What does the oxygen isotope composition of rodent teeth record? Aure ´ lien Royer a,e , Christophe Le ´ cuyer a,f,n , Sophie Montuire b,e , Romain Amiot a , Serge Legendre a , Gloria Cuenca-Besco ´s c , Marcel Jeannet d , Franc - ois Martineau a a Laboratoire de Ge´ologie de Lyon, UMR CNRS 5276, Universite´ Lyon 1 et Ecole Normale Supe´rieure de Lyon, 69622 Villeurbanne, France b Bioge´osciences, UMR CNRS 5561, Universite´ de Bourgogne, 6 Boulevard Gabriel, 21000 Dijon, France c Departamento de Ciencias de la Tierra, A ´ rea de Paleontologı ´a, Edificio de Geolo ´gicas, 50009 Zaragoza, Spain d LAMPEA, UMR 6636, MMSH, 5 rue du Chˆ ateau de l’Horloge, BP 647, 13094 Aix-en-Provence Cedex 2, France e Laboratoire Pale´obiodiversite´ et Evolution, EPHE-Ecole Pratique des Hautes Etudes, 21000 Dijon, France f Institut Universitaire de France, Paris, France article info Article history: Received 6 December 2011 Received in revised form 25 September 2012 Accepted 28 September 2012 Editor: J. Lynch-Stieglitz Available online 24 November 2012 Keywords: rodent fractionation equation oxygen isotope tooth phosphate palaeoclimate abstract Oxygen isotope compositions of tooth phosphate (d 18 O p ) were measured in 107 samples defined on the basis of teeth obtained from 375 specimens of extant rodents. These rodents were sampled from pellets collected in Europe from 381N (Portugal) to 651N (Finland) with most samples coming from sites located in France and Spain. Large oxygen isotopic variability in d 18 O p is observed both at the intra- and inter-species scale within pellets from a given location. This isotopic variability is partly explained by heterochrony in tooth formation related to the short time of mineralization for all rodent species as well as the duration of mineralization that is species-dependent. Consequently, tooth phosphate of rodents records a short seasonal interval in the oxygen isotope compositions of meteoric waters (d 18 O mw ). In addition, inter- species isotopic variability observed in the same pellets suggests behavioural differences implying distinct isotopic compositions for species living in the same location. At the scale of Europe, a robust linear oxygen isotope fractionation equation was determined for Muroidea between the midrange d 18 O p values and d 18 O mw values: d 18 O p ¼1.21( 70.20)d 18 O mw þ24.76( 72.70) with R 2 ¼0.79 (n¼9; po0.0001). & 2012 Elsevier B.V. All rights reserved. 1. Introduction The Quaternary is characterized by a succession of alternating glacial and interglacial stages leading to climatic changes that have deeply modified the floral and faunal associations (e.g. Anderson et al., 2007). At high-latitude, the oxygen isotope composition of ice provides a continuous and high-resolution temporal record of the continental climate (e.g. Dansgaard et al., 1993). At lower latitudes, continental sedimentary deposits only provide discontinuous tem- poral and spatial records associated with difficulties in the absolute dating of fossils or their host sediments. Most continental climate proxies are based on pollen, insect or mammal associations. Different methods are applied to reconstruct climate modes; they are based either on the analysis of faunal associations in relation to their preferential extant ecological niches (Delpech et al., 1983, 2000; Chaline et al., 1995; Griggo, 1996; Jeannet, 2010), or on faunal com- munities (e.g. Legendre, 1986, 1989; Montuire and Marcolini, 2002), or on diversity and species richness (Montuire et al., 1997, 2006; Herna ´ ndez Ferna ´ ndez, 2001; Legendre et al., 2005; van Dam, 2006; Herna ´ ndez Ferna ´ ndez et al., 2007; Cuenca-Besco ´ s et al., 2011). These different approaches generally involve the study of small mammals because of their high sensitivity to environmental parameters and climatic changes. Longinelli’s (1973, 1984) pioneer study demonstrated that the stable oxygen isotope composition of phosphatic tissues is a valuable proxy for investigating terrestrial climatic conditions. The oxygen isotope composition of biogenic phosphate (d 18 O p ) depends on both body temperature and oxygen isotope composition of body water (d 18 O bw ), itself related to the composition of ingested water most commonly of meteoric origin (Longinelli and Nuti, 1973; Kolodny et al., 1983; Luz et al., 1984). Indeed, most previous studies assumed that the oxygen isotope composition of vertebrate’s body water reflects the oxygen isotope composition of meteoric water (d 18 O mw ) depending on physiological parameters such as the general metabo- lism. These interdependent relationships allow the use of taxon- dependent oxygen isotope fractionation equations relating d 18 O p to d 18 O mw . At mid- and high-latitudes, the mean annual d 18 O mw is linearly related to the mean annual air temperature at the global scale (Dansgaard, 1964; Rozanski et al., 1993). Many oxygen isotope studies were performed on bones and teeth of large mammal fossils (e.g. Koch et al., 1989; Fricke et al., 1998; T¨ utken et al., 2007; Bernard et al., 2009; Chritz et al., 2009) as well as to small mammal fossils to estimate successfully past air temperatures (e.g. Lindars et al., 2001; Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/epsl Earth and Planetary Science Letters 0012-821X/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2012.09.058 n Corresponding author at: Laboratoire de Ge ´ ologie de Lyon, UMR CNRS 5276, Universite ´ Lyon 1 et Ecole Normale Supe ´ rieure de Lyon, 69622 Villeurbanne, France. E-mail address: clecuyer@univ-lyon1.fr (C. Le ´ cuyer). Earth and Planetary Science Letters 361 (2013) 258–271