3.2. Effects of mowing and abandonment on grassland communities

Continuous management strongly affected species composition in the Keminmaa meadow patches (II). The plant communities under continuous management consisted of species able to tolerate regular disturbance, in this case mowing. In contrast, the abandoned habitats around the Keminmaa churches included species with variable life-history strategies. The species composition was more variable in unmanaged than in managed meadows, because the managed sites were colonized by a small group of species with similar growth strategies. According to diversity indices, the species diversity was higher in abandoned meadow patches than in managed ones. However, the evenness values of abandoned and managed meadows suggested that the species dominance patterns were quite similar under both management practices.

The proportions of species origin, growth forms and life-history strategies varied between the habitats. Perhaps the most marked difference was seen in the comparison of two plant strategy theories, namely MacArthur-Fretwell’s scheme and Grime’s (1977, 1979) triangular strategy (CSR) theory. The proportion of grazing-tolerant plants (see Oksanen & Ranta 1992) was found to be larger in managed lawns than in abandoned meadows. Correspondingly, more K strategists inhabited abandoned meadows. Species representing the S and SR strategies (cf. Grime et al. 1988) did well on managed sites. The average cover value of ruderal species, which require open patches for regeneration, was low. At mown sites the ground layer was occupied by a thick cover of mosses, whereas at abandoned sites the closed field layer canopy did not allow ruderal species to regenerate.

The six-year study in Kiiminki (III) suggested that mowing late in the season causes only a few changes in grass-dominated meadows. The results showed that the average field cover decreased during the study period, and that the change was parallel in both mown and unmown plots of the meadow. According to matrix simulations, the cover of small herbs is expected to decrease in unmown plots and increase in mown plots, provided that the transition probabilities remain unchanged during the course of succession. Tall herbs are predicted to increase in both unmown and mown plots. During the simulation period of 30 years, the cover of grasses first declined and then remained unchanged in both unmown and mown plots. In all cases, the asymptotic stable distributions of cover classes corresponded well to the actual distributions observed in 1998.