Quaternary Lowlands, already deglaciated paleo-ecological con

37 Pages·2012·3.52 MB·English

LEIDSE GEOLOGISCHE MEDEDELINGEN, Vol. 38, 1966, pag. 91—128, published separately 1.9.1966
Late Quaternary vegetationand glacial history
in the St. Lawrence Lowlands, Canada
BY
Pierre Lasalle
Abstract
This paper presents data ofvarious kinds concerning the Quaternarygeologyofthe St. Lawrence Lowlands: pollendiagrams,
C14 dates, and diatomfloras.These data show thatthe highest parts ofthe St. Lawrence Lowlandswere already deglaciated
more than 12,000years ago, asappears fromthe existenceofglaciallakes aroundsome ofthe MonteregianHills.
The collectedevidenceseemstoconfirmthe data concerningthe age oftheChamplain Sea:itlastedfromapproximately 11,400
years B.P. tosomewhat before 9,500years B.P.
The datasupplyalso evidence foraloweringofsealevelduringthe ChamplainSeaepisode, relatedtotheSt.Narcissereadvance,
followed by asealevelrise (indicated by deeperwater) after the retreat ofthe St. Narcisseice.
As to the pollen diagramspresented here, theyshow that the New Englandpollenzonation canbe applied tothe lacustrine
pollensequence ofthe St.Lawrence Lowlands,andstrongly suggest (togetherwith C14dates) thatacorrelationexists between
the YoungerDryas ofnorthwestern Europe, the St.Narcisse readvance,and the firstpart ofthe pollen zone A4 from New
Englandasdescribed by Davis and other workers.
The diatom data ofLake Hertel appear toconfirm the present elevation ofthe maximum marine limit (level) in theMontreal
area asbeing approximately 570 feet or 171 meters. Finally the diatom floras supply informationonthe paleo-ecological con-
ditions ofthe lakes studied.
Résumé
Le présent travail rassemble des données de plusieurs sortes sur la géologiequaternaire desbasses terres duSt.Laurent: des
diagrammespolliniques, des mesures d’âgeauC14 et des flores dediatomées fossiles.
Onpeut enconclure d’abordque la déglaciationdespoints les plus élevés desbasses terres du St. Laurenteutlieu 12,000ans
avant aujourd’hui.
L’âge de l’épisode marin désigné sous le nomde „Mer Champlain” semble être confirmé:il s’étendraitde 11,400jusqu’àun
peu avant 9500 ansavant aujourd’hui. Les faits semblentaussi démontrerqu’un abaissement relatifdu niveau delamer s’est
produit pendantcette période,suivie d’une remontéerelativedu niveau des eauxmarines. Cet événement semble êtrerelié à
l’avancée glaciairede St.Narcisse, suivie de sonretrait.
Les diagrammespolliniques semblent démontrer que la séquence des zônes polliniques de la Nouvelle Angleterrepeut être
appliquéeauxséquencespolliniquesdes dépôtslacustres desbasses terres du St.Laurent.
Ces diagrammesdonnentdefortesindications (enconsidérant aussi les mesuresd’âgeauC14) queleDryas supérieurdel’Europe
septentrionalede l’ouest, laréavancée deSt.Narcisse etlazonepolliniqueA4 (Davis, 1958) sont corrélatifs.
Les diatoméesfossiles du Lac Hertel (danslapartietoutàfait inférieure dela section) semblentconfirmer l’altitude actuelle du
niveau maximum de la dernièreinvasion marine (Mer Champlain) dans la région de Montréal: environ570 piedsou 172 m
au-dessus du présent niveau moyen delamer.
Enfin les flores de diatoméesfossiles ont permis de tirer des conclusions concernant les conditions paléo-écologiques des lacs
étudiés.
Table of contents
Chapter I Introduction 92 4. The relation between the moraines and the
1. Purpose ofthe investigation 92 glacial lakes 98
2. Methods and equipment 92 5. TheTwo-Creeks,ValdersandtheChamplainSea 98
3. Acknowledgements 93
Chapter III Absolute chronology: C14 principles 100
Chapter II Historical geology and geomorphology 93 a. Marineshells 100
1. Glacialsedimentsin the St. Lawrence Lowlands 93 b. Lake sediments 101
2. Moraines in the St. Lawrence Lowlands ... 94 Chapter IV Sea level changes 101
3. The late-glaciallakes in theWestern part ofthe
St. Lawrence Lowlands 96 Chapter v Pollen diagrams 102
92 Lasalle: Late Quaternary vegetationand glacial history
1. Review of the literature 102 a. Particular cases 116
2. Location of the new deposits 103 b. Fossil diatoms ofthe St.Hilaire bog ... 116
3. The new pollen diagrams 105 c. Fossil diatoms ofthe Lake Hertel 117
a. General discussion 105 d. Fossil diatom of the St.Bruno bog 118
....
b. The pollen zones 107 3. Paleoecology of the lakes investigated .... 118
c. The St. Hilaire bog diagram 107 a. Comparison with the Crater Lake flora . . 118
d. The lake Hertel diagram 110 b. Paleoecology of the St.Hilaire bog 119
....
