Sedimentology and chronology of the advance and retreat of the last
British-Irish Ice Sheet on the continental shelf west of Ireland
Jared L. Peters a, *
, Sara Benetti a
, Paul Dunlop a
, Colm O Cofaigh b
, Steven G. Moreton c
,
Andrew J. Wheeler d
, Christopher D. Clark e
a
School of Geography and Environmental Sciences, University of Ulster, Cromore Road, Coleraine, BT52 1SA, Northern Ireland, UK
b
Department of Geography, Durham University, Durham, DH1 3LE, UK
c
Natural Environment Research Council, Radiocarbon Facility, East Kilbride, Scotland, G75 OQF, UK
d
School of Biological, Earth and Environmental Sciences, University College Cork, Cork, Ireland
e
Department of Geography, University of Sheffield, Sheffield, S10 2TN, UK
a r t i c l e i n f o
Article history:
Received 12 December 2015
Received in revised form
9 March 2016
Accepted 11 March 2016
Available online 4 April 2016
Keywords:
British-Irish Ice Sheet
Retreat rate
Grounding zone wedge
Readvance
Radiocarbon
Sedimentology
Geomorphology
Ice shelf
Killard Point Stadial
Nahanagan (Younger Dryas) Stadial
a b s t r a c t
The last British-Irish Ice Sheet (BIIS) had extensive marine-terminating margins and was drained by
multiple large ice streams and is thus a useful analogue for marine-based areas of modern ice sheets.
However, despite recent advances from investigating the offshore record of the BIIS, the dynamic history
of its marine margins, which would have been sensitive to external forcing(s), remain inadequately
understood. This study is the first reconstruction of the retreat dynamics and chronology of the western,
marine-terminating, margin of the last (Late Midlandian) BIIS. Analyses of shelf geomorphology and core
sedimentology and chronology enable a reconstruction of the Late Midlandian history of the BIIS west of
Ireland, from initial advance to final retreat onshore. Five AMS radiocarbon dates from marine cores
constrain the timing of retreat and associated readvances during deglaciation. The BIIS advanced without
streaming or surging, depositing a bed of highly consolidated subglacial traction till, and reached to
within ~20 km of the shelf break by ~24,000 Cal BP. Ice margin retreat was likely preceded by thinning,
grounding zone retreat and ice shelf formation on the outer shelf by ~22,000 Cal BP. This ice shelf
persisted for 2500 years, while retreating at a minimum rate of ~24 m/yr and buttressing a >150-km
long, 20-km wide, bathymetrically-controlled grounding zone. A large (~150 km long), arcuate, flattopped
grounding-zone wedge, termed here the Galway Lobe Grounding-Zone Wedge (GLGZW), was
deposited below this ice shelf and records a significant stillstand in BIIS retreat. Geomorphic relationships
indicate that the BIIS experienced continued thinning during its retreat across the shelf, which led
to increased topographic influence on its flow dynamics following ice shelf break up and grounding zone
retreat past the GLGZW. At this stage of retreat the western BIIS was comprised of several discrete,
asynchronous lobes that underwent several readvances. Sedimentary evidence of dilatant till deposition
suggests that the readvances may have been rapid and possibly associated with ice streaming or surging.
The largest lobe extended offshore from Galway Bay and deposited the Galway Lobe Readvance Moraine
by <18,500 Cal BP. Further to the north, an ice lobe readvanced at least 50 km offshore from Killary
Harbour, possibly by 15,100 Cal BP. The existing chronology currently does not allow us to determine
conclusively whether these readvances were a glaciodynamic (internally-driven) response of the ice
sheet during deglaciation or were climatically-driven. Following the <18,500 Cal BP readvance, the
Galway Lobe experienced accelerated eastward retreat at an estimated rate of ~113 m/yr.
© 2016 Elsevier Ltd. All rights reserved.
1. Introduction
Marine-terminating sectors of large ice sheets are considered to
be potentially inherently unstable and sensitive to climatic and
ocean forcing(s) (Cook et al., 2005; Rignot et al., 2010; Glasser et al.,
2011). Palaeo-glaciological data in the form of glacial geology and
* Corresponding author.
E-mail address: Peters-J@email.ulster.ac.uk (J.L. Peters).
Contents lists available at ScienceDirect
Quaternary Science Reviews
journal homepage: www.elsevier.com/locate/quascirev
http://dx.doi.org/10.1016/j.quascirev.2016.03.012
0277-3791/© 2016 Elsevier Ltd. All rights reserved.
Quaternary Science Reviews 140 (2016) 101e124
geomorphology is increasingly used to constrain numerical ice
sheet models of present and future ice sheet change (e.g. Bentley
et al., 2010; Robinson et al., 2011; Lecavalier et al., 2014). A logical
step towards improved model accuracy will come from advancing
our knowledge of previous ice sheet behaviour along marineterminating
margins. By developing a thorough understanding of
ice sheet behaviour during the Pleistocene, from maximum extent
to final retreat, palaeoglaciological assessments enable improved
understanding of the controls on modern ice sheet dynamics and
provide a means of testing the veracity of predictive models. The
last British-Irish Ice Sheet (BIIS) is a potential analogue for marinebased
ice sheet retreat because it consisted largely of marine-based
(~300,000 km3
) ice with termini fed by ice streams and flanked by
ice shelves (Clark et al., 2012a).
Ongoing efforts to reconstruct the size, dynamic behaviour and
chronology of the last BIIS have recently focused on its marineterminating
margins on the continental shelves around Ireland
and Britain (e.g. Bradwell et al., 2008; O Cofaigh et al., 2010; Peters
et al., 2015). This extends from over a century of largely
terrestrially-based research on the last BIIS (Clark et al., 2012a). In
the south of Ireland, these efforts have extended the maximum
southern position of the ice sheet (Fig. 1) hundreds of kilometres
beyond previous reconstructions (Scourse et al., 1990; O Cofaigh
and Evans, 2007; Praeg et al., 2015). Marine research north of
Scotland provides sedimentary evidence for a coalescence of the
BIIS with the Fennoscandian Ice Sheet (Sejrup et al., 1994) and
recent analyses of bathymetric data provide evidence for grounded
ice extending across the West Shetland Shelf (Bradwell et al., 2008).
West of Ireland, geomorphic analyses of new bathymetric data
provide compelling evidence for the extension of grounded ice to
the shelf break west of the Malin Sea and Donegal Bay during the
LGM and subsequent lobate readvances during deglaciation
(Benetti et al., 2010; Dunlop et al., 2010; O Cofaigh et al., 2010;
Fig. 1). Geomorphic, sedimentary and micropaleontological analyses
show that the last BIIS reached the Porcupine Bank, west of
Ireland (Fig. 1), and that its behaviour there was dynamic, fluctuating
in extent and forming an ice shelf over the Slyne Trough
(Peters et al., 2015). These offshore studies have led to a general
shift in the prevailing consensus on BIIS maximum extent from a
marine margin that did not extend far past the modern western
Irish coastline (Bowen et al., 1986) to one that predominantly
reached the edge of the continental shelf (Clark et al., 2012a; Fig. 1).
Despite recent advances in understanding of the marineterminating
sectors of the BIIS, a detailed reconstruction of their
retreat behaviour that incorporates chronologically constrained
marine sediment analyses has yet to be established. The aim of this
study is to reconstruct the dynamic behaviour of the marineterminating
margin of the last BIS on the continental shelf west
of Ireland and to provide chronological constraints on its retreat.
This is achieved by presenting new sedimentary data from fourteen
sediment cores (three were used in a previous study; Peters et al.,
2015; Table 1) sampled from the continental shelf west of
counties Mayo, Galway, Clare and Kerry, Ireland (Figs. 1 and 2).
Using new sedimentological data and detailed geomorphic analyses
of bathymetric data, we establish a regional stratigraphy for
the western Irish continental shelf that records the marine-based
advance, retreat and subsequent readvances of the last BIIS (Late
Midlandian; marine isotope stage 2). Five new accelerator mass
spectrometer (AMS) radiocarbon dates (Table 2) provide associated
chronological control including calculating retreat rates during
deglaciation of the shelf.
2. Methods
This study uses single beam bathymetric data compiled by the
Norwegian-developed Olex (www.olex.no) sonar data management
software to analyse the geomorphology of the continental
shelf west of Ireland. The sonar data are a compilation of
voluntarily-contributed, geolocated sonar measurements and, in
areas of adequate coverage, produces a raster of 5-m resolution
cells that typically convey vertical data with a resolution of 1 m
(www.olex.no; Bradwell et al., 2008). The bathymetric data is
presented as 2D hillshaded surfaces and a series of seafloor profiles
to analyse and characterise the geomorphology of glacial landforms
(Fig. 2).
Fourteen vibro-cores (see Table 1) were analysed, three of which
were used by Peters et al. (2015; Table 1). Cores with the designation
CE10008 (Table 1) were collected during the CE10008
research cruise conducted in 2010 with the RV Celtic Explorer; cores
labelled CE14004 were collected during the “WICPro” (West Ireland
Coring Program) research cruise conducted in 2014 with the RV
Celtic Explorer. X-radiographs of the cores were developed, usually
prior to splitting, using a CARESTREAM DRX Evolution system at
Ulster University, Jordanstown. The x-radiographs reveal sedimentary
structures and clast or dropstone presence in the cores
that may be unidentifiable from visual inspection of the core section
alone. Structures visible in the x-radiographs are displayed as
sketches in the sediment logs and illustrative x-radiograph examples
are also provided for most lithofacies. Sediment physical
properties (wet bulk density and magnetic susceptibility) were
measured prior to splitting using a GEOTEK©
multi sensor core
logger at the National University of Ireland, Maynooth. These data
are displayed in the sediment logs along with mean values calculated
for each core that allow intra-core trends to be seen and intercore
comparisons to be made. When large peaks (from individual
clasts or section breaks) in the data skew the magnetic susceptibility
mean, the data are smoothed by removing outlying values;
the excluded measurements are highlighted in the logs. After the
cores were split, sediment shear strength was measured using an
Impact©
shear vane, calibrated by MCC©
, at intervals typically
guided by lithofacies but generally 15 cm. Areas of high clast
density or with clast-supported deposits were avoided while collecting
shear strength measurements because clast contact with
the vane generates spurious results.
The sediment cores were split and stored at 4 C at Ulster University,
Coleraine. Visual and x-radiograph inspections, aided by
physical property analyses, identified twelve lithofacies from which
1-cm slabs were subsampled. Subsampling intervals were typically
guided by lithofacies descriptions (cf. Kilfeather et al., 2011), usually
with at least one subsample per lithofacies and often two, which
enabled improved definition of contacts and intra-lithofacies
changes. A sampling interval of 10 cm was maintained for most
of core 10-40 (the longest core examined; Table 1). Subsamples
were analysed for water content, grain size and relative abundance
of Elphidium excavatum forma clavatum (Feyling-Hanssen, 1972).
Lithic grains >1 mm were removed by dry sieving and reported as a
percent by weight of the total sediment sample. The <1 mm fraction
was analysed by laser granulometry using a MALVERN Mas-
tersizer©
at Trinity College, Dublin, at intervals guided by lithofacies
descriptions. In cores 10-40 and 14-59 the number of lithic grains
>1 mm were counted and are displayed as number/gram of sediment
(cf. Grobe, 1987), which allows a comparison to the lithic
grains >1 mm reported as mass percent.
Elphidium Clavatum specimens were identified using the
morphological characteristics outlined by Feyling-Hanssen (1972).
This species is used for this study because it is well-documented as
an opportunistic, arctic-subarctic species that often dominates
benthic foraminiferal populations in recently deglaciated marine
environments (Hald et al., 1994). E. clavatum relative abundance is
calculated as a percentage of the total number of well-preserved
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124102
foraminifera tests in an aliquot. Aliquots were split using a Green
Geological©
microsplitter to yield subsamples with !300 relatively
well-preserved benthic foraminifer tests (cf. Melis and Salvi, 2009)
and sieved to remove the <125 mm fraction because the sediment
below this size contained very few foraminifer tests, most of which
were unidentifiable, likely allochthonous fragments (cf. Peters
et al., 2015). E. clavatum relative abundances reported by Peters
et al. (2015) are referred to for core 10-45 and were produced
from aliquots with !200 benthic foraminifer tests.
Five calcareous biogenic samples were collected for AMS
radiocarbon analysis from sediment horizons interpreted to record
BIIS retreat or readvance across the continental shelf west of Ireland
and three previously-reported ages (Peters et al., 2015) are recalibrated
for use in this study. Samples were dated at the NERC
Radiocarbon Facility, Scotland, and the Poznan Radiocarbon Laboratory,
Poland. The dated material was unabraded and unbroken
and comprised coral fragments, paired and unpaired bivalve shells
and mixed benthic foraminiferal tests (sample photographs are
provided in sediment logs). Results are presented as conventional
radiocarbon (14
C BP) and calibrated (Cal BP) ages; the calibrated 2s
median results are referenced in the text. Radiocarbon ages were
calibrated at 2s confidence level with Calib©
7.1 software (Stuiver
and Reimer, 1993), using the Marine13 calibration dataset
(Reimer et al., 2013) and with the DR set as 0 ± 50 for radiocarbon
Fig. 1. Regional map locating the study area (shown in Fig. 2) and core locations amongst prominent glacial landforms, ice streams and the last BIIS's maximum extent (Sejrup et al.,
2005; Scourse et al., 2009; Clark et al., 2012a; Peters et al., 2015). The ‘Previously accepted Killard Point ice limit’ is compiled from reconstructions by McCabe et al. (1998) in Ireland
and the theoretical model of BIIS extent at approximately the 17 ka BP isochrone synthesised by Clark et al. (2012a).
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 103
ages that fall between the Younger Dryas and Heinrich 1 cold intervals
(~17.0e12.0 ka BP, Broecker et al., 1989; Bond et al., 1999;
Bowen et al., 2002) and a DR of þ300 ± 50 for radiocarbon ages
that fall between Heinrich 1 and Heinrich 2 (25.0e17.0 ka BP,
Dowdeswell et al., 1995).
3. Results
3.1. Geomorphology
Ridges with various morphological characteristics are found on
the continental shelf offshore of counties Mayo, Galway, Clare and
Kerry, Ireland (Fig. 2). The ridges on the Porcupine Bank and Slyne
Trough (the westernmost ridges) are sinuous and narrow (Fig. 2a,
b). These ridges were shown to have asymmetric profiles and orientations
that roughly align with corrugations on a larger, northsouth
trending arcuate ridge located <10 km to the east (Peters
et al., 2015; Fig. 2). This large (~80 km long), corrugated, arcuate
ridge on the outer continental shelf has an asymmetric profile that
narrows and steepens northward to its un-corrugated northernmost
extent (Figs. 2a and 3c). Its profile asymmetry is characterised
by short, steep western flanks and longer eastern flanks (Fig. 3b, c).
The ridge is truncated near its middle, but appears to be/have been
continuous based on the arc and orientations of the discrete segments
(Fig. 2a).
East (landward) of the corrugated ridge, the continental shelf is
covered by a complex of bisecting (i.e. truncated) arcuate ridges
(Fig. 2). These ridges have a similar, convex-seaward morphology to
the corrugated ridge but have disparate radii and profiles (Figs. 2a
and 3). The southernmost, largest of these ridges is poorly defined
to the south and is truncated by a smaller ridge to the north (Fig 2).
