Oxygen and hydrogen isotope ratios in tree rings: how well do models predict observed values?

We have measured annual oxygen and hydrogen isotope ratios in the α-cellulose of the latewood of oak (Quercus robur L.) growing on well-drained ground in Norfolk, UK. We compare the observed values of isotope ratios with those calculated using equations that allow for isotopic fractionation during the transfer of oxygen and hydrogen from source water taken by the tree to cellulose laid down in the cambium. The equations constitute a model in which isotopic fractionation occurs during evaporative enrichment within the leaf and during isotopic change between carbohydrates and water in the trunk during cellulose synthesis. From the relationship between isotope ratios in precipitation and α-cellulose, we deduce that the source water used by the tree comprises a constant mixture of groundwater and precipitation, chiefly from the months of May, June and July of the growth year. By selection of isotopic fractionation factors and the degree of isotope exchange within the trunk, we are able to model the observed annual values of oxygen isotope ratios of α-cellulose to a significant level (r=0.77, P<0.01). When we apply the same model to hydrogen isotope ratios, however, we find that, although we can predict the average value over the time series, we can no longer predict the year-to-year variation. We suggest that this loss of environmental signal in the hydrogen isotopes is caused by differences in the kinetic isotope effects of the biochemical reactions involved in the fixation of hydrogen in different positions of the glucose molecule. Owing to these effects, the hydrogen isotope ratios of cellulose can vary in a way not anticipated in current models and hence may induce non-climatic ‘noise’ in the hydrogen isotope time series.