Questions: Could the seed bank increase biodiversity during restoration of abandoned, species-poor, formerly cultivated vegetation? Is it possible to identify how climate, soil and former and present management and vegetation affected the seed bank?

Location: The study sites were eight abandoned grasslands, four in Orkdal, central Norway and four in Gaular, western Norway.

Methods: 144 seed bank samples were collected from three depths. Each sample was sown and placed in a greenhouse. After three months, the trays were dried and stored at 4°C in a dry place for two months. This was repeated twice.

Results: There was a separation of the two regions along the first DCA axis in both the seed bank and in the vegetation analysis and also a clear separation of the seed bank from the vegetation along the second axis. These results are caused by differences in former management as well as temperature, precipitation and soil type between Gaular and Orkdal. We found more annuals, short-lived species and species demanding light open conditions in the seed bank than in the vegetation probably because these species have the capacity for producing persistent seeds. Most of the species found only in the seed bank were found in very few samples and with few individuals.

Conclusion: These results suggest that it may be difficult to increase vegetation biodiversity through restoration of grasslands such as those investigated if the natural soil seed bank is the main seed source.

Nomenclature: Lid & Lid (1994).