Question: Is raising groundwater tables successful as a wetland restoration strategy?

Location: Kennemer dunes, The Netherlands; Moksloot dunes, The Netherlands and Bullock Creek fen, New Zealand.

Methods: Generalizations were made by analysing soil dynamics and the responsiveness of integrative plant traits on moisture, nutrient regime and seed dispersal in three case studies of rewetted vs. control wetlands with the same actual groundwater levels. Soil conditions included mineral (calcareous and non-calcareous) soils with no initial vegetation, mineral soils with established vegetation and organic soils with vegetation.

Results: The responsiveness of traits to raised groundwater tables was related to soil type and vegetation presence and depended on actual groundwater levels. In the moist-wet zone, oligotrophic species, ‘drier’ species with higher seed longevity occupied gaps created by vegetation dieback on rewetting. The other rewetted zones still reflected trait values of the vegetation prevalent prior to rewetting with fewer adaptations to wet conditions, increased nutrient richness and higher seed longevity. Moreover, ‘eutrophic’ and ‘drier’ species increased at rewetted sites, so that these restored sites became dissimilar to control wetlands.

Conclusions: The prevalent traits of the restored wetlands do not coincide with traits belonging to generally targeted plant species of wetland restoration. Long-term observations in restored and control wetlands with different groundwater regimes are needed to determine whether target plant species eventually revegetate restored wetlands.

Abbreviations: F = Mean moisture indicator values, lnRR = Natural log of response ratio, N = Nutrient indicator value, PCA = Principal Component Analysis, RDA = Redundancy analysis