Late Wisconsinan Glaciation of Southern Eureka Sound: Evidence for Extensive Innuitian Ice in the Canadian High Arctic During Th

Quaternary Science Reviews 19 (2000) 1319}1341

Late Wisconsinan glaciation of southern Eureka Sound: evidence for extensive Innuitian ice in the Canadian High Arctic during the Last Glacial Maximum Colm OD Cofaigh! *, John England!, Marek Zreda"

!Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Alberta, Canada T6G 2E3 "Department of Hydrology and Water Resources, University of Arizona, Tucson, Arizona 85721, USA

Abstract

Southern Eureka Sound was originally proposed as the centre of an Innuitian Ice Sheet in the Canadian High Arctic at the Last Glacial Maximum (LGM) based largely on the pattern of Holocene emergence. This paper focuses on the glacial geological evidence for such an ice sheet in the region. Granite dispersal trains and ice-moulded bedrock record regional, westward #ow of warm-based ice into Eureka Sound from SE Ellesmere Island. Regional ice was coalescent with local ice domes on inter-"ord peninsulas. Marine limit in the form of raised deltas, beaches and washing limits formed during deglaciation of the regional ice. Throughout southern Eureka Sound, marine limit dates )9.2 ka BP, indicating that ice commenced retreat during the early Holocene. Ice-divides were located along the highlands of central Ellesmere and Axel Heiberg islands, from which ice inundated Eureka Sound, #owing north and south along the channel. Regional radiocarbon dates on marine limit show that deglaciation occurred in two steps. Initial break-up and radial retreat of ice from Eureka Sound to the inner "ords was rapid and preceded stabilisation along adjacent coastlines and at "ord heads. Two-step deglaciation is also re#ected in di!erences in glacial geomorphology between the inner and outer parts of many "ords. A prominent belt of "ord-head glaciogenic landforms, long proposed to mark the last glacial limit, is re-interpreted to record initial, stabilisation of ice margins due predominantly to bathymetric control. ( 2000 Elsevier Science Ltd. All rights reserved.

1. Introduction (England, 1976). Most previous Quaternary investigations are from the northern part of the sound, and these This paper is the "rst detailed reconstruction of the have advocated a restricted Late Wisconsinan ice cover Late Wisconsinan glacial history of southern Eureka (England, 1987, 1990, 1992; Bell, 1992, 1996), although Sound, Queen Elizabeth Islands, Arctic Canada (Figs. 1 more recent work rejects that interpretation (Bednarski, and 2). Blake (1970) originally proposed the existence of 1998; England and OD Cofaigh, 1998; OD Cofaigh, 1998, a Late Wisconsinan Innuitian Ice Sheet in this region 1999a, b). based mainly on a corridor of maximum Holocene The paper focuses on the Ellesmere Island side of emergence extending from Eureka Sound to Bathurst southern Eureka Sound, but also includes Stor Island in Island. He argued that this emergence recorded the re- the central part of the channel (Figs. 1 and 2). Emphasis is moval of a pan-archipelago Innuitian Ice Sheet, which placed on Late Wisconsinan ice con"guration, dynamics was coalescent with the Laurentide Ice Sheet to the south and chronology. This reconstruction is based on sur"cial and the Greenland Ice Sheet to the northeast. A persist- mapping and sedimentological investigation of gla- ent problem with this reconstruction has been the lack of ciogenic and raised marine deposits and landforms. direct stratigraphic and chronologic evidence for such Chronological control is provided by radiocarbon dating an ice sheet along its proposed axis in Eureka Sound of marine shells and driftwood. In this paper the term Last Glacial Maximum (LGM) refers to the period of maximum ice cover, prior to the * Corresponding author. Current address: Bristol Glaciology Centre, School of Geographical Sciences, University of Bristol, Bristol BS8 1SS, onset of deglaciation. On western Ellesmere and Axel UK. Tel.: #44-117-928-9954; fax: #44-117-928-7878. Heiberg islands, radiocarbon dates on shell fragments E-mail address: [email protected] (C. OD Cofaigh). from till and outwash indicate that the LGM occurred

0277-3791/00/$- see front matter ( 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 2 7 7 - 3 7 9 1 ( 9 9 ) 0 0 1 0 4 - 3 1320 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Fig. 1. The Queen Elizabeth Islands and northwest Greenland showing location of the study area (box) and contemporary ice cover (dark shading).

Fig. 2. Southern Eureka Sound showing contemporary ice cover (dark sometime after 28}27 ka BP (Bednarski, 1998; this pa- shading) and location of placenames referred to in text. per), and possibly (20 ka BP based on an AMS date on sub-till organic detritus (Blake, 1992a). On eastern Elles- mere Island, AMS dates on shell fragments in till indicate sive ice"elds on Ellesmere and Axel Heiberg islands that the LGM may have occurred (19 ka BP (England, (Figs. 1 and 2). 1999). Throughout Ellesmere Island, deglaciation was underway by 10}11 ka BP (Hodgson et al., 1991; Be- 1.2. Previous work dnarski, 1995, 1998; Blake et al., 1996; England, 1998). Early investigations into the glacial history of Eureka 1.1. Regional setting Sound noted crystalline erratics (granite, gneiss and quartzite) derived from a source to the east, probably Eureka Sound separates Ellesmere and Axel Heiberg under the Prince of Wales Ice"eld (Troelsen, 1952; Fyles, islands, and extends from Norwegian Bay to Nansen in Jenness, 1962; Tozer, 1963), as well as evidence for high Sound (Figs. 1 and 2). Stor Island occupies the central relative sea levels (Schei, 1904; Farrand and Gadja, 1962). part of the channel (Fig. 2). Geologically the study area is Several workers (e.g., Fyles, in Jenness, 1962; Boesch, dominated by NE striking sedimentary rocks with ig- 1963; Hattersley-Smith, 1969) also noted elevational dif- neous dykes (Trettin, 1991). Precambrian granite out- ferences in the degree of weathering and preservation of crops 60 km to the east and underlies the Prince of Wales glacial landforms, and postulated that these re#ected Ice"eld (Fig. 2). Uplands in the study area reach multiple glaciations, with the last being thinner and more '1000 m asl and are dissected by "ords and valleys, restricted. These early observations broadly de"ne con- aligned both parallel to bedrock structure (e.g., Trold trasting reconstructions over the last three decades con- Fiord) and cross-cutting it (e.g., Bay Fiord). Contempor- cerning the extent of Late Wisconsinan glaciation in ary ice cover is limited to small, upland ice caps, although the Queen Elizabeth Islands. These have ranged from the study area is bordered to the east and west by exten- that of an extensive Innuitian Ice Sheet (e.g., Blake, 1970, C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1321

1972, 1992b, 1993; Blake et al., 1992; Tushingham, 1991; to fresh. However, no elevational di!erences were noted HaK ttestrand and Stroeven, 1996; Bednarski, 1998; Dyke, in weathering characteristics. Granite erratics and ice- 1998, 1999; England, 1998, 1999; Zreda et al., 1999), to moulded bedrock extend unevenly westwards across the a much more restricted ice cover, the Franklin Ice Com- study area from the margin of the Prince of Wales Ice"eld plex (e.g., England, 1976, 1983, 1987, 1990, 1992, 1996; (Figs. 3}5). At many sites, granites form part of a silty England and Bradley, 1978; Bednarski, 1986; Lemmen, diamict that also contains shell fragments. This diamict is 1989; Evans, 1990; Bell, 1996). interpreted as till on account of its regional continuity Eureka Sound has been a key area in this debate, and above marine limit, the presence of striated erratics, and illustrates the issues around which these contrasting re- its stratigraphic position overlying ice-moulded and stri- constructions have centred: ated bedrock. Granite-bearing Tertiary #uvial deposits have also (1) Postglacial emergence of up to 150 m along Eureka been reported on western Ellesmere Island (e.g., Fyles, Sound, and whether this represents a solely gla- 1989, 1990; Hodgson et al., 1991; Bell, 1992). They are cioisostatic response to the removal of a regional ice mostly sandy, but localised beds of rounded pebble to sheet (e.g., Blake, 1970) or that of a smaller ice load cobble gravel outcrop at the head of Strathcona Fiord with a possible neotectonic contribution (e.g., Eng- and contain granite (Fyles, 1989; John Fyles, pers. land, 1997). comm., 1998). These gravels lay in the path of ice advanc- (2) The signi"cance of a `drift belta (Hodgson, 1985) of ing westward through Bay Fiord (Fig. 2) from the Prince glacial landforms and sediments at many "ord heads of Wales Ice"eld, and therefore were likely to have been (e.g., Lemmen et al., 1994; OD Cofaigh, 1998), and partly glacially eroded and re-deposited. However, the whether this marks the limit of Late Wisconsinan #uvial gravels contrast texturally with the overlying re- glaciation, or a stillstand during retreat of a more gional till (2}3 m thick) which contains striated granite extensive ice cover. boulders up to 3 m in diameter (cf. Fyles, 1989; Hodgson (3) A sparsity of fresh glacial landforms beyond this drift et al., 1991; Bell, 1992). Therefore, granite erratics record belt in several locations (e.g., England, 1987; Lemmen glacial transport from the Prince of Wales Ice"eld, as well et al., 1994; OD Cofaigh, 1998), coupled with a greater as a subsidiary component of pebble to cobble-sized degree of weathering in many outer "ords and at clasts from Tertiary #uvial deposits. higher elevations (Boesch, 1963; England, 1987; Bell, The granites form a major dispersal train centred on 1996). Bay Fiord, extending westwards from the Prince of (4) The age of granite erratics and shelly till beyond the Wales Ice"eld to Stor Island and beyond (Figs. 3}5). drift belt which demonstrate past glacial inundation Along outer Bay Fiord, granites occur on summits reach- of "ords and inter-island channels. These have been ing 722 m asl, and extend as far west as Cape Chase (Fig. variously assigned to either the LGM or pre-Late 3). They also occur close to the summit of Stor Island Wisconsinan glaciation. (480 m asl) (Troelsen, 1952). Granites were observed on More recently, a consensus is emerging, based on "eld- the highest summit (764 m asl) separating Bay and Vesle work at several locations in the Queen Elizabeth Islands, Fiords (Fig. 3), and on uplands immediately north of the in favour of extensive ice during the LGM (Blake, 1992b; mouth of Vesle Fiord. The northward extension of this Bednarski, 1998; Dyke, 1998, 1999; England, 1998, 1999; dispersal train along the east coast of Axel Heiberg Is- Zreda et al., 1995, 1999; OD Cofaigh, 1998, 1999b). This land and western Fosheim Peninsula (Fig. 4) was mapped paper strengthens this consensus by demonstrating the by Bell (1992), who correctly suggested that it emanated style and chronology of glaciation in the previously from Bay Fiord. unstudied, southern part of Eureka Sound during the Abundant granites occur to at least 500 m asl in up- LGM. Emphasis is placed on regional granite erratic lands north of Trold Fiord (Fig. 3). By contrast, between dispersal trains that are the product of westward-#owing the head of this "ord and Star"sh Bay, only one granite Ellesmere Island ice into Eureka Sound. was observed. Granites were not observed on summits separating Trold and Blind "ords. However, they occur sparsely west of these "ord mouths (Fig. 3). Along sum- 2. Results mits bordering Eureka Sound, granites were observed only at Cape Chase, Trapper's Cove, and between Hare 2.1. Distribution of erratics Bay and Blind Fiord (Fig. 3). At the head of Trapper's Cove, one granite was found on a recessional moraine 2.1.1. Raanes Peninsula and Bay Fiord 15 m above marine limit, but they are common below The most conspicuous erratic in the study area is marine limit. granite. Granites range from pebbles to angular boulders In the southern part of the study area, granites are ('3 m in diameter) and many are striated. They exhibit ubiquitous above marine limit in Star"sh and Jaeger varying degrees of weathering, from partially grussi"ed bays (Fig. 3), and are common on intervening summits 1322 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Fig. 3. Distribution of granite erratics above marine limit and ice-#ow directional indicators (striae, #utings, ice-moulded bedrock) in the study area. Each circle represents a control point where granites were either observed (closed circle) or not observed (open circle). Elevation of key erratics in italics. Distribution of granitic till `Ga is from Hodgson (1979).

