This is the author’s version of a work that was submitted/accepted for pub-lication in the following source:

Trofimovs, Jessica, Fisher, Jodie K., MacDonald, Heather A., Talling, Pe-ter J., Spaks, R. Stephen J., Hart, Malcolm B., Smart, Christopher W.,Boudon, Georges, Deplus, Christine, Komorowski, Jean-Christophe, LeFriant, Anne, Moreton, Steven G., & Leng, Melanie J. (2009) Evidence forcarbonate platform failure during rapid sea-level rise; ca 14 000 year oldbioclastic flow deposits in the Lesser Antilles. Sedimentology, 57 (3), pp.735-759.

This file was downloaded from: http://eprints.qut.edu.au/52953/

c© Copyright 2009 The Authors.

Notice: Changes introduced as a result of publishing processes such ascopy-editing and formatting may not be reflected in this document. For adefinitive version of this work, please refer to the published source:

http://dx.doi.org/10.1111/j.1365-3091.2009.01117.x

Evidence for carbonate platform failure during rapid sea level rise; ca

14,000 year old bioclastic flow deposits in the Lesser Antilles.

Trofimovs, J.1*Ψ, Fisher, J.K.2, Macdonald, H.A.1, Talling, P.J.1Ψ, Sparks, R.S.J.1,

Hart, M.B.2, Smart, C.2, Boudon, G.3, Deplus, C.3, Komorowski, J-C.3, Le Friant, A.3,

Moreton, S.G.4, Leng, M.J.5

1 Department of Earth Sciences, University of Bristol, UK 2 School of Earth, Ocean and Environmental Sciences, University of Plymouth, UK 3 Institut de Physique du Globe de Paris & CNRS, 4 Place Jussieu, Paris Cedex 05,

France 4 NERC Radiocarbon Facility (Environment), Scottish Enterprise Technology Park,

East Kilbride, UK 5 NERC Isotope Geosciences Laboratory, British Geological Survey, Keyworth,

Nottingham, UK

* Corresponding Author Ψ Current address: National Oceanography Centre, Southampton, UK

Abstract

Bioclastic flow deposits offshore from the Soufrière Hills volcano on Montserrat in

the Lesser Antilles were deposited by the largest volume sediment flows near this

active volcano in the last 26 kyr. Their volume exceeds that of the world’s largest

historic volcanic dome collapse, which occurred on Montserrat in 2003. These flows

were most likely generated by a large submarine slope failure of the carbonate shelf

comprising the south-west flank of Antigua or the east flank of Redonda; adjacent

islands that are not volcanically active. The bioclastic flow deposits are relatively

coarse grained and either ungraded or poorly graded, and were deposited by non-

cohesive debris flow and high density turbidity currents. The bioclastic deposit often

comprises multiple sub-units that cannot be correlated between core sites; some

located just 2 km apart. Multiple sub-units in the bioclastic deposit result from either

flow reflection, stacking of multiple debris flow lobes, and (or) multi-stage collapse

of the initial landslide. This study provides unusually precise constraints on the age of

this mass flow event that occurred at ca14 ka. Few large submarine landslides have

been well dated, but the slope failures that have been dated are commonly associated

with periods of rapid sea-level change.

Keywords: Antigua, debris flow, geohazard, landslide, micropalaeontology,

Montserrat, radiocarbon dating, Redonda, turbidity current.

Introduction

Submarine slope failure has produced some of the world’s largest mass flows

(Masson et al., 2006). Recent work has indicated that large scale failure events are

more frequent than previously thought (Boudon et al., 2007; Coombs et al., 2007;

Oehler et al., 2008). For example, subaerial and submarine mapping has shown that

large landslides are a common occurrence within the Lesser Antilles island arc

(Deplus et al., 2001; Le Friant et al., 2004; Boudon et al., 2007). More than 47 events

have been identified, with 15 events occurring in the last 12,000 years (Boudon et al.,

2007). These landslide events are often very large, and are potentially a significant

hazard. One of these landslides has yet to be monitored closely as it occurs, and their

triggering and emplacement processes are therefore relatively poorly understood.

Careful analyses of past deposits must therefore play a key role in understanding these

slope failures.

Analysed here are submarine flow deposits comprising predominantly bioclastic

carbonate material found offshore from the island of Montserrat in the Lesser

Antilles. The eruption of the Soufrière Hills volcano on Montserrat has been ongoing

since 1995. Discovery of these bioclastic flow deposits was an unexpected result of

sediment coring during a scientific cruise in May 2005, which produced arguably the

most comprehensive set of sediment cores from around any island volcano (Fig. 1).

The predominantly (> 60%) bioclastic flow deposits record the largest volume mass

flows in the last 26 kyr around this active volcanic island. The minimum volume of

these bioclastic deposits (380 x 106 m3) indicates they resulted from slope failure.

This paper documents the extent and internal architecture of these bioclastic flow

deposits. The data includes volume estimates, particle distribution and component and

facies analyses. The first aim is to identify the origin of these bioclastic flow deposits.

Did they originate from the carbonate shelf around the active volcanic island of

Montserrat, or from either of the adjacent islands of Antigua or Redonda, which are

volcanically inactive? The second aim is to reconstruct the emplacement dynamics of

the original mass flow event(s). Did these events occur in multiple stages, and what

are the implications of these emplacement processes for tsunami generation? Finally,

precise age control potentially provides insight into what triggered the original

landslides. Did submarine slope failure coincide with periods of climatic change, or

with volcanic eruptions on Montserrat?

Understanding what triggers these events is important because they can generate

tsunamis whose magnitude is poorly constrained and highly controversial. Are the

largest tsunamis in island arcs generated by failure of volcanic or non-volcanic

slopes? Are the largest tsunamis associated with eruptions, and therefore potentially

easier to predict? Tsunamis can be a significant co-eruptive hazard on volcanic

islands. For example, entry of pyroclastic flows into the sea during a very recent

eruption of the Soufrière Hills volcano on Montserrat generated significant tsunamis

locally on Montserrat and smaller tsunamis on nearby and densely populated Antigua

and Guadeloupe (Herd et al., 2006; Mattioli et al., 2007). Much larger collapse

volumes could generate waves that reach several tens of metres in height, as modelled

for a collapse of Montagne Pelée on Martinique (Le Friant, 2001).

It is important to understand how landslides and associated mass flows are emplaced

as this determines the amplitude of consequent tsunamis. One extreme hypothesis

considers that large collapses occur as single failures with almost instantaneous

movement of the entire mass into the ocean (Ward & Day, 2001). Alternatively,

collapses may occur retrogressively with several closely spaced failures leading to

multiple sequential debris avalanches (Wynn & Masson, 2003; Masson et al., 2006).

Multi-stage failure significantly lowers the hazard of tsunami production (Wynn &

Masson, 2003).

Landslide deposits are relatively chaotic, and submarine deposits are yet to be cored

successfully. It is therefore difficult to determine the submarine emplacement process

and timing of these avalanches by studying the avalanche deposit itself. Deposits of

longer run out sediment flows generated by landslides can provide a simpler record of

debris avalanche emplacement dynamics, and such turbidites can be penetrated more

easily by coring. Previous work has shown how turbidites can be used to understand

landslide emplacement. For instance, large-scale flank collapse on the Canary Islands

and Hawaiian Islands generate distinctive turbidites that comprise multiple fining-

upward sub-units (Garcia & Hull, 1994; Wynn & Masson, 2003), suggesting that flank

collapse occurred in a number of stages separated by days to weeks. The cores

analysed here previously showed how inverse-to-normal grading turbidites closely

record waxing and waning of an initial mass flow; in this case pyroclastic flows

generated by volcanic dome collapse on Montserrat since 1995 (Trofimovs et al.,

2006, 2008).

Terminology

In this contribution, the term ‘landslide’ is used as a generic term for all forms of

slope failure. ‘Debris avalanche’ is used for a flow of cohesionless blocks, where

grain-to-grain interaction is inferred to be the dominant means mechanism of

fragmentation and particle support (Masson et al., 2006).

‘Debris flow’ is typically defined as laminar or very weakly turbulent flow in which

grain-to-grain interactions dominate particle support (Iverson, 1997; Mulder &

Alexander, 2001). If the flow also contains significant colloidal mud then the flow is

often termed ‘cohesive’. This definition is based upon support mechanisms that

cannot be measured directly for ancient flow events. This ensures that terminology

based on the support mechanism is difficult to apply to deposits without considerable

uncertainty. An alternative definition for ‘debris flow’ and ‘turbidity current’ is

adopted that is based on features within the deposit. A ‘debrite’ is defined as a ‘debris

flow’ deposit in which there is no segregation of larger and smaller particles. Large

and small particles are deposited all together (i.e. ‘en-masse’), producing debrites that

are ungraded and can contain chaotically distributed outsize clasts.

The term ‘turbidity current’ is often used to denote a more dilute flow in which fluid

turbulence is the dominant support mechanism (Kneller & Buckee, 2000; Mulder &

Alexander, 2001). An alternative definition is adopted here, in that deposits of

turbidity currents (‘turbidites’) accumulate progressively in a layer-by-layer fashion

such that larger and smaller particles become segregated (Kuenen, 1966; Walker,

1969; Allen, 1971). This layer-by-layer accumulation leads to vertical grading unless

the flow is steady. Cross bedding provides clear evidence of layer-by-layer

aggradation in a turbidite. Outsize clasts tend to be found along distinct horizons in

turbidites. Amy and Talling (2006) and Talling et al. (2007) provide a more detailed

description of this simple two-fold scheme for submarine flow deposit terminology.

