Quaternary stratigraphy and glacial history of the Peace ...qgrc/papers/Hartman and Clague CJES 2008.pdf · jointing, weathering, sliding planes, rupture surfaces, and the angle of

Quaternary stratigraphy and glacial history of thePeace River valley, northeast British Columbia1

Gregory M.D. Hartman and John J. Clague

Abstract: Two Cordilleran and three Laurentide glacial advances are recorded in Quaternary sediments and landforms inthe Peace River valley, northeast British Columbia. The advances are inferred from fluvial gravels, glaciolacustrine sedi-ments, and tills within nested paleovalleys excavated during three interglaciations and from the distribution of granitoidclasts derived from the Canadian Shield. Till of the last (Late Wisconsinan) Laurentide glaciation occurs at the surface, ex-cept where it is overlain by postglacial sediments. The advance that deposited this till was the most extensive in the studyarea, and the only advance definitively recognized in western Alberta south of the study area. Late Wisconsinan Cordil-leran till has not been found in the study area, but Cordilleran and Laurentide ice may have coalesced briefly during thelast glaciation. Support for this supposition is provided by the inferred deflection of Laurentide flutings to the southeast byCordilleran ice. The earliest Laurentide advance may have been the least extensive of the three Laurentide events recog-nized in the study area. Erratics attributed to this advance occur only east of the Halfway River – Beatton River drainagedivide.

Resume : Deux avancees glaciaires de la Cordillere et trois de l’Inlandsis laurentidien sont evidentes dans les sedimentset la topographie du Quaternaire de la vallee de la riviere de la Paix, dans le nord-est de la Colombie-Britannique. Lesavancees sont deduites a partir des graviers fluviatiles, des sediments glacio-lacustres et des tills dans des paleovalleesemboıtees excavees durant trois periodes interglaciaires et a partir de la distribution des clastes granitoıdes provenant duBouclier canadien. Un till de la derniere glaciation laurentidienne (Wisconsinien tardif) se retrouve a la surface, sauf laou il est recouvert de sediments post-glaciaires. L’avancee qui a depose ce till a ete la plus extensive dans le secteur al’etude et c’est la seule avancee definitivement reconnue dans l’ouest de l’Alberta, au sud du secteur a l’etude. Aucun tillde la Cordillere datant du Wisconsinien n’a ete trouve dans le secteur a l’etude. Cependant, la glace de la Cordillere etcelle de l’Inlandsis laurentidien pourraient s’etre unis durant une breve periode au cours de la derniere glaciation. Cettehypothese est soutenue par la deflexion inferee de rainures glaciaires laurentidiennes vers le sud-est par la glace de laCordillere. L’avancee glaciaire laurentidienne la plus precoce pourrait etre la moins etendue des trois evenements laurenti-diens reconnus dans le secteur a l’etude. Les blocs erratiques attribues a cette avancee ne se retrouvent qu’a l’est de laligne de partage des eaux entre les rivieres Halfway et Beatton.

[Traduit par la Redaction]

Introduction

The Quaternary stratigraphy of Peace Valley was docu-mented 30 years ago by Mathews (1978) but, since then,some researchers working in northeast British Columbia andnorthwest Alberta have come to conclusions incompatiblewith Mathews’ interpretation of the history of Cordilleran

and Laurentide glaciations of the area. Moreover, the Quater-nary history along the entire Canadian Rocky Mountainfront has, in recent years, been revised. Early researchersinferred up to four Laurentide glaciations of the AlbertaPlateau, but recent work has reduced this number, andsome workers have suggested that only one advancereached Edmonton, Simonette, and Watino (Fig. 1). Thenumber of Cordilleran advances has also been reducedfrom that proposed by early workers.

These revisions have been driven, in part, by improve-ments in describing and interpreting sediments, especiallydiamictons (Eyles et al. 1983). Equally important, Mac-Donald et al. (1987) demonstrated that radiocarbon ages ongyttja and some freshwater aquatic plants suffer an ‘‘oldcarbon effect.’’ These and other developments led to the re-evaluation of interpretations based on incorrectly describeddeposits or spurious radiocarbon ages. The work of early re-searchers however, cannot be completely discounted.Henderson (1959), Tokarsky (1967), St-Onge (1972), andMathews (1978) present evidence for multiple glaciationsthat is not dependent on lithofacies interpretations and radio-carbon ages.

Even recent researchers, however, do not agree on the ex-tent and times of Cordilleran advances and whether or not

Received 30 March 2007. Accepted 30 October 2007. Publishedon the NRC Research Press Web site at cjes.nrc.ca on 25 June2008.

Paper handled by Associate Editor R. Gilbert.

G.M. Hartman.2 Department of Earth Sciences, Simon FraserUniversity, 8888 University Drive, Burnaby, BC V5A 1S6,Canada; Present address: AMEC Earth and Environmental, 568170th Street, Edmonton, AB T6B 3P6, Canada.J.J. Clague. Department of Earth Sciences, Simon FraserUniversity, 8888 University Drive, Burnaby, BC V5A 1S6,Canada.

1This article is one of a selection of papers published in thisSpecial Issue on the theme Geology of northeastern BritishColumbia and northwestern Alberta: diamonds, shallow gas,gravel, and glaciers.

2Corresponding author (e-mail: [email protected]).

549

Can. J. Earth Sci. 45: 549–564 (2008) doi:10.1139/E07-069 # 2008 NRC Canada

Cordilleran glaciers ever coalesced with the Laurentide icesheet. Jackson et al. (1997), Bednarski (1999, 2001), Dykeet al. (2002), and Bednarski and Smith (2007) have arguedfor extensive Late Wisconsinan Cordilleran glaciation, in-volving coalescence of Cordilleran and Laurentide ice overa considerable distance of the Rocky Mountain front.Bobrowsky (1989), Bobrowsky et al. (1991, 1992), Bobrow-sky and Rutter (1992), and Catto et al. (1996), on the otherhand, concluded that the Late Wisconsinan Cordilleran ad-vance was restricted, and they attributed the most extensiveCordilleran advance to an earlier glaciation.

The Peace River valley in northeast British Columbia isideally located for a study of interactions of Cordilleran and

Laurentide glaciers. It was affected by both the Cordilleranand Laurentide ice sheets, and Quaternary sediments arewell exposed along Peace River and its tributaries. In addi-tion, interstadial and interglacial paleovalleys have been par-tially preserved, allowing an assessment of the stratigraphicrelations between valley-fill sediments of different age.

Study area

This study was conducted in the Charlie Lake map area(National Topographic System (NTS) 94A), which is locatedwithin the Alberta Plateau subprovince of the northernInterior Plains (Fig. 1). The area is a flat to gently rolling

Fig. 1. Study area. The map at the upper left shows adjacent areas where Quaternary research has been completed.

550 Can. J. Earth Sci. Vol. 45, 2008

# 2008 NRC Canada

plateau deeply incised by the valleys of the Peace River andits tributaries.

Mathews (1978) identified three distinct topographicprovinces in the study area (Fig. 2): (1) uplands with steepto gently rolling ridges and cuestas; (2) river valleys deeplyincised into the uplands; and (3) benches with little reliefadjacent to incised valleys. Uplands are areas of low tomoderate relief where near-surface bedrock controls thetopography. River valleys are incised into the upland topo-graphic province. They range from broad, flat-floored fea-tures to V-shaped and canyon-like forms. Benches adjacentto incised river valleys are nearly flat and are thus morpho-logically distinct from the uplands. Depths to bedrock be-neath the benches range from about 15 to 180 m.

Gently northeast-dipping Cretaceous sedimentary rocksunderlie the study area (Stott 1982). The rocks includemarine shale and siltstone of the Smokey Group, sandstoneand conglomerate of the Dunvegan Formation, and marineshale and minor sandstone of the Fort St. John Group. TheClear Hills and Chinchaga Hills are gravel-capped surfacesthat lie 300–600 m above the level of the surroundingInterior Plain and dip gently to the northeast (Fig. 1). Math-ews (1978) inferred that the hills are erosional remnants of apeneplain that once extended across the study area.