e. The St. Bruno bog diagram Ill c. Paleoecology of the Lake Hertel 119
j. The St.Antoine core diagram Ill d. Paleoecology of the St.Bruno bog .... 119
g. The lake Kenogami diagram 112
Chapter VIII The endemics and the late-glacial
Chapter VI Proposed correlation and vegetation corridor 120
history in the St. Lawrence Lowlands 113
Chapter IX General conclusions 120
Chapter VII Fossil diatom floras 115 References 123
1. Definition and literature 115 Appendix 127
2. Fossil diatoms 116 Figs. 7and 11 to 16 are folded in the back cover.
CHAPTER I
INTRODUCTION
The interest ofthe author for the Pleistocene geology in the cores analysed and special diatom data are
oftheSt. Lawrence Lowlands and the Montrealarea presented in a qualitative way.
developed in thecourse ofthe mapping ofthesurficial Some ecological conclusions are finally drawn and
sediments of the Beloeil map area during the field anattemptis made torelate the diatomdatawith the
season of 1960 (Lasalle and Elson, 1962). geological history of the area.
In the successive years, many more excursions were
made by the author alone or with other people to
2. METHODS AND EQUIPMENT
collect material. During the fall of 1963 the author
visited all the depressions that appeared onthe aerial The core ofthe St. Hilaire bog was collectedby the
photograph of the St. Hilaire mountain in the hope author with a Hillerborer. As the St. Hilaire bog is
of discovering some late-glacial sediments or at least completely filled now, and its surfaceforested, it has
sedimentsolder than the bogs lying onthe Champlain been possible to make a reconnaissance survey to
Sea deposits (Terasmae, 1960). The St. Hilaire bog find the maximum depth to refusal before collecting
was finally located, and it was thought that with the the core for pollen analysis. Dr. Terasmae kindly
9.5 meters ofsediments collectedit wouldbe possible collectedthe sample for C14 dating with a Livingstone
to definelate-glacial pollen zonesin the St. Lawrence borer.
Lowlands. The Lake Hertel core was obtained with a Living-
Pollen analysis ofthe core ofSt. Hilaire bog as well stone sampler operated by the partyofDr. Terasmae
as the core of Lake Hertel and St. Bruno bog has and the author. The sample for C14 dating was
indeed revealed an interesting sequence. obtained from the same core.
The core of the St. Bruno bog was collected by the
author, with a Hiller borer.
1. PURPOSE OF THE INVESTIGATION
The boring was made on the side of the partially
The purposeofthe present work is to definethe late- filled lake.
glacial pollen zonesinthe St. LawrenceLowlands and The coreoftheChamplain Seasedimentswas obtained
attempt correlationwith otherpollen zones definedin from the Ecole Polytechnique, University of Mon-
the North-easternUnited States, and with geological treal, through courtesy of Dr. Jean Bérard.
events in the area (Champlain Sea, morainesetc.). Preparations for pollen analysis ofthe St. Hilaire bog
On the basis ofthe new radiocarbon dates obtained were first made in the Palynology Laboratory ofthe
from these cores, it appeared also pertinent to discuss Geological SurveyofCanada, in Ottawa. The samples
some ofthe hypotheses concerning the glacial lakes were prepared again in the Palynology Laboratory
thatpreceded theChamplain Sea in the St. Lawrence ofthe University ofLeiden, The Netherlands, in the
Lowlands, and toattempt acorrelation ofthe morai- following manner: they were first boiled in KOH
nes thathave already been outlined. 10% for a few minutes, and then cooledand left in
Finally an abundantdiatomflorahasbeen discovered the ultra-sound apparatus at moderate intensity for
Introduction 93
5 minutes; mineral matter was removed in the more The author is also grateful to Dr. J. Terasmaeofthe
silty ones by using eitherbromoform or hydrofluoric Geological Survey ofCanada, who has collected the
acid; acetolysis was performed on all of them. They materialfor the C14 dating ofthe St. Hilaire bog and
were mounted in glycerine jelly. The same method also the core of Lake Hertel. His collaboration is
was followed for the St. Bruno bog samples. gratefully acknowledged.
For the Lake Hertel samples, bromoform was not The author has beenintroducedintothe study ofdia-
used to remove the mineralmatter;only hydrofluoric toms by Mr. A. du Saarand Mr. A. van derWerffof
acid was employed for the lowermost samples; they the Geologische Dienst Nederland. He also spent two
were mountedin liquid glycerine to allow the move- weeks at the Institute of Quarternary Geology and
ment of the grains under the microscope. Paleontology of the University of Uppsala working
The numbers of grains counted per slide range with Docent Maj-Britt Florin. The help of these
between 200 and 250; sometimes a larger number persons with the identificationof diatoms or confir-
was counted. mation of identificationis gratefully acknowledged.
The slidesfor diatom analysis were prepared by boi- But the authorbears all responsability for omission,
ling the materialwith diluteperoxide. To accelerate misidentification, or any conclusion drawn from the
theprocess ofoxydation, drops ofdilutepermanganate diatomdata.
of potassium were added. The mounting medium is He would like to thank also the following persons for
Gurr's Clearax. valuable discussions: Dr. S. Jelgersma, Dr. W. H.
Identificationsofthepollen grains and theircounting Zagwijn, Mr. A. van derWerffand Mr. A. du Saar
were madewithaLeitz Dialuxusing mostly a magni- ofthe Geologische Dienst Nederland, and Dr. J. G.
ficationof500 times. The critical identificationswere Vogel, of the University of Groningen.
made at 1250 times.
