermore, suitable targets for re-introductions could be identified by species surveys. Re-introductions are thus useful to counteract present species loss and decline of biodiversity. Furthermore, the results provide insights to knowledge about species regional dynamics, indicating dispersal limitation, and that this in turn may influence ongoing change in vegetation cover, due to altered land use. However, re-introductions are not uncontroversial, since they may be considered as negative for authenticity values. Further studies on the appreciation of artificial biodiversity should therefore be performed. Results from re-introductions may also add to risk assessments for endangered species. If re-introductions generally fails, this indicates that other factors than availability of apparently suitable sites, are responsible for species decline, i.e. the real suitability of sites has been misinterpreted. On the other hand, if the re-introductions are often successful, this implies that species are dispersal limited and that landscape configuration is essential, and must be taken into account in conservation planning. The fact that the degree of dispersal limitation varies among species indicates that the benefits of re-introductions will also vary among species. For agriculture the results imply that preservation of traditional management in semi-natural grasslands is essential. To promote dispersal and provide habitats for species, and thereby development of new populations, the surrounding landscape should also be used for extensive grazing. Conservation programmes should increase focus on whole landscapes. Economically sustainable systems for maintaining grazing, and for cooperation among farmers in order to increase the extent of grazed areas, should be developed. Such landscapes will also promote recreational values of landscapes. The results will also be valuable for research on regional plant population dynamics, since they add to a growing body of evidence that plant species in fragmented landscapes are dispersal limited.

Sampling served as a supporting activity to other deliverables of WP4 (clonal demography and measuring allocation pattern) and other WPs. It was thus the necessary component for the other deliverables in the WP4 and enabled also other work packages to extend their spatial scope. It consisted of site selection and sampling plant material. Sites were selected from floristic and habitat databases to satisfy the set of criteria that define dry species-rich grasslands and contain or are likely to contain the target species of the TRANSPLANT project. Initially 500 site records were considered, 100 sites were visited and ultimately 5 sites were chosen both for field experiments conducted within WP4 and for sampling of plant material both for WP4 and for other work packages. Clonal demography was performed directly at the sites selected during the initial screening phase. For measuring allocation pattern (WP4), seeds were collected at several localities in order to start allocation experiments performed in common garden experiments. Seeds were the material primarily sampled also for other work packages: WP2 (local adaptation study), WP3 (demography study), WP5 (terminal velocity and seed burial studies). Leaves were sampled for WP1 (regional genetic variability study).

Fragmentation in many European landscapes is accompanied by a slow deterioration of many of the remaining habitats. This deterioration is largely due to change in the landscape management, such as cessation of grazing or mowing, increased load of nutrients etc. which typically lead to increased intensity of competition from grasses and/or all tall forbs. While the rare plants are often able to survive in these deteriorated habitats for some time, competitive effects of a few tall species change their survival ability. This is particularly important for long-lived plants with vegetative reproduction, as their life cycles typically show large sensitivity to change in their survival parameters, and reproductive parameters are less important than in annual or short-lived plants. The second major task of the WP4 thus was an assessment of changes in ramet-level demography of such long-lived perennial plants. This was done by manipulating competitive stress and observing changes in plant survival in three selected grassland perennials (Centaurea scabiosa, Hypochoeris radicata and Succisa pratensis). Target plants were sampled over 3 vegetation seasons (2001-2003) in plots that differed in habitat quality (two levels of competition that was manipulated by clipping of neighbours). To be able to make a full account of species demography (in order the importance of survival can be assessed in the context of the whole life cycle of the plant), all plant size classes were included in the measurements. Using these data it became possible to assess the role of this survival to the overall demography of the plants. The results demonstrated a clear negative effect of habitat deterioration on population survival. Critical life cycle stages for population dynamics of these three species were identified. In species Centaurea scabiosa and Hypochoeris radicata, lower survival of vegetative and generative adults in plots with stronger competitive stress most negatively influenced population dynamics. In Succisa pratensis, there is a clear effect of competition on flowering and seed production patterns, but population dynamics of this species were only weakly affected by competitive stress.

