Synchronizing a sea-level jump, ﬁnal Lake Agassiz drainage, and abrupt cooling
8200 years ago
⁎, Torbjörn E. Törnqvist
, Johanna M. Nevitt
, Barry Kohl
Department of Earth and Environmental Sciences, Tulane University, New Orleans, LA 70118-5698, USA
Tulane/Xavier Center for Bioenvironmental Research, Tulane University, New Orleans, LA 70118-5698, USA
Accepted 18 May 2011
Available online 12 June 2011
Editor: P. DeMenocal
8.2 ka event
abrupt climate change
Freshwater pulses draining into the North Atlantic Ocean are commonly hypothesized to have perturbed the
Atlantic meridional overturning circulation (MOC), triggering abrupt climate changes such as Heinrich events,
the Younger Dryas, and the 8.2 ka event. However, dating uncertainties have prevented causal links between
freshwater pulses and climate events from being ﬁrmly established. Here we report a high-resolution relative
sea-level record from the Mississippi Delta that documents a sea-level jump that occurred within the 8.18 to
8.31 ka (2σ) time window and is attributed to the ﬁnal drainage of proglacial Lake Agassiz–Ojibway (LAO).
This age is indistinguishable from the onset of the 8.2 ka climate event, consistent with a nearly immediate
ocean–atmosphere response to the freshwater perturbation. This constitutes a rare currently available
example of a major abrupt climate cooling that can be directly linked to a well-documented freshwater source
with a temporal resolution on the order of a century. The total inferred eustatic sea-level rise associated with
the very ﬁnal stage of LAO drainage at 8.2 ka ranges from 0.8 to 2.2 m, considerably higher than previous
estimates. These new constraints on the timing and amount of ﬁnal LAO drainage permit signiﬁcantly
improved quantitative analysis of the sensitivity of MOC to freshwater perturbation, a crucial step toward
understanding abrupt climate change.
© 2011 Elsevier B.V. All rights reserved.
Abrupt climate change has received extensive interest for a wide
range of reasons, including its potential role in a future warming
world (Alley et al., 2003). Over the past few decades, the connection
between freshwater forcing and abrupt climate change due to
perturbation of the Atlantic meridional overturning circulation
(MOC) has enjoyed widespread popularity, since it offers a potential
mechanism to explain phenomena such as Heinrich events (Heinrich,
1988), the Younger Dryas (Broecker et al., 1989), and the 8.2 ka event
(Barber et al., 1999). However, the past few years have seen this
hypothesis becoming increasingly challenged (e.g., Broecker et al.,
2010; Fisher et al., 2008; Lowell et al., 2009), in part reﬂecting the fact
that very few abrupt climate events have been unequivocally linked to
a well-mapped and well-dated freshwater source (cf. Clement and
The 8.2 ka cold event is the most prominent abrupt North Atlantic
climate change of the Holocene and is increasingly recognized in
many other parts of the world (Alley and Ágústsdóttir, 2005; Cheng
et al., 2009). This event is often believed to have resulted from the
ﬁnal outburst of proglacial Lake Agassiz–Ojibway (LAO) when an ice
dam over Hudson Bay collapsed (Barber et al., 1999; Lajeunesse and
St-Onge, 2008) and the rapid drainage ﬂooded the North Atlantic
Ocean with freshwater and perturbed the Atlantic MOC (Ellison et al.,
2006; Kleiven et al., 2008), leading to widespread cooling. In addition,
the rerouting of western Canadian Plains runoff following the collapse
of the ice dam over Hudson Bay may have contributed to the 8.2 ka
climate event (Carlson et al., 2009). Despite the popularity of a causal
link between the ﬁnal LAO drainage and the 8.2 ka climate event, this
relationship has yet to be ﬁrmly demonstrated because the cata-
strophic LAO drainage remains poorly constrained in terms of its
timing and amount. The only available direct dating of the ﬁnal LAO
drainage yields an age range of 8.16 to 8.74 ka at the 1σlevel (Barber
et al., 1999). This large age uncertainty precludes an unequivocal
connection between LAO drainage and the 8.2 ka event and allows for
alternative hypotheses such as a role for solar forcing around this time
interval (Muscheler et al., 2004; Rohling and Pälike, 2005). Also, the
amount of LAO drainage is not well known as reﬂected by highly
variable estimates (e.g., Barber et al., 1999; Hijma and Cohen, 2010;
Leverington et al., 2002; Törnqvist et al., 2004a), inhibiting our
understanding of the sensitivity of MOC to freshwater perturbation.
This study seeks to reﬁne previous work (Törnqvist et al., 2004a)
that provided the ﬁrst evidence for a sea-level jump around 8.2 ka
based on stratigraphic data from the Mississippi Delta, Louisiana, USA.
Earth and Planetary Science Letters 315–316 (2012) 41–50
⁎Corresponding author at: School of Earth Sciences and Engineering, Nanjing
University, Nanjing 210093, China.
E-mail address: email@example.com (Y.-X. Li).
Present address: School of Earth Sciences, Stanford University, Stanford, CA 94305,
0012-821X/$ –see front matter © 2011 Elsevier B.V. All rights reserved.
