Soil seed banks in European grassland: does land use affect regeneration perspectives?

Summary 1. Soil seed banks of 38 grassland sites in western Europe were sampled and species composition was determined. All sites differed in geographical region, soil type and intensity of management.
2. Relationships between the composition of the seed bank, the established vegetation, and soil and management parameters were analysed using different ordination techniques.
3. Common species in the seed bank of extensively managed grasslands were Juncus without septa, Plantago lanceolata, Holcus lanatus and Cerastiumfontanum. General
differences in the species composition reflected the country of origin of the sites.
4. Differences between sites within a chronosequence were used to interpret the impact of management on the proportion of species in the seed bank representing different
functional groups. Seed banks were generally dominated by species of mid-range nutrient conditions. Species associated with poor nutrient conditions were relatively
scarce at most sites.
5. Intensive agricultural management had a negative effect on the soil seed bank of grassland species, whereas lack of management or abandonment resulted in the loss
of grassland species from both the vegetation and the soil seed bank owing to the invasion of woodland species. Restoration management, following long periods of
intensive use, took more than 20 years to establish a significant increase in species of poor nutrient conditions.
6. A regeneration index was calculated for each chronosequence to estimate the relative 'distance' between a degraded site and a species-rich target site in the same
series. These regeneration indices provide an indication of the timescales required for the restoration of sites towards the species-rich target grassland community types.
However, for most agriculturally improved grassland sites, the study indicated that restoration of unimproved or semi-natural types of grassland cannot depend on a
high level of recruitment from the seed bank of species which have disappeared from the established vegetation. This implies that maintenance of existing species-rich
grassland resources should be given high priority.
7. The results of this seed bank study show that grassland sites with a short history of agricultural improvement or biodiversity degradation are likely to be the most
worthwhile for restoration purposes.