Mercury in tundra vegetation of Alaska: Spatial and temporal dynamics and stable isotope patterns

Abstract

Vegetation uptake of atmospheric mercury (Hg) is an important mechanism enhancing atmospheric Hg deposition via litterfall and senescence. We here report Hg concentrations and pool sizes of different plant functional groups and plant species across nine tundra sites in northern Alaska. Significant spatial differences were observed in bulk vegetation Hg concentrations at Toolik Field station (52 ± 9 μg kg−1), Eight Mile Lake Observatory (40 ± 0.2 μg kg−1), and seven sites along a transect from Toolik Field station to the Arctic coast (36 ± 9 μg kg−1). Hg concentrations in non-vascular vegetation including feather and peat moss (58 ± 6 μg kg−1 and 34 ± 2 μg kg−1, respectively) and brown and white lichen (41 ± 2 μg kg−1 and 34 ± 2 μg kg−1, respectively), were three to six times those of vascular plant tissues (8 ± 1 μg kg−1 in dwarf birch leaves and 9 ± 1 μg kg−1 in tussock grass). A high representation of nonvascular vegetation in aboveground biomass resulted in substantial Hg mass contained in tundra aboveground vegetation (29 μg m−2), which fell within the range of foliar Hg mass estimated for forests in the United States (15 to 45 μg m−2) in spite of much shorter growing seasons.Hg stable isotope signatures of different plant species showed that atmospheric Hg(0) was the dominant source of Hg to tundra vegetation. Mass-dependent isotope signatures (δ202Hg) in vegetation relative to atmospheric Hg(0) showed pronounced shifts towards lower values, consistent with previously reported isotopic fractionation during foliar uptake of Hg(0). Mass-independent isotope signatures (Δ199Hg) of lichen were more positive relative to atmospheric Hg(0), indicating either photochemical reduction of Hg(II) or contributions of inorganic Hg(II) from atmospheric deposition and/or dust. Δ199Hg and Δ200Hg values in vascular plant species were similar to atmospheric Hg(0) suggesting that overall photochemical reduction and subsequent re-emission was relatively insignificant in these tundra ecosystems, in agreement with previous Hg(0) ecosystem flux measurements.