Contents lists available at ScienceDirect
Earth and Planetary Science Letters
journal homepage: www.elsevier.com/locate/epsl
We present a high-resolution relative sea-level (RSL) record around
this time interval using basal peat to track sea-level change. The
rationale of this approach is that rising seas drown the coastal
landscape and transform it into a peat-forming wetland that
accumulates over a consolidated, compaction-free Pleistocene base-
ment. Therefore, intertidal basal peats can be used to determine past
sea levels with high accuracy via precise measurements of their age
and elevation. The robustness of this approach has been demonstrated
in a variety of coastal settings (e.g., Donnelly et al., 2004; Jelgersma,
2. Study area
Coastal plains worldwide (e.g., the US Atlantic Coast) rarely capture
the age/depth range necessary to sample early Holocene sea-level
records that are more likely found in large, prograding deltas. However,
not all deltas contain basal peat and even fewer also occur in microtidal
settings which are particularly favorable for high-resolution sea-level
studies. Our sampling sites are located in the Bayou Sale area in the
western part of the Mississippi Delta (Fig. 1). The US Gulf Coast is
characterized by a microtidal regime with a present-day spring tidal
range typically b0.5 m in coastal Louisiana. In addition, the study area
has been tectonically relatively stable during the Holocene (Törnqvist et
al., 2006). Glacial isostatic adjustment (GIA) contributes signiﬁcantly to
RSL rise in this area around 8.2 ka (Kendall et al., 2008), but the GIA
component would be negligible during a short-lived sea-level jump.
This overall combination of circumstances makes our study area
exceptionally well suited to resolve dm-scale RSL changes for this
The Pleistocene basement in the study area consists of the
pervasively oxidized Prairie Complex (Autin et al., 1991) that is
capped by a few meters of Peoria Loess. Both units are highly
consolidated and essentially compaction-free due to prolonged
subaerial exposure. Overlying the Peoria Loess is an immature
paleosol consisting of an A-horizon enriched in highly decomposed
organic matter. This paleosol was classiﬁed as an Entisol, suborder
Aquent, by Törnqvist et al. (2004b) and is the result of transgression, a
rising groundwater table, and the initial transformation of the
landscape into a wetland environment. The continued rise of the
groundwater table eventually enabled the formation of basal peat. The
distinction between the paleosol and the basal peat is based on
(1) the dark gray matrix color for the paleosol vs. gray brown for the
basal peat; (2) the lesser degree of organic matter decomposition in
the peat as reﬂected by abundant herbaceous plant ﬁbers; and (3) the
massive structure of the paleosol compared to the faintly laminated
peat. Nevertheless, basal peat can have a signiﬁcant mud content and
occasionally contains distinct mud beds.
We collected cores with a Geoprobe system (model 6610 DT).
The early stage of coring aimed at mapping the stratigraphy along a
~6-km-long transect (Fig. 1), exhibiting a transgressive surface
associated with the Pleistocene–Holocene transition. Subsequent
efforts were focused on coring at key locations for detailed sampling
to improve the precision of depth measurements of this transgressive
Cores were initially described in the ﬁeld and then transported to
Tulane University for cold storage (~4 °C). In the laboratory,
representative cores containing basal peat were sampled for radio-
carbon dating, carbon isotope measurements, and foraminiferal
analysis to determine the chronology of basal peat and to constrain
depositional environments of both basal peat and adjacent strata.
Radiocarbon dating of terrestrial plant remains from basal peat was
performed by accelerator mass spectrometry (AMS) at the University
of California, Irvine. Stable carbon isotope and foraminiferal analyses
of two representative cores (sites Bayou Sale VI and IV) were
performed to characterize depositional environments. For δ
analysis, samples were ﬁrst dried at 60 °C for 24 h and acidiﬁed
with 10% HCl to remove carbonates. The residues were centrifuged
and the isolated organic material was then dried overnight at 60 °C.
C measurements were carried out at the Stable Isotope Laboratory
at the University of Miami. For the foraminiferal analysis, samples
were soaked in water for 24 h, wet sieved, and the N63 μm fraction
was examined under a microscope. Identiﬁcation of agglutinated
foraminifera was based mainly on pseudo-chitinous linings because
complete outer tests were often lacking due to poor preservation.
Optical surveys with an infrared TOPCON GTS-4B total station were
conducted between core sites and National Geodetic Survey (NGS)
benchmark T168 (UTM-coordinates: N= 3281.840; E= 645.980)
(Fig. 1) to determine the land surface elevation at the core sites. In
addition, temporary benchmarks were established between the NGS
benchmark and core sites. The temporary benchmarks (not shown in
Fig. 1) were located very close (typically b100 m) to the core sites. At
least two round-trip surveys were carried out between the NGS
benchmark and temporary benchmarks, and typically two round-trip
surveys were conducted between a temporary benchmark and a core
site. The cumulative error for a round-trip elevation survey between
the NGS benchmark and core sites is within 0.05 m.
We drilled 37 sites along the ~6-km-long transect to map the
stratigraphy in the Bayou Sale area (Fig. 1); key stratigraphic
information for all core sites is summarized in Table 1. Multiple
cores that capture the Pleistocene–Holocene transition were drilled at
the majority of the sites.