(up to 1036 m asl). Widespread granitic till extends east- extends across Braskeruds Plain to the contemporary ward to the margin of the Prince of Wales Ice"eld (Fig. 4; ice-margin (Hodgson, 1973, 1979) (Fig. 4). Hodgson, 1979). In summary, two granite dispersal trains are recognised. One is centred along Bay Fiord (Fig. 4), and 2.1.2. Adjacent areas extends northward along the coast of Axel Heiberg Is- South of Baumann Fiord, D. Hodgson (pers. comm., land and Fosheim Peninsula to Nansen Sound (Fyles, in 1996) recorded granite erratics above marine limit on Jenness, 1962; Bell, 1992). The other extends through southern Bjorne Peninsula (Fig. 2). Hodgson (1979) map- Star"sh and Jaeger bays, and discontinuously along the ped discontinuous granitic till between Makinson Inlet north side of outer Baumann Fiord (Fig. 4). The south and Baumann Fiord, as well as extensive areas of granitic side of this dispersal train is unde"ned. However, based till '300 m asl along the west and east sides of Vendom on the occurrence of granite erratics to the south, and Fiord (Fig. 2). Blake (1978) also recorded granite erratics between Makinson Inlet and Baumann Fiord (Hodgson, on Paleozoic carbonate bedrock in Makinson Inlet. 1973, 1979, pers. comm., 1996), it is likely that it lies south Granitic till can be traced to Vendom Fiord head and of Baumann Fiord and Bjorne Peninsula. C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1323

Fig. 4. Granite dispersal trains, western Ellesmere and eastern Axel Heiberg islands (grey stiple), distribution of granite erratics above marine limit throughout western Arctic islands (`Ga), and extent of Canadian Shield rocks on Ellesmere and Devon islands (black shading). Sources of erratic data (including this paper): St.-Onge (1965); Balkwill et al. (1974); Hodgson (1973, 1979, 1982, 1990, pers. comm., 1996); Bell (1992); Bednarski (1996); Dyke (1999). Source of bedrock data: Trettin (1991).

2.2. Glacial geomorphology and stratigraphy monly overlain by till containing striated and faceted clasts (Fig. 5). Below marine limit, the till is typically The pattern of glacial landforms in the study area is overlain by fossiliferous raised marine sediment in the illustrated in Fig. 6. Key aspects are highlighted in the form of beaches, deltas and blankets of silt (Fig. 7), following sections which describe "rstly Raanes and particularly at "ord heads. The silt commonly exceeds Svendsen peninsulas, followed by Bay Fiord and Stor 10 m in thickness and extends over several km. Island. West of Trold Fiord, lateral meltwater channels record concentric retreat towards the interior of Raanes Penin- 2.2.1. Raanes and Svendsen peninsulas sula (Fig. 6). Channels are particularly well developed in Fresh, "ord-parallel striae and roches moutonneH es on strike-aligned "ords and valleys where they grade to carbonate bedrock record westward-#owing trunk gla- raised deltas and beaches marking marine limit (Fig. 7). ciers in Star"sh and Jaeger bays (Fig. 6). A southwest- Lateral meltwater channels incised into till and bedrock ward de#ection in the orientation of striae at the mouth are also well developed on northern Raanes Peninsula, of Star"sh Bay implies coalescence with trunk ice in where they demonstrate onshore ice retreat, with de- Trold Fiord. At the mouth of Trold Fiord, striae indicate glacial #owlines generally perpendicular to the present that this coalescent ice then #owed into Baumann Fiord coastline (Fig. 6). where it too was further de#ected (to the SSW, Fig. 6). Ice-marginal landforms are rare along Trold Fiord Throughout the "eld area, ice-moulded bedrock is com- because of its cli!ed coastline. However, the passage of 1324 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Glacial landforms along outer Baumann Fiord, immediately west of Blind Fiord, are limited to proglacial meltwater channels that descend to marine limit. Locally, shelly till forms a patchy veneer on uplands. Sandstone tors, felsenmeer, and fresh to grussi"ed gabbro erratics of local provenance occur throughout the area. Well developed #ights of raised beaches (Fig. 7) extend to washing limits marking marine limit (138}142 m asl). Thick (granitic) till is limited to northern Raanes Pen- insula bordering Bay Fiord and Eureka Sound. Lime- stone-rich till containing granite erratics is common throughout Star"sh and Jaeger bays. Elsewhere, till is present as a patchy veneer, or limited to sparse erratics. Only one basal till was ever observed in stratigraphic section, and this always constitutes the lowermost unit.

2.2.2. Bay Fiord Bay Fiord extends eastwards from Stor Island and branches into three tributaries: Strathcona Fiord, and Irene and Augusta bays (Fig. 2). A prominent belt of ice-contact landforms and sediments at the heads of these tributaries (Hodgson, 1985) marks an ice-marginal position during deglaciation. Ice-moulded and striated carbonate bedrock, recording westward ice-#ow, occurs along Bay Fiord distal to the drift belt. Ice-moulded bedrock is commonly overlain by granitic till. A similar ice-#ow is also recorded by #uted till along the north shore of inner Bay Fiord (Fig. 6). Di!erences in the weathering of ice-moulded bedrock between the "ord heads of the tributaries and outer Bay Fiord were not observed. At the "ord mouth, evidence for the #ow of trunk ice into Eureka Sound is present on Hat Island, where granitic till overlies striated carbonate bedrock (Figs. 5 and 6). Stoss-side striae orientations record westward #ow towards Stor Island. South of Hat Island, ice- moulded and striated bedrock, overlain by granitic till and raised beaches documents southwestward #ow into Fig. 5. Examples of granite dispersal trains and striated bedrock, Eureka Sound (Fig. 6). southern Eureka Sound. (a) Granitic till, Bay Fiord. The till forms Lateral meltwater channels, graded to raised deltas a dispersal train extending westwards through Bay Fiord and north and beaches in central and outer Bay Fiord, record and south along Eureka Sound. (b) Striated bedrock overlain by gran- eastward receding trunk ice, and, along the south side of itic till, Hat Island (view taken looking eastwards up Bay Fiord towards the "ord, retreat towards the interior of Raanes Penin- the Prince of Wales Ice"eld). Striae and granites record westward ice-#ow from Bay Fiord into Eureka Sound. sula (Fig. 6). Farther east, in inner Bay and central Strathcona "ords, lateral channels graded to ice-contact deltas and gravel berms, both marking marine limit, document successive stages in the eastward retreat of ice through the outer "ord is recorded by till and "ord- trunk ice. parallel striae (Fig. 6). Lateral meltwater channels are On the south side of Bay Fiord (`xa in Fig. 6), a section also not widely developed along the cli!ed, west coast of (18 m high) of raised marine silt contains abundant Raanes Peninsula, particularly north and south of Trap- paired valves of Astarte borealis and occasional Mya per's Cove (Fig. 6). Where they do occur, however, they truncata. The silt has been over-folded from the SSE and demonstrate eastward ice retreat towards the interior of the top of the fold is truncated. Deformation decreases to Raanes Peninsula. Channels, arcuate moraines and asso- the north (coastwards) across section, but small-scale ciated glaciomarine deltas are well developed in Trap- deformation of individual silt and sand units in the form per's Cove, and similarly document the eastward retreat of chevron folding and normal faulting is common. This of ice from Eureka Sound (Fig. 6). deformation is interpreted as glaciotectonic because of C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1325