MethodsExposures of Quaternary sediments were visited and de-

scribed during the summers of 2002 and 2003 (Fig. 3). Thistask entailed locating intact exposures, defining and describ-ing sedimentary units, measuring elevations of unit contacts,and collecting samples.

Sediments were described and grouped using the litho-facies analysis method (Eyles et al. 1983). Characteristicsand features that were noted include grain size, sorting, sedi-mentary structures, colour, clast lithology, compaction,jointing, weathering, sliding planes, rupture surfaces, andthe angle of the exposure face. Basal and upper contacts ofunits were also described, and fossils, erratics, striated sur-faces, fabric, and deformation features, if present, werenoted.

The colours of moist sediment samples were determinedusing the Geological Society of America Rock Color Chart(Goddard et al. 1963). Pebble fabrics were determined bymeasuring the trends and plunges of 50 (Woodcock 1977)rod-shaped clasts (a:b ‡ 2) from a small area of exposure.Elevations of unit contacts were determined using a baro-metric altimeter and a laser range finder (Hartman 2005).Pink orthoclase-rich granite and granitoid gneiss pebbles,which are exotic lithologies in western Canada, were col-lected for examination in the laboratory.

Additional stratigraphic information was acquired fromwater well drill logs, and geotechnical drill logs. The drilllogs were obtained from a database maintained by the BCMinistry of Sustainable Resource Management. Geotech-nical drilling and water well logs also provided data on theelevation of the bedrock surface in drift-covered areas.

Stratigraphic data were entered in a geographic informa-tion system (GIS) to examine lithostratigraphic units in threedimensions. The GIS was also used to interpolate a bedrocksurface and discern the form of buried paleovalleys. The sur-

face was interpolated over known and estimated bedrocksurface points using inverse distance weighting.

Paleovalleys, planation surfaces, and deepbedrock basins

Three relict fluvial drainage systems are graded to differ-ent base levels, providing a framework for understanding theQuaternary stratigraphy and history of the study area. Post-glacial Peace River has incised almost 50 m deeper intobedrock than any of its precursors; the river today flowsover bedrock through most of the study area. Two incisedbedrock straths at different elevations, and thus differentages, are exposed in the valley walls of Peace River and itstributaries (Fig. 4). The straths and overlying fluvial sedi-ments have been identified in water well drill logs awayfrom valley wall exposures. A broad gravel-covered fluvialplanation surface, which is not incised, occurs locally onthe uplands at a higher elevation than the higher of the twoincised straths.

The two incised straths are the floors of paleovalleys thatare buried beneath up to 180 m of Pleistocene sediments. AsMathews (1978) suggested, the paleovalleys define drainagesystems that pre-date the glaciations responsible for theirburial. The straths are best delineated where they are nearand intersect the modern Peace River. Limited exposure alongtributaries of Peace River precludes detailed assessment ofthe tributary paleovalleys beyond their lower reaches.

The lower incised strath is locally well exposed about50 m above Peace River and has been encountered in waterwells elsewhere. It slopes 0.0006 down to the east, a gra-dient similar to that of modern Peace River. The higher in-

Fig. 2. Topographic provinces of the study area (modified fromMathews 1978; reproduced and adapted with the permission of theMinister of Public Works and Government Services Canada, 2007and courtesy of Natural Resources Canada, Geological Survey ofCanada).

Hartman and Clague 551

# 2008 NRC Canada

cised strath has the same gradient, is about 225 m abovePeace River, and is exposed on the north side of the valleybetween the confluence of Peace and Moberly rivers and theBritish Columbia – Alberta boundary. It has also been en-countered in water wells. The high broad gravel-coveredplanation surface is 330 m above Peace River. Its extent ispoorly known because it is exposed at only one locationand is intercepted by only a few water wells, but it appearsto slope 0.0005 to the east.

The lower paleovalley system is dominated by a large val-ley that trends eastward across the southern part of the studyarea near the modern Peace River valley. Its tributaries areclose to present-day Beatton, Pine, Moberly, Halfway, andDoig rivers. The lower paleovalley system diverges from themodern drainage along the canyons of Kiskatinaw, Beatton,Doig, and Halfway rivers and along two reaches of PeaceRiver—one near its confluence with Moberly River, and theother at the east edge of the study area. The lower paleoval-leys are two to four times wider than their modern counter-parts and have flat floors and fairly steep walls.

The higher paleovalley system is more difficult to delineatethan the lower one. It was only definitively identified on thenorth side of Peace River between Moberly River and theBritish Columbia – Alberta boundary. Although only a por-

tion of the upper paleovalley is preserved, what remains indi-cates that it was almost four times as wide as the modernPeace River valley and wider than the lower paleovalley. Thelower paleovalleys of Peace and Beatton rivers are incisedinto the higher paleovalley, as are the modern valleys ofPeace, Beatton, and Alces rivers and many smaller tributaries.

Deep bedrock basins occur below the Portage Mountainend moraine and near Lynx Creek (Fig. 5; Rutter 1977; Math-ews 1978; Hartman 2005). The basins are close to the RockyMountain front, where a former lobe of the Cordilleran icesheet flowed onto the Alberta Plateau. The location and depthof the basins suggest that they were carved by a Cordilleranoutlet glacier. This interpretation is supported by the presenceof diamicton near the base of a drill hole through the PortageMountain end moraine (Rutter 1977). The basins are local-ized and do not appear to be structurally controlled.

Quaternary stratigraphy

Fluvial deposits of the high planation surfaceFluvial gravel overlies a bedrock strath that stands slightly

higher than the adjacent plateau (Fig. 6; Table 1). Erosionhas removed most of this unit and lowered the surface ofthe uplands around its remnants.

Fig. 3. Locations of measured sections.


# 2008 NRC Canada

The unit is, on average, 6 m thick; its upper surface isabout 330 m above present-day Peace River. It is composedmainly of well-sorted, cross-bedded, pebble–cobble gravelwith minor lenses of silt and sand. Cross-beds at section M-3 (Imperial gravel pit) record northward flow. All clasts areof Cordilleran or local provenance.

Fluvial deposits of the upper paleovalleyA younger fluvial gravel unit overlies the higher of two in-

cised bedrock straths, within the upper paleovalley of PeaceRiver (Fig. 6; Table 1). The strath is located on the north sideof Peace River valley between the confluence of Peace andMoberly rivers and the British Columbia – Alberta boundary.

Fig. 4. Limits of buried paleovalleys. a.s.l., above sea level.

Fig. 5. Interpreted geology between drill-log M-1 near the Portage Mountain end moraine and drill-log M-2 at Lynx Creek. The bedrocksurface in this area is below the level of past fluvial erosion. a.s.l., above sea level.


# 2008 NRC Canada

The unit has an average thickness of about 8 m, and itstop is about 225 m above present-day Peace River. Thedominance of lithofacies interpreted to be channel depositssuggests that the sediments were deposited by a braidedriver (Miall 1977). The gravel consists mainly of Cordilleranlithologies, notably quartzite, chert, sandstone, granitic andvolcanic rocks, and carbonates. Eastern exposures nearAlces River also contain rare granite gneiss clasts derivedfrom the Canadian Shield. No gneiss clasts were found inwestern exposures of the unit near Old Fort.

Advance-phase glaciolacustrine deposits of the upperpaleovalley

Up to 66 m of glaciolacustrine silt and sand conformablyoverlie fluvial gravel of the upper paleovalley (Fig. 6;

Table 1). They occur up to 613 m above sea level (a.s.l.)within the upper paleovalley.

This lithostratigraphic unit is dominated by lithofaciesindicative of deposition from suspension and underflows ina lake, including laminated mud and sand, massive sand,rippled sand, and trough cross-bedded sand. Dropstones anddiamicton pods, some containing Canadian Shield clasts,increase in abundance toward the top of the unit. The tran-sition upwards from ice-distal to ice-proximal lithofaciesindicates the approach of Laurentide ice.

Pre-Late Wisconsinan Laurentide till?Mathews (1978) hypothesized that tills of two Laurentide

glaciations are present in exposures on the north side ofPeace River valley near Old Fort (Fig. 6). He assigned the

Fig. 6. Representative stratigraphic sections from the four paleo-drainage systems. a.s.l., above sea level.


