J Paleolimnol (2007) 37:331–348 DOI 10.1007/s10933-006-9051-4
ORIGINAL PAPER
Constraints on paleolake levels, spillways and glacial lake history, north-central Ontario, Canada Shawn R. Slattery Æ Peter J. Barnett Æ Darrel G. F. Long
Received: 30 September 2004 / Accepted: 1 July 2006 / Published online: 15 December 2006 Springer Science+Business Media B.V. 2006
Abstract The recognition of ice-marginal deltas constructed during the formation of the Nakina II moraine and a previously unrecognized spillway, in the vicinity of Longlac, northern Ontario, indicates that existing concepts of ancestral lake level history and drainage systems in the Lake Superior–Lake Nipigon region is inadequate. Based on isostatically corrected digital elevation maps, ice-marginal deltas of the Nakina II moraine probably formed at the level of glacial Lake Minong, most likely Minong III, and not glacial Lake Nakina as has been commonly suggested. In addition, the presence of a spillway near Longlac indicates that lake water drained southward through the Mullet Outlet–Pic River system
immediately following ice-marginal retreat from the Nakina II moraine and not eastward as previously proposed. Architectural-element analysis of exposures within the spillway indicates hyperconcentrated outbursts of meltwater produced thick channel-fill elements during flood conditions with peak-velocities exceeding 3 m/s. Subsequent retreat of ice from the Pic River valley to the east, may have allowed waters of Lake Agassiz, Lake Barlow–Ojibway, or both, to drain into post-Minong lake levels in the Lake Superior basin. These findings place major constraints on previously proposed concepts of northeastern or eastern outlets of Lake Agassiz. Keywords Nakina moraines Æ Lake Nakina Æ Lake Minong Æ Lake Agassiz Æ Lake Ojibway Æ Mullet Outlet Æ Pic River
S. R. Slattery (&) Lake Simcoe Region Conservation Authority, 120 Bayview Parkway, Newmarket, ON L3Y 4X1, Canada e-mail:
[email protected] P. J. Barnett Ontario Geological Survey, 933 Ramsey Lake Road, Sudbury, ON P3E 6B5, Canada e-mail:
[email protected] D. G. F. Long Department of Earth Science, Laurentian University, Sudbury, ON P3E 2C6, Canada e-mail:
[email protected]
Introduction Paleolake level studies of the Lake Superior Basin have been underway for well over a century (e.g. Lawson 1893; Taylor 1895, 1897; Leverett 1929; G.M. Stanley unpublished thesis; W.R. Farrand unpublished thesis; Zoltai and Herrington 1967; Zoltai 1967; Saarnisto 1974, 1975; Burwasser 1977; Teller and Thorleifson 1983; A.F. Bajc unpublished thesis; Teller and Mahnic 1988;
123
332
J Paleolimnol (2007) 37:331–348
Lewis and Anderson 1989; Teller et al. 1996; Bajc et al. 1997; Lowell et al. 1999; Breckenridge et al. 2004). Significant discussion and chronological revision over the last 35 years followed the recognition of the Marquette readvance, a late glacial event that moved ice into the Lake Nipigon and Lake Superior drainage basins approximately 10,025 BP (Black 1976; Hughes 1978; Clayton and Moran 1983; Clayton 1983; Drexler et al. 1983; Farrand and Drexler 1985; Lowell et al. 1999). Despite previous analyses, the chronology of deglaciation associated with the earliest post-glacial water levels in the Superior basin is not well established and remains poorly understood. This is due to the fragmented and complex deglaciation history of the Laurentide Ice Sheet (LIS), fluctuating water levels, coupled with rapid isostatic recovery, and a paucity of suitable study sites. The purpose of this paper is to re-examine the lake level history and drainage patterns in the
Lake Superior–Lake Nipigon region in light of new information from the Nakina II moraine east of Longlac, Ontario. First, the paleolake level history of glacial Lake Nakina is re-evaluated in light of the recognition of ice-contact deltas that formed the Nakina II moraine (S.R. Slattery unpublished thesis) and the discovery of a terraced spillway at the southwestern corner of Pagwachuan Lake (Fig. 1). Based on data obtained through Quaternary geological mapping of a 2,000 km2 area located immediately east of Longlac, detailed examination of sedimentoutcrop sections and landform analysis through the use of high-quality digital elevation models (DEMs), a date of approximately 9,000 BP is proposed for the Nakina II moraine. This is significantly older than suggested more recently by Thorleifson and Kristjansson (1993) and Breckenridge et al. (2004) and provides an alternate location of the LIS at approximately 9,000 BP.
86 30'W
50000'N
INSET
N
0
CCaa n aaddaa
Eskers
Eskers Ice-marginal deltas Nakina Moraine
Lake Nipigon INSET Nipigon Marathon Lake Superior Marquette
Highway 11 Highway 625
Wawa
Mullet Outlet-Pic River System
Gros Cap 86000'W
05
Lake Huron
10 Km 0
Lake Ontario
49 30'N
Lake Michigan Lake Erie Fig. 1 Map of Ontario and the Great Lakes. Study area is defined by black coloured box. Inset depicts hillshadeddigital elevation model of the study area. Arrows indicate eskers and several coalescing ice-marginal deltas that form a portion of the Nakina II moraine. The approximate areal
123
extent of the Nakina II moraine is indicated by dashed line. Elevations of ice-marginal deltas are approximately 360 masl. Note Mullet Outlet–Pic River spillway, location denoted by arrow
J Paleolimnol (2007) 37:331–348
Second, isostatically corrected DEMs and information obtained through a detailed sedimentological investigation on a previously undescribed spillway that incises the moraine, are used to recreate the paleogeography of the lake that existed during the formation of the Nakina II moraine. The isostatically corrected DEMs are then used to identify possible meltwater drainage routes between glacial lakes Nakina, Agassiz, Ojibway and a succession of glacial lakes that occupied the Lake Superior drainage basin. Lastly, a discussion is presented on the ramifications of this evidence on pre-existing chronologies. If valid, the proposed date of 9,000 BP for the Nakina moraines would have severe implications on the timing of glacial Lake Agassiz discharge into Lake Superior during the Nipigon Phase (~9,500 to 8,500 BP; Teller and Thorleifson 1983) and the deglacial history of the northern Great Lakes. Although considered to be influential, it is proposed in this study that Lake Agassiz discharge into Lake Superior through Lake Nipigon during the Nipigon Phase was not solely responsible for erosion of the Nadoway Point-Gros Cap sill, a morainic barrier responsible for maintaining high water levels within the Lake Superior basin during the Minong lake phase (~9,500 to 9,000 BP; Fig. 1). Perhaps another mechanism, such as the uncovering of the Pic River Outlet, may be responsible for downcutting of the Nadoway Point-Gros Cap sill to post-Minong levels in the Lake Superior basin.
