Beatrijs Bossuyt

Can the seed bank be used for ecological restoration? An overview of seed bank characteristics in European communities

Abstract Question: Can seeds in the seed bank be considered as a potential source of material for the restoration of European plant communities including forest, marsh, grassland and heathland? Methods: This study reviews seed bank studies (1990–2006) to determine if they provide useful and reliable results to predict restoration success. We formally selected 102 seed bank studies and analyzed differences between four plant community types in several seed bank characteristics, such as seed density, species richness and similarity between seed bank and vegetation. We also assessed the dominant genera present in the seed bank in each plant community. Results: We observed remarkably consistent trends when comparing seed bank characteristics among community types. Seed density was lowest for grassland and forest communities and highest in marshes, whereas species richness, diversity and evenness of the seed bank community was lowest in heathland and highest in grassland. Similarity between seed bank and vegetation was low in forest, and high in grassland. There was a lot of overlap of the dominant genera of seed bank communities in all studies. Conclusions: The absence of target species and the high dominance of early successional species, in particular Juncus spp., indicate that restoration of target plant communities relying only on seed germination from the seed bank is in most cases not feasible. The exceptions are heathland and early successional plant communities occurring after temporally recurring disturbances. Restoration of plant communities composed of late successional species, such as woody species or herbaceous species typical of woodland or forest rely mainly on seed dispersal and not on in situ germination.

Ecological perspectives for the restoration of plant communities in European temperate forests

Simultaneously with increasing afforestation efforts in western Europe, among conservationists the consciousness is growing that protecting areas to conserve biodiversity will not be sufficient in the long term, and that also the ecological restoration of more or less severely altered areas will be necessary. The probability that recently established forest stands develop towards their ecological reference (i.e. ancient forest) depends largely on the possibility of the target species to colonize them. We focused on the colonization ability of forest plant species and particularly on so-called ancient forest plant species. Major constraints for ecological forest restoration are the spatial characteristics of the target site (isolation, shape and area), imposing dispersal limitations, and in the duration and intensity of the historical land use, leading to changes in habitat characteristics influencing recruitment probability. We reviewed the ecological literature with respect to these constraints and conclude that it takes at least a century to restore the understorey layer of recent forests, even when the target stand is adjacent to a well-developed ancient forest. Both recruitment and dispersal limitation of the target species are responsible for this. Newly established forests should therefore be situated at a minimal distance of the ancient forest source. In other cases, forest plant species will not be able to colonize the newly established forest on a measurable time scale and artificial introduction of forest plant species can be taken into consideration. The negative effects of habitat characteristics, and mainly high soil nutrient values in the recent forest stand can be mitigated by soil nutrient lowering measures. Disturbances in the recent forest should be minimized to maintain a high canopy closure level, preventing light demanding, high competitive species from establishing a stable population. An additional negative consequence of soil disturbances is that it stimulates germination of species from the soil seed bank, which is mainly composed of highly competitive or ruderal species.

Seed bank and vegetation composition of forest stands of varying age in central Belgium: consequences for regeneration of ancient forest vegetation

Vegetation composition differs significantly between ancientand recent forest, due to slow colonization capacity of typical forest speciesand the higher abundance of early successional species in recent forest.However, little is known about differences in persistent seed bank compositionbetween ancient and recent forest and about the interaction between seed bankand vegetation in relation with forest age. We surveyed the seed bank and theunderstorey vegetationcomposition in transects from ancient to recent forest. Seed bank and fieldlayer vegetation characteristics and similarity between seed bank andvegetationwere analysed in relation to recent forest age and distance to the ancientforest. A total of 39 species and 14,911 seedlings germinated, whichcorresponds with a seed density of 12,426 seeds/m2.Total seed density is significantly higher in the youngest recent forest parcel(55 years). Also the seed bank composition in the youngest forest parceldifferssignificantly from the other parcels. After a longer period of reforestation,the seed bank approaches that of the ancient forest, suggesting seed bankdepletion, although the seed bank is permanently replenished to some extent byseed bank forming species from local disturbances. Seed bank composition doesnot change significantly with distance to the ancient forest. Similaritybetween seed bank and vegetation composition, nomatter the forest age, is very low, but decreases with increasing forest age.The most frequent species in the vegetation are absent in the seed bank andvice versa. The contribution of forest species is highin the vegetation and they almost not occur in the seed bank, while species offorest edges and clearings, and species of disturbed environments are morefrequent in the seed bank. The seed bank is mainly composed of earlysuccessional species of former forest stages or species which temporary occurinsmall-scale disturbances. The seed bank may enhance the negative effects ofearlysuccessional, mainly competitive species to the forest species richness in therecent forest. In this respect, forest management should minimise forestdisturbances, to prevent germination of competitive species form buriedseeds.

