Chronology and stratigraphy of the Magdalen Islands archipelago from the last glaciation to the

25 Pages·2016·10.65 MB·English

bs_bs_banner
Chronology and stratigraphy of the Magdalen Islands archipelago from
the last glaciation to the early Holocene: new insights into the glacial
and sea-level history of eastern Canada
(cid:1) (cid:1)
AUDREY M. REMILLARD, GUILLAUME ST-ONGE, PASCAL BERNATCHEZ, BERNARD HETU, JAN-PIETER BUYLAERT,
ANDREW S. MURRAYAND BENOIT VIGNEAULT
R(cid:1)emillard, A. M., St-Onge, G., Bernatchez, P., H(cid:1)etu, B., Buylaert, J.-P., Murray, A. S. & Vigneault, B.:
ChronologyandstratigraphyoftheMagdalenIslandsarchipelagofromthelastglaciationtotheearlyHolocene:
newinsightsintotheglacialandsea-levelhistoryofeasternCanada.Boreas.10.1111/bor.12179.ISSN0300-9483.
TheMagdalenIslands(Qu(cid:1)ebec,Canada)areakeylocationforunravellingtheglacialandsea-levelhistoryofthe
MaritimeProvincesofeasternCanada.Althoughmanysedimentarysequenceshavebeendescribedinthelitera-
ture,absoluteagesarelacking,impedinganaccurateinterpretationofthedepositsandtheestablishmentofapre-
cisechronologicalframework.Thisstudyprovidesadetaileddescriptionof21stratigraphicalsequenceslocated
throughoutthe archipelago, aswell asthe first comprehensiveluminescence chronology fromthe Last Glacial
Maximum(LGM)toc.10ka.Inadditiontothefivesamplescollectedforagecontrolpurposes,34luminescence
sampleswere taken from 17 different sites in glacial, periglacial and coastal deposits. The stratigraphical and
chronologicaldatarevealthattheislandswereatthecrossroadsoftwoicecapsduringtheLGM;thesouthern
islandswereglaciatedbytheEscuminacicecaplocatedinthewesternGulfofSt.Lawrencewhereasthenorthern
archipelagowasglaciatedattheendoftheLGMbyaniceflowfromNewfoundland.Theglacialdepositcovering
thenorthernMagdalenIslandswasassociatedwiththeNewfoundlandicecap;hereitisnamedtheGrande-Entr(cid:1)ee
tillandisdatedtoc.20ka.OSLagesbetweenc.23and17kaacquiredfromcryopedimentandcoastaldeposits
onthesouthernislandsindicatethatthispartofthearchipelagowasdeglaciatedshortlyaftertheLGMandwas
affectedbyahighsealevelandperiglacialprocesses.Around15ka,theentirearchipelagowasdeglaciatedand
partiallysubmergeduntilc.10ka.Thisdatasetisthefirstmajorcontributiontoadetailedchronologyofthe
MagdalenIslandsandconstitutesthefirststeptowardsinterpretingtheglacialandsea-levelhistoryofthecentral
area of the Gulf of St. Lawrence; this new understanding will provide input to regional marine and glacial
modelling.
AudreyM.R(cid:1)emillard([email protected])andGuillaumeSt-Onge,Institutdessciencesdelamerde
Rimouski(ISMER)&GEOTOP,Universit(cid:1)eduQu(cid:1)ebec(cid:3)aRimouski,310all(cid:1)eedesUrsulines,Rimouski,QCG5L
3A1,Canada;PascalBernatchezandBernardH(cid:1)etu,D(cid:1)epartementdebiologie, chimie etg(cid:1)eographie&Centre for
Northern Studies (CEN), Universit(cid:1)e du Qu(cid:1)ebec (cid:3)a Rimouski, 300 all(cid:1)ee des Ursulines, Rimouski, QC, G5L 3A1,
Canada;Jan-PieterBuylaertandAndrewS.Murray,NordicLaboratoryforLuminescenceDating,Departmentof
Geoscience,AarhusUniversity,RisøCampus,DK-4000Roskilde,Denmark;Jan-PieterBuylaert,CentreforNuclear
Technologies,TechnicalUniversityofDenmark,RisøCampus,DK-4000Roskilde,Denmark;BenoitVigneault,Min-
ist(cid:3)ere du D(cid:1)eveloppement durable, Environnement et Lutte contre les changements climatiques, 675 boul. Ren(cid:1)e-
L(cid:1)evesqueEst,Qu(cid:1)ebec,QCG1R5V7,Canada;received3rdAugust2015,accepted18thFebruary2016.
The glacial history of the Maritime Provinces of east- glaciation, either from local glaciers or the LIS,
ern Canada has been debated throughout the 20th whereas others argued that, during the last glaciation,
century. Two schools of thoughts have developed con- the archipelago was ice-free (e.g. Richardson 1881;
cerning the nature of the last glaciation in the area Chalmers 1895; Clarke 1911; Goldthwait 1915; Cole-
and these have become known as the ‘maximum’ and man 1919; Alcock 1941; Hamelin 1959; Prest et al.
the ‘minimum’ conceptual models (e.g. Grant 1989; 1976; Dredge & Grant 1987; Parent & Dubois 1988;
Stea et al. 1998, 2011). These models either argued Dredge et al. 1992). More recently, regional syntheses
that the Laurentide Ice Sheet (LIS) entirely covered have come to what is generally accepted in the litera-
the Maritimes to the end of the continental shelf (e.g. ture as the Appalachian Glacier Complex (AGC; e.g.