The survival probabilities of local populations in fragmented landscapes depend on the local demography of the species. The demography is determined by the vital rates of the plant species, i.e. the survival and growth of individuals, their reproduction and the establishment of new recruits. These vital rates vary among species. In particular we expect that survival of individuals versus seed dispersal and seed recruitment differ, and these vital rates are the main targets for study in TRANSPLANT. In addition, there may be systematic variation in demographic parameters between local populations as well as populations between different regions. Local differences are likely to be determined by the management of the site, soil properties and the history of the site determining selection on the properties of the plants in the past. Regional differences occur at the scale of the entire range of the species within Europe and differences are especially expected between populations at the centre of the distribution and populations at the margin. The latter populations are growing at their abiotic and biotic limits and are thus expected to be more vulnerable. It is unknown, however, how this vulnerability translates into specific demographic characteristics. These considerations have resulted into a unique European-wide monitoring scheme of plant demography. Four species have been involved. Carlina vulgaris and Succissa pratensis as species with low dispersal capacity, (seeds without adaptations for dispersal), and Hypochaeris radicata and Tragopogon pratensis as species with high dispersal capacity (seeds with pappus for wind dispersal). Carlina vulgaris and Tragopogon pratensis form short-lived monocarpic individuals that die after flowering, while Hypochaeris vulgaris and Succissa pratensis are longer-lived and polycarpic and clonal by which individuals survive after flowering. In close collaboration with other work packages, populations have been monitored for each of these species in at least three regions within Europe (including central and marginal populations), for at least two populations per region, for at least three growing seasons. Population sizes ranged from less than 100 up to several thousands of individuals, and distances to the closest populations ranged from 100 m up to several kilometres. Demographic field monitoring consisted of two parts. First, established plant individuals were censused in permanent quadrates within a site. In order to follow their fates over time, individuals were tagged and their coordinates within the quadrate recorded. Plant sizes were estimated by non-destructive measurement of the number of leaves and the leaf lengths. Growth, flowering and survival rates can thus be accurately determined if no less than about 100 individuals per population are censused at least once per growing season. Seed set was estimated by recording the number of capitula (flower heads) per flowering individual, accompanied by counting the number of mature seeds per flower head on individuals outside the quadrates. The second part of the demographic field monitoring involved an independent assessment of the germination and establishment probabilities of the species in the populations. To this aim a fixed number of seeds obtained from individuals in the population were sown in separate field plots. Germination and establishment (growth and survival) of seedlings are then recorded at regular intervals in these plots. Control plots that have not received any seeds are monitored to quantify background germination levels. These levels are generally low, however, and this is the reason that a separate evaluation of these important life cycle components is necessary. The results of the demographic measurements indicate that methods described above are appropriate for a concise assessment of the different plant life histories, both within the regional level and across the European scale. For some species (e.g. Tragopogon pratensis), our methods revealed large differences in plant sizes and reproductive schedules between nearby populations. The protocols may serve as a blueprint for many studies interested in the status of populations, either endangered or not, allowing an accurate evaluation of the effects of habitat quality, management and alike.