The transgressive succession at the stratigraphically deeper sites
(V, 32, VII, and VI) is characterized by a basal-peat bed that caps the
dark gray paleosol described above and is abruptly overlain by pale-
gray, shell-bearing muds (Figs. 2, 3). The basal-peat bed at the deepest
sites (V, 32 and VII) shows highly variable characteristics and
thicknesses among multiple cores at each site and is often absent in
this deeper portion of the record due to erosion (Table 1). This is
Bayou Sale Levee
Fig. 1. Map of the Bayou Sale study area and core sites. Filled circles indicate the sites
shown in Fig. 2 that contain basal peat. Asterisks represent sites with a sharp transition
from paleosol to lagoonal mud with no basal peat. Open circles denote the other core
sites listed in Table 1. The open diamond indicates the location of the NGS benchmark
(T168, elevation of +1.83 m with respect to North American Vertical Datum 88).
42 Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50
illustrated by site V where multiple cores (Fig. 3), all drilled within an
area of a few square meters, include a well preserved basal-peat bed
(core H) as well as clear signs of erosion and redeposition as
witnessed by mud clasts (core A), a completely reworked stratigraphy
(core F) or complete erosion of the basal peat (core D) and possibly
even the uppermost paleosol (cores D and F). This provides
compelling evidence that the initially formed basal peat was highly
susceptible to erosion associated with the rapid transgression.
The transgressive succession at the stratigraphically shallowest
sites (IV and 3) is characterized by a basal-peat bed that gradually
gives way to overlying brown-gray muds (Figs. 2, 3), comparable to
what was observed at all core sites with a Pleistocene–Holocene
transition b14 m below present sea level (Table 1). Unlike the deeper
sites discussed above, shells or shell fragments are completely absent
in the brown-gray mud that overlies the basal peat at the
stratigraphically shallower sites (IV and higher) (Figs. 2, 3). Instead,
the brown-gray mud is characterized by faintly laminated organic-
rich beds with plant matter reminiscent of the constituents of the
underlying basal peat.
The transgressive surface at intermediate depth intervals is
recorded at sites 25 and 29 and exhibits a unique signature not seen
in the remainder of the record (Table 1). At these sites, shell-rich
muds immediately onlap the underlying paleosol with no basal peat
between the paleosol and the muds (Figs. 2, 3). While the transition is
sharp, these deposits do not contain reworked organic matter, and,
hence, suggest that the transition is conformable (Fig. 3).
4.2. Paleoenvironmental reconstruction
The depositional environments associated with the facies de-
scribed above are reconstructed by means of foraminiferal and stable
carbon isotope analysis. Fig. 4a shows the succession of foraminiferal
assemblages and other microfossils at site Bayou Sale VI. The basal-
peat bed and the underlying paleosol are dominated by the
agglutinated taxa Haplophragmoides wilberti and Ammoastuta inepta.
This interval is interpreted to represent a brackish marsh environ-
ment. In the mud above the basal peat, the microfauna is dominated
by calcareous foraminifera of the taxa Ammonia beccarii sl. and
Elphidium gunteri, with Ammobaculites spp. being the next dominant
genus along with occurrences of H. wilberti.Ammobaculites spp. is
represented only by the early coiled portion of the test and therefore
no species identiﬁcation was possible. The calcareous foraminifera
(Ammonia and Elphidium) occur in open water where salinities are
generally greater than 10 ppt (Kane, 1967). The interval above the
peat is therefore interpreted to represent an open-water, brackish
lagoonal environment. Fragments of pelecypod taxa Rangia cuneata
and Macoma mitchelli, characteristic of shallow brackish environ-
ments with salinities of 2–15 ppt (Parker, 1959; Phleger, 1965; LaSalle
and de la Cruz, 1985) occur just above the basal peat (Fig. 1; sites VI, V,
VII, and 32), providing additional evidence that the brackish marsh
was abruptly replaced by a brackish lagoon.
Fig. 4b shows the succession of foraminiferal taxa and other
microfossils at site Bayou Sale IV. The section below 14.0 m is
dominated exclusively by H. wilberti and interpreted as a brackish
marsh environment. The interval above 14.0 m is represented by a
H. wilberti–A. inepta assemblage and is also interpreted as a brackish
marsh, possibly with a lower salinity due to the occurrence of A.
inepta.Scott et al. (1991) recorded A. inepta in Louisiana marshes with
salinities ranging from 3 to 5 ppt. H. wilberti and A. inepta are
represented in most samples only by their pseudochinous linings. In
one sample (13.76 m) whole specimens with the fragile test intact
were preserved, allowing for positive identiﬁcation of the species and
associated linings. The general environmental setting is similar to that
described by Kane (1967) where a H. wilberti–A. inepta assemblage
occurs as part of a fringe marsh with salinities less than 10 ppt.
The basal peat at sites VI and IV yielded δ
C values of −13.0‰and
−12.9 to −15.6‰, respectively, also indicative of a brackish marsh
environment (Chmura et al., 1987). The combined micropaleonto-
logical and geochemical data provide conclusive evidence that the
basal peat at both sites accumulated within the intertidal zone
(between mean tide level and mean spring high water). While at site
IV this environment persisted up section, at site VI the marsh was
abruptly replaced by an open-water, brackish lagoon. Given the
straightforward relationship between microfossil content and litho-
facies, all cores presented in this study (Table 1,Fig. 1) can be readily
interpreted in terms of depositional environments.
4.3. Elevation and sea-level relationship of basal peat
The elevation of past sea level was calculated using depth
measurements, elevation surveys, and the vertical indicative range
(sensu Van de Plassche, 1986) of basal peat with respect to sea level.
