LETTERS
PUBLISHED ONLINE: 7 APRIL 2013 | DOI: 10.1038/NGEO1782
Millennial-scale changes in atmospheric CO2
levels linked to the Southern Ocean carbon
isotope gradient and dust flux
Martin Ziegler1 *, Paula Diz1† , Ian R. Hall1 * and Rainer Zahn2,3
The rise in atmospheric CO2 concentrations observed at the end
of glacial periods has, at least in part, been attributed to the
upwelling of carbon-rich deep water in the Southern Ocean1,2 .
The magnitude of outgassing of dissolved CO2 , however, is
influenced by the biological fixation of upwelled inorganic
carbon and its transfer back to the deep sea as organic carbon.
The efficiency of this biological pump is controlled by the extent
of nutrient utilization, which can be stimulated by the delivery
of iron by atmospheric dust particles3 . Changes in nutrient
utilization should be reflected in the δ13 C gradient between
intermediate and deep waters. Here we use the δ13 C values of
intermediate- and bottom-dwelling foraminifera to reconstruct
the carbon isotope gradient between thermocline and abyssal
water in the subantarctic zone of the South Atlantic Ocean over
the past 360,000 years. We find millennial-scale oscillations
of the carbon isotope gradient that correspond to changes
in dust flux and atmospheric CO2 concentrations as reported
from Antarctic ice cores4,5 . We interpret this correlation as
a relationship between the efficiency of the biological pump
and fertilization by dust-borne iron. As the correlation is
exponential, we suggest that the sensitivity of the biological
pump to dust-borne iron fertilization may be increased when
the background dust flux is low.
Ocean–atmosphere gas exchange is a prime factor determining
fluctuations in the partial pressure of atmospheric carbon dioxide
(pCO2 atm ; ref. 6). This exchange critically involves two pathways
along the deep overturning circulation in the Southern Ocean7
(Fig. 1). First is a deep southern overturning domain (deep circuit in
Fig. 1), with deep- and bottom-water formation occurring close to
Antarctica, in which abyssal, CO2 -laden waters ascend to the surface
and release CO2 to the atmosphere. Previous studies have shown
that this deep circuit is important in determining the efficiency
of the so-called biological soft-tissue pump, which strips carbon
from the atmosphere and transfers it to the deep ocean, as it
provides a pathway for respired carbon from the deep ocean to
reach the surface7 . Modelling studies8 have indicated that it is
the strength of the deep overturning circulation that has a large
impact on pCO2 atm on glacial–interglacial timescales. Second is the
mid-depth overturning domain (mid-depth circuit in Fig. 1), which
combines the conversion of North Atlantic Deep Water (NADW) to
Circumpolar Deep Water (CDW) and the formation of Antarctic
Intermediate Water (AAIW) and Subantarctic Mode Waters
(SAMW) in the subantarctic zone (SAZ) of the Southern Ocean.
Palaeoceanographic reconstructions demonstrate that the divide
between the two circuits shoaled during past glacial periods and
Southern Component Water, an analogue of the Antarctic Bottom Waters, dominated the ocean below 2,000 m water depths9 .
Moreover, according to both theoretical7 and numerical models10
the deep circuit was isolated from the atmosphere during glacial
conditions (Fig. 1b), and this, in combination with increased sea
ice coverage and a more efficient biological pump, fuelled by
dust-derived iron fertilization, restricted outgassing of CO2 into the
atmosphere, hence promoting higher CO2 storage in the abyss2,7 .
The deep circuit is thought to have become operational again
during deglacial transitions, leading to a rise in pCO2 atm (ref. 7),
a process that was facilitated by reduced iron fertilization and an
inefficient biological productivity that was not capable of pumping
carbon back into the deep10 . The isolation of the glacial deep
water results in an increase in regenerated nutrients in the deep
and reduced preformed nutrients in intermediate water masses7 .