Fig. 6. Glacial geomorphology, southern Eureka Sound. Dashed line encompasses area after Hodgson (1979). the lateral decrease in deformation towards the coast, of poorly sorted, massive gravel, with occasional a-axis which is the opposite to that which might be expected transverse and b-axis imbricate fabric. This is succeeded from the onshore grounding of an iceberg. by 3 m of alternating beds of massive silty diamict and massive boulder gravel, which dip coastward and wedge 2.2.3. Stor Island out up-slope. These are capped by 3 m of horizontally The pattern of lateral and proglacial meltwater chan- bedded "ne sand with numerous paired valves of Mya nels (Fig. 6) demonstrates "nal ice retreat to the interior truncata, Astarte borealis, Hiatella arctica and occasional of the island. That this succeeded the break-up of ice in Clinocardium ciliatum. Two single valves of H. arctica Eureka Sound is suggested by prominent lateral channels dated 46.8 ka BP (AA-23585; Site 9, Fig. 9 and Table 1), graded to marine limit (145}151 m asl) along the north and 47.7 ka BP (AA-27489). coast of the island. The pro"le of these channels indicates The lowermost gravel was deposited from traction, that trunk ice from Bay Fiord wrapped around the north probably in a subaerial environment, given its imbrica- coast of the island. tion. Overlying wedges of diamict and boulder gravel are On the west coast of Stor Island, a 15 m-high section at inferred to be sediment gravity #ow deposits on account the mouth of a valley exposes a sequence of gravel, of their geometry and dips. This sequence could indicate diamict and sand. The lower 8 m consists predominantly an ice-proximal deglacial environment (L+nne, 1995), 1326 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Fig. 7. Examples of deglacial and raised marine landforms and sediments, southern Eureka Sound: (a) Ice-contact delta, Trold Fiord. (b) Raised beaches extending to 138 m asl, outer Baumann Fiord (view taken looking west to Bear Corner and Axel Heiberg Island). (c) Morainal bank composed of "ne-grained glaciomarine sediments, Trold Fiord. (d) Raised beaches, outer Bay Fiord. Beaches extend to marine limit at 116 m asl and trim ice-moulded bedrock. although till is absent. Uppermost "ne sands record more ted for #uctuations in atmospheric pressure and site distal sedimentation. Alternatively, the sediments could speci"c temperature. High tide level was used as the represent a non-glacial, transgressive sequence, with in- reference datum for sea level. itial subaerial #uvial sedimentation, followed by sedi- Throughout the study area, marine limit is marked by ment gravity #ows from the valley sides, subsequent one of the following: (1) the highest raised marine delta or submergence and faunal colonisation. beach which is connected upslope to subaerial outwash or meltwater channels (Fig. 7); (2) the lowest undisturbed 2.3. Marine limit till or felsenmeer (washing limit); or (3) the highest well- preserved marine shells. The latter criterion provides Marine limit is the maximum elevation attained by the only a minimum estimate on marine limit and was not sea along a glacioisostatically depressed coast. Its elev- used where shelly till outcrops. Washing limits are com- ation is a function of its distance from the former ice- monly marked by a notch cut in till with a well-sorted margin (hence ice thickness), the date of deglaciation, and sediment veneer or bedrock below, or by an abrupt subsequent eustatic sea level rise (Andrews, 1970). In textural transition between poorly sorted till/felsenmeer areas where former glaciers contacted the sea, marine and sorted sediment below. transgression behind the former ice limit occurs concur- Marine limit in the region ranges from 63 to 151 m asl rently with ice retreat, and hence accurate recognition (Fig. 8). The highest elevations occur on the north coast and dating of marine limit allows the timing of deglaci- of Stor Island at 145}151 m asl (Fig. 8). Marine limits ation to be established. The altitude of marine limit was '140 m asl also occur along the south coast of Raanes determined in the "eld using a Wallace and Tiernan Peninsula (Fig. 8). Fiords penetrating southern Raanes micro-altimeter (accuracy $2 m). Readings were correc- Peninsula and northwestern Svendsen Peninsula exhibit C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1327

Fig. 8. Marine limit elevations (m asl) marked by the uppermost delta, washing limit or raised beach, southern Eureka Sound. Italicised marine limit elevations are from Hodgson (1985).

a consistent drop in marine limit from mouth to head. ward to 65}87 m asl at the heads of Strathcona Fiord and For example, in Trold Fiord, marine limit falls from Irene Bay (Fig. 8). 143 m asl at the "ord mouth to 98 m asl at its head, whereas in Star"sh Bay it falls from 113 to 80 m asl. This 2.4. Chronology is the pattern expected with progressive up-"ord retreat by a trunk glacier (Andrews, 1970). 2.4.1. Pre-Holocene radiocarbon dates Along the east side of Eureka Sound, north of Hare Twelve radiocarbon dates were obtained by Acceler- Bay, marine limit is recorded by deltas at the mouths of ator Mass Spectrometry (AMS) on shell fragments from valleys. In inner Trapper's Cove, ice-contact deltas grade till and coarse outwash (Fig. 9 and Table 1). To avoid to relative sea levels at 118}120 m asl (Fig. 8). By con- a mixture of di!erent-aged shells, only single fragments trast, immediately south of the "ord mouth, along Eu- were dated. Finite dates on such shells provide a max- reka Sound, marine limit is marked by deltas at 99 m asl imum age on the glacier advance responsible for shell (Fig. 8). The north coast of Raanes Peninsula is charac- transport, and hence the youngest dates are the most terised by a variable marine limit ranging from 76 to instructive with respect to the timing of that event. Most 120 m asl, whereas marine limit falls progressively east- samples dated '30 ka BP, and are thus considered 1328 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Fig. 9. Pre-Holocene radiocarbon dates, southern Eureka Sound.

minimum ages, given the possibility of contamination by sinan marine deposits which survived glacial overriding younger carbon (Bradley, 1985). One sample dated during the LGM. (30 ka BP (27.3 ka BP; AA-23605; Site 3, Fig. 9 and Table 1), and there is no apparent reason why shells in 2.4.2. Holocene radiocarbon dates this age range should not be meaningful given the pre- Holocene radiocarbon dates obtained on marine shells cision of AMS analysis (Mangerud et al., 1981). and driftwood are listed in Table 2 and selected dates are Two samples collected from paired valves in growth shown in Fig. 10. The earliest deglacial date was obtained position provided pre-Holocene radiocarbon dates. Both on a sample of Portlandia arctica from the base of a mo- submitted samples utilised single valves of H. arctica rainal bank overlying till and striated bedrock in outer from 15 m asl at the top of the section on the west coast Trold Fiord. This dated 9.2 ka BP (TO-5604; Site 56, of Stor Island described above. The "rst sample dated Fig. 10 and Table 2). A second sample of P. arctica from 46.8 ka BP (AA-23585; Site 8, Fig. 9 and Table 1), where- the top of the bank dated 8.8 ka BP (TO-5592; Site 55). as the second dated 47.7 ka BP (AA-27489). There is no direct stratigraphic relationship between these samples (15 m asl) and local marine limit (124 m asl). These dates Calibrated ages are also presented for all radiocarbon dates in likely record a small pocket of older, pre-Late Wiscon- Table 2. C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1329

Table 1 Pre-Holocene radiocarbon dates, southern Eureka Sound

Site Location Laboratory Material " Age (yr BP) Enclosing Sample elev. Related RSL Comments dating no.! material (m asl) (m asl)

1 Bay Fiord AA-23607 Chlamys islandica 37,130$1000 Surface 128 na Glacially redeposited 78354N, 85300W fragment sample from granitic till veneer over ice moulded bedrock

2a Bay Fiord AA-23608 H.arctica fragment 33,030$610 Surface 137 na As above 78354N, 85300W

2b As above AA-23609 H. arctica fragment 38,490$1100 Surface 137 na As above

3a Bay Fiord AA-23605 M.truncata fragment 27,380$360 Surface 198 na Glacially redeposited 78351N, 84332W sample from sandy gravel outwash

3b Bay Fiord AA-23606 M.truncata fragment 37,910$960 Surface 198 na As above 78351N 84332W

4a Bay Fiord AA-23601 M.truncata fragment 34,830$850 Surface 176 na Glacially redeposited 7835050 N sample from sandy 8433300W gravel outwash

4b As above AA-23602 M.truncata fragment 30,930$420 Surface 176 na As above

4c As above AA-23603 H.arctica fragment 35,510$730 Surface 176 na As above

4d As above AA-23604 H.arctica fragment 41,690$1700 Surface 176 na As above

5 Bear Corner TO-5602 M.truncata fragment 35,310$400 Surface 191 na Glacially redeposited 78310N, 87324W sample from till veneer

6a Baumann Fiord TO-5600 M.truncata fragment 36,910$410 Surface 199 na Glacially redeposited 78309N, 87346W sample from till veneer

6b Baumann Fiord TO-5615 H.arctica ? fragment 36,160$430 Surface 190 na Glacially redeposited 78309N, 86346W sample from till veneer

7 Bjorne Peninsula GSC-2700 H.arctica 30,100$750# Surface 103-108 ? Whole valves and frag- 77329N, 85345W ments from gravelly silt on hilltop

8a Stor Island AA-27489 H.arctica" 47,790$3500 Sand 15 '15 Paired valves from 78358N, 86317W bedded sands

8b As above AA-23585 H.arctica" 46,850$2800 Sand 15 '15 As above

!Laboratory designations: GSC"Geological Survey of Canada; TO"IsoTrace Laboratory, University of Toronto; AA"University of Arizona. TO and AA samples were dated by accelerator mass spectrometry, and corrected for isotopic fractionation to a base of C"!25&. A reservoir correction of 410 yr was then applied, which is equivalent to corrections to a base of C"0&. GSC shell samples were dated conventionally and corrected for fractionation to a base of C"0&. "Denotes sample consisting of paired valves found within, or de#ating from, enclosing sediment. H. arctica"Hiatella arctica; M. truncata"Mya truncata. #GSC uncorrected date (Hodgson, 1985). This date has not been corrected for isotopic fractionation or a marine reservoir e!ect. 1330 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Fig. 10. Holocene radiocarbon dates, southern Eureka Sound.

Two fragments of H. arctica from beaches (133 and 131 m 24, respectively, Fig. 10 and Table 2). All these dates are asl) at the "ord mouth provided dates of 8.7 and 8.6 ka from deglacial sediments and thus provide closely limit- BP, respectively (AA-23583 and AA-23591; Sites 42 and ing minimum dates on ice-retreat. 43, Fig. 10 and Table 2). Further west, at Bear Corner, Along the north and west coasts of Raanes Peninsula whole valves and fragments of M. truncata and H. arctica deglacial dates as young as 8.2}7.1 ka BP were obtained from a raised beach (132 m asl) dated 8.7 ka BP (GSC- from ice-contact deposits (Fig. 10). The outer parts of 6028; Site 27). Localised patches of shelly, granitic till (see several "ords exhibit deglacial dates that are closely TO-5602; Site 5, Fig. 10 and Table 1) occur above marine similar to dates at the "ord heads, especially when the limit throughout this area. Paired valves of P. arctica standard errors are considered, and in some cases dates from ice-contact deltas in Trapper's Cove yielded ages of at the "ord mouths are actually younger than those at the 9.0 and 8.9 ka BP (AA-23587 and AA-23593; Sites 23 and heads. For example, Star"sh Bay has a date of 8.7 ka BP C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1331

Table 2 Holocene radiocarbon dates, southern Eureka Sound.!

