ABSTRACT: New perspectives on the NW Laurentide ice Sheet: implications for the NE extremity of Beringia, paleoclimate and the role of episodic ice shelves during deglaciation.

CGRG Bibliography of Canadian Geomorphology
Search Results

Author :

Date :

Title :

Publication :

Issue :

Page(s) :

Abstract

For the past half century, maps of NW North America during the Last Glacial Maximum (LGM) have portrayed widespread ice-free land in the western Canadian Arctic Archipelago lying distal to the limit of the Laurentide Ice Sheet (LIS). That these ice-free areas may constitute the NE limit of Beringia was recently proposed by Harrington (2005) who reported two solitary finds of mammoth tusk and bone, both dated ~22 ka BP, on Banks and Melville islands, respectively. Accordingly, these islands constitute the core of the proposed ice-free LGM terrain (Vincent 1982, 1983; Hodgson et al. 1984). Indeed, the NW corner of Banks Island purportedly contains a never-glaciated landscape, proposed to pre-date the Bruhnes Normal chron (>780 ka BP) during which the oldest and most extensive Banks Glaciation terminated along its southern flank. Two subsequent glaciations on Banks Island are shown to be progressively less extensive and both have been assigned to the pre-Late Wisconsinan (preserved distal to Late Wisconsinan ice). Similarly, to the north, on Melville Island, two undated Laurentide till sheets (Dundas and Bolduc tills) have been mapped beyond the limit of the Late Wisconsinan Winter Harbour Till (~10.4 ka BP; Hodgson and Vincent 1984). Here we report on fieldwork conducted across S. Melville Island during 2002–2004 that was extendedsouthward across intervening M’Clure Strait to N. Banks Island in 2005. These data require a fundamental revision of previous models concerning past advances of the NW Laurentide Ice Sheet into the western Canadian Arctic Archipelago (CAA), including its chronology, extent, dynamics and relative sea level adjustments. The maximum extent of Laurentide ice is marked by Dundas Till that crosses Dundas Peninsula (300 m asl) northward to Liddon Gulf. Laurentide ice coalesced with Melville ice exiting the Gulf, forming a trunk glacier that advanced westward through M’Clure Strait where granite erratics are found up to 275 m asl. Given adjacent water depths of 500 m in the Strait (~560 m including glacioisostatic depression) this thickness of Laurentide icewould have been fully grounded. This is consistent with flutings on the sea floor immediately to the east that are aligned with the Strait and support a M’Clure Strait Ice Stream (western limit undefined). Forty AMS radiocarbondates were obtained on individual shell fragments collected from Dundas and Bolduc tills on Dundas Peninsula. These dates indicate a common Late Wisconsinan age for the most extensive Laurentide advance across thepeninsula. Bolduc Till occurs downslope from Dundas Till, and is shell-rich and shares a similar range of ages on its ice-transported shells. It is interpreted to mark the initial arrival of the NW Laurentide Ice Sheet onto this coastline as an ice shelf (to ~100 m asl). Bolduc Till was subsequent overridden by grounded Laurentide ice that subsumed Dundas Peninsula, depositing abundant far-travelled granite erratics (from mainland Canada) as part of the coarser Dundas Till. Along the Melville Island coastline of M’Clure Strait, as well as around the perimeter of Dundas Peninsula, 41 AMS radiocarbon dates on deglacial marine limit are all immediately pre-Holocene, indicating Late Wisconsinan retreat. Furthermore, marine limit rises progressively to the southeast along M’Clure Strait, climbing in the direction of preceding Laurentide ice advance. The reinterpretation of the Late Quaternary record on Melville Island, favouring extensive Late Wisconsinan Laurentide ice, contrasts with the proposed ice-free conditions throughout most of adjacent Banks Island during the Late Quaternary. To further test the Melville Island revision, fieldwork was extended to N. Banks Island in 2005. Emphasis was placed on establishing the elevations and ages of deglacial sea levels that remained poorly documented along purported Laurentide ice margins