Variation in marine mammal sulfur stable isotope compositions (δ34S) is driven by sources of primary production, not trophic position

The homogeneous isotopic composition of marine sulfate (δ 34 S ≅ +21‰) leads to the expectation of an oceanic isoscape exhibiting little variation in δ 34 S values. Nevertheless, instances of variation in δ 34 S values have been identified in Arctic marine ecosystems that require clarification to explain the mechanisms driving unexpectedly low values. Walruses are frequently characterized by low δ 34 S values compared to other marine mammals, but it is unclear if these values are a function of the low trophic position of walruses or of a specialized foraging niche. Here we examined the 2 potential drivers of differences in sulfur isotopes at the species level: trophic position and foraging ecology. By analysing archaeological samples of walrus, ringed seal and polar bear from archaeological sites at 3 discrete locations in the Canadian Arctic (Smith Sound, Ellesmere Island; Lancaster Sound, Devon Island; and northwest Hudson Bay), we clarified and confirmed that in collagen-to-collagen comparisons, differences in δ 34 S values are not driven by trophic discrimination factors. Bacterial influences on the sulfur cycle, particularly symbiotic relationships between chemosynthetic bacteria and invertebrates in the benthos, may be a cause of the low δ 34 S values in walruses and likely represent an often-overlooked contribution of primary production to the shelf food web.