Dyke et al. 2002), or that local independent glaciers Stea et al. 1998, 2011; Josenhans & Lehman 1999;
were located mainly on land with ice margins ending Stea 2004; Shaw et al. 2006; Josenhans 2007). The
just offshore (e.g. Dyke & Prest 1987). The question AGC suggests that local glaciers, mostly located on
ofwhether theMagdalen Islands,locatedinthecentre the Maritime Provinces, were coalescing with each
of the Gulf of St. Lawrence (Fig. 1), were glaciated other and with the LIS that formed an ice stream in
during the Last Glacial Maximum (LGM) is key to the Laurentian Channel. This model results in a com-
resolving this debate (e.g. Grant 1989). Several studies plex sea-level history in the Gulf of St. Lawrence that
have focused on the palaeogeographical history of the is variable both regionally and locally (e.g. Quinlan &
islands but these have only served to strengthen the Beaumont 1981; Shaw et al. 2002; Bell et al. 2005).
dispute between the maximum and minimum models, Although the AGC is generally accepted, this concep-
because some of these studies recognized evidence of tual model extrapolates from several studies under-
DOI10.1111/bor.12179 ©2016CollegiumBoreas.PublishedbyJohnWiley&SonsLtd
2 AudreyM.R(cid:1)emillardet al. BOREAS
taken throughout the Maritimes; little is actually composedofquartzandisgenerallyporousandhighly
known of the glacial history in the centre of the Gulf. friable (Brisebois 1981). On the southern islands, the
Glacial, periglacial and marine deposits have been bedrock is overlain discontinuously by between 0.5
described in the literature (e.g. Prest et al. 1976; and 10 m thick Quaternary sediments (e.g. Dredge
Dredge et al. 1992; Paquet 1989; Vigneault 2012; et al.1992;R(cid:1)emillardet al.2013).Inthenorthernpart
R(cid:1)emillard et al. 2013, 2015a), but the Quaternary his- ofthearchipelago(thePointe-aux-LoupsandGrande-
tory of the islands remains poorly understood mostly Entr(cid:1)ee Islands), sedimentary bodies can be ~20 m
because of the lack of a robust chronological frame- thick(e.g.Dredgeet al.1992;Vigneault2012).
work. Nevertheless the archipelago offers a strategic
terrestrial record; it is likely to have experienced and Methods
recorded the last glaciation(s) and the inherent sea-
level and periglacialchangesin a keyarea. Sites
The oldest known record on the Magdalen Islands
is located on Havre-Aubert Island and corresponds to A total of 21 sedimentary outcrops mainly located in
a lagoonal marine deposit covering awoody peat. The coastal cliffs have been described in detail (Fig. 1).
wood provided U/Th Marine Isotopic Stage (MIS) 5 SevenarelocatedinthenorthernpartoftheMagdalen
ages of 106.4(cid:1)8.2 ka (UQT – 183), 101.7(cid:1)15.6 ka Islands (Old-Harry, Sandcove–Seacow, Bassin-aux-
(UQT – 182) and 89(cid:1)7.6 ka (UQT – 184). Thewoody Hu^ıtres west and east, Bluff-east, Pointe-aux-Loups
peat and the lagoon deposit are overlain by a littoral north and south), two on Havre-aux-Maisons Island
sandyandgravelridgeformedinacontextofsea-level (Airport and HAM), seven on Cap-aux-Meules Island
rise,whichhasitselfbeentruncatedbyalocaltillasso- (Fatima, Facterie, SAQ, Ars(cid:3)ene, Grader, Galet-Plat
ciated with the MIS 4 (Dredge et al. 1992). More and Gros-Cap) and five on Havre-Aubert Island
recently, based on a limited number of radiocarbon (ACW, AC-Lighthouse, AP-DEM, AP-CAM and
ages, R(cid:1)emillard et al. (2013, 2015a) suggested that the Clermont). Finally, two sites were studied for chrono-
southern part of the Magdalen Islands was glaciated logicalcontrolpurposesonly:Plaisance,locatedonthe
by the Escuminac icecap that flowed towards the tombolo connecting Cap-aux-Meules and Havre-
southeast at some point during MIS 2, and was Aubert Islands, and ACE, located on the southwest
affectedbytwomarinetransgressionsandtwoperigla- coastofHavre-AubertIsland.
cialphasesbothbeforeandaftertheMIS2glaciation.
Unfortunately, a similar chronological framework is Sedimentologicalanalyses
lackingforotherareas;thedepositsofthearchipelago
Ten clast fabrics (n = 30, 49 or 50) were measured in
require the construction of an absolute chronology
the field using A-axiswith an axial ratio of >1.5; eight
before an understanding of the relationships amongst
of these were re-analysed from the study of Vigneault
theglacial,periglacialandmarinephasescanbedevel-
(2012). Clast fabrics were all plotted as contoured
oped.
equal-area stereonets on the lower hemispheres using
Inthisstudy,weaimedto(i)describeandcharacter-
ize the sedimentology of 21 outcrops observed STEREO32 software and examined using eigenvector
analysis(Mark1973)(Table 1).Grain-sizedistribution
throughout the Magdalen Islands, (ii) improve the
of matrices was determined with a Beckman-Coulter
chronological framework by adding new OSL ages,
particle size analyser LS 13 320 (0.04–2000 lm) on 51
and(iii)increasetheunderstandingofthepalaeogeog-
raphy and sea-level history of a key area of eastern disaggregated samples and processed with GRADISTAT
v.8.0softwareusingthelogarithmicmethodofFolk&
Canada,fromtheLGMtoc. 10ka.
