Hydrogen isotope fractionation in leaf waxes in the Alaskan Arctic tundra

Leaf wax hydrogen isotopes (delta D-wax) are increasingly utilized in terrestrial paleoclimate research. Applications of this proxy must be grounded by studies of the modern controls on delta D-wax, including the ecophysiological controls on isotope fractionation at both the plant and landscape scales. Several calibration studies suggest a considerably smaller apparent fractionation between source water and waxes (epsilon(app)) at high latitudes relative to temperate or tropical locations, with major implications for paleoclimatic interpretations of sedimentary delta D-wax. Here we investigate apparent fractionation in the Arctic by tracing the isotopic composition of leaf waxes from production in modern plants to deposition in lake sediments using isotopic observations of precipitation, soil and plant waters, living leaf waxes, and waxes in sediment traps in the Brooks Range foothills of northern Alaska. We also analyze a lake surface sediment transect to compare present-day vegetation assemblages to epsilon(app) at the watershed scale. Source water and epsilon(app) were determined for live specimens of Eriophorum vaginatum (cottongrass) and Betula nana (dwarf birch), two dominant tundra plants in the Brooks Range foothills. The delta D of these plants' xylem water closely tracks that of surface soil water, and reflects a summer-biased precipitation source. Leaf water is enriched by 23 +/- 15% relative to xylem water for E. vaginatum and by 41 +/- 19% for B. nana. Evapotranspiration modeling indicates that this leaf water enrichment is consistent with the evaporative enrichment expected under the climate conditions of northern Alaska, and that 24-h photosynthesis does not cause excessive leaf water isotope enrichment. The epsilon(app) determined for our study species average -89 +/- 14% and -106 +/- 16% for B. nana n-alkanes and n-acids, respectively, and -182 +/- 10% and -154 +/- 26% for E. vaginatum n-alkanes and n-acids, which are similar to the epsilon(app) of related species in temperate and tropical regions, indicating that apparent fractionation is similar in Arctic relative to other regions, and there is no reduced fractionation in the Arctic. Sediment trap data suggest that waxes are primarily transported into lakes from local (watershed-scale) sources by overland flow during the spring freshet, and so delta D-wax within lakes depends on watershed-scale differences in water isotope compositions and in plant ecophysiology. As such, the large difference between our study species suggests that the relative abundance of graminoids and shrubs is potentially an important control on delta D-wax in lake sediments. These inferences are supported by delta D-wax data from surface sediments of 24 lakes where epsilon(app), relative to delta D-xylem, averages -128 + 13% and -130 +/- 8% for n-acids and n-alkanes, respectively, and co-varies with vegetation type across watersheds. These new determinations of plant source water seasonality and epsilon(app) for the Arctic will improve the delta D-wax paleoclimate proxy at high latitudes. (C) 2017 Elsevier Ltd. All rights reserved.