This produces a stronger chemical contrast between southernsourced deep and intermediate waters. Carbon isotopic gradients
from benthic and planktic foraminiferal calcite can be used to
reconstruct the chemical difference between bottom and upper
ocean waters11 . This approach exploits the recognition that the
δ13 C of inorganic carbon dissolved in sea water (δ13 CDIC ) traces
the nutrient cycling associated with the biological pump in the
ocean, because photosynthetic carbon fixation preferentially removes 12 C from the surface-layer DIC increasing surface-ocean
δ13 CDIC . 12 C-enriched carbon is later released to the deeper water
column as exported organic material is oxidized, lowering the
deep-ocean δ13 CDIC . Biological export productivity (that is, the
soft-tissue pump) together with upwelling and mixing of subsurface waters thus sets the magnitude of the vertical δ13 CDIC
gradient (1δ13 CDIC ) between bottom and upper ocean waters. An
increased chemical contrast has been documented during glacial
intervals, highlighting the importance of deep-reaching recoupling
processes in controlling pCO2 atm variability at orbital and millennial
timescales8,9,12 . The increased chemical contrast between the Southern Ocean deep and mid-depth was linked initially to lower CO2
during glacials13 . Subsequently, these changes were extended to the
surface ocean14 , and millennial-scale variations of the mid-depth
chemical stratification11 .
Here, we reconstruct at millennial-scale resolution the intermediate to deep 1δ13 CDIC gradient using co-registered δ13 C signals
recorded in benthic (Cibicides wuellerstorfi) and deep-dwelling
1 School
of Earth and Ocean Sciences, Cardiff University, Cardiff, CF10 3AT, UK, 2 Institució Catalana de Recerca i Estudis Avançats (ICREA), 08010
Barcelona, Spain, 3 Institut de Ciència i Tecnologia Ambientals (ICTA), y Departament de Física, Universitat Autònoma de Barcelona, 08193 Bellaterra,
Spain. † Present address: Department of Xeociencias Mariñas e Ordenación do Territorio, Facultade de Ciencias do Mar, University of Vigo, 36310 Vigo,
Spain. *e-mail: [email protected]; [email protected]
NATURE GEOSCIENCE | VOL 6 | JUNE 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
457
NATURE GEOSCIENCE DOI: 10.1038/NGEO1782
LETTERS
a
CO2
CO2
Buoyancy
Loss Gain
PFZ
0
SAZ
Dust
Sea ice
P
SAMW
Depth (m)
2,000
UCDW
Mid-depth circuit
(biologically productive)
B
3,000
AAIW
NADW
AABW
Dust flux
Upwelling
Deep circuit
(unproductive)
4,000
Modern
Sea ice
LCDW
NADW/CDW
B δ13CC. wuell.
Regenerated
nutrients
5,000
6,000
Uptake/export
1,000
SAMW/AAIW
Gradient Δδ13C
P δ13CG. trunc
Preformed
nutrients
Mid-depth
circuit
S
N
40° S
High dust flux
pCO atm
2
Weak upwelling/
Expanded stratified ocean
sea ice
SAMW/AAIW
13
P δ CG. trunc
NADW/CDW
Less
B δ13CC. wuell.
preformed
nutrients
Sequestration of
regenerated
nutrients
Mid-depth
Max. Δδ13C
circuit
gradient
c
0°
Low dust flux pCO atm
Stadial/
2
H-event Deep re-coupling/
Retreated less-stratified
ocean
sea ice
SAMW/AAIW
P δ13CG. trunc
NADW/CDW
More
B δ13CC. wuell.
preformed
Release of
nutrients
regenerated
nutrients
Mid-depth
Min.
Δδ13C
circuit
gradient
Efficent
uptake/export
Glacial
Efficent
uptake/export
b
20° S
Figure 1 | The South Atlantic overturning circulation. a, Schematic section of the South Atlantic overturning circulation, including nutrient
concentration/δ13 CDIC (grey shading), southern deep overturning circuit (blue arrows) mid-depth overturning circuit (red arrows), major water masses
(SAMW, AAIW, upper CDW (UCDW), lower CDW (LCDW), Antarctic Bottom Water (AABW) and North Atlantic Deep Water (NADW)). PFZ and SAZ
indicate the positions of the polar frontal zone and subantarctic zone. Circle B indicates the position of core MD02-2588 at Agulhas Plateau, north of the
SAZ (41◦ 19.900 S, 25◦ 49.700 E 2,907 m water depth) and circle P indicates where the deep-dwelling planktonic foraminifera G. truncatulinoides records
SAMW properties29 . The inset depicts 1δ13 CDIC as a function of the difference between the regenerated nutrients in the deep water and preformed
nutrients in SAMW. b, Schematic of a glacial situation with high dust flux, efficient nutrient uptake, a stratified Southern Ocean, decreased atmospheric
CO2 and a large 1δ13 CDIC . c, Schematic of a Heinrich-event type situation with decreased ocean stratification, low dust flux, inefficient nutrient uptake,
outgassing of CO2 and a small 1δ13 CDIC .
planktonic foraminifera (Globorotalia truncatulinoides) extracted
from single samples from the CASQ (Calypso square corer)
sediment core MD02-2588 from the southern Agulhas Plateau,
near the Subtropical Front (Fig. 1, 41◦ 19.900 S, 25◦ 49.700 E, 2,907 m
water depth; Fig. 2a–c).