This large ridge is ~150 km long and up to 22 km wide (Figs. 2 and
3bed). It is distinctly flat-topped and reaches heights above the
surrounding seafloor of up to 20 m (Fig. 3bed). The western edge of
this ridge (Fig. 3, bold grey line iii) is flanked by a continuous, lowrelief
(~11 m high, Fig. 3) landform that is also flat-topped but
distinctly lower than the main ridge crest (Figs. 2 and 3). East of and
nested within this large ridge is a smaller arcuate ridge that is
discernible in the Olex data despite its relatively poorly defined
boundaries (Fig. 2a). Although it is relatively ambiguous, this
smaller ridge seems to be ~120 km long and has a southern extent
that reaches ~10 km farther south than the poorly-defined southern
margin of the larger, flat-topped ridge (Fig. 2a).
North of the large, flat-topped ridge are two smaller
(~40e45 km long) arcuate ridges (Fig. 2). The smallest (i.e. tightest
radius) of these ridges is situated between the large, flat-topped
ridge and the northernmost ridge and appears to truncate both of
the neighbouring landforms (Fig. 2a). This small ridge is situated
east (landward) of the flat-topped, low-relief landform that flanks
the large ridge. The small ridge is ~20 m tall and reaches ~106 mbsl
in height, making it the bathymetrically highest ridge in the study
area (Fig. 3a). It is asymmetric and is characterised by relatively
steep western side but gentle eastern side (Fig. 3a). The northernmost
ridge has a wider radius and is truncated by the small ridge
but is otherwise geomorphically similar (Fig. 2a, d).
Much of the seafloor on the continental shelf is furrowed,
particularly areas shallower than ~350 mbsl located west of the
large, flat-topped arcuate ridge and in the Slyne Trough (Figs. 2 and
3d). The furrows are abundant on the continental shelf in areas
approximately <220 mbsl and cover both the arcuate ridges and the
surrounding seafloor (Fig. 2b, c, d). The largest and deepest furrows
occur in the Slyne Trough (Figs. 2b, c, 3d), where they have depths
of >10 m and terminate against bathymetric highs (Peters et al.,
2015). The furrows in the Slyne Trough often have relatively
poorly defined southern termini and distinct, kettle-like depressions
or pits at their northern termini (Fig. 2c, d). Some of the
large Slyne Trough furrows have corrugations at their base (Fig. 2c)
and are flanked by berms (Fig. 3d).
3.2. Sedimentology
Twelve lithofacies are identified by visual inspection, x-radiograph
analysis (Fig 4) and sediment physical property measurements.
These lithofacies are built into four lithofacies associations
as described in the following subsections.
3.2.1. LFA 1: Diamicton (Dmmc, Dmm and Dms) and compact,
deformed mud (Fm and Fl(s))
Highly consolidated diamicton (Dmmc) with shear strengths
from 50 to 197 kPa (note 197 kPa is the shear vane's maximum
measureable stress and thus measurements at this value are
regarded as minima) was recovered at the base of six cores (10-38,
10-42, 10-44, 14-36, 14-42, 14-53) and ranges from at least
35e70 cm in thickness (Fig. 5). This sediment consists of small to
large pebble-sized clasts with variable roundness supported by a
matrix of muddy sand or sandy mud with up to 64% sand in 10-42
and 14% sand in 14-36 (Fig. 5). The matrix is grey or nearly black in
colour (2.5Y 6/2, 2.5Y 5/2, 5Y 3/1, 5Y 2.5/1). Occasionally clasts
show a preferential alignment in x-radiographs with clast a-axes
oriented ~40 from horizontal (Fig. 4a). This lithofacies is characterised
by high wet bulk densities (!2.1 g/cm3
), low water content
(typically <15% and as low as 2.26% in core 10-38) and low magnetic
susceptibility (~13e21$10À8
SI)dwith the exception of core
14-42 (the southernmost core), which reaches 107$10À8
SI (Fig. 5).
Table 1
Sediment core information.
Core Abbreviation Latitude (N) Longitude (W) Water depth (m) Core length (m) Reference
CE10008_35 10-35 53
58.38090
11
13.78090
269.0 0.81 This paper
CE10008_36 10-36 53
58.36220
11
13.84630
234.4 2.01 This paper
CE10008_38 10-38 53
51.94920
10
42.34910
123.5 0.81 This paper
CE10008_40 10-40 53
40.20680
10
40.8039 134.8 4.0 This paper
CE10008_42 10-42 53
46.26550
12
38.65420
298.5 1.76 Peters et al., 2015
CE10008_44 10-44 53
38.91500
12
16.88180
295.1 2.30 Peters et al., 2015
CE10008_45 10-45 53
37.74870
12
08.07130
293.7 2.94 Peters et al., 2015
CE14004_36 14-36 53
02.36890
11
38.11940
155.6 1.65 This paper
CE14004_41 14-41 53
01.07130
10
52.58890
138.7 0.84 This paper
CE14004_42 14-42 52
27.96520
10
59.58060
126.0 0.81 This paper
CE14004_53 14-53 53
23.98690
11
01.12000
145.2 1.28 This paper
CE14004_54 14-54 53
38.11630
11
12.56200
174.4 1.47 This paper
CE14004_57 14-57 53
40.28680
10
53.08420
150.9 1.0 This paper
CE14004_59 14-59 53
35.83040
10
37.31720
132.0 3.43 This paper
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124104
In core 10-44 a well preserved coral fragment (Lophelia pertusa)
was recovered from the Dmmc lithofacies. However, in cores 10-38
and 14-36 this diamicton contains little or no observed biogenic
material (Fig. 5). Farther offshore (i.e. farther west), in cores 10-42
and 10-44 the Dmmc lithofacies contains highly abraded or broken
foraminifer tests but little to no poorly preserved macrofaunal test
material.
The base of cores 10-35, 10-40, 14-57 and 14-59 is composed of
at least 15e302 cm of poorly sorted, matrix-supported diamicton
with moderate consolidation (30e50 kPa) (Fig. 5). The same diamicton
overlies the Dmmc lithofacies in cores 14-36 and 14-53,
where it has a gradational lower contact (Fig. 5). The diamicton is
Fig. 2. (a) Study area map showing the Olex bathymetric dataset used for the geomorphic analyses and core locations relative to modern land. The topographic DEM (NASA, Shuttle
Radar Topography Mission) reveals regional, bedrock-controlled trends in geomorphology. The labelled transects delineate the seafloor profiles shown in Fig. 3. Isobaths are shown
every 100 mbsl and created from Irish National Seafloor Survey (INSS, www.infomar.ie) data and the United Kingdom Hydrographic Office (UKHO; www.ukho.gov.uk) data. Yellow
arrows delineate the western edge of the low, flat-topped ridge (Fig. 3ii); red arrows delineate the western edge of the large flat-topped ridge (Fig. 3iii); black dotted lines delineate
the eastern edge of the smaller, nested arcuate ridge. (bee) Classified (20-m) hillshaded Olex data showing: (b) detail of the Slyne Trough Moraines and the corrugated southern
section of the larger arcuate ridge on the outer shelf; the arrow marks an area of intense iceberg scouring and lodgement and dashed, white line highlights the geographic
relationship between the moraines and corrugations (Peters et al., 2015); (c) detail of a large seafloor furrow with a corrugated base, gradual, progressively-deepening southern
terminus and abrupt, pit-like northern terminus (marked by arrow); the white ‘X’ highlights noise in the Olex data that manifests horizontally across the data; (d) sketch of furrows
and pits in (c). Abbreviations: WIM ¼ West Ireland Moraine, GLGZW ¼ Galway Lobe Grounding Zone Wedge, GLRM ¼ Galway Lobe Readvance Moraine, BB ¼ Blacksod Bay,
CB ¼ Clew Bay, KB ¼ Killary Bay, GB ¼ Galway Bay, SRE ¼ Shannon River Estuary. (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 105
typically massive (Dmm) with the exception of crude stratification
(Dms) visible on x-radiographs at the base of cores 10-35 and 14-57
(Figs. 4 and 5). The beds within the Dms lithofacies are poorly
defined and are ~3 cm thick (Fig. 4g). Sandy, vertically oriented
intrusions that are devoid of shell fragments extend from 0.81 to
1.0 m bsf of core 14-53 and yield distinctly lower shear stress values
(Fig. 5). Small to large pebble sized clasts of variable roundness are
common in this lithofacies and occasionally clasts are rhythmically
concentrated in ~5-cm thick intervals (Fig. 4c). One 4-cm long (aaxis),
sub-rounded, dark blueish grey (Gley 2, 4/5PB), fine-grained
Table 2
Radiocarbon ages from the western Irish continental shelf.
Core Depth
(cm bsf)
Sample material 14
C age
(yrs. BP)
DR used Calibrated 2s
age range (yrs. BP)
Calibrated 2s age
median (yrs. BP)
d13
C (‰) Laboratory code
10-35 45 Paired bivalve shell 18,060 ± 120 þ300 ± 50 21,339e20,588 20,944 8.5 UCIAMS-144579
10-42 36 Paired bivalve shell 17,900 ± 89* þ300 ± 50 21,034e20,470 20,741 À4.7 Poz-66484
10-44 75 Coral fragment 13,614 ± 48 0 ± 50 16,109e15,646 15,876 À6.5 SUERC-48915
10-44 180 Coral fragment 20,710 ± 90* þ300 ± 50 24,351e23,783 24,067 À1.3 Beta-334419
10-45 94 Single bivalve shell 18,733 ± 107* þ300 ± 50 22,210e21,504 21,841 4.3 Poz#2-66430
14-53 70 Mixed benthic foraminifera 15,956 ± 87 þ300 ± 50 18,748e18,280 18,517 À6.9 Poz-66485
14-54 134 Mixed benthic foraminifera 18,222 ± 59 þ300 ± 50 21,437e20,891 21,161 À1.0 SUERC-58294
14-59 325 Single bivalve shell 13,121 ± 42 0 ± 50 15,349e14,844 15,148 À0.1 SUERC-58295
* From Peters et al., 2015.
Fig. 3. Seafloor profiles characterising the glacial landforms in the study area. Bold grey lines labelled i, ii and iii delineate the western flanks of seafloor ridges that extend between
profiles but do not represent ridge geography (Fig. 2a). (a) Profile X-X0 reveals the northern extension of the low-relief, flat-topped ridge (ii) that flanks the largest arcuate ridge in
the study area (Fig. 2a) and the steep-sided arcuate ridge that truncates the large ridge. The location of core 10-40 is shown. (b) Profile W-W0 shows the northern extent of the
westernmost arcuate ridge (WIM, i), the complex of large, flat-topped, arcuate ridges (ii and iii) and the ridge-free, furrowed seafloor in between. The location of cores 10-35, 10-36
and 14-53 are shown. (c) Profile Y-Y0 exposes the corrugated southern extent of the westernmost arcuate ridge (i), the complex of large, flat-topped, arcuate ridges (ii and iii) and the
ridge-free, furrowed seafloor in between. The location of core 14-54 is shown. (d) Profile Z-Z0 depicts the deep furrows with flanking berms that are abundant in the Slyne Trough
and the poorly-defined southern extent of the complex of large, flat-topped arcuate ridges. The location of core 14-36 is shown.
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124106
clast sampled from 3-m bsf in core 10-40 (Fig. 5) was found to have
multiple sets of striations. Grain-size analyses reveal highly variable
mass percentages of >1 mm grains, ranging from almost no
grains in core 14-36 to >85% in core 10-40 (Fig. 5). Sand, silt and
clay percentages in the matrix are also variable but typically
dominated by silt or sandy silt; sand percentages commonly increase
upwards (Fig. 5). In core 10-35, on the outer shelf (Fig. 2a),
the Dmm lithofacies is greyish brown (2.5Y 5/2 or 2.5Y 4/2) and
often sandier than the diamicton sampled from other parts of the
shelf (Fig. 5). In cores 10-40, 14-36, 14-57 and 14-59 the moderately
stiff diamicton is grey to very dark grey (2.5Y 5/1 to 5Y 3/1). This
lithofacies is characterised by highly variable wet bulk densities
(1.7e2.4 g/cm3
) and irregular water content that ranges from 4 to
19% (often covering a similar range within one core, Fig. 5). The
Dmm lithofacies sampled from west of the large arcuate ridges
(Fig. 2a) has lower magnetic susceptibility values (<36 10À8
SI;
Fig. 5) than those recorded from east of the ridges (>90e181 10À8
SI; Fig. 5).
Foraminiferal tests are rare within the moderately-consolidated
Dms lithofacies and are usually abraded and broken. Benthic foraminifera
populations in this lithofacies are characterised by high
percentages of E. clavatum ranging up to 65% in core 14-59 (Fig. 5).
The base of cores 10-36, 14-41 and 14-54 consists of consolidated
(!50 kPa) laminated mud lithofacies (Fl). Similar
consolidated Fl and massive mud (Fm) lithofacies overlie diamictons
at the base of cores 14-53 and 14-57 (Fig. 5). The mud is
typically laminated (Fl) but at the base of core 10-36 it is massive
(Fm); these lithofacies are further refined based on the presence of
dropstones and shear planes, which are henceforth abbreviated
-(d) and -(s) in the lithofacies abbreviations, respectively, following
the codes established by Eyles et al. (1983). Deformed Fl lithofacies
are also interbedded in the Dmm diamicton of core 14-59. Grainsize
analysis of this lithofacies typically reveals small percentages
(<1.5%) of >1 mm lithic grains and in core 14-41 no clasts were
detected. However at the base of core 14-54 there is a peak of
>1 mm grains that represents 14.5% of the sediment (by weight)
and in cores 10-36, 14-54 and 14-57 large, randomly oriented
lonestones are commonly observed (Fig. 5). Lonestone abundance
decreases upwards through the mud lithofacies in cores 10-36 and
14-54. The mud is well sorted and silty with mean percentages of
sand, silt and clay in cores 10-36, 14-41, 14-54 and 14-57 approximately
16%, 55% and 29%, respectively (Fig. 5). The mud is greyish
brown to greenish black in colour (2.5Y 5/2, 2.5Y 4/1, Gley 1 2.5/
10Y). Shear stress values range from moderate to high (22e80 kPa)
and typically decrease upwards through the cores (Fig. 5). This
lithofacies is characterised by relatively high wet bulk densities
(~2.1 g/cm3
), moderate water content (14e19.8%) and typically high
magnetic susceptibility (~62e90 10À8
SI) in cores 14-41, 14-53, 14Fig.
4. Representative x-radiograph facies. (a) X-radiograph and structure sketch of in Dmmc from core 10-38 exposing a ~40 preferential clast alignment and showing the location
of a shear stress measurement of 80 kPa. (b) X-radiograph, structure sketch and photograph of deformed Fl from core 14-59 revealing apparent clast alignment to strain and
showing the location of a shear stress measurement of 28 kPa. (c) Dmm from core 10-40 exposing rhythmic areas of high clast concentration. (d) Clast-rich Fl(s) from core 14-57
exposing a prominent shear plane (top ¼ white arrow, bottom ¼ black arrow) with a large (~3-cm long) clast (arrow labelled ‘c’) aligned in the direction of shear. (e) Fl(s) from core
14-41; well-defined laminae reveal a ~40 reverse fault with an offset of ~2 cm (top of offset ¼ white arrow, bottom of offset ¼ black arrow). (f) Dcm with a diffuse lower contact
overlying Dmm in core 14-59; black arrow marks an area devoid of large clasts but filled with interstitial sandy matrix. (g) Dms from core 10-35 exposing clast-rich, horizontal
stratification; black arrows mark contacts between select strata. (h) GO from core 10-38; black arrow marks an area of void space between clasts. (i) Diffuse contact between Sm and
underlying SH in core 14-54; large shells are exposed as light-grey curvilinear shapes. (j) Diffuse contact between Sm and underlying Suf in core 10-35; white dashed line marks the
location of the GSA analyses that yielded 0.3% > 1 mm lithic fragments; GSA analysis that yielded 16% > 1 mm lithic fragments was taken at the bottom of the core section shown;
white arrow marks an area of relatively low-density, poorly-sorted, shell-fragment-rich material interpreted as an infilled burrow (bioturbation); shell fragments are exposed as
light-grey/white.