Site Location Laboratory Material# Age(yrs BP) Enclosing Sample elev. Related RSL Calibrated age dating No." material (m asl) (m asl) (cal BP)%

1 Bay Fiord GSC-3823 M.truncata 6110$70$ Surface 92 *92})100 78343.5N, 82334W (outer fraction) #fragments

1a As GSC-3823 GSC-3823 M.truncata 7590$80$ Surface 92 *92})100 (inner fraction) #fragments

2 Bay Fiord GSC-5991 M.pseudoarenaria# 5300$70 Sand and silt 13 *13})96 5900}5580 7835330 N, 8232345 W

3a Irene Bay GSC-5897 Astarte borealis, 5200$70 Sand and silt 55 *55})78 5840}5460 7830318 N, H.arctica# 8132838 W

3b Irene Bay GSC-5966 Driftwood 6360$100 Sand 55 *55})78 7400}7010 7830318 N, Picea sp. 8132838 W

4 Irene Bay GSC-5955 Driftwood 1790$160 Sand 9 9 2050}1340 7830311 N, Picea sp. 8132852 W

5 Irene Bay GSC-1978 P. arctica# 8820$90 Silt 70}74 '80 9810}9340 79301N, 81331W

6 Irene Bay GSC-3397 H. arctica 7340$170 Surface 66}70 *75 8170}7500 79301N, 81328W

7 Augusta Bay GSC-118 H. arctica, 6370$100$ Silt 33 *37 78351N, 81348W M. truncata

8a Strathcona Fiord TO-5143 H. arctica 7740$70 Diamict 17 *17})74 8400}8060 7833620 N, fragment 8231810 W

8b Strathcona Fiord GSC-5937 M. truncata# 5220$80 Sand 13 *13})74 5870}5470 7833620 N, 8231830 W

9 Strathcona Fiord GSC-3765 M. truncata 6780$80 Surface 56 '56 7480}7170 78334N, 82319W

10 Strathcona Fiord TO-5144 P. arctica# 6700$100 Silt 54 '54})67 7450}7030 78332 30 N, 8231645 W

11 Strathcona Fiord GSC-175 Moss peat 7680$150$ Sand 395 na 78333N, 82320W

12 Strathcona Fiord TO-5145 Macoma calcarea# 5480$60 Silt 7 *7})82 6140}5760 78334 10 N, 8232000 W

13 Strathcona Fiord AA-23596 P. arctica# 7785$65 Silt 24 *24})84 8420}8110 78337N, 82342W

14a Strathcona Fiord AA-23589 P. arctica# 5945$70 Silty sand 16 *16})92 6620}6270 78340N, 82345W

14b Strathcona Fiord GSC-5960 Astarte borealis 6110$80 Surface 21 *21})92 6820}6410 7834030 N, 8234530 W 1332 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Table 2 (continued)

Site Location Laboratory Material# Age(yrs BP) Enclosing Sample elev. Related RSL Calibrated age dating No." material (m asl) (m asl) (cal BP)%

15 Strathcona Fiord GSC-170 M. truncata# 7750$160$ Silt 75 *77})98 78342N, 82351W 16a Strathcona Fiord AA-23599 M. truncata# 7480$65 Silty sand 53 *53})122 8130}7820 78345N, 83323W

16b Strathcona Fiord GSC-3728 Twig, Salix sp. 7280$90 Silty pebbly 70 *53})85 8250}7850 78345N, 83323W sand (122?) 17 Bay Fiord 78349N, AA-23586 P. arctica# 7140$55 Silty sand 48 *48})120 7780}7520 84303W 18 Bay Fiord AA-23598 H.arctica fragment 7180$65 Silt 6 '6})112 7840}7530 78351N, 84333W 19 Bay Fiord GSC-452 Whale skull 1380$130$ Surface 2.5 *2.5 78354N, 85310W 20 Bay Fiord AA-23588 H. arctica fragment 8190$60 Gravel 98 *98})120 8970}8540 78355N, 85313W 21 Eureka Sound AA-23590 H. arctica 7745$60 Surface 71 *71})76 8380}8090 78350N, 85344W 22 Stor Is. 78353N, AA-23584 M. truncata 7615$60 Gravel 49 '49})109 8280}7950 86314W fragment 23 Trapper's Cove AA-23587 P. arctica# 9030$70 Silt 72 '72})118 9930}9540 78339N, 86343W 24 Trapper's Cove AA-23593 P. arctica# 8925$70 Silt 97 '97})110 9850}9460 78332N, 86341W 25 Eureka Sound AA-23592 M. truncata 8245$90 Surface 99 '99 9140}8540 78331N, 87314W fragment

26 Bear Corner GSC-6067 M.truncata# 6040$60 Sand 23 '23 6710}6380 78308N, 87329W

27 Bear Corner GSC-6028 M.truncata, 8750$100 Surface 132 *132})142 9760}9200 78307N, 87327W H.arctica #fragments

28 Baumann Fiord TO-5610 Macoma calcarea 8160$70 Sand 65 '65 8960}8500 78308N, 86345W

29 Baumann Fiord TO-5862 M. truncata 8590$70 Gravel 128 *128})138 9460}9060 78308N, 86340W fragment

30 Blind Fiord GSC-5896 M. truncata, 8090$110 Surface 95 *95})120 8950}8370 78323N, 85340W H. arctica #fragments

31 Blind Fiord TO-5863 H. arctica fragment 7090$70 Surface 109 *109})125 7760}7440 78324N, 85348W

32 Blind Fiord GSC-6054 M. truncata, 8220$100 Surface 107 *107})124 9100}8470 78322N, 85348W H. arctica

33 Blind Fiord GSC-5924 M. truncata 7990$140 Surface 86 *86})124 8920}8190 7832045 N, #fragments 8535230 W

34 Blind Fiord TO-5608 H. arctica fragment 8310$80 Surface 123 *123})129 9210}8630 78322N, 85350W

35 Blind Fiord GSC-5990 M. truncata 8020$90 Surface 76 *76})129 8860}8310 7832010 N, H. arctica 8535058 W #fragments C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1333

Table 2 (continued)

Site Location Laboratory Material# Age(yrs BP) Enclosing Sample elev. Related RSL Calibrated age dating No." material (m asl) (m asl) (cal BP)%

36 Blind Fiord TO-5598 H. arctica 8670$190 Surface 122 *122})128 9850}8930 78319N, 85352W fragment 37 Blind Fiord TO-5862 H. arctica 5410$70 Surface 122 *122})133 6060}5670 78314N, 85356W fragment 38 Blind Fiord GSC-6047 H. arctica, 8550$90 Gravel 119 *119})133 9420}9000 78314N, 85357W M. truncata 39 Blind Fiord GSC-6102 H. arctica, 5640$110 Silt 31 *31})39 6360}5870 78313N, 86304W A. borealis# 40 Blind Fiord TO-5612 M. truncata 8510$80 Surface 127 *127})133 9400}8970 78311N, 86304W fragment 41 Baumann Fiord GSC-244 H. arctica, 8480$140$ Surface 116 *116 78306N, 85352W M. truncata #fragments 42 Baumann Fiord AA-23583 H. arctica 8725$65 Surface 133 *133})143 9600}9250 78306N, 85328W fragment 43 Trold Fiord AA-23591 H. arctica 8645$60 Gravel 131 *131})143 9500}9180 78306N, 85327W fragment 44 Trold Fiord BETA-91868 H. arctica, 6320$80 Sand 51 *51 7080}6650 78336N, 84330W A. borealis# 45a Trold Fiord TO-5594 P. arctica# 8250$450 Silts 83 '83})99 9830}7930 78337N, 84328W 45b Trold Fiord AA-23594 P. arctica# 8510$75 Silts 76 '76})99 9390}8980 78337N, 84328W 46 Trold Bay AA-23597 H. arctica# 8020$65 Silt 88 '88})92 8810}8350 78329N, 84334W 47 Star"sh Bay, GSC-6037 H. arctica, 7740$90 Silt 78 '78})101 8410}8000 78313N, 84334W M. truncata 48 Star"sh Bay GSC-2719 P. arctica# 8710$120 Silt 68 '72 9730}9060 78311N, 84308W 49 Star"sh Bay TO-5606 H. arctica# 5460$60 Silt 18 '18})86 6110}5740 78311N, 84301W 50 Star"sh Bay GSC-5907 M. truncata# 4800$120 Silt 6 '6})86 5460}4840 7831035 N, #fragments 8430200 W 51 Star"sh Bay TO-5596 H. arctica 7240$80 Silt 71 '71 7910}7570 78312N, 84300W 52a Star"sh Bay GSC-5936 M. truncata# 4630$80 Silt 3 '3})80 5240}4710 7831020 N, #fragments 8430315 W 52b Star"sh Bay GSC-5941 M. truncata 4950$70 Silt 9 '9})80 5570}5140 7831020 N, H. arctica# 8430315 W 53 Star"sh Bay, GSC-5967 M. truncata, 5130$90 Silt 7 '7})98 5760}5310 7831045 N, M. calcarea 8431820 W 54 Trold Fiord, TO-5590 H. arctica 7470$70 Surface 102 *102})110 8130}7800 7831045 N, fragment 8431820 W 55 Trold Fiord, TO-5592 P. arctica# 8840$80 Silt 71 '71 9810}9380 783080 N, 8435930 W 1334 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Table 2 (continued)

Site Location Laboratory Material# Age(yrs BP) Enclosing Sample elev. Related RSL Calibrated age dating No." material (m asl) (m asl) (cal BP)%

56 Trold Fiord, TO-5604 P. arctica# 9200$110 Silt 41 '41 10,150}9690 78308N, 84359W

57 Trold Fiord GSC-6034 M. truncata# 8210$90 Sand 75 *75})95 9050}8480 78307N, 85302W

58 Jaeger Bay AA-23600 H. arctica 7795$60 Surface 77 *77})106 8420}8120 78303N, 84347W fragment

59 Vendom Fiord GSC-1858 H. arctica# 7010$80 Sand 52 '54 7680}7370 78304N, 82306W

60 Vendom Fiord GSC-1957 H. arctica 6980$90 Surface 48}53 '53 ('63?) 7670}7320 78307N, 82308W

!Sources (including this paper): (Dyck and Fyles, 1964; Dyck et al., 1965; Lowdon and Blake, 1968, 1978; Hodgson, 1985) "Laboratory designations: GSC"Geological Survey of Canada; TO"IsoTrace Laboratory, University of Toronto; AA"University of Arizona. TO and AA samples were dated by accelerator mass spectrometry. These samples were corrected for isotopic fractionation to a base of C"!25&; a reservoir correction of 410 years was then applied, which is equivalent to correction to a base of C"0&; GSC shell samples were dated conventionally and corrected for fractionation to a base of C"0&. GSC terrestrial organic samples were dated conventionally and corrected for fractionation to a base of C"!25&. #Denotes sample consisting of paired valves found within, or de#ating from, enclosing sediment. H. arctica"Hiatella arctica; P. arctica"Portlandia arctica; M. truncata"Mya truncata; M. pseudoarenaria"Mya pseudoarenaria. $1960s GSC uncorrected dates (Hodgson, 1985). These dates have not been corrected for isotopic fractionation or a marine reservoir e!ect. Approximate corrections could be made for isotopic fractionation to a base of C"!25& by adding 400}410 yr to this uncorrected age (R. McNeely, unpublished communication to GSC clientele, 1991). A similar amount could then be subtracted to account for the marine reservoir e!ect. However, such a correction has not been applied as the result would be approximately the same as the uncorrected raw date reported here. GSC dates obtained during the course of this study (1990s) typically show di!erences between raw and corrected ages (to a base of C"0&) which are well within the reported standard errors of the individual dates. %Dates were calibrated using CALIB 3.0 (Stuiver and Reimer, 1993) and the calibrated date range reported here is that which yields 100% probability at 2p.

