Liquid water has been known to occur beneath the Antarctic ice sheet for more than 40 years, but only recently have these subglacial aqueous environments been recognized as microbial ecosystems that may influence biogeochemical transformations on a global scale. Here we present the first geomicrobiological description of water and surficial sediments obtained from direct sampling of a subglacial Antarctic lake. Subglacial Lake Whillans (SLW) lies beneath approximately 800 m of ice on the lower portion of the Whillans Ice Stream (WIS) in West Antarctica and is part of an extensive and evolving subglacial drainage network. The water column of SLW contained metabolically active microorganisms and was derived primarily from glacial ice melt with solute sources from lithogenic weathering and a minor seawater component. Heterotrophic and autotrophic production data together with small subunit ribosomal RNA gene sequencing and biogeochemical data indicate that SLW is a chemosynthetically driven ecosystem inhabited by a diverse assemblage of bacteria and archaea. Our results confirm that aquatic environments beneath the Antarctic ice sheet support viable microbial ecosystems, corroborating previous reports suggesting that they contain globally relevant pools of carbon and microbes that can mobilize elements from the lithosphere and influence Southern Ocean geochemical and biological systems.

[1] Subglacial lakes in East Antarctica can be separated into four categories specified by radar reflection properties. Definite lakes are brighter than their surroundings by at least 2 dB (relatively bright) and both are consistently reflective (specular) and have a reflection coefficient greater than À10 dB (absolutely bright). Dim lakes are relatively bright and specular but not absolutely bright, indicating nonsteady ice dynamics. Fuzzy lakes are both relatively and absolutely bright, but not specular, and may indicate saturated sediments or ''swamps.'' Indistinct lakes are absolutely bright and specular but no brighter than their surroundings. Lakes themselves and the different classes of lakes are not arranged randomly throughout Antarctica but are clustered around ice divides, ice stream onsets, and prominent bedrock troughs, with each cluster demonstrating a different characteristic lake classification distribution. The lake classification algorithm expands on previous studies and demonstrates a novel way to characterize icewater interactions in East Antarctica.

The dynamic response of the Greenland Ice Sheet (GrIS) depends on feedbacks between surface meltwater delivery to the subglacial environment and ice flow. Recent work has highlighted an important role of hydrological processes in regulating the ice flow, but models have so far overlooked the mechanical effect of soft basal sediment. Here we use a threedimensional model to investigate hydrological controls on a GrIS soft-bedded region. Our results demonstrate that weakening and strengthening of subglacial sediment, associated with the seasonal delivery of surface meltwater to the bed, modulates ice flow consistent with observations. We propose that sedimentary control on ice flow is a viable alternative to existing models of evolving hydrological systems, and find a strong link between the annual flow stability, and the frequency of high meltwater discharge events. Consequently, the observed GrIS resilience to enhanced melt could be compromised if runoff variability increases further with future climate warming.