| |
Communication at the NABS Annual meeting, Keystone, 2000
(95) INDEX OF BIOTIC INTEGRITY FOR A EUROPEAN RIVER BASIN, THE MEUSE : COMPARISON OF TWO METHODS TO CHARACTERIZE REFERENCE CONDITIONS.
N. Roset1, D. Goffaux2, J. Breine3, J.J. De Leeuw4, T. Demol5, T. Oberdorff1, C. Belpaire3, and P. Kestemont2. 1Conseil Supérieur de la Pêche, Metz, France, 2Facultés Universitaires Notre Dame de la Paix, Namur, Belgium, 3Instituut voor Bosbouw en Wildbeheer, Groenendaal, Belgium, 4Netherlands Institute for Fisheries Research, Ijmuiden, The Netherlands, 5Centre de Recherche de la Nature de la Forêt et du Bois, Gembloux, Belgium
In 1998 we assessed the ecological quality of the Meuse River basin (36,000 km2) fish assemblages. The Meuse (925 km) drains three countries (France, Belgium, The Netherlands) and ecological quality is generally poor due to pollution and hydrological modifications. A fish and ecological data base was assembled from 845 samples at 480 stations collected over 10 years in the three countries. Fish were sampled by electrofishing, gillnets, and beam trawling as river size increased. We estimated reference conditions in two ways: the original IBI method of trisection, and a multivariate model. The latter consists of modeling the fish assemblage metrics as functions of geomorphological and hydrological variables (e.g., catchment area, distance to source, width x slope, altitude) and ecological variables (water quality, physical habitat integrity). We used stepwise regression to attribute different weights to environmental variables based on their hierarchical influence on each metric. After selecting the best performing metrics, these statistical models, which explained >60% of data variability, were used to define reference conditions. They were validated on three independent data sets to evaluate index sensitivity to different types of degradation, temporal changes, and sampling methods.
Presented at 10:30 AM on Tuesday, May 30, 2000 in Determining Good Sites from Bad II
|
|
