In middle order (orders 3-6) streams, autochthonous (in-stream) production constitutes a significant portion of the whole stream energy budget. Primary producers include autotrophic bacteria, algae, bryophytes, and vascular plants; although benthic algae are often reported as the dominant producers in many streams. Activity associated with these organisms certainly influences stream nutrient retention, in particular the nutrient uptake length.

     However, metabolism of biofilm sediments is also affected by the increase of N and P loads. Moderate increase of nutrient input may subsidise the stream biota and enhance their activity and growth, and thereby reach scale retention of nutrients. Nevertheless, above a certain threshold, increased nutrient inputs may saturate the community response.

     The objective of this workpackage is to examine the influence of in-stream biological processes on nutrient retention in streams affected by high nutrient loads.

     We hypothesise that, under similar stream discharge conditions, nutrient retention in streams with high nutrient loads will be lower than in less polluted streams if biota mostly controls nutrient retention. Enhanced primary production due high nutrient loads represents an important source of energy and carbon to heterotrophic bacteria. Along with high nitrification rates, the processing of this organic pool can ultimately induce severe oxygen depletion and stimulate anaerobic processes, such as denitrification. This fact leads to heterogeneity in biological responses and physicochemical variability in the streambed as well as in the hyporheic sediments. Hence, depending on the "quality" of the nutrient loads, either autotrophic or heterotrophic processes will control the removal of dissolved nutrients from the water column. Overall, these metabolic changes at the sub-reach scale may affect stream nutrient retention at the reach scale. These effects and the associated consequences are also important to examine and quantify to be able to predict the direction of change in nutrient retention in front of a particular disturbance (e.g., nutrient load increases).

     To test these hypotheses we propose to measure the biofilm metabolism (autotrophic versus heterotrophic), biomass (polysaccharides and bacteria) and community structure (e. g. dominant algal taxa).

     We will measure streambed metabolism by conducting diurnal upstream-downstream dissolved oxygen budgets corrected by specific reaeration factors within each study site and estimating in situ denitrification rates on surface sediments. Since one of the major mechanisms of phosphate net retention results from its co-precipitating with calcium carbonate when pH increases due to high primary production rates, daily changes in pH joint with calcium, alkalinity, and dissolved phosphate will be also analysed during each sampling date.

     Of course, we are aware of possible macroinvertebrate top-down control on nutrient retention by biofilms. This might be particularly true for autotrophs. In fact, the benthic algal biomass declines in the presence of invertebrate grazers (snails, mayflies, midges) and biofilm nutrient uptake can decrease. Not only algae, but also bacteria and the matrix itself can also be of considerable trophic significance to stream invertebrates. However, invertebrate grazing on biofilm communities can also have dual effect on nutrient retention (i.e., inhibit and/or stimulate reach-scale solute removal).

     Since the efficiency of biofilm biomass removal depends on the composition of the invertebrate community, we propose to analyse invertebrate assemblages for functional feeding groups (FFG) and measure biomass distributions (in terms of C, N and P) to the FFG and dominant taxonomic entities. To test the effect of elevated nutrient loads on biofilms and on the structure of the invertebrate consumer compartments, we propose a compare of upstream versus downstream reaches relative to the point source. This would enable us to partition between human induced (e. g., elevated nutrient loads) and both geomorphological and biological factors.

     These biological measurements will be related to those from workpackage 2 to elucidate which biological processes play a major role on stream nutrient retention. These relationships will be compared to those observed between nutrient retention and physical parameters (WP2). Data analyses will be conducted for each particular site and also across-sites to develop an empirical model that will predict nutrient retention in front of changes in physical, chemical and biological features in polluted streams.

     All these relationships will be then used as the empirical basis for the knowledge base of the Expert System.