Geological Setting

The Lesser Antilles: Montserrat, Antigua and Redonda

The volcanically active island of Montserrat and the volcanically extinct islands of

Antigua and Redonda are situated in the northern section of the Lesser Antilles island

arc (Fig. 1 inset B). The Lesser Antilles arc is the result of westward subduction of

Atlantic oceanic lithosphere beneath the Caribbean Plate with a convergence rate of 2

to 4 cm/year (Bouysse et al., 1990; Zellmer et al., 2003; Grindlay et al., 2005). Arc

volcanism initiated at ca 40 Ma and continues today (Briden et al., 1979; Bouysse et

al., 1990). The Lesser Antilles island arc bifurcates north of Dominica forming two

island chains: the Outer (eastern) Limestone Caribbees and the Inner (western)

Volcanic Caribbees (Nagle et al., 1976; Briden et al., 1979; Bouysse et al., 1990).

South of the bifurcation the two island chains are superimposed.

The Early Tertiary Limestone Caribbees comprise extinct arc volcanoes. Volcanic

activity ceased on these islands during the Miocene (Martin-Kaye, 1969); thereafter

the islands were exposed to deep weathering and erosion. Extensive submarine

carbonate platforms have developed around the extinct volcanic bases forming

shallow submarine shelves that encircle entire islands (Le Friant et al., 2004). The

younger Late Miocene to recent Volcanic Caribbees include 12 active volcanoes.

Antigua is one of the larger islands of the Limestone Caribbees. The geology can be

divided into three broad units; the Basal Volcanic Suite (BVS), Central Plain Group

(CPG) and Antigua Formation (AF). The stratigraphic sequence documents Late

Oligocene volcanism followed by marine transgression (Weiss, 1994). The BVS is

dominantly comprised of andesite and basalt lavas and pyroclastic deposits

(Christman, 1972). The overlying CPG comprises intercalated lagoon and shallow

marine sediments, together with reworked volcanic material and subordinate primary

pyroclastic deposits that have been deposited within the marine environment (Martin-

Kaye, 1969). The uppermost AF dominantly comprises limestone with lenses of

reworked volcaniclastic deposits. A 1 to 5 km wide submarine shelf encircles the

island of Antigua. The shelf has an average depth of 15 to 20 m and is characterised

by reef structures and associated carbonate sediments (Wilkinson & Drummond,

2004).

The island of Montserrat, located within the Volcanic Caribbees, comprises three

volcanic massifs. The oldest of these (Silver Hills) is in the north, with the currently

active volcanic centre in the south (Fig. 1 inset C). Each massif comprises a central

andesite lava dome complex surrounded by talus aprons and pyroclastic deposits,

which are the result of dome failure. The northern-most, Silver Hills massif (ca 2600

to 1200 ka; Harford et al., 2002), has been exposed to long-term terrestrial and coastal

erosion. As a consequence, deep valleys incise the pyroclastic deposits and a wide (ca

5 km), flat, shallow (60 to 100 m depth) submarine shelf has formed around the

northern part of the island (Le Friant et al., 2004). The Centre Hills massif (ca 950 to

550 ka; Harford et al., 2002) has been subjected to less erosion than the Silver Hills

and consequently the adjacent submarine shelf narrows to 1 to 3 km wide. The

youngest Soufrière Hills-South Soufrière Hills complex (ca 170 ka to present;

Harford et al., 2002) has had no significant erosive influence. The submarine shelf in

the southern section of Montserrat is approximately 0.5 km wide (Le Friant et al.,

2004).

Located 56 km south-west of Antigua and 22.5 km north-west of Montserrat,

Redonda is an uninhabited andesitic rock outcrop approximately 1 km2 in area. It has

a well-developed shallow (20 to 100 m) submarine shelf up to 5 km wide (Le Friant et

al., 2004).

Methods

A research voyage of the RRS James Clark Ross (JR123; May 9 through 18, 2005)

used a Vibrocore system (developed by the British Geological Survey) to sample a

maximum of 6 m of unconsolidated marine sediment in water depths of less than

2000 m. Fifty-two sediment cores were recovered offshore from Montserrat, with

sampling focussed off the south-east coast (Fig. 1). Primary objectives of the research

voyage involved the sampling of modern (1995 through 2003) submarine pyroclastic

deposits from the Soufrière Hills volcano. However, some localities recovered thicker

(deeper) core samples intersecting older sedimentary horizons. This study is based on

19 cored localities that intersected biogenic sedimentary sequences stratigraphically

below the more recent Soufrière Hills volcano volcaniclastic material.

The recovered sediment cores were logged at appropriate scales and sampled for grain

size and component analysis. The coarse grained nature of the sediment necessitated

the use of nested sieve sets for grain size analysis. Half phi (φ) sieve intervals were

used between -1.0 φ (2mm) and 4.5 φ (0.045 mm). The cores were sampled at

centimetre-scale intervals and dried at 28°C for 48 hours before being gently

disaggregated and dry sieved. Bulk point counting a minimum of 500 grains for each

sample was used to classify the grain populations.

In order to assess the micropalaeontological component of these flow deposits, core

samples were taken at 5 cm intervals throughout the deposit. These samples were

washed in deionised water and sieved over 63 μm and 150 μm sieves. The >150 μm

size fraction was visually analysed to identify foraminifera species and other biogenic

components. The <63 μm size fraction was collected, filtered and dried in a cool (35

to 40°C) oven before being homogenised with an agate mortar and pestle and sent for

oxygen isotope analysis at the NERC Isotope Geosciences Laboratory, British

Geological Survey, Keyworth using standard methods. Overall analytical

reproducibility for these samples was better than 0.1‰ for δ13C and δ18O. Isotope

values (δ13C, δ18O) are reported as per mil (‰) deviations of the isotopic ratios

(13C/12C, 18O/16O) calculated to the Vienna Pee Dee Belemnite (VPDB) scale using a

within-run laboratory standard calibrated against National Bureau of Standards

isotope standards.

Samples were also taken for Accelerator Mass Spectrometry (AMS) 14C dating. These

samples were taken from hemipelagic sediment accumulations directly above and

below the targeted bioclastic unit. Each sample was disaggregated in deionised water

and washed over a 63 μm sieve. The samples were dried and dry sieved to collect the

>150 μm size fraction. The foraminifera species Globigerinoides ruber were hand

picked to provide a sample weight of 12 to 14 mg. Foraminifera specimens exhibiting

evidence of reworking and (or) diagenesis were excluded from the sample.

Known weights of cleaned, dried monospecific sample material (Globigerinois ruber

tests) were hydrolysed to CO2 using 85% orthophosphoric acid. The gas was

cryogenically isolated and a sub-sample of CO2 was analysed to determine 13C/12C

ratios, which were used to normalise 14C values to -25‰ δ13CVPDB. The remaining

sample CO2 was converted to graphite by iron/zinc reduction (Slota et al., 1987) and 14C activity determined by accelerator mass spectrometry at the Scottish Universities

Environmental Research Centre (SUERC) AMS Laboratory using either a NEC 5 MV

AMS (Xu et al., 2004) or a NEC 250 kV single stage AMS (Freeman et al., 2008).

The results are reported as conventional radiocarbon years BP (relative to AD 1950)

and % modern 14C, both expressed at the ±1σ level for overall analytical confidence.

The dates were calibrated against the Marine04 dataset using CALIB 5.0.2

Radiocarbon Calibration software. The Marine04 dataset calibrates ages between 0

and 26 ka, therefore dates older than this are presented in uncalibrated radiocarbon

years.

Sea floor Morphology and Submarine Stratigraphy

Bathymetric surveys between Redonda, Montserrat and Antigua reveal two areas of

hummocky topography on the ocean floor east of Redonda and south-west of Antigua

(Fig. 2). Le Friant et al. (2004) identified that these two areas, each covering 50 km2,

record relatively recent failures in the Redondan and Antiguan carbonate shelves.

West of Montserrat, and south-east of Redonda, the sea floor exhibits deep canyon

systems sloping away from Montserrat. To the east of Montserrat there is a relatively

flat sea floor that slopes from Antigua towards Montserrat with gradients generally

<2° (Figs 1 and 2).

East and south-east of Montserrat lies the fault-bounded Boulliante-Montserrat

Graben (Fig. 1), which slopes away from Montserrat towards Guadeloupe. The

graben floor has a relatively shallow gradient of ≤1° and exhibits generally flat

topography with isolated hummocks on a scale of ten’s of metres (Trofimovs et al.,

2008). The maximum depth of the graben is ca 1200 m. The eastern margin is bound

by steep topography and two seamounts with peaks around 800 m water depth,

whereas the west margin rises gradually from 1150 m to 900 m water depth with

gradients of 3 to 5°.

The May 2005 research cruise JR123 primarily cored within the Boulliante-

Montserrat Graben. Five cores were also taken from the south-west of Montserrat and

seven cores were recovered from the sea floor north of the graben between Montserrat

and Antigua (Fig. 1). The marine sedimentary sequence within the Boulliante-

Montserrat Graben comprises intercalated volcanic-sourced and biogenic-sourced,

dominantly sandy, mass flow deposits and hemipelagic sediment accumulations (Fig.

3; Trofimovs et al., 2007). The volcaniclastic deposits are the product of subaerial

block-and-ash pyroclastic flows from the Soufrière Hills volcano entering the ocean

(Trofimovs et al., 2006, 2008). The bioclastic deposits are the result of large-scale

collapse of the shallow carbonate platforms surrounding the Lesser Antilles islands.

Other sources of sediment density flows (e.g. river floods or reworking of shelf

sediment by ocean currents) would not produce such voluminous flows, or bioclastic

sediment composition in the case of flood deposits.

Grey-green hemipelagic silty clay deposits, ranging from centimetres to tens of

centimetres in thickness, separate the volcaniclastic and bioclastic flow deposits. The

hemipelagic sediment commonly contains 60 to 70% biogenic components and

evidence of bioturbation. In core, the hemipelagic sediment is visually identified by

the presence of randomly dispersed foraminifera within a muddy matrix. In contrast,

bioclastic flow deposits are mostly composed of well-sorted bioclastic sand with only

subsidiary amounts of mud matrix. Pelagic sedimentation rates have previously been

estimated from a single core to the south-west of Montserrat at 1 to 3 cm/kyr, with an

average of 2.3 cm/kyr (Le Friant et al., 2008). Additionally, Reid et al. (1996), show

variation from 1 to 3 cm/kyr in the backarc region, to 5 to 10 cm/kyr in the magmatic

arc platform, with lower sedimentation rates observed during interglacial periods.