(GSC-2719; Site 48, Fig. 10 and Table 2) in the inner Sound, all radiocarbon dates, many from ice-contact "ord, and 8.8 ka BP (TO-5592, Site 55) at the "ord deltas, indicate that marine limit is early Holocene (Fig. mouth; Blind Fiord has a date of 8.3 ka BP (TO-5608; 10). This requires that the ice responsible for deposition Site 34) in the inner "ord, and 8.6 ka BP (TO-5862, Site of the granite erratics retreated during the early Holo- 29) at the "ord mouth; whereas Irene Bay is 8.8 ka BP cene, and therefore the dispersal trains are Late Wiscon- (GSC-1978; Site 5) at the "ord head and 8.2 ka BP sinan. Thus, in southern Eureka Sound at the LGM, ice (AA-23588; Site 20) at the mouth of Bay Fiord (same extended far beyond the `drift-belta. A possible max- "ord system). imum age for ice advance is provided by the youngest pre-Holocene radiocarbon date in the "eld area, from central Bay Fiord, which, if "nite, implies that trunk ice advanced westwards into Eureka Sound (27.3 ka BP 3. Interpretation (AA-23605; Site 3, Fig. 9 and Table 1). This is consistent with work from Nansen Sound which indicates that ice 3.1. Age of granite dispersal trains in southern Eureka advance to the LGM occurred there (28 ka BP (Be- Sound dnarski, 1998), and also with similar evidence from eastern Ellesmere Island where ice advanced to the LGM Granite dispersal trains and associated shelly till and (20 ka BP (Blake, 1992a; England, 1999). ice-moulded bedrock record regional westward ice-#ow from the Prince of Wales Ice"eld to Eureka Sound 3.2. Ice conxguration and dynamics during the LGM (Figs. 3 and 4). Marine limit is superimposed onto both dispersal trains and formed during retreat of the granite- In southern Eureka Sound, granite dispersal trains carrying ice (Figs. 7 and 8). Lateral meltwater channels record regional, westward #ow of warm-based ice during extend to marine limit (Fig. 6) and therefore must be the LGM, (1) along the axis of Bay Fiord, and (2) across coeval with that retreat. Throughout southern Eureka northern Svendsen and southern Raanes peninsulas, C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1335

of Trold Fiord, and are rare westwards to Eureka Sound (Fig. 3). This decrease may record the displacement of granite-bearing ice by ice emanating from Trold Fiord, which forced granite-carrying ice #owing across Bras- keruds Plain and NW Svendsen Peninsula o!-shore into Baumann Fiord. Baumann Fiord ice would probably have coalesced with southward-#owing trunk ice exiting Eureka Sound, and #owed SW towards Norwegian Bay, although the limit of this ice is presently unknown (Fig. 11). An extensive ice cover is therefore inferred for the LGM in southern and central Eureka Sound. This con- sisted of expanded ice caps from Ellesmere and Axel Heiberg islands which coalesced along the length of Eu- reka Sound, and extended continuously northwards through Nansen Sound, where Bednarski (1998) also reports occupation by trunk ice during the LGM. The westward ice-#ows emanated from an ice divide to the east of the study area, probably close to the contemporary divide of the Prince of Wales Ice"eld (Fig. 11). Reeh (1984) proposed a similar divide in his model of ice extent at the LGM in the Canadian Arctic islands, and England (1999) also proposed westward #ow of Ellesmere Island Fig. 11. Proposed palaeogeography of the LGM in southern Eureka ice towards Eureka Sound during the LGM. Flow to- Sound, showing location of ice-divides (`Da), extent of local and re- wards Eureka Sound probably would have been mir- gional (granite-carrying) ice and associated principal ice-#ow direc- rored by eastward #ow from a divide over central Axel tions. Heiberg Island. No evidence was found for ice-#ow into southern Eureka Sound from Norwegian Bay. The re- both emanating from a divide in the vicinity of the construction of LGM ice advocated here for southern present Prince of Wales Ice"eld (Fig. 11). This #ow over- Eureka Sound most closely approximates the Innuitian topped summits at 1036 m asl between Star"sh and Ice Sheet model of Blake (1970). Jaeger bays, and at 764 m asl between Bay and Vesle Minimum estimates on ice thickness in Eureka Sound "ords. The intervening, predominantly granite-free area during the LGM are provided by the elevations of gran- of Raanes Peninsula was also ice-covered during the ite erratics: 722 m asl at the mouth of Bay Fiord, and LGM as recorded by the pattern of deglacial landforms 764 m asl between Bay and Vesle "ords. Water depths in (Fig. 11). outer Bay Fiord reach 422 m (Department of Fisheries The Bay Fiord dispersal train exhibits a convergent and Oceans, 1979), indicating a former ice thickness of at #ow pattern (Fig. 4). This implies that Bay Fiord was least 1200 m where Bay Fiord joins Eureka Sound. a major conduit for ice draining an expanded Prince of North of Stor Island, Bell (1992) reports granites on Wales Ice"eld during the LGM, and was bordered to summits at 725 m asl adjacent to Eureka Sound where the south by locally nourished ice on Raanes Peninsula water depth is up to 353 m, indicating a minimum ice (Fig. 11). The distribution of erratics and striae indicates thickness of &1080 m. By contrast, ice at the Wiscon- that granite-carrying ice exited Bay Fiord and #owed sinan divide over the Agassiz Ice Cap is inferred to have north and south along Eureka Sound (Fig. 11). Bell been only 200 m thicker than today (Koerner et al., 1987) (1992) mapped the northward extension of this #ow (currently 500}800 m thick, Koerner, 1989). This suggests along eastern Axel Heiberg Island and NW Fosheim that although the thickest ice may have been located in Peninsula (Fig. 4), and traced it to Nansen Sound. Else- Eureka Sound, an interpretation supported by the pat- where, granites are absent from summits along the tern of postglacial emergence (England and OD Cofaigh, Raanes Peninsula coast of Eureka Sound as far south as 1998; OD Cofaigh, 1999b, the ice divides which fed this Hare Bay. This suggests that granite-carrying ice was were situated on the alpine highlands to the east and west kept o!shore to the west of these summits by Raanes (Fig. 11). Peninsula ice (Fig. 11). Rare granites south of Hare Bay Such an interpretation is also supported by glaciologi- may record the onshore #ow of Eureka Sound ice, or cal modelling of a Late Wisconsinan ice-cover in the alternatively, they were deposited by Baumann Fiord ice. Canadian High Arctic (Reeh, 1984). Model results show Granite erratics are ubiquitous in Star"sh and Jaeger a main ice-divide, with ice thicknesses of &700}1000 m, bays, but they exhibit a dramatic decrease at the mouth located along the highlands of eastern Ellesmere Island, 1336 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 but with thicker ice (1500}2000 m) located west of the main divide in Eureka Sound and Norwegian Bay (ice- margins placed along the 200 and 600 m sea-depth con- tours, respectively) (Reeh, 1984, Fig. 4). The dispersal trains documented above demonstrate convergent ice-#ow into the major "ords of the study area. This #ow pattern suggests that these "ords existed prior to the Late Wisconsinan, and facilitated westward drainage of ice from the central Ellesmere Island ice divide during the LGM. Such an interpretation is also supported by the presence of Late Wisconsinan shelly till along these "ords. Therefore, physiography appears to have imposed a strong control on the pattern of ice-#ow during the LGM in southern Eureka Sound. Existence of these "ords prior to the LGM also provides indirect evidence for pre-Late Wisconsinan glaciation in the region, with evolution of the "ords during successive epi- sodes of glacial erosion. Di!erences in erratic weathering may also re#ect di!erences in the duration of weathering associated with age, implying some pre-Late Wiscon- sinan glacial transport of erratics.