# 2008 NRC Canada

lower of the two tills to a pre-Late Wisconsinan glaciation.An attempt was made to test this hypothesis by measuringthe clast fabric of the lower till. A clast fabric from thebase of section SS-26 has a polymodal to girdle distributionwith a dominant mode oriented 1208, perpendicular to theregional ice flow direction during the last glaciation(Mathews 1978). Striations on a boulder pavement at thebase of this unit trend 1128, similar to the direction of thedominant fabric mode. However, at section SS-24, a clastfabric from the base of a similar diamicton has a spread un-imodal distribution with a dominant mode of 2188, parallelto surface ice-flow indicators.

Two Laurentide tills directly overlie one another at sec-tion SS-50 near Old Fort. They can be distinguished by theirclast content, colour, and the presence of a recessive band atthe sharp undulating contact between the two units. Thelower till has a strong bimodal fabric with clusters oriented2488 and 688. A large anvil-shaped boulder at the base ofthe lower till appears to have been lodged and has a flatupper surface with two sets of striations, an older set ori-ented 2408 and a younger set oriented 2208. The fabric andstriations are roughly parallel to the surface ice-flow indica-tors. The upper till corresponds stratigraphically with theLate Wisconsinan Laurentide till.

Western-provenance tillUp to 60 m of till of western provenance is exposed in

lower Halfway River valley (Fig. 6; Table 1). The high clastcontent (about 30%), silty sand matrix, and greyish-orange(10YR 7/4) colour distinguish the unit from Laurentide till(clast content typically 10%, clay-rich matrix, and olivegrey (5Y 3/2) colour). Clasts are almost entirely of Cor-dilleran or local provenance; only two possible Laurentidepebbles were found at section SS-75. A clast fabric takennear the middle of the unit at section SS-75 has a strong un-imodal distribution with a primary eigenvector of 2218, par-allel to the orientation of a nearby gap in the mountainsthrough which Peace River once flowed (Beach and Spivak1943). The western-provenance till is overlain across anunconformable contact by Late Wisconsinan outwash, glacio-lacustrine sediments, and till, and thus is older than the lastglaciation.

Fluvial deposits of the lower paleovalleysFluvial gravel lies on the floors of the lower paleovalleys

(Fig. 6; Table 1). It is well exposed in the walls of the mod-ern Peace River valley and its larger tributaries (Fig. 7).

The unit is 13 m thick on average; its top is about 50 mabove modern river level. It is interpreted to be channeldeposits of a braided river (Miall 1977). Gravel transportedby the paleo-river is about the same size as that carried bythe modern river. Eastward flow is indicated by paleoflowmeasurements. The gravel is dominated by Cordilleran andlocal lithologies, but it contains rare clasts derived from theCanadian Shield in exposures between Lynx Creek and theeastern boundary of the study area.

Advance-phase glaciolacustrine deposits of the lowerpaleovalleys

Up to 100 m of glaciolacustrine sediments conformablyoverlie fluvial gravel of the lower paleovalleys (Figs. 6, 7,

8; Table 1). They record damming of the paleo-Peace valleyand its tributaries early during the last glaciation. The unitis dominated by laminated mud and horizontal and drape-laminated sands deposited from suspension and underflowsin ice-distal, deep-water environments (Gustavson et al.1975; May 1977; Ashley 1988; Wolfe and Teller 1995).

Coarser sediments are present in the glaciolacustrine unitnear paleovalley walls. They include clast-supported mass-ive diamicton deposited by subaqueous debris flows (May1977) and cross-bedded, horizontally bedded, and rippledsands deposited by energetic density underflows (Wolfe andTeller 1995). Similar sediments have been interpreted to bebasin-marginal glaciolacustrine deposits (Ashley 1988;Wolfe and Teller 1995).

The dominant, laminated mud facies generally contains nodropstones. However, at sections where the unit has notbeen extensively eroded by Laurentide ice, dropstones anddiamictic pods increase in abundance upward through theunit.

Advance-phase glaciolacustrine sediments may extendinto the Rocky Mountain Foothills between Portage andBullhead mountains. The log of the drill hole through thePortage Mountain end moraine records silt, silty sand, andclayey sand overlying interbedded gravel, sand, and silt(Campbell 1959). Based on its thickness and elevationrange, the gravel is assigned to fluvial deposits of the lowerpaleovalley. The overlying finer sediments likely recorddeposition in a lake dammed by advancing Laurentide ice.

The highest known elevation of the lower paleovalleyglaciolacustrine unit is 632 m a.s.l. The lake thus extendedto at least this elevation at its maximum stage.

Till of the last glaciationLaurentide till is a widespread surface or near-surface unit

in the study area. It is a massive matrix-supported diamictonwith <15% clasts and a silt and clay matrix. Many of theclasts are granite and granitoid gneiss derived from theCanadian Shield.

Drumlins and flutings mapped by Mathews (1978) suggestthat the Laurentide ice sheet flowed to the south-southwestduring the Late Wisconsinan. The flow indicators shift tothe south-southeast in the southeast part of the study area,indicating an interaction between Cordilleran and Lauren-tide ice. Previously mentioned pebble fabric orientationsand striated pavements measured at the base of the till attwo sections (SS-24 and SS-50) match the direction ofregional ice flow. However, at sections SS-26 and SS-91,the fabrics are parallel to the local paleovalley axis andtransverse to the regional flow direction. These data maybe interpreted in two ways—either till from two separateglaciations are superposed at sections SS-26 and SS-91(Mathews 1978), or initial Laurentide ice flow into theseareas was topographically controlled (Catto et al. 1996).

The ice sheet that deposited the till extended west beyondthe study area into the Rocky Mountain Foothills, but noterminal moraine related to this advance has been identified.Canadian Shield erratics occur above the maximum limit ofGlacial Lake Peace in east-draining valleys of the foothills(Bobrowsky et al. 1991; Catto et al. 1996).


# 2008 NRC Canada

Late-glacial glaciolacustrine depositsGlacial Lake Peace was impounded behind retreating

Laurentide ice at the end of the last glaciation (Fig. 6; Math-ews 1978, 1980). Sediments deposited in this lake consistmainly of laminated mud with <5% dropstones. Severalstages of Glacial Lake Peace have been identified from re-lict shorelines, deltas, and spillways (Rutter 1977, Mathews1980, Woolf 1993).

Postglacial depositsFluvial gravel underlies terraces cut during postglacial

incision of the modern Peace River valley (Fig. 6). Themost prominent terrace is 45 m above present-day PeaceRiver; its capping gravel is about 17 m thick. Postglacialgravel is similar to older gravel units, except for its muchgreater proportion of eastern-provenance clasts.

Landslide deposits are ubiquitous in Peace River valleyand its major tributaries. Most of the deposits are massivematrix-supported diamicton that is less compact than typicalLaurentide till and generally has a lower clast content.Recognition of postglacial landslide deposits, however, isbased less on lithofacies than on geomorphology. Landslidedeposits typically have hummocky surfaces with gentle tomoderately inclined slopes. The surfaces of active landslidesare marked by tension cracks, tilted trees, and sag ponds,and are located down-slope from scalloped head scarps.

Geologic historyFour drainage systems of different age, including the

modern drainage system, are recognized in the study area(Fig. 9). A fifth, older system is recorded by fluvial graveloverlying an eastward-sloping peneplain east of the studyarea (Mathews 1978; Liverman 1989). East-draining rivers

in the study area established themselves at successivelylower elevations during middle and late Cenozoic time fol-lowing the Laramide Orogeny. Mathews (1989) noted localdissected gravel-capped peneplains of early Oligocene tolate Tertiary age in the Rocky Mountain Foothills and west-ern Interior Plains from Montana north to the RichardsonMountains in Yukon Territory. Remnants of lower drainagesystems are probably Pleistocene in age. Successive drain-age systems cross-cut and nest within one another, produc-ing a palimpsest of valley in-fills that provide a frameworkon which a relative chronology of geologic events can beplaced.