333
the Nakina I moraine that was overridden by the younger Nakina II moraine (Fig. 2). According to Zoltai (1967) remnants of an overridden interlobate moraine exist south of Chipman Lake (Fig. 2). Both areas are composed of ice-contact stratified sand and gravel that is apparently capped by fine sandy till (Zoltai 1967). It has been over 35 years since these localities have been studied. During this time, several sand and gravel pit operations have opened and numerous forest access roads have been constructed within the area. These ongoing developments have uncovered unstudied exposures of Quaternary-aged strata and have permitted access to areas that were inaccessible to previous researchers (Zoltai 1965, 1967; Zoltai and Herrington 1967). These localities were revisited 0
0
88 00’W 51000’N
86 00’W 51000’N
Ogoki River Ogoki Lake Lake Barlow-Ojibway Nakina II Moraine O’Sullivan Lake Esnagami Lake Nakina I moraine Nakina Nakina II moraine Onaman moraine Chipman Lake Onaman Lake Lake Nakina overridden interlobate moraine overridden Nakina I moraine Wildgoose Lake Caramat Lake Kelvin McKay Lake Sturgeon River Long Lake Rosyln Lake Little Pic River Pic River
The Nakina moraines
Gravel River 0
Post-Minong Lake
49 00’N
The Nakina I and II moraines (Zoltai 1965, 1967) form two distinct, parallel ridges north and northeast of Lake Nipigon. The southernmost ridge, the Nakina I moraine, is considered to be the older of the two and is located approximately 3 km south of the younger Nakina II moraine. The ridges merge south of Esnagami Lake, forming a single morainic system that continues in a southeasterly direction (Fig. 2). North of Pagwachuan Lake (north of Caramat, Ontario), Zoltai (1967) identified a segment of
Lake Superior
0 4 8 12 16 mi
0
12.5
25 km
0
49 00’N
Terrace Bay Marathon
Moraine, overridden Moraine
Fig. 2 Location of the Nakina moraine and the Onaman interlobate moraines. Areas re-investigated during the present survey are enclosed by box. The approximate extent of glacial lakes Nakina, Kelvin, Barlow–Ojibway and a post-Minong lake are depicted in figure as envisioned by Zoltai (1967)
123
334
during the course of Quaternary mapping and associated sedimentological investigations during the summer months of 2001 and 2002 (Fig. 2), (Slattery 2003a, b; S.R. Slattery unpublished thesis). Both areas that Zoltai (1967) suggested to have been overridden contain two suites of genetically linked landforms. The first suite is represented by several eskers, representing former subglacial conduits. In both areas eskers can be traced in a down-ice direction into extensive, ice-contact shoal-water deltas, that represent the second suite of landforms (S.R. Slattery unpublished thesis). Morphologically the eskers in both areas are readily identifiable. They are characterized by a series of sharp-crested, sinuous knife-like ridges. If these landforms were subsequently overridden, pristine sharp-crested landforms would not be expected. Although subglacially derived, a stony sandy till is present along the lower flanks of the eskers. There is no evidence of a till overlying the sand and gravel core in any of the esker or deltaic sections examined. In the prodelta environment, topographic highs interpreted by Zoltai (1967) as components of the Nakina II moraine are bedrock outcrops, not morainic deposits. Northwest of Pagwachuan Lake, areas designated by Zoltai (1967) as the Nakina I moraine, are prodelta facies associated with extensive ice-marginal shoal-water deltas that form the Nakina II moraine. These deposits should not be separated and are part of the same landform. Similar discrepancies are also apparent south of Chipman Lake, where no evidence was found to support the presence of an overridden interlobate moraine, as suggested by Zoltai (1967). A possible rationale for suggesting a local readvance, as envisioned by Zoltai (1967), may be the flat-topped morphology of the delta-plain environments. However, examination of landforms in both localities (north of Pagwachuan Lake and south of Chipman Lake) indicate that these are primary in nature (Slattery 2003a) and are unaltered, meaning that they have not been overridden by an advancing ice-margin, and hence fail to support the idea of a local readvance. South of Chipman Lake, eskers commonly extend for 9 km and are typically greater than
123
J Paleolimnol (2007) 37:331–348
40 m in height. These conspicuous landforms are oriented roughly parallel to the overall direction of ice-flow within the area. This may explain why Zoltai (1967) interpreted this area as an interlobate deposit. A similar reinvestigation may also be required to test the validity of Zoltai’s (1965) interpretation of the Onaman interlobate moraine, southwest of Esnagami Lake. Although reinterpreted by Prest et al. (1968) as an ice marginal feature, it is possible that this landform represents a suite of subglacial conduits, similar to those located north of Chipman Lake. Until there are detailed Quaternary geology and sedimentological investigations on the Onaman interlobate moraine its origin should remain in question. There is no evidence to support, or any apparent reason to separate morainic deposits into the Nakina I and Nakina II moraines within the investigated areas. If an ice-marginal readvance occurred within the area, evidence for this event is not apparent. It is considered more plausible that these deposits are a single morainic belt. However, the question as to what specific advance or retreat (?) these deposits represent still remains. If the observed deposits are related to the Nakina I moraine, then the ice-marginal readjustment which produced the Nakina II moraine (Zoltai 1967), only occurred west of Esnagami Lake. An alternate explanation is that the observed deposits are associated with the younger of the two moraines. It is possible that sediments associated with formation of the Nakina II moraine buried the Nakina I moraine, making it difficult, if not impossible, to distinguish the Nakina I moraine based on morphology or morphological mapping exercises. Studies by S.R. Slattery (unpublished thesis) indicate that the part of the moraine north of Caramat (Fig. 2) is composed of numerous coalescing, shoal-water deltas. Delta topset beds are defined by distributary channel-fill deposits that resemble those associated with shallow, flashy, low-sinuosity braided rivers. Assemblages of low-angle, trough and planar cross-stratified sands and gravels that are incised by the lithotype assemblages that form delta topset beds are interpreted as gently dipping deltaic foreset beds. In the prodelta environment, 8 m thick successions of climbing
J Paleolimnol (2007) 37:331–348
ripple successions, planar-bedded deposits of silty, fine- to medium-grained sands, deposited primarily by bedload and suspension under lower-flow regime conditions, are interpreted as delta bottomset beds. The present-day elevations of topset beds, as defined through the use of a DEM, are approximately 360 m (Fig. 1: Inset). This is similar to elevations of other flat-topped, fan-shaped landforms (357–366 m) of probable deltaic origin, that compose the Nakina I and II moraines west of Esnagami Lake. This observation strongly suggests that observed deposits are associated with ancestral lakes that occupied the Lake Nipigon and Superior basins.
Paleogeographic reconstruction Tops of coalescing ice-contact deltas that form the Nakina II moraine (S.R. Slattery unpublished thesis) define an associated water plane. In order to assess the areal extent of the inundation for this and other proglacial lakes within the study area, isostatically corrected DEMs were created. These were developed in a series of steps in order to compensate for isostatic recovery. To do this, a digital grid was constructed from isobases established for the Lake Superior basin by W.R. Farrand (unpublished thesis). Isobases were plotted at right angles to the maximum direction of uplift (N 30 E) and assigned specific values (elevation in meters) in accordance with the tilt of each water plane as indicated by W.R. Farrand (unpublished thesis). For example, isobases assigned to the Minong III lake phase were constructed using a slope of 0.50 m/km (2.65 ft/mi). Isobases and their assigned elevations, were confirmed by determining whether or not the projected isobase intersected established morphological landforms such as deltas and beach ridges assigned to specific lake phases. The above process created a digital wedge that was subtracted from an unaltered DEM. DEMs for northern Ontario are derived from 1:20,000 scale Ontario Base Maps (OBM) which have a contour interval of 5 m and an elevation accuracy of ±2.5 m. Subtraction of the digital wedge from the unaltered DEM, in theory, restores present day elevations to elevations that existed during
335
shoreline feature formation prior to isostatic recovery. The end product is an isostatically corrected or tilted DEM. By extending the Lake Superior isobases of W.R. Farrand (unpublished thesis) in the same manner as Zoltai and Herrington (1967), A.F. Bajc (unpublished thesis), Geddes et al. (1987) and Teller and Mahnic (1988), from Nipigon, Ontario, in the direction of maximum uplift (N 30 E) to the Nakina II moraine, deltas that compose the moraine correlate to the Minong III water plane of Lake Superior (Fig. 3). Northward extrapolation of the Lake Superior–Lake Nipigon strandline diagram of Teller and Thorleifson (1983: Fig. 3), also demonstrates that water planes associated with the Minong phase of Lake Superior and Early Lake Nipigon stages were confluent. Considering that ice-contact deltas in the Nakina II moraine mark the highest water level present within Long Lake basin, it is plausible to consider the Nakina II moraine as the northernmost shore of the Minong III lake phase of Lake Superior. This reconstruction supports an earlier inference based on investigations in the Geraldton area by Thorleifson and Kristjansson (1993, p. 70) that the Minong levels of Lake Superior were confluent with lakes in the Nipigon and Long Lake basins. By inundating an isostatically corrected DEM using controls assigned to the Minong III lake phase by W.R. Farrand (unpublished thesis), the maximum areal extent of the Minong III lake phase of Lake Superior is portrayed in Fig. 4a. This paleolake level model demonstrates that water in the Lake Superior, Nipigon and Long Lake drainage basins coalesced to form glacial Lake Minong III. If this model is valid, ice-marginal deltas of the Nakina II moraine prograded into glacial Lake Minong III and not glacial Lake Nakina, as is commonly suggested (Fig. 2), (Zoltai and Herrington 1967; Zoltai 1967).
