Measurements of leaf sucrose to explain variability in hydrogen isotope composition of leaf cellulose

The hydrogen isotope composition (δ²H) of cellulose has been used to assess ecohydrological processes and carries metabolic information, adding new understanding to how plants respond to environmental change. However, experimental approaches to isolate drivers of δ²H variation is limited to the Yakir & DeNiro model (1990), which is difficult to implement and largely unvalidated. Notably, the two biosynthetic fractionation factors in the model, associated with photosynthetic (ε_A) and post-photosynthetic (ε_H) processes are currently accepted as constants, and the third parameter – the extent to which organic molecules exchange hydrogen (f_H) with local water – is usually tuned in order to resolve the difference between modelled and observed cellulose δ²H values. Thus, by virtue, the metabolically interpretable parameter is only f_H, whilst from theory, metabolic flux rates will also impact on the apparent fractionations. To overcome part of this limitation, we measured the δ²H of extracted leaf sucrose from fully-expanded leaves of seven species and a phosphoglucomutase ‘starchless’ mutant of tobacco to estimate the isotopic offset between sucrose and leaf water (ε_sucrose). Sucrose δ²H explained ~60% of the δ²H variation observed in cellulose. In general, ε_sucrose was higher (range: -203‰ to -114‰; mean: -151 ± 21‰) than the currently accepted value of -171‰ (ε_A) reflecting ²H-enrichment downstream of triose-phosphate export from the chloroplast, with statistical differences in ε_sucrose observed between species estimates. The remaining δ²H variation in cellulose was explained by species differences in f_H (estimated by assuming ε_H = +158‰). We also tested possible links between model parameters and plant metabolism. ε_sucrose was positively related to dark respiration (R²=0.27) suggesting an important branch point influencing sugar δ²H. In addition, f_H was positively related to the turnover time (τ) of water-soluble carbohydrates (R²=0.38), but only when estimated using fixed ε_A = -171‰. To decipher and isolate the “metabolic” information contained within δ²H values of cellulose it will be important to assess δ²H values of non-structural carbohydrates so that hydrogen isotope fractionation during sugar metabolism can be better understood. This study provides the first attempt at such measurements showing species differences in both source and sink processes are important in understanding δ²H variation of cellulose.