In a context of habitat fragmentation, plant populations decrease their size and increase their degree of isolation. From a genetic point of view, genetic diversity of fragmented plant populations is likely to decrease over time. Demographic and genetic mechanisms explain such a decrease in genetic diversity. Small populations with a low effective reproductive size (i.e. the number of reproductive individuals) exhibit high inbreeding and biparental inbreeding that causes inbreeding depression (i.e. the reduction in fitness of self compared to the outcrossed progeny). Inbreeding depression reduces plant fitness components, such as survivorship, growth and reproduction, enhancing therefore population extinction. Isolation increases the distance between populations, decreasing therefore the probability of gene flow between populations in a given area. Gene flow is effective in preventing inbreeding depression as increases population outcrossing rates. The implications of genetic diversity loss are manifold. For example, inbreeding increases homozygosity of extant individuals and inbreeding depression can eliminate several genotypes from the population. The loss of genetic diversity reduces the genetic potential of plants to adapt to new environmental conditions. In fact, the process of local adaptation to local population characteristics can be strongly enhanced by a situation of high inbreeding and no gene flow. Hence, if plants perform well only in their origin population, colonization of suitable habitats might not be successful. In any case, either progressive fitness decline or excessive adaptation to local conditions can significantly increase extinction probabilities of plant populations in a fragmented landscape. Genetic diversity represents the base for morphological trait variation on which natural selection acts. If genetic diversity reduces over time due to the factors mentioned above, the possibility to develop well-adapted phenotypes to changing environmental conditions is also reduced. The use of low variable molecular markers can be used to examine the genetic origin of plant populations at the regional scale. In the case of Central and Northern Europe, the current distribution pattern of plant species depends on the interglacial colonization of plant species from glacial refugia in Southern Europe. For this reason, other studies that addressed this question also found low levels of large-scale genetic variation in Northern latitudes. Chloroplast DNA (cpDNA hereafter) is commonly used in evolutionary and phylogenetic studies given that cpDNA evolves slowly and most cpDNA polymorphisms are thought to be caused by structural rearrangement or mutation. However, several other studies have used cpDNA to assess the large-scale genetic relationship among and between populations of plant species across a considerable part of their distribution area. Genetic variation of several study species (i.e. Succisa pratensis, Hypochaeris radicata, Tragopogon pratensis, Scabiosa columbaria, Pimpinella saxifraga, Ranunculus bulbosus and Carlina vulgaris) was analyzed using all cpDNA markers available (around 15 cpDNA markers) and the method of polymerase chain reaction - restriction fragment length polymorphism (PCRRFLP). Results showed that, except Pimpinella saxifraga, the rest of plant species exhibited no variation in cpDNA markers. Hence, results suggest that several plant species in Europe could share a common ancestor. In the case of Pimpinella saxifraga, the pattern of variation found (variation within and between regions) could be interpretated as (1) different origins since the last interglacial colonization, or (2) that this species is prone to variation in cpDNA. There are studies supporting each one of these possibilities. The population-level genetic variation requires more specific genetic markers, such as the microsatellites developed for Hypochaeris radicata.