Evidence for a new spillway A channel-like feature, 5.5 km in length and approximately 850 m wide, can be traced from Pagwachuan Lake in a southwesterly direction to highway 625 where it becomes concealed by eolian sediments in the form of parabolic dunes
123
336
J Paleolimnol (2007) 37:331–348 Profiles of Former Water Planes in Western and Northern Lake Superior Basin feature identified during present study any feature identified by Lawson or Stanley wave-cut notch, bluff beach ridge
1300 (396 m)
0 0
10
20 mi
15
30 Km
1100 (335 m)
GI MINON G III MINON
1000 (304 m)
G II MINON
NIPISSING
ALGOMA SAULT
Red Sucker Pt. Marathon Heron Bay
Pic Island
Jackfish
Terrace Bay
Schreiber
Cavers Rossport
Kama Bay
SUB-SAULT
Nipigon
i (0.50
m/km) m/km)
/km) i (0.42 m 2.20 ft/m 1.75 ft/mi (0.
33 m/km) m/km)
HOUGHTON
NIPISSING
700 (213 m)
2.65 ft/m
.33 1.75 ft/mi (0
DORION
800 (243 m)
i (0.50
INONG I POST-M
ONG II POST-MIN OG III POST-MIN
900 (274 m)
600 (182 m)
m 2.65 ft/
2.65 ft/mi (0.50 m/km):rate of uplift (slope of water plane)
0.65 ft/mi (0.12 m/km) 0.50 ft/mi (0.09 m/km) 0.33 ft/mi (0.06 m/km) 0.25 ft/mi (0.05 m/km)
Nakina II Moraine (Delta)
Feet Above Sea Level
delta 1200 (365 m)
Fig. 3 Northward extrapolation of former water planes of the Lake Superior Basin from Nipigon, Ontario to the Nakina II moraine in the direction of maximum uplift
(N 30 E). Symbols on isobases denote landforms (e.g. deltas, wave-cut notches) associated with former water planes (modified from W.R. Farrand unpublished thesis)
that are at least 5 m in thickness. This channel, herein named the Mullet outlet, becomes apparent again approximately 650 m west of highway 625, where it can be traced for an additional 700 m to the Pic River (Fig. 1: Inset). The uppermost surface of the channel is at a present-day elevation of 338 m. Paired terraces occur at a present-day elevation of 323 m whereas the base of the channel is at an elevation of 303 m. Architectural analysis was undertaken on sediments exposed in a trench constructed through the terrace of the proposed spillway at the southwestern corner of Pagwachuan Lake (Fig. 5a). The trench is oriented nearly perpendicular to the direction of paleoflow through the channel-like feature.
packages which can be related to specific depositional elements, that range in scale from bedforms, to macroforms (bars), channels and deposystems (Allen 1983; Miall 1985). Photomosaics were constructed from overlapping digital photographs of the trench face. Paleoflow measurements and directional attributes of bounding discontinuities were recorded directly onto the photomosaic during field investigations. Bounding surfaces were subsequently identified and later classified in hierarchical order (McKee and Weir 1953; Miall 1985). Directional attributes of all measured surfaces were subsequently re-plotted on line drawings made from the photomosaic, as symbols and summary rose diagrams, oriented so that measurements directed above the horizontal are directed away from the observer and those directed below are towards the observer (Fig. 5b). The strata between bounding surfaces were divided into genetically related architectural elements based on lithotype assemblage, internal geometry and
Architectural analysis Architectural analysis is based on the concept that strata can be divided into genetically related
123
J Paleolimnol (2007) 37:331–348
337 87000”W
88000”W
86000 W
85000 W N
A Laurentide Ice Sheet Onaman Lake 0
0
50 00”N
50 00”N
Initial formation of ice-marginal deltas that form the Nakina II moraine L. Nipigon Long Lake
49000”N
49000”N 0
Lake Superior 88000”W 0
88 00”W
25
50 Km
87000”W
86 00”W
85 00”W
87000”W
86000 W
85000 W
0
0
B
N
Laurentide Ice Sheet Onaman Lake
50000”N
50000”N ice-marginal deltas of the Nakina II moraine
L. Nipigon Long Lake
49000”N
49000”N 0
Lake Superior
88000”W
87000”W
86000”W
25
50 Km
85000”W
Fig. 4 Paleogeographic reconstructions. (a) Inundation of an isostatically corrected DEM to the Minong III lake phase of Lake Superior when ice rested at the Nakina II moraine. Areas in black depict the maximum areal extent of glacial Lake Minong III. Note initial formation of icecontact deltas that form part of the Nakina II moraine. The extent of present-day lakes is labelled and coloured white. (b) Inundation of the DEM model to the flat surface slightly above the Mullet outlet (a present-day elevation of 340 masl). It is possible that glacial Lake Minong III
remained connected to water in the Long Lake drainage basin through straits south of Lake Nipigon. This would call for a 20 m drop in water level in response to downcutting located at Nadoway Point, Michigan and Gros Cap, Ontario, which controlled the water levels of ancestral Lake Superior. This drop in water level may be related to glacial Lake Agassiz overflow into Lake Superior or progressive erosion of the barrier during isostatic recovery. Note the development of ice-contact deltas that form a part of the Nakina II moraine
directional attributes (Fig. 5c). Distinct elements were numbered and letter-coded (Miall 1985, 1996; Fig. 5d). All architectural elements with
representative lithotype codes (modified from Miall 1978), are indicated on the line drawings, and are described below.
123
338
J Paleolimnol (2007) 37:331–348 Profile Key
A
direction of 0th order surfaces th
direction and dip of 0 order and higher surfaces major boundary surface th azimuth of 0 order and higher surfaces number of measurements n V directional variance Sl lithofacies code 5B-CH architectural elements numbered in sequence for each profile 15
3
1
0
2m
B 4 3
3 3
3
N
324
0
Face Orientation
3
Paleoflow (0th order: foreset lamina) E
W
C =208 n=5 0 d=28
0
0 =214 =25 n=26 n=5 0 d=23 d=47.50 V=1046
=215 n=2 d =190
0
=440 n=10 d=460
st
=2180 n=5 d=210
144
D
S
nd
rd
1 , 2 , 3 order and higher boundary surfaces with angle of dip
=2320 n=4 0 d=27
=2100 n=6 d=17.50
0
0
=216 n=3 0 d=21
0
Gt 2-CH
St
1B-CH
1A-GLSE St
1A-GLSE St
1B-CH
Sr(A), Sr(B), Sr(T) St 1B-CH
1B-CH
Fig. 5 Mullet Outlet terrace trench. (a) Photomosaic of the trench constructed within the Mullet outlet. The orientation of trench is nearly perpendicular to the direction of paleoflow. (b) Directional attributes of observed bounding surfaces. Major boundary surfaces are numbered and circled. Measurements directed above the horizontal are away from the observer; those directed below are toward the observer. Solitary arrows indicate
the directional attributes of 0th order surfaces; pins indicate direction and dip of higher order surfaces. (c) Summary rose diagrams of directional attributes. Variance calculations are based on the methodology of Miall (1994). (d) Architectural element codes are displayed in boxes. Lithotype codes are displayed within architectural elements
Glaciolacustrine sheet element (GLSE)
In element 1-CH (Fig. 5d), dislocated and rotated remnants of units of Sr-A, Sr-B and Sr-T commonly occur as blocks within the element due to subsequent liquefaction or adjacent erosion. Blocks are typically 3 m thick and laterally continuous for at least 9 m (Fig. 5d). Vertical and lateral ripple-drift transitions (from type A to type B, and type B to type A) within element GLSE reflect fluctuating rates of deposition that can be directly related to the rate and mode of ripple migration (Jopling and Walker 1968; Allen 1969; Gustavson et al. 1975). Type-B ripple-drift sequences indicate a reduction in underflow current velocity. This subdued slip-face
GLSE represents remnants of drape-like, conformable, sheet-like bodies that form the flanks of the uppermost terraced surface (Fig. 5a). Where exposed above the uppermost terraced surface, GLSE is approximately 3–7 m thick and can be traced laterally for at least 90 m. The element is composed of ripple-drift sand sequences with cosets of type-A (lithotype SrA), type-B (lithotype Sr-B) and transitional ripple-drift varieties (Sr-T), of Jopling and Walker (1968), that are bounded by 2nd order surfaces (Fig. 5b, c).