Ward (1957) (Blott & Pye 2001; Table S1). Altitudes
were measured in situ using a Trimble RTK D-GPS
Geologicalandgeomorphologicalsettings ((cid:1)0.015 mverticaluncertainty).
The Magdalen Islands lie in the shallow waters of the
Magdalen Shelf (<100 m) near the centre of the Gulf Radiocarbondating
of St. Lawrence (47°N) (Fig. 1). Six of the seven
Samples of organic material were collected at the AC-
islands are connected to each other by Holocene
barrier beaches (tombolos) (R(cid:1)emillard et al. 2015b). Lighthouse and Clermont sites, both located on
The islands are built on Carboniferous–Visean vol- Havre-Aubert Island. The samples were prepared for
radiocarbon dating at the radiochronology laboratory
canicandsedimentarysubstrata(Brisebois1981;Giles
of Universit(cid:1)e Laval and analysed at the Keck Carbon
2008).Thegeomorphologyoftheislandsisdominated
Cycle AMS Facility at the University of California in
by basaltic conical hills surrounded by sandstone and
Irvine.Theconventional14Cageswerecalibratedusing
shaleplatformsslightlyinclinedtowardsthesea,which
areinterpretedascryopedimentsurfaces(Laverdi(cid:3)ere& the CALIB 7.1 program with the INTCAL13 calibra-
tiondataset(Reimeret al.2013).
Guimont1974; Paquet1989). The sandstoneismainly
BOREAS ChronologyandstratigraphyoftheMagdalenIslands,Canada 3
Fig.1. LocationoftheMagdalenIslandsineasternCanadaandtheGulfofSt.Lawrence.Samplesitesareshownasblackpushpins.
OSL81,OSL83,OSL84andOSL92).Allsampleswere
Optically-StimulatedLuminescence(OSL)dating
prepared under subdued orange light. The outer
Sampling, sample preparation and measurement proto- ~5 cm ends of the cylinder samples were used for
cols.– Thirty-nine OSL samples were collected water-content determination and dose-rate analysis.
throughout the archipelago from 19 sites (Tables 2, Samples were wet-sieved and the 180–250 lm (90–
S2). Generally,sedimentsweresampled byhammering 180 lm fraction for OSL08) fraction was etched with
opaque plastic cylinders of 5 cm diameter and 30 cm 10% HCl, 10% H O and 10% HF in the usual man-
2 2
in length into a homogeneous part of the deposit. For ner. Heavy liquid separation (2.58 g ml(cid:3)1) was then
the pebbly unitswhere the absence of sandy lenses did used to separate quartz from K-rich feldspar grains.
not allow this approach, sampleswere collected in the Finally, the quartz-rich extract was etched using con-
dark (at night) in opaque plastic bags (OSL80, centrated HF (40%) for 1 h to remove any remaining
4 AudreyM.R(cid:1)emillardet al. BOREAS
Table1. Clastfabricdatafromthisstudy.S=eigenvalues;K=shapeparameter;C=strengthparameter.
ClastfabricID n Meanvectors Eigenvalues
Azimuth Plunge S1 S2 S3 K C
PAL-North CF01 49 96 26 0.622 0.246 0.132 1.496 1.546
PAL-South CF02 50 77 20 0.558 0.322 0.120 0.560 1.537
Old-Harry CF03 50 194 13 0.675 0.253 0.071 0.773 2.248
Old-Harry CF04 50 21 2 0.705 0.186 0.109 2.499 1.865
BAH-West CF05 50 61 28 0.618 0.256 0.127 1.251 1.585
BAH-East CF06 49 87 20 0.663 0.215 0.123 2.016 1.688
BAH-East CF07 50 49 37 0.735 0.160 0.105 3.642 1.946
Grader CF08 50 22 14 0.635 0.285 0.080 0.634 2.071
Grader CF09 49 15 9 0.764 0.182 0.055 1.195 2.636
Gros-Cap CF10 30 54 12 0.665 0.236 0.099 1.197 1.902
feldspar and the outer alpha-irradiated layer from the Unfortunately, the 238U concentrations arenot known
quartzgrains. with sufficient precision to allow a discussion of the
All measurements were carried out using Risø TL/ stateofequilibriumbetween226Raand238Uonasam-
OSL readers (model DA-20) each equipped with blue ple-by-sample basis. However, it is interesting to note
LEDs (470 nm, ~80 mW cm(cid:3)2), infrared (IR) LEDs that the average 226Ra/238U ratio for all samples is
(870 nm, ~135 mW cm(cid:3)2) and with a calibrated 0.81(cid:1)0.09(n = 39),suggestingthattheremaybesome
90Sr/90Y beta source (Bøtter-Jensen et al. 2010). systematic disequilibrium in these samples. The most
Quartz OSL was detected through a 7.5-mm Schott obviousexplanationforalackof226Rainthisenviron-
U-340 (UV) filter. The quartz extracts were mounted mentismobilizationduetothehighsalinityatthetime
on ~8-mm-diameter stainless steel discs using ‘Silkos- of deposition (Webster et al. 1995). Although most of
pray’ silicone oil as a fixing agent. A single-aliquot these samples were collected from above sea level it is
regenerative-dose (SAR) protocol (Murray & Wintle speculated that the considerable accumulation of salt
2000,2003)wasusedforallequivalentdose(D )deter- sprayon the islands and subsequent down washing by
e
minations. Prior to OSL measurement, the purity of rain water is sufficient to continually remove some of
thequartzextractswasconfirmedbyanOSLIRdeple- the226Ra,maintainingthissmalldisequilibrium.How-
tion test (Duller 2003);allsampleshadOSL IR deple- ever, for these samples the U-series only contribute
tion ratios within 10% of unity, indicating that there ~8%tothetotaldoserateandsoasmallU-seriesdise-
was no significant contribution from feldsparorother quilibrium is not of significance. Dry dose rates were
IR-sensitive components to the blue-stimulated OSL calculated from the radionuclide concentrations using
signals. Quartz aliquots were stimulated at 125 °C the conversion factors given in Gu(cid:1)erin et al. (2011).
