| Summary: | Over the last thirty years the intensification of agricultural practices by drainage and conversion of grass lands to arable has led to the loss of hydrological and ecological functions of wetlands in the western marshes of France, in particular the conservation of animais characteristic of these wetlands. The goal of this study was to res tore at !east sorne of the ecological functions of wet grasslands by experimentally manipulating the water regime and grazing; the consequences for the plant and bird communities were measured. This was done in an experimental set-up of 63 fields covering 170 ha in the Marais de Brouage (Charente-Maritime). The water regime is driven by natural processes (rainfall and evaporation) and artificially, through the manipulation of the leve! of the main canal. By modifying the subsidiary canals and building sluice gates it was possible to isolate experimental fields from the influence of the main canal. The experimental fields were subjected to two treatments, dry (minimal flooding), intermediate (six months flooding) and wet (10 months). We worked at three hierarchical spatial levels, a field, a single catchment area and a set of catchment areas. For the restoration of biodiversity in this case particular bird species, we underline the importance of choosing the appropriate spatial level, which will depend on the objectives. In this welland system, which reacts fast, the time scale is not a major constraint. Particular attention was paid to birds because these are known to be good indicators of other animal groups, and because the changes in agricultural practices have led to declines in most of the species of farmland in Europe. Between 1997-2001 we counted breeding birds (principally the Skylark A lauda arvensis and the Yellow wagtail Motacilla jlava ). The former species declined and the latter increased in the set of catchment areas (170 ha). We tested the hypothesis that this was due to an increase in the duration of flooding : at the field level, the Skylark declined and Wagtails increased when the duration of flooding increased. It is therefore possible to experimentally induce rapid changes in the populations of the se birds at small spatial scales. In contrast for other species such as the breeding waders, and in particular the Lapwing Vanellus vanellus, no response was obtained at the se spatial and temporal scales. However, at the scale of the whole marsh (12000 ha) and 12 years major changes occurred in the ahundance and distribution of Lapwings, many of the birds moving from wet grasslands to arable for nesting. We propose hypotheses to explain this which require testing in the future. These results show the importance of choosing clear objectives in ecological restoration projects, and appropriate indicator species. Since the ecological processes involved occur at different spatial and temporal scales, the level of biological organisation selected and the species will modulate the expression of the processes Depuis une trentaine d'années, l'intensification de l'agriculture par le drainage et la mise en culture a entraîné la perte des fonctions hydrauliques et écologiques des marais de la façade atlantique. Conservation et réhabilitation de la biodiversité de ces milieux sont à l'origine de notre étude. Grâce à un dispositif expérimental de terrain (63 parcelles, 170 ha) situé dans le marais de Brouage (Charente-Maritime), nous avons cherché à restaurer le caractère humide des prairies naturelles en manipulant la gestion hydraulique et pastorale, et en mesurant les conséquences sur les oiseaux. Les parcelles en prairie ont été soumises à trois « scénarios » de gestion hydraulique : le scénario « sec » (durée d'inondation minimale), le scénario « intermédiaire » (inondation d'environ 6 mois) et le scénario « humide » (inondation d'environ 10 mois). Entre 1997 et 2001, des dénombrements exhaustifs des passereaux nicheurs (Alouette des champs Alauda arvensis et Bergeronnette printanière Motacilla flava), ont montré qu'à une échelle temporelle de cinq ans, il est possible d'enregistrer des évolutions d'abondance des populations (baisse de la densité des alouettes et augmentation de celle des bergeronnettes) en fonction du degré d'humidité dans la parcelle. Par ailleurs, le suivi de la dynamique des populations du Vanneau huppé (Vanellus vanellus) a montré que l'échelle spatiale choisie (170 ha) n'était pas adaptée et qu'il fallait prendre en compte l'ensemble du marais, soit 12000 hectares. Sur le plan temporel, des changements très significatifs dans 1'utilisation de 1'espace sont intervenus chez cette espèce, mais ces changements n'auraient pu être détectés en 3 années seulement
|
|---|