G. truncatulinoides is a non-symbiotic deep-dwelling planktonic
foraminifera that inhabits the thermocline (∼400 m water depth)
and its δ13 C is used here as a recorder of δ13 CDIC of SAMW (see
also Supplementary Information). The structure of this record
resembles planktonic δ13 C profiles that are found in the SAZ
(ref. 15) as well as in the low latitudes at sites influenced by
SAMW ventilation16,17 (Fig. 2c). This comparison confirms that
the signal in our record represents a large-scale water-mass signal.
Outstanding features of all these records are the distinct carbon
isotope minimum events (CIME) during glacial terminations15,16,18 .
The CIMEs reflect the breakdown of surface-water stratification,
renewed upwelling in the Southern Ocean of low-δ13 CDIC and
CO2 -laden aged deep waters to the surface, advection of these waters
458
to the convergence zone at the Subtropical Front by southwardshifting westerlies18 and their subsequent subduction to form lowδ13 CDIC SAMW. The prominence of CIMEs in our record provides
further evidence that the δ13 C record of G. truncatulinoides reliably
traces the advection of SAMW.
The benthic δ13 C record of MD02-2588 shows fluctuations that
closely mimic those observed in the benthic δ13 C stacks from the
mid-depth and deep Atlantic Ocean19 (Fig. 2b). During glacial intervals, low values in the benthic δ13 C result from changing mixing
ratios between the Northern and Southern component waters,
where an increased influence of isotopically low southern-sourced
CDW decreases the δ13 C values. The benthic δ13 C at our core
location therefore reflects ventilation changes in conjunction with
mode switches of the Atlantic meridional overturning circulation
and associated changes of deep nutrient content14 . Other influences
may relate to changing endmember compositions in these water
masses20 , potentially driven by, for example, air–sea exchange
variability (see also Supplementary Information). Processes that
NATURE GEOSCIENCE | VOL 6 | JUNE 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
NATURE GEOSCIENCE DOI: 10.1038/NGEO1782
LETTERS
a
Age (kyr)
0
20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360
I
II
IIIa IIIb
IV
¬380
¬400
3.0
b
δ18Obenthic
3.5
1.5
δ13Cbenthic
1.0
0.5
¬420
4.0
4.5
¬440
5.0
¬460
δDice
VPDB
2.5
0.0
¬0.5
c
1.4
1.0
0.2
0.6
¬0.2
0.2
¬0.6
¬1.0
¬0.2
0.0
Δδ13C
d
SAMW¬CDW
¬1.4
0.5
10
100
δDEPICA
¬380
f
¬400
1,000
¬420
101
10,000
2
3
4
5
6
102
103
h
20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360
Age (kyr)
2
pCO atm
200
180
H1 H2 H3
10
H4
H5 H5a H6
¬0.5
100
0.0
1,000
0.5
10,000
SAMW¬CDW
MD02¬2588 Δδ13C
220
EPICA dust flux (μg m¬2 a¬1)
260
240
104
2
0
pCO atm
300
280
260
240
220
200
180
Alkenones (μg g¬1)
ODP Site 1090
δ15N, E11-7
¬440
g
Dust flux
(μg m¬2 a¬1)
1.0
e
δ13CG. sacc.
δ13CG. trunc.
¬1.0
1.0
0
20
40
60
80
Age (kyr)
100
120
140
Figure 2 | The palaeocenaographic records of MD02-2588. a, Comparison of the benthic δ18 O record (yellow) with the Antarctic European Project for Ice
Coring in Antarctica (EPICA) ice-core δD (grey; construction of the age model is described in the Supplementary Information). b, Benthic δ13 C record of
MD02-2588 (solid yellow line) compared to the deep (lower boundary of yellow shading) and mid-depth (upper boundary of yellow shading) Atlantic
benthic δ13 C stacks19 . c, G. truncatulinoides δ13 C record of MD02-2588 (yellow) compared to a stack of Caribbean δ13 C planktonic records16 (grey).
d, 1δ13 CSAMW−CDW of MD02-2588 (grey, three-point running mean in red). e, EPICA dust record4 (logarithmic scale; blue). f, Alkenone record from ODP
site 1090 (ref. 30). Nitrogen isotope record of SAZ record E11-7 on their original age models23 . g, Atmospheric pCO2 measured in Antarctic ice cores6 .
h, 1δ13 CSAMW−CDW of MD02-2588 (red), EPICA dust4 (blue) and atmospheric pCO2 (ref. 5) comparison for the last glacial cycle. Grey bars indicate
glacial terminations (upper panel) and yellow bars Heinrich events (lower panel).
