ABSTRACT

Oxygen isotope ratios (δ¹⁸O) of tree-ring cellulose in pine (Pinus densiflora) and oak (Quercus serrate and Quercus variabilis) growing in central Japan from 1970 to 2011 share a common inter-annual variance among different cores in a tree, different trees in a species, and different species in a forest. The high mean correlation between the two radii (r = 0.738) of individual trees and the mean standard deviation of two measurements for the same year from the two radii of eight trees (σ = 0.361‰) reveals the reliability and representativeness of single cores. The significant correlation (r = 0.667, p < 0.01) between pine and oak cellulose δ¹⁸O suggests that δ¹⁸O in tree-ring cellulose is controlled by common external factors, and that it is possible to establish a unique δ¹⁸O chronology using different tree species. The δ¹⁸O values of tree-ring cellulose in pine and oak trees are negatively correlated with precipitation from June to August (correlation coefficients are −0.679 for pine and −0.583 for oak), which is attributable to the precipitation amount effect, mainly caused by the stagnant rain front (Baiu) in early summer and typhoons in late summer in this region. The δ¹⁸O values are negatively correlated with relative humidity from April to September for pine (r = −0.632, p < 0.01), and in June and July for oak (r = −0.437, p < 0.01); these correlations probably reflect different lengths of the growing season. Multiple regression analysis reveals that tree-ring δ¹⁸O values of pine and oak in this region are largely governed by summer precipitation. Relative humidity does not appear to greatly affect tree-ring δ¹⁸O of oak, in contrast to pine, which may be interpreted in terms of differences in leaf morphology and/or cellulose synthesis processes. Our results indicate that δ¹⁸O values of tree-ring cellulose can provide a potential crossdating tool, irrespective of the physiological differences between gymnosperms and angiosperms. In central Japan, tree-ring δ¹⁸O values of both pine and oak can act as reliable proxies of June–August precipitation, while pine trees tend to exhibit more potential for reconstruction of relative humidity.