123
J Paleolimnol (2007) 37:331–348
development and enhanced rate of deposition is due to increased deposition from suspension (Jopling and Walker 1968; Allen 1969, 1972). Type-A ripple-drift sequences were deposited by density underflow currents, largely through traction processes, under lower-flow-regime conditions at ~0–20 cm/s (Jopling and Walker 1968; Allen 1984). Soft-sediment deformation, in the form of liquefaction features and dislocation of blocks containing ripple-drift sequences, is not related to incipient ripple formation by density underflow currents. It is more likely to have occurred during rapid downcutting of the channel-like feature and associated emplacement of the adjacent and overlying CH elements (1-CH and 2-CH). Softsediment deformation is addressed in detail by Ankertell and Dzulynski (1968) and Allen (1984). Ripple-drift sequences similar to those observed within element GLSE have been documented by many authors in glaciolacustrine settings (Jopling and Walker 1968; Gustavson et al. 1975; Shaw 1975; Ashley 1975; Benn 1989). Ripple-drift successions in element GLSE are also interpreted as glaciolacustrine in origin. Channel-fill elements (CH) CH elements are 0.6–5.5 m thick, 25–95 m wide, sheet-like bodies that are bound by erosive concave-up basal 3rd and 4th order surfaces (Fig. 5a, b). They form concave-up, lenticular and tabular bodies that are predominantly composed of trough-cross-stratified gravel (Gt), sandy erosional scour fills (Ss) and lesser amounts of pebbly medium to very coarse grained trough-crossstratified sand (St). Element 1-CH (Fig. 5c, d) is composed of lithotypes St, Gt and reworked blocks of element GLSE. Typically, troughs associated with lithotype Gt are 0.2–0.6 m deep and 0.5–2.5 m wide. Basal contacts (1st order surfaces) are highly erosive, and in many cases are marked by lag deposits composed of boulder to cobble-sized material. Inverse to normal (coarse-tail) grading is apparent where boulder-sized material at the base of a trough passes vertically into cobblesized material at the top. Troughs typically occur
339
in grouped sets and are seldom present as isolated, solitary forms. Scours (Ss) are 0.25 m deep and 0.5 m wide and are overlain by medium to coarse grade sand. Lowermost 1st order surfaces are erosive and commonly incise the tops of lithotypes Gt and St. Pebbly medium to coarse trough-crossstratified sand (St) is rare within element 2-CH but is the dominant macroform in element 1-CH. Troughs (St) are typically 0.15–0.4 m deep and 0.20–1.5 m wide and lower 1st order surfaces, which are marked by lags of granule to large pebble grade material, commonly incise into underlying St and Ss units. Paleocurrent analysis within CH elements indicates strongly unimodal currents, with only minor dispersal of vectors. Tangential St and Gt 0th order surfaces (foresets) are directed in a southwesterly direction, towards the Pic River, coincident with 1st order set surfaces that dip in the same downcurrent direction at angles of approximately 28. Trough-cross-strata (St and Gt) between 3rd and 4th order element boundary surfaces are oriented nearly opposite to lowermost element boundary surfaces that dip in an up-current direction at an angle of approximately 46 (Fig. 5c).
Interpretation of architectural elements Architectural elements identified within the channel-like feature are here interpreted in terms of local deposition within a spillway. It is proposed that this spillway acted as a conduit for meltwater discharge originating from an ice-marginal lake that occupied a portion of Pagwachuan Lake basin (Fig. 1: Inset). Textural and geometrical variation amongst elements and individual lithotypes as observed in sediment-outcrop sections within the spillway, reflect deposition of sediments under fluctuating flow regimes, similar to that described by several authors (e.g. Allen 1984; Miall 1996). The directional attributes of elements and the depositional mechanisms responsible for their emplacement are addressed below. Element GLSE is clearly of glaciolacustrine origin, and represents part of a more extensive
123
340
sheet deposited prior to the development of the spillway. During deposition of overlying and physically adjacent CH elements, it acted as a weakly consolidated bank material. Element 1-CH was characterized by the migration of sinuous crested dunes, under upper lower-flow regimes (Allen 1984). It is likely that erosional scours (Ss) found within the element were formed by turbulent eddies or vigorous currents of transitional to upper flow conditions (Allen 1984). The presence of blocks of GLSE within element 1-CH, the oldest of the two channel elements, most likely reflects bank collapse during downcutting of the spillway. Blocks appear to have been dislocated, rotated and in part liquefied during incorporation and rapid burial in element 1-CH. Element 2-CH represents a second major pulse of meltwater flowing through the channel-like feature. Scoop-shaped units of Gt within element 2-CH are interpreted as the products of migrating slugs of gravel. Inverse, coarse-tail grading within Gt units is suggestive of deposition from turbulent eddies under transitional to upper flow conditions (Lowe 1982). Erosive, turbulent eddies are required to explain the up-current dipping lowermost 3rd and 4th order element boundary surfaces of elements 1-CH and 2-CH. Weak stratification within element 2-CH may be due to an increase in fluid turbulence (Lowe 1982), or local development of hyper-concentrated flow conditions. Element 2-CH which appears to be the younger of the two channel elements, incises all underlying units (element 1-CH and GLSE) and is interpreted to have formed during peak-flow conditions, during a period of exceptionally highsediment supply within the spillway. Rare troughcross stratified gravel within the element may represent small-scale channel fill sequences deposited during falling stage conditions. Following the method of Fetter (2001), peakflow velocities within element 1-CH and 2-CH are estimated to have reached 2.7 m/s and 3.8 m/s. Flow velocities within CH elements are comparable to those reported and interpreted by Matsch (1983) as flood related, hence CH elements within the channel-like feature are interpreted to have been deposited under flood-like conditions.
