Euro-limpacs Deliverables


Report on Integrated stochastic models evaluating the impact of global change at the catchment- scale

The aim of this deliverable is to study catchment−scale hydrological processes and biodiversity, and to contribute to integrated water management, in the context of the global changes. The following topics are covered:
i) hydrochemical modelling at the catchment scale with applications of land−cover data;
ii) effect of land−cover on the biodiversity of fish communities;
iii) effect of environmental changes in hydrochemistry on the biodiversity of macroinvertebrates, macrophytes and fish fauna;
iv) information supply and demand for decision−making on river basin water resources.

This deliverable is organised as follows:
a) The first part aims to determine the long term trends in water chemistry in the Adour−Garonne hydrographic network during the last three decades. Using both the spatial (45 studied sites, over the catchment scale) and temporal (more than 30 years of data, 1971 to 2004) data, models were developed to predict the general trends of variations of some hydro−chemical components. Using the GIS data, models predicting main hydrochemical changes were also developed.

b) The second part aims to model the impacts of land cover on fish distribution at the catchment scale. Using 192 studied sites over the whole Garonne basin, we initially developed the models predicting the community changes using to the first level of GIS data and some topographic data. This paper has been published in Science of the Total Environment (2006). The second step is to develop the models predicting species occurrence, using detailed GIS data. The contributions of parameters explaining fish occurrence, according to global changes is discussed in this part.

c) The third part concerns the models developed for the Lambourn basin. Using benthic macroinvertebrate data, annual variation of communities were evaluated using a self−organizing map (SOM). The ordination capacity of SOM was compared with traditional multivariate statistical analysis such as principal component analysis (PCA) and non−metric multidimensional scaling (NMS).

Download deliverable report [1.6 MB]

« back to Deliverables