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 107
54 and 14-57. The westernmost occurrence of this lithofacies, in
core 10-36 (Table 1, Fig. 2a), has a markedly lower magnetic susceptibility
of 11 10À8
SI (Fig. 5).
Shear planes are common in the compact muds and displacement
of laminae records both extensional and compressional strain
(Fig. 4d, e). Compressional strain (thrust faulting) is only identified
in compact mud with shear strength values >50 kPa. Compressional
deformation in compact Fl sediment from core 14-41 can be
seen in x-radiographs to have a displacement of >2 cm along a
shear plane tilted 40 from horizontal (Fig. 4e). Clasts proximal to
the shear planes occasionally show preferential a-axis alignment
parallel to the plane (Fig. 4d, e). In core 14-59 the Fl lithofacies
consists of rhythmic, ~1 cm thick, sandy laminae with sharp lower
contacts that fine upwards to mud; these laminae are deformed
and show distinct inhomogeneous ductile folding (Benn and Evans,
1996; Fig. 4b).
The benthic foraminiferal population in the consolidated mud
lithofacies of cores 10-36, 14-53 and 14-54 contains high percentages
(relative abundance up to 38%) of well-preserved tests of
Elphidium excavatum forma clavatum (Feyling-Hanssen, 1972)
(Fig. 5). However this lithofacies is otherwise poor in biogenic
material and in several locations (cores 14-41 and 14-57), no
benthic foraminifera tests were identified in this lithofacies. In
cores 14-41 and 14-57 the stiff mud is devoid of shell fragments and
very few foraminifer tests were discovered (Fig. 5).
Fig. 5. Plots of core data from the western Irish continental shelf. From left: true-colour photograph, sketch of prominent structures discerned by x-radiograph analyses, log of
lithology and sedimentary structures, lithofacies abbreviations and extents of LFAs, grain size data, sediment density (wet bulk) and (in cores 10-40 and 14-59) counts of lithic
fragments >1 mm, sediment shear strength and water content (percent by weight), magnetic susceptibility and relative abundance of the benthic foraminifer Elphidium excavatum
forma clavatum. AMS radiocarbon sample locations (m bsf) are marked by horizontal, dashed yellow lines; ages are provided in Cal BP and photographs of the dated materials are
shown. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124108
3.2.2. LFA 2: Loose diamicton (Dmm) and mud (Fl(d), Fm)
Massive, muddy diamicton (Dmm) with moderate to low shear
strengths (~5e40 kPa, typically <20 kPa) occupies the base of core
10-45 and overlies the consolidated and deformed sediment of LFA
1 in cores 10-35, 10-40, 10-42, 10-44, 14-36, and 14-59 (Fig. 5). This
sediment is lithologically and structurally similar to the underlying,
consolidated diamicton, but is less compact and often has a higher
water content and lower density than the basal deposits (Fig. 5).
Like the underlying, moderately consolidated diamictons, the loose
Dmm lithofacies contain variable amounts of pebble-sized clasts
that comprise several lithologies. The loose Dmm contains higher
amounts of sand than the underlying diamictons and the sand
content typically increases up core (Fig. 5). There is a distinct
decrease in the number of lithic grains >1 mm in core 10-40 between
2.05 m bsf and 1.92 from 91.3 grains/g to 1.7 grains/g; this
decrease roughly correlates with a drop in the mass percentage of
>1 mm grains (Fig. 5). In cores 10-35 and 10-45 (on the outer shelf;
Fig. 2a) the loose Dmm lithofacies is characteristically greyish
brown (2.5Y 5/2 or 2.5Y 4/2) and often sandier than the diamicton
sampled from other parts of the shelf (Fig. 5).
Fig. 5. (continued).
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 109
The benthic foraminifera population of the loose Dmm lithofacies
contains high amounts of E. clavatum, including the highest
percentage observed in the analysed cores (73%, core 10-40; Fig. 5).
The loose Dmm lithofacies typically contains more shell fragments
than the underlying diamictons and the abundance of shell fragments
often increases upwards (Fig. 5). Visual inspection reveals
vertical to sub-vertical, sandy, granule and shell-fragment rich irregularities
that commonly extend downwards into the top of the
Dmm diamicton from the overlying lithofacies.
Loose to moderately-consolidated (20e50 kPa) mud lithofacies
overlies consolidated mud lithofacies in cores 10-36, 14-41, 14-54,
and 14-57 (Fig. 5). The loose or moderately-consolidated mud
lithofacies are typically laminated (Fl) but massive in core 10-36.
This sediment is lithologically and structurally similar to the underlying,
consolidated mud deposits, but is less compact and
usually has a higher water content and lower density than the basal
deposits (Fig. 5). Lonestones of variable lithologies are common in
this lithofacies, but are typically less abundant than in the underlying
mud deposits and often decrease in abundance up core
(Fig. 5). Shear planes are common in the loose mud lithofacies in
cores 14-41, 14-54 and 14-57 (Fig. 5). At ~0.5 m bsf in core 14-54 the
loose Fl(s) lithofacies contains an unconsolidated, sandy, softsediment
clast (Fig. 5).
The benthic foraminifera population of the loose mud lithofacies
is dominated by E. clavatum and reaches relative abundances from
30% to 38% (Fig. 5). Vertical to sub-vertical, sand and shell-fragment
rich irregularities commonly extend downwards into the top of the
loose mud lithofacies from the overlying deposits (Fig. 5).
Fig. 5. (continued).
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124110
Fig. 5. (continued).
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 111
3.2.3. LFA 3: Coarse, clast-supported lithofacies (Dcm, GO) and shell
hash (SH and GSH)
Coarse, massive, clast-supported diamicton (Dcm) or openwork
gravel (GO) overlies the muddy and diamictic lithofacies of LFA 1
and LFA 2 in eight of the cores examined in this study (10-35, 10-38,
10-42, 10-44, 10-45, 14-42, 14-57, 14-59). These facies are relatively
thin, ranging from 5 to 25 cm in thickness (Fig. 5). Lower contacts
are typically diffuse, except above the Dmmc lithofacies in core 10-
38, the Fl(s) lithofacies in core 14-57 and the Dmm lithofacies in
core 14-59 (Fig. 5). This lithofacies consists of small to large pebblesized
clasts of variable lithologies and roundness, surrounded by
varying amounts of coarse sandy matrix. A clast in the Dcm facies in
core 10-45 exhibited striations, a “rounded nose” and a “plucked
lee” (Peters et al., 2015; Fig. 5). In core 10-38, the Dmmc lithofacies
is overlain by GO gravel with rare granules occupying the interpebble
cavities (Fig. 4h). Grain-size analyses reveal moderate to
high percentages of lithic fragments >1 mm (typically >20% and up
to 46% in core 10-45, Fig. 5). Matrix material is dominated by sand
(>60e90%) and typically contains shell fragments. The colour of
this lithofacies is highly variable, but the matrix is often similar in
colour to the underlying diamicton; conversely, in cores 10-38, 14-
57 and 14-59 (where these lithofacies have sharp lower contacts),
the GO and Dcm matrix material is more similar to the overlying
sediment. This lithofacies is characterised by high wet bulk densities
(up to 2.5 g/cm3
in cores 10-35 and 14-59), low water content
(13.5e14.5%) and highly variable magnetic susceptibility, often with
Fig. 5. (continued).
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124112
irregular peaks above or below the core mean (Fig. 5).
The foraminifera sampled from the matrix material of the Dcm
facies were often broken and abraded (cf. Peters et al., 2015). Well
preserved E. clavatum are rare in the Dcm sediment (>5% relative
abundance), but occasionally high numbers were are recovered in
the surrounding deposits (31.1e44.8% relative abundance in core
14-59, Fig. 5). Shell fragments are common in the Dcm and GO
lithofacies and lower contacts are commonly bisected by poorlysorted,
vertical to sub-vertical, sandy, granule and shell-fragment
rich inclusions.
A shell hash (SH) or gravelly shell hash (GSH) consisting primarily
of highly fragmented bivalve shells overlies the diamictic
and muddy lithofacies in cores 10-40, 14-41, 14-53, 14-54, 14-59
(Fig. 5). This lithofacies has both diffuse and sharp lower contacts
and varies from 7 to 148 cm in thickness (Fig. 5). The shell hash
occurs most commonly and typically with the greatest thickness in
the easternmost cores. Rounded, pebble sized clasts of variable lithologies
are common in the GSH deposits in cores 10-40 and 14-41
(Fig. 5). Grain-size analyses of the <1 mm sediment fraction of the
SH lithofacies in core 14-59 reveal the small amounts of matrix
sediment to be 100% sand (Fig. 5). In core 10-40, the GSH sediment
contains a high percentage of granules >1 mm (64.5e80.8%, Fig. 5).
The colour of this lithofacies is variable, but typically light yellowish
brown to light olive brown (2.5Y 6.4 to 2.5Y 5/6). This lithofacies is
characterised by highly variable wet bulk densities with erratic
peaks that usually fall well above the core mean, magnetic susceptibility
values below the core mean, low shear stress measurements
(<6 kPa), and low water content (<17.3%, typically <7% and
as low as 3.9% in core 10-40, Fig. 5).
The foraminifera sampled from the sand matrix of the shell hash
Fig. 5. (continued).
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 113
deposits were often broken and abraded. The indicator species
E. clavatum is relatively rare in the SH and GSH lithofacies and
shows an upwards decrease in abundance (e.g. cores 10-40 and 14-
59, Fig. 5). In cores 14-41 and 14-54, lower contacts are bisected by
vertical to sub-vertical sandy, granule and shell-fragment rich
irregularities.
3.2.4. LFA 4: Sandy lithofacies (Sh, Suf, Sm)
The top of each of the cores analysed for this study, with the
exception of core 14-59 (the easternmost core), is composed of
massive (Sm), upward fining (Suf) or crudely horizontally bedded
sand (Sh) lithofacies; the sand is 6e50 cm thick and usually has
sharp lower contacts (Fig. 5). These lithofacies are well sorted with
grain-size analyses revealing very high amounts of sand (up to 94%
in core 14-41, Fig. 5). The sandy deposits from the westernmost
cores (10-35, 10-36, 10-42, 10-44) are greyish brown (2.5Y 5/2); in
the remaining cores, the sand is light yellowish brown (2.5Y 6/3) to
olive yellow (2.5Y 6/6) in colour. This lithofacies is characterised by
low wet bulk densities (often <2.0 g/cm3
) that decrease upwards,
low shear strengths (<5 kPa), moderate to high water contents
(usually >20%) that increase upwards, and often low magnetic
susceptibility values (usually <40 10À8
SI) that are less than core
mean values (Fig. 5). Moderate magnetic susceptibility values from
~62 to as much as ~70 10À8
SI are recorded from cores 14-41, 14-42,
14-54, 14-57 and 14-59 (the easternmost cores, Fig. 2a).
The sand lithofacies yielded few E. clavatum tests, often with no
individuals found near the core tops (Fig. 5). Shell fragments are
abundant near the bottom of these lithofacies, but usually decrease
in size and occurrence upwards (e.g. 20e5 cm bsf in core 10-35,
Fig. 4j). Irregularly shaped areas of relatively low density, composed
of poorly-sorted, shell-fragment rich material are common (Fig. 4j).
3.3. Chronology
Five samples of calcareous, marine biogenic material were dated
using AMS radiocarbon analyses (Table 2). All samples were well
preserved, with little to no abrasion, breakage or discolouration.
The samples were taken from glacigenic material collected on the
continental shelf. Three AMS radiocarbon ages from a previous
study (Peters et al., 2015) on the Porcupine Bank and Slyne Trough
are also recalibrated for use in this study (Table 2). The oldest age
on the shelf west of Ireland (24,067 Cal BP) is sampled from core 14-
44 in the Slyne Trough (one of the westernmost cores). The age
distribution across the shelf reveals a typical eastward trend of
younger sediment deposition, with the youngest age of
15,148 Cal BP from the easternmost core (14-59, Table 2).
4. Interpretation
4.1. Geomorphology
4.1.1. Seafloor ridges
The arcuate and nested shapes, scale and orientation of the
seafloor ridges are inconsistent with the regional, bedrockcontrolled
geomorphic trends discernible in the topographic DEM
(Fig. 2a), but they are comparable to submerged moraines (e.g.
Bradwell et al., 2008; O Cofaigh et al., 2010). This suggests that the
ridges are likely formed in association with marine-terminating ice
margins during glacial advance or retreat on the shelf. The asymmetrical
profile and arcuate shape of the westernmost corrugated
ridge (Figs. 2a, b, 3b, c) is consistent with deposition as a push
moraine with a steep ice-distal flank to the west (Boulton, 1986).
The corrugations on its southern end are aligned with the sinuous
ridges in the Slyne Trough, which have been interpreted as a record
of grounded ice extent, followed by ice shelf recoupling during
retreat (Peters et al., 2015). The ridge-corrugation alignment
(Fig. 2b) suggests that the corrugations may have been formed
contemporaneously and by the same mechanism as the ridges in
the Slyne Trough. Thus we tentatively interpret the arcuate ridge to
be a push moraine that was subsequently overridden by an ice shelf
that extended over the Slyne Trough (Peters et al., 2015). This
interpretation is supported by the southward broadening of the
arcuate ridge, which is consistent with moraines that were overridden
and smoothed by readvancing ice (e.g. Bentley et al., 2007;
Jonsson et al., 2014). This partially overridden, westernmost
arcuate ridge is henceforth referred to as the West Ireland Moraine
(WIM).
The arcuate shape of the largest, southernmost ridge in the
study area (Fig. 2a) is consistent with ice-marginal deposition
against BIIS marine termini (cf. Bradwell et al., 2008; O Cofaigh
et al., 2010). Its flat-topped profile (Fig. 3b, c) suggests deposition
below a floating ice mass that constrained vertical accretion
(Dowdeswell et al., 2008; Batchelor and Dowdeswell, 2015). Thus,
we interpret this landform as a grounding-zone wedge that formed
against an ice lobe with a floating ice shelf. Because of the location
of the ridge on the continental shelf, we interpret the ice lobe to
have most likely emanated from south of the Maumturk Mountains,
roughly centred on Galway Bay (Fig. 2a); thus, this feature is
henceforth referred to as the Galway Lobe Grounding-Zone Wedge
(GLGZW). The low-relief, flat-topped landform that forms the
western flank of the GLGZW (Figs. 2 and 3) is interpreted to most
likely record an earlier phase of grounding-zone wedge deposition
because of its geomorphic similarities to the GLGZW. Therefore, we
interpret both flat topped landforms to be components of a complex,
grounding zone deposit that was built under ice shelves with
variable thicknesses along an apparently bathymetrically
controlled grounding zone (cf. O Cofaigh et al., 2005). The dimensions
of the GLGZW (~150 km long and 22 km wide) are similar
to some large Antarctic grounding-zone wedges (Evans et al., 2005)
and indicate that it likely records a significant (i.e. centuries long)
stillstand during BIIS retreat (cf. Dowdeswell et al., 2008).