123
J Paleolimnol (2007) 37:331–348
Paleogeographic reconstruction of the lake or lakes controlled by the Mullet outlet The recognition of the spillway that breached the Nakina II moraine at the southwestern end of Pagwachuan Lake (Fig. 1: Inset), referred to herein as the Mullet outlet, places constraints on the possible configuration of glacial Lake Nakina. The Mullet outlet is assumed to be related to rapid meltwater discharge from a local ice-marginal lake that occupied a portion of Pagwachuan Lake basin. If ice became pinned on topographically high areas (bedrock ridges), it is plausible to assume that small reservoirs of meltwater would collect and pond within the Pagwachuan Lake basin. Once meltwater reached and subsequently exceeded the elevation of the Nakina II morainic barrier, erosion of the barrier would inevitably commence. Continued erosion coupled with an influx of meltwater from the ice-front, would direct overflow into the Mullet outlet, an area of low threshold across the continental divide. Paleocurrent analysis of deltaic and prodeltaic facies associated with the Nakina II moraine demonstrates that during the drainage of glacial Lake Minong III, flow convergence occurred south of the moraine and was subsequently directed or funnelled through the Mullet outlet toward the Pic River. The Pic River was also considered by Saarnisto (1975) to be an area of low threshold across the continental divide. Detailed sedimentological investigations of strata within the Mullet outlet support this finding and clearly point to the Pic River as the conduit for meltwater discharge. For the Mullet outlet to have been operational, two conditions must have been met. First, the outlet must have been free of ice, indicating that the ice-margin retreated northeastwards from the current position of the outlet and/or the Nakina II moraine. Secondly, water levels north of the drainage divide were at or slightly above the elevation of the uppermost terraced surface of the outlet. Inundation of the DEM model at a present-day elevation of 340 m, an elevation which corresponds to the uppermost flat-surface of the Mullet outlet and approximately 20 m lower than deltas
J Paleolimnol (2007) 37:331–348
which form the Nakina II moraine suggests that water levels of the Minong III lake phase may have remained connected to water within the Long Lake drainage basin east of Lake Nipigon (Fig. 4b). If valid, a 20 m drop in water level occurred during isostatic recovery and prior to the activation of the Mullet outlet. This drop in water level may reflect a combination of differential isostatic recovery of the Superior basin, and downcutting of the morainic barrier at Nadoway Point, Michigan. This barrier controlled the lake levels of the Lake Superior basin, possibly in response to glacial Lake Agassiz overflow through Lake Nipigon (Teller and Thorleifson 1983). The absence of distinct morphological features indicative of phase-like evacuation or pauses in lake drainage (draw down), including wave-cut notches, bluffs or beach ridges on the delta plain environments (Fenton et al. 1983; Kehew and Clayton 1983), indicate unimpeded drainage of glacial Lake Minong III to an elevation at or just below the Mullet outlet. This implies that water levels in the Long Lake basin dropped approximately 37 m (from the delta plain to the terraced surface of Mullet outlet) prior to activation of the outlet which is assumed to have commenced shortly after ice-marginal retreat from the Nakina II moraine. Inundation of the DEM model to the terraced surface of the Mullet outlet, demonstrates the maximum areal extent of glacial Lake Nakina following the drainage of glacial Lake Minong III. The areal extent and upper limit of inundation for glacial Lake Nakina is re-interpreted in accordance with Fig. 6a. Following ice-marginal retreat from the Nakina II moraine, the Mullet–Pic River outlet may have continued to be used as waters of Lakes Agassiz and/or Ojibway inundated the low terrain north of the Nakina moraines (Fig. 6b). A preexisting sediment dam across the Pic River valley located upstream of the junction between the Mullet and Pic River outlets (Boissonneau 1965; Gartner and McQuay 1980) would have to be eroded before southward drainage into Lake Superior switched from the Mullet outlet to the main outlet along the Pic River. The subsequent, substantial erosion that has occurred along the Pic River (A.F. Bajc unpub-
341
lished thesis; Slattery 2003b) may have been the result of catastrophic drainage of waters from Lake Agassiz, Lake Ojibway or both (Fig. 6b).
Discussion Following the Marquette readvance at ~10,025 BP, deglaciation of the Lake Superior drainage basin is assumed to have been rapid, facilitated by a calving ice-margin (A.F. Bajc unpublished thesis; Bajc et al. 1997). Continued northward recession of the ice margin (north of the Keweenaw Peninsula) allowed glacial Lake Minong to expand in a westward direction, invading newly deglaciated parts of the Lake Superior drainage basin. By ~9,500 BP, glacial Lake Minong was established throughout the Lake Superior basin (W.R. Farrand unpublished thesis; Zoltai and Herrington 1967; Prest 1970; Saarnisto 1974, 1975; Drexler et al. 1983). Glacial Lake Minong was thus the first and highest in a succession of proglacial lakes which occupied the whole Lake Superior basin following the Marquette readvance (W.R. Farrand unpublished thesis). According to W.R. Farrand (unpublished thesis) and Zoltai (1965) glacial Lake Minong was bordered by ice at the Nipigon moraine and along the north shore of Lake Superior. Later work indicates that the water level of glacial Lake Minong was controlled by a morainic barrier located at Nadoway Point, Michigan and Gros Cap, Ontario (Prest 1970; Saarnisto 1975; Cowan 1978) and not by ice as envisaged by W.R. Farrand (unpublished thesis). To date, no datable material has been located in beach or deltaic sediments associated with the lake along the north shore of Lake Superior (A.F. Bajc unpublished thesis). Three distinct water levels have been assigned to the Minong phase of Lake Superior: Minong I, II and III (W.R. Farrand unpublished thesis). It has been suggested by A.F. Bajc (unpublished thesis) that the two lower water levels, Minong II and III, formed shortly after glacial Lake Minong I, in response to downcutting of the morainic threshold. In contrast, Prest (1970) suggested that the Minong lake phase extended north of Lake Superior along the Pic River valley to the Nakina
123
342
J Paleolimnol (2007) 37:331–348 0
0
0
0
86 00 W
87 00”W
88 00”W
A
85 00 W N
Laurentide Ice Sheet
ice-marginal deltas
Onaman Lake 50000”N
50000”N
glacial Lake Nakina
ice-marginal delta Pagwachuan Lake Mullet outlet L. Nipigon Long Lake Pic River Valley
49000”N
49000”N 0
Lake Superior 88 00”W
87000”W
86 00”W
88000”W
87000”W
86 00 W
0
50 Km
0
85 00”W
0
85 00 W
0
0
Laurentide Ice Sheet
O'Sullivan Lake Esnagami Lake
B
25
N
Lake Agassiz overflow 0
0
50 00”N
50 00”N Onaman Lake
Chipman Lake
ice-marginal delta
Pagwachuan Lake
L. Nipigon Lake Ojibway overflow Long Lake White Otter River Pic River valley
49000”N
49000”N Lake Superior 0
88 00”W
0
87000”W
0
86 00”W
25
50 Km 0
85 00”W
Fig. 6 Paleogeographic reconstructions during and after Mullet Outlet operation. (a) Areas in black depict the maximum areal extent of glacial Lake Nakina. Following ice-marginal retreat from the Nakina II moraine the Mullet outlet was activated. Glacial Lake Nakina II is assumed to have drained through the Mullet outlet into the lower reaches of the Pic River valley and into Lake Superior (denoted in figure by dashed white line with arrow). (b) Following ice-marginal retreat from the Nakina II moraine and deglaciation of the Pic and White
Otter river valleys, glacial Lake Agassiz overflow may have been directed behind the Nakina II moraine into Lake Superior through the Pic River. This drainage route is depicted by white line with arrowhead. At this time overflow from glacial Lake Ojibway may have been partially funnelled into the White Otter River and into post-Minong lake levels in Lake Superior through the Pic River. Severe downcutting that has occurred along the Pic River may have been the result of catastrophic drainage of waters from Lake Agassiz, Lake Ojibway or both
moraines. He further speculated that partial drainage of glacial Lake Barlow–Ojibway into the early stages of glacial Lake Minong occurred at
approximately 9,500 BP (see Figure XII-16 of Prest 1970). According to Saarnisto (1974), if glacial Lake Barlow–Ojibway drained into Lake