previously assigned to multiple glaciations. Southward (inland) retreat from the north-central coast of Banks Island is recorded by a previously unreported marine limit at the head of Mercy Bay. This ice configuration is equivalent to Vincent’s (1982, 1983) Thomsen Glaciation, purportedly predating the last interglacial (MIS 5e; Sangamonian, locally Cape Collinson). Marine sand extends downslope from marine limit (Mercy Bay) where it overlies fossiliferous mud at 1 m asl. Both Cyrtodaria kurriana and Hiatella arctica collected from this unit provide five concordant AMS dates immediately prior to the Holocene. Furthermore, all available radiocarbon dates along the Banks Island coast of M’Clure Strait are associated with marine limits whose elevation, gradient and age match those on the opposing coast of Melville Island. Collectively, this records synchronous retreat of the previous M’Clure Strait Ice Stream. Along the opposing coasts of Melville and Banks islands, a low-elevation till sheet (Winter Harbour Till) and corresponding moraines (< 100 m asl) mark the late glacial readvance of the Viscount Melville Sound Ice Shelf. These deposits mark the limit of a 400 km advance of Laurentide ice from its previously grounded margin on central Victoria Island (Hodgson 1993). There, streamlined bedforms mark a former ice stream that inundated NE Victoria Island, feeding into the Viscount Melville Sound Ice Shelf that advanced to eastern M’Clure Strait 100,000 km2. The age of the ice shelf was previously debated (10.4 to 9.6 ka BP, Hodgson and Vincent 1984 vs. 11.3–11.0 ka BP, Dyke 1987). Fossiliferous marine sediments over-run by the ice shelf at widely separated sites show concordant dates for ice shelf arrival that support Hodgson and Vincent 1984 (at least 20 dates are indistinguishable within the standard errors of their reported ages). Radiocarbon dates constraining ice shelf chronology also include distal marine sediments tectonized by the advance as well as shells in growth position living on its till and moraines after abandonment. The ice shelf broke up by 10 ka BP, based on several deglacialdates. Outflow to maintain this ice shelf, together with its subsequent breakup that promoted further drawdown, served to exhaust the M’Clintock Ice Divide of the NW Laurentide Ice Sheet, and by 9.2 ka BP the sea had gutted its axis, penetrating southward to the mainland. Conclusion and implications based on this revision include: 1) the NW LIS crossed N. Banks Island and S. Dundas peninsula during the LGM, filling at least eastern M’Clure Strait and possibly advancing farther west to the polar continental shelf; 2) a more extensive margin for the NW LIS in the western CAA harmonizes this record with maximum LGM extent for the LIS to the south where it crossed the Mackenzie Mountains to impound proglacial lakes in the Yukon (= 22 ka BP; Morlan et al. 1990); 3) the extent of ice recognized on N. Banks and S. Melville islands during the LGM precludes the extension of NE Beringia into the same area asrecently proposed (Harrington 2005); 4) this vigorous outflow of Laurentide ice across what was considered to be an arid and ice-free landscape lends support for a strong moisture source consistent with the hypothesis of a split jet stream whose northern branch advected storm tracks from the Pacific into the CAA (Bromwich et al. 2002); 5) during deglaciation, the Pacific mollusc C. kurriana arrived in N. Banks Island by at least 11.5 14C ka BP, suggesting the resubmergence of Bering Strait at least 1000 years earlier than previous estimates (Elias et al. 1996; Keigwin et al. 2006), with important dynamical and paleoclimatic implications for the Arctic Ocean and the Younger Dryas (Bradley and England 2007, this volume); 6) the catastrophic readvance of the NW LIS, forming the Viscount Melville Sound Ice Shelf (10.4 to 10.0 ka BP), corresponds to the end of the Younger Dryas and marks a glaciological event that immediately precedes and likely triggered the removal of the marine-basedM’Clintock Ice Divide, possibly providing an important analogue for the modern day West Antarctic Ice Sheet under a warming global climate.

Bibliography of Canadian Geomorphology