3.3. Deglaciation

Deglaciation of southern Eureka Sound commenced prior to 9.2 ka BP (Fig. 10 and Table 2), and trunk ice had vacated the sound by 9.0 ka BP. Furthermore, sev- Fig. 12. Early Holocene deglacial dates from Eureka Sound and Nan- ` a" eral "ords exhibit early Holocene radiocarbon dates at sen Sound, Ellesmere and Axel Heiberg islands. ES Eureka Sound, `NSa"Nansen Sound. Sources (including this paper): Hodgson et al. their heads (Fig. 10 and Table 2), indicating that they (1991); Hein and Mudie (1991); Bell (1996); Bednarski (1995, 1998). experienced rapid retreat of trunk glaciers as early as 8.7}8.8 ka BP. Along the north and west coasts of Raanes Peninsula, "ord mouth to head, a pattern more indicative of pro- geomorphic evidence demonstrates concentric ice-retreat gressive, slow retreat (Andrews, 1970). However, these into peninsula interiors following break-up in adjacent observations (fast retreat and up-"ord decline in marine "ords and Eureka Sound. Radiocarbon dates from ice- limit) are compatible because regional deglaciation of contact sediments indicate a period of ice-marginal southern Eureka Sound took place down-isobase, with stabilisation in the vicinity of the present coastline of eastward retreat from the sound (where ice was thickest, Raanes Peninsula until at least 7.1 ka BP, prior to retreat see above) to the `drift belta of the inner "ords. Thus into the interior. This is consistent with low marine limits restrained rebound was minimised at the former loading at the mouths of some valleys (e.g., 76 m asl, Fig. 8) which centre in Eureka Sound but not at the "ord heads, record the presence of locally persisting glaciers that resulting in maximum ampli"cation of di!erential re- delayed re-entry of the sea. bound as recorded by marine limit (England and The style of deglaciation reconstructed from available OD Cofaigh, 1998). radiocarbon dates and geomorphic evidence indicates In contrast to southern Eureka Sound, earlier de- a two-step retreat pattern. Initial concentric retreat by glacial dates are reported from the northern part of the calving of trunk glaciers was rapid in many "ords (cf. channel in Otto Fiord and outer Nansen Sound (11.6 and OD Cofaigh, 1998). This was followed by terrestrial 10.3 ka BP, respectively; Bednarski, 1995, 1998; Fig. 12 stabilisation in the narrower and shallower inner "ords, and Table 3). Hodgson et al., (1991) also report a date of and along coastlines adjacent to the larger channels, 10.6 ka BP from the interior of Fosheim Peninsula prior to "nal retreat. Regionally, evidence for a two-step (Fig. 12 and Table 3), although this is regionally anomal- pattern to early Holocene deglaciation has been present- ous. This chronological di!erence between southern and ed for western Axel Heiberg and eastern Ellesmere is- northern Eureka Sound implies either (1) an asymmetry lands (Lemmen et al., 1994; England, 1999). in the timing of initial deglaciation at the two ends of In several "ords where radiocarbon dates show rapid Eureka Sound, or (2) that marine shells as old as those to early Holocene deglaciation (e.g., Star"sh Bay, Bay the north are present but have not been collected yet in Fiord), marine limit exhibits a decline in elevation from the southern part of the sound. Because outer Nansen C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1337

Table 3 Early Holocene radiocarbon dates, Eureka Sound and Nansen Sound!

Site Location Laboratory Material# Age (yr BP) Enclosing Sample elev. Related RSL Calibrated age dating no." material (m asl) (m asl) (cal BP) $

1 Axel Heiberg Shelf TO-1148 Foraminifera 9955$80 Mud Marine core 11,220}10,550 &80358N, 97325W 2 Nansen Sound GSC-4740 Yoldia sp. 10,300$100 Silt 45 *45 11,950}11,010 81301N,91340W 3 Otto Fiord TO-1110 Portlandia sp. 11,660$80 Silt 72 '104 13,510}13,040 81303N,88341W 4 Otto Fiord S-2646 M.truncata 9115$130 Surface 77 '77}115 10,100}9490 81307N,87304W 5 Fosheim TO-2240 P. arctica 9560$90 Mud 90 '88 )141 10,780}10,030 Peninsula 79356N,84317W 6 Fosheim GSC-4784 H. arctica? 10,600$90 Surface 122}123 *123 12,420}11,610 Peninsula 79358.2N, 84329W 7 Trapper's Cove AA-23587 P. arctica 9030$70 Silt 72 '72)118 9930}9540 78339N, 86343W 8 Bear Corner GSC-6028 M. truncata 8750$100 Surface 132 *132! 9760}9200 78307N, 87327W H. arctica )142 #fragments 9 Trold Fiord TO-5604 P. arctica 9200$110 Silt 41 '41)143 10,220}9630 78308N,84359W

!Sources (including this paper): (Hodgson et al., 1991; Hein and Mudie, 1991; Bell, 1996; Bednarski, 1995, 1998). "Laboratory designations: GSC"Geological Survey of Canada; TO"Iso Trace Laboratory, University of Toronto; S"Saskatchewan Research Council; AA"University of Arizona. TO and AA samples were dated by accelerator mass spectrometry. These samples were corrected for isotopic fractionation to a base of C"!25&; a reservoir correction of 410 yr was then applied, which is equivalent to a correction to a base of C"0&; GSC and S shell samples were dated conventionally and corrected for fractionation to a base of C"0&. #H. arctica"Hiatella arctica; P. arctica"Portlandia arctica; M. truncata"Mya truncata. $Dates were calibrated using CALIB 3.0 (Stuiver and Reimer, 1993) and the calibrated date range reported here is that which yields 100% probability at 2p.

Sound is located in a more distal part of the former landform/sediment assemblages found in the inner parts ice-cover over the Queen Elizabeth Islands during the of "ords, and a sparsity or absence of such features in the LGM, earlier deglaciation of Nansen Sound is expected outer "ords (Hodgson, 1985; England, 1987, 1990; Lem- (cf. Blake, 1970; Bednarski, 1998). Deglaciation of Eureka men et al., 1994; Bell, 1996; OD Cofaigh, 1998, 1999a, Sound was delayed until )9.5 ka BP and the channel OD Cofaigh et al., 1999). The signi"cance of this landscape was ice-free by 9.0 ka BP. It is possible that the driving zonation has been the subject of much debate. Most mechanism for ice retreat through the sound was eustatic discussion has centred around whether the sparse sea level rise between &9.5 and 10.2 ka BP and 11- geomorphic record of the outer "ords implies a restricted 11.5 ka cal BP (Fairbanks, 1989; Bard et al., 1990; Blanc- Late Wisconsinan ice cover (e.g., England, 1987, Bell, hon and Shaw, 1995). Furthermore eustatic sea level rise 1996), or alternatively re#ects a more extensive cold- occurred in concert with an abrupt regional increase in based ice-cover, which inhibited the formation of glacial temperature commencing &10-10.2 ka BP (Koerner landforms, preserving pre-existing weathered terrain (e.g., and Fisher, 1990; Alley et al., 1993; Kapsner et al., 1995). Sugden and Watts, 1977; Hughes, 1987). A third alternative to explain this landscape zonation 3.4. Wider implications is that rapid deglaciation, as indicated by glacial geomorphology and radiocarbon dating in several "ords in 3.4.1. Signixcance of landscape zonation in inner vs. outer southern Eureka Sound, may have occurred regionally xords across western Ellesmere and Axel Heiberg islands. On western Ellesmere and Axel Heiberg islands many Rapid deglaciation by calving in the larger "ords and authors have noted a contrast between prominent glacial inter-island channels (Lemmen et al., 1994; Funder and 1338 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Hansen, 1996; OD Cofaigh, 1998; Bednarski, 1998; Eng- However, ratios were not obtained on shell fragments land, 1999) would have promoted extensional #ow and from tills in stratigraphic section, and thus the chrono- thinned the ice pro"le, facilitating extensive crevassing logic subdivision of tills into discrete glaciations lacks and meltwater drainage. This could have inhibited a stratigraphic basis. Furthermore, many ratios were ob- formation of deglacial landforms and sediments until tained from surface collections. Surface shells can yield ice-margins stabilised on-land. Extensive deglacial higher D/L ratios than samples of the same age which landform/sediment assemblages at many "ord heads in have been buried '1 m for most of their history due to the region are inferred to record this stillstand (cf. their di!erent thermal histories since deposition (Miller Hodgson, 1985; OD Cofaigh, 1998; OD Cofaigh et al., and Brigham-Grette, 1989). Thus the ratios of some of 1999). these surface samples may be maximum estimates. The Several workers have proposed that this regional, "ord thermal history of all samples is further complicated by head belt of glacial landforms and sediments marks a cli- the fact they were transported and covered for an un- matically-driven change in the style of early Holocene known length of time by warm-based ice which was at deglacial sedimentation, associated with a transient the pressure melting point and thus much warmer than switch in the basal thermal regime of trunk glaciers from the current mean annual air temperature in the area cold-based to warm (e.g., Lemmen, 1990; Stewart, 1991). (!203C). As a result, samples may have been subjected This was largely based on the interpretation of a re- to signi"cant oscillations in temperature throughout stricted Late Wisconsinan ice cover, characterised by their history that could have a!ected their epimerisation limited expansion of cold-based glaciers. However, Late rate and resulting ratios (cf. Dyke, 1984). Wisconsinan dispersal trains and associated ice-moulded Finally, deposits interpreted as raised marine shore- bedrock in southern Eureka Sound demonstrate that lines (160}170 m asl) in northern Eureka Sound lack some trunk glaciers were warm-based throughout much associated "ne-grained sediments with in-situ marine of their length, and associated early Holocene ground- macrofauna, and thus their age and genesis warrant ing-line fans and morainal banks indicate that they re- reconsideration. As on Raanes Peninsula, the highest mained so during deglaciation. Together with glacial raised marine shorelines recognised on Fosheim Penin- geomorphic evidence for spatial variation in basal ther- sula (as deduced from their association with included in mal regime between retreating trunk glaciers (OD Cofaigh, situ raised marine fauna) all date early Holocene, imply- 1998), this argues against any climatically-driven re- ing that the area was inundated by Late Wisconsinan ice. gional switch to warm-based thermal conditions during Thus, given the direct glacial geological evidence for the early Holocene (cf. OD Cofaigh et al., 1999). extensive Late Wisconsinan ice to the north and south of Fosheim Peninsula, that reached at least 1200 m thick- 3.4.2. Late Wisconsinan glaciation of northern Eureka ness in much of Eureka Sound, we reject earlier inter- Sound pretations favouring only pre-Late Wisconsinan regional Late Wisconsinan granite-carrying ice exiting Bay ice along northern Eureka Sound. Fiord, advanced northwards along Eureka Sound to Nansen Sound (Fig. 4). Recent work from Nansen Sound 3.4.3. Late Wisconsinan glaciation of the western Arctic (Bednarski, 1998) reports stratigraphic and chronologic Archipelago evidence of only one glacial cycle there, which is assigned During the LGM, regional granite-carrying ice #owing to the Late Wisconsinan/early Holocene. During that across northern Svendsen and southern Raanes penin- interval, Nansen Sound was "lled by northward-#owing sulas, coalesced with southward-#owing trunk ice exiting ice emanating from Eureka Sound which is tributary to Eureka Sound (Fig. 11). Such coalescent ice would have it. Hence extensive ice in northern Eureka Sound during #owed southwestwards towards Norwegian Bay. Several the last glaciation is implicit (Bednarski, 1998), and this is workers (e.g., St.-Onge, 1965; Balkwill et al., 1974) have consistent with the proposed Late Wisconsinan diver- documented granite erratics on Amund Ringnes and gence of ice (both north and south) in central Eureka Ellef Ringnes islands in the western Arctic. These erratics Sound (this study). have been ascribed to either an early Quaternary north- On Fosheim Peninsula, northern Eureka Sound, Bell ward advance of Laurentide ice from the mainland (1992) concluded that two regional glaciations occurred, (Hodgson, 1989), or to an ice advance from the eastern or one dating from the late Tertiary/early Quaternary, the southeastern parts of the Arctic Archipelago, including other from the early-mid Quaternary. This interpretation Ellesmere Island (Craig and Fyles, 1960; St.-Onge, 1965). was predominantly based on amino acid ratios (al- If the granite erratics on the Ringnes Islands were depos- loisoleucine to isoleucine) on shell fragments collected ited by Laurentide ice advancing northwards across the from surface tills, and on the interpretation that high archipelago, this ice should also have crossed Bathurst elevation (160}170 m asl) outwash deposits were pre- Island, as well as Grinnell Peninsula on northwestern Holocene raised marine shorelines (Holocene marine Devon Island. With the exception of one shield erratic limit is )150 m asl; Bell, 1996). close to marine limit on Ile Vanier (o! western Bathurst C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1339