Mostofthephotographs weremadeattheGeologische The following organizations and persons havehelped
DienstNederlandinHaarlemwithaLeitz microscope with the C14 datings:
automatic camera; Agfa AGEPE FF Dokumenten- The Geological Survey ofCanada;
filmwas used (12Din). Asmall numberofthe diatom The Centre National de la Recherche Scientifique in
photographs have also been made in Uppsala, with Paris and Madame G. Delibrias;
a Wild microscope equipped with a photo-tube; the The Groningen C14Laboratoriesand Dr. J. C.Vogel;
same type offilm was used. Finally, the enlargements TheYaleGeochronology Laboratoriesand Dr. Minze
of the fossils have been calculated individually for Stuiver;
each print as it revealed difficult to get a uniform Dr.J.A.Elson,ofMcGillUniversity, whohasallowed
enlargement fromthephotographic shop. the author to quote some ofhis C14 data.
3. ACKNOWLEDGMENTS The author would also like to thank the following
Theauthorwouldlike to thankProf. Dr. F. P. Jonker persons who havehelped inone way oranother: Miss
and Prof. Dr. A. J. Pannekoek for their interest in Els Schilstrawhohasprepared with greatpatience the
the present work. pollen slidesofthe marinesediments, and Mr. T. A.
He wouldlike also tothank specially Dr. Th. van der Wijmstra and Mr. F. J. J. van Heyst, ofthe Palyno-
Hammenunder whose direction the present research logy Laboratory at Leiden.
wasdone.Dr. van der Hammenhasbeen most encou- Lastly, the financial participation of the Quebec
raging during all phases of the work, and without Department of Natural Resources is here gratefully
his help and support, the present paper could not acknowledged.
have come to completion.
CHAPTER II
HISTORICAL GEOLOGY AND GEOMORPHOLOGY
1. GLACIAL SEDIMENTS AND STRATIGRAPHY IN view. The interested reader is refered to the work
THE ST. LAWRENCE LOWLANDS ofGadd (1960) and Mac Clintock and Stewart(1965)
The basic scheme oftheglacial stratigraphy ofthe St. for a complete historical background.
Lawrence Lowlands is mainly the result of recent According torecent studies, twotillsheets have been
studies (Gadd, 1955, 1960; Terasmae, 1955, 1958; recognized in the Bécancour area, located approxi-
Gadd and Karrow, 1959). Mac Clintockand Stewart mately half-way between Montrealand Quebec City.
(1965, p. 11 to 24) have ably resumed the literature The lower till sheet is called the Bécancour, and is
dealing with the Pleistocene studies in the St. Law- separated from the upperone, the Gentilly till, by the
rence Lowlandsforthelastonehundredyears,or so. St. Pierre sediments. The latter are sands and peat
It is not our intention here to take up again that re- offreshwater origin indicating a climate cooler than
94 Lasalle: Late Quaternary vegetation and glacial history
the present (Gadd, 1960, p. 12). As material from 11,500 yrs. B. P. onaccount ofG14 dates obtainedon
thesesediments has yielded ages possibly ofthe order fossil marine shells.in the St. Lawrence Lowlands.
of60,000 years B. P. (Gadd, 1960,p. 24), the deposit Indeed, the sea could not have occupied the St. Law-
has been assigned stratigraphically to the early Wis- rence Lowlandsunlessthe areawas deglaciated; hence
consin (Flint, 1963, p. 403). the maximumageobtainedonshellsofthe Champlain
In adjacent areas of New England and New York Sea would give a minimumage forits déglaciation.
State, Mac Clintock and Terasmae (1960) and Mac On the otherhand, amaximumagehas been assig-
Clintock and Stewart (1965) have also reported two ned to the St. Antonin moraine (GSC-102, 12,770
tills: The Malone till deposited by anice sheet mo- ±170 yrs. B. P.; Dyck and Fyles, 1963, p. 18). This
ving from the northeast, and the Fort Covington till datehas been performed on marine shells that were
deposited by anicesheetmoving fromthe northwest. „part of a marine-deltaic-outwash deposit formed
Mac Clintock and Stewart (1965) have suggested the when wastage of the ice sheet permitted the sea to
Fort-Covington is ofPort-Huron (Mankato) age and penetrate into this part of the St.Lawrence basin"
that the Malone tillis ofCary age. This is based in (Dyck and Fyles, 1963, p. 18). This dateis then more
part on C14 dating, but mainly on the fact that the likely to be related to animprovement ofthe climate,
„rearrangement" ofthe ice that permitted two read- since the deltaic materialhas been deposited during a
vances from so divergent directionsis ofa magnitude „wastage" ofthe icesheet. Theseawaters at that time
that deserves a substage status (Mac Clintock and appear to haveoccupied only a re-entrant in theice.
