Questions: How is succession on ex-arable land affected by sowing high and low diversity mixtures of grassland species as compared to natural succession? How long do effects persist?

Location: Experimental plots installed in the Czech Republic, The Netherlands, Spain, Sweden and the United Kingdom.

Methods: The experiment was established on ex-arable land, with five blocks, each containing three 10 m × 10 m experimental plots: natural colonization, a low- (four species) and high-diversity (15 species) seed mixture. Species composition and biomass was followed for eight years.

Results: The sown plants considerably affected the whole successional pathway and the effects persisted during the whole eight year period. Whilst the proportion of sown species (characterized by their cover) increased during the study period, the number of sown species started to decrease from the third season onwards. Sowing caused suppression of natural colonizing species, and the sown plots had more biomass. These effects were on average larger in the high diversity mixtures. However, the low diversity replicate sown with the mixture that produced the largest biomass or largest suppression of natural colonizers fell within the range recorded at the five replicates of the high diversity plots. The natural colonization plots usually had the highest total species richness and lowest productivity at the end of the observation period.

Conclusions: The effect of sowing demonstrated dispersal limitation as a factor controlling the rate of early secondary succession. Diversity was important primarily for its ‘insurance effect’: the high diversity mixtures were always able to compensate for the failure of some species.

Abbreviations; ED = Euclidian distance; HD = High diversity; LD = Low diversity; NC = Natural colonization