usingbluelight(90%power)for40 s.QuartzD values Field water contents were measuredwhen the samples
e
were calculated using the first 0.32 s of the signal and were first opened in the laboratory. Unfortunately,
a background based on the following 0.32–0.64 s to material was insufficient for saturation water content
minimizeanypossiblecontributionofnon-fastcompo- measurements on 14 out of the 39 samples. To over-
nents (Cunningham & Wallinga 2010). The sample D come this problem, we used the saturation water con-
e
usedforthefinalagedeterminationcorrespondstothe tent mean value of the other 25 available samples
unweighted average of all accepted aliquots (between (numbersinitalicsinTableS2).Manyofthesesamples
21 and 48 aliquots per sample). As part of the investi- were deposited under water but were, compared with
gation of the reliability of the quartz OSL ages, the their subsequent burial time, uplifted shortly after
feldspar extracts were also measured using infrared deposition. Other sampleswere deposited sub-aerially.
stimulation at elevated temperature and a SAR proto- Allweresubsequentlywelldrained,andsoforallsam-
col. Details of these measurements are given in ples (except the modern analogues), a lifetime water
AppendixS1. content of 50% of the saturation value was assumed
withanabsoluteuncertaintyof(cid:1)6%,e.g.forasample
Radionuclide analysis and dosimetry.– Approximately with a saturation water content of 40% we adopted a
250–300 g of material was dried, ground and ignited lifetimeaveragewatercontentof20(cid:1)6%.Forthethree
(24 h at 450 °C) and subsequently cast in wax in a modern analogues where the likely fraction of satura-
fixedcup-shapedgeometry.After3weeksofstorageto tion could be estimated more accurately, we adopted
let 222Rn reach equilibrium with its parent 226Ra, the full saturation for sample OSL78 (subtidal) and 20%
cups were counted on a high-resolution gamma spec- of saturation for samples OSL87 and OSL88 (beach).
trometer for at least 24 h following Murray et al. These assumed lifetime water contents were used to
(1987). The radionuclide activities, water contents, dry correct the calculated dry dose rates as described
dose rates and total dose rates are given in Table S2. by Aitken (1985). Finally a cosmic ray dose rate was
BOREAS ChronologyandstratigraphyoftheMagdalenIslands,Canada 5
e
‘’eofndis- Highlyprobabl XXXXXXX–– X––– XXX–– XXXXXXXXXXXXXX– XXX
agis d?
averages ache nt
ulatedisthequartzOSL Wellble Confide X XXXX –– X––– XX–– XXXXXX X X XXX– X X
uivalentdosetabreliabilityofthe (cid:1)rAge1(ka) (cid:1)414(cid:1)444(cid:1)11.60.7(cid:1)383(cid:1)9.80.6(cid:1)11.10.8(cid:1)11.70.9(cid:1)17.41.4(cid:1)18.01.4>150>115(cid:1)15.11.1>150>156>110(cid:1)12.61.0(cid:1)11.30.8(cid:1)10.10.6>130>159(cid:1)11.40.9(cid:1)11.00.7(cid:1)323(cid:1)14.81.0(cid:1)23.01.6(cid:1)12.20.9(cid:1)0.140.02(cid:1)10.90.8(cid:1)17.91.4(cid:1)21.11.7(cid:1)14.91.2(cid:1)17.61.6(cid:1)0.050.03(cid:1)0.050.04>174(cid:1)19.81.4(cid:1)19.51.4(cid:1)14.51.1(cid:1)12.80.9
qe
eh 477016311–– 3––– 211–– 31811434832284– 8614
eT n 222322222 3 322 32222222222342 2222
Th.D0
13.92 (cid:3)1)
eferstoFig.isbasedon Totaldoserate(Gyka (cid:1)1.160.06(cid:1)1.270.07(cid:1)1.990.11(cid:1)2.180.12(cid:1)2.130.12(cid:1)1.760.09(cid:1)1.820.10(cid:1)1.370.07(cid:1)1.760.09(cid:1)1.540.09(cid:1)1.990.10(cid:1)1.980.11(cid:1)1.470.08(cid:1)1.410.08(cid:1)2.090.12(cid:1)1.710.10(cid:1)1.550.08(cid:1)1.450.08(cid:1)1.770.10(cid:1)1.450.09(cid:1)1.310.07(cid:1)1.150.06(cid:1)1.60.2(cid:1)1.530.08(cid:1)1.370.07(cid:1)1.400.07(cid:1)1.030.05(cid:1)1.600.05(cid:1)3.70.2(cid:1)3.80.2(cid:1)3.040.17(cid:1)3.80.2(cid:1)1.570.09(cid:1)1.910.11(cid:1)1.320.07(cid:1)1.600.09(cid:1)2.610.15(cid:1)1.770.10(cid:1)1.660.09
re
