Quaternary glacial and marine environmental history of northwest Greenland: a review and reappraisal

Quaternary Science Reviews 18 (1999) 373—392

Quaternary glacial and marine environmental history of northwestGreenland: a review and reappraisal

Michael Kelly!,*, Svend Funder", Michael Houmark-nielsen#, Karen Luise Knudsen$,Christian Kronborg$, Jon Landvik%, Lennart Sorby&

! Department of Environmental Science, Lancaster University, Lancaster, LA1 4YQ, UK" Geologisk Museum, K~benhavns Universitet, DK-1350 K~benhavn K, Denmark# Geologisk Institut, K~benhavns Universitet, DK-1350 K~benhavn K, Denmark

$ Geologisk Institut, Aarhus Universitet, DK-8000 Aarhus C, Denmark% Universitetsstudiene pa> Svalbard, N-91750 Longyearbyen, Norway

& Kvarta( rgeologiska avdeling, S-223 62 Lund, Sweden

Abstract

New information on Middle and Late Quaternary deposits of the Thule region, northwest Greenland necessitates the revisionof the chronostratigraphy of the area and allows fuller understanding of the marine environmental changes in Baffin Bay.The stratigraphic record is interpreted as the product of three marine events (Saunders ", Qarmat and Nuna, in decreasingage), during which marine sedimentation occurred on coastal areas now land, and three or four glacial events of decreasing ageand ice cover (Agpat, Narsaarsuk, and Kap Abernathy/Wolstenholme Fjord), to which the marine events were glacio-isostaticallyrelated.

Subarctic Atlantic water reached this northern part of Baffin Bay during all three marine events but the warmest conditions,warmer than the present, occurred in the Qarmat event, correlated with the deep sea Oxygen Isotope Stage 5e. The question of lateStage 5 events is considered. The Nuna event dates to latest Stage 2 and the Holocene, whilst the Saunders " event is of uncertain agebut at least Stage 6. A number of glacial events with advances confined to the fjords are dated to Stage 2 (Wolstenholme Fjord), andlate Stage 6 (Narsaarsuk) and an intermediate advance (Kap Abernathy) is undated in the range of Stage 2—5. Extension of glaciationto the shelf edge is dated to Stage 6 (Agpat event). ( 1999 Elsevier Science Ltd. All rights reserved.

1. Introduction

Changes in the ice cover of the land around northernBaffin Bay during the Quaternary depended not only onthe thermal climate regime but also on the nature of theoceanic circulation through its influence on moisturesupply (Miller et al., 1992; Funder et al., 1992). The Thulearea of northwest Greenland, lying today at the northernlimit of the warm West Greenland current, is well placed,in theory, to investigate this interaction of climate andcirculation on ice sheet history but it requires a wellestablished chronostratigraphy to accomplish this.The present paper reviews the extent to which this ispossible, reviewing recent evidence and reappraisingpast information.

*Corresponding author. E-mail address: [email protected]

The Thule area (Fig. 1) has long been known to havea stratigraphical record of multiple glacial and marineevents extending back beyond the Late Weichselian gla-ciation maximum, from the occurrence of marine sedi-ments with thermophilous faunas 14C dated to '30 kain association with tills (Krinsley, 1963; Blake, 1975,1977, 1987; Kelly, 1980a). A subsequent investigation byourselves of old and new localities in 1986 led to therecognition of three glacial events and three marineevents (Funder, 1990). However, fieldwork in 1989 andsubsequent extensive laboratory investigations enablea fuller palaeoenvironmental reconstruction as well asnecessitating a revision of the age estimates for theseevents.

The area comprises the mainland around Wolsten-holme Fjord, the islands within the fjord and the offshoreislands of the Carey "er. The bedrock is unmetamor-phosed Proterozoic sedimentary and volcanic rocksoverlying Precambrian crystalline basement rocks

0277-3791/99/$ — see front matter ( 1999 Elsevier Science Ltd. All rights reserved.PII: S 0 2 7 7 - 3 7 9 1 ( 9 8 ) 0 0 0 0 4 - 3

Fig. 1. Locality map of the Thule area and reconstructed ice margins (described sections are indicated by letters, other sites by numbers and icemargins by Roman numerals).

(Dawes et al., 1982, 1991). The topography is one ofmoderate relief, with rounded low hills and plateauxrising to ca. 1000 m in Steensby Land. Along the northcoast, a marked bedrock platform lies along the coast atvarious elevations up to ca. 50 m and with widths of up toseveral kilometres. Elsewhere, the coastline mainly con-sists of steep cliffs except at valley mouths. The mainbathymetric features are the '200 m deep troughs in-cised into the SW branch of Wolstenholme Fjord and the'1000 m deep trough separating the Carey "er fromthe mainland, which is a continuation of the deepertrough of Hvalsund and Inglefield Bredning to the north.The present glacierisation consists mainly of local icecaps and their small outlet glaciers. However, HaraldMoltke Bræ, which occupies the inner part of Wolsten-holme Fjord, is a calving glacier draining partly from icecaps and partly from the main Greenland Ice Sheet,which otherwise lies to the east of the area.

2. Absolute and relative dating

(i) Thermoluminescence (TL) and optically stimulatedluminescence (OSL) dating of shallow marine and fluvialsediments were carried out by the Nordic Laboratory for

Luminescence Dating, University of Aarhus, using themethods described in Kronborg and Mejdahl (1990) andMejdahl and Funder (1994) (Table 1) (all uncertaintiesquoted for ages are 1 sigma). TL dates are corrected forthe effect of shallow traps, including uncorrected datespreviously reported in Kronborg and Mejdahl (1990).This correction compensates for the underestimation ofage arising from the difference between natural dose ratesand those used in laboratory irradiation of samples andthe presence of shallow traps in feldspars. This is basedon experiments with feldspars from the Thule area in theuncorrected age range of 40—120 ka (Mejdahl et al.,1992), although three older ages have also been corrected.

The luminescence dates show a continuous distribu-tion without stratigraphically significant breaks, al-though with only a few departures from the predictedstratigraphic sequence (Fig. 2). However, ages of depositsassigned to the middle marine event (Qarmat) havea wide range of 91—154 ka and a mean of 127$16 ka.Removing the two upper outliers reduces the range to91—133 ka and mean to 116$11 ka, which agrees wellwith the anticipated age discussed below. This restrictedgroup are all from littoral sands, which have the bestchance of signal zeroing during transport in the near-shore zone, with frequent wave resuspension. The two

374 M. Kelly et al. / Quaternary Science Reviews 18 (1999) 373—392

Table 1New and revised thermo- and optically stimulated luminescence dates

Sample no. Locality (section) Member Lab. no. TL date & CorrectedOSL date! (ka) date (ka)

68-143 Qarmat (I) Q R-861009 86 $10" 11568-144 Qarmat (I) Q R-861010 80$10" 10768-147 Qarmat (I) Q R-861011 80$10" 10768-185 Qarmat (I) ? R-861012 61$6" 8077-126 Dundas (P) Q? R-891021 69$10 9177-127 Dundas (P) N R-891022 30 $5 36

16$1!

76-898 Narsaarsuk (G) Q R-891008 92$10 12376-965 Narsaarsuk (E S R-891010 130$10 17776-967 Narsaarsuk (E) Q R-891011 89$10 11976-876 Kap Abernathy (M) Qr R-891005 88$10 11876-886 Kap Abernathy (O) ? R-891006 125$20 17076-894 Kap Abernathy (M) Q R-891007 99$10 13377-104 Iterlak (K) S R-891013 107$10 14477-113 Iterlak (K) N R-891014 31$5 38

9$1!

77-121 Iterlak (L) Q R-891016 88$10 11868-005 Saunders " (A) S R-861001 136$15" 18568-006 Saunders " (B) Q R-861002 114$10" 15468-007 Saunders " (A) S R-861003 119$10" 16168-008 Saunders " (B) Q R-861004 89$10" 11968-009 Saunders " (B) Q R-861005 113$10" 15368-010 Saunders " (B) N R-861006 14$2" 1468-011 Saunders " (C) N R-861007 36$4 45

6.5$0.5!

68-012 Saunders " (C) Q R-861008 69$10" 9176-853 Nordvest+ (X) Q R-891001 96$10 12976-871 Nordvest+ (X) ? R-891004 53$5 6976-863 Isbj+rn " (Z) ? R-891003 34$5 59

44$5!

! OSL dates."uncorrected dates from Kronborg and Mejdahl (1990)Member:- S: Saunders "; Q: Qarmat; N: Nuna; r: reworked

Fig. 2. Age distribution of luminescence dates according to stratigraphic member (1986 and 1989 surveys).

M. Kelly et al. / Quaternary Science Reviews 18 (1999) 373—392 375

Table 2New conventional and AMS radiocarbon dates

Sample no. Locality (section or no.) Member Lab. no.! Lab. age (conv. BP) Corrected d13C (ppt) Speciesage"(Rcorr BP)

77-002 Dundas (P) Qr? AAR-1401 '37,000 1.5 Mya truncata77-005 Dundas (P) N AAR-1402 9330$140 8930 1.8 Mya truncata76-901 Iterlak (93) N SRR-3759 9235$55 8835 #0.21 Mya truncata76-905 Iterlak (93) N SRR-3760 9110$55 8710 !0.3 Mya truncata77-028 Iterlak (91) N SRR-3761 9044$55 8644 0.0 Mya truncata77-101 Iterlak (K) Q SRR-3762 42,390#1700 0.0 Hiatella arctica

!2160 #Mya truncata77-110 Iterlak (K) S SRR-3763 42590#1700 0.0 Hiatella arctica

!2160 #Mya truncata77-115 Iterlak (L) N SRR-3764 8925$65 8525 #0.6 Hiatella arctica

#Mya truncata76-803B Granville Fjord (62) Q? T-8720 38900#750 Hiatella arctica

!600 #Mya truncata76-961 Drown Bugt (24) Qr? AAR-138 '40,700 Hiatella# Mya

#Serripes77-022B Drown Bugt (48) N T-8723 8580$100 #1.0# Hiatella arctica

#Mya truncata76-952 Booth Sund (58) N T-8721 9010$125 #1.0# Mya truncata77-021 Booth Sund (41) N T-8722 9150$60 #1.0# Hiatella arctica76-852 Nordvest+ (X) Qr AAR-140 '34,000 Chlamys islandica76-852 Nordvest+ (X) N AAR-1399 7440$90 7040 1.8# Mya truncata76-855 Mellem+ (9) Q? AAR-136 '39,000 Chlamys islandica76-856 Fire+ (13) Q? AAR-139 '34,000 Chlamys islandica76-859 Isbj+rn " (Z) N AAR-135 8530$160 8130 0.0# Hiatella arctica76-861 Isbj+rn " (Z) N AAR-1400 8420$110 8020 1.5# Mya truncata76-866 Isbj+rn " (Y) Q AAR-141 '33,000 Chlamys islandica

!AAR are AMS determinations."Finite age corrected for reservoir effect when not included in laboratory age.#Assumed value. Member — S: Saunders "; Q: Qarmat; N: Nuna; r: reworked.

outlying dates are for deposits of a more turbid environ-ment, with mud deposition interrupted by rapidlysedimented thin turbidite sands, all of which could resultin retention of a residual signal and an overestimated age.