Bioclastic Mass Flow Deposit offshore from the SE Montserrat coast

A voluminous bioclastic-rich mass flow deposit is preserved within the Montserrat

submarine stratigraphy (Unit 5, Fig. 3). This deposit typically contains 55 to 90%

bioclasts, and 10 to 45% volcanic material and is dominantly comprised of massive

sands. This deposit is not observed in all of the 52 collected cores, some of which

only penetrated into overlying flow units.

Stable Isotope Analysis and Geochronology

Stable isotope (δ18O, δ13C) analysis and AMS radiocarbon dating was performed on

selected cores to produce a chronological framework for the recorded stratigraphic

units. The hemipelagic sediment accumulations in cores 35-V, 17-V and 6-V (Fig. 4)

provide a reliable δ18O stratigraphy. Mass flow deposits within these cores show a

compromised stable isotope signature due to the erosion and incorporation of older

sediment as well as vertical mixing within the flow before deposition. Analysis of the

stable isotope records, therefore, avoids the data from within the mass flow deposits.

The δ18O values obtained from the hemipelagic sediments in these three cores

compare well to the published oxygen isotope curves of Imbrie et al. (1984) and Prell

et al. (1986). It is also comparable to δ18O data from core CARMON2 of Le Friant et

al. (2008) taken to the south-west of Montserrat. This indicates a consistent isotope

record is preserved in the marine sediment around Montserrat during the Late

Pleistocene and into the Holocene.

Core 35-V, located north-east of Montserrat (Fig. 1), contains the longest hemipelagic

sediment record. Stable isotope stage boundaries can be identified down to Marine

Isotopic Stage 5 (MIS 5.5), indicating that this core preserves a sediment record back

to ca 120 ka (Martinson et al., 1987). Figure 5 shows an age versus depth plot for

core 35-V, from which the average rates of sediment accumulation were calculated.

The accumulation rate varies within the core between 3.70 and 6.88 cm/kyr, which is

similar to values recorded by Reid et al. (1996) for the magmatic arc platform region.

The core contains a single deposit from the bioclastic mass flow event, the top of

which lies below the boundary between Marine Isotopic StagesMIS 1 and 2 (dated at

12,050 years before present; Martinson et al., 1987). Using the age versus depth plot

the average sedimentation rate was used to determine the age of the top of the flow

deposit. The sedimentation rate above the flow event, (although only determined by a

single 14C date and comparison of the oxygen isotope curve for core 35-V with that of

the published oxygen isotope curves of Imbrie et al. (1984) and Prell et al. (1986)

(Fig. 4),) was used as a morethe most reliable representation of the post-glacial

sedimentation rate at the end of MIS 2, and therefore used to obtain the age. Evidence

of erosion at the base of the flow, together with the use of sedimentation rates nearly

20,000 years prior to the flow event in MIS 3, excluded the use of sedimentation rates

calculated from sediment below the flow deposit. Therefore, assuming a

sedimentation rate of 5.84 cm/kyr, the top of the flow deposit equates to an age of

14,171 years before present (BP). The base of the flow deposit lies within the upper

part of Marine Isotopic StageMIS 3, just above MIS 3.13 (ca 43 Ka; Martinson et al.,

1987). Cores 17-V and 6-V contain less hemipelagic sediment than core 35-V.

However, both cores confirm that the top of the bioclastic mass flow deposit is

situated below the boundary between Marine Isotopic StagesMIS 1 and 2.

Samples for AMS radiocarbon dating were collected from levels that would

complement the stable isotope profiles (Table 1; Fig. 4). Hemipelagic material was

sampled from above and below the targeted flow deposits. Sampling these intervals

provided relative dates for the flow units. Dating the top of the flow deposit provides

the closest estimate for the timing of the end of deposition as it represents the material

sedimented directly after the flow event. Dates at the base of the flow deposits

provide an estimate of how much material (if any) may have been eroded by the

passage of the flow. The radiocarbon dates show that the bioclastic-rich mass flow

deposit is slightly older than 13,000 to 13,500 years BP. This supports the age

estimate obtained from the oxygen-isotope profiles and the age versus depth plot of

core 35-V.

Large age differences are observed between the ages immediately above and below

the flow deposit. This observation is consistent with erosion and incorporation of the

substrate as the flow moved over the sea floor. For example, if the timing of the flow

is placed at 14,000 years BP, the base of the flow deposit in core 35-V has eroded

down to approximately 38,000 years BP. Using the average pre-event sedimentation

rate estimated from the sequence directly below the flow deposit, this suggests around

88 cm of sediment has been removed by erosion. This thickness is calculated by the

age difference (24,000 years) multiplied by the average rate of sedimentation (3.7

cm/kyr). Core 6-V also exhibits significant erosion down to approximately 39,000

years BP, equating to 92 cm of removed sediment. Although the age at the base of the

deposit in 6-V uses an uncalibrated radiocarbon date and therefore is less well

constrained than 35-V, it is still remarkable that the distal reaches of the flow were so

erosive. Less erosion is observed in core 17-V with the base of the flow younger than

24,000 years BP, implying less than 37 cm of erosion at this location.

ca 14 ka Bioclastic Deposit Distribution and Volume

The biogenic flow deposit is, at its thickest cored interval, greater than 3.42 m

adjacent to the eastern Montserrat coast (Vibrocore 16-V). The deposit thins both

towards Antigua in the north and Guadeloupe in the south-east (Fig. 2). The bulk of

the sediment is confined within the Boulliante-Montserrat Graben, with significant

deposit thinning up the topography that defines the graben margins.

Reconstruction of the deposit thickness using the available sediment core data

provides a volume estimate of 380 x 106 m3. Limited core coverage, the failure to

intersect the base of the deposits in the northern part of the Boulliante-Montserrat

Graben, and absence of samples directly adjacent to the south-east coast of Montserrat

where the overlying volcanic strata is thickest make this a minimum estimate.

Sediment Composition

The distribution and abundance of different biogenic and volcanic material within the

bioclastic mass flow deposit varies both horizontally with flow direction, and

vertically within an individual depositional unit. In general, well-preserved planktonic

foraminifera, typical of open ocean environments (e.g., Globigerinoides ruber and

Globigerinoides sacculifer) have highest abundances at the top of the depositional

unit, whilst benthonic foraminifera (predominantly shallow water reworked

specimens such as Peneroplis spp. and Amphistegina spp.), coral debris, shell

fragments and calcareous sand are concentrated at the base of the units. The volcanic

component (10 to 45%) comprises angular to sub-rounded, variably altered,

intermediate composition, lava fragments, together with broken plagioclase,

hornblende and orthopyroxene crystals. This volcanic material is typical of that

derived from the Soufrière Hills volcano on Montserrat (Harford et al., 2002). The

deposits typically exhibit concentrations of volcanic-sourced sand at their base, with

abundances decreasing up-sequence.

Deposit Characteristics: North of the Boulliante-Montserrat Graben

Variations in deposit characteristics are observed from north to south (Fig. 6). On the

relatively flat (<2°) sea floor north of the Boulliante-Montserrat Graben core 35-V

exhibits a single deposit, 60 cm in thickness (Fig. 7). The poorly sorted (1.08 to

1.38σφ), normally graded, fine to coarse sand (1.07 to 2.38 Mφ) deposit exhibits a

scoured bottom contact with underlying hemipelagic sediment. The base of the

deposit dominantly (71%) comprises angular volcanic lithic clasts and broken

hornblende and plagioclase mineral grains. The remainder comprises abraded

foraminiferid tests, angular pieces of shell and coral, and carbonate sand. Centimetre-

scale planar laminations, defined by volcanic mineral and lithic concentrations

alternating with finer grained bioclastic-rich silty horizons, are preserved within the

base of the unit. Overlying this basal region is massive, dominantly coarse bioclastic

sand, which contains less than 2% volcanic clasts. An upper set of millimetre to

centimetre-scale planar stratification is preserved at the top of the deposit and is

defined by alternating bands of coarse grained (0.5 to 0.7 mm) dominantly biogenic

sand, and horizons with an increased abundance of finer grained (0.25 to 0.125

mm) volcanic material (ca 33%). The top of the deposit shows well-developed normal

grading and contains well-preserved planktonic foraminifera.

Deposit Characteristics: Northern Boulliante-Montserrat Graben

The thickness of the deposited material increases markedly in the northern Boulliante-

Montserrat Graben. Coring did not intersect the base of the bioclastic unit within

cores 36-V to 8-V of the north-south transect (Fig. 6); minimum thicknesses are

consistently greater than 1 m. A transect perpendicular to the inferred flow direction

(Transect A, Fig. 8) illustrates the difference in deposit thickness within the graben

(cores 32-V and 33-V) compared with core 17-V which is located 200 m above the

basin floor.

The bioclastic mass flow deposits within the graben comprise multiple sub-units ten’s

to hundred’s of centimetres in thickness, separated by erosive bases, grain size breaks

and abrupt component abundance variations. For example, within core 16-V (Figs 6

and 9) four distinct coarse-grained, massive sand sub-units are separated from one

another by thin (ca 3 cm) deposits of fine to medium sand with increased silt content.

As the bottom of the fourth, basal sub-unit was not intersected, granulometry was

focussed on the three uppermost sub-units (Fig. 9).

Grain size analysis on the bioclastic deposit in core 16-V in general reveals a large

variation in sorting, from moderately well-sorted to poorly-sorted (0.57 to 2.80 σφ).

Grain sizes are mostly in the sand range from -1.0 φ (2 mm) to 4.5 φ (0.045 mm).