The objective of this task was to provide a basis for analyses of different aspects of landscape configuration on species distribution and diversity. A GIS model is constructed as different layers of information each bearing a specific content, where the information is spatially explicit. The information layers were species distributions, vegetation, geomorphology and land use. Species distributions were based on field surveys where each of the four target species, Succisa pratensis, Tragopogon pratense, Ranunculus bulbosus and Carlina vulgaris were mapped in the study area, Nynäs nature reserve, southern Sweden. The maps of the target species where imported to a GIS-framework using the software Arcview. Present day vegetation was inferred from aerial photographs, and the classification was verified in the field. Geomorphology consisted of topography, soil and bedrock, and was imported from existing geomorphological maps. Land use history was extracted from cadastral maps dating back to the 17th century and from aerial photos from the 1940s and onwards. The history of the landscape includes a general decline in traditionally managed grasslands, which has either developed into forest or been used for agricultural fields. During the last decades many fields have again been transformed to pastures, which implies that there are different categories of grasslands, traditionally managed (un-fertilized and not ploughed) and recent grasslands, i.e. former fields. This land use change has severe effects on the distribution and dynamics of species, and these effects have been subject to both descriptive and analytical studies; the latter using population models. Species response to land use change is basically a result of changing amount of suitable habitats in the landscape, and the actual configuration of habitats in the landscape. Species may however, due to time lags in either colonization or extinction processes, have a distribution that deviates from what would be expected from the landscape structure. Colonization time lags are due to dispersal limitation, i.e. species are not able to colonize all available suitable sites. Extinction time lags are due to persistent life-cycle stages, for example dormant seeds, clonal propagules or long-lived vegetative plants. Descriptive analyses of species distributions in relation to habitat characteristics suggest that the target species mainly occur in traditionally managed sites or in sites that have been managed, mainly by grazing, over the last 50 years. For one species, Succisa pratensis, the distribution pattern today reflects the distribution of semi-natural grasslands in the 1940s, whereas the occurrences of the species in relation to present vegetation is approximately equally distributed among grasslands and forest. This implies that there is likely a time lag in the response of this grassland species to ongoing land use change. Such a time lag has also been supported by simulation models performed in collaboration with work package 4 (Tomás Herben, University of Prague), which are further described in the summary of the task "Analysis of landscape configuration". There are several implications of these results. For assessment and monitoring of plant biodiversity in the landscape, it is essential that the non-equilibrium state of species occurrence be acknowledged, otherwise surveys may be misleading, and conservation programmes may fail to reach their goals. Moreover, the results suggest that conservation of biodiversity must account for landscape level processes and not only focus on target sites, "biodiversity hotspots". The results suggest that land use history must be much more in focus of spatial analyses of plant biodiversity, and that explicit knowledge of historical change provide an indirect assessment of the time-scale of the response of plant species to land use change. For agriculture the results imply that it would be favourable to preserve traditional management, e.g. grazing on all still existing semi-natural grasslands, but also that the surrounding landscape should be used for extensive grazing, to promote dispersal of species and development of new populations. Conservation programmes should increase focus on whole landscapes. Economically sustainable systems for maintaining grazing, and for cooperation among farmers in order to increase the extent of grazed areas, should be developed. Such landscapes will also promote recreational values of landscapes. The results will also be valuable for research on regional plant population dynamics, as there are still relatively few analyses combining demography, regional dynamics and land use history. We foresee that there is also a value for education, both for schools and for higher education at universities. The interface between ecology, history and cultural geography is also enhanced by the results of these studies.

In the fragmented landscapes of Europe today plants occur more and more in small and isolated populations. The survival of endangered species in this landscape configuration depends on the persistence of the local populations and the ability to colonise vacant habitat patches. Much research attention over the past years has been devoted to dispersal and colonisation and on the issue of how to define a vacant site. Much less time and effort has been spent on the seed source strength, the number of seeds that remaining populations produce. Insight in the factors that determine local seed production will improve our understanding of the functioning of metapopulation processes for plants. We do know that local populations may begin to deteriorate if local conditions change due to succession, climate change or external impacts such as eutrophication. Malfunctioning begins when local seed recruitment comes to a halt because vegetations get higher and denser and small-scale disturbances no longer open the vegetation. This generally reduces population viability because seed germination of many species is light dependent. However, mature perennial individuals may persist for a long time under these conditions. Unknown until now is whether these individuals remain flowering and set seed and thus remain serving as a significant seed source within the metapopulation structure. Based on the measurements described under demography we have made seed production measurements of the populations that we studied. The calculations are based on detailed monitoring of numbers of flowers per plant and the seed set. One of our species, Tragopogon pratensis, may serve as an example. Germination experiments indicated that seeds of this species are unable to germinate under light conditions that simulate canopy shading. In more productive habitats with very few open patches, seedling recruitment will therefore be negligible. Mature plants of Tragopogon, however, may grow and flower profusely and have an abundant seed set in such environments. Indeed, we have monitored relatively small populations that produce tens of thousands of seeds per year, albeit with very little chances for local recruitment. We conclude that there is a high potential for small populations as a seed source in the landscape, even when local conditions reduce population viability due to reduced establishment of sexual offspring.

Searching for OpenAIRE data...

Exploitable results

Share this page

Download