TL dates for the youngest marine sediments (Nunaevent), 14—69 ka, are generally older than expected. How-ever, re-analysis of these using OSL, which is much moresensitive to brief exposure (Mejdahl and Funder, 1994),has produced two dates (10 ka and one of 16 ka, inbetter agreement with the 14C ages. The error in the TLages is probably due to incomplete zeroing affectingyoung sediments proportionally more than older (V.Mejdahl, pers. comm. 1994). The two remaining olderdates, one of which also has an old OSL date (44 ka), arefrom coarse beach sediments on the Carey "er whichhave produced a variety of dates by different methods(see below) and which appear to be an admixture ofNuna and possibly Qarmat material, both fossils andsediment matrix. The luminescence dates are, therefore,not necessarily an indication of the age of the marineconditions.

The few TL dates from sediments assigned to theoldest marine event (Saunders ") range from 144 to

177 ka with a mean of 167$16 ka, which, on stratig-raphical grounds, may not be reliable.

(ii) Radiocarbon dating was carried out on bivalvemolluscs at the Scottish Universities Reactor Centre,Scotland (SRR); Trondheim Laboratory for RadiologicalDating, Norway (T); and the AMS 14C Dating Facility,Aarhus University, Denmark (AAR). The results areshown in Table 2, which supplements the results pre-viously reported by Morner and Funder (1990). An agecorrection for the oceanic reservoir effect is applied to thelaboratory’s reported dates when this has not been al-lowed for, using an apparent age of 400 years (Mornerand Funder, 1990).

(iii) Uranium series dating was carried out on bivalvemollusc species at Lancaster University, U.K., based onthe ingrowth of 230Th from 234U and 238U. U and Thisotope determinations were made by alpha spectrometryafter ion-exchange separation of the elements (Table 3).Significant levels of 232Th are present in most samplesindicating contamination by detrital 230Th. A correctionfor this source of error was made on one sample by theuse of an isochron method. For the other samples, themean 230Th: 232Th ratio of three modern (living) samples


https://www.researchgate.net/publication/230034509_Luminescence_dating_of_Late_Quaternary_sediments_East_Greenland?el=1_x_8&enrichId=rgreq-3860f3f4-938d-4249-ae32-f28aabfd5b5d&enrichSource=Y292ZXJQYWdlOzIyODM2ODkxMjtBUzo5OTcwNDA5MjIzMzcyOEAxNDAwNzgyNzE3MTQw

Table 3Uranium series dates on molluscs

Sample Locality Member Species 232Th 238U 234U/232Th 234U/238U 230Th/232Th Age (ka)no. (mg kg~1) (mg kg~1) $1 sigma

68-143 Qarmat (I) Q Mt 2.183$0.049 6.568$0.081 10.929$0.281 1.185$0.021 5.843$0.167 69.5$2.668-112A Narsaarsuk

(E)Sr Mt! 0.102$0.005 1.671$0.029 62.321$3.480 1.235$0.030 22.880$1.254 106 $60"

B Mt! 0.223$0.008 1.675$0.029 29.773$1.159 1.290$0.030 10.274$0.388C Mt! 0.180$0.011 1.121$0.043 25.635$1.780 1.344$0.068 6.248$0.416D Mt! 0.139$0.009 1.542$0.021 42.392$2.707 1.249$0.023 17.589$1.13368-126 Narsaarsuk

(F) SrHa 2.699$0.059 5.009$0.108 7.603$0.234 1.337$0.041 4.266$0.124 69.1#3.8!3.7

68-005 Saunders S ? 2.291$0.052 7.225$0.100 13.512$0.355 1.398$0.027 6.788$0.207 63.4#2.6!2.5" (A)

68-006 Saunders Q Ha 0.423$0.014 7.265$0.116 75.512$2.552 1.359$0.031 31.163$1.076 57.5$1.7" (A)

77-110 Iterlak (K) S Mt 0.504$0.017 1.594$0.052 13.568$0.624 1.400$0.065 9.467$0.352 105.6#8.1!7.4

76-861 Isbj+rn N Ha 0.094$0.005 2.289$0.030 73.333$3.777 0.980$0.018 4.738$0.264 5.7$0.2" (Z)

77-128 Narsaarsuk Mt 0.126$0.010 0.260$0.010 7.133$0.630 1.132$0.060 0.876$0.106 modern77-129 Narsaarsuk Ha 0.396$0.018 0.501$0.016 4.516$0.241 1.163$0.050 0.895$0.059 modern77-1331 Godhavn, Ha 0.791$0.037 0.409$0.023 1.956$0.124 1.232$0.087 0.918$0.062 modern

Disko

Ha: Hiatella arctica, Mt: Mya truncata.! single valves." Isochron date.Member — S: Saunders "; Q: Qarmat; N: Nuna; r: reworked.

was used as the basis for correction. Theoretically, thelatter method is much less satisfactory, although theisochron based result has a large uncertainty.

One sample gives an age appropriate to the Nunamarine event, agreeing with the results of other datingmethods. All the other dates in the range 58—106 ka aresubstantially younger than TL dates on the same sedi-ments or inferred dates. This is considered to be due todiagenetic uptake of U.

(iv) Analysis of the D-alloioleucine : L-isoleucine ra-tios (aIle : Ile) in bivalve molluscs were carried out atBergen University, Norway (Sejrup, 1990). Several valvesfrom each sample were analysed and replicate analysescarried out on many of these. The means and standarddeviations of the results are shown in Table 4, whichsupplements that published previously by Sejrup (1990).

The ratios in the mean total hydrolysate fraction fromsamples from a given site show a continuous distributionwhich does not distinguish clearly between the variousstratigraphic units (Fig. 3). Thus, the two oldest marineunits (Saunders " and Qarmat) are even less separatedthan by the luminescence dates, presumably due to sitespecific factors which have influenced the thermal historyexperienced by the shells. The overall poor time resolu-tion may also be caused by extremely low diagenetictemperatures during the cold intervals before and afterthe marine events, as suggested by Sejrup (1990). How-ever, there is usually a clear distinction between Qarmat

and Nuna event samples, with Holocene age shells fromthe latter having total hydrolysate aIle:Ile ratios (0.017and a non-detectable free fraction. One anomaloussample from Carey "er beach sediments, with a mean of0.019 and range of 0.013—0.032, apparently includesshells of different ages, giving Holocene and infinite AMS14C dates. A similar explanation holds for a sample witha high mean and range of values from reworked shells atthe surface of a beach gravel. Typically, high ratios havebeen obtained from glacially reworked shells in tills.

As indicated above, there is sometimes a lack of corres-pondence between dates established by the differentmethods, TL, OSL, amino acid and 14C, especially atsites on the islands of the Carey "er (Isbj+rn ", Nor-dvest+ and Fire+). The deposits there, located in smallbays and inlets, are unusually coarse boulder gravels withsand matrices which appear to be mixed, in terms ofsediments and/or faunas, by mass movement, marinereworking and by the re-occupation by faunas of oldsubstrates.

3. Faunal analysis

3.1. Macroinvertebrates

Samples analysed for macroinvertebrates were gener-ally 5 kg, although some samples consist of shells picked


Table 4New amino acid ratios (aIle : Ile) in total hydrolysate and free fractions in Mya truncata

Sample no. Locality (section or no.) Member Lab. no. Alloisoleucine : isoleucine ratio

HYD mean FREE mean$1 sigma $1 sigma(no. analyses) (no. analyses)

77-002 Dundas (P) N BAL 2427 0.020$0.002 (6) 0.203$0.087 (6)77-005 Dundas (P) N BAL 2428A 0.011$0.002 (2) ND76-877 Kap Abernathy (M) Qr BAL 2424 0.034$0.002 (8) 0.306$0.028 (8)76-878 Kap Abernathy (M) Q BAL 2425 0.035$0.005 (7) 0.303$0.048 (7)76-880 Kap Abernathy (M) Q BAL 2426 0.052$0.013 (6) 0.359$0.031 (6)77-101 Iterlak (K) Q BAL 2429 0.019$0.002 (6) 0.187$0.020 (6)77-102 Iterlak (K) Q BAL 2430 0.021$0.002 (6) 0.171$0.025 (6)77-110 Iterlak (K) S BAL 2431 0.034$0.006 (6) 0.324$0.013 (6)77-115 Iterlak (L) N BAL 2432 0.011$0.001 (6) ND76-803 Granville Fjd. (62) Q? BAL 2137 0.025$0.001 (6) 0.238$0.031(1)76-852 Nordvest+ (X) N#Qr BAL 2138 0.019$0.006 (6) ND76-872 Nordvest+ (15) r BAL 2144 0.062$0.014 (7) 0.350$0.054 (5)76-855 Mellem+ (9) Q? BAL 2139/40 0.026$0.005 (6) 0.152$0.029 (6)76-856 Fire+ (13) N BAL 2141 0.015$0.001 (5) ND76-861 Isbj+rn " (Z) N BAL 2142 0.014$0.002 (7) ND76-866 Isbj+rn " (Y) Q? BAL 2143 0.035$0.007 (4) 0.206$0.023 (4)

Member — S: Saunders "; Q: Qarmat; N: Nuna; r: reworked.

Fig. 3. Distribution of amino acid ratios (total hydrolysed D-alloisoleucine:L-isoleucine) according to stratigraphic member (1986 and 1989 surveys).


from the surface. Table 5 lists the species collected duringfield work in 1989, supplementing the published listsfrom sites investigated in 1986 (Feyling-Hanssen andFunder, 1990). The following assemblages have been re-cognised in the fossil macroinvertebrate faunas. Theirecological interpretation and biostratigraphy are dis-cussed more fully later.