Individual sub-horizons show similar characteristics, with normally graded,

moderately sorted tops and weakly reverse graded bases. An overall upward fining

pattern is observed within the deposit, with the lowest sub-units containing larger

grain sizes than the upper sub-unit. The coarse grained bases of each sub-unit exhibit

two grain size populations, with grain size distribution peaks at -1φ and 2-2.5φ. The

finer grained material (2 to 2.5φ; 0.177 to 0.25 mm) is present in high abundances

throughout the deposit. However, the coarser-grained size fraction (0.0 to -1.0φ; 0.5 to

1.0 mm) is predominantly found at the base of the sub-units and mainly comprises

large (1 mm to 5 mm), platy, porous coral fragments. Volcanic material is also

concentrated at the base of each sub-unit and decreases up-sequence within the bulk

of the flow deposit. Both the large, platy coral fragments and the high abundance of

volcanic material are absent in the fine to medium sand deposits that cap and define

the upper boundaries of each sub-unit. A systematic decrease in volcanic-sourced

material is observed from the base to the top of the entire deposit (Fig. 9).

Micropalaeontology varies within the deposit as a whole, and within each individual

sub-unit (Fig. 10). Open water column planktonic foraminifera species such as

Globigerinoides ruber and Globigerinoides sacculifer, together with fine sponge

spicules and pteropods, occur as a minor component throughout the deposit, but

become dominant in the fine grained top. Conversely, large coral fragments and

carbonate sand, together with increased abundances of benthonic foraminifera such as

Amphistegina spp. and Cyclorbiculina spp., dominate the base of the deposit.

Deposit Characteristics: Central Boulliante-Montserrat Graben

Centrally within the graben the deposits are still comprised of sand sized particles, yet

become finer grained on average, with sizes ranging from 0.0φ (1 mm) to 4.0φ

(0.0625 mm) (Fig. 11). Compared to the northern graben cores, the deposits are better

sorted, ranging from moderately well-sorted to moderately-sorted (0.53 to 0.84 σφ).

Multiple depositional units are identified. Some individual sub-units are recognised

by variations in volcanic versus biogenic clast abundances, but exhibit no sharp grain

size breaks (e.g., Core 12-V; Fig. 11). The deposits typically preserve massive bases

and normally graded tops with indistinct planar stratification defined by horizons of

planktonic foraminifera (ca 95%) alternating with the more typical mix of carbonate

sand, shell and coral fragments, and abraded benthonic foraminifera. Volcanic

material is concentrated at the base of the deposits (up to 70%) together with an

increased abundance of benthonic versus planktonic foraminifera.

Transects perpendicular to flow, show the deposit is thickest within the Boulliante-

Montserrat Graben and thin markedly up the margins of the basin (Transect B; Fig.

8). Grain size is similar in all deposits along the transect (0.0φ to 4.5φ; 0.044 to 1.0

mm), with planktonic foraminifera representing the largest grain size fraction

measured. Benthonic foraminifera, coral and shell fragments are dominantly

preserved within the thicker deposits located within the central section of the

Boulliante-Montserrat Graben.

Deposit Characteristics: Southern Boulliante-Montserrat Graben

Only a single depositional unit is associated with the ca 14 ka bioclastic flow event

(Unit 5; Fig. 3) within the southern extent of the Boulliante-Montserrat Graben (cores

7-V to 5-V on the north-south transect; Fig. 6). Here the deposits are characterised by

moderately well-sorted to moderately-sorted (0.60 to 0.86 σφ) silt and sand sized

particles (0.0φ to 4.5φ; mean diameters <3φ). Thicknesses between 5 and 30 cm are

recorded. A massive base is common, overlain by millimetre-scale planar laminations

and rare cross-stratified laminations. The top of the deposit is normally graded and

capped by mud and silt. The volcanic component of the deposits decreases in

comparison with more northern deposits, to ca 15%. A transect perpendicular to the

inferred flow direction (Transect C; Fig. 8) shows the deposit thins and fines up the

graben margins.

Discussion

Source of the ca 14 ka Bioclastic Deposit

Initiation by slope failure: It is inferred that these bioclastic flow events resulted from

slope failure, as they are too voluminous and widespread to have been produced by

river floods or reworking of shelf sediment by ocean currents.

Shallow water: Analysis of the micropalaeontology of bioclastic-rich units provides

information on the source of the flow. The bioclastic flow deposits are dominated by

coarse-grained biogenic material. The presence of bryozoa, coral and molluscs

indicate a shallow water source. However, foraminifera species present in the deposit

indicate a more precise depth of less than 70 m. For example, observed Amphistegina

spp. and Peneroplis spp. both inhabit the photic zone within depth ranges of 0 to 130

m and 0 to 70m respectively (Murray, 2006). Amphistegina spp. are indicative of

lagoons and coral reefs, whereas Peneroplis spp. are typically found in lagoon and

innermost shelf environments. At least part of the original slope failure therefore most

likely occurred in very shallow water. The shallow water fauna could have been

reworked by preceding events into deeper water, but this is less likely given the

abundance of shallow water species present.

Montserrat, Antigua or Redonda? The majority of the islands in the Lesser Antilles

have well developed carbonate platforms that could have provided the source

environment for the components preserved in the bioclastic deposit (Adey & Burke,

1979; Le Friant et al., 2004; Picard et al., 2006). The eastern Limestone Caribbees of

the Lesser Antilles arc preserve the widest carbonate platforms, as they have been

subjected to over 5 Myr of weathering and erosion of their volcanic cores, and the

growth of reef structures (Bouysse et al., 1990). The western Volcanic Caribbees

typically have less well-developed carbonate platforms, particularly around active

volcanoes where eruptions build up steep island flanks (Le Friant et al., 2004).

The distribution of the bioclastic flow deposits, in particular the deposition of the

thickest units within the Boulliante-Montserrat Graben, suggest the flow originated

from the carbonate shelf around either Antigua, Redonda or Montserrat. A possible

source for the bioclastic flow is a large debris avalanche deposit that was identified by

Le Friant et al. (2004) off the south-western shelf of Antigua; ca 8 km north of core

35-V (labelled ‘Source 1’ in Fig. 2). A second, similar source (‘Source 2’ in Fig. 2) is

represented by the debris avalanche east of the Redondan carbonate shelf; ca 22 km

north-west of core 35-V. These deposits contain large mega-blocks that are 100 to

400 m in diameter and 10 to 40 m high, and each cover an area of 50 km2 (Le Friant

et al., 2004). The bioclastic deposits most proximal to these debris avalanches are

relatively thin (60 cm) and fine-grained (up to 2 mm) with a single fining up flow unit

preserved in core 35-V (Fig. 7). At this location the flow is relatively erosive, such

that ca 88 cm of underlying hemipelagic sediment was removed.

It might be expected that the most proximal part of the turbidite would comprise the

coarsest and thickest deposit. However, recent work in other turbidite systems

(Masson et al., 2002; Talling et al., 2007) have shown that some turbidity currents

can leave behind relatively thin and fine grained deposits in proximal locations. Large

volume turbidity currents in the Moroccan Turbidite System deposited little sediment

for over 150 km on slopes of more than 0.05°, and this thin proximal deposit was

relatively fine grained (Talling et al., 2007). Flows in the Moroccan Turbidite system

eroded more deeply into underlying sediment in these proximal locations, with

negligible erosion in distal locations (Talling et al., 2007). In other locations in the

same system, turbidity currents bypassed across steeper slopes with no sediment

deposition (Masson et al., 2002). Comparison of the bioclastic deposits described in

this contribution with the observations from Masson et al. (2002) and Talling et al.

(2007) are consistent with the ca 14 ka bioclastic flow having been generated by the

debris avalanche from the wholly bioclastic shelf offshore from either Antigua or

Redonda. The volcanic components observed within the bioclastic deposits are typical

of those from the submarine pyroclastic flow deposits resulting from lava dome

collapse into the ocean at the Soufrière Hills volcano (Trofimovs et al., 2006, 2008).

Therefore, the volcanic component in the flow deposits was most likely eroded and

entrained as the flow ran out past Montserrat. Without further coring, particularly

north of Montserrat towards Redonda, and dating of the debris avalanche deposits we

are unable to conclusively determine which of the collapse events, offshore from

either Redonda or Antigua, produced the ca 14 ka bioclastic deposit.

The third potential source is from the volcanically active island of Montserrat, where

a small embayment is observed in the north-east part of Montserrat’s coast (labelled

‘Source 3’ in Fig. 2). The 60 to 100 m depth of the carbonate shelf in this area (Le

Friant et al., 2004) could produce the shallow water fauna seen in the bioclastic flow

deposits. However, a high-resolution multibeam bathymetry and shallow seismic

survey offshore from Montserrat in December 2007 (cruise JC18 of the RRS James

Cook; unpublished data) found no evidence for a significant recent landslide in this

area. It is possible that evidence of such a landslide has been buried by subsequent

eruptive events from the Soufrière Hills volcano. However, the large volume and

young age of the bioclastic deposit suggests that the scarp features and proximal

blocky deposits would be easy to recognise. It is therefore less likely that Montserrat

is the source.

Depositional process – High density turbidity current or non-cohesive debris flow

The bioclastic flow deposits often lack sedimentary structures, and are weakly graded

or ungraded. The depositional process for this type of massive sand has been the

subject of vigorous debate (Shanmugam & Moiola, 1995; and the seven replies to

their paper in AAPG Bull. 81, 1997).

The bioclastic flow deposits differ in a number of ways from ‘classic’ turbidites that

comprise a full Bouma sequence (Bouma, 1962). The flow deposits lack an upper

mud division or Bouma E division, and there is a sharp grain size discontinuity

separating hemipelagic mud from the uppermost sandy part of these flow deposits.

The lack of a muddy interval is either because disintegration of the initial slope

failures produced little mud sized sediment, or that the mud bypassed through the

study area to be deposited elsewhere.

Sedimentary structures are rare in the flow deposits, and comprise rather poorly

developed planar laminations. Cross-laminated intervals (Bouma C divisions) are

generally absent. The lack of cross lamination suggests that these flows had relatively

high sediment concentrations, and sediment fallout rates that suppressed bedform

development. However, this lack of cross bedding may be related, at least in part, to

the relatively coarse nature of sediment within these flows. Cross lamination in

turbidites is most typically associated with finer grained intervals.