Forest biodiversity : lessons from history for conservation

This book focuses on the diverse impact of forest history in general, and of forest continuity, fragmentation and past management in particular, on the diversity and distribution of species. The implications for the conservation of biodiversity in forests are also addressed. Chapters have been developed from papers presented at a conference held in Leuven in January 2003. The emphasis is on temperate forests in Europe and North America, but the information may also be applicable to other regions or biomes. The book will be of significant interest to researchers working within the areas of forestry, ecology, conservation and environmental history.

Uncertainties in the Environmental Risk Assessment of Metals

As life has evolved in the presence of metals, the assessment of the potential adverse effects of metals on ecosystems requires a different approach than those presently used for man-made organic substances. This article provides a brief review of applications and limitations of current techniques and presents, based on recent research results, suggestions for improving the scientific relevance and accuracy of environmental risk assessments of metals. The importance of the following factors responsible for major uncertainties in current environmental risk assessments of metals are discussed: factors affecting metal bioavailability and toxicity, the potential importance of deficiency effects (for essential metals), and field extrapolation of laboratory toxicity data. Possible (regulatory) consequences of inaccurately assessing the natural background concentrations of metals and acclimatization/adaptation potential of laboratory organisms and resident communities are illustrated using examples of recent research, hypothesis development, and a probabilistic environmental risk assessment.

Inflow of seeds through the forest edge: evidence from seed bank and vegetation patterns

To determine the influence of the proximity of a forest edge on seed bank composition and diversity, we performed a seed bank sampling at ancient deciduous forests bordering intensive arable fields. Also vegetation patterns were taken into account. We hypothised that forest edges may facilitate the entrance of diaspores of invasive species into the forest and the successive incorporation of these species in the forest seed bank. We noticed a substantial influence of the proximity of an edge on seed bank composition at as well the forested side of the edge as the field side. The forest edge zone was limited to 3 m into the forest and the field edge zone extended 3m into the field. The seed bank samples of field and forest edge are characterised by a higher species diversity and seed density and a higher similarity between seed bank and vegetation, compared to field or forest samples. The forest edges contains fewer pioneer species in comparison with the forest interior and more competitive species and species of edges and clearings compared with field and forest samples. The seed longevity index increases towards the forest interior. We can conclude from our data that the forest and edge seed bank are composed by both seeds from recent dispersal processes and local seed set and by seeds originating from past vegetation on the site. Near the edge, actual seed input seems of primal importance. Further towards the forest interior seed input decreases and long-living seeds of past vegetation become more important. Ancient forest edges thus act as a barrier for seeds of species of the surrounding arable field.

Seed bank assembly follows vegetation succession in dune slacks

Abstract Question: Is the seed bank in dune slacks during the whole successional range mainly composed of early successional species or does it vary according to the changing vegetation? Location: Belgium, western part of the coast. Methods: We investigated the soil seed bank composition using a seedling germination method in a chronosequence of 20 dune slacks, ranging in age from five to 55 yr. Results: Both seed density and species richness in the seed bank were very low in the first successional stages and increased with age, mainly as a result of increasing seed production. The similarity between seed bank and vegetation was higher in older slacks. A comparison of characteristics between seed bank and vegetation showed that the seed bank was, to a large extent, composed of later successional species. Occurrence patterns of individual species also showed that seeds become incorporated in the soil after the species has established in the vegetation. Conclusion: The seed bank is not likely to be the driving force for successional changes in the vegetation, and successional changes rely on dispersal. Some early successional species persist in the seed bank, but generally only in low numbers. The results also confirm that most typical dune slack species do not form persistent seeds, so that re-establishment from the seed bank is not to be expected when the species has disappeared from the vegetation.

