Item

Abstract

Seaweed consumption by wild, feral and domesticated animals in coastal areas world-wide is currently likely widely underestimated. Seaweed consumption on the Orkney Islands by domesticated animals has become an established part of the archaeological literature, but the extent of seaweed consumption elsewhere is still largely unknown in archaeological contexts. The identification of small amounts of seaweed consumption by collagen δ13C and δ15N values remains problematic, as it is unclear to what extent seaweed consumption is reflected in skeletal tissues, and how results may vary between different tissues. In this study, modern sheep consuming known seaweed (predominantly kelp) and terrestrial diets on the Orkney Islands were analysed for δ13Ccollagen, δ15Ncollagen, δ13Cbone apatite and δ13Cenamel to provide a reference for archaeological studies. Seaweed and terrestrial vegetation were also analysed for δ13C and δ15N (n = 122). Seaweed δ15N values did not differ significantly from terrestrial vegetation on North Ronaldsay, indicating that δ15N is not a reliable indicator of seaweed consumption. In contrast, we confirmed that δ13C is a suitable marker for substantial seaweed consumption in all studied tissues in herbivorous diets in the absence of C4 plants. The consumption of both seaweed and terrestrial vegetation led to a large degree of variability in δ13C results (−19.1 to −11.5‰) within one herd kept under a consistent management system, due to differences in the amount of seaweed consumed by the individual sheep. However, when only small amounts of seaweed are consumed (<25%), this may not be evident in the δ13Ccollagen data. In contrast, when seaweed-consumption occurs primarily in winter, spring-born lambs may be expected to have substantially higher δ13C values than their mothers. This study emphasises the need for modern reference data in archaeology, and may aid the identification of seaweed consumption by herbivores globally.