(i) Macoma calcarea assemblage, with Mya truncata,Hiatella arctica, Clinocardium ciliatum, Serripes groenlan-dicus, Balanus balanus and Strongylocentrotusdroebachensis present as common species: this is con-sidered to be identical to the arctic Macoma calcareacommunity which dominates the present benthic com-munities in the area (Vibe, 1950).

(ii) Nuculana-Chlamys assemblage, with one or moretaxodont species such as Nuculana minuta, N. pernula,Bathyarca glacialis, Portlandia arctica, as well as Chlamysislandica and the polar water species Palliolum greenlan-dicum: this is a soft bottom community, developed belowwave base probably indicating conditions slightly coolerthan at present.

(iii) Mytilus-Balanus assemblage, typically composedof worn shell-fragments of Mytilus edulis and compart-ment plates of Balanus, especially Balanus balanoides: thisis a shallow water fauna indicating a subarctic influencegreater than that of today.

3.2. Foraminifera

For analysis of the foraminifera, 100 g samples wereprepared according to Meldgaard and Knudsen (1979).At least 300 specimens were identified from eachsample, where ever possible. The percentage distributionof the more frequent species recorded from the 1989survey is given in Table 6. For poor samples the actualnumbers obtained are indicated with italics. The tablesupplements the records given in Feyling- Hanssenand Funder (1990). The most frequent of theforaminiferal species indicating warmer conditions thanat present in the area are marked with ‘SA’ (cf. Gudinaand Evserov, 1973, Feyling-Hanssen, 1980, Feyling-Hanssen and Funder, 1990). In addition, the faunaldiversity index of Walton (1964) has been used for theenvironmental interpretation. The following sixforaminiferal assemblages have been recognised in thedeposits; three (i)—(iii) indicating water temperaturescomparable to the present and three (iv)—(vi) warmerconditions:

(i) The Cibicides lobatulus-Astrononion gallowayi as-semblage is dominated by C. lobatulus with A. gallowayiand Elphidium hallandense as frequent associated speciesand with a low faunal diversity index. This compositionof species indicates normal marine salinity in a highenergy environment and the assemblage presumably re-sembles the present faunas in the area (Phleger, 1952;

Loeblich and Tappan, 1953; Feyling-Hanssen and Fun-der, 1990).

(ii) The Astrononion gallowayi assemblage is domin-ated by A. gallowayi together with Islandiella helenae, I.norcrossi, I. islandica and C. reniforme (see Feyling-Han-ssen and Funder, 1990). The faunal diversity is low and sois the content of subarctic species. This assemblage typereflects ice-distal arctic conditions in an area of normalmarine salinity (see below).

(iii) The Islandiella helenae assemblage is dominatedby I. helenae with associated species such as I. norcrossi,A. gallowayi, C. reniforme, E. excavatum and C. lobatulus.This type of fauna is found in ice-distal sediments in theCanadian Arctic (Osterman and Andrews, 1983; Oster-man and Nelson, 1989) and is similar to that found inpresent assemblages of the Thule area (Phleger, 1952;Feyling-Hanssen and Funder, 1990).

(iv) The Astrononion gallowayi-Gavelinopsis praegeriassemblage is dominated by C. lobatulus together withA. gallowayi. A content of the subarctic speciesGavelinopsis praegeri and Rosalina vilardeboana, togetherwith a relatively high faunal diversity index, point toameliorated temperature conditions. A. gallowayi isfound associated with relatively warm, but still arctic,saline bottom waters in the recent arctic (Mudie et al.,1984). C. lobatulus was considered a subarctic species byGudina and Evzerov (1973), but this species seems to bestrongly connected with coarse sediments and high en-ergy conditions rather than temperature (Murray, 1991).This assemblage is interpreted as indicating open normalsaline waters and temperatures higher that at present inthe area.

(v) The Nonionella auricula assemblage is character-ised by a high faunal diversity index and a relatively highcontent of subarctic species such as N. auricula, Melonisbarleeanus, Islandiella inflata and in some samples alsoG. praegeri. The dominant species are E. excavatum,Cassidulina reniforme, I. islandica, I. helenae and I. nor-crossi. Shallow water species such as E. albiumbilicatum,Haynesina orbiculare, E. hallandense and the Buccellafrigida group are usually relatively common. This assem-blage also reflects conditions warmer than at present (e.g.Gudina and Evzerov, 1973; Feyling-Hanssen and Fun-der, 1990).

(vi) The Islandiella inflata-Cibicides lobatulus assem-blage is characterised by the subarctic I. inflata (seeGudina, 1966; Feyling-Hanssen, 1990) together witha relatively high amount of shallow water species such asE. albiumbilicatum and E. hallandense. The content of thesubarctic species I. inflata, E. albiumbilicatum, N. auriculaand G. praegeri and the high faunal diversity indicespoint to warmer temperatures. Dominant species in theassemblage are C. lobatulus, E. excavatum, C. reniformeand I. helenae. This assemblage also indicates warmerconditions than at present in a relatively shallow waterenvironment.


Table 5Macrofossil marine faunas from new localities in the Thule area

Sam

ple

No.

Mem

ber

mab

ove

sea

leve

l

Ass

embla

ge

Nuc

ula

tenu

isR

eeve

Nuc

ula

pern

ula

Cost

iger

a

Nuc

ulan

am

inut

a(O

.F.M

olle

r)

Por

tlan

dia

arct

ica

(Gra

y)

Bat

hyar

cagl

acia

lisG

ray

Mod

iola

ria

nigr

a(G

ray)

?Cre

nella

faba

Mul

ler

Myt

ilus

edul

isLin

ne

Chl

amys

isla

ndica

(Mul

ler)

¸im

atul

ahy

perb

orea

Jense

n

Pal

liolu

mgr

eenl

andi

cum

(Sow

erby)

Indicator sp. A! SA# SA# A# A!

Nordvest+76-852 Q 3.2 — — — — — — — — s — —

Mellem+76-856 Q 9.9 — — — — — — — — s — —

Fire+Y 76-856 Q 9.2 chu-nu — — — — — — — — s — 1

Isb+m+Y 76-860 Q 13.2 — — — — — — — — — — —

76-861 Q 9.4 ch-nu — — — — — — — — s 1 —76-870 Q 9.4 1 — — — — — — — s — —76-862 Q 8.6 — — — — — — — — — — —

Z 76-866 Q 7.3 my-ba — 2 — — — — — s c — —Iterdlak

K 77-101 Q 25.6 my-ba — — — — — — — c — — —77-102 Q 25.5 my-ba — — — — — — — c — — —77-105 S 10.0 ch-nu — — s 2 2 — — 1 s — 277-106 Q 9.0 ch-nu — — s ?s — — — — s — —77-110 S 1.0 ch-nu — s c f — s 2 — s — s

L 77-115 N 15.5 — — s — — — — — — — —77-118 N 14.5 — — 1 — — — — — — — —

Kap AbernathyM 76-876 Q 5.8 my-ba — — — — — — — s — — —

76-877 Q 5.5 my-ba — — — — — 1 — f — — —76-894 Q 5.0 my-ba — — — — — — — f — — —76-878 Q 4.3 my-ba — — — — — — — f — — —76-879 Q 3.3 my-ba — — — — — — — c — — —76-880 Q 1.5 my-ba — — — 1 — — — c c — —

DundasP 77-003 N 7.3 — — — — — — — — — — —

77-002 N 7.2 — — — — — — — — — — —77-015 N 6.0 — — — — — — — — — — —77-013 N 5.5 — — — — — — — — — — —77-006 Q 3.7 my-ba — — — — — — — c — — —

NarsaarsukN 76-866 S 1.5 ch-nu — — — — — — — — s — 3

4. Lithostratigraphic and biostratigraphic interpretation

During the two field surveys, detailed examination,logging and sampling was conducted of sections ex-cavated in coastal exposures, together with reconnais-

sance mapping of the geomorphology and deposits of thesurrounding areas.

Lithostratigraphic logs for key sections and theircorrelation are shown in Fig. 4a and b (with samplenumbers from 1986 and 1989 surveys). Additional litho-


Table 5. Continued

¹rido

nta

bore

alis

Shum

ache

r

¹rido

nta

mon

tagu

iD

illw

yn

¹rido

nta

mon

tagu

iva

r.st

riat

aL

each

¹rido

nta

sp.

Axi

nops

ida

orbi

cula

taSar

s

Clin

ocar

dium

ciliat

um(F

abrici

us)

Serr

ipes

groe

nlan

dicu

s(B

rugu

iere

)

Mac

oma

calc

area

(Gm

elin

)

Hia

tella

arct

ica

(Lin

ne)

Myo

trun

cate

Lin

ne

¸ep

eta

caec

a(M

ulle

r)

Mar

garite

she

lici

nus

(Fab

rici

us)

Put

illa

cf.tu

rgid

a(J

effre

ys)

Sol

arie

llaob

scur

a(C

outh

ouy)

Alv

ania

cf.j

anm

ayen

i(F

riel

e)

¹ro

phon

trun

catu

s(S

trom

)

Buc

cinu

msp

.(?bel

cher

iR

eeve

)

Ret

usa

obtu

sa(M

onta

gu)

¹on

icel

lam

arm

orea

(Fab

rici

us)

Bal

anus

bala

noid

esL

inne

Bal

anus

cf.ba

lano

ides

Bal

anus

bala

nus

(Lin

ne)

Bal

anus

cren

atus

Bru

gier

e

Bal

anus

cf.cr

enat

us

Bal

anus

sp.

Stro

ngly

loce

ntro

tus

droe

bach

ensis

(Mulle

r)

Par

adex

iosp

ira

vitr

ea(F

abrici

us)

Spiror

bis

sp.

Byr

ozoa

SA! SA! SA! SA# SA# SA!