Incision of the drainage systems is probably due to acombination of epeirogenic uplift (Pazzaglia et al. 2001;McMillan and Heller 2002), increased erosion due to cli-matic forcing in the late Cenozoic (Dethier 2001), andbase-level lowering resulting from drainage integration andglaciation (Pederson and Pazzaglia 2002).

Pre-Laurentide river systemFluvial gravel of the high planation surface is the oldest

post-Cretaceous lithostratigraphic unit in the study area. Itlacks eastern-provenance clasts and thus was deposited priorto the first Laurentide glaciation of the study area. Thegravel records a period of deposition following planation ofbedrock surfaces by an east- to north-flowing river system.Its wide distribution on low-relief upland bedrock surfacessuggests deposition on a broad peneplain.

The gravel is younger than higher gravels of middle tolate Tertiary that cap peneplains at the west edge of theInterior Plains, including the Clear Hills, Saddle Hills, andChinchaga Hills (Mathews 1978; Liverman 1989; Stott andAitken 1993). It is tentatively assigned a late Tertiary or


# 2008 NRC Canada

Table 1. Characteristics of lithostratigraphic units recognized in the study area.

Lithostratigraphic unitAvg. elev. abovePeace River (m)

Thickness(m) Areal extent Dominant lithofacies

Postglacial deposits Variable 17 (mean) valley slopes andfloors

gravel, matrix-supporteddiamicton, massive silt

Late-glacial glaciolacustrinedeposits

Variable 10 (mean) various stages ofGlacial Lake Peace

laminated mud

Till of the last glaciation Variable 12 (mean) Regional Matrix-supported diamicton

Advance-phase glaciolacustrinedeposits of the lower paleovalley

Variable 18 (mean)100 (max)

Lower paleovalley Laminated mud, massive,rippled, and trough cross-bedded silt and sand

Fluvial deposits of the lowerpaleovalley

50 13 (mean),67 (max)

Lower paleovalley Horizontally bedded gravel

Western-provenance till Variable 40 (mean),60 (max)

Lower Halfwayvalley

Matrix-supported diamicton

Pre-middle Wisconsinan eastern-provenance till?

Variable 32 m (onelocalityonly)

Local Matrix-supported diamicton

Advance-phase glaciolacustrinedeposits of the upper paleovalley

Variable 33 (mean),66 (max)

Upper paleoalley Laminated mud, massive,rippled, and trough cross-bedded silt and sand

Fluvial deposits of the upperpaleovalley

225 8 (mean) Upper paleovalley Horizontally bedded gravel

Fluvial deposits of the high plana-tion surface

330 6 (mean) Regional Horizontally bedded gravel

early Pleistocene age because it pre-dates the oldest Lauren-tide glaciation of the study area and because it lies on highlydissected, fluvially eroded bedrock. Drainage subsequentlybecame incised into this broad planation surface.

Early nonglacial river systemThe next younger deposit in the study area is fluvial

gravel of the upper paleovalley. It contains the same litholo-gies as the older, pre-glacial gravel unit, with one importantexception. Rare clasts of granite and granitoid gneiss de-rived from the Canadian Shield occur in exposures nearAlces River and Golata Creek and at section SS-109, butnot farther west. The Canadian Shield clasts must havebeen transported westward up the regional slope of theInterior Plains during an early Laurentide glaciation. Math-ews (1978) concluded that the clasts were transported as farwest as Golata Creek by Laurentide ice. Alternatively, theycould have been rafted on icebergs in a lake dammed by theLaurentide ice sheet somewhere to the east, or they couldhave been transported into paleo-Peace River valley by anorthern tributary or by meltwater diverted along theLaurentide ice margin. At the very least, the presence ofthese clasts indicates that an early Laurentide advanceapproached the study area.

This early glaciation may be younger than the MatuyamaReversed Polarity Chron (0.78 Ma). Barendregt and Irving(1998) found that all glacial sediments on the central Interi-or Plains are normally magnetized and concluded that theLaurentide ice sheet first reached the western Interior Plainsafter 0.78 Ma ago.

Glaciation may have been partly responsible for entrench-ment of the high bedrock planation surface. Pederson andPazzaglia (2002) concluded that glacial repositioning and

integration of drainage were important causes of base-levellowering.

Penultimate glaciation

Laurentide advanceAdvance-phase glaciolacustrine sediments of the upper

paleovalley conformably overlie nonglacial fluvial gravel.They were deposited when Laurentide ice advanced up theregional slope of the western Interior Plains and blockedeast-flowing paleo-Peace River. The glaciolacustrine sedi-ments occur in outcrop west to Old Fort, but the lakereached to at least 613 m a.s.l. and thus extended fartherwest.

Advance-phase glaciolacustrine sediments have not beendated, but they are older than the lower paleovalley systeminto which they are cut. The lower paleovalley is no youngerthan Middle Wisconsinan (see later in the text), con-sequently the glaciolacustrine sediments filling the upperpaleovalleys, as well as the Laurentide ice sheet that im-pounded the lake in which these sediments accumulated, areEarly Wisconsinan or older.

Older advance-phase glaciolacustrine sediments andCanadian Shield clasts within the lower paleovalley fluvialgravel provide definitive evidence for the penultimateLaurentide advance. Although uncommon, shield clastswere recovered from lower paleovalley gravel as far westas Lynx Creek, corroborating Mathew’s assertion that thepenultimate Laurentide glaciation transported eastern-provenance clasts farther west than an earlier Laurentideadvance.

The western limit of the penultimate Laurentide advanceis unknown. No definitive landforms or till of the pen-

Lithofaciesassociation

Lowercontact Bounding units (above / below) Shield clasts Event

Fluvial, colluvial Erosional None above / all units Yes, west to RockyMountain front

Holocene fluvial andcolluvial deposition

Glaciolacustrine Conformableover till

None above / till of the last glaciation Yes, west to RockyMountain front

Late Wisconsinan Lauren-tide glacial retreat

Glacial Erosional Late-glacial glaciolacustrine deposits ornone above / many different units below

Yes, west to RockyMountain front

Late WisconsinanLaurentide glaciation

Glaciolacustrine Conformable Till of the last glaciation / fluvial depositsof the lower paleovalley

Yes, west to PortageMountain endmoraine

Late Wisconsinan Lauren-tide glacial advance

Fluvial Erosional Advance-phase glaciolacustrine deposits ofthe lower paleovalley / bedrock

Yes, west to LynxCreek

Middle Wisconsinan riversystem

Glacial Erosional Outwash and advance-phase glaciolacus-trine deposits of the lower paleovalley /bedrock

No Pre-Middle WisconsinanCordilleran glaciation

Glacial Erosional Till of the last glaciation / advance-phaseglaciolacustrine deposits of the upperpaleovalley

Yes, west to OldFort?

Pre-Middle WisconsinanLaurentide glaciation

Glaciolacustrine Conformable Till of the last glaciation / fluvial depositsof the upper paleovalley

Yes, west to OldFort

Pre-Middle WisconsinanLaurentide glacialadvance

Fluvial Erosional Advance-phase glaciolacustrine deposits ofthe upper paleovalley / bedrock

Yes, west to GolataCreek

Early or middlePleistocene

Fluvial Erosional Till of the last glaciation / bedrock No Pliocene or earlyPleistocene river system


# 2008 NRC Canada

ultimate glaciation have been found, thus the possibility thatCanadian Shield clasts were transported into the study areaby an agent other than a glacier must be considered. Perhaps

the clasts were rafted on icebergs or carried by streams tothe positions they are found today.

The penultimate and last glaciations were separated by aninterglaciation during which the lower paleovalley systemdeveloped. Based on this evidence, the penultimate event isEarly Wisconsinan or older.

Cordilleran advance

A till in lower Halfway River valley is the only western-provenance glacial unit found in the study area. It may havebeen deposited by the glacier that eroded the bedrock basinsbeneath the Portage Mountain end moraine and Lynx Creek.The basins contain thick glaciolacustrine sediments that aretruncated by fluvial deposits of the lower paleovalley. Thesestratigraphic relations suggest that the early Cordilleranadvance is older than the lower paleovalley. Furthermore,Cordilleran till is not present in any of the exposures be-tween the Rocky Mountain front and Halfway River, makingit unlikely that the Cordilleran till at Halfway River is LateWisconsinan in age. This inference is in agreement withBobrowsky (1989), Bobrowsky and Rutter (1992), and Cattoet al. (1996), who assigned the early Cordilleran till to anEarly Wisconsinan or older glaciation.