Stewart, 1965, p. 41). The correlationofthe Malone Later, the ice that deposited the St. Antoninmoraine
and Fort Covington tills with the Bécancour and overrodetheoutwash deposit. Thecorrelationproblem
Gentilly tills respectivelly, though probable (because ofthis morainewith other areaswillbe taken up later
of closeness oftheir type areas), does not yet rest on in this section. But is it important to rememberthat
firm ground. thesouth-westernpartoftheHighland Front Moraine
The glacial sediments ofthe St. Lawrence Lowlands System may be younger than the north-eastern part
are overlaininplaces by varvedclays, separating them as it may have been deposited during aretreat ofthe
from the marine clays above. The latter are for the ice from a more advanced southern position. Indeed,
greater partsilty clays, fossiliferous inplaces. Upward the nature of the deposits suggests that they were
in thesection, they are replaced by interbeddedsands deposited during a wastage ofthe ice sheet.
and clays and then by sands, that are the estuarine The Drummondville moraine (fig. 1) occupies a
shallow-water sediments of the early St. Lawrence. position subparallel to the Highland Front Moraine
The latter have been in places channelled,together System. It consists ofan elongated ridge andis inter-
with the underlying clays: that river erosion was preted (Gadd, 1960, 1964) as a „halt" oftheice as it
produced as the St. Lawrence River was gradually retreated across the St. Lawrencevalley. Aminimum
retreating to itspresent position. Some ofthese chan- age has been assigned to this moraine because the
nelsare occupied today by large peatbogs and under- basal layer ofa bog resting onit has yielded ages of
fit streams (Verchères and Sorel areas). The distri- ca. 9500 yrs. B. P. (L-441-C, 9,550±600 years;Teras-
butionofthe peatbogs onaerial photographs depicts mae, 1960, p. 16). However, the moraine, because
very well the pattern of those braided streams. of its position with respect to the only plausible
The St. Lawrence River may have retreated to its entrance for the marine waters (Quebec narrows) in
presentposition some6000(?) years agowheneustatic the St. Lawrence Lowlands, is most probably a few
sea-levelreached its presentpositionorapproximately hundred years older, than this minimum reported
so (see also Terasmae, 1960, p. 21). age (above), ifwe take into account the maximum
Therivervalley itselfinwhich theSt. LawrenceRiver ages reported on fossil marineshells in the St.Law-
is flowing is undoubtedly a geologic featurethat is of rence Lowlands. In fact, the Drummondvillemoraine
Tertiaryage or even older. It has beenintermittently may be as old as the maximumreported age ofthe
occupied bycontinentalglaciers during thePleistocene marineshells,or even slightly older, as noindicationof
and presumably filledin part by glacial and marine calving into the sea waters has been reported.
sediments, before becoming the major river course The shells that were found moreor less incorporated
that it is today. into it are attributed to reworking by waves (Gadd,
1960,p. 16). Presumably, theicefront mayhavebeen
standing in the waters of proglacial Lake Frontenac,
2. MORAINES IN THEST. LAWRENCE LOWLANDS
just prior to the opening of the Quebec narrows to
Three moraines or systems of moraines have been the sea waters. This would make the moraineslightly
reported in the St. Lawrence Lowlands. The several older than 11,400 yrs. B. P.
students who have contributed to their descriptions The northernmostknown moraine,i.e. theSt. Narcisse
and definitionswill be mentionedbelow. morainewasfirstreported byOsborne (1950).However
The southernmostofthose is Gadd's Highland Front thefirstsegment ofithadbeenreported by Mawdsley,
Moraine System (Gadd, 1964). In fig. 1 this system (fig. 1) in hismemoirof 1927, dealing withthebedrock
has been generalized. It includes variousgravelly geology oftheSt. Urbainarea, Quebec. We quotethe
deposits, and the St. AntoninmoraineofLee (1962). following paragraph from Mawdley's paper (p. 39) :
Gadd (1964) considers that this system is older than „Two well defined terminal moraines are known
Historical geology and geomorphology 95
Fig.1.Moraines intheSt.Lawrance Lowlands
within thearea. One dams the outletofFourvalley, collecting sites of the shells. This distance is of the
outofwhichflows thebrook whichfeeds lake Cygnes, order of200 kilometers. Hence thereal ageofthe St.
and the other dams the outletof the valley through Narcisse readvance must be in the 10,500 years B. P.
which the upperpart ofMoucheriver flows". In the range.
last few years several workers have found extensions Theoscillation ofthe St. Narcisse readvance appears
ofthismoraine:LaverdièreandCourtemanche(1959), to fit the low Quercus and strong herb zone (A4a) of
Elson and Karrow (reported in Lasalle, 1965), Gadd the St. Hilaire bog pollen diagram. The low Quercus
and Karrow (1959). The concensus of geologists is very prominent in most late-glacial diagrams of
having worked in theareaofthe St. Lawrenceis that New England and the several dates obtainedon zone
the stratigraphie position ofthis moraineis equivalent A4 (low Quercus) range from approximately 10,800 to
to the Valders readvanceofthe Mid-continent.How- 10,400 years B.P. (see Davis, 1965, p. 389). These
ever, only a minumumage had been available so far, dates suggestvery strongly that the St. Narcisse read-
thesameasfortheDrummondvillemoraine(Terasmae, vance is contemporaneouswith the Younger Dryas of
1960, p. 20), and ofthe order of9,500 yrs. B. P. European stratigraphers. That readvance is certainly
C14 assays performed on marine shells collected at also contemporaneouswith some phase ofthe Cham-
the marinelimitalong the Saguenay Riverand inthe plain Sea. Indeed, in the Trois-Rivières area, it has
Lake St. John area (see fig. 1) and locatedbehind the been demonstrated (Gadd and Karrow, 1959) that
inferredposition ofthe ice front(Lasalle, 1965),have the ice overrode the marine clays at the timeof the
yielded ages from 9,400 to 10,250 approximately (cf. deposition of the St. Narcisse moraine.
appendix, table 3: GrN-4811, Gif-423, and Gif-424; Furthermore,inthe Verchèresarea (Lasalle, 1964), a
see also Lasalle, 1966). clay layer containing marine fossils (exact location:
The timeof the St.Narcisse readvance (and also the 46°46' latitude, 73°16' longitude, in an excavation
Drummondville„halt") is then olderthan the above for a gravel pit) has been foundoverlying agranule
datesbytheamountoftimeneededfortheicetoretreat gravel and coarse sand layer containing a mixed
from the St.Narcisse moraineto the position of the fossil faunaoffresh water and marine mollusks.