— — — — — — — — f s 3 — — — — — — — — — — 2 f — — — — — —

— — — — — — — — f s 3 — — — — — — — — — — c — 2 — f — — 1

1 — — — 1 — — — c c — — — — — — — — — — — 1 — — — 3 — — 2

— — — — — — — — 1 1 — — — — — — — — — — — 2 — — — c — — —s — — — — — — — f s f 3 — 3 2 1 — 2 — — — f — — — f — — f1 — — — — — — — f 2 s — — — — — — — — — 1 f — — — f — s —— — — — — — — — 1 1 1 — — — — — — — — — — — — — — 2 — — 4— — — — — — — — c c — — — — — — — — — — — f — — — f — s c

— — — — — — — 2 s c — — — — — — — — — — — — — — — — — — —— — — — 1 — — s f f — — — — — — — — — — — 1 1 — f 1 — — 12 s — — — s — — c c — — — — — — — — — — — s s — — 3 — c s— 2 — — — 1 — — — — — — — — — — — — — — — — — — s — — s —— — s — — 3 1 — f f — — — — — — — — — — — ? — — — 2 — 3 s— — — — — 2 — s c c — — — — — — — — — — — c 2 — — 1 — — —1 — — — — — — — s s — — — — — — — — — — — ?c — — — — — — —

— — — — — — — — 1 c — — — — — — — — — — 2 — — — s — — — —— — — — — — — — s f — — — — — — 1 — — — 3 — 1 — f — — 1 —3 — — — — — — — 3 f — — — — — — — — — f f 1 — — — 2 — — —— — — — — — — — — f — — — — — — — — — c c — — — — 1 — — —c — — — — — — — — f — — — — — — — — — f f 1 — — — c — 1 —— — — — — — — — s s — — 1 — — — — — f f — s c — f 3 f s —

— — — — — 2 — s 1 f 2 — — — — — — — — — — — — — — — — — —— — — — — — — — s c — — — — — — — — — — — s — — — — — — —— — — — — — — — 2 — — — — — — — — — — — — — — — — — — — —— — — — — — — — — ?c — — — — — — — — — — — — — — — — — — —— — — — — — — — — 3 — — — — — — — — — — — — — — — 1 — — —

— — — — — — — 1 3 c — — — — — — — — — — — c c — — — — — 3

f: frequent ('20 fragment/valves), c: common (10—20), s: scarce (3—10); r: rare (1—3); 1: identified by D. Eibye—Jacobsen; Indicator sp.: see text for key;Member—S: Saunders 0, Q: Qarmat, N: Nuna.

and biostratigraphic information for the sites investi-gated in 1986 (Qarmat, Narsaarsuk and Saunders ") hasbeen published in Houmark-Nielsen et al. (1990), Funderand Houmark-Nielsen (1990), and Feyling-Hanssen and

Funder (1990). With one exception (see the Kap Aber-nathy event), the lithostratigraphy established by theearlier work is applicable to the wider area covered bythe later fieldwork.


Fig. 4. (a) Lithostratigraphic logs and their correlation, eastern part of the Thule area (b) Lithostratigraphic logs and their correlation, western part ofthe Thule area (see Fig. 4a for key).

Six major lithostratigraphic units, i.e. members, havebeen recognised in the sections, comprising three units ofmarine facies sediments (ice distal) and 3—4 units ofglacial facies sediments (including ice proximal marine).The marine units represent marine events defined by theinundation of areas now land. They include transgressiveand regressive phases responding to glacio-isostaticevents modified by eustatic controls. The glacial unitsrepresent glacial events defined by an increase in icecover relative to the present. The two types of event arenot necessarily separated in time since, beyond the mar-gin of an extended ice cover, a glacial event could berepresented in time by a marine event. Furthermore, theboundaries of lithostratigraphic units and the events theyrepresent will be spatially diachronous.

The key to the stratigraphic interpretation is the dis-tinction of the three marine members seen in the sectionsand their correlation across the area. In order of decreas-ing age, these constitute the deposits of the Saunders ",Qarmat and Nuna marine events. The youngest, Nunamember is recognised by young 14C dates (see below) andthe amino acid criteria given earlier. The primary evid-ence for the distinction of the oldest marine members istheir relative stratigraphic position at 3 sites (Fig. 4A/B,

E and K). In addition, the new evidence for a faunaldifference is used to correlate isolated occurrences withthe main sites, i.e. the restriction of the Mytilus-Balanusmollusc assemblage to the Qarmat member. By theirrelationships to the marine members, three or four glacialmembers and events are distinguished: Agpat, Narsaar-suk, and Kap Abernathy/Wolstenholme Fjord, in orderof decreasing age. Their chronostratigraphy is discussedlater but an indication of the proposed correlation withOxygen Isotope Stages is given in the headings below.

4.1. Agpat glacial event (OIS 6?)

Sediment of this event (Agpat member) is present inonly one section (A), at Saunders ", where a mud matrixsupported diamicton with predominantly exotic coarseclasts interpreted as till lies beneath the marine Saunders" member.

4.2. Saunders Ø marine event (OIS 6?)

At the type site at the island of Saunders" (Fig. 4A—C),the Saunders " member consists of 18 m of fossiliferoussediments. The lowest subunit is interbedded sandy or


Table 6Distribution of fossil foraminifera from new localities in the Thule area. Numbers are percentages (X for (0.5) or actual numbers when small (italics). (Abbreviations}SA: subarctic group;S: Saunders " event, Q: Qarmat event, N: Nuna event; assemblages}Cib:Ast: Cibicides lobatulus}Astrononion gallowayi, Ast-Gav: Astrononion gallowayi}Gavelinopsis praegeri, Isl. hel:Islandiella helenae, I. in#-Cib: Islandiella in-ata}Cibicides lobatulus, Non. aur: Nonionella auricula)

Fig. 4. Continued

muddy diamicton and marine shell bearing muds andsands interpreted as a deep water sediment. Above, isa thick sequence of mainly foreset bedded, clast sup-ported gravels of local origin, interbedded with sands.The upper part is a horizontally stratified gravel indicat-ing littoral conditions. The whole is interpreted as repres-enting the progradation of a small coastal delta or spitduring a period of sea level regression following theAgpat glacial event. At Iterlak (Fig. 4K—L), this membercomprises 13 m of fossiliferous glaciomarine diamictoninterrupted by 1 m of clast supported gravel with aninfiltrated mud matrix. Unless represented by this gravel,a regressive episode has not been identified at Iterlak. AtNarsaarsuk, the results of the revisit to the site indicatedthat the lowest unit (Fig. 4E) is a glacially disturbedfossiliferous marine diamicton assigned to the Saunders" event.

The macrofossil faunas at all three sites displayNuculana-Chlamys assemblages but in association withsome subarctic species. Iterlak has the highest frequencyof the polar water species Portlandia arctica and Pal-liolum greenlandicum encountered in the area. Theforaminiferal faunas at these localities include two assem-blages with subarctic elements indicating warmer condi-

tions than at present. Of these, the Islandiella inflata-Cibicides lobatulus assemblage at Iterlak indicates highenergy conditions in relatively shallow water, while theNonionella auricula assemblage at Narsaarsuk and Saun-ders " (Feyling-Hanssen and Funder, 1990) representsa deeper water facies.

4.3. Narsaarsuk glacial event (OIS 6?)

At the type locality at Narsaarsuk (Fig. 4E—G), proxi-mal glacial facies sediments lie between deposits of theSaunders " and Qarmat marine members. They com-prise a sandy diamicton, interpreted as a basal till, to-gether with coarse gravels and boulder gravels which, inturn, pass laterally into cross bedded sands. Deformationstructures at the base of the unit are consistent with an iceflow from the NW. At Iterlak (Fig. 4K), 11 m of verycoarse clast supported boulder gravels with a silt to sandmatrix occur between the Saunders " and Qarmat mar-ine members. These sediments are also considered to beproximal glacial deposits although, because of the insta-bility of the section, they were not closely examined. AtSaunders " (Fig. 4A—B), 0.5 m of clast supported graveland diamicton overlying the shallow marine beds of the


Saunders " member is now interpreted as a glacial prox-imal deposit belonging to this event (Houmark-Nielsen,pers. comm.). The bed above, 0.5 m of mud matrix sup-ported diamicton with shell fragments, represents thetransition to deep water conditions at the beginning ofthe next marine event.

4.4. Qarmat marine event (OIS 5)

Typically, the Qarmat member consists of coarseningupwards, regressive facies sequences typified by the suc-cessions present at Narsaarsuk and Qarmat. They beginwith fossiliferous marine muddy diamictons and muds,locally containing burrowing molluscs in life position,which are considered to be deposits of relatively deepwaters. The general coincidence of warm faunas (seebelow) and the clast content of the muds, which is highestat the Qarmat section (Fig. 4I), indicates the existence ofa calving glacier in inner Wolstenholme Fjord, givingconditions similar to those that existed during the Nunaevent in the Holocene. The muddy sediments are suc-ceeded by sands associated with shallower water, withthe position of the transition depending on local depthconditions. At Narsaarsuk (Fig. 4E—G), the coarser sedi-ments are 3—8 m of horizontally bedded sands withplanar and cross-lamination. Laminae of heavy mineralconcentrates and lag gravel horizons indicate wave re-working. There is abundant evidence of the existence ofbenthic infaunas in the form of bioturbation, burrowingand in situ molluscs. These sands thicken towards thesouth (G) where the uppermost subunit is cross-beddedsands with laminae of plant debris, representing a smalldeltaic deposit. At Qarmat, the thinner sequence presentthere ends with a beach gravel.

Other localities show departures from the basic se-quence described above. At Iterlak, only the coarser unitsare present, with 3 m of shallow marine sands andgravels. A thicker sequence exists at Kap Abernathy but,unlike other Qarmat sites, it has been strongly glacio-tectonised and the overall stratigraphy is unclear (Fig. 5M—O). However, 4.5 m of shell rich sands and gravels,which overly muds and are interpreted as beach deposits,can be assigned to this event by their fauna (M). Thestratigraphical position of other units is unclear. Shallowwater coarse sediments with reworked warm faunal ele-ments at Dundas (Fig. 4, P) possibly also belong to thisunit. More notably, at Saunders " (Fig. 4 A—C), muds aresucceeded by 15 m of cross-bedded sands and gravelscoarsening upwards to beach gravels, considered to bethe deposits of a prograding delta or coastal spit. Impor-tantly, this unit is interpreted as overlying and distinctfrom the similar sequence belonging to the Saunders" event and, therefore, a second regression episode.Overall, all the Qarmat member successions agree witha model of marine regression related to glacio-isostaticuplift.