The flow deposits typically show relatively poor vertical grading, in comparison to

turbidites seen in many other sequences (Nilsen et al., 2008). In some cores, these

flow units show no statistically significant vertical grading for considerable

thicknesses (e.g. over 1 m in core 10-V; Fig. 11). This lack of strong grading suggests

deposition by either rather coarse grained and temporally steady turbidity currents, or

by higher sediment concentration non-cohesive debris flows. It is inferred that

intervals with weak but distinct normal grading were deposited by turbidity currents

with high sediment concentration (e.g. the upper part of the deposit in core 16-V; Fig.

9). Ungraded intervals were most likely deposited by even higher sediment

concentration non-cohesive debris flow (e.g. the deposit in core 10-V; Fig. 11).

Deposition of these ungraded intervals in a layer-by-layer fashion by remarkably

steady turbidity currents would be a less likely occurrence.

Deposit architecture

The deposits within the northern and central part of the Boulliante-Montserrat Graben

comprise multiple sub-units ten’s to hundred’s of centimetres in thickness, separated

by erosive bases, grain size breaks and component abundance variations. These sub-

units are not separated by hemipelagic sediment, suggesting either emplacement in a

series of events spaced relatively closely in time or significant basal erosion removing

all trace of intervening background sediment accumulations.

Variations in component abundance from the base to the top of each sub-unit reflect

particle segregation according to density and grain size in the original flow. For

example, benthonic foraminifera, which generally have thick shell walls, are

predominantly found towards the base of the deposit whereas thin-walled planktonic

foraminifera are observed at the top. Small, dense (2000 kg m-3) volcanic clasts

dominate the base of the deposits. The less dense carbonate sand and larger

foraminifera are proportionally dominant at the top.

High concentration flows have been observed to show vertical stratification with the

basal part behaving as a debris flow, whilst the more dilute upper parts behave as a

turbidity current (Postma et al., 1988). Kinetic sieving along the bottom of the flow

would allow the small, dense volcanic clasts to fall through openings between the

larger particles and therefore accumulate at the base of the deposit (c.f. Gray &

Hutter, 1997; Schwarzkopf et al., 2005). ), whereas tTurbulent suspension of the

slightly more dilute top of the flow would segregate all particles according to density

and grain size.

Origin and significance of multiple bioclastic sub-units

The thickness and character of bioclastic sub-units in the central Boulliante-

Montserrat Graben contrasts with the much simpler architecture of younger flow

deposits, such as those produced by volcanic dome collapse during the 1995 through

2006 eruption of the Soufrière Hills volcano on Montserrat (Trofimovs et al., 2006,

2008). The complex sub-units in the bioclastic deposit cannot be traced between cores

that are in some cases only 2 km apart. What is the origin and significance of this

more complex stacking of multiple sub-units in the bioclastic deposit?

The first hypothesis is that the multiple bioclastic sub-units result from repeated flow

reflection from topography on either side of the Boulliante-Montserrat Graben.

However, the sub-units at a given location do not appear to get thinner and finer

upwards in the core, as might be expected to occur if they were produced by

reflection of a single initial pulse. The second hypothesis is that multiple bioclastic

sub-units are due to complex flow reflection from the hummocky topography

produced by the underlying debris avalanche deposits within the Boulliante-

Montserrat Graben. Some avalanche blocks are several tens of metres high (Le Friant

et al., 2004). Flow reflected from such blocks could result in a complex series of

stacked depositional units. However, no such effects are observed in the younger flow

deposits produced by dome collapse from the Soufrière Hills volcano, which have

encountered the same blocky sea floor topography (Trofimovs et al., 2006, 2008). A

third hypothesis is that the bioclastic sub-units result from complex superposition of

debris flow lobes, which were emplaced during a single event. Debris flows typically

produce deposits with complex stacking of individual lobes (e.g. Iverson, 1997; Major

and Iverson, 1999; Ferrucci et al., 2005). The final hypothesis is that the multiple

bioclastic sub-units record multiple stages of failure in the original landslide.

Multistage failure might be expected to produce a similar pattern of sub-units in

adjacent cores. This hypothesis would not explain why the bioclastic sub-units cannot

be easily correlated between closely spaced cores, unless flow reflection or debris

flow lobe stacking also occurred. If the landslide originated from Antigua or

Redonda, then only one of the multiple stages of failure was recorded in the proximal

core 35-V.

Towards the southern end of the Boulliante-Montserrat Graben the deposit is again

represented as a single depositional unit that thins and fines away from source. Well-

developed planar stratification and rare cross stratification indicate deposition from a

low sediment concentration turbidity current. The observed normally graded top

suggests deposition from a waning flow.

The complex nature of the deposits within the Boulliante-Montserrat Graben makes

unequivocal interpretation of the flow dynamics difficult. However, a combination of

the above processes could plausably have occurred. A potential sequence of events

(Fig. 12) includes multistage failure of the initial landslide to produce the debris

avalanche deposits proximally. Progressive disintegration of the avalanche blocks

with distance from source produced debris flows. In the vicinity of core 35-V the

debris flows were dominantly erosive and all but one flow bypassed the north-east of

Montserrat. These debris flows eroded and incorporated older volcaniclastic turbidites

sourced from the Soufrière Hills volcano. The break in slope at the Boulliante-

Montserrat Graben floor abruptly decreased the flow momentum, causing rapid

deposition. Sub-units within the deposit in this area likely represent multiple flows

from the source landslide, overlap of debris flow lobes and possible reflection around

large blocks on the sea floor and from the basin margins. It is likely that the debris

flows were vertically stratified with a more dilute upper part supporting and

segregating particles in the turbulent flow. Upon further sedimentation and water

incorporation the high sediment concentration, stratified flows evolved into more

“typical” dilute turbidity currents. These turbidity currents remained high energy at

their furthest cored extent, as indicated by the large amount of erosion underneath the

single depositional unit in core 6-V.

Timing and triggers of initial landslide

Sea-level rise: The ca 14 ka date coincides with the rapid sea-level rise associated

with the change from glacial to interglacial conditions following the Last Glacial

Maximum at ca 20 ka. Ice volumes approached maximum values around 30 ka and

remained nearly constant for 11,000 years (Lambeck et al. 2002a). During the Last

Glacial Maximum eustatic sea level was 125 +/- 5 m lower than present (Fleming et

al. 1998). However, as the ice sheets began to decline, sea-level began to rise. Initially

sea-level rose relatively slowly at around 3.3 mm/yr from about 19 to 16 ka. From 16

to 12.5 ka the rate increased to 16.7 mm/yr according to Lambeck et al. (2002a).

Fleming et al. (1998) support this change in rate, although they determined slightly

lower values. Lambeck et al. (2002a) note a gap in their record at ca 14 ka and

suggested a short duration of very rapid rise in sea level, an event also recorded by

Fairbanks (1989) as ‘meltwater pulse 1A’ and Bard et al. (1990) as ‘catastrophic rise

event 1’.

A temporal association large volume landslides and glacial stage terminations has

been recorded in a number of geodynamic settings globally (Quidelleur et al., 2008).

Accurate dates are rare for submarine landslide deposits. However, Quidelleur et al.

(2008) suggest that landslides in the Canary Islands, offshore Hawaii, the Lesser

Antilles and Tahiti, correspond to rapid transitions from high to low δ18O values in

the global record compiled by Lisiecki and Raymo (2005). High δ18O values

correspond to colder global climate conditions when there is abundant continental ice.

Low δ18O values are indicative of warmer global conditions with a lesser volume of

continental ice (Lambeck et al., 2002b). The data presented here suggests that the

bioclastic flow occurred during a glacial to interglacial transition. This transition

released melt water from the ice caps causing a rapid rise in sea level. This sea-level

rise may have promoted slope failure due to increased physical erosion of the

coastline, which increased the sediment load and sedimentation rates on the

submarine shelf.

Earthquake: Masson et al. (2006) suggest that a number of contributing factors may

combine to initiate a submarine collapse. For example, study of the deposits from the

Storegga slide off the west coast of Norway indicated that fine-grained marine

sedimentary layers deposited during interglacial periods acted as planes of weakness

along which the landslide moved preferentially (Kvalstad et al., 2005). In addition to

the presence of these layers, it was hypothesised that an earthquake was ultimately

necessary to trigger the collapse (Bryn et al., 2005; Kvalstad et al., 2005). There are

several examples in the literature of earthquake-triggered landslides (Heezen &

Ewing, 1952; Tappin et al., 2001; Gracia et al. 2003; Fine et al., 2005).

The Lesser Antilles island arc is a known area of seismic hazard, with several large

(M>7) earthquakes occurring historically (Feuillet et al., 2002). Tectonic earthquakes

are associated with the subducting North American Plate, whilst shallow earthquake

activity is associated with large, east-west orientated normal faults. These faults have

formed a series of grabens, such as the Boulliante-Montserrat Graben, that are

characteristic of the northern Lesser Antilles arc (Douglas et al., 2006).

Volcanic eruptions: Magma movement and volcanic activity is commonly cited as a

potential cause of flank collapse on volcanic islands (Moore et al., 1989; Masson et

al., 2002). If the ca 14 ka bioclastic unit originated from Antigua or Redonda, a

volcanic influence can be ruled out for triggering the collapse as all volcanic activity

ceased before the Miocene. If the landslide originated from Montserrat there is the

possibility of volcanic activity induced collapse. However, the flow deposit is bound

above and below exclusively by hemipelagic sediment. There is no evidence for

eruptive activity from the Soufrière Hills volcano around this time preserved in any of

the 80 marine sediment cores recovered during recent cruises. Similarly, studies of the

stratigraphy on land (Harford et al., 2002; Smith et al., 2007; Le Friant et al., 2008)

show volcanic activity at 4 ka and at 24 ka, but no eruption around 14 ka. This implies

that if the bioclastic unit did originate from Montserrat, it did so during a period of

volcanic quiescence from the Soufrière Hills volcano.