96 Lasalle: Late Quaternary vegetation and glacial history
FIG.2.PROPOSED CORRELATION BETWEEN THE MORAINES AND POLLEN ZONES.
Mollusksidentifiedby Dr. C. O. van Regteren Altena readvance. According to Hughes, it is only of local
oftheRijksmuseum van Natuurlijke HistorieinLeiden importance (p. 564).
are the following: As for the Lake St. John area:what appears to be a
gravelly moraine (Lasalle, 1966) in the Hébetville
Fresh water mollusks: Marine mollusks: area, deserves no status at thepresent timeas a clima-
Valvata sp. Portlandia arctica (Gray) tic oscillation. It appears to be only a forced halt
Amnicola sp. Macoma balthica (L) becauseofthe depression oftheLake St.John basin.
Pleurocera sp. Hiatella sp. Considering the newpollen stratigraphy atSt. Hilaire
Physa sp. Mya sp. bog and other sites and taking the C14 dates at face
Lymnaea (Stagnicola) sp. value, we have proposed a tentativecorrelationofthe
Sphaerium sp. moraine in the St. Lawrence Lowlands (in fig. 2).
Portlandiaarcticaand Sphaerium, thespecies ofthe latter 3. THE LATE-GLACIAL LAKES IN THE WESTERN
PART OF THE ST. LAWRENCE LOWLANDS
being probably nitidum, which has been found else-
where inthe St. LawrenceLowlands,are indicativeof Several authors have written on thelate-glacial lakes
cold conditions. in the St. Lawrence Lowlands. They cannot all be
As for the granule gravel, it is interpreted (together mentionedhere.Mostoftherecent additionshavebeen
with the mixed fauna) as indicating a shoaling and made on the basis ofthe framework set up by Chap-
freshening of the water at the time of deposition of man in his paper of 1937. It appears thatin the light
the St.Narcisse moraine. The layer overlying the of the new C14 date (G.S.C-419) this story can be
sand layer is indicativeofa deepening ofthe water, made more precise or at least reinterpreted. Several
following the deposition of the St. Narcisse moraine radiocarbon dates in the New York area, south of
and retreat ofthe icefrom the present locationofthe Lake Ontario indicate now thatLake Iroquois was in
same moraine. existence invarious phases from 12,500 to 11,500 ap-
North oftheposition ofthe St. Narcisse moraine, only proximately (cf. appendix, table 2). According to
the much controversialCochrane readvance has been Rubin and Berthold (1961, comments on W-883)
reported. Recent work by Hughes (1965) appears to the Niagaraescarpmentwas freeof iceby some 12,000
have shaken the status of this moraine as being a years B. P., and the „Valders" ice never advanced
Historical geology and geomorphology 97
south of the area. Other dates quoted in the table 2 free ofice when the marine waters invaded the St.
and G.S.C.-419 tend to support the idea that dégla- LawrenceLowlands, butalsoa large areanorthofthe
ciationwas rapid in the Adirondacksand that a large Champlain Valley and also the St. Lawrence Valley
body of meltwaterexisted in the Lake Ontario basin west ofMontreal(see alsoElson, 1963). Indeedpartof
priortotheinvasionoftheSt.LawrenceLowlandsbythe the St. Lawrence Lowlandsinthe west mayhavebeen
marinewaters. Glacial Lake Iroquois (as this body of deglaciated rapidly and above the sea level (Lake
water is called) was draining first via the Mohawk Ontario:Karrowetal., 1961)beforethe marinewaters
Valley and Hudson River south of the Adirondack reached the area; a strong argumentfor this,reported
Highlands. Upon retreat of the Fort Covington ice also by Mac Clintock and Terasmae (1960, p. 238), is
(Mac Clintock and Terasmae, 1960) Lake Iroquois the fact that the varved clays and uppermostpart of
drained successively at lower levels via the Covey underlying till (both underlying the marine clays in
Hill drainage way into Chapman's Lake Vermont places), bear evidenceofsub-aerialexposure, in exca-
(present elevationofshore line:725feet or220meters vations made in the New-York part of the St. Law-
above sea level). Furthermore, geological evidence rence Seaway and Power project. This would not be
brought by the work of Mac Clintock and Stewart the first time that there is partial déglaciation ofthe
(1965) and Mac Clintock and Terasmae (1960), St. Lawrence Lowlands without the presence of the
appears to demonstrate clearly the existence of a marinewaters in the immediatevicinity. Gadd (1960,
glacial lake in the Cornwall areaof Ontario (fig. 2) p. 12) reports that there is „no evidence in any area
prior to the marine invasion ofthe samearea. These studied by the writer, of marine or brackish water
last two authors may well be quoted at this point: „a deposition during the St. Pierre interval".Theglacial
large body ofwater was confluentin the Champlain topography existing during the St. Pierre interval
and St. Lawrence Lowlands with shore lines at pre- after the retreat ofthe ice, could not have been very
sent levelof725'inCovey Hillarea"(Mac Clintockand different from the topography emerging at the final
Terasmae, 1960, p. 237—238). All the evidence déglaciation of the St. Lawrence Lowlands, both
appears to show that not only St. Hilaire Mtn. was permitting fresh water deposition and subaerial ero-
Fig.3.Possible late-glacial lake stagesin theSt.Lawrance Lowlands, justprior tothemarine invasion.