The last interstadeThe lower paleovalley fluvial deposits record an east-

Fig. 7. Contact between fluvial deposits of the lower incised paleovalley and Glacial Lake Mathews sediments in an exposure along PeaceRiver between Halfway and Moberly rivers.

Fig. 8. Fluvial deposits of the lower paleovalley conformably over-lain by advance-phase glaciolacustrine sediment, section SS-13.


# 2008 NRC Canada

flowing river system. They are the oldest sediments in thestudy area that have been securely dated. Mathews (1978)recovered a mammoth tooth that yielded a radiocarbon ageof 27 400 ± 580 14C years BP (GSC-2034) from paleovalleygravel beneath postglacial terrace gravel at Taylor. A bisonvertebra from the same unit at section SS-29 gave an age of26 450 ± 310 14C years BP (Beta-188305). The two ages in-dicate that Peace River was aggrading its channel within thelower paleovalley late during the Middle Wisconsinan.

Ancestral Peace River flowed north of Portage Mountainbefore the last glaciation. Fluvial gravel encountered between527 and 547 m a.s.l. during drilling through the PortageMountain end moraine (Campbell 1959) is the same heightabove modern Peace River as gravel of the last interstade else-where in the study area and probably was deposited by ances-tral Peace River during the Middle Wisconsinan. Drainagenorth of Portage Mountain was blocked when Cordilleran icebuilt the end moraine there. During deglaciation, Peace Riverwas forced through a saddle south of Portage Mountain, whereit excavated Peace River Canyon (Beach and Spivak 1943).

Late Wisconsinan glaciation

Glacial Lake MathewsAdvance-phase glaciolacustrine sediments record dam-

ming of east-flowing rivers by advancing Laurentide ice.The lake in which the sediments accumulated is here termed‘‘Glacial Lake Mathews’’ in recognition of the many contri-butions of the late Dr. W.H. Mathews to our understandingof the Quaternary geology of western Canada. The sedi-ments have not been directly dated, but they are younger

than 26 450 ± 310 14C years BP, which is the age of a bisonvertebra recovered from underlying fluvial gravel.

The approach of the Laurentide ice sheet is indicated byan upward transition from ice-distal to ice-proximal glacio-lacustrine sediments in sections in the western part of thestudy area. Diamictic pods containing Canadian Shield clastsare restricted to the upper part of the glaciolacustrine se-quence.

The waters of Glacial Lake Mathews appear to have risenquickly, inundating paleo-Peace valley without significantinterruption. The transition from fluvial channel sedimenta-tion to lacustrine sedimentation spans <1 m in most sections(Fig. 8).

Fluvial gravel at some sections (SS-18, SS-19, SS-44, andpossibly SS-47) records a minor lake regression followed byrenewed glaciolacustrine sedimentation shortly after GlacialLake Mathews inundated the sites. The gravel occurs about3–4 m above the base of the glaciolacustrine sequence atthese sites, and it sharply and unconformably overliesbedded and laminated mud. Possible explanations for a lakeregression include (1) minor retreat of Laurentide ice, ex-posing lower lake outlets that previously had been blockedby ice, (2) erosion of the lake outlet or development of newand lower spillways when the lake overtopped drainagedivides across which channels were subsequently eroded byout-flow, and (3) draining of the lake during an outburstflood (jokulhlaup). No other lake-lowering event is recog-nized in Glacial Lake Mathews deposits.

To our surprise, we found no exposures of deltaic sedi-ments deposited by paleo-Peace River in Glacial Lake Math-ews. However, a buried delta may be present beneath the

Fig. 9. Cross-section A–A’ through Beatton and Peace River valleys showing the fluvial deposits of three relict drainage systems and over-lying valley fills.


# 2008 NRC Canada

end moraine in ancestral Peace River valley north of PortageMountain (Beach and Spivak 1943). Well sorted graveloccurs from 481 to 527 m a.s.l. in the borehole through theend moraine (Campbell 1959). We correlate this unit withthe lower paleovalley gravel. It is overlain by 66 m of inter-stratified silt, sand, and gravel, which are likely GlacialLake Mathews deposits.

Late Wisconsinan Laurentide glaciationBobrowsky et al. (1992) and Catto et al. (1996) found

Canadian Shield clasts up to about 950 m a.s.l. in the RockyMountain Foothills west of the study area and argued thatthe Laurentide ice sheet could not have ascended valleys inthe foothills if it had been opposed by east-flowing Cordil-leran ice. This observation is consistent with the conclusionsof Bobrowsky (1989) and Bobrowsky and Rutter (1992) thatthe Cordilleran advance was restricted and that its limit inPeace River valley was at Portage Mountain. Radiocarbonages from Finlay River valley, a major tributary of Peacevalley, bracket the time of Late Wisconsinan Cordilleranglaciation there to between 15 180 ± 100 14C years BP (TO-708) and 10 100 ± 90 14C years BP (GSC-2036) (Bobrowsky1989). A series of Middle Wisconsinan radiocarbon ages onsediments underlying till indicate that only one advanceoccurred in Finlay River valley during the Late Wiscon-sinan.

These data are inconsistent, however, with the interpreta-tion that flutings east of Fort St. John and adjacent to themountain front formed beneath coalescent Laurentide andCordilleran ice during the last glaciation (Mathews 1978). Ifthe flutings are indeed Late Wisconsinan, Cordilleran iceprobably reached east of Portage Mountain, albeit perhapsonly for a short time. Restricted Cordilleran glaciation inPeace valley also appears to be inconsistent with the evi-dence presented by Bednarski (1999, 2001) and Bednarskiand Smith (2007) for overtopping of high ridges in theRocky Mountains northwest of the study area by the Cordil-leran ice sheet (see below).

Glacial Lake Peace formed behind the retreating Lauren-tide ice sheet. Wood recovered from the contact betweenLate Wisconsinan Laurentide till and overlying Lake Peacesediments near Fort St. John yielded a radiocarbon age of13 970 ± 170 14C years BP (TO-2742) (Catto et al. 1996),which is a minimum for local Laurentide deglaciation.

Glacial Lake Peace drained in stages, perhaps with rapidlowering of the lake between stages. The lake was east ofFort St. John before 10 770 ± 120 14C years BP (SFU-454),which is the age of the lowest cultural layer at the CharlieLake cave archaeological site (Fladmark et al. 1988). Geert-sema and Jull (2002) and Jull and Geertsema (2006) report aseries of radiocarbon ages on charcoal layers in a debris-flow fan on a prominent river terrace at Bear Flat. The char-coal layers increase in age consistently with depth to about10 500 14C years BP, indicating that Glacial Lake Peace haddrained and Peace River had incised its modern valley to thelevel of the terrace before 10 500 14C years BP.

Postglacial Peace River valleyPeace River may have excavated its modern valley rapid-

ly, soon after deglaciation. Entrenchment did not commenceuntil Glacial Lake Peace drained from the study area, some-

time after 13 970 ± 170 14C years BP (TO-2742) (Catto et al.1996) and was well advanced or complete before 10 500 14Cyears BP (Geertsema and Jull 2002). The rock-walled, in-cised canyons of Beatton, Halfway, and Kiskatinaw rivershave a meandering planform. Meandering rivers typicallyare alluvial; lateral channel migration produces the meanders(Miall 1992). The walls of the Beatton, Halfway, and Kiska-tinaw canyons are bedrock, and the meandering planformcould not have developed after the rivers incised to theirpresent levels. Rather, it developed when the rivers flowedover the floor of Glacial Lake Peace soon after the lakedrained. Incision was so rapid that the rivers did not migratelaterally, preventing the canyons from being widened andpreserving the meandering form.