The depth is deﬁned as the contact between the basal-peat bed and
the underlying paleosol. Since basal peats were deposited on the
highly consolidated Pleistocene substrate, this essentially eliminates
elevation errors induced by post-depositional compaction. Multiple
cores were collected from each site to determine the measurement
error of the depth level of basal-peat beds (Table 2).
Summary of the Pleistocene–Holocene transition features in the Bayou Sale area,
western Mississippi Delta.
UTM coordinates Transition
N/A 0792.002 3281.500 646.980 1
N/A 0892.011 3282.320 647.320 2
N/A 0892.028 3279.040 644.000 1
−12.22 0892.013 3281.960 647.080 M 3
−12.34 0892.003 3282.140 647.180 M 2
−12.51 0892.002 3282.020 647.100 M 4
−12.76 0892.014 3282.400 647.380 M 6
−13.10 0892.015 3282.380 647.440 M 2
−13.11 0892.012 3282.300 647.300 M 1
−13.12 0892.017 3282.360 647.460 M 1
−13.14 0892.010 3282.260 647.280 M 5
−13.25 0892.018 3282.500 647.420 M 2
−13.45 0792.003 (3) 3283.420 647.640 M 14
−13.57 0792.001 3282.320 647.520 M 1
−13.68 0892.023 3279.120 643.620 M 1
−14.01 0892.027 3279.020 643.860 M 1
−14.05 0892.022 3279.100 643.560 M 1
−14.07 0892.026 3279.020 643.880 M 6
−14.08 0892.004 (IV) 3279.020 643.580 M 10
−14.11 0892.030 3279.000 643.780 M 1
−14.14 0892.025 (25) 3279.000 643.960 L 6
−14.16 0892.024 3279.020 643.840 M
−14.18 0892.031 3279.040 643.960 L 1
−14.39 0892.029 (29) 3279.040 643.980 L 6
−14.41 0892.007 (VI) 3281.100 646.080 LP 7
−14.75 0892.020 3279.260 644.460 LE 3
−15.09 0892.019 3279.160 644.080 LE 3
−15.42 0892.006 3278.960 642.820 LP 2
−15.44 0892.008 3280.040 645.060 LE 1
−15.54 0892.035 (VII) 3279.920 644.980 LP,LE 3
−15.57 0892.001 3281.700 646.880 LE 4
−15.61 0892.009 3279.800 644.820 LP 1
−15.69 0892.034 3279.940 645.000 LE 1
−15.87 0892.032 (32) 3279.800 644.840 LP, LE 5
−15.89 0892.021 3279.100 644.780 LE 3
−16.02 0892.033 3279.880 644.920 LP 1
−16.07 0892.005 (V) 3279.400 644.760 LP, LE 10
M = marsh mud/peat on paleosol, L = lagoonal mud on paleosol, LP = lagoonal mud
on peat on paleosol; LE = lagoonal mud erosive into paleosol or underlying strata.
Elevation is calculated with respect to NAVD 88; Elevation “N/A”indicates that the
transition is not well deﬁned at the site.
Number in parentheses indicates sites shown in Fig. 2.
Elevation is based only on one observation and is not as precise as those at sites
0892.025 and 0892.029 where multiple observations were obtained.
43Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50
The indicative range of basal peat refers to the vertical interval in
which basal-peat formation takes place with respect to mean sea
level. Van de Plassche (1982) showed that basal-peat accumulation in
coastal settings often occurs between mean sea level and mean high
water. As shown by the brackish signature of the δ
foraminiferal data discussed above, our basal-peat samples formed
of 0.33 ± 0.23 m
Weighted mean C ages
Marsh peat Shellfragments
Marsh peat of variable thickness
VI-1: 7395 ± 10
VI 1: 7395 ± 10
VII-3: 7605 ± 20
VII-2: 7665 ± 35
VII-1: 7710 ± 25
7645 ± 15
V-3: 7 545 ± 15
V-2: 7 590 ± 10
V-1: 7745 ± 15
7610 ± 10
(8390-8420) IV-3: 7325 ± 15
IV-2: 7440 ± 15
IV-1: 7440 ± 10
IV-1 / 2 / 3:
V-1 / 2 / 3:
VII-1 / 2 / 3:
7405 ± 10
Depth below NAVD 88 (m)
Fig. 2. Stratigraphy and chronology of selected cores from the Bayou Sale area (the complete set of cores is listed in Table 1). Note the striking difference between the deeper portion
of the record (sites V, 32, VII, and VI) that features rapid ﬂooding by means of lagoonal mud and the shallower portion (sites IV and 3) that is characterized by persistent marsh facies.
Sites VI, 29, 25 and IV capture an abrupt ﬂooding event where the basal-peat beds at sites VI and IV record its onset and end, respectively. Calibration of
C ages was performed with
OxCal (v4.0) for each basal-peat bed, both independently without considering its stratigraphic order (shown in italics in upper box) and by taking into account the stratigraphic
order of basal-peat beds (Table 3). NAVD 88 = North American Vertical Datum 88.
Sharp transition from basal peat to lagoonal mud
Sharp transition from paleosol to lagoonal mud Gradual transition from basal peat to marsh mud
Marsh peat Mud clast
Fig. 3. Photos of representative cores from sites IV, V, VI, and 25. The corresponding stratigraphic column is shown to its left. Photos are arranged to show the relative stratigraphic
levels and are not to scale. The diameter of each core is 3.8 cm. Radiocarbon ages of site V were obtained from core H that contains a well preserved basal-peat bed. Note that subtle
color differences between cores are partly due to variable light conditions when photos were taken in the ﬁeld. See text for further details.