123
J Paleolimnol (2007) 37:331–348
Superior, it occurred after 9,000 BP, following northward recession of the ice-margin from the Nakina moraines, which he estimated to have occurred after 9,000 BP. Coincidentally, the Nipigon Phase of glacial Lake Agassiz had commenced (Teller and Thorleifson 1983). During the Nipigon Phase, the level of glacial Lake Agassiz was lowered approximately 200 m, from the lower Campbell level to the Gimli level between 9,500 BP and 8,000 BP (Teller 1985; Leverington and Teller 2003; Breckenridge et al. 2004). Teller and Thorleifson (1983) suggest that glacial Lake Agassiz drained catastrophically through present day Lake Nipigon and into the Lake Superior drainage basin as progressively lower outlets, located west of Lake Nipigon, became ice-free. Farrand and Drexler (1985) suggest that these floods caused downcutting of the morainic barrier at Nadoway Point, Michigan, which controlled the post-Minong levels of the Lake Superior basin. Wood collected by Zoltai (1965) near Rosslyn, from below beach sediments, assumed to be related to the uppermost post-Minong lake phase of the Lake Superior basin, was radiocarbon dated at 9,380 ± 150 (Geological Survey of Canada-GSC-287). According to A.F. Bajc (unpublished thesis), this post-Minong shoreline is representative of the post-Minong I water level, and can be traced inland to the Nakina moraines, suggesting that ice had retreated beyond the north shore of Lake Superior well before 9,380 BP. The elevations of fossil shell assemblages collected below the post-Minong I water plane, near Caramat, have been used by Zoltai and Herrington (1967) to infer that the Nakina moraine formed the northern shore of this lake. A date of 8,200 BP for the Nakina moraines has been proposed by Teller and Thorleifson (1983) and Teller et al. (1996), that coincides with glacial Lake Agassiz discharge into Lake Superior. This is significantly younger than the ~9,400 BP suggested by Zoltai (1967) and A.F. Bajc (unpublished thesis), and the 9,000 BP age suggested by Saarnisto (1974, 1975), based on pollen stratigraphy and radiocarbon dating techniques. A key factor is thus the age of the Nakina moraines. According to Saarnisto (1975), A.F. Bajc (unpublished thesis) and Thorleifson and Kristjansson
343
(1993), high-Minong and post-Minong lake levels were maintained in Lake Superior until ~9,000 BP. If the Nakina II moraine marks the Minong III or postMinong I lake phase of Lake Superior it is plausible to consider these moraines as older than the postMinong III-IV (Dorion) transition which occurred at 8,200 BP (Bajc et al. 1997). A date of 9,060 ± 150 BP of basal gyttja from Jock Lake, west of Marathon, dates the fall of water levels below the earliest postMinong shoreline (Saarnisto 1975). Consequently an age of 8,200 BP (Teller and Thorleifson 1983; Teller et al. 1996) seems far too young for the Nakina moraines. A tentative date of ca. 9,000 BP, as suggested by Saarnisto (1975) is feasible for the Nakina moraines, an age that is comparable to Zoltai’s (1965, 1967) original date of 9,400 BP. As isostatic recovery continued, the White Otter–Pic River junction near Caramat, Ontario (Fig. 2) (the present position of the Great LakesHudson Bay drainage divide) was uplifted enough to sever the connection between glacial lakes that formed south of the drainage divide from those that formed north of the drainage divide (glacial lakes Nakina and Kelvin), (Zoltai 1965, 1967; Zoltai and Herrington 1967). Zoltai (1967) referred to the lake north of the drainage divide as glacial Lake Nakina, and tentatively referred to water located south of the drainage divide as a postMinong lake. Zoltai (1967) suggests that glacial Lake Nakina drained through the Jellicoe and Pikitigushi spillways into the low stages of glacial Lake Kelvin (ancestral Lake Nipigon). Prest et al. (1968) later suggested that the Jellicoe spillway extended north of Longlac, Ontario. Recent investigations by Thorleifson and Kristjansson (1993) indicate that glacial Lake Nakina drained east to Long Lake and that the Jellicoe spillway did not act as an outlet for glacial Lake Nakina. The paleogeographic reconstruction (Fig. 6b) indicates that the main connections between glacial lakes Kelvin in the Lake Nipigon basin and Nakina are in the Onaman Lake area. According to Teller and Thorleifson (1983) and Teller et al. (1996), overflow from glacial Lake Agassiz during the Nipigon Phase entered glacial Lake Nakina through several outlets associated with the Pikitigushi system. As the icemargin receded northward from the Nakina and Agutua moraines, overflow from glacial Lake
123
344
J Paleolimnol (2007) 37:331–348
Agassiz into glacial Lake Nakina ceased, initiating the Ojibway Phase of glacial Lake Agassiz at ~8,500 BP (Teller and Thorleifson 1983). During this phase, glacial Lake Agassiz drained eastwards, north of the Nakina moraines, coalescing with the Kinoje´vis phase (~9,000 to 8,200 BP) of glacial Lake Ojibway (Teller and Thorleifson 1983; Teller 1985). Lake Ojibway drained south into the Ottawa River (Hughes 1965; Vincent and Hardy 1979; Lewis and Anderson 1989). Perhaps there is still some merit to Prest’s (1970) speculation that envisages a connection between Lake Ojibway and Lake Superior. Following ice-marginal retreat from the Nakina II moraine, sometime between ~8,200–9,000 BP (Saarnisto 1975; Thorleifson and Teller 1983; Teller et al. 1996), several small ice-contact lakes existed, such as the one that occupied the Pagwachuan Lake basin. These lakes later coalesced with glacial Lake Ojibway (Zoltai 1965), which was rapidly inundating newly deglaciated terrain in northeastern Ontario (Vincent and Hardy 1979). Evidence of glacial Lakes Agassiz–Ojibway is provided by 3–5 m thick exposures of rhythmically bedded silts and clays located approximately 9 km north of the Nakina moraine (Slattery 2003a); (Fig. 7).
0
50 00'N
86030'W
Fig. 7 One of several varved sediment-outcrop sections located north of the Nakina II moraine. Encircled letter on map depicts location of inset. Inset depicts rhythmically bedded (varved) silts (lighter coloured beds) and clays (darker coloured beds) of glacial Lake Agassiz–Ojibway. Note upward thinning of silt beds and the equal thickness of clay beds
Following ice-marginal retreat from the Nakina II moraine, Lake Ojibway overflow could have been directed into the Pic River through the White Otter River and into a post-Minong lake phase of the Lake Superior basin (Fig. 6b). If valid, this model may explain the unaccounted-for 60 m drop in water level of glacial Lake Ojibway as reported by R.C. Paulen (unpublished thesis) during the Angliers phase and early Kinoje´vis phase, between 9,000 BP and 8,200 BP. This substantial discharge event would explain the presence of overturned varves and contorted beds of silt and clay observed south of the Nakina II moraine (A.F. Bajc unpublished thesis). Interclasts of openwork gravel, interpreted as beach facies within dislocated units of glaciolacustrine silts and clays by A.F. Bajc (unpublished thesis) along the upper reaches of the Pic River (north of Marathon, Ontario) could also be explained by Lake Ojibway discharge. Lake Superior varved sequences collected by Teller and Mahnic (1988) indicate enhanced rates of sedimentation following the Dorion phase of Lake Superior at approximately 8,200 BP (Thorleifson and Kristjansson 1993; Teller et al. 1996; Breckenridge et al. 2004). Although these sequences have been interpreted as temporal
A
N
A
Ice-marginal deltas Nakina Moraine Highway 11
INSET Highway 625
0
86 00'W
0
123
5
10 Km
0
49 30'N
J Paleolimnol (2007) 37:331–348
equivalents of the Nakina moraines (Thorleifson and Kristjansson 1993; Teller et al. 1996; Breckenridge et al. 2004), they could easily be explained by Lake Ojibway or combined Ojibway– Agassiz discharge into the Lake Superior basin. Continued erosion of the morainic threshold at Nadoway Point, Michigan and Gros Cap, Ontario caused the water level in the Lake Superior drainage basin to stabilize at the low water Houghton level at ~8,000 BP (Farrand and Drexler 1985). Further discussion regarding the Nipissing, Algoma, Sault and Sub-Sault lake levels is addressed by Karrow (1989).
345
3.
Conclusions From our research, several plausible scenarios have been developed concerning the deglacial chronology of the Nakina moraines and glacial Lake Nakina. These scenarios are described below. 1.
2.