98 Lasalle: Late Quaternary vegetation and glacial history
sion. In the second case however, the déglaciation in
theeastern partwasmorecompleteallowing themarine
waters to penetrate far inland.
4. THERELATIONBETWEENTHEMORAINESAND
THE VARIOUS GLACIAL LAKES
It can be then accepted as a fact that the ice must
have been comparatively thin in the Richelieu-Lake
Champlain Valley some 12,500yearsago(G.S.C.-419).
Indeed, it couldnot havebeen more than 400 meters
thick (ifwe remove the marinesediments), otherwise
the St. Hilaire bog could not have been open to
organic sedimentationat that time (cf. fig. 5). As the
Fig. 4 StratigraphyoftheTwoCreekstypesection;sketchedafterThwaitesandBertrand, 1957,p.857
hypothesis is set up above, and if the icewas calving (nottoscale)
in part in a re-entrant in the ice in the Champlain
Valley, this wouldexplain why we do not apparently
theTwo-Creeks forest bed in the Great Lakes area?
hMaoveunantyYatrmacaeskoaf.thTehDeruarmeamownasdvtihlleenmoocrcauinpieedwebsyt oaf Itappears to be fairlyand reasonably wellestablished
that the retreat after the Mankatoice readvance and
glacial lakeas shown on the map (fig. 3). the improvement ofthe climatethat is in part corre-
The various phases ofthe history can thenbe recon-
lative with it must have been initiatedconsiderably
structed as follows: At the time ofthe deposition of
beforethe growing ofthe Two-Creeks forest bed and
the Highlands Front Moraine System (Gadd, 1964)
itskilling by the risingofthe water ofLake Michigan.
the ice was calving into a glacial lake in the Cham-
(Thwaites andBertrand, 1957; Broeckerand Farrand,
plain Valley; ifnot all the time, this was at least true
1963). Indeed, as will be seen below, the growing of
during the retreat ofthe ice from thatposition. Lake
Iroquois was already inexistencein the Lake Ontario the Two-Creeks forestitselfmay have been oneofthe
area as suggested by Mason (1960), and as shown on several events that can be expected during a short
warm episode ofthis sort.
the map (fig. 3), when the seawater proceeded from
the east to meet the fresh water in the west; thereby As reported by Jelgersma (1962) and as quoted by
a large area had already been deglaciated, including Broecker and Farrand (1963, p. 799), the Mankato
part ofthe Ottawa Valley. The marinewaters never ice must have been in „full retreat" by 12,500 years
tained access to the Lake Ontario basin because it ago (W. 824, 12,650±350 years B.P.). As this date
ad already been elevatedabovesea levelby isostatic was obtained from the bottom sediments ofa kettle
hole well inside the Mankato drift border, its strati-
rebound following its earlier déglaciation as dates in
table 2 indicate and also probably because ofthe low graphic position does not appear in doubt(Jelgersma,
sea level ofthetime (see alsoKarrow et al., 1961). 1962, p. 523 and 526).
It canbeaddedherethatthe large body ofgravel and Hence, the improvement of the climate that was
sand south of St. Hilaire (plate 3, fig. 2) may have cthoendsuhcotrieveotfoLtahkeegMroiwchiniggaonfisthseosmperuwcheatfooreldsetratlhoangn
been inpartoutwash deposited ina glacial lake south
the time of the growing of the forest itself (v. d.
ofthemountain(or ice-contact drift),beforebeing re-
Hammen, 1951). This conclusion is warranted also
worked by the waters ofthe Champlain Sea.
by the stratigraphy of the Two-Creeks type section
itself, as reported by Thwaites and Bertrand (1957).
5. THETWO-CREEKS,VALDERS,AND THECHAM- A sketch (not to scale) ofthis stratigraphy has been
PLAIN SEA
prepared from the original paper of Thwaites and
Thepresentauthorwas tempted tonamethe following Bertrand (1957) and appears in fig. 4.
short and speculative discussion „The Two-Creeks The poorly preserved varved clays underlying the
problem". However, is there a Two-Creeks problem forest bed represent an unknown number of years,
any more thanthere is aValdersproblem? The solu- and were deposited during the retreat oftheMankato
tion lies only in revising the correlationas the infor- ice froma position further tothe south, whilethesand
mation is brought to light. above the forested bed was deposited by advancing
The age of the Two-Creeks forest bed appears to Valders ice, the forest at Two-Creeks having been
have been settled by Broecker and Farrand (1963), drownedas a consequenceofthe closing of the strait
as being ofthe order of years B. P. (p. of Mackinac in northern Michigan by the Valders
796, average ofseveral dates). A slightly older age is ice (according to Thwaites and Bertrand, 1957).
quoted by Broecker and Kaufman (1965, p. 556), The material that constitutes the Two-Creeks forest
viz. ll,900±100yearsB.P. bed proper has been studied by many authors. The
However the main problem that remains to us for latest ofthose studies is by West (1961). The macro-
the area thatwe are concerned with is the following: fossils as reported by Thwaites and Bertrand (1957,
Is there in the eastern part of North America any p. 586) indicate the presence of abundant Picea
equivalent of the Valders readvance that overrode mariana,andalsothat atleastone otherspecies ofPicea.