The occurrence of this unit in the Carey "er is moreproblematic because of the mixing of sediments andfaunas from the Nuna and an older marine event, as thedates discussed earlier indicate. The sediments in ques-tion, occurring in bays and inlets of the four investigatedislands, are unusually coarse-grained sediments, formedof boulders and cobbles with interstitial sands. On theevidence of the fauna at Isbj+rn " (see below) and a TLdate from Nordvest+, all the deposits which have yieldedpre-Holocene amino acid ratios and infinite 14C datesare considered to be in part from the Qarmat event, i.e.occurrences on Isbj+rn " (Fig. 4, Y—Z), including cobbleand boulder beds with attached Balanus balanus whichhave a non-finite 14C age (Blake, 1977); Nordvest+(Fig. 4X); Fire+ (0.5 m of shelly gravel); and Mellem+(25 m of boulders with a local shell gravel matrix).

The littoral facies of the Qarmat member characteristi-cally contains elements of the shallow marine Mytilus-Balanus assemblage, including locally extinct species.Amongst these species, B. balanoides occurs at Qarmat,Narsaarsuk, Kap Abernathy and possibly on Isbj+rn ".Individual sites also have other locally extinct subarcticspecies: the gastropod ¹rophon truncatus, a serpulid Par-adexiospira vitrea and a bryozoan Berenicea arctica. OnSaunders ", the fauna is represented by the frequentoccurrence of Mytilus edulis. The more subarctic charac-ter of the fauna indicates a higher water temperature thanat present (see below).

The foraminiferal assemblages include two which indi-cate warm conditions, e.g. the shallow marine Islandiellainflata-Cibicides lobatulus assemblage found at KapAbernathy and the Nonionella auricula assemblage atQarmat (Feyling-Hanssen and Funder, 1990) and Nar-saarsuk, at the latter together with a relatively highcontent of shallow water species. However, water temper-atures comparable to the present are indicated by otherassemblages. The Islandiella helenae assemblage, repres-enting high-arctic, but ice-distal conditions is present inone sample from Iterlak and one from Qarmat (Feyling-Hanssen and Funder, 1990). Another arctic assemblage,the Astrononion gallowayi assemblage, occurs at the baseof Qarmat event deposits on Saunders " (Feyling-Han-ssen and Funder, 1990). The assemblage in the underly-ing sample appears to be a transitional fauna between theAstrononion gallowayi and Islandiella helenae assem-blages. On the Carey "er, the Cibicides lobatulus-As-trononion gallowayi assemblage in the Qarmat memberindicates a high energy environment and the planktonicspecies point to open, ice-free conditions.

4.5. Kap Abernathy and Wolstenholme Fjord glacial events(OIS 4? and 2)

Glacial sediments of the Wolstenholme Fjord member,as originally defined (see below), are only present insections in the vicinity of the inner fjord, i.e. at Qarmat,


Fig. 5. Sketch of the stratigraphy and structure of the coastal exposures at Kap Abernathy.

where they overlie the sediments of the Qarmat marinemember and an intervening fluviatile sand, and beneaththe Nuna event marine sediments at Nunatarsuup nuua(Fig. 4J). At Dundas, close to the postulated ice marginfor this event, 3.5 m of gravels with reworked Qarmatevent fossils may be a proximal glacial facies or an older,talus influenced, littoral marine deposit.

The latest survey has shown that glacial sedimentsyounger than the Qarmat event also occur 10 km west ofthe Wolstenholme Fjord ice margin, at Kap Abernathy.There, a basal till, comprising a sandy diamicton withsediment clasts and shells reworked from older sedi-ments, overlies glacio-tectonised sediments which includethe Qarmat member (Fig. 5). In turn, it is overlain by thinbeach gravels presumed to belong to the Nuna event. Therelationship between the Kap Abernathy and Wolsten-holme Fjord events is not known and the possibilitiesrange from them being separate events, with the absenceof a marine event associated with the earlier being due toeustatic low sea levels, to both being phases of the sameglacial event. For the present, both terms are retained.

4.6. Nuna event (OIS 2? and 1)

Marine facies sediments from this event, constitutingthe Nuna member, cap many of the sections (Fig. 4).Coarsening upwards sequences, comprising fossiliferousmarine diamicton overlain by sand and then gravel, oc-cur at Saunders", Narsaarsuk, Iterlak and Nunatarsuupnuua. At Dundas, the sequence is modified by the upper-most unit being thick angular gravels interpreted asa prograding talus from the adjacent hillslope and, at theCarey "er, clast supported boulder gravels with sandmatrix are present.

The mollusc faunas generally include the Macoma cal-carea and the Nuculana-Chlamys assemblages. More no-tably, the shallow marine facies include the Mytilus-Bal-anus assemblage, but without Balanus balanoides and theother exotic species present in this assemblage in theQarmat event.

The arctic Islandiella helenae assemblage is found indeposits from the Nuna event at Dundas and Iterlak.A content of 3% Islandiella inflata in the Iterlak deposit

(118, Table 6) may be considered as reworked Qarmatevent deposits. Similar Islandiella helenae assemblagesoccur on Saunders " whereas the Astrononion gallowayiassemblage occurs at Narsaarsuk (Feyling-Hanssen andFunder, 1990). These two assemblage types do not con-tain subarctic species and indicate temperature condi-tions similar to those found in the area today. The corres-ponding Nuna event deposits on the Carey Islandscontain the high energy Cibicides lobatulus-Astrononiongallowayi assemblage, which also indicates similarconditions to the present.

5. Extent of ice cover

Consideration of the extent of the ice cover during thevarious glacial events requires correlation of the sectionevidence with the regional geology and geomorphology.Reconciling the two sets of evidence is problematical.

5.1. Shelf glacial phase

There is clear evidence in the upland areas for theirglaciation by an ice sheet which would have reached theshelf margin (Fig. 6). In all cases, these glacial landscapeshave a very weathered appearance. For example, in thearea between Iterlak and Granville Fjord (Fig. 1), theplateau surfaces above 200—300 m are either deeplyweathered bedrock, felsenmeer developed mainly in situbut with scattered erratics, or heavily soliflucted tills.Further west, around Booth Sund, soliflucted tills andfelsenmeer occur on surfaces between 200 and 400 m. Inboth areas, exotic crystalline erratics with a source to theeast or north occur at the margins of present day smallice caps, indicating that they developed on surfaces onceglaciated by a major icesheet.

A similar situation exists on the Carey "er, where theupland areas are covered with a deeply weathered felsen-meer with scattered erratics. These are dominated bysandstones and dolomites from the Wolstenholme Fjorddrainage area to the east while the less frequent reddishgranite may come from Inglefield Land to the north(P. Dawes, personal communication 1991). Striations


Fig. 6. Reconstructed ice sheet margins in northern Baffin Bay (600 m margin and ice divide from Reeh, 1984).

and chatter-marks, occurring on bedrock ‘islands’ pierc-ing the regolith, also indicate ice movement from thenorth (Chamberlain, 1895; Bendix- Almgreen et al., 1967;

Blake, 1977; Blake et al., 1996). For this to have occurredrequires a considerable expansion of the Greenland icesheet, resulting in a major ice stream flowing out of


Hvalsund-Inglefield Bredning and occupying the deepsubmarine trough which separates the Carey "er fromSteensby Land (Fig. 1).

The deep weathering of the upland surfaces suggests anold age for the shelf glaciation, unless it is argued that theweathered surfaces were preserved beneath a youngercold based ice-sheet. However, it is consistent with theoverall stratigraphic interpretation discussed below ifthis phase is correlated with the Agpat event.

5.2. Outer fjord glacial phase

A phase of fjord glaciers with ice margins in the outerfjords is indicated by degraded lateral moraines whichoccur near to sea-level in the middle reaches of theWolstenholme Fjord complex (II, Fig. 1). Thus, at Mori-usaq, a lateral moraine, kame terraces and other ice-marginal features lie along the fjord side from 70—210 m.Further east, at Iterlak, meltwater channels occur at250—350 m and glaciolacustrine sediments exist in thevalley bottom, where a lake was dammed up by the ice inthe fjord. Similarly, two large moraines at 70—113 m, alsoassociated with glaciolacustrine sediments, cut across themouth of the valley at Narsaarsuk on the fjord system’ssouthern side. Further west, ice marginal features indi-cate the presence of ice correlated with this phase occupy-ing Granville Fjord and in small valleys of the BoothSund area but there is no evidence for an ice stream alongthe outer coast, i.e. between southern Steensby land andthe Carey "er.

The keys to the relationship between the regionaland section evidence is the correlation between the outerfjord moraine at Narsaarsuk and the Narsaarsukmember present in the adjacent coastal exposures (Hou-mark—Nielsen et al., 1990) and the fact that ice contactsediments between the Qarmat and Nuna marine mem-bers have not yet been recognised in the sections atIterlak, Saunders " and Narsaarsuk. This is taken toindicate that the outer fjord moraines predate the Qar-mat event and mark the limit of the Narsaarsuk ice sheetand, furthermore, that the inner fjord phase is of post-Qarmat age.

5.3. Inner fjord glacial phase/s

This phase was located by Houmark-Nielsen et al.(1990) at the Wolstenholme Fjord lateral moraine systemand, hence, correlates with the member and event of thisname. The moraine stretches for 12 km along the south-ern side of the inner part of the fjord and is consistentwith an ice front lying at the mouth of the inner fjord (IV,Fig. 1). A qualitative distinction was made between thedegree of weathering on these moraines and on those ofthe outer fjord phase. The newly discovered Kap Aber-nathy till indicates an ice cover of at least intermediateextent between the outer and inner phases but has been

given a theoretical ice margin in the vicinity of KapAbernathy (III, Fig. 1) because of its stratigraphical posi-tion (post-Qarmat).

5.4. Historic glacial phase

This is marked by the fresh moraine laid down by theNeoglacial readvance which culminated in this century(V, Fig. 1) (Kelly, 1980b).

5.5. Marine limits

A relatively well marked marine limit occurs in iso-lated localities throughout the area, correlated by14C dates (reservoir corrected) which demonstrate that itbelongs to the Nuna marine event. It has a low altitudewhich slightly increases from east to west, 35 m a.s.l. atSalisbury Gletscher, 36 m at Narsaarsuk, 40 m on Saun-ders ", 46 m at Iterlak and Booth Sund. The highestdated marine molluscs have an age of 8.9 ka, from 35 ma.s.l. at Booth Sund, but dates from lower localities up to9.2 ka (Booth Sund and Salisbury Gletscher) give anolder minimum age for the marine limit (Table 2; Mornerand Funder, 1990).