NATURE GEOSCIENCE | VOL 6 | JUNE 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
459
NATURE GEOSCIENCE DOI: 10.1038/NGEO1782
LETTERS
impact the whole-ocean δ13 C, such as the transfer of relative
low-δ13 C terrestrial organic material to the oceans during glacial
periods, are without consequence for the 1δ13 CDIC gradient as
they impact equally the entire water column. The records of
pCO2 atm and oceanic vertical 1δ13 CDIC are therefore linked in the
Southern Ocean surface, where nutrient utilization, the product of
biological productivity and water-column stratification, determines
the amount of CO2 outgassing to the atmosphere and drives
1δ13 CDIC (ref. 18; Fig. 1).
Our reconstruction of the δ13 CDIC gradient between SAMW and
CDW (1δ13 CSAMW−CDW ), shows that the largest difference coincides
with full glacial conditions indicating a stronger chemical gradient
between thermocline and deep waters, when pCO2 atm was lowest
(Fig. 2f,h). This observation is in line with the idea of an isolated
old deep-water mass in the Southern Ocean2,14 . A distinctive feature
of the 1δ13 CSAMW−CDW record is the persistent millennial-scale
variability (Fig. 2h) that correlates directly with similar timescale
features seen in Antarctic ice-core records. The magnitude of the
1δ13 CSAMW−CDW shifts at the millennial scale is substantial, being
almost equivalent to the observed glacial–interglacial range. This
suggests rapid changes of a large magnitude in the Southern Ocean
carbon cycle, and implies that other processes contribute to the
glacial–interglacial difference in atmospheric CO2 .
Our record reveals insights into the past dynamics of nutrient
utilization in the SAZ. Modern nutrient utilization in the SAZ
is low, as is evident by the high levels of preformed (that is,
unutilized) nutrients in newly formed SAMW and AAIW (ref. 8),
hence causing an inefficient CO2 uptake in the SAZ. A viable
mechanism for increasing nutrient utilization, and consequentially
pCO2 atm drawdown, during glacial periods relates to the increase
in iron supply through enhanced dust flux to the SAZ (refs 3,
6). Moreover, an increased SAZ nutrient drawdown and carbon
transfer to the deep ocean stimulates an alkalinity feedback through
the dissolution of carbonate in the abyss that would further
lower pCO2 atm (ref. 21). In addition, a reduction in the biologically
unutilized (or preformed) nutrient content of SAMW may reduce
the low-latitude carbonate pump and raise whole-ocean alkalinity,
hence also contributing to a further pCO2 atm drawdown22 . In this
context, it is intriguing to note that the 1δ13 CSAMW−CDW shows
a tight exponential correlation with changes in the dust flux at
millennial to orbital timescales. Although we note that such a
strong correlation between the 1δ13 CSAMW−CDW and dust flux
does not in itself demonstrate a mechanistic link between both
parameters, we suggest that dust-induced iron fertilization is a
likely mechanism that pervasively drives SAZ nutrient uptake
variability, which is reflected in the 1δ13 CSAMW−CDW gradient.
Such an explanation is consistent with evidence for increased
nutrient utilization during glacial maxima in the SAZ that was
derived from isotope records of diatom-bound nitrogen23 (Fig. 2e).
Although other as yet undetermined isotope effects may impact
nitrogen isotope ratios, these data suggest a higher consumption
of nitrate, and hence higher nutrient utilization in the SAZ, with
implications for pCO2 atm drawdown10 . Other palaeo-records from
the SAZ also indicate significantly higher biological productivity
during full glacial conditions, coinciding with elevated dust flux
and a pCO2 atm that was lowered by some 50 ppm below interglacial
levels24 . An exponential relationship with dust flux could be related
to an increase in C/Fe elemental ratio in diatom assemblages under
iron-limited conditions25 . Such a high sensitivity of the Southern
Ocean biological pump to the local iron fertilization would explain
the tight coupling between Southern Ocean dust flux and pCO2 atm
variability over a broad range of timescales26 .