mbercolumnminimumdos (Gy)De (cid:1)473(cid:1)553(cid:1)23.10.5(cid:1)835(cid:1)21.60.5(cid:1)19.60.9(cid:1)21.31.2(cid:1)23.81.4(cid:1)31.71.8>230>230(cid:1)1.129.8>230>230>230(cid:1)21.71.0(cid:1)17.50.7(cid:1)14.70.5>230>230(cid:1)14.90.7(cid:1)12.70.4(cid:1)36.31.8(cid:1)22.70.8(cid:1)31.61.3(cid:1)17.10.8(cid:1)0.140.02(cid:1)17.40.8(cid:1)663(cid:1)814(cid:1)453(cid:1)675(cid:1)0.080.05(cid:1)0.100.07>230(cid:1)31.71.3(cid:1)512(cid:1)25.81.3(cid:1)21.30.8
nuhe
xT
resultingquartzOSLages.Theindes).Samplesinitalicweresaturated. Long.(W)Elev.Depth(ma.s.l.)(cm) 61.98982914.848061.98982915.640061.995980247061.9729991230061.9729991310061.86515413.515061.86515414.57561.8697521315061.86975212.512561.548475370061.5484751115061.548475167561.4907836100061.4907831510061.48894010.530061.4889401125061.4889401215061.488940135061.7123131810061.7123138130061.793613217561.79361337561.9126954.525061.912695610061.75506099061.49078315.5100(cid:3)61.7936130.35061.501402410061.8752555.55061.87300745061.873500107561.9279175027561.9169404.81061.9169400.52061.4654711130061.46547112.515061.934643505061.71231315.535061.71231318100
dot
nu
D)aealiq N) 517517500810810281281646646753753753937937720720720720185185373373613613967937373772933785198529720720836836770185185
doses(easured Lat.( 47.21947.21947.21947.21347.21347.24047.24047.23847.23847.54547.54547.54547.56847.56847.57047.57047.57047.57047.53147.53147.42347.42347.41447.41447.41647.56847.42347.56147.35447.35447.37547.37647.26947.26947.56947.56947.39347.53147.531
ntm
quivaleof802on. Indexno. –– 1– 23456–– 7––– 8910–– 111213141516– 171819––––––– 2021
h,eoutecti
plelocation,elevation,deptalof798acceptedestimatesReliabilityoftheOSLagess SampleEnvironmentID OSL03BeachOSL04BeachOSL06SubtidalOSL08SubtidalOSL12SubtidalOSL16SubtidalOSL17SubtidalOSL18SubtidalOSL19SubtidalOSL20MarineOSL22MarineOSL23SubtidalOSL26MarineOSL29MarineOSL30MarineOSL31SubtidalOSL32SubtidalOSL33SubtidalOSL34MarineOSL36MarineOSL44SubtidalOSL46SubtidalOSL47SubtidalOSL48SubtidalOSL63SubtidalOSL72SubtidalOSL78SubtidalOSL79SubtidalOSL80BeachOSL81BeachOSL83ColluvialOSL84ColluvialOSL87Beach(dune)OSL88Beach(tidal)OSL89MarineOSL90GlacialOSL92ColluvialOSL96SubtidalOSL97Subtidal
mote
Table2.Saestimates(tcussedinth Site ACEACEACWAC-LightAC-LightAP-DEMAP-DEMAP-CAMAP-CAMBluff-eastBluff-eastBluff-eastSandcoveSandcoveSeacowSeacowSeacowSeacowPAL-northPAL-northAirportAirportFatimaFatimaHAMSandcoveAirportBAH-eastGalet-PlatGalet-PlatSAQGraderPlaisancePlaisanceOld-HarryOld-Harry(cid:3)ArsenePAL-northPAL-north
6 AudreyM.R(cid:1)emillardet al. BOREAS
deposited coastal environments were collected as
A described above at two different sites: OSL78 from a
subtidal environment (Airport site) and OSL87 and
OSL12 OSL88 from a beach (Plaisance site; Table 2). At two
sites(ACEandAC-Lighthouse),luminescenceagesare
9.8 ±0.6 ka
also compared with published radiocarbon ages and
one new AMS 14C age obtained on plant fragments
(Fig. 2, Table 3). Note that the ACE site is not pre-
10.20±0.03 14C cal. ka BP sented in the Results and interpretations section
PS48
because the exposure has already been described and
discussedinR(cid:1)emillardet al.(2013);itisusedhereonly
forchronologicalcontrolpurposes.
Quartz luminescence characteristics. – Figure 3A illus-
trates a typical sensitivity-corrected growth curve with
anaturalOSLstimulationdecaycurve(inset),together
with a decay curve from a quartz calibration standard
(Hansenet al.2015).Thereproducibilityoflaboratory
B
~45-50 14C ka BP measurements using this material is illustrated by the
twomeasurementsofthesensitivity-correctedsignalat
26 Gy. The average of this recycling ratio for all avail-
able measurements is 1.079(cid:1)0.015 (n = 588). The
growth curve also passes very close to the origin; the
average recuperation is 0.09(cid:1)0.08% of the natural sig-
nal (n = 588). The D for this aliquot (Fig. 3A) is
0
115(cid:1)14 Gy,suggestingthatwecanusethismaterialto
estimate D smaller than ~230 Gy (Wintle & Murray
e
2006).
