Search

The study of large-scale pressure-impact relationships involves questions of hierarchy and scales. Answers to these questions will help managers define priorities for action to achieve the good ecological status' required by the Water Framework Directive (WFD). The main objectives of our study were 1) to establish the relative impact of pressures that degrade ecological status, especially those caused by agriculture and urbanization, 2) to identify regional patterns in these pressure-impact relationships and 3) to evaluate the relative weight of the pressures acting at the basin and riparian corridor scales, and the possible buffering effect of riparian areas. We developed large-scale models linking invertebrate indices of ecological quality to river basin and riparian land cover in France, Slovakia, Estonia and UK. Invertebrate indices, transformed to Ecological Quality Ratios (EQR), were taken from national monitoring networks. We based the models on Partial Least Squares (PLS) regressions at national and a hydro-ecoregion (HER) scales. The HERs provided a framework for grouping data in terms of natural river features and human activities. The different national methods provided consistent results that indicated the hierarchy of pressures impacting river invertebrates at the European scale. The most salient result was that artificial land cover (e.g. urban and industrial sites) in the river basin represented the pressure with the most negative impact on invertebrate indices, in all countries and regions. The impact of agricultural land cover was more variable. Arable land had a smaller impact than urban areas, and it was even insignificant in some models. The impact of vineyards depended on the natural geographical context. The effect of pastures seemed to be related to the intensity of the livestock they carried. These results supported the concept of regional pathologies for river ecosystems, as land use and anthropogenic influences are closely linked to physical landscape features. The proportion of arable land in the river basin appeared to be a weak predictor of agricultural impacts by itself; the type of cultivation and intensity as well as the proximity to the river must be taken into account. At the riparian corridor scale, the negative impact of artificial areas or arable land and the positive effects of forests and pastures were demonstrated in many regions. The protective effect of riparian forests against mixed agricultural and urban pressures was demonstrated in three regions in France. Riparian corridors appear to be manageable areas, and these results strongly support the idea of including their restoration in priority actions for achieving good ecological status.

International audience; Spatio-temporal variability of the physical conditions in streams is known to influence freshwater population dynamics. However, the complex relationship between these physical factors (mainly available habitat, water temperature, discharge, sediment transport) and the biological response is not yet clearly understood in either natural or regulated streams. In order to investigate this relationship, it is essential to understand how the influences of these physical factors differ in time and space. In the present study, we analyzed the spatio- emporal patterns of fluctuation in 31 brown trout populations spread across France, in both natural and regulated (i.e. by-passed sections downstream of hydroelectric water intakes) sites. This data is a result of annual fish monitoring surveys carried out continuously over a minimum of 5 years. Using statistical methods, we estimated the degree of synchrony among brown trout cohort densities (young-of-the-year, juveniles and adults) and evaluated the spatial scale at which synchrony is best identified. We also investigated whether synchrony patterns differ between natural and regulated sites. Results are then discussed according to fluctuations in local physical factors. This study demonstrates the utility of multi-site approaches for a better understanding of freshwater population dynamics. It is indeed of major importance to identify the most structuring physical factors at regional or local scales, in order to design consistent management schemes.

La mise en ½uvre de la Directive-cadre européenne sur l'eau implique le développement de modèles pressions/impacts permettant d'identifier les actions nécessaires à la restauration. Nous avons développé des modèles " large échelle ", reliant l'indice invertébré aquatique (IBGN) à l'occupation du sol du bassin versant. La méthode des arbres de décision produit des classifications permettant l'extrapolation spatiale des résultats. Des modèles d'extrapolation spatiale validés ont été employés pour représenter à l'échelle nationale l'état écologique probable des cours d'eau, basé sur l'IBGN. L'urbanisation du sol est apparue comme premier facteur négatif d'impact suivi de l'agriculture (terres arables) ; les espaces naturels sont apparus comme facteur positif pour l'état écologique. Nous avons également utilisé ce modèle pour tester deux hypothèses pour la limite de bon état écologique et pour simuler les résultats à l'échelle du territoire national. Beaucoup de secteurs soumis à une pression agricole élevée (plaines et région méditerranéenne) sont classés en limite de " bon état " suivant l'hypothèse prise en considération. / For the implementation of the European Water Framework Directive, there is a need of pressures/impacts models allowing to set the priority actions for the restoration. We developed large scale pressures/impacts models, linking aquatic invertebrate index (IBGN) to watershed land cover. We used decision trees method to generate classifications allowing spatial extrapolation. Validated spatial extrapolation models were used to represent on a national scale the probable current ecological status of rivers based on IBGN. Urban land cover appeared as the first negative impact factor followed by agriculture (arable land); natural areas appeared as a positive factor on ecological status. We used this model to test two hypothesis for the good status boundary, and to simulate the results at the national scale. Many areas under high agricultural pressure (lowlands and Mediterranean region) are in borderline situation according to the good status hypothesis.