These dates associate the Nuna event and its marinelimit with the glacioisostatic response of a late Weich-selian glacial event, identified with the WolstenholmeFjord/inner fjord phase advance. In which case, overmost of the area it was a transgressive event occurring inthe flexural zone of depression outside the ice margin.

On the Carey"er, the marine limit is estimated to be at55 m, but the highest dated Holocene sediments are from15 m a.s.l. Whether the marine limit here can also becredibly associated with the Nuna event, in part dependson the location of the Late Weichselian ice margin in theuninvestigated area to the north (discussed later).

At Booth Sund, there is some evidence of older shore-lines above the Nuna marine limit up to 65 m a.s.l., buttheir origin is uncertain. Demonstrably older marinesediments are not known above 26 m a.s.l. (Iterlak).

6. Age of events

A suggested chronostratigraphy of the Thule area issummarised in Table 7. The most secure point in thisscheme is the age of the Qarmat marine event, welldefined by the revised mean TL date of 116$11 kadiscussed previously, and the yet more selective date forsites which have direct evidence of conditions warmerthan present of 114$14 ka (Fig. 2). The range of dates(91—133 ka) suggests that this event may have been longlasting, correlating with the whole of OIS 5 rather thanthe period of optimum conditions in OIS 5e (Eemian),although the precision of the dating technique does notallow a confident conclusion to be reached. This dates


TABLE 7. Quaternary succession in the Thule area of northwestGreenland

the preceding outer fjord glaciation (Narsaarsuk event)to OIS 6 or Saalian, probably late Stage 6 (see below).

The age of the preceding marine event (Saunders ") isproblematical, with the mean TL date of 167$16 kafalling within OIS 6 (Martinson et al., 1987), when thedeep sea oxygen isotope record implies glacial conditionsexisted. Either the TL dates are generally too young andthe event should be correlated with the warm OIS7 ('190 ka), or they are too old (because of incompletezeroing) and it belongs to the Stage 5/6 boundary, ora Stage 6 age is accepted. Consequently the age of theolder Agpat glacial event and shelf glaciation is alsouncertain and it is tentatively given a minimum age ofStage 6.

The youngest marine event (Nuna) has so far yielded14C dates confined to the Holocene, (9.2 ka, but itcould extend back into the late Weichselian. As discussedabove, this dates the inner fjord glaciation/Wols ten-holme Fjord event to the late Weichselian (OIS 2). Theminimum date for the end of this phase, with retreat ofthe ice behind the present margin is given by a 9.0 ka datefrom Nunatarsuup nuua (Morner and Funder, 1990).Broadly similar 14C dates have been obtained by earlierworkers (Crane and Griffin, 1959; Goldthwaite, 1960;Weidick, 1978). Dating of the Kap Abernathy glacialevent is poorly constrained by the above age estimates, tobetween early Stage 2 to late Stage 5.

7. Palaeo-oceanographic conditions

As described previously (Funder, 1990), the Thule areatoday lies at the northern boundary of penetration bysubarctic Atlantic water into Baffin Bay, which flowsnorthwards along the Greenland coast as the WestGreenland current. The area is therefore well suited tomonitor palaeo-oceanographic changes in Baffin Bay.

Within the macroinvertebrate fossil faunas are speciesor groups of species which can be regarded as indicatorsof particular water masses, based on their modern watermass requirements, as deduced from Vibe (1950), Thor-son (1951), Macpherson (1971), Lubinsky (1980), Dale(1985), and Dale et al. (1989):

(i) Subarctic species, close to their northern limits inthe Thule area today (SA#, Table 5), e.g. Mytilus edulis,Chlamys islandica, Balanus crenatus, ½oldia hyperborea,which indicate that Atlantic water was being advectedinto Baffin Bay at least as far north as Thule.

(ii) Other subarctic species which do not reach theThule area today (SA-, Table 5), e.g. Balanus balanoides,¹rophon truncatus, Paradexiospira vitrea, Berenicea arc-tica, which show that the influx of Atlantic water wasgreater than today’s.

(iii) Polar water indicators are more problematical,because the shallow water Arctic species often extendtheir distribution to the south in deeper waters. Thus, therare ¸imatula hyperborea known from the fossil recordprobably occurs in the area today, although it has notbeen recorded (A#, Table 5). It is more surprising thatthe investigation of the recent fauna in the area (Vibe,1939, 1950) has failed to find Portlandia arctica andPalliolum greenlandicum (A-, Table 5), which are ubiqui-tous in Arctic shallow water benthic communitiesand which lived in the fjords in the Thule area in theEarly Holocene (Funder, 1990). Portlandia arcticatoday is absent from western Greenland, except for anisolated occurrence south of Thule, while Palliolum oc-curs in northern West Greenland, south of Thule (Thor-son, 1951). Both are common in the fjords of EllesmereIsland and Baffin Island (Syvitski et al., 1989; Aitken andFournier, 1993). Therefore, even though systematic in-vestigations might turn up these two arctic species in thefjords of the Thule area, the area can be considered to becritical for them and their occurrence in fossil faunasindicates periods with a larger influx of Polar water thantoday.

Faunas with subarctic species wholly or largely extinctin the Thule area today (SA-, Table 5) occur only insediments from the Qarmat event. Balanus balanoideswas abundant then along the coasts and Mytilus eduliswas common. The latter species also occurred in the earlyHolocene, Nuna event but is now rare in the area (e.g.Theisen, 1973). It was absent from the Saunders " event,although non-extinct subarctic species did occur, to-gether with the extinct arctic species Portlandia arcticaand Palliolum greenlandicum.

Thus, the evidence is that subarctic Atlantic waterreached northern Baffin Bay during the three marineintervals known from the Thule area. However, condi-tions during the Qarmat event were somewhat warmerthan at present, whilst the preceding Saunders " eventwas either slightly cooler, based on the macrofauna, or


warmer than at present, according to the foraminifera.The presence of an intertidal-littoral fauna during theQarmat event, whereas the beach sediments from theSaunders " event are devoid of shells, supports the for-mer conclusion, since the present northern boundary ofthe subarctic region in the vicinity of Thule marks thenorthern boundary for stationary littoral benthos due tothe development of shoreline ice (ice foot) (Madsen, 1940;Dale et al., 1989). Notably, in the Nuna event, the onset ofthe subarctic conditions due to the West Greenland Cur-rent had occurred by the relatively early date of 9.2 kaBP.

In addition, the remains of locally extinct plants andinsects in the deltaic sediments from the upper part of theQarmat member indicate that land temperatures thenwere also high, with estimated July temperatures 4°Cabove those of the present (Bennike and Bocher, 1992).

8. Discussion

The stratigraphy of the Thule area, summarised inTable 7, raises a number of wider questions. The first ofthese is whether, over the same time interval, there havebeen other periods of high relative sea-level with an‘‘interglacial’’ type circulation in Baffin Bay, transportingsubarctic water northwards along the West Greenlandcoast, than the three already identified.

One particular point is whether the multiple oscil-lations in the marine environment known to have occur-red elsewhere in OIS 5 should be represented similarly innorthwest Greenland. For example, the abyssal sedi-ments from the Labrador Sea south of Baffin Bay showa conventional development of the sequence, in particu-lar showing several high sea level/warm phases in Subst-ages 5e and 5a (Aksu et al., 1989). However, the oxygenisotope record from within Baffin Bay itself is stronglyaffected by local factors related to inputs of glacialmeltwater and only one incursion of warm water hasbeen identified, in Substage 5e (De Vernal et al., 1987;Hilaire-Marcel et al., 1989). Again, the oxygen isotoperecord from the Summit ice core from the Greenland IceSheet records several warm oscillations in Stage 5, with5e warmer than the Holocene and less warm intervals in5c and 5a, as well as a number of higher frequencyoscillations in 5e (Dansgaard et al., 1993; GRIP Mem-bers, 1993; Bond et al., 1993).

In the Greenland terrestrial stratigraphic record, a se-quence of marine and glacial events that correlate withall the substages of Stage 5 is known only from theAtlantic sector of the Greenland Ice Sheet, at ScoresbySund, east Greenland (Funder et al., 1994). At Thule,according to the new interpretation, these substages havenot been identified within the period of the Qarmatmarine event, which the TL dates indicate extended overall of Stage 5.

Elsewhere in northwest and west Greenland, numer-ous isolated occurrences of marine sediments of probableMiddle or Late Quaternary age are known and they havebeen used locally to define at least seven marine events.However, attempts to correlate and date them faunisti-cally and by amino acid ratios (Funder and Sımonarson,1984; Kelly, 1986; Bennike et al., 1994) have yieldedresults that are at best tentative, due to the uncertaintyassociated with amino-acid dating (Sejrup, 1990; Funderet al., 1992). Only the ice-proximal glaciomarine Lak-sebugt marine event has been luminescence dated, to154—164 ka (4 dates) (Bennike et al., 1994), which com-pare with the Saunders " dates and the oldest dates fromthe Qarmat event suspected of being over-estimates. Sub-arctic conditions were present during several of theseevents but none were clearly warmer than at present.Nevertheless, the Kaffehavn and Svartenhuk marineevents were suggested to be of Stage 5 age (Kelly, 1986;Bennike et al., 1994). Overall, these sites do not provideunequivocal evidence for more periods of interglacialtype circulation in Baffin Bay than recognised at Thule.

A similar conclusion can be reached about the evid-ence from the Canadian side of Baffin Bay. The correla-tion of marine intervals on both sides of Baffin Bay hasbeen discussed several times (e.g. Andrews et al, 1986;Funder, 1989; Funder, 1990; Funder et al., 1992; Miller etal., 1992), based on the recognition that there was wide-spread faunal evidence for a pre-Holocene period withwarm water circulation in Baffin Bay. Due to the limita-tions of the dating methods, as in west Greenland, it isnot possible to advance beyond the general conclusionsof Funder (1990) that the Qarmat event correlates withthe Kogalu aminozone and perhaps even the Loks Landaminozone of Baffin Island.