The dependency of the D and the dose recovery
e
OSL04 -44±4ka ratio(Murray&Wintle2003)onthermalpretreatment
for sample OSL03 was tested by varying the preheat
OSL03 -41±4 ka
temperature; the thermal pretreatment employed after
giving the test dose (the ‘cut-heat’) was 40 °C below
the preheat temperature (except for 160 and 180 °C
Fig.2. Comparisonofluminescenceageswithradiocarbonagesfor preheat for which the cut-heat was kept fixed at
agecontrol.A.TheAC-Lighthousesite;OSL12iscomparedwitha 160 °C;Fig. 3B,C).BothD estimatesanddoserecov-
newcalibratedAMS14Cageobtainedonplantfragments.B.OSL03 e
andOSL04arecomparedwithradiocarbonagesalreadypublished ery ratios have a pronounced dependence on preheat
inR(cid:1)emillardetal.(2013). temperature with the dose recovery ratio only
consistent with unity for preheats ≤220 °C (Fig. 3C).
The D values are also relatively insensitive to preheat
e
estimated from the assumed lifetime burial depth, fol- temperatures<220 °C.Becauseofthis,wechoseapre-
lowing the equations given by Prescott & Hutton heat/cut-heatcombinationof200/160 °C.Furtherdose
(1994). recovery measurements (Murray & Wintle 2003) were
made using six aliquots of all samples (except OSL80,
Age control. – In order to test the degree of bleaching OSL83, OSL88 and OSL97 because of lack of mate-
of the coastal deposits, three samples of recently rial) and given doses varying between 4 and 100 Gy
Table3. Listofradiocarbonagesdiscussedinthispaper.
Site LaboratoryID Age(14CaBP(cid:1)1r) Calibratedage((cid:1)2r) Reference
ACE UCIAMS-74416 >46000 n/a R(cid:1)emillardetal.(2013)
ACE UCIAMS-41189 47100(cid:1)2700 n/a R(cid:1)emillardetal.(2013)
ACE UCIAMS-74417 50100(cid:1)3300 n/a R(cid:1)emillardetal.(2013)
ACE UCIAMS-84792 47100(cid:1)2300 n/a R(cid:1)emillardetal.(2013)
ACE UCIAMS-84793 47800(cid:1)2500 n/a R(cid:1)emillardetal.(2013)
AC-Lighthouse UCIAMS-134737 8995(cid:1)25 10197(cid:1)32 Thispaper
Clermont UCIAMS-134729 9430(cid:1)25 10656(cid:1)73 Thispaper
BOREAS ChronologyandstratigraphyoftheMagdalenIslands,Canada 7
Fig.3. A.QuartzOSLSARgrowthcurvefromOSL08(siteAC-Lighthouse).Sensitivity-correctedregeneratedsignalsareshownasfilledcir-
cles,theunfilledcirclerepresentsarepeatpoint(recycling)andtheopentriangletheresponsetozerodose(recuperation).Thesensitivity-cor-
rectednaturalOSLsignalisinterpolatedontothegrowthcurvetogivetheequivalentdose,D (inthiscase,76.5Gy).Insetshowsatypical
e
naturalOSLdecayfromthesamesampletogetherwithadecaycurveofanaliquotofcalibrationquartz(dashed);abackgroundsignalfrom
theendofthestimulationcurvewassubtractedbeforenormalization.B.PreheatplateautestcarriedoutonsampleOSL03;eachpointrepre-
sentstheaverageofthreealiquots.ThedashedlinerepresentstheaverageD overthetemperaturerange160–200°C.C.Doserecoverypla-
e
teautestcarriedoutonsampleOSL03;eachpointrepresentstheaverageofsixaliquots.Thedashedlinerepresentsadoserecoveryratioof
unity.D.Summaryofdoserecoverydataforallaliquots(n=150)of25samplesmeasuredwithapreheatof200°Cfor10sandcut-heatof
160°C.Insetshowsthemeasureddosesplottedagainstthegivendoses.
depending on the approximate D of each sample toryluminescencecharacteristicsandthedoserecovery
e
(Fig. 3D); the average dose recovery ratio is ratiosareindistinguishablefromunity(Fig. 3).
