) clumps on gravel and mussels. Species (M. modiolus or S. rustica), tissue (foot and posterior adductor muscles in M. modiolus) and site (8.1 km apart) factors had strong effects on isotope ratios, whilst the presence of a potential competitor had no effect on SIR in either species. We conclude that the diets of M. modiolus and S. rustica are not affected by co-occurrence and do not overlap much, giving no evidence of competition. Further research should take in account the substantial difference in SIR between the 2 M. modiolus muscular tissues studied: −21.582 ± 0.048 (δ 467: 113-120, 2012 size, likely due to technical limitations of the approach based on direct counting of tagged food particles consumed by each individual.Mar Ecol Prog SerStable isotope ratios of carbon and nitrogen are widely used to indirectly assess the relative contribution of different food sources to the diet of animals (see Boecklen et al. 2011 for review). Isotopic signatures are expressed as δ, the ratio of heavy (e.g. . Based on the measurements of these ratios in potential food sources (which usually include particulate organic matter of different origin or certain prey species for marine benthic consumers), one may apply a mixing model to estimate their contribution to the diet of a particular species (Phillips et al. 2005). Stable isotope composition of consumer tissues does not reflect the most recent diet. Instead, it is affected by feeding patterns of several weeks to few months preceding sampling, depending on the metabolic rate of a particular organism and tissue (Dalerum & Angerbjörn 2005). Accurate mixing models thus require potential food sources sampled in advance according to previously studied metabolic rates.Interspecific differences in 13 C and 15 N signatures have been unhesitatingly interpreted as indicators of feeding segregation between potentially competitive suspension-feeding species (e.g. Dubois et al. 2007). High variation of stable isotope signatures observed among the tissues of a single organism may, however, compromise these conclusions (see Lorrain et al. 2002, Yokoyama & Ishihi 2006, CabanellasReboredo et al. 2009, Deudero et al. 2009, Aya & Kudo 2010 for data on bivalves). In coastal Arctic waters, relatively slow metabolic rates of benthic invertebrates are combined with high seasonal changes in potential food sources for suspensionfeeders due to the presence of ice cover and variation in river discharge. Tissue-specific isotope incorporation models (Boecklen et al. 2011) are not yet developed for most species. Therefore, the evidence of interspecific feeding interference or its absence based on mixing models demands multiple repeated sampling of the potential food sources at least several months prior to sampling the consumers. This may not always be technically possible or logistically feasible due to the ice freezing and melting seasons.Interspecific competition for food alters the diet at least of an inferior competitor in relation to the diet observed in the absence of a superior competit...