The second question relates to the position of the lateWeichselian glacial maximum corresponding to lowestglobal sea levels. Over the southern two-thirds of theBaffin Bay sector of the Greenland Ice Sheet the marginreached out onto the inner shelf, defining the Sisimiutstade (Kelly, 1985; Funder and Hansen, 1996). However,the reconstruction of a much less advanced ice margin atThule, relative to the present margin, is in line with theevidence along the northern and northeastern GreenlandIce Sheet margins (Kelly and Bennike, 1992; Funder andHansen, 1996) as well as in the Melville Bugt area im-mediately to the south. The actual position at Thule isuncertain, located either at the Wolstenholme Fjord mo-raine (Fig. 1, IV) or in the vicinity of Kap Abernathy10 km further west (Fig. 1, III).

The apparent extent of the Holocene deglaciation inthe Thule area and adjacent Melville Bugt region to thesouth, relative to the present ice margin, is very slightcompared to the southern sector of the western Green-land Ice Sheet. Both this, and to some extent the appar-ent small Late Weichselian advance, may be partly anexpression of a greater expansion of the north western


sector of the Greenland Ice Sheet in the late Holocenethan elsewhere (Kelly, 1980b), as a result of the area’sproximity to the open water source of moisture providedby the West Greenland Current in Melville Bugt duringa period of lowered air temperatures.

The final point of regional interest is the age of theglaciation of the outer shelf and the islands on it (Carey"er) by the Baffin Bay margin of the Greenland Ice Sheet.In western Greenland generally the shelf glaciation,termed the Fiskebanke Glacial (Kelly, 1985), is tentative-ly given a Saalian/OIS 6 age, as it is at Thule (Agpatevent). In Fig. 6, the ice margin suggested for the shelfglaciation is located at the 400 km bathy- metric contour,equivalent to a grounding line in approximately 300 m ofwater, allowing for eustatic and isostatic effects, which istypical of sections of the East Antarctic Ice Sheet margin(Drewry, 1983). This would indicate an ice thickness ofc. 1000 m in the region of the Carey "er. Reeh (1984)showed the three dimensional form required for a similarice sheet with a grounding line at 600 m present depth, aswell as for a more restricted ice cover with a 200 m(present depth) grounding line, which does not require iceextending south of Smith Sound.

The age of the shelf ice sheet configuration has beendebated by Blake (1992a,b) and Blake et al. (1992, 1996).Their evidence from the Canadian and Greenland side ofSmith Sound clearly indicates the presence there ofa much younger ice stream dated to the end of theWeichselian, from evidence for substantial isostatic upliftcommencing in the early Holocene, with up to 88 m ofemergence on the Greenland side and up to 108—135 mon Ellesmere Island. A shell bearing till cover in the areais identified with this advance. The ice stream originatedat an ice saddle occupying Kane Basin, fed by the coales-cence of expanded Greenland and Inuitian ice sheets oneither side. Support for this configuration of the LateWeichselian ice sheets comes also from evidence ofa northward flowing ice stream north of Kane basin, atthe Arctic Ocean end of Nares Strait, on Hall Land(Kelly and Bennike, 1992) and on Hans" in the middle ofKennedy Channel (De Freitas, 1990).

However, the question remains as to whether thesoutherly flowing ice stream had its margin in the Weich-selian at the entrance to Smith Sound (Fig. 6) or whetherit reached much further south. Blake et al. (1996) haverecorded ice moulded bedrock features from the shelf onboth sides of the trough south of Smith Sound, at waterdepths of 200—400 m in the vicinity of the Carey "er andat 440 m off southeast Ellesmere Is. More importantly,they obtained a core from 823 m depth east of the Carey"er which gave dates of 10.8—10.9 ka for foraminiferaat the contact between upper laminated marine sedi-ments and a lower diamicton. This correlates theupper sedimentary unit with the Nuna event. However,the diamicton could be either a till of late Weichselianage or a glaciomarine sediment bearing an increased

content of ice rafted debris from ice margins whichwere situated much further north and east, within theinner fjords. The latter explanation would be compatiblewith the Saalian (or older) date for the shelf glaciationsuggested here.

In conclusion, our present view is that, since the gla-ciation of the shelves around northern Baffin Bay in OIS6, ice advances of the Thule sector of the Greenland IceSheet have terminated within the fjords, with the ad-vances having amplitudes (beyond the present ice mar-gin) of (50 km. This limited spatial scale makes theirdifferentiation difficult but the record apparently in-cludes advances both late and early in glaciations (Stages6, 4, 2 and potentially the Neoglacial), which may relateto the coincidence of low temperatures and the openwater conditions associated with the presence of thewarm West Greenland current. This type of circulation inBaffin Bay developed on three occasions during this timespan (Stages 6, 5e and 1) with the middle one being thewarmest.

Acknowledgements

The assistance of Keith Bradshaw with the uraniumand thorium analysis is gratefully acknowledged.

References

Aitken, A. E., & Fournier, J. (1993). Macrobenthos communities ofCambridge, McBeth and Itirbilung fiords, Northwest Territories,Canada. Arctic. 46, 60—71.

Andrews, J. T., Aksu, A., Kelly, M., Klassen, R., Miller, G.H., Mode,W.N., & Mudie, P. (1986) Land/ocean correlations during the lastinterglacial/glacial transition, Baffin Island, northwestern NorthAtlantic: A review. Quaternary Science Reviews, 4, 333—355.

Aksu, A. E., de Vernal, A. E., & Mudie, P. J. (1989). High resolutionforaminifer, palynologic and stable isotopic records of Upper Pleis-tocene sediments from the Labrador Sea: Paleoclimatic andpaleoceanographic trends. In S.P. Srivastava et al. (Eds), Proceed-ings of the Ocean Drilling Project, Scientific Results (Vol. 105, pp.617—652).

Bendix-Almgreen, S. E., Bristrup, B., & Nichols, R. L. (1967). Notes onthe geology and geomorphology of the Carey "er, North-westGreenland. Meddelelser om Gr~nland, 164(8), 19.

Bennike, O., & Bocher, J. (1992). Early Weichselian interstadial biotasat Thule, northwest Greenland. Boreas, 21, 111—117.

Bennike, O., Hansen, S. B., Knudsen, K.-L., Penney, D. N., & Rasmus-sen, K. L. (1994). Quaternary marine stratigraphy and geochronol-ogy in central West Greenland. Boreas, 23, 194—215.

Blake, W. Jr. (1975) Glacial geological investigations in northwesternGreenland. Geological Survey of Canada Paper, 75-1A, 435—439.

Blake, W. Jr. (1977). Radiocarbon age determinations from the CareyIslands, northwest Greenland. Geological Survey of Canada Paper,77-1A, 445—454.

Blake, W. Jr. (1987). Geological Survey of Canada radiocarbon datesXXVI. Geological Survey of Canada Paper, 86-7, 60.

Blake, W. Jr. (1992a). Holocene emergence at Cape Herschel, eastcentral Ellesmere Island, Arctic Canada: Implications for ice sheetconfiguration. Canadian Journal of Earth Science, 29, 1958—1980.


Blake, W. Jr. (1992b) Shell-bearing till along Smith Sound, EllesmereIsland—Greenland: Age and significance. Sveriges Geoligiska ºnder-so( kning Serie, Ca 81, 51—88.

Blake, W. Jr., Boucherle, M. M., Fredskild, B., Janssens, J. A., & Smol,J. P. (1992). The geomorphological setting, glacial history andHolocene development of ‘Kap Inglefield S+’, InglefieldLand, northwest Greenland. Meddelelser om Gr~nland, Geoscience,27, 42.

Blake, W. Jr., Jackson, H. R., & Currie, C. G. (1996). Seafloor evidencefor glaciation, northernmost Baffin Bay. Bulletin of the GeologicalSurvey of Denmark, 43, 157—168.

Bond, G., Broecker, W., Johnson, S., McManus, J., Labeyrie, L.,Jouzel, J., & Bonani, G. (1993). Correlation between climate recordsfrom North Atlantic sediments and Greenland ice. Nature, 365,507—508.

Chamberlin, T. C. (1895). Geology. In H. G. Bryant (Ed), The Pearyauxiliary expedition of 1894. Bulletin of the Geographical Club ofPhiladelphia, 1, 167—194.

Crane, H. R., & Griffin, J. B. (1959). University of Michigan radiocar-bon dates IV. Radiocarbon, 1, 173—198.

Dale, J. E. (1985). Recent intertidal molluscs from the east-central coastof Ellesmere Island, Northwest Territories. Geological Survey ofCanada Paper, 85-1B, 319—324.

Dale, J. E., Aitken, A. E., Gilbert, R., & Risk, M. J. (1989) Macrofauna ofCanadian arctic fjords. Marine Geology, 85, 331—358.

Dansgaard, W., Johnson, S. J., Clausen, H. B., Dahl-Jensen, D., Gun-destrup, N. S., Hammer, C. S., Steffenson, J. P., Sveinbj+rnsdottir,A. E., Jouzel, J., & Bond, G. (1993). Evidence for general instabilityof past climate from a 250-kyr ice-core record. Nature, 364,218—220.

Dawes, P. R. (1991). Geological Map of Greenland Sheet 5 ¹hule. Copen-hagen: Greenland Geological Survey.

Dawes, P. R., Frisch, T., & Christie, R. L. (1982). The Proterozoic Thulebasin of Greenland and Ellesmere Island, importance to the NaresStrait debate. In P. R. Dawes & J. W. Kerr, (Eds.), Nares Strait andthe drift of Greenland: A conflict in plate tectonics. Meddelelser omGr~nland, Geoscience, 8, 89—104.

De Freitas, T. A. (1990). Implications of glacial sculpture on HansIsland between Greenland and Ellesmere Island (Nares Strait).Journal of Glaciology, 36, 129—130.

De Vernal, A., Hillaire-Marcel, C., Aksu, A. E., & Mudie, P. J. (1987).Palynostratigraphy and chronostratigraphy of Baffin Bay deepsea cores: Climatostratigraphic implications. Palaeogeography,Palaeoclimatology, Palaeoecology, 62, 96—106.

Drewry, D. J. (Ed.). (1983). Antarctica: Glaciological and geophysicalfolio. Cambridge: Scott Polar Research Institute.

Feyling-Hanssen, R. W. (1980). Microbiostratigraphy of young Ceno-zoic marine deposits of the Qivituq Peninsula, Baffin Island. MarineMicropaleontology, 5, 153—183.

