Resistance and Resilience of Species Composition: Thirty Years of Experimental Mismanagement and Subsequent Restoration in a Species Rich Meadow

Traditionally managed grasslands are among the most species-rich communities, which are threatened by land use changes-management intensification or abandonment. The resistance of their species composition to mismanagement and ability to recover after re-establishment of traditional management is of prime conservational interest. In a manipulative experiment in a wet meadow, we simulated mismanagement by a factorial combination of abandonment of mowing and fertilization. The dominant species <i>Molinia caerulea</i> was removed in half of the plots to assess its role in community dynamics. The 21 years' mismanagement period was followed by the re-establishment of the traditional management. The plots were sampled yearly from 1994 (the baseline data, before the introduction of the experimental treatments), until 2023. Estimates of cover of all vascular plant species provided the species richness and effective number of species. For each year, the chord distances to baseline species composition and to corresponding control plot were calculated. The compositional data were analyzed by constrained ordination methods, and the univariate characteristics by Repeated Measures ANOVA. All the plots, including those with traditional management throughout the whole experiment, underwent directional changes, probably caused by a decrease in groundwater level due to global warming. Both fertilization and abandonment led to a loss of competitively weak, usually low-statured species, due to increased asymmetric competition for light. The effect of fertilization was faster and stronger than that of abandonment demonstrating weaker resistance to fertilization. The removal of dominant species partially mitigated negative effects only in unmown, non-fertilized plots. The recovery following mismanagement cessation was faster (signifying higher resilience) in unmown than in fertilized plots, where it was slowed by a legacy of fertilization. In a changing world, two reference plot types are recommended for assessment of resistance and resilience, one original state and one reflecting compositional changes independent of treatments.