Nine biological indices based on the macroinvertebrate community inhabiting rivers have been calculated in order to find out which of them is more appropriate to evaluate the qualitative status of the running water for public environmental authorities. The BMWP' score (Alba-Tercedor & Sánchez-Ortega 1988) has been chosen because it is both accurate and precise, and easy to calculate. Some problems derived from sampling strategies are discussed.

Every month 2 samples of the Rhine and 1 sample of the Rhone are analysed for about 20 chemical and physical variables at 7 stations by two monitoring agencies.This paper presents some of our conclusions on the accuracy of these programmes and on the solubility of the calcium/carbonate/sulphate/phosphate system. In total we have analysed 916 data sets ; they were screened for imprecision before being used.At all stations the water is saturated with CaCO3 , but the "apparent" solubility product depends on the pH. The regression line can be described by IPca = 8.73 x 10-15. pH15.667.In both the Rhine and the Rhone an acidification is seen and an increase of the sulphate concentration when going downstream. We have shown that the increase in sulphate concentration is due to the disposal of CaSO4. This addition of sulphate influences the quotient Ca/HCO3. The data fit very well the theoretical regression line Ca/HCO3 = 0,5(SO4) + 0.5.The acidification originates from the mineralization of organic matter. For 6 stations dowstream from the great lakeswe found that the two rivers are saturated with hydroxy-apatite, the -log of the ionic product being 49.9 (Rhine ; σ = 0.6 ;N = 259) and 50.0 (Rhone ; σ = 0.3 ; N = 479). The o-phosphate concentration does not therefore depend on the phosphate loading of the rivers, but on the Ca concentration, the temperature and the pH. This has great influence on the transport of phosphate through the delta's of the two rivers. The disturbance of acidity and buffering capacity influences heavy metal behaviour as well.

L'évaluation de la valeur descriptive des peuplements de Cladocères vis-à-vis du milieu perçu comme entité spatio-temporelle (temps et espace hiérarchisés et leurs interactions) a comporté trois phases successives : classification des stations par leur fonctionnement hydrologique (origine, qualité et renouvellement de l'eau, mesurés par des paramètres physico-chimiques), analyse des interactions espace-temps au niveau des peuplements de Cladocères puis classification des stations par les peuplements de Cladocères. L'étude a été réalisée sur 3 bras-morts du Rhône qui diffèrent par leur mode de relation avec les eaux courantes du fleuve ; chacun de ces 3 biotopes comprend plusieurs stations de mesures physico-chimiques et faunistiques. La comparaison entre la classification hydrologique et la classification faunistique montre que les Cladocères possèdent un pouvoir descripteur fonctionnel et non stationnel ; en effet, à l'exception des grandes zones d'eau libre, profondes et dépourvues de macrophytes, chaque peuplement transmet une information non pas sur sa station et sa végétation, mais sur son bras-mort et son fonctionnement hydrologique global.