The ridge situated directly east of the GLGZW (Fig. 2a) is interpreted
to most likely be a recessional or readvance moraine based
on its smaller size and nested relationship to the larger ridge (cf. O
Cofaigh et al., 2010). Because of its similar orientation, which is
indicative of ice extension roughly from Galway Bay, this moraine
was likely formed by ice draining a similar area to the GLGZW. Thus,
this moraine is henceforth referred to as the Galway Lobe Readvance
Moraine. However, the term ‘readvance’ is used tentatively
on the basis of only the geomorphic evidence.
The two arcuate ridges north of the GLGZW have steep, asymmetric
flanks (Figs. 2 and 3a) consistent with deposition as push
moraines (Boulton, 1986). The smaller, southern ridge truncates the
GLGZW and the northern ridge, indicating that it likely formed
during ice overriding of the surrounding landforms following their
deposition. Based on their relatively close proximity to land (within
~25 km, Fig. 2a), their well-preserved, steep-sided morphology, and
cross-cutting nature of these ridges, they are interpreted to have
been formed during the readvance of ice lobes across the shelf (cf.
Bradwell et al., 2008). The smaller, evidently younger ridge is
henceforth referred to as the Connemara Lobe Moraine because it is
positioned offshore of the mountainous Connemara district
(Fig. 2a). The arcuate ridge situated north of the Connemara Lobe
Moraine is interpreted as an older recessional or readvance
moraine that likely extended onto the shelf from ice sources that
drained past Clew or Blacksod bays from County Mayo based on its
geographic position (Figs. 1 and 2a). This landform is henceforth
referred to as the Mayo Lobe Moraine.
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124114
4.1.2. Seafloor furrows
The seafloor furrows are interpreted as iceberg scours based on
their irregular, occasionally meandering, trajectories and the
presence of adjacent lateral berms (Belderson et al., 1973;
Dowdeswell et al., 1993; O Cofaigh et al., 2002, 2010; Fig. 3d).
Their abundance on the shelf in water depths approximately
<350 mbsl suggests a local iceberg source and supports an interpretation
of ice shelf formation and subsequent break up west of
the GLGZW (Peters et al., 2015). When identified, the gradual scour
termini mark the deepest water depths of the scours and the
abrupt, pit-like depressions mark the shallowest points (cf.
Andreassen et al., 2014; Fig. 2c). Thus, the gradual and abrupt
termini are interpreted to record the inception of iceberg contact
with the seafloor and iceberg grounding pits, respectively (cf. Hill
et al., 2008; Andreassen et al., 2014). The abrupt scour termini in
the Slyne Trough are typically located at the northern ends of the
scours, indicating a northward or northeastward palaeocurrent (cf.
Peters et al., 2015) or a combination of current and wind interaction.
The corrugated troughs that characterise some scours are
interpreted as the signature of tidal action on the grounded icebergs
(cf. Jakobsson et al., 2011). Rare parallel furrows (Fig. 2c, d)
indicate the grounding of large icebergs with multiple keels and
suggests calving from a collapsing ice shelf (Andreassen et al.,
2014).
4.2. Sedimentology
Four Lithofacies Associations (LFAs) are identified in the cores
and record the sedimentary history from BIIS advance to Holocene
postglacial marine sedimentation. Without sub-bottom seismic
stratigraphic data, these lithofacies are assumed to sample the
glacial landform-creating sediment in cores 10-35, 10-38, 10-40, 10-
42, 10-44, 10-45, 14-36, 14-42, 14-54, 14-57 and 14-59 because of
their penetration depths (up to 4.0 m) and close proximity (usually
in contact with, occasionally 440 m) to the landforms (Fig. 2a).
This assumption is reinforced by sedimentological similarities to
glacially-derived sediment described by previous studies from
formerly glaciated continental shelves (e.g. O Cofaigh and
Dowdeswell, 2001; Dowdeswell et al., 2004; Evans et al., 2005; O
Cofaigh et al., 2005, 2011; Hillenbrand et al., 2010).
4.2.1. Lithofacies association 1 (ice-contact deposition/reworking)
LFA 1 consists of Dmmc, Dmm, Fl(s) and consolidated or
deformed Fm lithofacies. The Dmmc diamicton at the base of cores
10-38, 10-42, 10-44, 14-36, 14-42 and 14-54 (Fig. 5) is interpreted as
an overconsolidated subglacial traction till based on its high shear
strength, massive structure, high wet bulk density and low water
content (cf. Wellner et al., 2001; Evans et al., 2005; O Cofaigh et al.,
2005; D. Evans et al., 2006a). Striated clasts within the Dmmc
diamicton also support a subglacial depositional environment
(Sharp, 1982). Areas of high shear strength (50e>197 kPa, Fig. 5)
suggest thick ice, low pore-water pressures and potentially low iceflow
velocities (non-streaming) (Sættem et al., 1996; cf.
Dowdeswell et al., 2004). Preferentially-oriented clast a-axes
indicate alignment with strain during subglacial till formation
(Benn and Evans, 1996; Bennett and Glasser, 2011; van der Meer
et al., 2003; J. Evans et al., 2006b). Microfossil damage is consistent
with the cannibalisation and subglacial reworking of preglacial
marine sediment (O Cofaigh et al., 2011); sediment devoid
of foraminiferal tests suggests a terrigenous sediment supply. A
coral fragment from core 10-44 is AMS radiocarbon dated to
24,067 Cal BP. This date is interpreted to constrain the maximum
age of the till deposited during the advance of the last BIIS (Peters
et al., 2015; Table 2).
The Dmm at the base of cores 10-35, 14-57 and 14-59 and
overlying the overconsolidated basal till in cores 14-36 and 14-53
(Fig. 5) is also interpreted as a subglacial traction till, associated
with deformation of dilatant sediment, because of its massive or
crudely stratified structure and moderate shear strengths that
gradually decrease upwards (Benn and Evans, 1996; J. Evans et al.,
2005, 2006b; D. Evans et al., 2006a). Sandy, vertically oriented inclusions
with moderate shear strength (40 kPa) in core 14-53
(Fig. 5) are interpreted as dewatering structures and, along with the
increase in water content (up to 19%) relative to the underlying, stiff
basal till, indicate higher porosity and pore-water pressures
(Rijsdijk et al., 1999; van der Meer et al., 1999). Furthermore, clast
lithology, magnetic susceptibility and foraminiferal content is
comparable between the Dmm and Dmmc lithofacies, suggesting
that the former is derived from the reworking of the latter (cf.
Dowdeswell et al., 2004; Evans et al., 2005; Kilfeather et al., 2011).
An unabraided, unbroken, parasitized bivalve shell in the till near
the base (325 cm bsf) of core 14-59 (Fig. 5) is AMS radiocarbon
dated to 15,148 Cal BP (Table 2). This age indicates that the ice
advanced over and incorporated material that was biomineralized
late in MIS 2. This suggests that either an ice mass readvanced to
this point or icebergs grounded against moraines on the shelf at
least 1000 years after the Killard Point Stadial.
The highly-consolidated, silt-rich Fl(d) and Fl(s) deposits overlying
the subglacial tills in cores 10-36, 14-41, 14-53, 14-54 and 14-
57 are interpreted as glacitectonised glaciomarine sediment based
on their preserved primary structures (laminae), thrust faults and
high shear strengths (Benn and Evans, 1996). The sediments were
likely originally deposited proglacially by suspension settling from
sediment plumes based on their high silt content, parallel laminae
and dropstones (O Cofaigh and Dowdeswell, 2001). Areas devoid of
foraminiferal tests and with few dropstones in cores 14-41 and 14-
57 support an interpretation of terrigenous sediment supply in an
ice-proximal, possibly sub-ice shelf environment. These sediments
were likely glacitectonised penecontemporaneously during recoupling.
Conversely, glacitectonised suspension plume sediment
(Fl(d) and Fl(s)) with abundant E. clavatum tests in cores 10-36, 14-
53 and 14-54 (Fig. 5) is interpreted to have been deposited proglacially
prior to being overridden (cf. O Cofaigh et al., 2011). The
sharp lower contacts and normal grading of the folded Fl lithofacies
in core 14-59 (Figs. 4b and 5) indicate a glaciomarine primary
depositional environment. The laminae are interpreted as having
been deposited by suspension settling from meltwater plumes
(plumites) based on their rhythmic nature, sharp lower contacts
and upward fining (Dowdeswell et al., 2000). The ductile, inhomogeneous
folding identified in the Dms lithofacies of core 14-59 is
interpreted to record compressional strain. The evidence for
compressional strain and moderate shear strength (up to 28 kPa) in
the folded sediment are consistent with glaciotectonisation from an
overriding ice mass (Benn and Evans, 1996; O Cofaigh et al., 2011).
An interpretation of ice readvance over these sites is supported by
two AMS radiocarbon dates acquired from relatively wellpreserved
(i.e., light surface abrasion but little discolouration or
fracturing) mixed benthic foraminifera (dominantly E. clavatum)
that were sampled from within the glaciotectonite of cores 14-53
and 14-54 (Table 2, Fig. 5). The ages constrain the biomineralization
of tests within the glaciotectonite to 21,161 Cal BP (core 14-54,
Table 2) west of the GLGZW and 18,517 Cal BP (core 14-53,
Table 2) east of the GLGZW. These ages suggest a readvance of the
ice sheet across the shelf either associated with glaciodynamic
adjustments during deglaciation or possibly related to climate
forcing during the Killard Point (Heinrich event 1) stadial (McCabe
and Clark, 1998; McCabe et al., 2005, 2007; Clark et al., 2009a,
2009b, 2012b, and references therein).
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 115
4.2.2. Lithofacies association 2 (Glaciomarine deposition and
iceberg turbation)
This LFA consists of Dmm, Fl(d) and Fm lithofacies with a deposit
of Sm lithofacies in core 14-53 (Fig. 5). The loose, greyish-brown
Dmm diamicton overlying the till and glaciotectonite in cores 10-
35, 10-42, 10-44 and 14-36 and at the bottom of cores 10-40, 10-45
and 14-59 (Fig. 5) is interpreted as the signature of meltwaterderived
sediment-plume suspension settling (plumite) with high
IRD input (rain-out sediment) based on its low to moderate shear
strength (6e25 kPa), high water content, abundance of outsized
lonestones, and highly variable wet bulk density and magnetic
susceptibility (cf. Hillenbrand et al., 2005; Lucchi et al., 2013).
Similarly, the loose, greyish-brown mud overlying the glaciotectonite
in cores 10-36, 14-41, 14-54 and 14-57 (Fig. 5) is interpreted
as proglacial, retreat-phase suspension plume sediment
with more meltwater-derived fines and less IRD than the massive
diamicton, based on its laminated structure, high silt content,
moderate shear strength (20e40 kPa), and variable wet bulk density
and magnetic susceptibility (Dowdeswell et al., 2000; O
Cofaigh and Dowdeswell, 2001). The massive, shelly sand (Sm)
from 32 to 49 cm bsf in core 14-53 is highly deformed (indicated by
large areas of loose sediment or void space against the core liner;
Fig. 5) but tentatively interpreted as a postglacial current deposit or
sandy rain-out deposit that dewatered and mixed with the overlying
shell hash during core acquisition. The rain-out sediment in
the three westernmost cores from the Slyne Trough (10-42, 10-44
and 10-45, Fig. 2a) has previously been interpreted as in situ, subice
shelf and ice-proximal rain-out sediment based on micropaleontological
data, clast abundance and lithologic similarity (Peters
et al., 2015). Variability in the wet bulk density and magnetic susceptibility
is interpreted as a signature of the high sediment heterogeneity.
The mud component in LFA 2 typically decreases
upwards (Fig. 5), which is also consistent with an interpretation of
increasingly ice-distal glaciomarine sedimentation (cf. Smith et al.,
2011). Zones of high clast density within the cores (e.g. core 10-40,
Fig. 4c) are probably related to increased delivery of IRD to the core
sites and their rhythmic occurrence is likely a record of a seasonal
IRD production (Cowan et al., 1997) or seasonal sea-ice cover that
suppressed iceberg activity (Dowdeswell et al., 2000; O Cofaigh and
Dowdeswell, 2001). This rain-out deposit is dated to 21,841 Cal BP
(core 10-45, Table 2) in the Slyne Trough (Peters et al., 2015) and an
AMS radiocarbon date from a well preserved (unbroken and unabraded),
paired bivalve shell sampled from 45 cm bsf in core 10-35
(Fig. 5) provided an age of 20,944 Cal BP (Table 2). This date constrains
the age of deposition for glaciomarine sedimentation and
BIIS retreat from the Slyne Trough.
The genesis of plumite and rain-out sediment is further refined
based on IRD abundance, secondary structures, biogenetic content,
and the presence of vertical to sub-vertical shell-fragment rich sand
intrusions interpreted as related to bioturbation. Grounding lineproximal
sediment is identified at the base of the proglacial deposits
based on high IRD abundance, a lack of bioturbation or
macro-shell fragments, and a benthic foraminiferal population
dominated by E. clavatum (cf. Lucchi et al., 2013). Conversely,
grounding line-distal deposits are differentiated by the presence of
unabraded bivalve shells and fragments, reduced or absent IRD
sedimentation and bioturbation towards the top of the deposit (cf.
L€owemark et al., 2015).
Post-depositional reworking, likely from iceberg turbation, or
debris flows is identified above the till in cores 14-41, 14-42 and 14-
57 and overlying the proglacial plumite in core 14-54 based on
occasional shear planes and soft sediment rip-up clasts in loose
sandy mud (cf. Hillenbrand et al., 2013). These deposits are differentiated
from the subglacial till by their typically lower shear
strengths (<40 kPa) and higher water content (Fig. 5). Loose
diamicton overlying an iceberg-rafted deposit in core 10-44 is also
interpreted as reworked by local iceberg turbation or possibly mass
wasting on the flank of a Slyne Trough Moraine based on its lithologic
similarity to the underlying sediment, abruptly lower shear
strength and increasing sand content (cf. Vorren et al., 1983;
Dowdeswell et al., 1994; Fig. 5). Iceberg turbate formation in the
Slyne Trough is supported by previous micropalaeontological and
geomorphic research that documents an increase in foraminiferal
test damage and intense iceberg scouring in the interval we interpret
as reworked (Peters et al., 2015). An anomalously young
radiocarbon age (15,876 Cal BP) from this reworked interval in core
10-44 is comparable to other chronologic inconsistencies in iceberg
turbates (e.g. Hillenbrand et al., 2010) and suggests that on-ridge
reworking occurred <16,000 yr BP.
4.2.3. Lithofacies association 3 (glacial to postglacial transition)
Lithofacies association 3 consists of four lithofacies (Dcm, GO,
SH and GSH) deposited between the overlying sandy lithofacies of
LFA 4 and the underlying glacial sediments of LFAs 2 and 1. Clasts
within the Dcm diamictons and GO gravel are lithologically similar
to those in the underlying deposits, suggesting a similar sediment
source. Therefore the reduction in matrix material in LFA 3 suggests
either increased rates of IRD production (cf. Kilfeather et al., 2011)
or sediment winnowing (Eyles, 1988). The Dcm diamicton from the
western Slyne Trough (cores 10-42, 10-44 and 10-45, Fig. 2a) that
overlies the sub-ice shelf sediment has previously been interpreted
as intense IRD rain-out sediment that records ice shelf breakup and
subsequent current winnowing (Peters et al., 2015).
