Laurentide Ice Sheet Meltwater and the Atlantic Meridional Overturning Circulation During the Last Glacial Cycle: A View From the Gulf of Mexico - Abrupt Climate Change: Mechanisms, Patterns, and Impacts - page 43

laminated sediment, indicating no obvious changes in oxy-
genation associated with the cessation event.
Overall, the rapidity and magnitude of the cessation event
appears consistent with a cessation of meltwater to the GOM
[Broecker et al., 1989] but not a sudden, complete diversion.
It is dif

years [Hughen et al., 1998, 2000; Keigwin, 2004; McMa-
nus et al., 2004; Keigwin and Boyle, 2008].

18 Osw USING PAIRED 18 O

5. ISOLATING δ δ

AND Mg/Ca DATA

cult to differentiate whether the cessation indicates a

“ of LIS meltwater. Because sea

level continued to rise during the Younger Dryas [Fairbanks,
1989; Bard et al., 2010], it seems unlikely that melting
stopped along the entirety of the southern LIS margin unless
significant melting continued elsewhere including Antarctica.
There was at least a local readvance in the Lake Superior
region [Lowell et al., 1999]. Although Lake Agassiz may
have ceased its southward flow during the Marquette read-
vance that corresponds closely with the Younger Dryas,
summer melting along the southern margin of the LIS may
have continued. However, the rapidity of the cessation event
(possibly <200 years) is dif

diversion ” or a “ reduction ” One complication in this analysis is the fact that forami-

niferal δ

18 O is a function of temperature and ice volume, as

18 Osw. Recent studies have attempted to separate
the temperature and

well as δ

18 Osw components of Orca Basin

δ

18 O and Mg/Ca mea-
surements [Flower et al., 2004; Hill et al., 2006; Williams et
al., submitted manuscript, 2011). At least 30 specimens of G.
ruber from the 250 - 355 µm size fraction were gently
crushed and aliquots taken for stable isotopic (~80 µg) and
elemental (~300 µg) analyses. Stable isotope data were gen-
erated on a ThermoFinnigan Delta Plus XL light stable
isotope ratio mass spectrometer (SIRMS) with a Kiel III
automated carbonate preparation device at the College of
Marine Science, University of South Florida. Long-term
external precision for this instrument is 0.09 and 0.06

18 O by paired δ

planktic foraminiferal δ

cult to explain by shutting off
meltwater sources. In combination with recent evidence for
northward flow during the Younger Dryas [Murton et al.,
2010], a diversion of LIS meltwater remains the leading
possibility. Indeed, recent records from the North Atlantic
are interpreted to indicate eastern flow of LIS meltwater
[Thornalley et al., 2010]. Regardless of whether the cessa-
tion event indicates a diversion or a reduction of LIS melt-
water, the timing (11,000 14 C years ago, 12.9 ka) seems to
coincide with the onset of the Younger Dryas, whether based
on comparison to 14 C ages in European lake sequences
[Ammann and Lotter, 1989] or to layer-counted ages in a
German lake [Brauer et al., 2008] and in Greenland ice
[Rasmussen et al., 2006].
On the other hand, independent evidence from the Car-
iaco Basin for reservoir age changes may complicate age
determination of the cessation event. Comparison of the
German tree ring 14 C chronology and the Cariaco Basin
14 C chronology suggest the oceanic reservoir age was re-
duced by 100

‰ for

13 C, respectively, based on >4000 analyses of
NBS-19 run since July 2000.
The elemental splits were cleaned for Mg/Ca analysis
using the Cambridge method that does not include a reduc-
tive cleaning step [Barker et al., 2003]. Samples were dis-
solved in weak HNO 3 to yield calcium concentrations of
~20 ppm to minimize calcium concentration effects. Elemen-
tal ratio data were generated on a Perkin Elmer 4300 dual-
view inductively coupled plasma (ICP)-optical emission
spectrometer. A standard instrument-drift correction tech-
nique [Schrag, 1999] was routinely used. Precision based on
replicates was 0.16 mmol mol 1 (±0.4°C) for EN32PC6
[Flower et al., 2004] and 0.09 mmol mol 1 (±0.3°C) for
MD02-2551 [Hill et al., 2006]. Our new data on MD02-
2550 were generated on an Agilent 7500cx ICP

18 O and

δ δ

mass spec-
trometer (MS), and precision was 0.08 mmol mol 1 (±0.3°C)
[Williams et al., 2010]. Al/Ca and Mn/Ca data showed no
correlation to Mg/Ca, indicating minimal in

-

400 years near the start of the Younger
Dryas event [Kromer et al., 2004; Muscheler et al., 2008;
Reimer et al.,2009;Hua et al., 2009]. A reduction in
reservoir age would increase the age of the cessation event
by a similar amount and increase its lead over the Younger
Dryas (Williams et al., submitted manuscript, 2011). To the
extent that the reservoir age in the Gulf of Mexico was
close to that in Cariaco Basin as it is today [Wagner et al.,
2009], the ages of the cessation event and AMOC reduction
(based on

-

uence of insuf-

ficient clay removal or Mn-Fe overgrowths on our Mg/Ca
values [Williams et al., 2010]. The δ

18 O and Mg/Ca data will
be available at the National Climate Data Center, NOAA at
http://www.ncdc.noaa.gov/paleo/paleo.html.
Salinity has long been a potential problem in the calibra-
tion of foraminiferal Mg/Ca data to temperature. Early cul-
ture experiments on Globigerinoides sacculifera revealed a
substantial increase in Mg/Ca of 110% when salinity was
increased by 10 psu [Nürnberg et al., 1996]. Smaller salinity
changes (<3 psu) produced no change in foraminiferal Mg/Ca.
Other culturing experiments with Globigerinoides bulloides
and Orbulina universa found a 4% increase in Mg/Ca per

14 C in Cariaco Basin) would remain similar.
However, the reduced reservoir age may be unique to
Cariaco Basin [Reimer et al., 2009; Hua et al., 2009]. By
whatever route the meltwater entered the North Atlantic
Ocean (discussed above), it seems that the cessation event
coincided with AMOC reduction within a few hundred

Δ