Restoration of the understorey layer of recent forest bordering ancient forest

Recently established forests are commonly char- acterized by an impoverished understorey. Restoration is mostly based on spontaneous secondary succession, but little is known about the time period needed to achieve a community species pool with species composition equal to that of ancient forests. Vegetation in transects of 197 plots in 13 recent forest stands contiguous to the Meerdaalwoud ancient forest com- plex was surveyed. The recent forest stands ranged in age from 36 to 132 yr. The community species pool was described with an ecological, functional and phytosociological approach and based on groups derived from a CCA. Differences in commu- nity species pool between age classes of recent forest stands were analysed. During establishment of a new forest competitive species, forest edge species and species with high Ellenberg values for light and nitrogen and a more persistent seed bank will domi- nate the understorey. After 90 yr of succession the cover by these species decreases and reaches equal values to ancient forest after ca. 105 yr. A large number of forest species will be able to colonize the forest in less than 90 yr. Some typical forest species, however, have very low colonization rates and still have low cover in recent forest more than 105 yr old, so that complete restoration of the understorey requires a time period of over a century. Anthropogenic introduction of forest plant species may reduce the time required for ancient forest vegetation equality.

An island biogeographical view of the successional pathway in wet dune slacks

Abstract Wet dune slacks occur as small, naturally fragmented systems in the dune landscape, isolated from other slacks. We studied the effects of slack isolation and area on the rate and direction of primary succession in a chronosequence of dune slacks. The results indicate that important changes occur in community characteristics over a period of 50 yr. Total cover and number of species increase as a result of the endogenous succession process, during which organic matter and nutrients accumulate. Consequently, competitive interactions shift from competition for nutrients to competition for light. Local factors thus determine, at least partially, the community composition in the slack. However, differences in community composition with increasing age are smaller when the slack is more isolated or smaller, suggesting a slower succession rate and biomass accumulation. Together with a lower contribution of slow dispersing species in more isolated slacks, this indicates that species accumulation is dispersal limited and thus influenced by regional factors. The stochastic variation resulting from this dispersal limited species accumulation causes a divergent successional pathway. Nomenclature: Lambinon et al. (1998). Abbreviation: MIV = Mean indicator value.

Interactions between plant life span, seed dispersal capacity and fecundity determine metapopulation viability in a dynamic landscape

Classical metapopulation models do not account for temporal changes in the suitability of habitat patches. In reality, however, the carrying capacity of most habitat types is not constant in time due to natural succession processes. In this study, we modeled plant metapopulation persistence in a successional landscape with disappearing and emerging habitat patches, based on a realistic dune slack landscape at the Belgian–French coast. We focused on the effects of the variation of different plant traits on metapopulation persistence in this changing landscape. Therefore, we used a stage based stochastic metapopulation model implemented in RAMAS/Metapop, simulating a large variation in plant traits but keeping landscape characteristics such as patch turnover rate and patch lifespan constant. The results confirm the conclusions of earlier modeling work that seed dispersal distance and seed emigration rate both have an important effect on metapopulation persistence. We also found that high population growth rate or high recruitment considerably decreased the extinction risk of the metapopulation. Additionally, a long plant life span had a strong positive effect on metapopulation persistence, irrespective of the plants dispersal capacity and population growth rate. Plant species that invest in life span require less investment in offspring and dispersal capacity to avoid extinction, even in dynamic landscapes with deterministic changes in habitat quality. Moreover, metapopulations of long-lived plant species were found to be much less sensitive to high levels of environmental stochasticity than short-lived species.