The SH and GSH lithofacies are composed largely or entirely of
upward-fining bivalve shells and fragments within very small
amounts of sand matrix, suggesting fine sediment winnowing in a
biologically active palaeoenvironment (Flemming et al., 1992). The
shell-rich deposits are interpreted to record sea level transgression
based on their sharp lower contacts, upward fining and winnowed
fines (cf. Chang et al., 2006). The shell hash of core 14-53 contains
an 8-cm thick unit of massive, shell-rich sand (Fig. 5) that we
interpret to likely record a period of relative, local palaeocurrent
quiescence.
4.2.4. Lithofacies association 4 (postglacial deposition/reworking)
This LFA consists of Sh, Suf and Sm lithofacies. The poorlysorted,
yellowish sand at the top of all cores except core 14-59
(Fig. 5) is interpreted as postglacially reworked sediment based on
its low mud content, low shear strength, typically low wet bulk
density and high water content (cf. Hillenbrand et al., 2013). Upward
fining sand at or near the top of cores 10-35, 10-40, 10-42, 10-
44 and 10-45 (Fig. 5) indicate reworking by gradually weakening
bottom currents (Bishop and Jones, 1979; Fyfe et al., 1993; Viana
et al., 1998), which is in accord with interpretations of winnowing
in the underlying transitional sediments of LFA 3. Previous
micropaleontological research on the Porcupine Bank is also
compatible with an interpretation of winnowing and reworking by
bottom currents (Smeulders et al., 2014; Peters et al., 2015). This
upwards waning in palaeocurrent activity is interpreted as a record
of postglacial sea level transgression across the continental shelf
(cf. Amorosi et al., 1999; Barrie and Conway, 2002).
4.3. Core correlation and regional stratigraphy
Core correlations are based on LFA interpretations and from this
four stratigraphic transects are produced that extend from west to
east across the study area (Fig. 6). These transects cross all of the
major glacigenic landforms in the study area and reveal glacial and
postglacial depositional trends west of Ireland. The stratigraphic
sequence is generally consistent between the cores and shows the
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124116
following vertical stratigraphic sequence: subglacial traction till
which shows a progressive decrease in shear strength upwards,
glaciotectonite, glaciomarine deposits (plumite and iceberg-rafted
sediment), a transgressive lag and current-reworked sand (Fig. 6).
However, the following lateral patterns in sediment distribution
across the shelf are also identified: (1) a general eastward
(landward) increase in glaciotectonite, (2) a prevalence of reworked
sediment near the GLGZW, (3) a general eastward thickening of the
transitional, transgressive deposits (LFA 3), and (4) an occasional,
marked thickening of the glaciomarine and postglacial deposits to
the east of large moraines (e.g. cores 10-40, 10-45, 14-59, Fig. 6).
The ~2e4 m thick sedimentary record of the western Irish
Fig. 6. Core correlations revealing continental shelf stratigraphy west of Ireland and exposing the glacial and postglacial depositional history associated with the last BIIS. Sediment
core data is shown in detail in Fig. 5. Abbreviations: WIM ¼ West Ireland Moraine; KLM ¼ Killary Lobe Moraine; GLGZW ¼ Galway Lobe Grounding-Zone Wedge; PBM ¼ Porcupine
Bank Moraine; STMs ¼ Slyne Trough Moraines.
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 117
continental shelf is dominated by the glacigenic deposits of LFA 1
(Fig. 6). Subglacial traction tills are common at the base of the
Midlandian glacial sedimentary record west of Ireland and their
occurrence confirms previous research documenting grounded BIIS
extension across the continental shelf (e.g. O Cofaigh et al., 2010;
Peters et al., 2015). A less-compact till with a diffuse lower contact
often overlies the lowermost stiff basal till in cores from the
shelf. This relatively soft till records increased pore water pressures
that may have been caused by the underlying clay-rich, and thus
lower porosity, stiff till which would have acted to impede pore
water expulsion from the soft till (Fischer et al., 1999; Lian and
Hicock, 2000; J. Evans et al., 2006b). In places the tills are covered
by glaciotectonite up to !1 m thick (Fig. 6). This glaciotectonite is
most common towards the east of the study area (i.e. landward)
and frequently occurs on the flanks of the Galway Lobe Moraine
(Fig. 6).
A glaciomarine deposit with a diffuse lower contact overlies the
till and glacitectonite in all cores except 10-38 (Fig. 6). It is interpreted
as a product of suspension settling from meltwater plumes
during BIIS retreat. It ranges in thickness from 12 to 224 cm thick
with a mean thickness of ~75 cm. The thickest glaciomarine deposits
are massive, muddy, IRD-rich diamictons; these deposits are
>100 cm in the Slyne Trough (Cores 10-44 and 10-45) and just east
of the Connemara Lobe Moraine (core 10-40, Fig. 6). These areas of
thick glaciomarine deposition are interpreted as a record of either
prolonged sub-ice shelf (cf. Kilfeather et al., 2011) or iceberg rafted
sedimentation. The lack of glaciomarine sediment in core 10-38 is
interpreted as an erosional unconformity caused by current winnowing,
possibly facilitated by the bathymetric high formed by the
Connemara Lobe Moraine (Figs. 2a and 3a).
Areas of remobilised glaciomarine sediment overly the plumite
near moraines (cores 10-44, 14-42, 14-54 and 14-57) and, less
frequently, on the inner shelf (core 14-41) (Fig. 6). Most of these
reworked deposits likely record iceberg turbation on or near
moraine ridges. This interpretation is consistent with a previous
examination of iceberg plough mark geomorphology in the Slyne
Trough, which documents iceberg grounding against moraine
ridges and a roughly northward palaeocurrent (Peters et al., 2015).
Postglacial mass wasting, which likely would have been most
intense on the morainic ridges, is another potential source for
sediment reworking (cf. McCabe, 1986).
A unit of shell-bearing, bottom-current-winnowed sediment
interpreted to record a transitional period from glacial to postglacial
conditions overlies the glaciomarine deposits in each core
except core 10-36 (Fig. 6). These transitional sediments range in
thickness from 6 to 146 cm (cores 10-35 and 14-59, respectively)
and are thickest to the east (Fig. 6). This sediment also typically
contains progressively more biogenic material towards the east
(Fig. 6). The thickest (up to 21 cm) of the lithic-dominated, winnowed
sediment units are found in the Slyne Trough; this is interpreted
as a result of coarse sub-ice shelf or ice-proximal sediment
supply that led to increased IRD production after ice shelf break up.
The relative abundance of large lithic clasts was likely increased by
winnowing from a northward Slyne Trough palaeocurrent that
developed following ice shelf break up (Peters et al., 2015). Another
thick, lithic, winnowed sediment unit overlies the highlyconsolidated,
subglacial till at the base of core 10-38 (Figs. 5 and
6). This area (the crest of the Connemara Lobe Moraine, Figs. 2a
and 3a) is interpreted to have experienced particularly high
amounts of winnowing based on a lack of preserved glaciomarine
sediment in core 10-38, the thickness of the winnowed transitional
unit (20 cm), and the general absence of fine-grained sediment in
the GO facies (Fig. 5). The eastern, shell-dominated, winnowed
sediment is thickest within 50 km of shore (cores 10-40 and 14-59,
Fig. 2a) where it is largely composed of biogenic material (Figs. 5
and 6). The eastern thickening of the unit and increasing shell
content is consistent with reworking during sea-level transgression
(e.g. Saito et al., 1998), thus, we refer to these deposits collectively
as a transgressional lag (cf. Chang et al., 2006).
Mean magnetic susceptibility values show a distinct eastward
(i.e. shoreward) increase (Fig. 6). Values for cores >40 km west of
the GLGZW and Connemara Lobe Moraine (cores 10-35, 10-36, 10-
42, 10-44, 10-45; Fig. 2a) range from 11 to 36 10À8
SI (Figs. 5 and 6)
with an average of 19.75 10À8
SI. Conversely, the magnetic susceptibility
of cores from near to or east of these large moraines
(Fig. 2a) ranges from 21 to 135 10À8
SI (Figs. 5 and 6) with a mean of
83.86 10À8
SI. This discrepancy is interpreted as the signature of
increasing amounts of glacially-mobilised terrigenous sediment
supply to the shelf (Robinson et al., 1995; Shevenell et al., 1996).
This interpretation is supported by previous sedimentary interpretations
in the Slyne Trough and Porcupine Bank (Peters et al.,
2015).
Elphidium clavatum was found to be the dominant benthic
foraminifera below LFA 3 in most of the micropalaeontologicallyexamined
cores, usually near the contact with the overlying
transgressional lag (Figs. 5 and 6). The dominance of E. clavatum
typically diminishes upwards in cores from the shelf (Fig. 5). This
E. clavatum range indicates a palaeoenvironment with variable
salinity, high sedimentation rates, cold (<1 C) average sea temperatures,
and potential sea ice cover (Mudie et al., 1984; Hald and
Korsun, 1997; Polyak et al., 2002; Stalder et al., 2014). This inferred
micropaleontological environment supports our sedimentological
interpretation of glaciomarine deposition from meltwater and
iceberg rafting with possible areas of seasonal sea ice cover. The
bottom of the rain-out sediment in core 14-36 and the top of the
subglacial sediment in cores 10-38, 14-41 and 14-57 are devoid of
any foraminiferal tests (Fig. 6), indicating a depositional environment
dominated by terrigenous sediment supply where little or no
pre-glacial marine sediment was incorporated into the till (cf. O
Cofaigh et al., 2011; McCabe and Clark, 2003).
5. Discussion
5.1. LGM ice sheet extent, configuration and chronology on the shelf
west of Ireland
The moraines on the Porcupine Bank and Slyne Trough (Peters
et al., 2015) and the large, arcuate WIM (Fig. 2) provide geomorphic
evidence for an extensive grounded BIIS on the shelf west of
Ireland. This geomorphic interpretation is also supported by the
presence of subglacial traction tills in cores from the shelf (Fig. 6).
These moraines appear to be restricted to less than ~350 mbsl
(Fig. 2) and are oriented roughly parallel to the shelf break, suggesting
that at maximum extent the BIIS consisted of a grounded
ice mass that was calving into deep water (cf. Sejrup et al., 2005);
this large ice mass was likely fed by ice draining from counties
Mayo, Galway, Clare and Kerry, Ireland (Greenwood and Clark,
2009; Fig. 2a). Based on the geomorphic and sedimentary evidence
documented in this study, the ice sheet at its maximum
extended to within ~20 km of the shelf break. This is compatible
with previous geomorphic reconstructions and conceptual models
of the western BIIS (e.g. Sejrup et al., 2005; Greenwood and Clark,
2009).
The presence of overconsolidated subglacial traction tills with
low water content at the base of the sedimentary sequence across
the shelf west of Ireland (Fig. 6) suggests that initial BIIS advance
was extensive and may have occurred largely without ice streaming
(cf. Wellner et al., 2001; Shipp et al., 2002; Dowdeswell et al., 2004;
Evans et al., 2005; O Cofaigh et al., 2005, 2007). This interpretation
is also consistent with delayed IRD sedimentation on the Porcupine
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124118
Bank in relation to the Rosemary Bank (Scourse et al., 2009).
The till blanket that covers the continental shelf (Fig. 6) is dated
to 24,067 Cal BP (Table 2) in the Slyne Trough, within 50 km of the
shelf break (cf. Peters et al., 2015, Fig. 2a). This is compatible with
other radiocarbon ages that constrain initial advance of Irish Sea ice
in the Celtic Sea (Fig. 1) to 24,000 Cal BP (O Cofaigh and Evans,
2007; O Cofaigh et al., 2012). An AMS radiocarbon age from the
glaciomarine sediment that drapes the WIM (Figs. 6 and 7) constrains
the timing of ice-proximal, deglacial deposition to
20,944 Cal BP (core 10-35, Table 2). This age indicates that the WIM
is blanketed by glaciomarine sediment from Midlandian deglaciation
and its stratigraphic relationship to the underlying subglacial
till suggests in turn that the WIM was deposited by the last BIIS
during MIS 2. The 20,944 Cal BP age is ~900 years younger than
glaciomarine sedimentation in the Slyne Trough (21,841 Cal BP, core
10-45, Table 2); this suggests that initial ice retreat may have
started over the bathymetrically deep Slyne Trough.
5.2. Ice shelf chronology and BIIS dynamics during GLGZW
formation
Sedimentological and micropalaeontological evidence in the
Slyne Trough (Peters et al., 2015) indicate local BIIS uncoupling and
ice shelf formation by 21,841 Cal BP (core 10-45, Table 2, Figs. 6 and
7). Ice shelf formation precedes the rapid sea-level rise recorded at
Kilkeel Steps (Clark et al., 2004; Fig. 1) by ~2800 years and follows
the Greenland Interstadial 2 warming event (~21,000 Cal BP)
recorded by the GRIP ice core d18
O record (Dansgaard et al., 1993;
Bj€orck et al., 1998) by <200 years. This chronologic sequence suggests
that initial BIIS thinning over the Slyne Trough may have been
the result of climate amelioration or variations in Atlantic Meridional
Overturning Circulation (cf. Clark et al., 2012b).
The presence of the GLGZW indicates that the ice sheet underwent
a still-stand on the mid-shelf following initial recession from
the WIM. This must have occurred after the formation of an ice
shelf over the Slyne Trough at 21,841 Cal BP, (core 10-45, Table 2;
Peters et al., 2015) and prior to the biomineralization of
Fig. 7. Schematic map of study area showing isochrones derived from AMS radiocarbon data (Table 2) and analyses of geomorphology and stratigraphy that constrain the BIIS
marine-terminating margin during retreat. AMS radiocarbon ages are displayed in coloured text that depicts the LFA that was sampled to provide the date; LFA 4 ¼ postglacial
reworking, LFA 2 ¼ glaciomarine deposition, LFA 1 ¼ advance-phase till or glacitectonite. Cosmogenic ages referred to in the text are shown with ages and abbreviations introduced
in Table 3. Flowlines are generalised and determined geomorphically. The labelled blue arrow illustrates ice shelf retreat at a rate of ~24 m/yr and the labelled red arrows illustrate
grounding line retreat at ~74 m/yr west of the Galway Lobe Grounding-Zone Wedge (GLGZW) and ~113 m/yr east of the GLGZW. (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article.)
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 119
foraminiferal tests found in core 14-54 (21,161 Cal BP, Figs. 6 and 7).
This means that the grounding zone retreated eastward across
~50 km of seafloor over a period of ~680 years at a rate of ~74 m/yr.
This ~50 km section of seafloor is devoid of any morainic ridges,
suggesting that grounding zone retreat back to the mid-shelf and
the position of the GLGZW was probably continuous and uninterrupted
by stillstands or readvances (Dowdeswell et al.,
2008). Comparable retreat rates are recorded for modern marineterminating
outlet glaciers on the Antarctic Peninsula (Cook et al.,
2005) and similar episodic retreat has been proposed for other
areas of the western BIIS marine margin (Bradwell et al., 2008; O
Cofaigh et al., 2010).