44 Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50
within the intertidal zone. The average present-day spring tidal range
in coastal Louisiana is 0.47 m (González and Törnqvist, 2009).
Assuming the early Holocene tidal range was comparable to the
modern tidal range, the indicative range of the basal peats in our study
area would be 0.24 m (cf. González and Törnqvist, 2009). We convert
this value to a “two-sided error”of ± 0.12 m to be consistent with
error designations for depth and elevation measurements.
The cumulative uncertainty of the sea-level elevation inferred
from basal peat at each site can be computed with the following
LITHOLOGY KEY MICROFOSSIL KEY
CVA VRR FA
CVA VRR FA
CVA VRR FA
CVA VRR FA
CVA VRR FA
CVA VRR FA
CVA VRR FA
CVA VRR FA
VR - Very Rare (1-3)
R - Rare (4-10)
F - Few (11-20)
C - Common (21-40)
A - Abundant (41-100)
VA - Very Abundant (101-300)
CVA VR RFA
CVA VR RFA
LITHOLOGY KEY MICROFOSSIL KEY
Depth below NAVD88 (m)
Depth below NAVD88 (m)
Ammoast uta i nep t a
VR -Very Rare (1-3)
R - Rare (4-10)
F -Few (11-20)
C - Common (21-40)
A -Abundant (41-100)
VA - Ver
Fig. 4. Summary of the foraminiferal data of representative cores at site VI (A) and site IV (B). See text for further details.
45Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50
where Eis the total uncertainty; E
is the depth measurement error;E
the surveying error; and E
is the indicative range error. The elevation
measurements and uncertainties are summarized in Table 2.
For each basal-peat bed, different types of terrestrial botanical
macrofossils from mostly 2-cm-thick peat intervals were selected for
C dating. We obtained 21 AMS
C ages from sites Bayou Sale IV, V,
VI, and VII (Table 3). Since cores from sites V and VII show a highly
variable stratigraphy within a short distance and some cores (e.g.,
core D in Fig. 3) even display erosional features, utmost caution was
exercised and only well-preserved basal peats were chosen for
C measurement (Bayou Sale V-1a) was rejected because
it provided a younger age than all stratigraphically higher samples.
C ages were calibrated to calendar years Before
Present (BP =AD 1950) using OxCal (v4.0) (Bronk Ramsey, 1995) and
Summary of the elevation measurements and uncertainties of the basal peat/paleosol
C dated sites.
Site Mean elevation
Bayou Sale IV −14.08 10 0.09 0.05 0.12 0.16
Bayou Sale V −16.07 10 0.10 0.05 0.12 0.16
Bayou Sale VI −14.41 7 0.10 0.05 0.12 0.16
Bayou Sale VII −15.53 3 0.05 0.05 0.12 0.14
Radiocarbon ages of basal peat from the present study in the Bayou Sale area.
Material dated UCIAMS
Radiocarbon age Calibrated age (cal yr BP)
(N) (E) (
3279.02 643.58 0.31 14.37–14.39 11 Scirpus spp.
51101 7435 ± 15 7440 ± 10 IV-1 8210 8180–8255 79.8
14.37–14.39 25 herbaceous
14.35–14.37 2 Scirpus spp.
51103 7450 ± 60 7440 ± 15 IV-2 8205 8180–8235 60.5
14.35–14.37 9 herbaceous
14.33–14.35 15 Scirpus spp.
51105 7315 ± 15 7325 ± 15 IV-3 8180 8155–8205 84.4
14.33–14.35 7 Scirpus spp.
7405±10 8240 8180–8310 N/A
3279.40 644.76 0.33 16.38–16.40 9 Scirpus spp.
7745±15 V-1 8445 8425–8470 20.4
16.38–16.40 14 herbaceous
16.36–16.38 17 Scirpus spp.
51109 7525 ± 15 7590 ± 10 V-2 8405 8395–8415 59.6
16.36–16.38 14 herbaceous
16.34–16.36 12 Scirpus spp.
51111 7500 ± 15 7545 ± 15 V-3 8400 8390–8410 5.3
16.34–16.36 7 herbaceous
7610±10 8405 8390–8420 N/A
3281.10 646.08 0.55 14.94–14.96 24 Scirpus spp.
51113 7300 ± 15 7395 ± 10 VI-1 8260 8185–8310 78.1
14.94–14.96 4 large herbaceous
14.94–14.96 N30 small herbaceous
7395±10 8230 8175–8305 N/A
3279.92 644.98 0.41 15.93–15.94 1 large unidentiﬁed
59674 7710 ± 35 7710±25 VII-1 8395 8385–8405 0.1
15.93–15.94 2 Scirpus spp. achenes,
9 charcoal fragments
15.91–15.93 10 Scirpus spp.
59676 7665 ± 35 7665±35 VII-2 8395 8385–8405 22.1
15.91–15.93 9 Scirpus spp.
15.89–15.91 11 Scirpus spp.
59678 7600 ± 25 7605±20 VII-3 8395 8385–8405 99.7
15.89–15.91 30 small herbaceous
7645±15 8420 8400–8450 N/A
UTM coordinates (UTM zone 15R) with reference to North American Datum of 1983 (NAD83).
