Organic carbon export from the Greenland ice sheet

Abstract

Glacial meltwater exports a unique type of organic carbon to marine systems, distinct from non-glacially derived riverine export, potentially capable of stimulating downstream marine primary productivity. Here, we describe for the first time the bulk-level dissolved organic carbon (DOC) and particulate organic carbon (POC) isotopic composition of glacial meltwater draining the Greenland ice sheet (GrIS). These data, in conjunction with an earlier study that investigated the molecular-level composition of GrIS dissolved organic matter, collectively describe the concentration, radiocarbon content, and lability of organic carbon in subglacial discharge from a land-terminating outlet glacier during a melt season. By scaling up our measurements across the ice sheet, we estimate that the annual DOC flux from the GrIS (0.08 Tg/y) is equivalent to that from a small Arctic river (discharge (Q) < 50 km³/y), and that the annual POC flux from the GrIS (0.9 Tg/y) may be comparable to that of a large Arctic river (Q > 200 km³/y). The DOC flux is derived primarily from beneath the glacier (subglacial) (>75%) in the early season, and from surface ice-melt (up to 100%) transmitting through the subglacial environment at the peak of the meltseason. The POC flux is primarily derived from the subglacial environment throughout the meltseason. The early season (low flow) glacier discharge contains higher DOC concentrations (0.5–4.1 mg L⁻¹), and exports more enriched carbon (Δ¹⁴C_DOC ∼ −250‰) compared to the peak season (high flow) discharge, when the concentrations are lower (0.1–0.6 mg L⁻¹) and the Δ¹⁴C is more depleted (Δ¹⁴C_DOC ∼ −400‰). Conversely, the POC export (1.4–13.2 mg L⁻¹, Δ¹⁴C_POC ∼ −250‰) shows no temporal variation in either concentration or radiocarbon content throughout the meltseason. Dissolved ion loads in concomitant samples reflected the seasonal evolution of the subglacial drainage system, confirming the influence of subglacial hydrology on the composition of the bulk carbon pools. Based on this work, we conclude that (1) different mechanisms control the DOC and POC flux from glacial systems; (2) chemically-distinct DOC pools are accessed by seasonally-evolving hydrological flow-paths; and (3) the GrIS can deliver labile carbon, which may also be ¹⁴C-depleted, to downstream proglacial and marine environments.

Introduction

Glacial environments possess a dynamic and reactive carbon system (Hood and Scott, 2008, Hood et al., 2009, Pautler et al., 2011). From a glacial–interglacial perspective, in situ microbial metabolism of subglacial organic carbon beneath large continental (e.g. the Laurentide) ice sheets could produce CO₂ and CH₄ (Skidmore et al., 2000, Wadham et al., 2008) that may have been released following deglaciation. From a present-day perspective, recent work reveals that modern glacier runoff along the Gulf of Alaska (GOA) is capable of exporting ancient, labile dissolved organic carbon to surrounding coastal ecosystems (Hood et al., 2009). This hypothesis by Hood et al. (2009) has important implications for the coastal waters surrounding Greenland, where the GrIS contributed an estimated ∼400 km³ meltwater runoff in 2010 (Bamber et al., 2012) comparable to the average annual discharge from a large Arctic river (e.g. the Ob River; Dittmar and Kattner, 2003). Despite the fact that GrIS runoff is increasing, particularly to the North Atlantic Ocean (Bamber et al., 2012), there are very few studies of organic carbon export from large ice sheets.

Extant studies have focused primarily on end-member carbon pools found on the ice sheet surface and the bed, rather than on bulk meltwater runoff. These studies reveal that in comparison to riverine, marine, and estuarine environments, organic carbon from the ice sheet surface (i.e. supraglacial snow, ice, and meltwater) and base (basal ice) is nitrogen-rich, containing proteinaceous and other biologically-derived compounds (Bhatia et al., 2010, Dubnick et al., 2010, Pautler et al., 2011). The source of these compounds is presumed to be in situ microbial communities on the glacier surface and at the ice-bed interface (Carpenter et al., 2000, Skidmore et al., 2000, Bhatia et al., 2006, Hodson et al., 2008). Other recent studies have suggested that aerosol particles, specifically anthropogenic combustion products (Stubbins et al., 2012), deposited on the glacial surface, are another potential source of organic material to glacial systems (Stibal et al., 2008). A recent review of glacial ice dissolved organic matter (DOM) from 26 glaciers in the European Alps, however, suggests that combustion products are not quantitatively significant in glacially-derived DOM, at least for that region (Singer et al., 2012). Clearly, the origin and nature of the labile component of glacially-derived DOM remains unresolved.

Recent studies have shown that the majority of the meltwater draining the Greenland ice sheet drains first to the bed, and is then discharged via a seasonally-evolving subglacial (beneath the ice) drainage system (Das et al., 2008, Bartholomew et al., 2010, Bhatia et al., 2011). Thus, any study of the contribution of carbon from ice sheets to surrounding coastal oceans needs to be undertaken within the proper context for meltwater outflow (hydrology and volume). Previous work on Alaskan glaciers indicates that glacial runoff possesses a radiocarbon (¹⁴C)-depleted signature (Hood et al., 2009). Compositional studies of runoff organic carbon have highlighted the presence of both terrestrial and microbial-derived components (Barker et al., 2006, Bhatia et al., 2010, Dubnick et al., 2010). One potential explanation for these observations is that terrestrial material may be derived from overridden soils and vegetation while the presence of uniquely-adapted subglacial microbes at the glacier bed may provide metabolic by-products (Sharp et al., 1999, Skidmore et al., 2005, Cheng and Foght, 2007). The metabolite pool would retain a ¹⁴C-depleted signature if these organisms utilize relict organic carbon sources (Petsch et al., 2001).

Previously, we investigated glacially-derived organic carbon using molecular-level analyses (i.e. ultra-high resolution mass spectrometry; (Bhatia et al., 2010)), providing insight into DOM source and lability. Here, we combine these observations with bulk-level analyses of abundance and radiocarbon content of organic carbon in glacial meltwater to establish a comprehensive description of cycling and export of organic carbon from the Greenland ice sheet. We examine both the dissolved and particulate organic carbon (DOC, POC) pools, since they are likely influenced by different dynamics in the subglacial system, and have different fates in the marine environment. We combine prior results from an isotope mixing model (Bhatia et al., 2011) with the major ion chemistry to gain insight into the evolution of the subglacial drainage system, and the dominant subglacial chemical weathering regimes at our study site. With these results, we test the hypothesis that glacier hydrology plays an important role in dictating the type (concentration, radiocarbon content, source, lability) of organic matter released from glacial systems.