The arcuate shape of the GLGZW closely mimics the orientation
of the 200-m isobath (Fig. 2a), suggesting that grounding zone
stabilization was controlled largely by bathymetry (cf. Ship et al.,
1999). Thus, the GLGZW likely records a period of BIIS reconfiguration
and stabilization along the shallower, mid continental shelf.
This reconfiguration likely resulted from increased ice buoyancy on
the outer shelf (i.e. areas of modern depths >200 m, Fig. 2a), either
from sea level rise (Eyles and McCabe, 1989, McCabe et al., 2005) or
ice thinning caused by AMOC variations (Clark et al., 2012b). The
composite shape of the GLGZW, with at least two distinct flattopped
ridges (Fig. 3ii, iii), most likely records minor oscillations
of the grounding zone. Specifically, the westernmost flat-topped
ridge likely records a period of grounding-zone wedge deposition
below a thicker ice shelf that created a smaller vertical accommodation
space (Batchelor and Dowdeswell, 2015). Further evidence
of a dynamic grounding zone is provided by glaciotectonised plumite
deposits near the flanks of the GLGZW. That this glaciotectonism
occurred without the formation of distinct ice-terminal
landforms indicates periods of minor readvance or recoupling
along the grounding zone.
Extensive iceberg scouring west of the GLGZW indicates that the
ice shelf that flanked the western BIIS marine margin retreated via
calving. This interpretation is consistent with geomorphic investigations
on the shelf west of Donegal Bay, north of the study
area (Fig. 1). There, extensive zones of iceberg scouring are interpreted
to record rapid ice loss by calving following BIIS maximum
extension to the shelf break (Benetti et al., 2010; Dunlop et al., 2010;
O Cofaigh et al., 2010). A period of increased iceberg production
following ice shelf formation is also evidenced by the massive clastrich
diamicton commonly overlying till on the continental shelf
(Fig. 6). Radiocarbon dates from the iceberg-rafted deposits in cores
10-35 on the WIM and 10-45 in the Slyne Trough restrict this phase
of calving retreat to (21,000 Cal BP (Table 2, Fig. 6), which is
roughly coincident with post- Heinrich Event 2 peaks in BIIS-IRD
production in cores from the Porcupine Seabight and Rockall
Trough (Peck et al., 2006; Scourse et al., 2009). The ice shelf disintegrated,
or at least its terminus retreated to a point on the
continental shelf east of the GLGZW by 18,517 Cal BP, when
foraminiferal tests were biomineralized in core 14-53 (Table 2). This
allows an estimate for ice shelf duration of (2500 years and a
minimum rate of ice shelf retreat over the ~60-km long expanse
between the Slyne Trough and the GLGZW, to be calculated at
~24 m/yr. However, it is likely that this retreat was not steady state
and that rapid sea-level rise occurring ~19,000 Cal BP (Clark et al.,
2004) may have exacerbated the calving rate of this ice margin.
Grounding-zone wedges with similar dimensions to the GLGZW
that occupy other glaciated continental shelves offshore of
Antarctica and Norway (e.g. Shipp et al., 2002; Ottesen et al., 2005)
are associated with stillstands during episodic ice retreat (Howat
and Domack, 2003; Dowdeswell et al., 2008). Although no sediment
flux rate can yet be calculated for the GLGZW, groundingzone
wedges with comparable dimensions have typically been
shown to record stillstands that lasted decades or centuries
(Batchelor and Dowdeswell, 2015, and references therein). Thus, it
is likely that the GLGZW formed along a grounding zone that was
stable for a relatively long time while its vertical accretion was
constrained by a large, buttressing ice shelf. It is evident that the
stillstand marked by the GLGZW occurred after the deposition of till
in the Slyne Trough dated to ~24,067 Cal BP (Peters et al., 2015;
Table 2, Figs. 6 and 7) and prior to the formation of the nearby
Galway Lobe Readvance Moraine <18,517 Cal BP (core 14-53,
Table 2, Figs. 5e7). This ~5500 year period represents the maximum
duration of grounding-zone proximal deposition along the GLGZW.
However, it is more likely that the GLGZW was deposited over a
period of <3300 years, defined by the age of ice-shelf inception over
the Slyne Trough (21,841 Cal BP) and the age of foraminiferal tests
within glacitectonised sediment to the east (18,517 Cal BP).
5.3. BIIS dynamics and chronology following ice shelf break up
A glacitectonised plumite in core 14-53 (Figs. 5 and 6) contains
calcareous benthic foraminifera that provide an AMS radiocarbon
date of 18,517 Cal BP (Table 2). The date provides a maximum age
for the glaciotectonisation. This date is approximately 2700 years
older than previous estimates of the timing of deglaciation from
surface exposure dating from the west coast of Ireland, which was
calculated to ~15.85 ka BP from four cosmogenic ages (Table 3)
(Bowen et al., 2002; Ballantyne et al., 2008). Collectively these dates
constrain the formation of the Galway Lobe Readvance Moraine to a
~2700 year window from 18,400 Cal BP to >15.85 ka BP (Fig. 7).
Radiocarbon ages on ice sheet readvance(s) in eastern Ireland
during the Clogher Head Stadial (<18,200e17,100 Cal BP, McCabe
et al., 2007; Clark et al., 2012b) and cosmogenic ages interpreted
to constrain the age of readvances in western Ireland during the
Killard Point Stadial (Table 3; !17.1e!16.0 ka BP, Clark et al., 2012b)
fall within the range of age constraints for this readvance on the
Irish continental shelf (Fig. 7). This suggests that climactic forcing
during those stadials may have influenced BIIS readvance(s)
offshore of western Ireland; however, based on our existing chronology
the role of internal glaciodynamic forcing mechanisms (e.g.
changes in subglacial bed conditions) cannot be ruled out.
The duration of ice occupation at the Galway Lobe Readvance
Moraine is unknown, but the mean post-LGM deglaciation age for
the west coast of Ireland (~15.85 ka BP, Table 3) indicates that the
BIIS retreated from this position across ~130 km of continental shelf
to the shore of County Clare (Figs. 2a and 7) at a minimum rate of
~48 m/yrdcalculated assuming moraine formation at the time of
age-constraining biomineralization (18,517 Cal BP). If the Galway
Ice Lobe retreated following a Killard Point readvance at
~17,000 Cal BP, this would provide a retreat rate of ~113 m/yr.
The general southward progression of the maximum extent of
the three largest ice marginal features (WIM, GLGZW and Galway
Lobe Readvance Moraine) indicates a possible increase in topographic
constraint (likely from the Maumturk Mountains, Fig. 2a)
on ice dynamics imposed on the thinning, retreat-phase BIIS (cf.
Bradwell et al., 2008; Greenwood and Clark, 2009; Clark et al.,
2012a). Unlike the three largest ice-terminal landforms in the
study area, the Connemara Lobe and Mayo Lobe moraines have an
arcuate shape that is unrelated to local bathymetry and suggestive
of outflow from local, terrestrial source areas (Fig. 2a). These lobate
moraines likely record the development of topographicallyrestricted
ice-flow outlets on the west of Ireland that developed
following BIIS thinning (cf. Bradwell et al., 2008). The moraines are
well preserved, truncate the GLGZW and are situated roughly
within the radius of the WIM (Figs. 2a and 7), suggesting that these
smaller moraines record BIIS readvances during overall retreat.
These readvances identify the evolution of deglacial ice drainage
regimes west of Ireland (cf. Greenwood and Clark, 2008). The
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124120
Connemara Lobe Moraine is at least partially composed of soft,
probably dilatant, subglacial till overlain in areas by glaciotectonised
glaciomarine sediment (cores 10-40 and 14-59; Fig. 6). The
dilatant till suggests periods of accelerated ice flow or ice streaming
over bed material with high pore water pressures (O Cofaigh and
Evans, 2001; O Cofaigh et al., 2007). The till sampled in core 14-
59 incorporates calcareous marine biogenic material that provides
an age for the deposition of the top 4 m of the Connemara Lobe
Moraine of 15,148 Cal BP (Table 2; Fig. 7). This suggests that the
Connemara Lobe advanced to this point on the continental shelf by
~15,000 Cal BP. Alternatively, the moderately stiff (~40 kPa) diamicton
at the base of core 14-59 could have been formed by iceberg
turbation or debris flows; however the high density, low water
content and folding of this deposit (Fig. 5) are more compatible
with and interpretation of subglacial till. This indicates that the
western BIIS had a marine margin 1000 years after the Killard Point
Stadial. We suggest that internal (glaciodynamic) forcing mechanism(s)
are most likely to have triggered this readvance. However,
we cannot rule out the possibility that readvance was climaticallydriven
and associated with the Nahanagan Stadial (Younger Dryas)
although this would imply significantly more extensive glacier
growth during this period in western Ireland than has hitherto
been proposed.
Although no direct age constraints are available for the Mayo
Lobe Moraine, it was likely deposited after the formation of the
GLGZW at 21,841e18,517 Cal BP. An estimated age of deposition of
<21,000 Cal BP for the Mayo Lobe Moraine is ~5400 years older
than the cosmogenic ages that define the Killard Point Readvance at
Furnace Lough (Clark et al., 2009b; Table 3). Thus, the Mayo Lobe
Moraine is interpreted as pre-dating the Killard Point Stadial.
The readvances identified here on the basis of the core sedimentology
and associated geochronology post-date break-up of the
ice-shelf farther offshore in the Slyne Trough (Peters et al., 2015,
Fig. 7). This suggests that the loss of the buttressing, floating ice
mass could have initiated a period of accelerated westward flow
similar to that which has been observed for ice sheet outlets around
the Antarctic Peninsula (cf. Scambos et al., 2004).
6. Conclusions
 The ~80-km long, arcuate West Ireland Moraine reaches to
within 20 km of the shelf break west of Ireland and records the
minimum westward, grounded extension of the BIIS
24,067 Cal BP. This moraine is constructed, at least in part,
from subglacial till and is capped by glaciomarine sediment that
was deposited ~2100 Cal BP. This age constraint supports an
interpretation of corrugation genesis from the periodic
grounding of an overriding ice shelf that extended over the
Slyne Trough.
 Overconsolidated subglacial till common at the base of the
stratigraphic sequence across the continental shelf indicates
that the BIIS advanced to its maximum offshore position
without streaming or surging. However, after the initiation of
retreat across the continental shelf west of Ireland
( 21,841 Cal BP), the BIIS likely experienced accelerated flow
over a dilatant till with increased pore water pressures. These
periods of accelerated flow formed readvance moraines.
 An ice shelf formed over the Slyne Trough and extended over the
West Ireland Moraine following the advance of grounded ice
24,067 Cal BP. AMS radiocarbon dated sub-ice shelf deposits
indicate that this ice shelf formed 21,841 Cal BP over the Slyne
Trough and within ~800 years of the Greenland Interstadial 2
warming event (~21,000 BP), suggesting that the westernmost
BIIS may have uncoupled and formed a floating ice shelf after
thinning that was induced by variations in the Atlantic Meridional
Overturning Circulation. Two AMS radiocarbon dates from
ice-proximal, IRD-rich glaciomarine sediment constrain the
calving-dominated, eastward retreat of this ice shelf to have
begun by approximately <21,000 Cal BP.
 The BIIS grounding zone retreated during a single event,
without forming moraines or grounding-zone wedges, across
~50 km of seafloor. This retreat likely took place over a period of
~680 years at a rate of ~74 m/yr. The grounding zone stabilised
near and parallel to the 200-m isobath, suggesting a bathymetric
control on its retreat.
 The bathymetrically-controlled grounding zone most likely
persisted in roughly the same location for a maximum of <3300
Table 3
Referenced, cosmogenic nuclide ages from near the coast of counties Mayo, Galway and Clare, Ireland.
Age (ka) Avg. age
(ka)a
Isotope Location (abbreviation)a
Elevation
(m asl)
Interpretation Reference (age) Reference
(interpretation)
14.5 ± 0.9 13.73 10
Be Lough Acorrymore (LA) 287 Nahanagan readvance age constraint Ballantyne et al. (2008) Ballantyne et al. (2008)
11.7 ± 0.7 10
Be 190
15.0 ± 1.0 10
Be 198
15.1 ± 1.0 14.15 10
Be Mweelrea (MW) 305 Post-LGM deglaciation age constraint Ballantyne et al. (2008) Ballantyne et al. (2008)
13.2 ± 0.8 10
Be 650
15.3 ± 1.0 e 36
Cl Kilkee (KE) 66 Post-LGM deglaciation age constraint Bowen et al. (2002) Bowen et al. (2002)
15.7 ± 1.0 15.6 ± 0.4 10
Be Furnace Lough (FL) 14-74 Killard Point readvance age constraint Clark et al. (2009b) Clark et al. (2009b)
15.2 ± 1.0
14.3 ± 0.9
17.3 ± 1.0
16.4 ± 1.1
16.1 ± 0.7
15.4 ± 0.9
14.1 ± 0.7
16.7 ± 1.0 e 10
Be Farnaght Hill (FH) 125 Post-LGM deglaciation age constraint Ballantyne et al. (2008) Ballantyne et al. (2008)
16.4 ± 1.9 16.95 36
Cl Lough Nakeeroge (LN) 5 Killard Point readvance age constraint Bowen et al. (2002) Ballantyne et al. (2008)
17.5 ± 3.7 36
Cl 5
17.1 ± 1.1 17.15 10
Be Anaffrin East Col (AC) 440 Post-LGM deglaciation age constraint Ballantyne et al. (2008) Ballantyne et al. (2008)
17.2 ± 1.1 10
Be 440
a
Used in Fig. 7.
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 121
years. Whilst in this position, the flanking ice shelf buttressed
ice flow and added to the stability of the retreat stillstand.
During this stillstand, subglacial sediment delivery along the
grounding zone resulted in the formation of a large (~150-km
long, up to 22-km wide, and up to 20-m thick) groundingzone
wedge. This grounding-zone wedge is termed the Galway
Lobe grounding-zone wedge (GLGZW) because it is interpreted
to have formed against ice draining from Galway Bay,
Ireland (the Galway Lobe).
 During the stillstand that formed the GLGZW, the position of the
grounding zone on the shelf was relatively stable, but several
minor fluctuations in ice dynamics formed the GLGZW as a
composite landform. Glaciotectonised glaciomarine deposits are
common on the flanks of the GLGZW, and record periods of local
ice recoupling or readvance, one of which is dated to
21,161 Cal BP. Retreat from the GLGZW was complete by
<18,517 Cal BP.
 BIIS marine margin retreat was characterised by a
topographically-driven modification of ice flow dynamics
probably associated with overall ice sheet thinning. This
increased topographic influence on the retreating BIIS is evidenced
by migrating positions of moraines and a general eastwards
decrease in moraine size. These landform characteristics
record the increasingly lobate structure of the BIIS during
retreat, which is likely the signature of ice sheet thinning.
 The Galway Lobe Readvance Moraine was formed following
grounding zone retreat <18,517 Cal BP. Although no direct age
constraints are available, the smaller Mayo Lobe Moraine was
likely deposited after the formation of the Galway Lobe Readvance
Moraine and is at least younger than the
21,841e18,517 Cal BP GLGZW. The Connemara Lobe moraine was
formed by an ice lobe that advanced onto the continental shelf
after the formation of the GLGZW, likely <15,148 Cal BP and
truncated surrounding morainic landforms. This indicates that
the ice sheet underwent readvances on the shelf during deglaciation
although our existing chronology does not yet allow us
to determine conclusively if these readvances were glaciologically
(internally) driven or a response to climatic forcing.