Historical geology and geomorphology 99
However, Picea marianaappears to be the mostabun- The first two dates appear certainly to come from
dant.The presenceofbalsam fir (Abies) is alsorepor- materialcorrelative withthe Two-Creeks, and related
ted but above the main forest bed. This position of to thesame area(Broecker and Farrand, 1963,p.799).
Abies above the main forest bed appears to suggest The third one appears to have a more controversial
that some species of that genus may prefer cooler stratigraphie position, and its correlation with the
temperature than the spruce species mentioned. The warm episode of the Two-Creeks can be only very
microfossil study of deposits presumably correlative tentative.
of the Two-Creeks forest bed made by West (1961) In the present state ofknowledge, the terms Valders
reports a greater abundanceofpollen grains ofPicea and Two-Creeks should not be used outside their
glauca¡thanofpollen of.Piceamariana.As theidentificat- type locality (Flint, 1963) or they should be used
ion ofthepollen species is based on size, West (1961, in a very informal way, as van der Hammen and
p. 767) has only suggested that the identificationof Vogel (1966) have suggested by including both
the species may have validity. The presence of the Boiling and Allerod in a Two-Creeks episode „sensu
pollen ofAbies is not reported by West (p. 771). lato". They feel there isno good moraineasa logical
The study of the tree rings ofthe same material or candidatefor correlation with the EarlierDryas.
material correlative with it (Thwaites and Bertrand, Furthermore, another possibility must be examined,
1957, p. 856) indicates that they were thinner in the especially ifone feels that the evidence for a deterio-
last twenty years of the life ofthe trees, indicating ration of the climate indicated by the microfossil
less than optimal conditions for growth. The maxi- assemblage oftheTwo-Creeks forest bed is notstrong
mum number of tree rings counted, as reported by enough. Indeed, the sand above the Two-Creeks
Thwaites and Bertrand (1957, p. 856), is 142. forestmay indicate a shallowwater deposit; the forest
West (1961) has also given some indicationofa dete- bed may then have been buried under a sort ofdel-
riorationofthe climateinhisstudy ofthe pollenassem- taic or other shore deposit during a low water stage.
blage of the Two-Creeks bed. A large percentage of This idea has come in the course of discussions with
herb pollen is very clearly shown in the upper part Dr. van der Hammenand with Dr. Saskia Jelgersma
ofhis pollen diagram (p. 771). whohas visited the site (see also Jelgersma, 1966). In
Hence, the evidence brought together above from thatcase, the ageofthe Two-Creeksforest bedmaybe
published sources indicate that the improvement of considered as fitting somewhere in the middle ofan
the climate that is conducive to the retreat of the interstadial. Consequently, the lowest possible strati-
Manicatoice is somewhatolderthan 12,000 yearsago graphic position ofthe Valders readvance (maximum
by an amount that may be of the order of several possible age) is delimited by the age of the Two-
hundredyears. It is certainly older than the dates of Creeks bed (sensu stricto, 11,900i 100 years B. P.,
the Two-Creeks forest bed as reported by Broecker Broecker and Kaufman, 1965), but its highest stra-
and Farrand (1963) and Broecker and Kaufman tigraphie position (minimum age) remains in doubt
(1965) and as these authors (Broecker and Farrand, (see fig. 3).
1963) have argued themselves. As suggested elsewhere in this paper and with the
On the other hand, the deteriorationofthe climate above discussion in mind, it would appear that the
that ultimately brought about the killing ofthe trees Highland Front Moraine system may correspond to
that grew atTwo-Creeks is somewhatolder than the the Earlier Dryas of van der Hammen and Vogel
dateofthekilling ofthe trees itselfand is presumably (1966) asthis system mustbeolder than themaximum
inlargepart contemporaneouswith theirliving. These age obtained on shells ofthe Champlain Sea (Gadd,
trees appear to have been growing in the last spell of 1964). The general and rapid déglaciation ofthe St.
the climatic improvement that preceded the Valders Lawrence Lowlands may correspond to the warmer
and in part during the deterioration ofthe climate. Allerodepisode ofthe European workers.
Hence, theconclusionsofBroeckerandFarrand(1963) Broecker, Ewing and Heezen (1960) have discussed
that the Two-Creeks forest bed (or its correlative) the evidence brought about by C14 dating of North
grewduring a climatic improvement equivalent tothe American material (and also other material), for a
Boiling appears to be warranted. The difference in general improvement of the climate around 11,000
age may be only apparent and due to the type of years ago. However they used a certain number of
sediments used for dating. In Northern Europe, lake datesthat fallin therange between 11,000and 11,500
sediments have been used for dating, in which the years ago to buildup theirargument; hence the im-
change ofvegetation type proper reflectsmore directly provement ofthe climate that they are dealing with
achange in climate(Broecker and Farrand, 1962). must have started somewhat before 11,500 years ago,
Other peat dates corresponding approximately with and culminated around 11,000 years ago, as the
Boiling datesreported by BroeckerandFarrand (1963, Allerad episode. The St. Narcisse readvance, which
p. 799) and Fries et al. (1961, p. 691) deserve to be overrode the Champlain Sea sediments in the Trois-
mentionedhere: Rivièresarea(Gadd and Karrow, 1959) may then be
correlativeoftheYoungerDryas.As fortheChamplain
W—641, 12,000±500 yrs. B. P. Sea itself, it is „definitely" younger than the Two-
W—762, 12,200±250 yrs. B. P. Creeks „sensu stricto" (Broecker and Farrand, 1963,
W—354, 12,030±200 yrs. B. P. p. 801).
