Production,nutrient availability,and elemental balances of two meadows affected by different fertilization and water table regimes in The Netherlands

The restoration of degraded peat-grasslands is an important nature conservation goal in The Netherlands. We investigated the effects of ceased fertilization (15 years) combined with a groundwater-raised water table (6 years) on the production of the peat-grassland vegetation and soil nutrient availability in a meadow. Furthermore, we evaluated whether and how this difference between meadows affected the balances between nutrient inputs and outputs in the ecosystem. We used an adjacent fertilized meadow in which the water table followed agricultural practice as a control. Yield of the grassland vegetation was significantly lower in the wet than in the control meadow. The tissue concentrations of N, P, and K in the harvested vegetation were significantly lower, but those of Ca higher in the wet than in the control meadow. The difference between both meadows significantly affected the annual nitrification rate, but not the annual C and N mineralization rates and the annual net P and K release rates. The difference between both meadows also significantly affected the seasonal nitrification and K release rates. Season exerted a significant effect on the seasonal C and N mineralization and nitrification rates. The elemental balances and relative contributions of the balance terms to elemental inputs and outputs varied considerably with element. Annually, the wet meadow lost N, P and K, while the control meadow gained these elements. The elemental demand of the grassland vegetation in the wet meadow was met for N for a large part by mineralization and for the remainder by atmospheric deposition, for P it was in the same order of magnitude as the net soil-P release, as it was for K. It is to be expected that the soil resources of N, P and K will continue to decrease under a continued regime of ceased fertilization and a raised water table, with those of N decreasing with the same rate, of P morerapidly and of K more slowly than estimated from regressions.