 Estimated rates of BIIS retreat vary drastically. Following ice
shelf formation over the Slyne Trough, the BIIS's grounding zone
retreated at a rate of ~74 km/yr before stabilizing and depositing
the GLGZW. The ice shelf terminus retreated from the Porcupine
Bank to the GLGZW at a minimum rate of ~24 m/yr before the
grounding zone withdrew from the stillstand marked by the
GLGZW. Following GLGZW deposition, BIIS marine margin
retreat was punctuated by at least one readvance. During this
continued retreat the marine-terminating ice sheet comprised
of several discreet lobes, the largest of which formed the Galway
Lobe Readvance Moraine prior to retreating ~130 km to the west
coast of Ireland at a likely rate of ~113 m/yr. This highlights a
drastic increase in marine margin retreat following the loss of
the ~50-km wide ice shelf.
Acknowledgements
JLP acknowledges a University of Ulster PhD studentship and
Vice-Chancellor's Scholarship from 2013 to 2016. This work was
supported by the NERC Radiocarbon Facility NRCF010001 (allocation
number 1722.0613), NERC consortium grant NE/J009768/1, for
radiocarbon dating. Further radiocarbon dating support was provided
by the International Association of Sedimentologists (Postgraduate
Research Grant, 1st session, 2014) and the Raidio Teilifís
Eireann (RTE) broadcasting company for the television programme
‘The Investigators’. Thanks to Derek Fabel at the Scottish Universities
Environmental Research Centre, University of Glasgow, for his
help during radiocarbon calibrations. Grateful thanks to Christian
Wilson of Ocean DTM for Olex data-acquisition support, Robin
Edwards at Trinity College, Dublin for the use of his MALVERN
Mastersizer©
for grain size analyses and sample processing, and
Stephen McCarron at the National University of Ireland, Maynooth
for the use of the GEOTEK multi-sensor core logger. We would like
to thank the Captain and Crew of the RV Celtic Explorer, cruise
participants and Mr Aodhan Fitzgerald of the Marine Institute, for
their invaluable assistance during cruises CE10008 and CE14004
(WICPro). The research surveys were carried out under the Sea
Change strategy with the support of the Marine Institute and the
Marine Research Sub-programme of the National Development
Plan 2007e2013. The authors also thank two anonymous reviewers,
whose insightful comments have helped to improve this
manuscript.
References
Amorosi, A., Colalongo, M., Pasini, G., Preti, D., 1999. Sedimentary response to Late
Quaternary sea-level changes in the Romagna coastal plain (northern Italy).
Sedimentology 46 (1), 99e121.
Andreassen, K., Winsborrow, M.C., Bjarnadottir, L.R., Rüther, D.C., 2014. Ice stream
retreat dynamics inferred from an assemblage of landforms in the northern
Barents Sea. Quat. Sci. Rev. 92, 246e257.
Ballantyne, C.K., Stone, J.O., McCarroll, D., 2008. Dimensions and chronology of the
last ice sheet in western Ireland. Quat. Sci. Rev. 27 (3e4), 185e200.
Barrie, J.V., Conway, K.W., 2002. Rapid sea-level change and coastal evolution on the
Pacific margin of Canada. Sediment. Geol. 150 (1), 171e183.
Batchelor, C., Dowdeswell, J., 2015. Ice-sheet grounding-zone wedges (GZWs) on
high-latitude continental margins. Mar. Geol. 363, 65e92.
Belderson, R.H., Kenyon, N.H., Wilson, J.B., 1973. Iceberg plough marks in the
northeast Atlantic. Palaeogeogr. Palaeoclimatol. Palaeoecol. 13, 215e224.
Benetti, S., Dunlop, P., O Cofaigh, C., 2010. Glacial and glacially-related features on
the continental margin of northwest Ireland mapped from marine geophysical
data. J. Maps 6 (1), 14e29.
Benn, D.I., Evans, D.J., 1996. The interpretation and classification of subglaciallydeformed
materials. Quat. Sci. Rev. 15 (1), 23e52.
Bennett, M.M., Glasser, N.F., 2011. Glacial Geology: Ice Sheets and Landforms. John
Wiley & Sons.
Bentley, M., Evans, D., Fogwill, C., Hansom, J., Sugden, D., Kubik, P., 2007. Glacial
geomorphology and chronology of deglaciation, South Georgia, sub-Antarctic.
Quat. Sci. Rev. 26 (5), 644e677.
Bentley, M.J., Fogwill, C.J., Le Brocq, A.M., Hubbard, A.L., Sugden, D.E., Dunai, T.J.,
Freeman, S.P., 2010. Deglacial history of the west Antarctic ice sheet in the
Weddell Sea embayment: constraints on past ice volume change. Geology 38
(5), 411e414.
Bishop, P., Jones, E.J.W., 1979. Patterns of Glacial and Post-Glacial Sedimentation in
the Minches, North-west Scotland. In: Elsevier Oceanography Series, vol. 24,
pp. 89e194.
Bj€orck, S., Walker, M.J., Cwynar, L.C., Johnsen, S., Knudsen, K.L., Lowe, J.J.,
Wohlfarth, B., 1998. An event stratigraphy for the Last Termination in the North
Atlantic region based on the Greenland ice-core record: a proposal by the
INTIMATE group. J. Quat. Sci. 13 (4), 283e292.
Bond, G.C., Showers, W., Elliot, M., Evans, M., Lotti, R., Hajdas, I., Bonani, G.,
Johnson, S., 1999. The North Atlantic's 1-2 Kyr Climate Rhythm: Relation to
Heinrich Events, Dansgaard/Oeschger Cycles and the Little Ice Age. In: Mechanisms
of Global Climate Change at Millennial Time Scales, pp. 35e58.
Boulton, G.S., 1986. Push-moraines and glacier-contact fans in marine and terrestrial
environments. Sedimentology 33, 677e698.
Bowen, D.Q., Rose, J., McCabe, A.M., Sutherland, D.G., 1986. Correlation of quaternary
glaciations in England, Ireland, Scotland and Wales. Quat. Sci. Rev. 5,
299e340.
Bowen, D.Q., Phillips, F.M., McCabe, A.M., Knutz, P.C., Sykes, G.A., 2002. New data for
the last glacial maximum in Great Britain and Ireland. Quat. Sci. Rev. 21, 89e101.
Bradwell, T., Stoker, M.S., Golledge, N.R., Wilson, C.K., Merritt, J.W., Long, D.,
Everest, J.D., Hestvik, O.B., Stevenson, A.G., Hubbard, A.L., Finlayson, A.G.,
Mathers, H.E., 2008. The northern sector of the last British ice sheet: maximum
extent and demise. Earth Sci. Rev. 88 (3), 207e226.
Broecker, W.S., Kennett, J.P., Flower, B.P., Teller, J.T., Trumbore, S., Bonani, G.,
Wolfli, W., 1989. The routing of Laurentide ice-sheet meltwater during the
younger dryas cold event. Nature 341, 318e321.
Chang, T.S., Flemming, B.W., Tilch, E., Bartholom€a, A., W€ostmann, R., 2006. Late
Holocene stratigraphic evolution of a back-barrier tidal basin in the East Frisian
Wadden Sea, southern North Sea: transgressive deposition and its preservation
potential. Facies 52 (3), 329e340.
Clark, P.U., McCabe, A.M., Mix, A.C., Weaver, A.J., 2004. Rapid sea level rise at 19,000
years ago and its global implications. Science 304, 1141e1144.
Clark, J., McCABE, A., Schnabel, C., Clark, P.U., Freeman, S., Maden, C., Xu, S., 2009a.
10Be chronology of the last deglaciation of County Donegal, northwestern
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124122
Ireland. Boreas 38 (1), 111e118.
Clark, J., McCabe, A.M., Schnabel, C., Clark, P.U., McCarron, S., Freeman, S.P.H.T.,
Maden, C., Xu, S., 2009b. Cosmogenic Be-10 chronology of the last deglaciation
of western Ireland, and implications for sensitivity of the Irish Ice Sheet to
climate change. Geol. Soc. Am. Bull. 121 (1e2), 3e16.
Clark, C.D., Hughes, A.L.C., Greenwood, S.L., Jordan, C., Sejrup, H.P., 2012a. Pattern
and timing of retreat of the last British-Irish ice sheet. Quat. Sci. Rev. 44,
112e146.
Clark, J., McCabe, A.M., Bowen, D.Q., Clark, P.U., 2012b. Response of the Irish ice
sheet to abrupt climate change during the last deglaciation. Quat. Sci. Rev. 35,
100e115.
Cook, A., Fox, A., Vaughan, D., Ferrigno, J., 2005. Retreating glacier fronts on the
Antarctic Peninsula over the past half-century. Science 308 (5721), 541e544.
Cowan, E.A., Cai, J., Powell, R.D., Clark, J.D., Pitcher, J.N., 1997. Temperate glacimarine
varves: an example from Disenchantment Bay, southern Alaska. J. Sediment.
Res. 67 (3).
Dansgaard, W., Johnsen, S.J., Clausen, H.B., Dahl-Jensen, D., Gundestrup, N.S.,
Hammer, C.U., Hvidberg, C.S., Steffensen, J.P., Svelnbj€ornsdottir, A.E., Jouzel, J.,
Bond, G., 1993. Evidence for general instability of past climate from a 250-kyr
ice-core record. Nature 364 (6434), 218e220.
Dowdeswell, J.A., Villinger, H., Whittington, R.J., Marienfeld, P., 1993. Iceberg
scouring in Scoresby Sund and on the east Greenland continental shelf. Mar.
Geol. 111, 37e53.
Dowdeswell, J.A., Whittington, R.J., Marienfeld, P., 1994. The origin of massive diamicton
facies by iceberg rafting and scouring, Scoresby Sund, East Greenland.
Sedimentology 41 (1), 21e35.
Dowdeswell, J., Maslin, M., Andrews, J., McCave, I., 1995. Iceberg production, debris
rafting, and the extent and thickness of Heinrich layers (H-1, H-2) in North
Atlantic sediments. Geology 23 (4), 301e304.
Dowdeswell, J., Whittington, R., Jennings, A., Andrews, J., Mackensen, A.,
Marienfeld, P., 2000. An origin for laminated glacimarine sediments through
sea-ice build-up and suppressed iceberg rafting. Sedimentology 47, 557e576.
Dowdeswell, J., O Cofaigh, C., Pudsey, C., 2004. Continental slope morphology and
sedimentary processes at the mouth of an Antarctic palaeo-ice stream. Mar.
Geol. 204 (1), 203e214.
Dowdeswell, J., Ottesen, D., Evans, J., O Cofaigh, C., Anderson, J., 2008. Submarine
glacial landforms and rates of ice-stream collapse. Geology 36 (10), 819e822.
Dunlop, P., Shannon, R., McCabe, M., Quinn, R., Doyle, E., 2010. Marine geophysical
evidence for ice sheet extension and recession on the Malin shelf: new evidence
for the western limits of the British Irish ice sheet. Mar. Geol. 276 (1), 86e99.
Evans, J., Pudsey, C.J., OCofaigh, C., Morris, P., Domack, E., 2005. Late Quaternary
glacial history, flow dynamics and sedimentation along the eastern margin of
the Antarctic Peninsula ice sheet. Quat. Sci. Rev. 24, 741e774.
Evans, D.J.A., Phillips, E.R., Hiemstra, J.F., Auton, C.A., 2006a. Subglacial till: formation,
sedimentary characteristics and classification. Earth Sci. Rev. 78 (1),
115e176.
Evans, J., Dowdeswell, J.A., O Cofaigh, C., Benham, T.J., Anderson, J.B., 2006b. Extent
and dynamics of the West Antarctic ice sheet on the outer continental shelf of
Pine Island Bay during the last glaciation. Mar. Geol. 230 (1), 53e72.
Eyles, C.H., 1988. A model for striated boulder pavement formation on glaciated,
shallow-marine shelves: an example from the Yakataga Formation, Alaska.
J. Sediment. Res. 58 (1), 62e71.
Eyles, N., McCabe, A.M., 1989. The Late Devensian (< 22,000 BP) Irish Sea Basin: the
sedimentary record of a collapsed ice sheet margin. Quat. Sci. Rev. 8 (4),
307e351.
Eyles, N., Eyles, C.H., Miall, A.D., 1983. Lithofacies types and vertical profile models;
an alternative approach to the description and environmental interpretation of
glacial diamict and diamictite sequences. Sedimentology 30 (3), 393e410.
Feyling-Hanssen, R.W., 1972. The foraminifer Elphidium excavatum (Terquem) and
its variant forms. Micropaleontology 18 (3), 337e354.
Fischer, U., Clarke, G.K.C., Blatter, H., 1999. Evidence for temporally varying “sticky
spots” at the base of Trapridge Glacier, Yukon Territory, Canada. J. Glaciol. 45,
352e360.
Flemming, B., Schubert, H., Hertweck, G., Müller, K., 1992. Bioclastic tidal-channel
lag deposits: a genetic model. Senckenberg. Marit. 22, 109e129.
Fyfe, J.A., Long, D., Evans, D., 1993. The Geology of the Malin-hebrides Sea Area.
British Geological Survey, United Kingdom. Offshore Regional Report. HMSO,
London.
Glasser, N.F., Scambos, T.A., Bohlander, J., Truffer, M., Pettit, E., Davies, B.J., 2011.
From ice-shelf tributary to tidewater glacier: continued rapid recession, acceleration
and thinning of R€ohss Glacier following the 1995 collapse of the Prince
Gustav ice shelf, Antarctic Peninsula. J. Glaciol. 57 (203), 397e406.
Greenwood, S.L., Clark, C.D., 2008. Subglacial bedforms of the Irish ice sheet. J. Maps
4 (1), 332e357.
Greenwood, S.L., Clark, C.D., 2009. Reconstructing the last Irish ice sheet 2: a
geomorphologically-driven model of ice sheet growth, retreat and dynamics.
Quat. Sci. Rev. 28 (27e28), 3101e3123.
Grobe, H., 1987. A simple method for the determination of ice-rafted debris in
sediment cores. Polarforschung 57 (3), 123e126.
Hald, M., Korsun, S., 1997. Distribution of modern benthic foraminifera from fjords
of Svalbard, European Arctic. J. Foraminifer. Res. 27 (2), 101e122.
Hald, M., Steinsund, P.I., Dokken, T., Korsun, S., Polyak, L., Aspeli, R., 1994. Recent and
late Quaternary distribution of Elphidium excavatum f. clavata in Arctic seas.
Cushman Found. Spec. Publ. 32, 141e153.
Hill, J.C., Gayes, P.T., Driscoll, N.W., Johnstone, E.A., Sedberry, G.R., 2008. Iceberg
scours along the southern US Atlantic margin. Geology 36 (6), 447e450.
Hillenbrand, C.D., Baesler, A., Grobe, H., 2005. The sedimentary record of the last
glaciation in the western Bellingshausen Sea (West Antarctica): implications for
the interpretation of diamictons in a polar-marine setting. Mar. Geol. 216 (4),
191e204.
Hillenbrand, C.D., Larter, R.D., Dowdeswell, J.A., Ehrmann, W., O Cofaigh, C.,
Benetti, S., Grobe, H., 2010. The sedimentary legacy of a palaeo-ice stream on
the shelf of the southern Bellingshausen Sea: clues to west Antarctic glacial
history during the Late Quaternary. Quat. Sci. Rev. 29 (19), 2741e2763.