Ice-marginal deltas of the Nakina II moraine probably formed into a level of glacial Lake Minong, most likely Minong III, and not glacial Lake Nakina as is commonly suggested (Zoltai and Herrington 1967; Zoltai 1967; Farrand and Drexler 1985; Dredge and Cowan 1989). Even if these deltas formed into a lower, post-Minong lake phase (postMinong I), water in the Long Lake basin is related to the Lake Superior drainage basin and is not representative of an independent water plane, previously interpreted to be that of glacial Lake Nakina. If the Nakina moraines represent the northernmost shore of the Minong III or the earliest post-Minong I lake phase, a date of approximately 9,000 BP (Saarnisto 1975) is plausible. However, this would have severe implications on the timing of glacial Lake Agassiz discharge into Lake Superior during the Nipigon Phase. If ice were resting at the Nakina moraines at ~9,000 BP this would imply that all Lake Agassiz’s eastern outlets, those of which carried ‘catastrophic’ overflow into Lake Superior and have been deemed responsible for ‘step-like’ erosion of the
4.
5.
Nadoway Point-Gros Cap sill, would have been operational. Although influential, Lake Agassiz discharge into Lake Superior through Lake Nipigon would not be solely responsible for erosion of the Nadoway Point-Gros Cap sill. Lake Agassiz overflow through Lake Nipigon may have only resulted in downcutting of the Nadoway Point-Gros Cap sill to water levels represented by the Minong lake phase of the Lake Superior basin. Perhaps another mechanism, such as the uncovering of the Pic River, may be responsible for downcutting of the sill to post-Minong levels in the Lake Superior basin. Following ice-marginal retreat from the Nakina moraines, waters of Lake Agassiz and Ojibway may have been directed towards the Pic River. This influx of meltwater into the Superior basin would have resulted in severe downcutting of the Nadoway Point-Gros Cap sill, resulting in low post-Minong water levels in the Lake Superior basin. Substantial erosion that has occurred along the Pic River may have been the result of catastrophic drainage of waters from Lake Agassiz, Lake Ojibway or both. Following the drainage of water within the Long Lake basin to a level represented by the Mullet outlet and ice-marginal retreat from the Nakina moraines, only small ice-marginal lakes persisted east of Long Lake (Fig. 6a, b). Significant ponding of meltwater appears to have occurred northwest of Long Lake as envisioned by Zoltai (1965, 1967). By definition, these lakes represent glacial Lake Nakina. The possibility exists that these lakes could have coalesced with glacial lakes Ojibway and/or Agassiz and partially drained through the Pic River into Lake Superior, prior to Lake Ojibway overflow through the Ottawa River. Perhaps Coleman’s (1909) initial interpretation that glacial Lake Nakina was a precursor for glacial Lake Ojibway should not have been abandoned. Interpretations by Thorleifson and Kristjansson (1993) which indicate that the Jellicoe spillway did not act as an outlet for glacial Lake Nakina are considered valid. However, eastward drainage of glacial Lake Nakina into Long Lake (Thorleifson and Kristjansson
123
346
J Paleolimnol (2007) 37:331–348
1993) is not supported by the present study. Observations in accordance with Fig. 6a indicate that glacial Lake Nakina drained southward through the Mullet outlet–Pic River valley into Lake Superior. Spillways observed by Thorleifson and Kristjansson (1993) that exceed present-day elevations of 338 m are most likely associated with the drainage of a Minong lake phase from the Long Lake basin and are not representative of glacial Lake Nakina. The revised paleolake reconstructions presented in this paper should help in piecing together the commonly neglected, complex deglaciation history of northwestern Ontario. Our studies indicate that future research should be directed toward the possible connections between glacial Lake Kelvin and the final stages of glacial Lake Nakina, a direct date of the Nakina moraines, and a more precise reconstruction of the Minong and postMinong waterplanes based on detailed sedimentological investigations, morphological mapping and high-quality DEMs. Acknowledgements The project was made possible by financial assistance from the Lake Simcoe Conservation Authority, the Sedimentary Geoscience Section of the Ontario Geological Survey, and an NSERC discovery grant to D.G.F. Long. We would like to thank T.F. Morris, L.H. Thorleifson, A. Bajc, R.I. Kelly and C.L. Baker for discussions on the deglaciation history of the Lake Superior basin. Comments made by C.F.M. Lewis, P.F. Karrow, D. Sharpe and A. Breckenridge improved the manuscript. The DEM used in this communication was provided by staff of the Ontario Ministry of Natural Resources. We also acknowledge J. Shirota for his help in the isostatic adjustment process.
References Allen JRL (1969) On the geometry of current ripples in relation to stability of fluid flow. Geogr Ann 51A:61–96 Allen JRL (1972) A theoretical and experimental study of climbing ripple cross lamination with a field application to the Uppsala Esker. Geogr Ann 53A:157–187 Allen JRL (1983) Studies in fluviatile sedimentation: bars, bar-complexes and sandstone sheets (low-sinuosity braided streams) in the Brownstones (L. Devonian). Welsh Borders. Sed Geol 33:237–293 Allen JRL (1984) Sedimentary structures, their character and physical basis: developments in sedimentology 30. Elseveir, New York
123
Ankertell JM, Dzulynski S (1968) Patterns of density controlled convolutions involving statistically homogenous and heterogenous layers. Ann Soc Ge´ol Pol 3:401–409 Ashley GM (1975) Rhythmic sedimentation in glacial Lake Hitchcock, Massachusetts-Conneticut. In: Jopling AV, McDonald BC (eds) Glaciofluvial and glaciolacustrine sedimentation, SEPM, Spec Pub 23, pp 282–303 Bajc AF, Morgan AV, Warner BG (1997) Age and paleoecological significance of an early postglacial fossil assemblage near Marathon, Ontario, Canada. Can J Earth Sci 34:687–698 Benn D (1989) Controls on sedimentation in a late Devensian ice-damned lake, Achnasheen, Scotland. Boreas 18:31–42 Black RF (1976) Quaternary geology of Wisconsin and contiguous upper Michigan. In: Mahaney WC (ed) Quaternary stratigraphy of North America. Dowden Hutchison and Ross Inc., Stroudsburg, Pennsylvania, pp 93–117 Boissonneau AN (1965) Surfical geology, Algoma-Cochrane. Ontario Department of Lands and Forests, Map 5365, scale 1:506 880 Breckenridge A, Johnson TC, Beske-Diehl S, Mothershill JS (2004) The timing of regional lateglacial events and post-glacial sedimentation rates from Lake Superior. Quat Sci Rev 23:2355–2367 Burwasser GJ (1977) Quaternary geology of the city of Thunder Bay and vicinity, District of Thunder Bay. Ontario Geol Surv Report GR-164, p 70 Clayton L (1983) Chronology of Lake Agassiz drainage into Lake Superior. In: Teller JT, Clayton L (eds) Glacial Lake Agassiz. Geol Assoc of Canada, Spec Paper 26, pp 291–307 Clayton L, Moran SR (1983) Chronology of Late Wisconsinan glaciation in middle North America. Quat Sci Rev 1:55–82 Coleman AP (1909) Lake Ojibway; last of the great glacial lakes. Ontario Bureau of Mines, 18th Annual Report, Part. 1, p 284 Cowan WR (1978) Radiocarbon dating of Nipissing Great Lakes events near Sault Ste. Marie, Ontario. Can J Earth Sci 15:2026–2030 Dredge LA, Cowan WR (1989) Quaternary geology of the southwestern Canadian Shield. In: Fulton RJ (ed) Chapter 3 of Quaternary geology of Canada and Greenland. Geol Surv of Canada, Geology of Canada, no.1 pp 189–318 Drexler CW, Farrand WR, Hughes JD (1983) Correlation of glacial lakes in the Superior Basin with eastward discharge events from Lake Agassiz. In: Teller JT, Clayton L (eds) Glacial Lake Agassiz. Geol Assoc of Canada, Spec Pap 26, pp 261–290 Farrand WR, Drexler CW (1985) Late Wisconsinan and Holocene history of the Lake Superior basin. In: Karrow PF, Calkin PE (eds) Quaternary evolution of the Great Lakes. Geol Assoc of Canada, Spec Pap 30, pp 17–32 Fenton MM, Moran SR, Teller JT, Clayton L (1983) Quaternary stratigraphy and history in the southern part of the Lake Agassiz basin. In: Teller JT, Clayton L (eds) Glacial Lake Agassiz. Geol Assoc of Canada, Spec Pap 26, pp 49–74