UCIAMS = University of California, Irvine, accelerator mass spectrometry; Weighted means were obtained with the “combination”function of OxCal (v4.0) (Bronk Ramsey, 1995).
Calibrated ages shown in italic were obtainedwith OxCal by treating each basal peat bed independently without considering their stratigraphic order. For theOxCal sequence analysis
approach, the stratigraphicorder of basal peat beds is taken into account and a typical 2 cm interval within a basal-peatbed is considered a ‘phase’for calibration (VII-1, 1 cm thick).
V-1a is rejected and calibration for V-1 was thus based on V-1b only; Calibrated ages are rounded to the nearest 5 years.
46 Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50
the IntCal09 calibration curve (Reimer et al., 2009). Since each basal-
peat bed contains multiple
C ages, we derived calibrated ages using
the combination feature of OxCal that calculates weighted mean
ages prior to calibration (Table 3). Together with sites I and II from
previous studies in the Bayou Sale area (Törnqvist et al., 2004a, 2006),
these calibrated ages were used to reconstruct the RSL history for a
~600 yr time span around 8.2 ka (Fig. 6a). In addition, we used the
OxCal sequence analysis feature that takes into account the
stratigraphic order of basal-peat beds by means of a model scheme
(Fig. 5). A quantitative measure of how well the calibrated ages agree
with the model scheme is indicated by the “A index”. Calibrated ages
with an A index over 60% are considered reliable (Bronk Ramsey,
1995). The calibrated ages of peat beds at sites IV through VII are
shown in Table 3; the calibrated ages of peat beds at sites I and II are
shown in Table 4.
5.1. Identifying a sea-level jump
While the entire data set (Table 1) exhibits evidence of
transgression and RSL rise, only one portion of the record (including
sites 25 and 29; Figs. 2, 3) features open-water lagoonal muds that
conformably onlap the paleosol with no basal peat. Collectively, sites
VI, 25, 29, and IV record an abrupt ﬂooding event that is unlike
anything seen elsewhere in our record (Table 1). The sharp transition
from basal peat to lagoonal mud at the deeper elevation of site VI
marks the onset of this ﬂooding event, while the re-emergence of
basal peat at the shallower elevation of site IV registers its end. The
absence of basal peat between these two elevations at sites 25 and 29
represents the ﬂooding event itself, when rapidly rising seas
prevented coastal marsh from developing and caused direct (con-
formable) deposition of lagoonal mud over the underlying paleosol.
The stratigraphy at sites 25 and 29 is distinctly different from the
remainder of the record (Table 1) and suggests a short pulse of near-
instantaneous ﬂooding due to extremely rapid sea-level rise. It is
unlikely that the distinctive stratigraphy of sites VI, 25, 29, and IV
resulted from gradual RSL rise or normal faulting. Had sea level risen
gradually, basal peat would occur at sites 25 and 29 as well.
Furthermore, recent work (Törnqvist et al., 2006) has shown that
fault activity within the study area during the Holocene has been
The OxCal combination approach shows that the basal-peat beds
at sites VI and IV yield indistinguishable ages of 8175–8305 and 8180–
8310 (2σ) cal yr BP (Fig. 2,Table 3), respectively, indicating that this
ﬂooding event occurred within the 8.18–8.31 ka time window. The
OxCal sequence analysis approach provides almost similar ages for the
basal-peat bed at site VI and the lowermost 2-cm interval (IV-1) of the
basal-peat bed at site IV of 8185–8310 and 8180–8255 (2σ) cal yr BP,
respectively (Fig. 2,Table 3). This similarity shows that our timing of
8.18–8.31 ka for the sea-level jump is robust.
The mean elevation difference of the basal-peat beds at sites VI and
IV is 0.33 m (Table 2) and the associated uncertainty was calculated
are the total uncertainty of the inferred sea level at
sites VI and IV, respectively. Since E
=0.16 m, ΔE= 0.23 m.
Therefore, the magnitude of the sea-level jump recorded between
sites VI and IV is 0.33± 0.23 m.
5.2. Final Lake Agassiz–Ojibway drainage
The reconstructed early Holocene RSL history (Fig. 6a) suggests
slightly higher rates of RSL rise before than after the sea-level jump
Bayou Sale VI
Bayou Sale IV
Bayou Sale VII
Bayou Sale V
Bayou Sale II
Bayou Sale I
Fig. 5. Model scheme for calibration of radiocarbon ages with the OxCal (v4.0) sequence
analysis feature (Bronk Ramsey, 1995). Following the deﬁnition for ‘sequence’and
‘phase’in OxCal, a peat bed is considered a sequence while each 2 cm interval within a
peat bed represents a phase. Thus, the bottom, middle, and top 2 cm intervals of a 6-cm-
thick peat bed (i.e., a sequence) constitute three phases (e.g., IV-1, IV-2, and IV-3) that
are contiguous (i.e., one phase starts immediately after the previous phase has ended,
without a time gap). Individual
C ages from a 2 cm interval (i.e., a phase) are
combined to produce a combined age (i.e., a weighted mean) for the corresponding
phase using the “
C date combination”option of OxCal. To account for the abrupt
ﬂooding event (see Section 5.1), the basal-peat beds at sites VI and IV were grouped
together to be considered as one sequence.
