Quaternary Science Reviews 23 (2004) 977–991 Carbon and nitrogen stable isotope analysis of northwest European horse bone and tooth collagen, 40,000BP–present: Palaeoclimatic interpretations Rhiannon E. Stevens*, Robert E.M. Hedges Research Laboratory for Archaeology and the History of Art, University of Oxford, 6 Keble Road, Oxford OX1 3QJ, UK Received 19 February 2003; accepted 13 June 2003 Abstract Carbon and nitrogen stable isotope ratios of bone and tooth dentine collagen from 142 horses were analysed to consider the potential of isotopic analysis of faunal remains as a palaeoenvironmental and palaeoclimatic proxy. A general trend through time towards lighter d 13 C values was observed, but with an obvious depletion at the Late Glacial/Holocene transition. Horse d 13 C values appear to track plant d 13 C. Although phenomenon such as the canopy effect may influence faunal d 13 C values, we believe that the influence on plants of increasing atmospheric CO 2 is the primary factor in the change in faunal d 13 C values at the Late Glacial/ Holocene transition. A post-glacial depletion in horse d 15 N values is observed in areas proximal to glacial retreat although no data is available in these areas during the Last Glacial Maximum. Horse d 15 N values are presumed to reflect plant d 15 N values. Depleted plant nitrogen may be attributed to a number of factors e.g. water availability, temperature, cycling of nitrogen in the soil, soil nutrient availability (especially phosphorus in the post-glacial period), and the type of mycorrhizal associations formed with plant. These effects may be interrelated, and be linked such that a positive feedback can give a non-linear response to a single parameter. r 2004 Elsevier Ltd. All rights reserved. 1. Introduction Isotopic analysis of bone collagen has long been used as a quantitative technique for reconstructing archae- ological human and faunal diet (Schoeninger and DeNiro, 1984; Ambrose and DeNiro, 1986; Iacumin et al., 1997; Bocherens et al., 1999; Richards and Hedges, 1999; Richards et al., 2001; Privat et al., 2002). These palaeodietary reconstructions are based on the principle that food sources contain different isotopic signatures, which are passed along the food chain to their consumers. In recent years, it has been suggested that although diet is the primary control determining bone collagen isotopic values, other para- meters such as climate and local environment may also create small-scale variability (Van Klinken et al., 1994; Cormie and Schwarcz, 1996; Gr . ocke et al., 1997; Schwarcz et al., 1999). Hedges et al. (2004) highlight the potential of bone collagen stable isotope ratios as a palaeoclimatic proxy, but recognise that a lack of continuity of data, both temporally and spatially, from certain species can confuse reconstructions. Different species occupy different ecological niches, thus isotopic analysis of individual species can provide an insight into specific palaeoenvironments. This case study focuses on stable isotope analysis of horse collagen (Equus ferus and Equus caballus) from North- west Europe. The aims of this study were: (1) to confirm temporal trends in faunal collagen isotope values observed by Richards and Hedges (2003) and Hedges et al. (2004); (2) to establish if the rate and magnitude of change in horse collagen isotope values is uniform across Northwest Europe; (3) to consider the influence of different climatic parameters on faunal isotope values; and (4) to determine what palaeoenvironmental information can therefore be gained from these results. Equus ferus (pre-domesticated horses from the British Isles) and Equus caballus (pre-domesticated or domes- ticated horses from the continent, and domesticated horses from the British Isles) are analysed in this study. Although horse was absent in the British Isles between ca 9000 BP and the start of the Neolithic due to climatic warming and woodland development (Yalden, 1999), it ARTICLE IN PRESS *Corresponding author. E-mail address: rhiannon.stevens@rlaha.ox.ac.uk (R.E. Stevens). 0277-3791/$-see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.quascirev.2003.06.024