RATIONALE: We evaluated the applicability of tree-ring δ13 C and δ18 O values in bulk wood - instead of the more time and lab-consuming α-cellulose δ13 C and δ18 O values, to assess climate and physiological signals across multiple sites and for six tree species along a latitudinal gradient (35°97'N to 45°20'N) of the northeastern United States. METHODS: Wood cores (n = 4 per tree) were sampled from ten trees per species. Cores were cross-dated within and across trees at each site, and for the last 30 years. Seven years, including the driest on record, were selected for this study. The δ13 C and δ18 O values were measured on two of the ten trees from the bulk wood and the α-cellulose. The offsets between materials in δ13 C and δ18 O values were assessed. Correlation and multiple regression analyses were used to evaluate the strength of the climate signal across sites. Finally the relationship between δ13 C and δ18 O values in bulk wood vs α-cellulose was analyzed to assess the consistency of the interpretation, in terms of CO2 assimilation and stomatal conductance, from both materials. RESULTS: We found offsets of 1.1‰ and 5.6‰ between bulk and α-cellulose for δ13 C and δ18 O values, respectively, consistent with offset values reported in the literature. Bulk wood showed similar or stronger correlations to climate parameters than α-cellulose for the investigated sites. In particular, temperature and vapor pressure deficit and standard precipitation-evaporation index (SPEI) were the most visible climate signals recorded in δ13 C and δ18 O values, respectively. For most of the species, there was no relationship between δ13 C and δ18 O values, regardless of the wood material considered. CONCLUSIONS: Extraction of α-cellulose was not necessary to detect climate signals in tree rings across the four investigated sites. Furthermore, the physiological information inferred from the dual isotope approach was similar for most of the species regardless of the material considered.