J Paleolimnol (2007) 37:331–348 Fetter CW (2001) Applied hydrology. Prentice-Hall Inc., New Jersey, USA Gartner JF, McQuay DF (1980) Obakamiga Lake area, National Topographic Series 42F/SW, Districts of Algoma and Thunder Bay. Ontario Geol Surv, Northern Ontario Engineering Geology Terrain Study 45, p 16 Geddes RS, Kristjansson FJ, Teller JT (1987) Quaternary features and scenery along the north shore of Lake Superior. Nat Res Council of Canada, XIIth INQUA Congress Field Excursion Guide C-12, p 62 Gustavson TC, Ashley GM, Boothroyd JC (1975) Depositional sequences in glaciolacustrine deltas. In: Jopling AV, McDonald BC (eds) Glaciofluvial and glaciolacustrine sedimentation. SEPM, Spec Pub 23, pp 264–280 Hughes JD (1978) A post-Two Creeks buried forest in Michigan’s northern peninsula. Proceedings, 24th annual meeting, Inst on Lake Superior Geology, p 16 Hughes OL (1965) Surficial geology of part of the Cochrane District, Ontario, Canada. In: Wright HE, Frey DG (eds) International studies on the Quaternary. Geol Soc of Am, Spec Paper 84, pp 535–565 Jopling AV, Walker RG (1968) Morphology and origin of ripple drift cross-lamination, with examples from the Pleistocene of Massachusetts. J Sed Petrol 38:69–84 Karrow PF (1989) Quaternary geology of the Great Lakes subregion. In: Fulton RJ (ed) Chapter 4 of Quaternary geology of Canada and Greenland. Geol Surv of Canada, Geology of Canada no. 1; also Geol Soc of Am, The Geology of North America, v. K-1, pp 321– 440 Kehew AE, Clayton L (1983) Late Wisconsinan floods and development of the Souris-Pembina spillway system in Saskatchewan, North Dakota, and Manitoba. In: Teller JT, Clayton L (eds) Glacial Lake Agassiz. Geol Assoc of Canada, Spec Pap 26, pp 261–290 Lawson AC (1893) Sketch of the coastal topography of the north side of Lake Superior with special reference to the abandoned strands of Lake Warren. Geol and Natural History Surv of Minnesota, 20th Ann Report 1891, pp 181–289 Lewis CFM, Anderson TW (1989) Oscillations of levels and cool phases of the Laurentian Great Lakes caused by inflows from glacial Lakes Agassiz and BarlowOjibway. J Paleolimnol 2:99–146 Leverett F (1929) Moraines and shorelines of the Lake Superior basin. US Geol Surv Professional Paper 154-A, p 72 Leverington D, Teller JT (2003) Paleotopographic reconstructions of the eastern outlets of glacial Lake Agassiz. Can J Earth Sci 40:1259–1278 Lowe DR (1982) Sediment gravity flows II: depositional models with special reference to the deposits of highdensity turbidity currents. J Sed Petrol 5:279–297 Lowell TV, Larson GJ, Hughes JD, Denton GH (1999) Age verification of the Lake Gribben forest bed and the Younger Dryas advance of the Laurentide Ice Sheet. Can J Earth Sci 36:383–393
347 Matsch CL (1983) River Warren, the southern outlet of Glacial Lake Agassiz. In: Teller JT, Clayton L (eds) Glacial Lake Agassiz. Geol Assoc of Canada, Spec Pap 26, pp 231–244 McKee ED, Weir GM (1953) Terminology for stratification and cross-stratification in sedimentary rocks. Geol Soc Am Bull 64:381–389 Miall AD (1978) Lithofacies types and vertical profile models in braided river deposits: a summary. In: Miall AD (ed) Fluvial sedimentology. Can Soc of Petroleum Geol Memoir 5, pp 597–604 Miall AD (1985) Architectural-element analysis: a new method of facies analysis applied to fluvial deposits. Earth Sci Rev 22:261–308 Miall AD (1994) Reconstructing fluvial macroform architecture from two-dimensional outcrops: examples from the Castelgate Sandstone, Book Cliffs, Utah. J Sed Res 64:146–158 Miall AD (1996) The geology of fluvial deposits. SpringerVerlag, Italy Prest VK (1970) Quaternary geology of Canada. In: Douglas RJW (ed) Geology and economic minerals of Canada, 5th edn. Geol Surv of Canada, Economic Geology Report 1, pp 676–764 Prest VK, Grant DR, Rampton VN (1968) Glacial map of Canada. Geological Survey of Canada, Map 1253A, scale 1:5 000 000 Saarnisto M (1974) The deglaciation history of the Lake Superior region and its climatic implications. Quat Res 12:316–339 Saarnisto M (1975) Stratigraphical studies on the shoreline displacement of Lake Superior. Can J Earth Sci 12:300–319 Shaw J (1975) Sedimentary successions in Pleistocene icemarginal lakes. In: Jopling AV, McDonald BC (eds) Glaciofluvial and glaciolacustrine sedimentation. Soc of Econ Paleontologists and Mineralogists, Spec Pub 23, pp 281–303 Slattery SR (2003a) Quaternary geology of the Castlebar Lake area, Districts of Thunder Bay and Cochrane. Ontario Geol Surv, Map P.3523, Geological Series – Preliminary Map, scale 1:50 000 Slattery SR (2003b) Quaternary geology of the Pagwachuan Lake area, District of Thunder Bay. Ontario Geol Surv, Map P.3524, Geological Series – Preliminary Map, scale 1:50 000 Taylor FB (1895) The Nipissing beach on the north Superior shore. Am Geol 15:304–314 Taylor FB (1897) Notes on the abandoned beaches of the north coast of Lake Superior. Am Geol 20:111– 128 Teller JT (1985) Glacial Lake Agassiz and its influence on the Great Lakes. In: Karrow PF, Calkin PE (eds) Quaternary evolution of the Great Lakes. Geol Assoc of Canada Spec Pap 30, pp 1–16 Teller JT, Mahnic P (1988) History of sedimentation in the northwestern Lake Superior basin and its relation to Lake Agassiz overflow. Can J Earth Sci 25:1660– 1673
123
348 Teller JT, Thorleifson LH (1983) The Lake AgassizSuperior connection. In: Teller JT, Clayton L (eds) Glacial Lake Agassiz. Geol Assoc of Canada, Spec Pap 26, pp 261–290 Teller JT, Thorleifson LH, Matile G, Brisbin WC (1996) Sedimentology, geomorphology and history of the central Lake Agassiz Basin. Field Trip Guidebook B2, Geol Assoc of Canada/Minerol Assoc of Canada, p 84 Thorleifson LH, Kristjansson FJ (1993) Quaternary geology and drift prospecting, Beardmore-Geraldton area, Ontario. Geol Surv Can Memoir 436, p 146
123
J Paleolimnol (2007) 37:331–348 Vincent JS, Hardy L (1979) The evolution of glacial lakes Barlow and Ojibway, Quebec and Ontario. Geol Surv of Canada, Bulletin 316, p 18 Zoltai SC (1965) Glacial features of the Quetico-Nipigon Area, Ontario. Can J Earth Sci 2:247–269 Zoltai SC (1967) Glacial features of the north-central Lake Superior region Ontario. Can J Earth Sci 4:515–518 Zoltai SC, Herrington HB (1967) Late glacial molluscan fauna north of Lake Superior, Ontario. J Paleontol 40:439–446