Owing to the thermal bipolar see-saw behaviour of the Atlantic
Ocean27 , the Southern Ocean warms during millennial-scale North
Atlantic cold events (that is, Heinrich Events). A warmer Southern
Ocean is accompanied by a retreat of the sea ice cover, potentially
460
leading to an increase in pCO2 atm through increased air–sea
exchange28 . Further outgassing may be a consequence of a poleward
shift in the position of the Southern Hemisphere westerlies and
hence increased upwelling in the Southern Ocean, bringing old,
carbon-rich deep water back to the surface of the Southern Ocean7
(Fig. 1c). As dust flux to the Southern Ocean reduces during these
warming phases of the Southern Hemisphere4 , nutrient utilization
is limited and in consequence facilitates outgassing. As such the iron
fertilization processes may be viewed as a positive feedback process
that is a response to an initial forcing related to changes in the
interhemispheric ocean–atmosphere system. The inferred highly
sensitive response of the Southern Ocean biological pump to iron
fertilization on millennial timescales should be further examined
using climate modelling to fully understand natural variability
in atmospheric CO2 .
Received 30 November 2012; accepted 28 February 2013;
published online 7 April 2013
References
1. Anderson, R. F. et al. Wind-driven upwelling in the southern ocean and the
deglacial rise in atmospheric CO2 . Science 323, 1443–1448 (2009).
2. Skinner, L. C., Fallon, S., Waelbroeck, C., Michel, E. & Barker, S. Ventilation
of the deep southern ocean and deglacial CO2 rise. Science 328,
1147–1151 (2010).
3. Martin, J. H. Glacial–interglacial CO2 change: The iron hypothesis.
Paleoceanography 5, 1–13 (1990).
4. Lambert, F. et al. Dust-climate couplings over the past 800,000 years from the
EPICA Dome C ice core. Nature 452, 616–619 (2008).
5. Ahn, J. & Brook, E. J. Atmospheric CO2 and climate on millennial time scales
during the last glacial period. Science 322, 83–85 (2008).
6. Petit, J. R. et al. Climate and atmospheric history of the past 420,000 years from
the Vostok ice core, Antarctica. Nature 399, 429–436 (1999).
7. Toggweiler, J. R., Russell, J. L. & Carson, S. R. Midlatitude westerlies,
atmospheric CO2 , and climate change during the ice ages. Paleoceanography
21, PA2005 (2006).
8. Marinov, I., Gnanadesikan, A., Toggweiler, J. R. & Sarmiento, J. L. The
Southern Ocean biogeochemical divide. Nature 441, 964–967 (2006).
9. Lynch-Stieglitz, J. et al. Atlantic meridional overturning circulation during the
last glacial maximum. Science 316, 66–69 (2007).
10. Hain, M. P., Sigman, D. M. & Haug, G. H. Carbon dioxide effects of Antarctic
stratification, North Atlantic Intermediate Water formation, and subantarctic
nutrient drawdown during the last ice age: Diagnosis and synthesis in a
geochemical box model. Glob. Biogeochem. Cycle 24, GB4023 (2010).
11. Charles, C. D. et al. Millennial scale evolution of the Southern Ocean chemical
divide. Quat. Sci. Rev. 29, 399–409 (2010).
12. Fischer, H. et al. The role of Southern Ocean processes in orbital and millennial
CO2 variations: A synthesis. Quat. Sci. Rev. 29, 193–205 (2010).
13. Ninnemann, U. S. & Charles, C. D. Changes in the mode of Southern Ocean
circulation over the last glacial cycle revealed by foraminiferal stable isotopic
variability. Earth Planet. Sci. Lett. 201, 383–396 (2002).
14. Hodell, D. A., Venz, K. A., Charles, C. D. & Ninnemann, U. S. Pleistocene
vertical carbon isotope and carbonate gradients in the South Atlantic sector of
the Southern Ocean. Geochem. Geophys. Geosyst. 4, 1004 (2003).
15. Ninnemann, U. S. & Charles, C. D. Regional differences in quaternary
subantarctic nutrient cycling: Link to intermediate and deep water ventilation.
Paleoceanography 12, 560–567 (1997).
16. Curry, W. B. & Crowley, T. J. The δ13C of equatorial Atlantic
surface waters: Implications for ice age pCO2 levels. Paleoceanography 2,
489–517 (1987).
