Variation in oxygen isotope fractionation during cellulose synthesis: intramolecular and biosynthetic effects

The oxygen isotopic composition of plant cellulose is commonly used for the interpretations of climate, ecophysiology and dendrochronology in both modern and palaeoenvironments. Further applications of this analytical tool depends on our in-depth knowledge of the isotopic fractionations associated with the biochemical pathways leading to cellulose. Here, we test two important assumptions regarding isotopic effects resulting from the location of oxygen in the carbohydrate moiety and the biosynthetic pathway towards cellulose synthesis. We show that the oxygen isotopic fractionation of the oxygen attached to carbon 2 of the glucose moieties differs from the average fractionation of the oxygens attached to carbons 3–6 from cellulose by at least 9%, for cellulose synthesized within seedlings of two different species (Triticum aestivum L. and Ricinus communis L.). The fractionation for a given oxygen in cellulose synthesized by the Triticum seedlings, which have starch as their primary carbon source, is different than the corresponding fractionation in Ricinus seedlings, within which lipids are the primary carbon source. This observation shows that the biosynthetic pathway towards cellulose affects oxygen isotope partitioning, a fact heretofore undemonstrated. Our findings may explain the species-dependent variability in the overall oxygen isotope fractionation during cellulose synthesis, and may provide much-needed insight for palaeoclimate reconstruction using fossil cellulose.

Oxygen isotope ratios of cellulose-derived phenylglucosazone: An improved paleoclimate indicator of environmental water and relative humidity

Oxygen atoms within fossil wood provide high-resolution records of climate change, particularly for the Quaternary. However, current analysis methods of fossil cellulose do not differentiate between different positions of the oxygen atoms. Here, we propose a refinement to tree-cellulose paleoclimatology modeling, using the cellulose-derived compound phenylglucosazone as the isotopic substrate. Stem samples from trees were collected at northern latitudes as low as 24°37′N and as high as 69°00′N. We extracted stem water and cellulose from each stem sample and analyzed them for their 18O content. In addition, we derived the cellulose to phenylglucosazone, a compound which lacks the oxygen attached to the second carbon of the cellulose–glucose moieties. Oxygen isotope analysis of phenylglucosazone allowed us to calculate the 18O content of the oxygen attached to the second carbon of the cellulose–glucose moieties. By way of these analyses, we tested two hypotheses: first, that the 18O content of the oxygen attached to second carbon will more closely reflect the 18O content of the stem water, and will not resemble the 18O content of either cellulose or its derivative phenylglucosazone. Second, tree-ring models that incorporate the variable oxygen isotope fractionation shown here and elsewhere are more accurate than those that do not. Our first hypothesis was rejected on the basis that the oxygen isotope ratios of the oxygen attached to the second carbon of the glucose moieties had a noisy isotopic signal with a large standard deviation and gave the poorest correlation with the oxygen isotope ratios of stem water. Related to this isotopic noise, we observed that the correlation between oxygen isotope ratios of phenylglucosazone with both stem water and relative humidity were higher than those observed for cellulose. Our hypothesis about tree-ring models which account for changes in the oxygen isotopic fractionation during cellulose synthesis was consistent only for the 18O content of phenylglucosazone. We showed that the tree-ring model based on the 18O content of phenylglucosazone was an improvement over existing models that are based on whole cellulose. Additionally, this approach may be used in other cellulose based archives such as peat deposits and lacustrine sediments.