Implications for Geohazards within the Lesser Antilles

Coring through the ca 14 ka bioclastic unit revealed that a minimum of 380 x 106 m3

of calcareous material was deposited within, and adjacent to the Boulliante-

Montserrat Graben. This volume of material far exceeds any single volcanic collapse

from the Soufrière Hills volcano. Historically, the July 2003 dome collapse on

Montserrat was the most voluminous for any volcano; 210 x 106 m3 of material

entered the ocean (Herd et al., 2006). The Soufrière Hills volcano is recognised as the

major hazard associated with Montserrat. However, the bioclastic deposit shows that

the largest collapse preserved within the offshore sediment record is not related to an

eruption. The lack of a volcanic trigger makes such a landslide difficult to forecast.

A clear link has been established between submarine landslides and tsunamis

(Bonaccorso et al., 2003; Gracia et al. 2003; Harbitz et al., 2006). Therefore it is

conceivable that a significant tsunami wave was associated with this ca 14 ka event.

The magnitude of a tsunami depends on the volume, initial acceleration, velocity, and

failure dynamics (retrogressive or single collapse) of the source landslide (Harbitz et

al., 2006). Submarine landslides generally have large wave run-up heights close to

source and more limited effects distally (Okal & Synolakis, 2004). However, the close

proximity of islands in the Lesser Antilles favours the effects of even a small-scale

tsunami being felt throughout the island chain. For example, the tsunami associated

with the July 2003 210 Mm3 dome collapse at the Soufrière Hills volcano produced

wave heights of 3 to 4 m on Montserrat and between 0.5 to 1 m on neighbouring

Guadeloupe (Pelinovsky et al., 2004). It is likely Tthat the more voluminous mass

flow at ca 14 ka may have produced a somewhat larger tsunami. More accurate

estimates of the volume for the ca 14 ka event could help to provide more precise

estimates of tsunami magnitude, however, unknown parameters such as knowledge of

the failure dynamics and flow velocities would be necessary to ascertain the true

magnitude of any resulting tsunami. .

Conclusions

Bioclastic flow deposits offshore from the Soufrière Hills volcano on Montserrat in

the Lesser Antilles were deposited by the largest volume sediment flows near this

active volcano in the last 26 kyr. These flows were most likely generated by a large

submarine slope failure of the carbonate shelf comprising the south-western flank of

Antigua or eastern flank of Redonda; adjacent islands that are not volcanically active.

The flows are less likely to have been generated by debris avalanches that formed an

embayment on Montserrat. However, the debris avalanches offshore Antigua,

Redonda and Montserrat need to be cored and dated to determine conclusively the

origin of this bioclastic mass flow.

The bioclastic flow deposits are coarse grained and either ungraded or poorly graded,

and were deposited by non-cohesive debris flow and high density turbidity currents.

The proximal part of the bioclastic deposit (assuming that the event originated from

Antigua or Redonda) comprises a single flow unit. This proximal deposit is underlain

by up to 88 cm of erosion, and is finer grained and thinner than some distal deposits.

The more distal bioclastic deposit often comprises multiple sub-units that cannot be

correlated between sites located only 2 km apart. This architecture is much more

complex than that of submarine deposits from recent volcanic dome collapses that

comprise a single unit. Multiple sub-units in the bioclastic deposit result from either a

greater degree of reflection due to larger flow volume and complex palaeotopography,

stacking of multiple debris flow lobes, and (or) multi-stage collapse of the initial

landslide.

This study provides unusually precise constraints on the age of this mass flow event

that occurred around 14,000 years ago. This event is not associated with any evidence

for a major volcanic eruption on Montserrat, showing that large landslides and

associated tsunamis need not be triggered by magma movement. The bioclastic flow

event occurred at a change from glacial to interglacial conditions and coincides with a

period of rapid sea-level rise.

Acknowledgements

This work was supported by the NERC grant NER/A/S/2002/00963. The authors

would like to thank the British Geological Survey for their technical support and

expertise and the captain, crew and scientific team on the RRS James Clark Ross for

their invaluable assistance. R.S.J.S. acknowledges a Royal Society-Wolfson Merit

Award. M.B.H., C.W.S. and J.K.F. acknowledge the support of a research grant (F/00

568/P) from The Leverhulme Trust. Radiocarbon dating and stable isotope analysis

was supported by NERC (grants RCL/1164.0306, RCL/1229.0407 and IP/992/1107

respectively). D. Masson and R. Urgeles are thanked for their insightful reviews.

References

Adey, W.H. and Burke, R. 1979. Holocene bioherms (algal ridges and bank-barrier

reefs) of the eastern Caribbean. Geological Society of America Bulletin, 87 (1), 95-

109.

Allen, J.R.L. 1971. Instantaneous sediment deposition rates deduced from climbing-

ripple cross-lamination. Journal of the Geological Society London, 127, 553-561.

Amy, L.A. and Talling, P.J. 2006. Anatomy of turbidites and linked debrites based

on long distance (120 x 30 km) bed correlation, Marnoso Arenacea Formation,

Northern Apennines, Italy. Sedimentology, 53 (1), 161-212.

Bard, E., Hamelin, B., Fairbanks, R.G. and Zindler, A. 1990. Calibration of the 14C timescale over the past 30,000 years using mass spectrometric U-Th ages from

Barbados corals. Nature, 345, 405-410.

Bonaccorso, A., Calvari, S., Garfi, G., Lodato, L. and Patane, D. 2003. Dynamics

of the December 2002 flank failure and tsunami at Stromboli volcano inferred by

volcanological and geophysical observations. Geophysical Research Letters, 30 (18),

19-41. doi: 10.1029/2003GL017702

Boudon, G., Le Friant, A., Komorowski, J-C., Deplus, C. and Semet, M.P. 2007.

Volcano flank instability in the Lesser Antilles arc: Diversity of scale, processes, and

temporal recurrence. Journal of Geophysical Research – Solid Earth, 112 (B8),

B08205.

Bouma, A.H. 1962. Sedimentology of some Flysch Deposits: A Graphic Approach to

Facies Interpretation. Elsevier, Amsterdam, p. 168.

Bouysse, P., Westercamp, D. and Andreieff, P. 1990. The Lesser Antilles Island

Arc. In: Moore, J.C. & Mascle, A. (eds) Proceedings of the Ocean Drilling Program,

scientific results, 110, Ocean Drilling Program, College Station, TX, 29–44.

Briden, J.C., Rex, D.C., Faller, A.M. and Tomblin, J.F. 1979. K-Ar Geochronology

and palaeomagnetism of volcanic rocks in the Lesser Antilles Island Arc.

Philosophical Transactions of the Royal Society of London: Series A, 291(1383),

485-528.

Bryn, P., Berg, K., Forsberg, C.F., Solheim, A. and Kvalstad, T.J. 2005.

Explaining the Storegga slide. Marine and Petroleum Geology, 22, 11-19.

doi:10.1016/j.marpetgeo.2004.12.003

Christman, R.A. 1972. Volcanic geology of southwestern Antigua, B.W.I.

Geological Society of America Memoir, 132, 439-448.

Coombs, M.L., White, S.M. and Scholl, D.W. 2007. Massive edifice failure at

Aleutian arc volcanoes. Earth and Planetary Science Letters, 256 (3-4), 403-418.

Deplus, C., Le Friant, A., Boudon, G., Komorowski, J-C., Villemant, B.,

Harford, C., Ségoufin, J. and Cheminée, J-L. 2001. Submarine evidence for large-

scale debris avalanches in the Lesser Antilles Arc. Earth and Planetary Science

Letters, 192, 145-157.

Douglas, J., Bertil, D., Roulle, A., Dominique, P. and Jousset, P. 2006. A

preliminary investigation of strong-motion data from the French Antilles. Journal of

Seismology, 10, 271-299. Doi: 10.1007/s10950-006-9016-0.

Fairbanks, R. G. 1989. A 17,000-year glacio-eustatic sea level record: influence of

glacial melting dates on the Younger Dryas event and deep ocean circulation. Nature,

342, 637-642.

Ferrucci, M., Pertusati, S., Sulpizio, R., Zanchetta, G., Pareschi, M.T. and

Santacroce, R. 2005. Volcaniclastic debris flows at La Fossa Volcano (Vulcano

Island, southern Italy): Insights for erosion behaviour of loose pyroclastic material on

steep slopes. Journal of Volcanology and Geothermal Research, 145 (3-4), 173-191.

Feuillet, N., Manighetti, I., Tapponnier, P. and Jacques, E. 2002. Arc parallel

extension and localization of volcanic complexes in Guadeloupe, Lesser Antilles.

Journal of Geophysical Research, 107 (B12), 2331, doi: 10.1029/2001JB000308.

Fine, I.V., Rabinovich, A.B., Bornhold, B.D., Thomson, R.E. and Kulikov, E.A.

2005. The Grand Banks landslide-generated tsunami of November 18, 1929:

preliminary analysis and numerical modelling. Marine Geology, 215, 45-57.

doi:10.1016/j.margeo.2004.

Fleming, K., Johnston, P., Zwartz, D. Yokoyama, Y., Lambeck., K. and Chappell,

J. 1998. Refining the eustatic sea-level curve since the Last Glacial Maximum using

far- and intermediate-field sites. Earth and Planetary Science Letters, 163, 327-342.

Freeman, S.P.H., Dougans, A., McHargue, L., Wilcken, K.M. and Xu, S. 2008.

Performance of the new single stage accelerator mass spectrometer at the SUERC.

Nuclear Instruments and Methods in Physics Research, B 266, 2225–2228.

Garcia, M.O. and Hull, D.M. 1994. Turbidites from giant Hawaiian landslides:

Results from Ocean Drilling Program Site 842. Geology, 22, 159-162.

Gracia, E., Danobeitia, J., Verges, J. and PARSIFAL Team. 2003. Mapping active

faults offshore Portugal (36N-38N): Implications for seismic hazard assessment along

the southwest Iberian margin. Geology, 31 (1), 83-86.

