Influence of physiology and climate on δD of leaf wax n-alkanes from C3 and C4 grasses

We measured hydrogen isotope compositions (613) of high-molecular-weight n-alkanes (C-27-C-33) from grasses grown in greenhouses and collected from the US Great Plains. In both cases, n-alkanes from C-4 grasses are enriched in D by more than 20 parts per thousand relative to those from C-3 grasses. The apparent enrichment factor (epsilon C-29-GW) between C-29 n-alkane and greenhouse water is -165 +/- 12 parts per thousand for C-3 grasses and -140 +/- 15%. for C-4 grasses. For samples from the Great Plains, delta D values of C-29 n- alkanes range from -280 to -136 parts per thousand with values for C-4 grasses ca. 21%. more positive than those for C-3 grasses from the same site. Differences in C-3 and C-4 grass n-alkane delta D values are consistent with the shorter interveinal distance in C-4 grass leaves, and greater backdiffusion of enriched water from stomata to veins, than in C-3 grass leaves. Great Plains' grass n-alkane isotopic ratios largely reflect precipitation delta D values. However, the offset or apparent fractionation between n-alkanes and precipitation is not uniform and varies with annual precipitation and relative humidity, suggesting climatic controls on lipid delta D values. The dryer sites exhibit smaller absolute apparent fractionation indicative of D-enrichment of source waters through transpiration and/or soil evaporation. To explore the relationship between climate and n-alkane delta D values, we develop three models. (1) The 'direct analog' model estimates delta Dc(29) values simply by applying the apparent enrichment factors, CC29-GW, observed in greenhouse grasses to precipitation delta D values from the Great Plains. (2) The 'leaf-water' model uses a Craig-Gordon model to estimate transpirational D-enrichment for both greenhouse and field sites. The transpiration-corrected enrichment factors between C-29 and bulk leaf-water, epsilon C-29-GW, calculated from the greenhouse samples (-181%, for C-3 and -157%. for C-4) are applied to estimate delta D-C29 values relative to modeled bulk leafwater delta D values. (3) The 'soil- and leaf-water' model estimates the combined effects of soil evaporation, modeled by analogy with a flow-through lake, and transpiration on delta D-C29 values. Predictions improve with the addition of the explicit consideration of transpiration and soil evaporation, indicating that they are both important processes in determining plant lipid delta D values. D-enrichment caused by these evaporative processes is controlled by relative humidity, suggesting that important climatic information is recorded in leaf wax n-alkane delta D values. Calibration studies such as this one provide a baseline for future studies of plantwater-deuterium systematics and form the foundation for interpretation of plant wax hydrogen isotope ratios as a paleo-aridity proxy. (c) 2005 Elsevier Inc. All rights reserved.