Comparison with previous workTwo models have been developed to explain the Quater-

nary stratigraphy and landforms of the westernmost AlbertaPlateau (Bobrowsky et al. 1991). One model is based on thework of Rutter (1976, 1977) and Mathews (1978, 1980). Itfeatures two or three Laurentide advances, three or four Cor-dilleran advances, and coalescence of the Cordilleran andLaurentide ice sheets during one glaciation. The secondmodel is based on Bobrowsky (1989) and includes oneLaurentide advance and two Cordilleran advances withoutcoalescence of the Laurentide and Cordilleran ice sheets.

The stratigraphy presented in this study (Fig. 10) supports,with some modification, the interpretation of Mathews(1978, 1980), specifically that there were three Laurentideand two Cordilleran advances in the region. This stratigra-phy is discussed in the following two sections in the contextof previous stratigraphic work.

Laurentide stratigraphySeveral researchers have argued that the Laurentide ice

sheet reached Edmonton, Watino, and areas to the southand west only once in the Pleistocene, during the Late Wis-consinan (Westgate et al. 1971, 1972; Catto 1984; Liverman1989; Liverman et al. 1989; Young et al. 1994). Thestratigraphy reported in this study requires three Laurentideglacial events separated by interglaciations or long inter-stades. Only the last of the three glacial events is Late Wis-consinan.

Laurentide glaciations were centred in northern Quebecand Labrador, Keewatin, and the Foxe Basin (Dyke et al.2002; Kleman et al. 2002; Dyke 2004). To reach the studyarea, ice had to flow southwest to south-southwest from theKeewatin ice dome. Flutings in the study area that matchthis flow direction were noted by Mathews (1978, 1980).Ice flowing into the study area from the northeast or north-northeast during pre-Late Wisconsinan glaciations may haveterminated short of Edmonton, 500 km to the southeast, andWatino, 160 km to the east-southeast.

Two, pre-Late Wisconsinan Laurentide advances are in-ferred in the study area from the presence of CanadianShield erratics in gravels of the upper and lower paleovalleysystems and from glaciolacustrine sediments. The penulti-mate Laurentide advance is interpreted to have been themore extensive of the two based on the distribution of shielderratics in the paleovalley gravels.


# 2008 NRC Canada

The limits of these pre-Late Wisconsinan advances mustbe north and east of Edmonton and Watino, yet closeenough to the study area to introduce shield clasts into thepaleo-Peace River drainage and to impound lakes in its val-leys. No Laurentide glacier extended beyond the most distalshield clasts that it transported, but the locations of theclasts do not necessarily delineate the ice margin. The pres-ence of eastern-provenance clasts in paleovalley fluvialgravels merely indicates that Laurentide ice entered thewatershed of the paleo-drainage system.

The interpretation presented in this study can be recon-ciled with those of Catto (1984), Liverman (1989), Livermanet al. (1989), and Young et al. (1994) by considering the pa-leogeography of the rivers that deposited the sediments onwhich these researchers have based their stratigraphies andwithin which Canadian Shield clasts are absent. All stratigra-phies from which only a single Laurentide advance has beenpostulated come from north- and east-draining watershedssouth and west of Watino and Edmonton. Pre-Late Wiscon-sinan advances that terminated north and east of these water-

sheds would not have introduced shield clasts into thesediments on which the stratigraphies are based.

The distribution of the Canadian Shield clasts on thewesternmost Interior Plains and the paleogeography of thewatersheds in which they occur constrain pre-Late Wiscon-sinan Laurentide ice limits. An early Laurentide glacialadvance was extensive enough to deposit shield clasts in thewatersheds of northern tributaries of ancestral Peace River,including ancestral Beatton River. This early Laurentideadvance did not cross the drainage divide that separates an-cestral Beatton and Halfway rivers, nor did it extend southand west of Watino. The penultimate Laurentide glacialadvance extended west at least as far as the British Colum-bia – Alberta boundary because ancestral Peace River in thestudy area was dammed by this advance. In addition, Cana-dian Shield clasts are present in the lower paleovalley gravelas far west as Lynx Creek.

The pre-Late Wisconsinan, Laurentide glacial history pro-posed here is similar to that of Mathews (1978). It is alsoconsistent with the finding of Liverman et al. (1989) that

Fig. 10. Stratigraphy of the study area (modified from Mathews 1978; reproduced and adapted with the permission of the Minister of PublicWorks and Government Services Canada, 2007 and courtesy of Natural Resources Canada, Geological Survey of Canada).


# 2008 NRC Canada

the early Laurentide advances recognized by Mathews(1978) did not reach Edmonton and Watino.

Cordilleran stratigraphyThe stratigraphy presented in this paper argues for two

Cordilleran glacial advances—an earlier advance sometimeprior to the Middle Wisconsinan, and a Late Wisconsinanadvance that reached at least as far east as Portage Moun-tain. Cordilleran till has not been found east of PortageMountain, but Mathews (1978) argued that the Cordilleranglacier in Peace valley during the Late Wisconsinan wasextensive and that it merged with the Laurentide ice sheet,depositing a single till consisting of mixed eastern and west-ern lithologies. He was unable to distinguish Cordilleran andLaurentide tills, which he attributed to the absence of a dis-tinct boundary between the two till sheets and to a shiftingzone of interaction between the two glaciers. He noted thatthe eastern-provenance clasts decrease significantly in anarrow zone in the western third of the study area,coincident with an increase in the abundance of western-provenance clasts. This area coincides with the zone of co-alescence of Cordilleran and Laurentide ice delineated byflutings (Mathews 1978). The flutings suggest that Cordil-leran ice emanating from ancestral Peace valley was de-flected to the southwest by Laurentide ice.

Several researchers have presented evidence for an exten-sive Cordilleran advance onto the plains and foothills adja-cent to the Rocky Mountains during the Late Wisconsinan(Jackson et al. 1997; Bednarski 1999, 2001; Dyke et al.2002; Dyke 2004; Bednarski and Smith 2007). Jackson etal. (1997) showed that the Foothills Erratics Train wasdeposited during the Late Wisconsinan by coalescence ofLaurentide and Cordilleran ice on the westernmost InteriorPlains. The erratics extend more than 1000 km from Atha-basca valley in the north to Lethbridge in the south. Dykeet al. (2002) suggested that coalescence of the ice sheetsover such a distance would not be possible without the icesheets also coalescing in the vicinity of Peace River. Thisassertion is, however, debatable. Significant coalescence ofthe Cordilleran and Laurentide ice sheets south of Athabascavalley does not require coalescence more than 400 km northof the head of the Foothills Erratics Train.

Bednarski (1999, 2001) and Bednarski and Smith (2007)have argued for an extensive Late Wisconsinan Cordilleranice sheet in the Rocky Mountains northwest of the studyarea. They found striae on mountain ridges that trend per-pendicular to the ridge crests. The implication is that thestriae were produced by ice flowing unhindered by topogra-phy. 36Cl exposure dating of the striated surfaces suggeststhat they were deglaciated between 14 020 ± 760 and13 100 ± 1560 cal years ago. The presence of limestone andschist erratics of western provenance on the plains east ofthe Rocky Mountains supports extensive Late WisconsinanCordilleran glaciation. Bednarski (1999, 2001) and Bednar-ski and Smith (2007) also noted eastern-provenance erraticsand Laurentide till in east-draining valleys of the RockyMountain Foothills, and suggested that Laurentide iceadvanced into these valleys shortly after Cordilleran ice re-treated from them.

Bobrowsky (1989) presented arguments against extensiveLate Wisconsinan Cordilleran glaciation in the Finlay River

valley northwest of the study area. The upper part of Finlayvalley remained ice-free until at least 15 180 ± 100 14Cyears BP (TO-708). Cordilleran ice thus advanced to itseastern limit and retreated significantly in <5000 years.

Bobrowsky (1989) also argued that the western-provenance clasts noted by Mathews (1978) probably weretransported and deposited by a pre-Late Wisconsinan Cordil-leran glacier and were subsequently entrained by LateWisconsinan Laurentide ice and deposited as part of its till.Catto et al. (1996) suggested that the marked decrease ineastern-provenance lithologies in Mathews’ mixing zone isdue more to entrainment of a greater amount of surficialsediment by the Laurentide ice sheet in this area than to co-alescence of Laurentide and Cordilleran ice.

