Ecological hazard assessment of chemicals for aquatic ecosystems requires the use of biological assays in order to establish relationships between contamination and effect existing in the field. This step is an essential condition for the implementation of programs meant the restoration of aquatic environments. Therefore the in situ experimentations based on the use of calibrated organisms from one control population seeking to limit biological confounding factors, appears today as essential to improve the diagnosis of aquatic systems. In this context, we developed a suite of markers addressing neurotoxicity (acetylcholinesterase), genotoxicity (Comet assay), digestive enzymes, feeding rate and reproduction impairments and capacity to osmoregulate (Xuereb et al., 2009; Geffard et al., 2010; Issartel et al., 2010) in the freshwater sentinel species Gammarus fossarum. Then, for each marker, benchmark values taking into account their natural variability in relation to environmental confounding factors (temperature, alkalinity…) have been determined and validated in field experiments. These values lead to an accurate interpretation of marker levels in terms of contamination and toxicity of aquatic systems (Coulaud et al., 2011). However, the diagnosis using a sentinel species should be relevant towards the diversity of native populations and related species with their own history. Furthermore, the relevance of the population and the species chosen as surrogate test organisms is questionable. The potential divergence of life histories and sensitivities to pollutants between experimental and native populations is in fact an open question. The Project GAMMA aims to assess (1) the natural variability of molecular responses (AchE, digestive enzymes) and life-history traits (feeding rate, reproduction and osmoregulation) in relation to geographical location of control gammarid populations and species used as sources organisms for in situ tests, and (2) the divergence of the sensitivities of these markers to contaminants between populations and species. Results will be presented and their implication for ecotoxicological assessment of aquatic systems will be discussed. Funding was provided by ANR CESA (n°021 04), GAMMA, 2012-2015.