11:00 〜 13:00
[U05-P09] Proxy assessments of tree-ring intra-annual δ18O for precipitation reconstruction in Ashiu Forest Research Station, Kyoto.
Tree-rings width, oxygen and carbon isotopic ratios have potential for reconstructions of paleoclimate. Especially, oxygen isotopic ratios are highly correlated among individuals of different tree species and have been studied as a good paleoclimate proxy for precipitation and relative humidity. However, most of these studies have measured only inter-annual isotopic ratios, and only a few studies have conducted paleoclimate reconstructions based on intra-annual isotopic ratio variations. In this study, we examined whether isotopic ratios analysis of intra-annual tree-rings can be used to reconstruct paleoclimate with higher resolution than conventional analysis of inter-annual tree-rings.
We used Cryptomeria japonica as a sample. After cellulose extraction, each tree-ring was cut into six sections and the oxygen and hydrogen isotopic ratios were measured. We measured isotopic ratios during 1918-2017 from Cryptomeria japonica.
The age of each sample was determined by cross-dating with master chronology of oxygen isotopic ratios in central Japan. We confirmed that there were no problems about reproducibility of six-division of tree-rings by two times six-division and measurement from 1988 to 1997 of Cryptomeria japonica.
Inter-annual oxygen and hydrogen isotopic ratios of Cryptomeria japonica were positively correlated, but intra-annual oxygen and hydrogen isotopic ratios were inversely correlated in Cryptomeria japonica. This may be due to the intra-annual variation in the ratio of secondary isotope exchange between organic matter produced in the leaves and conduit water prior to cellulose synthesis in Cryptomeria japonica.
We numbered six-section No.1 to No.6. The most inner part (earlywood) of the six-sections is numbered 1, and the outer (latewood) parts are numbered sequentially up to 6. There was a strong positive correlation between the oxygen isotopic ratio variations of adjacent numbers. However, the correlation with the other numbers was weak for number 6, suggesting that the growth period of this tree differs from the other numbers.
Both inter and intra-annual oxygen isotopic ratio variations were negatively correlated with summer precipitation and relative humidity, and positively correlated with temperature. The period of strongest inverse correlation with precipitation and each of six-divisions δ18O gradually shifted from spring to the latter summer.
We attempted a multiple regression analysis using the inter-annual oxygen isotopic ratios as explanatory variables and meteorological data such as precipitation as objective variables. Because there was a strong positive correlation between the oxygen isotopic ratio variations of No.1 to No.5, multicollinearity was avoided by principal component analysis, and then multiple regression analysis was conducted. The results showed that the multiple regression was more accurate over a wider period from spring to summer than the single regression analysis using the inter-annual oxygen isotopic ratios. This indicates that the analysis of intra-annual tree-ring cellulose allows us to reconstruct the paleoclimate with higher resolution than the analysis of inter-annual tree-ring cellulose.
We used Cryptomeria japonica as a sample. After cellulose extraction, each tree-ring was cut into six sections and the oxygen and hydrogen isotopic ratios were measured. We measured isotopic ratios during 1918-2017 from Cryptomeria japonica.
The age of each sample was determined by cross-dating with master chronology of oxygen isotopic ratios in central Japan. We confirmed that there were no problems about reproducibility of six-division of tree-rings by two times six-division and measurement from 1988 to 1997 of Cryptomeria japonica.
Inter-annual oxygen and hydrogen isotopic ratios of Cryptomeria japonica were positively correlated, but intra-annual oxygen and hydrogen isotopic ratios were inversely correlated in Cryptomeria japonica. This may be due to the intra-annual variation in the ratio of secondary isotope exchange between organic matter produced in the leaves and conduit water prior to cellulose synthesis in Cryptomeria japonica.
We numbered six-section No.1 to No.6. The most inner part (earlywood) of the six-sections is numbered 1, and the outer (latewood) parts are numbered sequentially up to 6. There was a strong positive correlation between the oxygen isotopic ratio variations of adjacent numbers. However, the correlation with the other numbers was weak for number 6, suggesting that the growth period of this tree differs from the other numbers.
Both inter and intra-annual oxygen isotopic ratio variations were negatively correlated with summer precipitation and relative humidity, and positively correlated with temperature. The period of strongest inverse correlation with precipitation and each of six-divisions δ18O gradually shifted from spring to the latter summer.
We attempted a multiple regression analysis using the inter-annual oxygen isotopic ratios as explanatory variables and meteorological data such as precipitation as objective variables. Because there was a strong positive correlation between the oxygen isotopic ratio variations of No.1 to No.5, multicollinearity was avoided by principal component analysis, and then multiple regression analysis was conducted. The results showed that the multiple regression was more accurate over a wider period from spring to summer than the single regression analysis using the inter-annual oxygen isotopic ratios. This indicates that the analysis of intra-annual tree-ring cellulose allows us to reconstruct the paleoclimate with higher resolution than the analysis of inter-annual tree-ring cellulose.