Radiocarbon ages of basal peat from previous studies (Törnqvist et al., 2004b, 2006) in the Bayou Sale area.
Sample Surface elevation
Radiocarbon age Calibrated age (cal yr BP)
CyrBP±1σ) Weighted mean (±1σ) Phase Weighted mean 2σrange A index (%)
Bayou Sale I-1
0.27 11.56–11.58 0.33 6997 ± 40 I-1 7865 7755–7940 102.3
Bayou Sale I-1 6995±40 7835 7725–7935 N/A
Bayou Sale II-1a
0.48 13.53–13.55 0.35 7480 ± 110 7280±30 II-1 8075 8020–8145 99.9
Bayou Sale II-1b
Bayou Sale II-2
13.50–13.53 7315±60 7315 ± 60 II-2 8070 8010–8140 109.9
Bayou Sale II-1/2 7290±25 8100 8025–8170 N/A
Calibrated ages from sites I and II of the previous studies were obtained in the same way as those from sites in this study.
Cores were hand-drilled and the vertical errors also include non-vertical drilling errors of 0.02 m per meter drilled. Weighted mean and calibrated ages are rounded to the
nearest 5 years.
47Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50
(~0.74 and ~0.68 cm/yr, respectively). While this difference is subtle
and not statistically signiﬁcant, it is consistent with the markedly
different stratigraphic successions prior to and after the sea-level
jump. The sharp transition from basal peat to lagoonal mud at sites V,
32, VII, and VI resulted from sudden ﬂooding and marsh drowning.
The early portion of our RSL record (sites V, 32, and VII) often exhibits
poorly preserved basal peat with large spatial variability over very
short distances and abundant evidence of transgressive erosion
(Fig. 3,Table 1). We therefore entertain the possibility that episodes
of near-instantaneous sea-level rise may have punctuated this phase
as well. The sea-level jump recorded at sites VI, 25, 29, and IV
represents the ﬁnal episode of rapid sea-level rise in the early
Holocene. The gradual transition from basal peat to marsh mud at
sites IV and 3 (Fig. 2) suggests that marsh accretion kept pace with RSL
rise after the sea-level jump, which is consistent with the slightly
lower rate of RSL rise during this time.
We interpret the sea-level jump within the 8.18–8.31 ka time
window as the result of the ﬁnal LAO outburst. This age range
represents the tightest available radiocarbon age constraint on the
timing of the ﬁnal LAO drainage (Fig. 6b). It is important to note that
this 130 year time interval arises from the radiocarbon calibration
procedure and is thus merely associated with the intrinsic limitations
of the dating technique. In addition to the stratigraphic evidence for
near-instantaneous drowning, hydraulic modeling has suggested that
the ﬂooding associated with the ﬁnal LAO outburst would have lasted
for as little as six months (Clarke et al., 2004). Therefore, this sea-level
jump must have occurred as a brief event at any time between 8.18
and 8.31 ka, not as a gradual ﬂooding that persisted for up to
The LAO drainage is often believed to have taken place in at least
two steps (Dominguez-Villar et al., 2009; Ellison et al., 2006;
Leverington et al., 2002). A high-resolution marine record from the
North Atlantic reveals two distinct episodes of surface ocean
freshening and associated cooling at 8.18–8.34 ka and ~8.49 ka,
respectively, suggesting two pulses of freshwater discharge (Ellison
et al., 2006). The striking concordance in the timing of the 8.18–
8.31 ka sea-level jump and the 8.18–8.34 ka climate anomaly in the
North Atlantic suggests that the sea-level jump very likely corre-
sponds to the younger pulse of the LAO drainage (i.e., the ﬁnal stage of
LAO drainage). Since RSL rise prior to the sea-level jump in our study
area occurred too rapidly for brackish marsh to be sustained, we
cannot rule out the presence of earlier pulses of LAO drainage. For
example, the similar ages of basal-peat beds at sites V and VII (Fig. 6a)
may indicate such an earlier pulse of freshwater drainage around
8.4 ka, which could potentially correspond to the earlier pulse of
~8.49 ka reported by Ellison et al. (2006) (It should be noted that their
age estimate is based on interpolation of limited
C dating evidence.).
We also note that our oldest two samples are consistent with the age
of the onset of a sea-level jump (8.54–8.38 ka, 2σ) recently recognized
in the Rhine–Meuse Delta (Hijma and Cohen, 2010). Such an earlier
freshwater pulse may have pre-conditioned the ocean–atmosphere
system (Wiersma and Jongma, 2009), setting the stage for the
principal climate event triggered by the ﬁnal stage of LAO drainage.
Our interpretation of a sea-level jump resulting from the ﬁnal LAO
drainage is also consistent with a reconstruction of the properties of
the Atlantic inﬂow that exhibits a pronounced, abrupt freshening of
the sub-thermocline at 8.2 ka, interpreted to result from glacial
freshwater discharge (Thornalley et al., 2009).
5.3. Volume of the freshwater drainage
The elevation data for sites VI and IV show that the sea-level jump
amounted to 0.33 ±0.23 m (Fig. 2). A maximum of 1.2±0.2 m of
abrupt sea-level rise was previously estimated in the study area
(Törnqvist et al., 2004a) and was subsequently considered to be
dominated by glacial isostatic adjustment (GIA) (Kendall et al., 2008)
which is now conﬁrmed by our reﬁned RSL record. These previous
studies lacked the stratigraphic details that are currently available
(particularly the abrupt ﬂooding evidence from sites 25 and 29) and
while GIA was indeed a signiﬁcant contributor to the overall high
rates of early Holocene RSL rise in this region, it was not an
appreciable factor for the short-lived sea-level jump identiﬁed here.