0.974(cid:1)0.011 (n = 150), confirming that the chosen Three sampleswere taken to represent modern ana-
SARprotocolissuitableforthesesamples. logues of the various coastal and subtidal deposits:
OSL78 from a subtidal lagoon deposited in ~50 cm of
Reliability of the OSL ages. – Quartzfrom theCana- water at low tide; OSL88 taken from just below the
dianShieldinmainlandQuebecisnotusuallyconsidered high tide limit on a beach, but deposited underwater;
suitableforluminescencedatingbecauseofthelowyield OSL87, aeolian sediment from an incipient dune just
of sensitive grains (M. Lamothe, pers. comm. 2015). above the high tide limit. Samples OSL87 and OSL88
However, the sediment on and around the Magdalen both giveagesindistinguishablefromzero(30(cid:1)30and
Islands is not derived from the Canadian Shield, but 50(cid:1)40 a) and even the subtidal lagoonal sediment has
ratherfromaCarboniferoussandstoneplatformderived an apparent age of only 160(cid:1)20 a. These quartz ages
from aeolian and fluvial sand (Brisebois 1981). This are very small compared with all those in Table 2,
appearstohaveresultedinlocalquartz-richsedimentfor indicating that all the coastal/subtidal samples can
whichtheOSLsignalisdominatedbyafast-component probably be considered as well bleached. The limited
(e.g.Fig. 3A).TheresultingOSLsignalshavesatisfac- independentagecontrolcanalsobeusedtodiscussthe
8 AudreyM.R(cid:1)emillardet al. BOREAS
completeness of bleaching of some of the samples at In summary, the agreement of the quartz OSL ages
thetimeofdeposition.OSL12hasaquartzOSLageof with modern analogues and radiocarbon age control,
9.8(cid:1)0.6 ka and this sample sits immediately above a the internal consistency and the comparison with dif-
radiocarbon sample of 10.20(cid:1)0.03 cal. ka BP ferential bleaching rates of IR and pIRIR signals
50 150
(Fig. 2A). Figure 2B shows the stratigraphical rela- allows us to conclude that all the samples in this data
tionship between five radiocarbon ages lying between setwereveryprobablywellbleachedatdepositionand
46 and 50 ka (R(cid:1)emillard et al. 2013) and the chronos- arethusreliable.
tratigraphically consistent samples OSL03 (41(cid:1)3 ka)
and OSL04 (43(cid:1)3 ka). The agreement between these
Resultsandinterpretations
OSL ages and the independent age control in both of
thesegroupsstronglysuggeststhatthequartzOSLsig-
Pointe-aux-Loups(PAL)northandsouth
nal in all three of these subtidal/beach OSL samples
was well bleached at the time of deposition. Further- Pointe-aux-Loups (PAL) Island is located on the
more, samples collected in a single unit are almost northwest side of the Magdalen Islands (Fig. 1). On
always consistent with each other (e.g. OSL03 and the western coast of the island, two active cliffs were
OSL04; OSL31, OSL32 and OSL33; OSL44 and described on each side of the wharf: PAL-north and
OSL46; Table 2). The only exceptions are OSL08 and PAL-south.
OSL12,whichweresampledatthetopandthebaseof Sediments of the PAL-north exposure are ~18 m
a single unit and give ages of 38(cid:1)3 and 9.8(cid:1)0.6 ka, thick and ~60 m wide, filling a bedrock depression
respectively. For safety reasons (very steep cliff), this (Fig. 4A). The firstunit (U1) lies directlyon the sand-
outcrop was sampled at two different sections, one at stone bedrock and is composed of imbricated well-
the top and one at the base of the unit; it was not rounded pebbles and boulders, alternating with sand
possible to completely connect these two sections and gravel beds slightly inclined (~3–4°) southwards
stratigraphically. Thus it now seems likely that the (Fig. 4B). These abrupt grain-size variations in U1
sections were misidentified in the field, and that this reflect several energy changes during deposition with
explainstheapparentstratigraphicalinconsistency. periodsofveryhighenergy;somebouldersreachmore
Further evidence for the completeness of bleaching than a metre in diameter. A petrographic count
ofthequartzOSLsignalatdepositionisgivenbycom- (n = 100) carried out in U1 revealed that 66% of the
paring the results with the less-bleachable K-feldspar pebbles and boulders are fossiliferous limestone from
(post-IR) IRSL signals (Murray et al. 2012; Buylaert AnticostiIsland locatedinthenorthoftheGulfofSt.
et al. 2013) (analytical data and summary of feldspar Lawrence, and 30% correspond to the St. Lawrence
results are given in Appendix S1). Feldspar IRSL sig- RiverNorthShoregeology(gneiss,granite,metaquart-
nals are significantly more difficult to bleach than the zite, anorthosite) (Vigneault 2012). With the exception
OSL fast-component signal from quartz used here of a few sandstone pebbles, the composition of U1 is
(Godfrey-Smith et al. 1988; Thomsen et al. 2008; entirely erratic. U1 evolves gradually into a stratified
Fig. S2). This provides an independent test of the deposit (U2) composed of centimetric sandy beds
degree of bleaching ofour quartz. In Fig. S3 we com- alternating with millimetric silty beds inclined gener-
paretheagesderivedfromthemoredifficult tobleach ally towards the southwest (Fig. 4C, D; Table S1);
feldspar signals with those from quartz. As discussed inclination of the beds increases from the base (~3–4°)