17. Pena, L. D., Cacho, I., Ferretti, P. & Hall, M. A. El Nino Southern Oscillation
like variability during glacial terminations and interlatitudinal teleconnections.
Paleoceanography 23, PA3101 (2008).
18. Spero, H. J. & Lea, D. W. The cause of carbon isotope minimum events on
glacial terminations. Science 296, 522–525 (2002).
19. Lisiecki, L. E., Raymo, M. E. & Curry, W. B. Atlantic overturning responses to
Late Pleistocene climate forcings. Nature 456, 85–88 (2008).
20. Curry, W. B. & Oppo, D. W. Glacial water mass geometry and the distribution
of δ13 C of CO2 in the western Atlantic Ocean. Paleoceanography 20,
PA1017 (2005).
21. Boyle, E. A. Vertical oceanic nutrient fractionation and glacial/interglacial CO2
cycles. Nature 331, 55–56 (1988).
22. Matsumoto, K., Sarmiento, J. L. & Brzezinski, M. A. Silicic acid leakage from
the Southern Ocean: A possible explanation for glacial atmospheric pCO2 .
Glob. Biogeochem. Cycles 16, 1031 (2002).
NATURE GEOSCIENCE | VOL 6 | JUNE 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
NATURE GEOSCIENCE DOI: 10.1038/NGEO1782
23. Robinson, R. S. et al. Diatom-bound 15 N/14 N: New support for enhanced
nutrient consumption in the ice age subantarctic. Paleoceanography 20,
PA3003 (2005).
24. Kohfeld, K. E., Quere, C. L., Harrison, S. P. & Anderson, R. F. Role of marine
biology in glacial–interglacial CO2 cycles. Science 308, 74–78 (2005).
25. De Baar, H. J. W., Gerringa, L. J. A., Laan, P. & Timmermans, K. R. Efficiency of
carbon removal per added iron in ocean iron fertilization. Mar. Ecol. Prog. Ser.
364, 269–282 (2008).
26. Martinez-Garcia, A. et al. Southern Ocean dust-climate coupling over the past
four million years. Nature 476, 312–315 (2011).
27. Barker, S. et al. 800,000 years of abrupt climate variability. Science 334,
347–351 (2011).
28. Stephens, B. B. & Keeling, R. F. The influence of antarctic sea ice on
glacial–interglacial CO2 variations. Nature 404, 171–174 (2000).
29. Mulitza, S. et al. Late Quaternary δ13 C gradients and carbonate
accumulation in the western equatorial Atlantic. Earth Planet. Sci. Lett. 155,
237–249 (1998).
30. Martinez-Garcia, A. et al. Links between iron supply, marine productivity,
sea surface temperature, and CO2 over the last 1.1 Myr. Paleoceanography 24,
PA1207 (2009).
Acknowledgements
We thank the International Marine Past Global Changes Study (IMAGES) and Institut
Polaire Français Paul Emile Victor (IPEV) for making the RV Marion Dufresne available
LETTERS
and for technical support. We are indebted to G. Uenzelmann-Neben for generously
providing geophysical survey data. J. Becker and H. Medley are thanked for technical
assistance. We acknowledge financial support from the UK Natural Environment
Research Council (NERC) and NERC Radiocarbon Laboratory (I.R.H. and P.D.), the
Spanish Ministerio de Educación y Ciencia (MEC) (R.Z. and P.D., postdoctoral research
grant EX-2004–0918), The Seventh Framework Programme PEOPLE Work Programme
Grant 238512 (I.R.H., R.Z. and M.Z., Marie Curie Initial Training Network
‘GATEWAYS’, www.gateways-itn.eu) and the Climate Change Consortium of Wales
(M.Z. and I.R.H.; www.c3wales.org).
Author contributions
I.R.H. and R.Z. collected the core material. P.D. performed preparation for foraminiferal
stable isotope analysis. Stable isotope analyses were carried out in Cardiff University
under the guidance of I.R.H. All authors contributed to data analysis and interpretation.
M.Z. wrote the manuscript with contributions from all other authors.
Additional information
Supplementary information is available in the online version of the paper. Reprints and
permissions information is available online at www.nature.com/reprints.
Correspondence and requests for materials should be addressed to M.Z. or I.R.H.
Competing financial interests
The authors declare no competing financial interests.
NATURE GEOSCIENCE | VOL 6 | JUNE 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
461