Island), recent "eldwork in these locations does not re- Mayewski, P.A., Zielinski, G.A., 1993. Abrupt increase in Greenland port shield erratics above marine limit (Bednarski, 1996; snow accumulation at the end of the Younger Dryas event. Nature Dyke, 1999). 362, 527}529. Andrews, J.T., 1970. A geomorphological study of postglacial uplift This raises the possibility that during the Late Wiscon- with particular reference to Arctic Canada. Institute of British sinan, granite-carrying ice from SW Ellesmere Island Geographers, London, England, Special Publication No. 2, 156pp. advanced into Norwegian Bay and extended westwards Balkwill, H.R., Roy, K.J., Hopkins, W.S., Sliter, W.V., 1974. Glacial across the Ringnes Islands. Dyke (1999) has documented features and pingos, Amund Ringnes Island, Arctic Archipelago. ice-#ow directional indicators of Late Wisconsinan age Canadian Journal of Earth Sciences 11, 1319}1325. Bard, E., Hamelin, B., Fairbanks, R.G., 1990. U-Th ages obtained by on northern Grinnell Peninsula that demonstrate north- mass spectrometry in corals from Barbados: sea level during the ward #ow. If Ellesmere Island ice had extended into past 130,000 years. Nature 346, 456}458. Norwegian Bay, it would likely have coalesced with this Bednarski, J., 1986. Late Quaternary glacial and sea-level events, Clem- northerly #owing Devon Island ice and been de#ected ents Markham Inlet, northern Ellesmere Island, Arctic Canada. northwestwards across the Ringnes islands. North- Canadian Journal of Earth Sciences 23, 1343}1355. Bednarski, J., 1995. Glacial advances and stratigraphy in Otto Fiord westerly ice-#ow directional indicators along the east and adjacent areas, Ellesmere Island, Northwest Territories. Cana- coast of Amund Ringnes Island (Balkwill et al., 1974) dian Journal of Earth Sciences 32, 52}64. may re#ect #ow of this ice through Massey Sound (Fig. Bednarski, J., 1996. Sur"cial geology and sea level history of Bathurst 1). The validity of this hypothesis awaits more detailed Island, Northwest Territories. Current Research Part B, Geological "eldwork in Norwegian Bay and on the Ringnes Islands. Survey of Canada, Paper 96-B, pp. 61}66. Bednarski, J., 1998. Quaternary history of Axel Heiberg Island bordering Nansen Sound, Northwest Territories, emphasising the Last Glacial Maximum. Canadian Journal of Earth Sciences 35, Acknowledgements 520}533. Bell, T., 1992. Glacial and sea level history of western Fosheim Penin- This paper is part of the senior author's Ph.D. thesis, sula, Ellesmere Island, Arctic Canada. Unpublished Ph.D. thesis, University of Alberta, Edmonton, 172pp. written at the University of Alberta. The research was Bell, T., 1996. Late Quaternary glacial and sea level history of Fosheim supported by the Natural Sciences and Engineering Re- Peninsula, Ellesmere Island, Canadian High Arctic. Canadian Jour- search Council of Canada, Grant A6680 to J. E., the nal of Earth Sciences 33, 1075}1086. Canadian Circumpolar Institute (BAR Grant) and the Blake, W., Jr., 1970. Studies of glacial history in arctic Canada. I. Quaternary Research Association (Young Research Pumice, radiocarbon dates, and di!erential postglacial uplift in the eastern Queen Elizabeth Islands. Canadian Journal of Earth Workers Award) to C. OD C., and the National Science Sciences 7, 634}664. Foundation, Grant OPP-9530857 to M.Z.. Logistical Blake, W., Jr., 1972. Climatic implications of radiocarbon-dated drift- support was provided by the Polar Continental Shelf wood in the Queen Elizabeth Islands, Arctic Canada. In: Vasari, Y., Project, Natural Resources Canada. Radiocarbon dates HyvaK rinen, H., Hicks, S. (Eds.), Climatic Changes in Arctic Areas were determined by the Geological Survey of Canada during the Past Ten-Thousand Years. Acta Universitatis Ouluensis, Series A. Scienti"ae Rerum Naturalium, No. 3, Geologica No. 1, pp. (Ottawa), IsoTrace Laboratory (University of Toronto), 77}104. University of Arizona (Tucson), and BETA Analytic (Mi- Blake, W., Jr., 1978. Aspects of glacial history, southwestern Ellesmere ami). We particularly thank Dr. Roger McNeely, GSC Island, District of Franklin. Geological Survey of Canada Paper Radiocarbon Dating Laboratory, for assistance and dis- 78-1A, pp. 175}182. cussion concerning radiocarbon dates. B. TenbruK ggen, C. Blake W., Jr., 1992a. Shell-bearing till along Smith Sound, Ellesmere Island - Greenland: age and signi"cance. Sverigies Geologiska Horvath and S. Gordon provided dedicated "eld assist- UndersoK kning 81, 51}58. ance. Assistance by A. Podor during initial "eldwork is Blake W., Jr., 1992b. Holocene emergence at Cape Herschel, east also gratefully acknowledged. Discussion and/or reviews central Ellesmere Island, Arctic Canada: implications for ice sheet of earlier drafts of the manuscript by M.J. Sharp (Univer- con"guration. Canadian Journal of Earth Sciences 29, sity of Alberta), D.S. Lemmen and J. Bednarski (Geologi- 1958}1980. Blake, W., Jr., 1993. Holocene emergence along the Ellesmere Island cal Survey of Canada, Calgary), M. Allard (UniversiteH coasts of northernmost Ba$n Bay. Norsk Geologisk Tidsskrift 73, Laval), N.W. Rutter and D. Vitt (University of Alberta), 147}160. and D.A. Hodgson (Geological Survey of Canada, Ot- Blake, W., Jr., Boucherle, M.M., Fredskild, B., Janssens, J.A., Smol, J.P., tawa) are also gratefully acknowledged. Jonathan Tooby, 1992. The geomorphological setting, glacial history and Holocene ` +a Photographic Services, School of Geographical Sciences, development of Kap Ingle"eld S North-West Greenland. Med- delelser om Gr+nland, Geoscience, 27, 41pp. University of Bristol, assisted with preparation of the Blake, W., Jr., Jackson, H.R., Currie, C.G., 1996. Sea#oor evidence for photographs. Formal reviews by A.S. Dyke and W. glaciation, northernmost Ba$n Bay. Bulletin of the Geological Blake, Jr. improved the paper. Society of Denmark 43, 157}168. Blanchon, P., Shaw, J., 1995. Reef-drowning events during the last deglaciation: evidence for catastrophic sea-level rise and ice-sheet collapse. Geology 23, 4}8. References Boesch, H., 1963. Notes on the geomorphological history. Preliminary Alley, R.B., Meese, D.A., Shuman, C.A., Gow, A.J., Taylor, K.C., Report 1961}1962. McGill University, MontreH al, Axel Heiberg Is- Grootes, P.M., White, J.W.C., Ram, M., Waddington, E.D., land Research Reports, pp. 163}168. 1340 C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341