Gray, J.M.N.T. and Hutter, K. 1997. Pattern formation in granular avalanches.

Continuum Mechanics and Thermodynamics, 9 (6), 341-345.

Grindlay, N.R., Hearne, M. and Mann, P. 2005. High risk of tsunami in the

Northern Caribbean. EOS Transactions, American Geophysical Union, 86 (12), 121-

132.

Harbitz, C.B., Lovholt, F., Pedersen, G. and Masson, D.G. 2006. Mechanisms of

tsunami generation by submarine landslides: a short review. Norwegian Journal of

Geology, 86, 255-264.

Harford, C.L., Pringle, M.S., Sparks, R.S.J. and Young, S.R. 2002. The volcanic

evolution of Montserrat using 40Ar/39Ar geochronology. In: Druitt, T.H., Kokelaar,

B.P. (eds) The eruption of Soufrière Hills volcano, Montserrat, from 1995 to 1999.

Geological Society of London, Memoir, 21, 93-113.

Heezen, B.C. and Ewing, M. 1952. Turbidity currents and submarine slumps, and the

1929 Grand Banks Earthquake. American Journal of Science, 250, 775-793.

Herd, R.A., Edmonds, M. and Bass, V. 2006. Catastrophic lava dome failure at

Soufrière Hills Volcano, Montserrat 12-13 July 2003. Journal of Volcanology and

Geothermal Research, 148 (3-4), 234-252.

Imbrie, J., Hays, J.D., Martinson, D.G., McIntyre, A., Mix, A.C., Morley, J.J.,

Pisias, N.G., Prell, W.L. and Shackleton, N.J. 1984. The orbital theory of

Pleistocene climate: support from a revised chronology of the marine δ18O record. In:

Berger, A.L., Imbrie, J., Hays, J.D., Kukla, G., Saltzmann, B. (eds) Milankovitch and

Climate, Part 1. Reidel, Hingham, MA, 269-305.

Iverson, R.M. 1997. The physics of debris flows. Rev. Geophys., 35, 245-296.

Kneller, B.C. and Buckee, C. 2000. The structure and fluid mechanics of turbidity

currents: a review of some recent studies and their geological implications.

Sedimentology, 47, 62-94.

Kuenen, P.H. 1966. Matrix of turbidites: experimental approach. Sedimentology, 7,

267-297.

Kvalstad, T.J., Andresen, L., Forsberg, C.F., Berg, K., Bryn, P. and Wangen, M.

2005. The Storegga slide: evaluation of triggering sources and slide mechanics.

Marine and Petroleum Geology, 22, 245-256. doi:10.1016/j.marpetgeo.2004.10.019.

Lambeck, K., Yokoyama, Y. and Purcell, T. 2002a. Into and out of the Last Glacial

Maximum: sea-level change during Oxygen Isotope Stage 3 and 2. Quaternary

Science Reviews, 21, 343-360.

Lambeck, K., Esat, T.M. and Potter, E-K. 2002b. Links between climate and sea

levels for the past three million years. Nature, 419, 199-206.

Le Friant, A. 2001. Les déstabilisations de flanc des volcans de l’arc des Petites

Antilles: origines et conséquences. Unpublished PhD thesis, University of Paris,

France.

Le Friant, A., Harford, C.L., Deplus, C., Boudon, G., Sparks, R.S.J., Herd, R.A.

and Komorowski, J-C. 2004. Geomorphological evolution of Montserrat (West

Indies): importance of flank collapse and erosional processes. Journal of the

Geological Society London, 161, 147-160.

Le Friant, A., Lock, E.J., Hart, M.B., Boudon, G., Sparks, R.S.J., Leng, M.,

Smart, C.W., Komorowski, J.C., Deplus, C. and Fisher, J.K. 2008. Late

Pleistocene tephrochronology of marine sediments adjacent to Montserrat, Lesser

Antilles volcanic arc. Journal of the Geological Society of London, 165, 279-289.

Lisiecki, L.E. and Raymo, M.E. 2005. A Pliocene-Pleistocene stack of 57 globally

distributed benthic δ18O records. Paleoceanography, 20, PA1003.

doi:10.1029/2004PA001071

Major, J.J. and Iverson, R.M. 1999. Debris-flow deposition: Effects of pore-fluid

pressure and friction concentrated at flow margins. Geological Society of America

Bulletin, 111 (10), 1424-1434.

Martin-Kaye, P.H.A. 1969. A summary of the geology of the Lesser Antilles. In:

Matson, P. H. (ed), West Indies Island Arc. 263-286.

Martinson, D.G., Pisias, N.G., Hays, J.D., Imbrie, J., Moore, T.C. and

Shackleton, N. 1987. Age dating and the orbital theory of the ice ages: Development

of a high-resolution 0 to 300,000-year chronostratigraphy. Quaternary Research, 27,

1–30.

Masson, D.G., Watts, A.B., Gee, M.R.J., Urgeles, R., Mitchell, N.C., Le Bas, T.P.

and Canals, M. 2002. Slope failures on the flanks of the western Canary Islands.

Earth Science Reviews, 57, 1-35. doi:10.1016/S0012-8252(01)00069-1

Masson, D.G., Harbitz, C.B., Wynn, R.B., Pedersen, G. and Lovholt, F. 2006.

Submarine landslides: processes, triggers and hazard prediction. Philosophical

Transactions of the Royal Society of America, 364, 2009-2039.

doi:10.1098/rsta.2006.1810

Mattioli, G.S., Voight, B., Linde, A.T., Sacks, I.S., Watts, P., Widiwijayanti, C.,

Young, S.R., Hidayat, D., Elsworth, D., Malin, P.E., Shalev, E., Van Boskirk, E.,

Johnston, W., Sparks, R.S.J., Neuberg, J., Bass, V., Dunkley, P., Herd, R., Syers,

T., Williams, P. and Williams, D. 2007. Unique and remarkable dilatometer

measurements of pyroclastic flow-generated tsunamis. Geology, 35, 25-28.

Moore, J.G., Clague, D.A., Holocomp, R.T., Lipman, P.W., Normark, W.R. and

Torresan, M.E. 1989. Prodigious submarine landslides on the Hawaiian Ridge.

Journal of Geophysical Research, 94, 17465-17484.

Mulder, T. and Alexander, J. 2001. The physical character of subaqeous

sedimentary density currents and their deposits. Sedimentology, 48, 269-299.

Murray, J. 2006. Ecology and Applications of Benthic Foraminifera. Cambridge

University Press, Cambridge.

Nagle, F., Stipp, J.J. and Fisher, D.E. 1976. K-Ar geochronology of the Limestone

Caribbees and Martinique, Lesser Antilles, West Indies. Earth and Planetary Science

Letters, 29, 401-412.

Nilsen, T.H., Shew, R.D., Steffins, G.S. and Studlick, J.R.J. (eds) 2008. Atlas of

deep-water outcrops. American Association of Petroleum Geologists, 504 pp.

Oehler, J.F., Lenat, J.F. and Labazuy, P. 2008. Growth and collapse of the Reunion

Island volcanoes. Bulletin of Volcanology, 70 (6), 717-742.

Okal, E.A. and Synolakis, C.E. 2004. Source discriminants for near-field tsunamis.

Geophysics Journal International, 158, 899-912.

Pelinovsky, E., Zahibo, N., Dunkley, P., Edmonds, E., Herd, R., Talipova, T.,

Kozelkov, A. and Nikolkina, I. 2004. Tsunami generated by the volcano eruption on

July 12-13, 2003 at Montserrat, Lesser Antilles. Science of Tsunami Hazards, Vol.

22, Issue 1, 44-58.

Picard, M., Schneider, J.L. and Boudon, G. 2006. Contrasting sedimentary

processes along a convergent margin: the Lesser Antilles arc system. Geo-Marine

Letters, 26 (6), 397-410.

Postma, G., Nemec, W. and Kleinspehn, K.L. 1988. Large floating clasts in

turbidites: a mechanism for their emplacement. Sedimentary Geology, 58, 47-61.

Prell, W.L., Imbrie, J., Martinson, D.G., Morley, J.J., Pisias, N.G., Shackleton,

N.J. and Streeter, H.F. 1986. Graphic correlation of oxygen isotope stratigraphy

application to the Late Quaternary. Palaeoceanography, 1, 137-162.

Quidelleur, X., Hildenbrand, A. and Samper, A. 2008. Casual link between

Quaternary paleoclimatic changes and volcanic island evolution. Geophysical

Research Letters, 35, L02303, doi:10.1029/2007GL031849.

Reid, R.P., Carey, S.N. and Ross, D.R. 1996. Late Quaternary sedimentation in the

Lesser Antilles island arc. Geological Society of America Bulletin, 108 (1), 78-100.

Schwarzkopf, L.M., Schmincke, H.U. and Cronin, S.J. 2005. A conceptual model

for block-and-ash flow basal avalanche transport and deposition, based on deposit

architecture of 1998 and 1994 Merapi flows. Journal of Volcanology and Geothermal

Research, 139 (1-2), 117-134.

Shanmugam, G. and Moiola, R.J. 1995. Reinterpretation of depositional processes

in a classic Flysch sequence (Pennsylvanian Jackfork Group), Ouachita Mountains,

Arkansas and Oklahoma. AAPG Bulletin – American Association of Petroleum

Geologists, 79 (5), 672-695.

Slota, P.J., Jull, A.J.T., Linick, T.W. and Toolin, L.J. 1987. Preparation of small

samples for C-14 accelerator targets by catalytic reduction of CO. Radiocarbon, 29,

303–306.

Smith, A.L., Roobol, M.J., Schellekens, J.H. and Mattioli, G.S. 2007. Prehistoric

stratigraphy of the Soufrière Hills-South Soufrière Hills volcanic complex,

Montserrat, West Indies. Journal of Geology, 115, 115-127.