100 Lasalle: Late Quaternary vegetation and glacial history
In the Southwestern UnitedStates, Martinand Meh- van der Hammen is the following: the deterioration
ringer (1965, p. 440) appearto be skeptical about the of the climate caused by the incoming cold phase
existence, in the pollen diagrams ofthat area, ofthe killed forest trees, at least some of them, so that they
evidence forthe oscillations observed elsewhere inthe became an easy prey forfire. This is a very plausible
temperate regions. However there appears to be a interpretation because climatic changes are thought
striking correspondence in the datesof the charcoal to occur over wideareas and that actually this pheno-
horizon (A-378, 10,940±100 yearsB.P.) ofthe Lehner menon (presence ofcharcoal in late-glacial sediments)
Mammothsite as reported by Mehringer and Haynes has been observed in widely separated places. Some
(1965, p. 20), and various other datesmade on char- charcoal isreported in thepresent work in the pollen
coal found elsewhere in the world and reported by diagramofSt. Hilaireand St. Bruno. Its quantitative
van der Hammen ( 1965). For example, a charcoal expression, however, must be interpreted cautiously.
layer foundin the Mondoñedoformationat the edge Whether the charcoal horizon of the Lehner site
ofthe Sabana de Bogotá has yielded agesof 10,840± deserves the same ecological interpretation as the
110 years B. P. and 10,760±160 years B. P. (GrN- otheroccurrences is a problem in itself, as the expec-
4209 and GrN-4210 respectively; van der Hammen, tancy offorest fire in the southwest (arid?) may be
1965). greater than in the northernmore temperate regions
Datesfor a charcoal horizon in northwesternEurope ofthe world. However the coincidenceof the dates is
have yielded ages corresponding to the end of the certainly positive and cannot be ruled out, especially
Allerodperiod. Theyarereported byvan derHammen ifone agrees with the idea that deteriorationsof the
(1965) as follows: 10,880±160 years B. P. (Y-139/2) climate (ice readvances in higher latitudes) are corre-
and 10,795 years B.P. (GrN-937). lative with a more humid climate (higher precipi-
As these agesappearto datea change from awarmto tation) in the lowerlatitudes(Maarleveld andvan der
a cold phase, the ecological interpretation made by Hammen, 1959;Broecker and Kaufman, 1965).
CHAPTER III
ABSOLUTE CHRONOLOGY
Before discussing sea level changes it appears oppor- Only marine shells and lake sediments (gyttja) will
tune to say a few words here about the C14 dating be considered here.
principles and the reliability ofthe results.
The content of this paragraph has arisen from dis- a. Marine shells
cussions with Dr. J. C. Vogel during a short visit to Dates obtained from marine shells are considered
the C14laboratoriesin Groningen andhas beenwritten today as fairly reliable, especially if the shells dated
with his collaboration. It is mostly a resumé of the areshallowwaterforms(Mac ClintockandTerasmae,
literature, and partofit isbased onDr. Vogel's paper: 1960,p. 239). TheC14/C12ratioofthesemollusksappears
„Use ofCarbonisotopesingroundwaterstudies"(1963). tobethesameas thatofcontemporaneouswood; these
Though Dr. Vogel has critically read this paragraph, conclusions are based on the measurement ofC14/C12
the application of the principles to the particular ratio ofliving mollusks in bodiesofwater comparable
situation dealt with here or any other error remains to a certain extent with the ancient ones (Broecker
with the author. and Orr, 1958), andalso ongeneraor species thatare
The C14 dating method itself relies on three basic the same as the fossil ones.
assumptions: However, though we know now the approximate
1) Theorganic matter,whenlivinghadaC14/C12ratio salinity ofthe ancientwater bodies at various phases
thatis approximately the same as surrounding atmos- of their existence (Elson and Elson, 1959), there is
phere, where this ratio isthe result ofanequilibrium still a discrepancy between those and the model stu-
between the bombardment of nitrogen atoms by died;forexample, the Champlain Seahasnocounter-
cosmic-ray-produced neutrons and radioactive decay. part in existence today, that would be comparable
2) The C14/C12atmospheric ratiois the same today as to itasfar aslocation, extentor depth.
at the timethe organic matter dated was living. Hence, the dates obtained on marine shells of the
3) The C14/C12 ratio is not changed by any other Champlain Seamaystill bein error forthreereasons:
means than thatofradioactive decay ofC14 after the 1) Surface waters around Spitzbergen where glaciers
burial or fossilization of the organic matter. are still existing, have yielded ages of ± 400 years
With those assumptions in mind, it can be seen that (Berglund, 1964, p. 7). This is due to upwelling of
C14 datingcannot beaccepted blindly ifresultsappear waterfrom greater depth. However, thismay not be
anomalous.Especially thethird assumption is depend- serious in a body of water like the Champlain Sea,
ent of the nature of the material dated and is the that had a depthofthe orderof200 meters orless.
reason for differencesin the reliability of the results. 2) Admixtureoffresh waters with the marinewaters