Hillenbrand, C.-D., Kuhn, G., Smith, J.A., Gohl, K., Graham, A.G., Larter, R.D.,
Klages, J.P., Downey, R., Moreton, S.G., Forwick, M., Vaughan, D.G., 2013.
Grounding-line retreat of the west Antarctic ice sheet from inner Pine Island
Bay. Geology 41 (1), 35e38.
Howat, I.M., Domack, E.W., 2003. Reconstructions of western Ross Sea palaeo-icestream
grounding zones from high-resolution acoustic stratigraphy. Boreas 32
(1), 56e75.
Jakobsson, M., Anderson, J.B., Nitsche, F.O., Dowdeswell, J.A., Gyllencreutz, R.,
Kirchner, N., Mohammad, R., O'Regan, M., Alley, R.B., Anandakrishnan, S.,
Eriksson, B., Kirshner, A., Fernandez, R., Stolldorf, T., Minzoni, R., Majewski, W.,
2011. Geological record of ice shelf break-up and grounding line retreat, Pine
Island Bay, West Antarctica. Geology 39 (7), 691e694.
Jonsson, S.A., Schomacker, A., Benediktsson, I.€O., Ingolfsson, O., Johnson, M.D., 2014.
The drumlin field and the geomorphology of the Múlaj€okull surge-type glacier,
central Iceland. Geomorphology 207, 213e220.
Kilfeather, A.A., O Cofaigh, C., Lloyd, J.M., Dowdeswell, J.A., Xu, S., Moreton, S.G.,
2011. Ice-stream retreat and ice-shelf history in Marguerite Trough, Antarctic
Peninsula: sedimentological and foraminiferal signatures. Geol. Soc. Am. Bull.
123 (5e6), 997e1015.
Lecavalier, B.S., Milne, G.A., Simpson, M.J., Wake, L., Huybrechts, P., Tarasov, L.,
Kjeldsen, K.K., Funder, S., Long, A.J., Woodroffe, S., Dyke, A.S., 2014. A model of
Greenland ice sheet deglaciation constrained by observations of relative sea
level and ice extent. Quat. Sci. Rev. 102, 54e84.
Lian, O.B., Hicock, S.R., 2000. Thermal conditions beneath parts of the last Cordilleran
ice sheet near its centre as inferred from subglacial till, associated sediments,
and bedrock. Quat. Int. 68, 147e162.
L€owemark, L., Chao, W.S., Gyllencreutz, R., Hanebuth, T.J., Chiu, P.Y., Yang, T.N.,
Su, C.C., Chuang, C.K., Dominguez, D.C.L., Jakobsson, M., 2015. Variations in
glacial and interglacial marine conditions over the last two glacial cycles off
northern Greenland. Quat. Sci. Rev. (in press), ISSN: 0277e3791.
Lucchi, R., Camerlenghi, A., Rebesco, M., Colmenero-Hidalgo, E., Sierro, F.,
Sagnotti, L., Urgeles, R., Melis, R., Morigi, C., Barcena, M.-A., Giorgetti, G.,
Villa, G., Persico, D., Flores, J.-A., Rigual-Hernandez, A.S., Pedrosa, M.T., Macri, P.,
Caburlotto, A., 2013. Postglacial sedimentary processes on the Storfjorden and
Kveithola trough mouth fans: significance of extreme glacimarine sedimentation.
Glob. Planet. Change 111, 309e326.
McCabe, A.M., 1986. Glaciomarine facies deposited by retreating tidewater glaciers:
an example from the Late Pleistocene of Northern Ireland. J. Sediment. Res. 56
(6).
McCabe, A.M., Clark, P.U., 1998. Ice-sheet variability around the North Atlantic
Ocean during the last deglaciation. Nature 392 (6674), 373e377.
McCabe, A., Clark, P.U., 2003. Deglacial chronology from county Donegal, Ireland:
implications for deglaciation of the BritisheIrish ice sheet. J. Geol. Soc. 160 (6),
847e855.
McCabe, M., Knight, J., McCarron, S., 1998. Evidence for Heinrich event 1 in the
British Isles. J. Quat. Sci. 13 (6), 549e568.
McCabe, A., Clark, P., Clark, J., 2005. AMS C-14 dating of deglacial events in the Irish
Sea Basin and other sectors of the BritisheIrish ice sheet. Quat. Sci. Rev. 24 (14),
1673e1690.
McCabe, A.M., Clark, P.U., Clark, J., 2007. Radiocarbon constraints on the history of
the western Irish ice sheet prior to the Last Glacial maximum. Geology 35 (2),
147e150.
Melis, R., Salvi, G., 2009. Late Quaternary foraminiferal assemblages from western
Ross Sea (Antarctica) in relation to the main glacial and marine Lithofacies. Mar.
Micropaleontol. 70 (1e2), 39e53.
Mudie, P.J., Keen, C.E., Hardy, I.A., Vis, G., 1984. Multivariate analysis and quantitative
paleoecology of benthic foraminifera in surface and Late Quaternary shelf
sediments, northern Canada. Micropaleontology 8, 283e313.
O Cofaigh, C., Dowdeswell, J.A., 2001. Laminated sediments in glacimarine environments:
diagnostic criteria for their interpretation. Quat. Sci. Rev. 20 (13),
1411e1436.
O Cofaigh, C., Evans, D.J., 2001. Deforming bed conditions associated with a major
ice stream of the last British ice sheet. Geology 29 (9), 795e798.
O Cofaigh, C., Evans, D.J., 2007. Radiocarbon constraints on the age of the maximum
advance of the BritisheIrish ice sheet in the Celtic Sea. Quat. Sci. Rev. 26 (9),
1197e1203.
O Cofaigh, C., Taylor, J., Dowdeswell, J.A., Rosell-Mele, A., Kenyon, N.H., Evans, J.,
Mienert, J., 2002. Geological evidence for sediment reworking on high-latitude
continental margins and its implications for palaeoceanography: insights from
the Norwegian-Greenland Sea. In: Dowdeswell, J.A., O'Cofaigh, C. (Eds.), Glacierinfluenced
Sedimentation on High-latitude Continental Margins, vol. 203.
Geological Society of London, Special Publication, pp. 325e348.
O Cofaigh, C., Dowdeswell, J.A., Allen, C.S., Hiemstra, J.F., Pudsey, C.J., Evans, J.,
Evans, D.J.A., 2005. Flow dynamics and till genesis associated with a marinebased
Antarctic palaeo-ice stream. Quat. Sci. Rev. 24, 709e740.
O Cofaigh, C., Evans, J., Dowdeswell, J.A., Larter, R.D., 2007. Till characteristics,
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124 123
genesis and transport beneath Antarctic paleo-ice streams. J. Geophys. Res.
Earth Surf. 112 (F3).
O Cofaigh, C., Dunlop, P., Benetti, S., 2010. Marine geophysical evidence for Late
Pleistocene ice sheet extent and recession off northwest Ireland. Quat. Sci. Rev.
44, 147e159.
O Cofaigh, C., Evans, D.J.A., Hiemstra, J.F., 2011. Formation of a stratified subglacial
‘till’ assemblage by ice-marginal thrusting and glacier overriding. Boreas 40 (1),
1e14.
O Cofaigh, C., Telfer, M.W., Bailey, R.M., Evans, D.J., 2012. Late Pleistocene chronostratigraphy
and ice sheet limits, southern Ireland. Quat. Sci. Rev. 44,
160e179.
Ottesen, D., Rise, L., Knies, J., Olsen, L., Henriksen, S., 2005. The Vestfjorden-Trænadjupet
palaeo-ice stream drainage system, mid-Norwegian continental shelf.
Mar. Geol. 218 (1), 175e189.
Peck, V., Hall, I., Zahn, R., Elderfield, H., Grousset, F., Hemming, S., Scourse, J., 2006.
High resolution evidence for linkages between NW European ice sheet instability
and Atlantic meridional overturning circulation. Earth Planet. Sci. Lett.
243 (3), 476e488.
Peters, J.L., Benetti, S., Dunlop, P., O Cofaigh, C., 2015. Maximum extent and dynamic
behaviour of the last BritisheIrish ice sheet west of Ireland. Quat. Sci. Rev. 128,
48e68.
Polyak, L., Gataullin, V., Gainanov, V., Gladysh, V., Goremykin, Yu, 2002. Kara Sea
expedition yields insight into LGM ice sheet extent. Eos 83 (46), 525e529.
Praeg, D., McCarron, S., Dove, D., O Cofaigh, C., Scott, G., Monteys, X., Facchin, L.,
Romeo, R., Coxon, P., 2015. Ice sheet extension to the Celtic Sea shelf edge at the
last glacial maximum. Quat. Sci. Rev. 111, 107e112.
Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Ramsey, C.B., Buck, C.E.,
Hai, C., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., Haflidason, H.,
Hajdas, I., Hatte, C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G., Hughen, K.A.,
Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A.,
Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M., van der Plicht, J., 2013.
IntCal13 and Marine13 radiocarbon age calibration curves 0-50,000 years cal
BP. Radiocarbon 55 (4), 1869e1887.
Rignot, E., Koppes, M., Velicogna, I., 2010. Rapid submarine melting of the calving
faces of West Greenland glaciers. Nat. Geosci. 3 (3), 187e191.
Rijsdijk, K.F., Owen, G., Warren, W.P., McCarroll, D., van der Meer, J.J., 1999. Clastic
dykes in over-consolidated tills: evidence for subglacial hydrofracturing at
Killiney Bay, eastern Ireland. Sediment. Geol. 129 (1), 111e126.
Robinson, S.G., Maslin, M.A., McCave, I.N., 1995. Magnetic susceptibility variations in
late Pleistocene deep-sea sediments of the N.E. Atlantic: implications for icerafting
and paleocirculation at the last glacial maximum. Paleoceanography
10, 221e250.
Robinson, A., Calov, R., Ganopolski, A., 2011. Greenland ice sheet model parameters
constrained using simulations of the Eemian interglacial. Clim. Past 7 (2),
381e396.
Saito, Y., Katayama, H., Ikehara, K., Kato, Y., Matsumoto, E., Oguri, K., Oda, M.,
Yumoto, M., 1998. Transgressive and highstand systems tracts and post-glacial
transgression, the East China Sea. Sediment. Geol. 122 (1), 217e232.
Scambos, T.A., Bohlander, J.A., Shuman, C.U., Skvarca, P., 2004. Glacier acceleration
and thinning after ice shelf collapse in the Larsen B embayment, Antarctica.
Geophys. Res. Lett. 31 (18), 1e4.
Scourse, J.D., Austin, W.E.N., Bateman, R.M., Catt, J.A., Evans, C.D.R., Robinson, J.E.,
Young, J.R., 1990. Sedimentology and Micropalaeontology of Glacimarine Sediments
from the Central and Southwestern Celtic Sea, vol. 53(1). Geological
Society, London, pp. 329e347. Special Publications.
Scourse, J.D., Haapaniemi, A.I., Colmenero-Hidalgo, E., Peck, V.L., Hall, I.R.,
Austin, W.E., Knutz, P.C., Zahn, R., 2009. Growth, dynamics and deglaciation of
the last BritisheIrish ice sheet: the deep-sea ice-rafted detritus record. Quat.
Sci. Rev. 28 (27), 3066e3084.
Sejrup, H.P., Haflidason, H., Aarseth, I., King, E., Forsberg, C.F., Long, D.,
Rokoengen, K., 1994. Late Weichselian glaciation history of the northern North
Sea. Boreas 23 (1), 1e13.
Sejrup, H.P., Hjelstuen, B.O., Dahlgren, K.I.T., Haflidason, H., Kuijpers, A., Nygård, A.,
Praeg, D., Stoker, M.S., Vorren, T.O., 2005. Pleistocene glacial history of the NW
European continental margin. Mar. Pet. Geol. 22, 1111e1129.
Sharp, M., 1982. Modification of clasts in lodgement tills by glacial erosion. J. Glaciol.
28 (100), 475e481.
Shevenell, A.E., Domack, E.W., Kernan, G.M., 1996. Record of Holocene palaeoclimate
change along the Antarctic Peninsula: evidence from glacial marine
sediments, Lallemand Fjord. In: Papers and Proceedings of the Royal Society of
Tasmania, vol. 130(2), pp. 55e64.
Ship, S., Anderson, J., Domack, E., 1999. Late PleistoceneeHolocene retreat of the
West Antarctic ice-sheet system in the Ross Sea: part 1dgeophysical results.
Geol. Soc. Am. Bull. 111 (10), 1486e1516.
Shipp, S.S., Wellner, J.S., Anderson, J.B., 2002. Retreat Signature of a Polar Ice Stream:
Sub-glacial Geomorphic Features and Sediments from the Ross Sea, Antarctica,
vol. 203. Special Publication-Geological Society of London, pp. 277e304.
Smeulders, G.G.B., Koho, K.A., de Stigter, H.C., Mienis, F., de Haas, H., van
Weering, T.C.E., 2014. Cold-water coral habitats of Rockall and Porcupine Bank,
NE Atlantic Ocean: sedimentary facies and benthic foraminiferal assemblages.
In: Deep Sea Research Part II: Topical Studies in Oceanography, vol. 99,
pp. 270e285.
Smith, J.A., Hillenbrand, C.D., Kuhn, G., Larter, R.D., Graham, A.G., Ehrmann, W.,
Moreton, S.G., Forwick, M., 2011. Deglacial history of the West Antarctic ice
sheet in the western Amundsen Sea embayment. Quat. Sci. Rev. 30 (5),
488e505.
Stalder, C., Spezzaferri, S., Rüggeberg, A., Pirkenseer, C., Gennari, G., 2014. Late
Weichselian deglaciation and early Holocene development of a cold-water coral
reef along the Lopphavet shelf (northern Norway) recorded by benthic foraminifera
and ostracoda. In: Deep Sea Research Part II: Topical Studies in
Oceanography, vol. 99, pp. 249e269.
Stuiver, M., Reimer, P.J., 1993. Extended 14C database and revised CALIB radiocarbon
calibration program. Radiocarbon 35, 215e230.
Sættem, J., Rise, L., Rokoengen, K., By, T., 1996. Soil investigations, offshore mid
Norway: a case study of glacial influence on geotechnical properties. Glob.
Planet. Change 12 (1), 271e285.
van der Meer, J.J., Kjaer, K.H., Krüger, J., 1999. Subglacial water-escape structures and
till structures, slettj€okull, Iceland. J. Quat. Sci. 14 (3), 191e205.
van der Meer, J.J., Menzies, J., Rose, J., 2003. Subglacial till: the deforming glacier
bed. Quat. Sci. Rev. 22 (15), 1659e1685.
Viana, A., Faugeres, J.-C., Stow, D., 1998. Bottom-current-controlled sand depositsda
review of modern shallow-to deep-water environments. Sediment. Geol. 115
(1), 53e80.
Vorren, T.O., Hald, M., Edvardsemn, Lind-Hansen, O.W., 1983. Glacigenic sediments
and sedimentary environments on continental shelves: general principles with
a case study from the Norwegian shelf. In: Ehlers, J. (Ed.), Glacial Deposits in
North- West Europe. Balkema, Rotterdam, pp. 61e73.
Wellner, J., Lowe, A., Shipp, S., Anderson, J., 2001. Distribution of glacial geomorphic
features on the Antarctic continental shelf and correlation with substrate: implications
for ice behavior. J. Glaciol. 47 (158), 397e411.
J.L. Peters et al. / Quaternary Science Reviews 140 (2016) 101e124124