Feyling-Hanssen, R. W. (1990). A remarkable foraminiferal assemblagefrom the Quaternary of northeast Greenland. Bulletin of the Geo-logical Society of Denmark, 38, 101—107.

Feyling-Hanssen, R. W., & Funder, S. (1990). Fauna and flora, 19—32.In S. Funder (Ed), Late Quaternary stratigraphy and glaciology inthe Thule area, northwest Greenland. Meddelelser om Gr~nland,Geoscience, 22, 19—23.

Funder, S. (1989). Quaternary geology of the ice-free areas and adjacentshelves of Greenland. In R. J. Fulton (Ed), Quaternary geology ofCanada and Greenland. ¹he geology of Canada (Vol. 1, pp. 775—783).Montreal: Geological Survey of Canada. (also ¹he Geology of NorthAmerica, K-1. Geological Society of America).

Funder, S. (Ed.). (1990). Late Quaternary stratigraphy and glaciology inthe Thule area, northwest Greenland. Meddelelser om Gr~nland,Geoscience, 22, 63 pp.

Funder, S., & Sımonarson, L.A. (1984). Bio- and aminostratigraphy ofsome Quaternary marine deposits in West Greenland.Canadian Journal of Earth Science, 21, 843—852.

Funder, S., & Houmark-Nielsen, M. (1990). Local correlation and eventstratigraphy. In S. Funder (Ed), Late Quaternary stratigraphy andglaciology in the Thule area. Meddelelser om Gr~nland, Geoscience,22, 40—45.

Funder, S., & Hansen, L. (1996) The Greenland ice sheet — a model forits culmination and decay during and after the last glaciation max-imum. Bulletin of the Geological Society of Denmark, 42, 137—152.

Funder, S., Hjort, C., & Kelly, M. (1992). Isotope Stage 5 (73—130 ka) inGreenland. Quaternary International, 10-12, 107—122.

Funder, S., Hjort, C., & Landvik, J. (1994). The last glacial cycles in EastGreenland, an overview. Boreas, 23, 283—293.

Goldthwaite, R. P. (1960). Study of an ice cliff in Nunatarssuaq, Green-land. ¹echnical Report Cold Regions Research and Engineering¸aboratory, 47, 63 pp.

GRIP Members (1993). Climate instability during the last interglacialperiod recorded in the GRIP ice core. Nature, 364, 203—207.

Gudina, V. I. (1966). Foraminifera and stratigraphy of the NorthwestSiberian Quaternary. Akademia Nauk SSSR, Siberian Department,Institute of Geology and Geophysics, 132. (In Russian).

Gudina, V. I., & Evserov, U. Y. (1973). Stratigraphy and foraminifera ofLate Pleistocene of Kola Peninsula. Akademia Nauk SSSR, SiberianDepartment, Institute of Geology and Geophysics, 175, 148 (in Rus-sian).

Hilaire-Marcel, C., de Vernal, A. E., & Aksu, A. E. (1989). High resolu-tion isotopic and micropaleontological studies of Upper Pleistocenesediments at ODP site 645, Baffin Bay. In S.P. Srivastava, et al.(Eds.), Proceedings of the Ocean Drilling Project, Scientific Results,(Vol. 105, pp. 599—616).

Houmark-Nielsen, M., Kelly, M., Landvik, J., & Sorby, L. (1990). Thepre-Holocene Quaternary: Lithostratigraphic and geomorphic evid-ence. In S. Funder (Ed.), Late Quaternary stratigraphy and glaciologyin the Thule area. Meddelelser om Gr~nland, Geoscience, 22, 8—18.

Kelly, M. (1980a). Preliminary investigations of the Quaternary ofMelville Bugt and Dundas, north-west Greenland. RapportGr~nlands Geologiske ºnders~gelse, 100, 33—38.

Kelly, M. (1980b). The status of the Neoglacial in Greenland. RapportGr~nlands Geologiske ºnders~gelse, 96, 24.

Kelly, M. (1985). A review of the Quaternary geology of westernGreenland. In J.T. Andrews (Ed.), Quaternary environments easternCanadian Arctic, Baffin Bay and western Greenland (pp. 461—501).Boston: Allen and Unwin.

Kelly, M. (1986). Quaternary, pre-Holocene, marine events of westernGreenland. Rapport Gr~nlands Geologiske ºnders~gelse, 131, 23.

Kelly, M., & Bennike, O. (1992). Quaternary geology of western andcentral Greenland. Rapport Gr~nlands Geologiske ºnders~gelse,153, 34.

Krinsley, D. B. (1963). In W. E. Davies, D. H. Krinsley & A. H. Nichol,(Eds.), Geology of the North Star Bugt area, northwest Greenland.Meddelelser om Gr~nland, 162, 48—66.

Kronborg, C., & V. Mejdahl, (1990). Thermoluminescence dating ofmarine sediments from Saunders " and Qarmat. In S. Funder (Ed.),Late Quaternary stratigraphy and glaciology in the Thule area.Meddelelser om Gr~nland, Geoscience, 22, 33—36.

Loeblich, A. R., & Tappan, H. (1953). Studies of Arctic Foraminifera.Smithsonian Miscellaneous Collections, 121, 1—150.

Lubinsky, I. (1980). Marine bivalve molluscs of the Canadian centraland eastern Arctic: Faunal composition and zoogeography. Cana-dian Bulletin of Fisheries and Aquatic Sciences, 207, 111.

Macpherson, E. (1971). The marine molluscs of Arctic Canada. Prosob-ranch Gastropods, Chitons and Scaphopoda. National Museum ofNatural Sciences Publications in Biological Oceanography, 3, 149.

Madsen, H. (1940). A study of the littoral fauna of northwest Green-land. Meddeleser om Gr~nland, 124, 24.

Martinson, D. G., Pisias, N. G., Hays, J. D., Imbrie, J., Moore, T. C.,& Shackleton, N. J. (1987). Age dating and the orbital theory of theice ages: Development of a high-resolution 0—300,000 year chrono-stratigraphy. Quaternary Research, 27, 1—29.


Mejdahl, V., Shlukov, A. I., Shakhovets, S. A., Voskovskaya, L. T.,& Lyashenko, H. G. (1992). The effect of shallow traps: a possiblesource of error in TL dating of sediments. Ancient ¹¸, 10, 22—25.

Mejdahl, V., & Funder, S. (1994). Luminescence dating of Late Quater-nary sediments from East Greenland. Boreas, 23, 525—535.

Meldgaard, S., & Knudsen, K. L. (1979). Metoder til indsamling ogparbejdning af pr+ver til foraminifer—analyser. Dansk Natur— Dansk Skole, As rsskrift, 1979, 48—57.

Miller, G. H., Funder, S., de Vernal, A., & Andrews, J. T. (1992). Timingand character of the last interglacial-glacial transition in the easternCanadian Arctic and northwest Greenland. In P.U. Clark & P. Lea,(Eds.), The last interglaciation-glaciation in North America. Geo-logical Society of America, Special Paper, 270, 223—231.

Morner, N.-A., & Funder, S. (1990). C-14 dating of samples collected bythe NORDQUA 86 expedition, and notes on the marine reservoireffect. In S. Funder (Ed.), Late Quaternary stratigraphy and glaciol-ogy in the Thule area. Meddelelser om Gr~nland, Geoscience, 22,57—59.

Mudie, P. J., Keen, C. E., Hardy, I. A., & Vilks, G. (1984). Multivariateanalysis and quantitative paleoecology of benthic foraminifera insurface and Late Quaternary shelf sediments, northern Canada.Marine Micropaleontology, 8, 283—313.

Murray, J. W. (1991). Ecology and Palaeoecology of Benthic Foraminif-era. 397 pp. London: Longman.

Osterman, L. E., & Andrews, J. T. (1983). Changes in glacial-marinesedimentation in core HU77-159, Frobisher Bay, Baffin Island,N.W.T: A record of proximal, distal, and ice-rafting glacial- marineenvironments. In B. F. Molnia (Ed.), Glacial-marine sedimentation,(pp. 451—493). New York: Plenum Press.

Osterman, L.E., & Nelson, A.R. (1989). Latest Quaternary and Holo-cene paleoceanography of the eastern Baffin Island continental

shelf, Canada: Benthic foraminiferal evidence. Canadian Journal ofEarth Science, 26, 2236—2248.

Phleger, F. B. (1952). Foraminifera distribution in some sediment sam-ples from the Canadian and Greenland Arctic. Cushman Foundationof Foraminiferal Research, Contribution, 3, 80—89.

Reeh, N. (1984). Reconstruction of the glacial ice covers of Greenlandand the Canadian Arctic Islands by three dimensional, perfectlyplastic ice-sheet modelling. Annals of Glaciology, 5, 115—121.

Sejrup, H.-P. (1990). Amino acid geochronology. In S. Funder, (Ed.),Late Quaternary stratigraphy and glaciology in the Thule area.Meddeleser om Gr~nland, Geoscience, 22, 36—39.

Syvitski, J. P. M., Farrow, G. E., Atkinson, R. J. A., Moore, P. G.,& Andrews, J. T. (1989). Baffin Island fjord macrobenthos: Bottomcommunities and environmental significance. Arctic, 42, 232—247.

Theisen, B. F. (1973). The growth of Mytilus edulis L. (Bivalvia) fromDisko and Thule district, Greenland. Opheila, 12, 59—77.

Thorson, G. (1951). The Godthaab expedition 1928, Scaphopoda,Placophora, solenogastres, gastropoda, prosobranchiata, lamellib-ranciata. Meddeleser om Gr~nland, 81, 117.

Vibe, C. (1939). Preliminary investigations on shallow water animalcommunities in the Upernavik- and Thule-districts (northwestGreenland). Meddeleser om Gr~nland, 124, 42.

Vibe, C. (1950). The marine mammals and the marine fauna in theThule district (northwest Greenland) with observations on ice con-ditions in 1939-41. Meddeleser om Gr~nland, 150, 115 pp.

Walton, W.R. (1964). Recent foraminiferal ecology and paleoecology, InJ. Imbrie & N.D. Newell, (Eds.), Approaches to P.aleoecology (pp.151—237). New York: Wiley.

Weidick, A. (1978). Comments on radiocarbon dates from northernGreenland made during 1977. Rapport Gr~nlands Geologiskeºnders~gelse, 90, 124—128.


Quaternary glacial and marine environmental history of northwest Greenland: a review and reappraisal

Documents