This project was an effort to restore species-rich flood-plain meadows from abandoned arable fields in the valley of the river Meuse, in northeast France. The project area was sown with a commercial seed mixture of Phleum pratense, Festuca pratensis, Lolium perenne and Trifolium repens. Before restoration, the above-ground fallow vegetation was dominated by ruderal and annual species, with only a few meadow species present. Sowing led to tall, dense vegetation, mainly dominated by sown species. Ruderal and annual species had decreased three years after the restoration, but target species still remained underrepresented. Species richness was significantly lower in the sown site than in the semi-natural target meadows and the vegetation had a different composition. Further analysis of the soil seed bank of the restored soil seed bank had a different composition: one of only a few meadow species, but dominated by a few ruderal species. Three years of the restoration project and the site is moving toward the target communities but impoverished seed sources limit the success of the restoration.
Meuse, France, 49.0824319, 5.282399700000042
Country or Territory:
Freshwater Ponds & Lakes, Grasslands & Savannas - Temperate
Area being restored:
University / Academic Institution
Primary Causes of DegradationAgriculture & Livestock, Fragmentation
While previously used for low intensity hay-making, these flood-plain meadows were, over the last fifty years, converted into more intensively used fields. They were drained, protected against flooding, fertilized more, cut earlier and more frequently and often converted into arable land. The combination of high productivity lands (a consequence of increase nutrient availability) and an absence of viable propagules of characteristic species appear to cause bottlenecks for restoration. Part of the reason for this is that conversion to arable fields that are tilled, plowed, and grow crops result in the depletion of the soil seed bank, as well a complete modification of the chemical composition of the soil. The restoration site was tilled for maize cultivation from 1988 to 1993, but was flooded too often to be suitable for regular cultivation. The area was abandoned for 5 years and recolonized naturally by species such as Cirsium arvense and Urtica dioica.
Reference Ecosystem Description
Flood-plain meadows are ecosystems are heavily influenced by spring and winter flooding. Typically, they have high productivity with a substantial exchange of matter and energy with the river, and may support typical, species-rich plant communities. These flood-plain meadows were flooded frequently and were used for hay-making under a low-intensity management regime. In the semi-natural meadows of the Meuse, three associations were distinguished along the wetness gradient: Colchico-Festucetum pratensis mesophilic meadow, occurring in higher areas of the valley. This is rarely flooded. In contains mesoxerophilic (Primula veris, Avenula pubescens) and more mesophilic or meso-hygrophilic species (Peucedanum carvifolia, Galium verum, Ranunculus acris).; Senecioni-Oenanthetum mediae mesohygrophilic meadow, situated at intermediate topographic positions, where the floods last longer and occur more frequently. This is a meadow community dominated by grasses (Phleum pratense, Festuca pratensis, Anthoxanthum odoratum) but with mesohygrophilic and mesophilic species (Glechoma hederacea, Galium palustre, Achillea ptarmica, Centaurea jacea, Senecio aquaticus).; Gratiolo-Oenanthetum fistulosae hygrophilic meadow, located in the lower areas of the valley. This type is flooded most frequently and for long periods and is characterized by some mesophilic and mesohygrophilic meadow species, mixed with heleophytes (Oenanthe fistulosa, Eleocharis palustris). Based on different topographic levels within the site, which control both flooding frequency and duration, all three communities could potentially occur in the restored meadow.
The goals of the project were: 1) to analyze vegetation changes after sowing; 2) to compare the vegetation after sowing with that before sowing and with the composition of characteristic species-rich target meadows; 3) to estimate the contribution of the soil seed bank in the restoration of species-rich flood-plain meadows.
The project does not have a monitoring plan.
This project was conducted as part of a large scale practical restoration scheme to discuss the interest of restoration attempts realized in France. It was decided that such a project would have to be low-impact and could not cost a lot of money as one element of the project was to keep the farmers managing the land. The sown species were chosen because they are present in flood-plain meadows of the Meuse and allow for a sufficiently high production that produces high-quality hay, even with late cutting dates. Festuca pratensis and Lolium perenne establish rapidly and are highly productive, while Trifolium repens is low and provides structural heterogeneity to the vegetation.