Canadian Shield erratics occur in east-draining valleys inthe Rocky Mountain Foothills to 950 m a.s.l. (Bobrowsky etal. 1991, 1992; Catto 1991; Catto et al. 1996). This eleva-tion is higher than the upper limit of Glacial Lake Peace,therefore the erratics must have been deposited by the Lau-rentide ice sheet. In addition, Laurentide ice could not haveascended the valleys in contact with Cordilleran ice.

If Bobrowsky (1989), Bobrowsky and Rutter (1992), andCatto et al. (1996) are correct, the flutings noted by Math-ews (1978) can only be a product of coalescence of the Lau-rentide and Cordilleran ice sheets prior to the LateWisconsinan. This scenario, however, requires that the flut-ings survive overriding by the Laurentide ice sheet duringthe Late Wisconsinan. An argument to this effect was madeby Catto et al. (1996) to explain the Redwillow Fluting fansouth of the study area. They argued that the RedwillowFluting fan formed at the base of a tongue of pre-LateWisconsinan Cordilleran ice that extended from the RockyMountains through Redwillow valley and onto the AlbertaPlateau. The flutings are overlain locally by Late Wiscon-sinan Laurentide till.

Any coalescence of the Cordilleran and Laurentide ice inthe study area during the Late Wisconsinan must have beenbrief and would have ended with rapid retreat of Cordilleranice to Portage Mountain. At the same time, the Laurentideice sheet may have spread westward into the Rocky Moun-tain Foothills (Bednarski and Smith 2007).

In summary, the pattern of restricted Late WisconsinanCordilleran glaciation envisioned by Bobrowsky (1989),Bobrowsky and Rutter (1992), and Catto et al. (1996) is in-compatible with the more extensive glacial model ofBednarski (1999, 2001) and Bednarski and Smith (2007).The area where Bednarski and Smith worked and the RockyMountain Trench, studied by Bobrowsky and Rutter, areseparated by the Rocky Mountains, but such a fundamentaldifference in extent of Late Wisconsinan Cordilleran icecannot be explained by the different locations of the twostudy areas. Our study fails to resolve this issue. Late Wis-consinan Cordilleran ice may have reached only as far eastas Portage Mountain, or it may have advanced about 50 kmfarther east, where it coalesced briefly with Laurentide ice.

ConclusionFour relict drainage systems have been identified in the

Peace River area in northeast British Columbia. The twoyoungest relict drainages are buried valleys. Each of the


# 2008 NRC Canada

two valley fills comprises a basal fluvial gravel unit and anoverlying, thicker glaciolacustrine mud unit. The glacio-lacustrine sediments are overlain across an erosional contactby till or postglacial fluvial deposits.

The valley-fill sediments of the study area provide evi-dence for three Laurentide and two Cordilleran glaciations.The glaciations are separated by interglaciations or lengthyinterstades, during which the east-flowing drainage wasre-established and rivers achieved maturity. The earliestLaurentide advance was less extensive than the penultimateadvance, and the penultimate advance, in turn, was lessextensive than the Late Wisconsinan advance. An advanceof the Cordilleran ice sheet reached beyond the RockyMountain front, but the extent of the Cordilleran glacier inPeace River valley during the Late Wisconsinan remainsuncertain.

AcknowledgementsFinancial support for this study was provided by the Natural

Sciences and Engineering Research Council of Canada(NSERC) through a Discovery Grant to John Clague and bythe Geological Survey of Canada courtesy of Stephen Evans.Virginia Pow, Bryan Hartman, John Orwin, Kenna Wilkie,Britta Jensen, and Tracy Arsenault assisted in the field.Jonathan Driver identified the dated interstadial bison verte-bra. Reviews of the paper by Norm Catto, Duane Froese, Rob-ert Gilbert, and Vic Levson contributed to its improvement.

ReferencesAshley, G.M. 1988. Classification of glaciolacustrine sediments. In

Genetic classification of glacigenic deposits. A.A. Balkema,Rotterdam, The Netherlands, pp. 243–260.

Barendregt, R.W., and Irving, E. 1998. Changes in the extent of NorthAmerican ice sheets during the late Cenozoic. Canadian Journal ofEarth Sciences, 35: 504–509. doi:10.1139/cjes-35-5-504.

Beach, H.H., and Spivak, J. 1943. The origin of the Peace RiverCanyon, British Columbia. American Journal of Science, 241:366–376.

Bednarski, J.M. 1999. Preliminary report on mapping surficialgeology of Trutch map area, northeastern British Columbia. InCurrent research. Geological Survey of Canada, Paper 1999-A,pp. 35–43.

Bednarski, J.M. 2001. Drift composition and surficial geology ofthe Trutch map area (94G), northeastern British Columbia. Geo-logical Survey of Canada, Open File Report D3815.

Bednarski, J.M., and Smith, I.R. 2007. Laurentide and montaneglaciation along the Rocky Mountain Foothills of northeasternBritish Columbia. Canadian Journal of Earth Sciences, 44: 445–457. doi:10.1139/E06-095.

Bobrowsky, P.T. 1989. Late Cenozoic geology of the northernRocky Mountain Trench, British Columbia. Ph.D. thesis, Uni-versity of Alberta, Edmonton, Alta.

Bobrowsky, P.T., and Rutter, N.W. 1992. The Quaternary geologichistory of the Canadian Rocky Mountains. Geographie physiqueet Quaternaire, 46(1): 5–50.

Bobrowsky, P.T., Catto, N.R., and Levson, V. 1991. Reconnais-sance Quaternary geological investigations in Peace RiverDistrict, British Columbia. British Columbia Ministry of Energy,Mines and Petroleum Resources, Paper 1991-1, pp. 345–358.

Bobrowsky, P.T., Smith, C.P., Grant, B., and Newell, J.M. 1992.Quaternary studies in the Peace River District, 1990; stra-tigraphy, mass movements and glaciation limits (94P). British

Columbia Ministry of Energy, Mines and Petroleum Resources,Paper 1992-1, pp. 363–374

Campbell, D.D. 1959. Portage Mountain Dam (sire 3A) drill logs.In Peace River Hydro-Electric Project — Geology. Peace RiverPower Development Company Limited, Vancouver, BC.

Catto, N.R. 1984. Glacigenic deposits at the Edmonton ConventionCentre, Edmonton, Alberta. Canadian Journal of Earth Sciences,21: 1473–1482.

Catto, N.R. 1991. Quaternary geology and landforms of the easternPeace River region, British Columbia NTS 94A/1,2,7,8. BritishColumbia Ministry of Energy, Mines and Petroleum Resources,Open File Report 1991-11.

Catto, N.R., Liverman, D.G.E., Bobrowsky, P.T., Rutter, N., andMandryk, C.A.S. 1996. Laurentide, Cordilleran, and Montaneglaciation in the western Peace River – Grande Prairie region,Alberta and British Columbia, Canada. Quaternary International,32: 21–32.

Dethier, D.P. 2001. Pleistocene incision rates in the western UnitedStates calibrated using Lava Creek B Tephra. Geology, 29: 783–786. doi:10.1130/0091-7613(2001)029<0783:PIRITW>2.0.CO;2.

Dyke, A.S. 2004. An outline of North American deglaciation withemphasis on central and northern Canada. In Quaternary glacia-tions: extent and chronology. Edited by J. Ethers and P.L.Gibbard. Elsevier, Amsterdam, The Netherlands, pp. 373–424.

Dyke, A.S., Andrews, J.T., Clark, P.U., England, J.H., Miller, G.H.,Shaw, J., Veillette, J.J., and Mix, A.C. 2002. The Laurentide andInnuitian ice sheets during the last glacial maximum. QuaternaryScience Reviews, 21: 9–31. doi:10.1016/S0277-3791(01)00095-6.

Eyles, N., Eyles, C.H., and Miall, A.D. 1983. Lithofacies types andvertical profile models; an alternative approach to the descrip-tion and environmental interpretation of glacial diamict anddiamictite sequences. Sedimentology, 30: 393–410. doi:10.1111/j.1365-3091.1983.tb00679.x.