Bradley, R.S., 1985. Quaternary Paleoclimatology: Methods Fyles, J.G., 1989. High terrace sediments, probably of Neogene age, of Paleoclimatic Reconstruction. Chapman & Hall, London, west-central Ellesmere Island, Northwest Territories. Current Re- 472pp. search, Part D, Geological Survey of Canada, Paper 89-1D, pp. Craig, B.G., Fyles, J.G., 1960. Pleistocene Geology of Arctic Canada. 101}104. Geological Survey of Canada, Paper 60}10, 21pp. Fyles, J.G., 1990. Beaufort Formation (Late Tertiary) as seen from Department of Fisheries and Oceans, 1979. Eureka Sound and south- Prince Patrick Island, Arctic Canada. Arctic 43, 393}403. ern approaches including Baumann Fiord. Canadian Hydrographic Hattersley-Smith, G., 1969. Glacial features of Tanquary Fiord and Service, Chart 7940, scale 1: 3,00 000. adjoining areas of northern Ellesmere Island, NWT. Journal of Dyck, W., Fyles, J.G., 1964. Geological Survey of Canada radiocarbon Glaciology 8, 23}50. dates III. Radiocarbon 6, 167}181. HaK ttestrand, C., Stroeven, A.P., 1996. Field evidence for wet-based Dyck, W., Fyles, J.G., Blake, W., Jr., 1965. Geological Survey of Canada ice-sheet erosion from the south-central Queen Elizabeth Islands, radiocarbon dates IV. Radiocarbon 7, 24}46. Northwest Territories, Canada. Arctic and Alpine Research 28, Dyke, A.S., 1984. Multiple deglaciations of the Hudson Bay Lowlands, 466}474. Canada, since deposition of the Missinaibi (last interglacial?) Hein, F.J., Mudie, P.J., 1991. Glacial-marine sedimentation, Canadian Formation: Discussion. Quaternary Research 22, 247}252. polar margin, north of Axel Heiberg Island. GeH ographie Physique et Dyke, A.S., 1998. Holocene delevelling of Devon Island, Arctic Canada: Quaternaire 45, 213}227. implications for ice sheet geometry and crustal response. Canadian Hodgson, D.A., 1973. Landscape and late-glacial history, head of Journal of Earth Science 35, 885}904. Vendom Fiord, Ellesmere Island. Report of Activities, Part B, Dyke, A.S., 1999. The Last Glacial Maximum and deglaciation of Geological Survey of Canada Paper 73-1B, pp. 129}136. Devon Island, Arctic Canada: support for an Innuitian Ice Sheet. Hodgson, D.A., 1979. Sur"cial materials, south-central Ellesmere Is- Quaternary Science Reviews 18, 393}420. land, NWT. Geological Survey of Canada Open File 635. England, J., 1976. Late Quaternary glaciation of the eastern Queen Hodgson, D.A., 1982. Sur"cial materials and geomorphological pro- Elizabeth Islands, Northwest Territories, Canada: alternative mod- cesses, western Sverdrup and adjacent islands, District of Franklin. els. Quaternary Research 6, 185}202. Geological Survey of Canada, Paper 81-9, 37pp. England, J., 1983. Isostatic adjustments in a full glacial sea. Canadian Hodgson, D.A., 1985. The last glaciation of west-central Ellesmere Journal of Earth Sciences 20, 895}917. Island, Arctic Canada. Canadian Journal of Earth Sciences 22, England, J., 1987. Glaciation and the evolution of the Canadian high 347}368. arctic landscape. Geology 15, 419}424. Hodgson, D.A., 1989. Quaternary geology of the Queen Elizabeth England, J., 1990. The late Quaternary history of Greely Fiord and its Islands. In: Fulton, R.J. (Ed.), Quaternary Geology of Canada and tributaries, west-central Ellesmere Island. Canadian Journal of Greenland, Geological Survey of Canada, Geology of Canada, no. Earth Sciences 27, 255}270. 1, pp. 441}477. England, J., 1992. Postglacial emergence in the Canadian High Arctic: Hodgson, D.A., 1990. Were erratics moved by glaciers or icebergs to integrating glacioisostasy, eustasy and late deglaciation. Canadian Prince Patrick Island, western Arctic Archipelago, Northwest Terri- Journal of Earth Sciences 29, 984}999. tories? Current Research, Part D, Geological Survey of Canada, England, J., 1996. Glacier dynamics and paleoclimatic change during Paper 90-1D, pp. 67}70. the last glaciation of eastern Ellesmere Island, Canada. Canadian Hodgson, D.A., St-Onge, D.A., Edlund, S.A., 1991. Sur"cial materials of Journal of Earth Sciences 33, 779}799. Hot Weather Creek basin, Ellesmere Island, Northwest Territories. England, J., 1997. Unusual rates and patterns of Holocene emergence, In Current Research, Part E, Geological Survey of Canada Paper Ellesmere Island, Arctic Canada. Journal of the Geological Society 91-1E, pp. 157}163. London 154, 781}792. Hughes, T., 1987. Ice dynamics and deglaciation models when ice sheets England, J., 1998. Support for the Innuitian Ice Sheet in the Canadian collapsed. In: Ruddiman, W.F., Wright Jr., H.E. (Eds.), North Amer- High Arctic during the Last Glacial Maximum. Journal of Quater- ica and adjacent oceans during the last deglaciation. Geological nary Science 13, 275}280. Society of America, The Geology of North America, v.K-3, Boulder, England, J., 1999. Coalescent Greenland and Innuitian ice during the Colorado, pp. 183}220. Last Glacial Maximum: revising the Quaternary of the Canadian Kapsner, W.R., Alley, R.B., Shuman, C.A., Anandakrishnan, S., High Arctic. Quaternary Science Reviews 18, 421}456. Grootes, P.M., 1995. Dominant in#uence of atmospheric circulation England, J., Bradley, R.S., 1978. Past glacial activity in the Canadian on snow accumulation in Greenland over the past 18,000 years. High Arctic. Science 200, 265}270. Nature 373, 52}54. England, J., OD Cofaigh, C., 1998. Deglacial sea level along Eureka Koerner, R.M., 1989. Queen Elizabeth Islands glaciers. In: Fulton, R.J. Sound: the e!ects of ice retreat from a central basin to alpine (Ed.), Quaternary Geology of Canada and Greenland, Vol. 1, Geo- margins. Geological Association of Canada and Mineralogical As- logical Survey of Canada, Geology of Canada, pp. 464}473. sociation of Canada Annual Meeting, abstract volume, Quebec Koerner, R.M., Fisher, D.A., 1990. A record of Holocene summer City, A-52. climate from a Canadian high-Arctic ice core. Nature 343, 630}631. Evans, D.J.A., 1990. The last glaciation and relative sea history of Koerner, R.M., Fisher, D.A., Patterson, W.S.B., 1987. Wisconsinan and northwest Ellesmere Island. Canadian High Arctic. Journal of pre-Wisconsinan ice thicknesses on Ellesmere Island, Canada: in- Quaternary Science 5, 67}82. ferences from ice cores. Canadian Journal of Earth Sciences 24, Fairbanks, R.G., 1989. A 17,000 year glacio-eustatic sea level record: 296}301. in#uence of glacial melting rates on the Younger Dryas event and Lemmen, D.S., 1989. The last glaciation of Marvin Peninsula, northern deep ocean circulation. Nature 342, 637}642. Ellesmere Island, High Arctic, Canada. Canadian Journal of Earth Farrand, W.F., Gadja, R.T., 1962. Isobases on the Wisconsin marine Sciences 26, 2578}2590. limit in Canada. Geographical Bulletin 17, 5}22. Lemmen, D.S., 1990. Glaciomarine sedimentation in Disraeli Fiord, Funder, S., Hansen, L., 1996. The Greenland ice sheet * a model for its High Arctic Canada. Marine Geology 94, 9}22. culmination and decay during and after the Last Glacial Max- Lemmen, D.S., Aitken, A.E., Gilbert, R., 1994. Early Holocene deglaci- imum. Bulletin of the Geological Society of Denmark 42, ation of Expedition and Strand "ords, Canadian High Arctic. Cana- 137}152. dian Journal of Earth Sciences 31, 943}958. Fyles, J.G., 1962. In: Jenness, S.E. (Ed.), Field Work, 1961; Geological L+nne, I., 1995. Sedimentary facies and depositional architecture of Survey of Canada Information Circular No. 5, pp. 4}6. ice-contact glaciomarine systems. Sedimentary Geology 98, 13}43. C. O! Cofaigh et al. / Quaternary Science Reviews 19 (2000) 1319}1341 1341

Lowdon, J.A., Blake, W., Jr., 1968. Geological Survey of Canada St. Onge, D., 1965. La geomorphologie de l'Ile Ellef Ringnes Territoires radiocarbon dates VII. Radiocarbon 10, 207}245. du Nord-Ouest, Canada. Geographical Branch, Paper 38, Lowdon, J.A., Blake, W. Jr., 1978. Geological Survey of Canada 58pp. radiocarbon dates XVIII. Geological Survey of Canada, Paper 78-7, Stewart, T.G., 1991. Glacial marine sedimentation from tidewater gla- 20pp. ciers in the Canadian High Arctic. In: Anderson, J.B., Ashley, G.M. Mangerud, J., Gulliksen, S., Larsen, E., Longva, O., Miller, G.H., (Eds.), Glacial Marine Sedimentation; Paleoclimatic Signi"cance. Sejrup, H.P., S+nstegaard, E., 1981. A Middle Weichselian ice-free Geological Society of America Special Paper 261, period in Western Norway: The As lesund Interstadial. Boreas 10, pp. 95}105. 447}462. Stuiver, M., Reimer, P.J., 1993. Extended C data base and revised Miller, G.H., Brigham-Grette, J.B., 1989. Amino acid geochronology: CALIB 3.0 C age calibration program. Radiocarbon 35, resolution and precision in carbonate fossils. Quaternary Interna- 215}230. tional 1, 111}128. Sugden, D.E., Watts, S.H., 1977. Tors, felsenmeer, and glaciation in OD Cofaigh, C., 1998. Geomorphic and sedimentary signatures of early northern Cumberland Peninsula, Ba$n Island. Canadian Journal Holocene deglaciation in High Arctic "ords, Ellesmere Island, Ca- of Earth Sciences 14, 2817}2823. nada: implications for deglacial ice dynamics and thermal regime. Tozer, E.T., 1963. Trold Fiord. In: Fortier, Y.O., Blackadar, R.G., Canadian Journal of Earth Sciences 35, 437}452. Glenister, B.F., Greiner, H.R., McLaren, D.J., McMillan, N.J., Nor- OD Cofaigh, C., 1999a. Late Quaternary glaciation and postglacial ris, A.W., Roots, E.F., Souther, J.G., Thorsteinsson, R., Tozer, E.T. emergence, southern Eureka Sound, High Arctic Canada. Unpub- (Eds.), Geology of the North-central Part of the Arctic Archipelago, lished Ph.D. thesis, University of Alberta, Edmonton, Alberta, Northwest Territories (Operation Franklin). Geological Survey of 153pp. Canada Memoir 320, pp. 370. OD Cofaigh, C., 1999b. Holocene emergence and shoreline delevelling, Trettin, H.P., 1991, Geology of the Innuitian Orogen and Arctic Plat- southern Eureka Sound, High Arctic Canada. GeH ographie physique form of Canada and Greenland. Geological Survey of Canada. et Quaternaire, 53, 1}13. Geology of Canada, No. 3, 569pp. OD Cofaigh, C., Lemmen, D.S., Evans, D.J.A., Bednarski, J., 1999. Glacial Troelsen, J.C., 1952. Geological investigations in Ellesmere Island. landform/sediment assemblages in the Canadian High Arctic and Arctic 5, 199}210. their implications for Late Quaternary glaciation. Annals of Gla- Tushingham, A.M., 1991. On the extent and thickness of the Innuitian ciology, 28, 195}207. Ice-Sheet: a postglacial-adjustment approach. Canadian Journal of Reeh, N., 1984. Reconstruction of the glacial ice covers of Greenland Earth Sciences 28, 231}239. and the Canadian Arctic islands by three-dimensional, Zreda, M., Phillips, F., Elmore, D., 1995. The timing of retreat of the perfectly plastic ice-sheet modelling. Annals of Glaciology 5, Greenland Ice-Sheet from Nares Strait based on chlorine-36 buil- 115}121. dup dating . American Geophysical Union, 1995 Fall Meeting: Eos Schei, P., 1904. Preliminary account of the geological investigations Supplement 76 (46), 717. made during the second Norwegian Polar Expedition in the Fram, Zreda, M., England, J., Phillips, F., Elmore, D., Sharma, P., 1999. 1898}1902. In: Sverdrup, O. (Ed.), New Land, Vol. 2. Longhams, Unblocking of the Nares Strait by Greenland and Ellesmere ice- Green and Company, London, pp. 455}466. sheet retreat about 10,000 years ago. Nature 398, 139}142.