Description of Project Activities:
In 1998 the restoration project got under way, when the above-ground vegetation was mown and plowed and then in autumn the site was sown with the commercial seed mixture of Phleum pratense, Lolium perenne, Festuca pratensis, and Trifolium repens. The initial sowing density was increased by the farming, rendering the exact density unknown, however it is estimated to have been 30-35 kg-seed/ha. Once sowed the vegetation was cut twice a year (late June, late August) and fertilization was no longer allowed. Soil samples were collected from target and experimental meadows (before and after sowing) to assess the chemical composition of the site. Nine permanent plots of 4 sq m were established in 1998 before sowing, then in 1999, 12 plots were added. All 21 plots were surveyed during 3 years following sowing. Vegetation dynamics were monitored on these plots with species frequency assessed from its frequency in 25 hand-fulls of vegetation. To compare changes in the floristic composition over the four year project, the species were arranged into five groups: sown-mixture species, annual species, ruderal species, meadow grasses and other meadow species. In June 2001, phytosociological releves of 25 sq m adjacent to the 21 plots were done in order to allow for comparison with the vegetation of target meadows. Ten releves in each of the three grassland communities present in the Meuse valley. These grasslands were semi-natural meadows where agricultural management had been extensive and stable for at least ten years. Soil seed bank composition during the fallow phase was analyzed at 11 sites, with ten replicates taken from each study site from the top 10 cm with a 4-cm diameter drill. Each sample represented a soil volume of 1256 cubic cm. Those samples were taken in April 1998, stored, cleaned, and sieved, and then put in soil in a greenhouse. Watered everyday, the seedlings that emerged were identified, counted and remove. Unidentifiable seedlings were transplanted until they flowered and could be identified.
Ecological Outcomes Achieved
Eliminate existing threats to the ecosystem:
The above-ground vegetation at the restoration site were tall, dense, and dominated by the sown species after three years of restoration. The ecological species groups clearly reflected the different successional stages of the area, i.e. first grassland, then cultivation, next abandonment and finally sowing. Sown species took the place of ruderal species, such as Cirsium arvense and Epilobium sp., which dominated the vegetation during the period of abandonment. Re-establishment of target meadow species, such as Plantago lanceolate and Leucanthemum vulgare, occurred in only a few plots. Only the most common meadow grasses were present and abundant, especially Poa trivialis and Elymus repens, probably due to their high vegetative reproduction. The plowing and sowing initially led to a reduction in meadow species, but these species quickly reappeared in the above-ground vegetation. These disturbances also led to an increase in frequency of annual species and the maintenance of ruderal species. Both grouns of species decreased slowly with time, the cutting regime and the competition of meadow species. The overall floristic differences between 2000 and 2001 were small, indicating that vegetation dynamics were then slower than at the beginning of the restoration It was also clear that an abandonment stage of 5 years after cultivation has a positive impact on species richness and on the number of established species. The mean species richness was significantly lower in the sown meadow that in the target communities and the floristic compositions were different. Three years after sowing, common meadow species were present and well-represented, but target species were scarce and meso-oligotrophic species such as Anthoxanthum odoratum and Briza media were completely absent. Although the majority of characteristic species did not colonize the site, the presence of some does suggest a development towards the target communities. The similarity between restored and target meadows increased significantly with time, which suggests that the restoration of the sown meadow will proceed in the direction of the target meadow. Ultimately, further success of the restoration operations will depend on seed sources.
Factors limiting recovery of the ecosystem:
It may have been possible to recreate a meadow directly from the cultivation stage by mowing the established vegetation without plowing and sowing. Additionally, the floristic composition of the cultivation fallow was dominated by ruderal species but meadow species were equally present, however, sowing was considered essential in order to provide quality hay so that farmers would be willing to manage the area. The results of the soil seed bank also indicated limits to this type of restoration as it was dominated by only a few species, while many of the target meadow species were not found. This indicated that seed mortality is directly related to cultivation period. It is likely that the soil seed bank is partly destroyed during cultivation, probably due to physical damage with agricultural machines, through chemical destruction by pesticides, and my burial of seeds by plowing. The distance from target communities is a distinct limitation to the success of the restoration. The window of a three year project was insufficient to gauge the effects of the target communities, as they were unable over that length of time to cover the distance of several hundred meters between species-rich source areas and restored sites. Without seeds in the seed bank, they need to disperse to the restored area. The limitations of seed dispersal indicates that it is a key limitation in the restoration of species richness, unless the site is surrounded by species-rich meadows.
Socio-Economic & Community Outcomes Achieved
Economic vitality and local livelihoods:
The project illustrated that such efforts to restore flood-plain meadows can actually restore hay meadows, although it is limited in restoring species-richness. The use of sown commercial seed mixes provided farmers with a means for restoring their previous source of hay, and was a means for altering the ecological trajectory of the system in an alternative direction.
Vecrin, M.P., R. van Diggelen, F. Grevilliot, and S. Muller. 2002. Restoration of species-rich flood-plain meadows from abandoned arable fields in NE France. Applied Vegetation Science, 5: 263-270.