Is cellulose extraction necessary for developing stable carbon and oxygen isotopes chronologies from Callitris glaucophylla ? Louise E. Cullen * , Pauline F. Grierson Ecosystems Research Group, School of Plant Biology M090, University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia Received 2 March 2005; received in revised form 2 August 2005; accepted 3 November 2005 Abstract We examined the necessity of extracting cellulose for the development of stable carbon and oxygen isotope chronologies from Callitris glaucophylla J. Thompson and L.A.S. Johnson (Cupressaceae). We compared relationships among isotopes of whole wood, cellulose and climate for the period 1979–1999 and using a moving interval technique with a 10-year window. Isotopes of whole wood and cellulose were significantly correlated and had similar relationships with climate when considered over the entire period. In general, variation in d 13 C and d 18 O was related to summer temperature relative humidity and rainfall. However, d 13 C and d 18 O of whole wood generally exhibited weaker and more variable correlations with climate through time than cellulose. In addition, isotopes of whole wood did not deviate from 1979 to 1999 mean values in the same way as isotopes of cellulose. The instability of whole wood–climate relationships may be due to the inclusion in whole wood of a climate signal from lignin that is more variable between tree rings than cellulose. Overall, it appears that C. glaucophylla at least whole wood does not record climate in the same way as cellulose and stable isotopes of cellulose provide a more temporally stable proxy of past climate. We suggest that when developing isotope chronologies for other species the suitability of whole wood vs. cellulose as a proxy for climate should be assessed with consideration of the temporal stability of their relationship with climate. D 2005 Elsevier B.V. All rights reserved. Keywords: d 13 C; d 18 O; Dendrochronology; Precipitation; Temperature; Relative humidity 1. Introduction The utility of stable isotope methodologies for den- drochronology has greatly increased over recent years due to significant advances in understanding the frac- tionation and physiological processes that contribute to isotope variations and development of online methods of analysis (McCarroll and Loader, 2004). Isotope ra- tios of tree rings provide information on trends in temperature, rainfall and relative humidity, patterns in atmospheric circulation, as well as the response of ecophysiological processes of trees to changes in cli- mate (Gray and Thompson, 1977; Leavitt and Long, 1989; Robertson et al., 1997; Anderson et al., 1998; Edwards et al., 2000; Barbour et al., 2002). Despite the growing number of studies measuring stable isotopes of tree rings, there is ongoing debate as to whether analysis of whole wood or cellulose is most appropriate for dendroclimatological studies. Wood is composed of several components, which vary in their isotopic signatures because of individual pathways of formation (Wilson and Grinsted, 1977). Analyzing whole wood may, therefore, increase the bnoiseQ in 0031-0182/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.palaeo.2005.11.003 * Corresponding author. Tel.: +61 8 6488 3445; fax: +61 8 6488 7925. E-mail address: lcullen@plants.uwa.edu.au (L.E. Cullen). Palaeogeography, Palaeoclimatology, Palaeoecology 236 (2006) 206–216 www.elsevier.com/locate/palaeo