Talling, P.J., Wynn, R.B., Masson, D.G., Frenz, M., Cronin, B.T., Schiebel, R.,

Akhmetzhanov, A.M., Dallmeier-Tiessen, S., Benetti, S., Weaver, P.P.E.,

Georgiopoulou, A., Zuhlsdorff, C. and Amy, L.A. 2007. Onset of submarine debris

flow deposition far from original giant landslide. Nature, 450 (22), 541-544.

doi:10.1038/nature06313

Tappin, D.R., Watts, P., McMurty, G.M., Lafoy, Y. and Matsumoto, T. 2001. The

Sissano, Papua New Guinea tsunami of July, 1998 – offshore evidence on the source

mechanism. Marine Geology, 175, 1-24. doi:10.1016/S0025-3227(01)00131-1

Trofimovs, J., Amy, L., Boudon, G., Deplus, C., Doyle, E., Fournier, N., Hart,

M.B., Komorowski, J.C., Le Friant, A., Lock, E.J., Pudsey, C., Ryan, G., Sparks,

R.S.J. and Talling, P.J. 2006. Submarine pyroclastic deposits formed at the Soufrière

Hills Volcano, Montserrat (1995-2003): what happens when pyroclastic flows enter

the ocean? Geology, 34 (7), 549-552. doi: 10.1130/G22424.1

Trofimovs, J., Sparks, R.S.J. and Talling, P.J. 2007. Transport and deposition of

submarine pyroclastic flows; Soufrière Hills volcano, Montserrat, West Indies.

Abstracts, IUGG XXIV General Assembly, July 2-13, Perugia, Italy, p. 6637.

Trofimovs J., Sparks, R.S.J. and Talling, P.J. 2008. Anatomy of a submarine

pyroclastic flow and associated turbidity current: July 2003 dome collapse, Soufrière

Hills volcano, Montserrat, West Indies. Sedimentology, 55, 617-634.

Walker, R.G. 1969. Geometrical analysis of ripple-drift cross-lamination. Canadian

Journal of Earth Science, 6, 683-691.

Ward, S.N. and Day, S.J. 2001. Cumbre Vieja volcano – potential collapse and

tsunami at La Palma, Canary Islands. Geophysical Research Letters, 28, 3397-3400.

doi:10.1029/2001GL013110

Weiss, M.P. 1994. Oligocene limestones of Antigua, West Indies: Neptune succeeds

Vulcan. Caribbean Journal of Science, 30 (1-2), 1-29.

Wilkinson, B.H. and Drummond, C.N. 2004. Facies mosaics across the Persian Gulf

and around Antigua – Stochastic and deterministic products of shallow-water

sediment accumulation. Journal of Sedimentary Research, 74 (4), 513-526.

Wynn, R.B. and Masson, D.G. 2003. Canary Island landslides and tsunami

generation: can we use turbidite deposits to interpret landslide processes. In: Locat, J.,

Mienert, J. (eds) Submarine mass movements and their consequences. Dordrecht

Netherlands: Kluwer Academic Publishers, 325-332.

Xu, S., Anderson, R., Bryant, C., Cook, G.T., Dougans, A., Freeman, S.,

Naysmith, P., Schnabel, C. and Scott, E.M. 2004. Capabilities of the new SUERC

5MV AMS facility for 14C dating. Radiocarbon, 46, 59–64.

Zellmer, G.F., Hawkesworth, C.J., Sparks, R.S.J., Thomas, L.E., Harford, C.L.,

Brewer, T.S. and Loughlin, S.C. 2003. Geochemical evolution of the Soufrière Hills

Volcano, Montserrat, Lesser Antilles Volcanic Arc. Journal of Petrology, 44 (8),

1349-1374.

Table 1. Radiocarbon dates from the submarine stratigraphy off the east coast of

Montserrat

Figure 1. A) Seafloor map around Montserrat showing bathymetry, core locations and

the position of the Boulliante-Montserrat Graben. B) Regional map of the Lesser

Antilles island arc showing the area depicted in 1A. C) The four volcanic centres

identified on Montserrat with ages from Harford et al., 2002.

Figure 2. Bathymetric map adapted from Le Friant et al. (2004), showing the

hummocky sea floor topography that defines the extent of the debris avalanche

deposits from carbonate shelf collapses south-west of Antigua (Source 1) and east of

Redonda (Source 2). Source 3 shows the embayment off the north-east of Montserrat

carbonate shelf. Isopachs on the map show the thickness distribution of the bioclastic

mass flow deposits. Isopach contours are marked in centimetres. Individual core

thickness measurements are given in centimetres.

Figure 3. Correlative stratigraphic logs showing the submarine stratigraphic sequence

within the Boulliante-Montserrat Graben. The numbers shown against core 6-V are

dates in ‘years before present’ from the radiocarbon analyses shown in Table 1. Inset

map shows the location of the stratigraphic transect. Detailed unit descriptions are as

yet unpublished data from current studies by the authors. The yellow ‘Unit 5’

represents the bioclastic flow deposit.

Figure 4. Oxygen isotope (δ18O) stratigraphy measured for cores 35-V, 17-V and 6-V

(see Fig. 1 for locations). Each core is presented with the Marine Isotope Stages (1-5)

labelled on the right hand side of each profile and sub-stages are shown against the

profile. Radiocarbon dates (Table 1) are shown in ‘years before present’ with 2σ

standard deviation. Arrows marked against the logs show the dated sample’s location.

Samples taken from within flow units are shaded grey. Analysis of the stable isotope

record in the text avoids this data. Stratigraphic legend as in Figure 3, except where

marked for the oldest three units of core 6-V. As a comparison, the complete oxygen

isotope stratigraphy down to MIS 7, constructed using data from Imbrie et al. (1984),

Prell et al. (1986) and Martinson et al. (1987), is presented on the left hand side of the

figure.

Figure 5. Age versus depth plot from core 35-V, showing the variation in rate of

sedimentation observed within the core. Age estimates are from the Marine Isotope

Stages outlined in Martinson et al. (1987). Sedimentation rates are shown as average

rates determined through the length of the core. Sedimentation was thought to be

occurring at a rate close to 5.84 cm/kyr at the time the bioclastic flow was deposited.

Extrapolating this rate to the top of the bioclastic deposit gives a date of 14,171 years

BP for flow deposition. Further explanation of this calculation is provided in the text.

Figure 6. Detailed stratigraphic log transect along the inferred flow axis of the

bioclastic deposit. Inset map traces the logged profile down the main flow axis.

Figure 7. Grain size and component analysis of the bioclastic mass flow deposit in

Vibrocore JR123-35-V. Legend as in Figure 6.

Figure 8. Correlative stratigraphic logs for three transects located perpendicular to the

main flow axis. Deposits thin and fine upslope towards the margins of the Boulliante-

Montserrat Graben. Inset map shows the location of the transects.

Figure 9. Grain size and component analysis of the three uppermost bioclastic mass

flow sub-horizons in Vibrocore JR123-16-V. Legend as in Figure 6.

Figure 10. Detailed component and microfossil analysis for the three uppermost

bioclastic mass flow sub-horizons in Vibrocore JR123-16-V. Scanning Electron

Microscope images show a selection of the observed biota; the core depth the sample

was taken from is given in metres. a) Amphistegina lessonii 1.575 m; b) Archaias

angulatus 2.715 m; c) Peneroplis proteus 1.375 m; d) Globorotalia menardii 3.275

m; e) Fragment of coral 3.275 m. The scale bar in all photographs is 200 µm.

Figure 11. Grain size and component analysis of the bioclastic mass flow deposits in

Vibrocores JR123-12-V and JR123-10-V. Only the sub-units where the top and base

of the deposit was intersected have been used for analysis. Legend as in Figure 6.

Core photo A shows the top contact between the uppermost sub-unit of the bioclastic

deposit and overlying hemipelagic sediment. Core photo B shows one of the

bioclastic sub-unit boundaries, which are defined by distinct changes in the

abundance of volcanic clasts, yet no sharp grain size breaks. The sediment in the

upper (darker) half of the photo contains ca 75% volcanic clasts (25% carbonate

clasts), whereas the sediment in the lower (lighter) half of the photo has ca 45%

volcanic clasts (55% carbonate clasts). Core photo C shows the sharp basal contact of

the bioclastic sand deposit with underlying hemipelagic sediment.

Figure 12. Schematic reconstruction of the processes involved in the submarine

landslide that produced the ca 14 ka bioclastic deposits.

Table 1.

Publication Code Core

Number

Depth (cm)

below sea

floor

Radiocarbon

Age (Years

BP +/- 1σ)

δ13CVPDB‰

+/- 0.1

Calibrated Ages

1σ 2σ

SUERC-12986 JR123-6-V 17.5 2051 +/-35 0.9 1243.5 +/-87.5 1231 +/-189

SUERC-12987 JR123-6-V 29.0 2912 +/-35 0.8 2219.5 +/-102.5 2212 +/-228

SUERC-12988 JR123-6-V 46.5 3762 +/-35 1.1 3267 +/-108 3243.5 +/-227.5

SUERC-18388 JR123-6-V 93.5 4502 +/-38 1.0 4214 +/-128 4193.5 +/-246.5

SUERC-18389 JR123-6-V 98.5 8888 +/-39 1.0 9151 +/-119 9185 +/-217

SUERC-18392 JR123-6-V 138.5 11830 +/-41 0.9 12989 +/-76 13000 +/-140

SUERC-18393 JR123-6-V 158.5 39354 +/-659 0.9 Too old for Marine04 calibration curve

SUERC-18378 JR123-17-V 72.0 12377 +/-41 0.4 13462 +/-111 13479 +/-198

SUERC-18379 JR123-17-V 137.0 20792 +/-78 1.1 23987.5 +/-160.5 24005 +/-327

SUERC-18383 JR123-35-V 60.0 9762 +/-38 0.9 10265.5 +/-100.5 10280.5 +/-214.5

SUERC-18377 JR123-35-V 148.5 38447 +/-588 1.1 Too old for Marine04 calibration curve