Fladmark, K.R., Driver, J.C., and Alexander, D. 1988. The Paleo-indian Component at Charlie Lake Cave (HbRf-39), BritishColumbia. American Antiquity, 53: 371–384. doi:10.2307/281025.

Geertsema, M.E., and Jull, A.J.T. 2002. Charcoal and sedimen-tation record from a fan at Bear Flat, northeastern BritishColumbia. Geological Society of America, Abstracts with Pro-grams, 34(6): 319.

Goddard, E.N., Trask, P.D., DeFord, R.K., Rove, O.N., Singewald,J.T., Jr., and Overbeck, R.M. 1963. Rock Color Chart. Geologi-cal Society of America, New York, NY.

Gustavson, T.C., Ashley, G.M., and Boothroyd, J.C. 1975. Deposi-tional sequences in glaciolacustrine deltas. In Proglacial fluvialand lacustrine environments. Society of Economic Paleontolo-gists and Mineralogists, Special Publication 23, pp. 264–280.

Hartman, G.M.D. 2005. Quaternary stratigraphy and geologichistory of the Charlie Lake (NTS 94A) map-sheet Area, BritishColumbia. M.Sc. thesis, Simon Fraser University, Burnaby,B.C., 165 p.

Henderson, E.P. 1959. Surficial geology of Sturgeon Lake map-area, Alberta. Geological Survey of Canada, Memoir 303.

Jackson, L.E., Jr., Phillips, F.M., Shimamura, K., and Little, E.C.1997. Cosmogenic 36Cl dating of the Foothills Erratics Train,Alberta, Canada. Geology, 25: 195–198. doi:10.1130/0091-7613(1997)025<0195:CCDOTF>2.3.CO;2.

Jull, A.J.T., and Geertsema, M. 2006. Over 16,000 years of firefrequency determined from AMS radiocarbon dating of soilcharcoal in an alluvial fan at Bear Flat, northeastern BritishColumbia. Radiocarbon, 48: 435–450.

Kleman, J., Fastook, J.L., Stroeven, A.P., Fabel, D., and Marshall,S.J. 2002. Geologically and geomorphologically constrained


# 2008 NRC Canada

numerical model of Laurentide ice sheet inception and build-up.Quaternary International, 95–96: 87–98. doi:10.1016/S1040-6182(02)00030-7.

Liverman, D.G.E. 1989. The Quaternary geology of the GrandePrairie area, Alberta. Ph.D. thesis, University of Alberta,Edmonton, Alta.

Liverman, D.G.E., Catto, N.R., and Rutter, N.W. 1989. Laurentideglaciation in west-central Alberta; a single (late Wisconsinan)event. Canadian Journal of Earth Sciences, 26: 266–274.

MacDonald, G.M., Beukens, R.P., Kieser, W.E., and Vitt, D.H.1987. Comparative radiocarbon dating of terrestrial plant macro-fossils and aquatic moss from the ‘‘ice-free corridor’’ of WesternCanada. Geology, 15: 837–840. doi:10.1130/0091-7613(1987)15<837:CRDOTP>2.0.CO;2.

Mathews, W.H. 1978. Quaternary stratigraphy and geomorphologyof Charlie Lake (94A) map-area, British Columbia. GeologicalSurvey of Canada, Paper 76-20.

Mathews, W.H. 1980. Retreat of the last ice sheets in northeasternBritish Columbia and adjacent Alberta. Geological Survey ofCanada, Bulletin 331.

Mathews, W.H. 1989. Development of Cordilleran landscapes dur-ing the Quaternary. In Quaternary geology of Canada andGreenland. Edited by R.J. Fulton. Geological Survey of Canada,Geology of Canada, no. 1, pp. 32–34.

May, R.W. 1977. Facies model for sedimentation in the glacio-lacustrine environment. Boreas, 6: 175–180.

McMillan, M.E., and Heller, P.L. 2002. Comparison of long-termrates of incision across the Rocky Mountain orogenic plateau.Geological Society of America, Abstracts with Programs, 34:409.

Miall, A.D. 1977. A review of the braided-river depositionalenvironment. Earth-Science Reviews, 13: 1–62. doi:10.1016/0012-8252(77)90055-1.

Miall, A.D. 1992. Alluvial deposits. In Facies models, response tosea level change. Edited by R.G. Walker and N.P. James. Geo-logical Association of Canada, St. Johns, Newfoundland,pp. 119–142.

Pazzaglia, F.J., Wisniewski, P.A., and Zaprowski, B. 2001. The flu-vial record of Rocky Mountain epeirogeny; a review. GeologicalSociety of America, Abstracts with Programs, 33: 51.

Pederson, J.L., and Pazzaglia, F.J. 2002. Middle Cenozoic qui-escence and late Cenozoic exhumation; how base-level changethrough drainage development and epeirogenesis has sculptedthe West. Geological Society of America, Abstracts with Pro-grams, 34: 408–409.

Rutter, N.W. 1976. Multiple glaciation in the Canadian RockyMountains with special reference on north-eastern BritishColumbia. In Quaternary stratigraphy of North America. Editedby W.C. Mahaney. Hutchinson and Ross, Stroudsburg, Penn-sylvania, pp. 409–440.

Rutter, N.W. 1977. Multiple glaciation in the area of WillistonLake, British Columbia. Geological Survey of Canada, Bulletin273.

St-Onge, D.-A. 1972. La stratigraphie du quaternaire des environsde Fort-Assiniboine, Alberta, Canada. La Revue de Geographiede Montreal, 26: 153–163.

Stott, D.F. 1982. Lower Cretaceous Fort St. John Group and UpperCretaceous Dunvegan Formation of the foothills and plains ofAlberta, British Columbia, District of Mackenzie and YukonTerritory. Geological Survey of Canada, Bulletin 328.

Stott, D.F., and Aitken, J.D. (Editors). 1993. Sedimentary cover ofthe craton in Canada. Geological Survey of Canada, Geology ofCanada, no. 5.

Tokarsky, O. 1967. Geology and groundwater resources (Qua-ternary) of the Grimshaw area, Alberta (Canada). M.Sc. thesis,University of Alberta, Edmonton, Alta.

Westgate, J.A., Fritz, P., Matthews, J.V., Jr., Kalas, L., and Green,R. 1971. Sediments of mid-Wisconsinan age in west-centralAlberta; geochronology, insects, ostracodes, Mollusca, and theoxygen isotopic composition of molluscan shells. GeologicalSociety of America, 3: 419.

Westgate, J.A., Fritz, P., Matthews, J.V., Jr., Kalas, L., Delorme,L.D., Green, R., and Aario, R. 1972. Geochronology and palaeo-ecology of mid-Wisconsinan sediments in west-central Alberta,Canada. International Geological Congress, 24, p. 380.

Wolfe, B., and Teller, J.T. 1995. Sedimentation in ice-dammedglacial Lake Assiniboine, Saskatchewan, and catastrophic drai-nage down the Assiniboine Valley. Geographie physique et Qua-ternaire, 49: 251–263.

Woodcock, N.H. 1977. Variations in till macrofabric over veryshort distances. Geological Society of America Bulletin, 88:1231–1236. doi:10.1130/0016-7606(1977)88<1231:SOFSUA>2.0.CO;2.

Woolf, K.E. 1993. Radiocarbon chronology for glacial Lake Peace.M.Sc. thesis, Simon Fraser University, Burnaby, B.C.

Young, R.R., Burns, J.A., Smith, D.G., Arnold, L.D., and Rains,R.B. 1994. A single, late Wisconsin, Laurentide glaciation,Edmonton area and southwestern Alberta. Geology, 22: 683–686.doi:10.1130/0091-7613(1994)022<0683:ASLWLG>2.3.CO;2.


# 2008 NRC Canada

Documents

Quaternary stratigraphy and glacial history of the Peace ...qgrc/papers/Hartman and Clague CJES 2008.pdf · jointing, weathering, sliding planes, rupture surfaces, and the angle of