in the Appendix S1, the feldspar data generally to the top (>10°) (Vigneault 2012). U2 also contains
confirm the conclusion that our samples are well scattered dropstones, which decrease in number from
bleached, although there remain four samples for the base to the top. The top of U2 is characterized by
which, based onlyon feldspar, we cannot be confident many load structures (balls and pillows, flames) and
of the completeness of bleaching of the quartz OSL deformations(faultsandfolds)(Fig. 4E).Thequartzin
signal prior to deposition. Amongst them, the two U2iscompletelysaturatedandsoonlyaminimumage
samples with the largest IR age overestimate com- of >130 ka can be derived based on the 29D value
50 0
pared to quartz (Fig. S3C) are OSL03 and OSL04; >230 Gy (OSL34 and OSL36; Table 2). The entire U1
both of these samples have quartz D of <60 Gy and andU2sequenceistruncatedbyacompactandpoorly
e
so are well below the ~230 Gy dating limit. Fortu- sorted diamict (U3; Table S1) that appears either as a
nately, OSL03 and OSL04 are the two samples for stone line or as remnant pockets. Figure 4E shows a
which we have independent age control based on remnantpocketinwhichclastfabricmeasurementssug-
radiocarbon (Fig. 2B); the agreement between the gestanE–Waxis(insetFig. 4E;Table 1).Somepebbles
radiocarbonandquartzOSLagesclearlyindicatesthat fromU3arestriatedandhavecontouredshapes.Apet-
the quartz OSL signal from both these samples must rographiccount(n = 50)revealsthattheunitismostly
havebeensufficientlybleachedatdeposition.Thiscali- erratic, composed of80%metaquartzite, gabbro, schist
brates our IR signal for use in testing quartz bleach- andanorthosite,and14%typicalAppalachianlithology
50
ing. (jasper, phyllite, chert) (Vigneault 2012). U3 is overlaid
BOREAS ChronologyandstratigraphyoftheMagdalenIslands,Canada 9
OSL97
A OSL96 B
U4
Abrupt
energy
U3
changes
U1
U2
U1
E W
Sandstone
PS02
C D
U2
Dropstone
Sca(cid:2)ered PS01 U2
dropstones
U1
E F
U4
U3
OSL97 –12.8±0.9 ka
CF01
PS03
Deforma(cid:3)ons
OSL96 –14.5±1.1 ka
(sampling not shown)
Fig.4. RepresentativepicturesofthePAL-northsite.U1=proximalfluvioglacialdepositwithalithologyfromAnticostiIslandandQue-
bec’sNorthShore;U2=stratifiedmarinedepositofmoderateenergy;U3=tillwithAppalachianlithology;U4=stratifiedsubtidalsand.
StarscorrespondtoOSLsamplelocations.Numberedsymbols(e.g.PS01)areparticlesizesamples.A.OverallpictureofthePAL-northsite.
Asscale,apersonof155cmheightstandsintheoutcrop.B.Close-upofU1;imbricatedwell-roundedpebblesandboulders,alternatingwith
sandandgravel,reflectingabruptenergychanges.C.GradualtransitionbetweenU1andU2andpresenceofdropstoneswithinU2.D.Close-
upofU2;stratifieddepositcomposedofcentimetricsandybedsalternatingwithmillimetricsiltybedsandpresenceofdropstones.E.Close-
upofU3whereitappearsasatill;compactandpoorlysorteddiamictcomprisingstriatedandcontoured-shapepebbles.Insetshowstheclast
fabricmeasuredinU3.F.Close-upofU4;stratifiedunitcomposedofwell-sortedandwell-roundedredandwhitesand.Asscale,thediameter
oftheblackcylinderis5cm.Note:OSL96wascollectedatthebaseofU4(notshownonthepicture).
10 AudreyM.R(cid:1)emillardet al. BOREAS
Indentation
A
NE SW
CF02
B C
Aeolian sand
Glacial diamict Aeolian sand
Thrust towards 250o
Fold axis 340o Glacial diamict
Thrust towards 265o
Fold axis 175o
Sandstone
Sandstone
Facing southeast Facing northwest
Fig.5. RepresentativepicturesofthePAL-southsite;atillwithAppalachianlithologyoverliesthefoldedsandstonebedrock.A.Locationof
theoutcroponthecoastalcliff.Thetotalwidthofthepictureisapproximately50m.B.Outcropfacingthesoutheast.C.Outcropfacingthe
northwest.Insetshowstheclastfabricmeasuredinthetill.
by a ~3-m-thick stratified unit (U4) composed of well- tively 250 and 265°) (Fig. 5B, C). A clast fabric mea-
sorted andwell-rounded red andwhite sand (Fig. 4F). sured in the overlying diamict suggests a NE–SWaxis
U4ischaracterizedbyanalternationoffinewhitesand (insetFig. 5C;Table 1).Apetrographiccount(n = 50)
andreddishsilty-sand(TableS1). Thecontactbetween indicates that this unit is mostly allochthonous, with
the different beds exposes micro-scale ripple marks. 76% metaquartzite, gabbro and anorthosite, and 22%
Two samples were collected in U4, OSL96 at the base varied erratic (jasper, granite). The diamict is topped
(+15 m) and OSL97 (+18 m) at the top, giving ages of byrecentaeoliansediment.
14.5(cid:1)1.1 and 12.8(cid:1)0.9 ka, respectively (Fig. 4F;
Table 2). Interpretations.– Owing to the clastic support, the
ThePAL-southexposureislocatedinanindentation imbrication, the very well-rounded shape of pebbles
of the sandstone platform (Fig. 5A), allowing the out- and boulders (regardless of the nature) and the metric
crop tobestudiedontwoopposingsides (Fig. 5B,C). sizeofseveralboulders,U1isinterpretedasaproximal
The top of the sandstone bedrock is folded and over- fluvioglacialunit(highenergy;Dredgeet al.1992).U2
laidbya~80-cm-thickcompactandpoorlysorteddia- is identified as a marine deposit of moderate energy
mict. The fold axes measured on each side of the (shallow water); the rhythmicityof beds, probably due
exposure are consistent with each other (340 and to an alternating energy, might reflect seasonality. The
175°), and reveal a thrust towards the WSW (respec- conformable contact and the gradual decrease in