The sea-level rise of 0.33± 0.23 m would mathematically deﬁne a
range of 0.10 to 0.56 m for the sea-level jump, a value that must be
viewed in conjunction with ecological information on marsh
resiliency. It is unlikely that 0.1 m of sudden sea-level rise would
leave such a widespread stratigraphic signature. Studies of modern
7.7 7.8 7.9
8.1 8.2 8.3 8.4
8.6 8.7 8.8
7.7 7.8 7.9
8.1 8.2 8.3 8.4
8.6 8.7 8.8
0.33 ± 0.23 m
(Cheng et al., 2009)
Final drainage of Lakes Agassiz
and Ojibway (Barber et al., 1999)
Greenland ice cores
(Thomas et al., 2007)
(Kobashi et al., 2007)
Abrupt flooding in
the Mississippi Delta
~ 0.68 cm/yr
Abrupt flooding in
the Rhine-Meuse Delta
(Hijma & Cohen, 2010)
(Törnqvist et al., 2004a)
~ 0.74 cm/yr
Depth below NAVD88 (m)
Fig. 6. Relative sea-level (RSL) record and timing of the sea-level jump. (a) The early
Holocene RSL record shows a sea-level jump (red arrow) of 0.33± 0.23 m between sites
VI and IV and RSL rise rates slightly higher before than after the sea-level jump. Note
that the identiﬁcation of this sea-level jump is based on the distinct stratigraphy at sites
VI, 25, 29, and IV (Fig. 1). The sea-level index points (SLIPs) in green represent sites
where a gradual transition from basal peat to marsh mud occurs. The SLIPs in blue
indicate sites where the basal peat was abruptly overlain by brackish-lagoonal mud.
Each SLIP is deﬁned by the weighted mean age and the mean elevation of the
paleosol/basal peat contact. SLIP ages were obtained with OxCal (v4.0) that calibrated
C ages for each basal peat bed independently without considering its stratigraphic
order. Age error bars indicate the calibrated 2σage range; calculations of vertical errors
are discussed in the text. The green line is a linear regression for sites I, II and IV; the
blue line is a linear regression for sites VI, VII, and V. (b) Timing of the onset of the 8.2 ka
event (yellow bar) determined from Greenland ice cores and speleothem records from
China, Oman, and Brazil, compared to the inferred chronology of 8.18 to 8.31 ka for the
ﬁnal outburst of Lakes Agassiz and Ojibway. The present study reduces the age
uncertainty from about 600 years (1σrange) to about 130 years (2σrange). The results
of this study reﬁne the previous RSL record (Törnqvist et al., 2004a) around 8.2 ka in the
48 Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50
coastal ecosystems in the Mississippi Delta (Sasser, 1977) have shown
that Scirpus spp.-dominated marshes (i.e., comparable to our
reconstructed brackish marsh paleoenvironment) occur in the upper
portion of the tidal frame and are ﬂooded much less frequently
compared to Spartina alterniﬂora-dominated salt marshes (~20 to 160
vs. ~300 times per year, respectively). In other words, given the
average present-day spring tidal range for coastal Louisiana of 0.47 m,
a substantial sea-level rise is needed to permanently convert a
brackish marsh into an open-water lagoon. In addition, coastal marsh
plants in microtidal environments like our study area typically have
elevation ranges of 0.2 to 0.4 m (Silvestri et al., 2005), and, thus would
require a sea-level jump larger than these elevation ranges to enable
complete drowning of such ecosystems. In light of these observations,
we conservatively adopt a value of 0.2 m as the minimum amount of
abrupt sea-level rise. Thus, the sea-level jump around 8.2 ka in our
study area amounts to 0.20–0.56 m.
Since the catastrophic LAO drainage would perturb the gravita-
tional ﬁeld and lead to non-uniform changes in sea level (Kendall
et al., 2008), sea-level rise observed in the Mississippi Delta would
measure only a fraction of the eustatic sea-level rise (this fraction is
known as the ﬁngerprint). The ﬁngerprint value could range from 0.2
if the drainage consisted of LAO freshwater only, to 0.4 if the drainage
occurred exclusively as rapidly disintegrating ice over Hudson Bay
and Hudson Strait (Kendall et al., 2008). As the contribution from
disintegrating ice has been proposed to be relatively small (Clarke
et al., 2009), we assume a ﬁngerprint value of 0.25 for the ﬁnal LAO
drainage. The observed sea-level rise of 0.20 to 0.56 m at the
Mississippi Delta would then correspond to 0.8 m to 2.2 m of eustatic
sea-level rise associated with the ﬁnal LAO outburst (equivalent to ~ 3
), exceeding previous estimates for the ﬁnal LAO
drainage (e.g., Barber et al., 1999; Leverington et al., 2002; Törnqvist
et al., 2004a). Given the uncertainties in the position of the ice margin
of the retreating Laurentide Ice Sheet, the LAO volume may have been
larger than the reconstructed 0.45 m sea-level equivalent (SLE)
(Leverington et al., 2002) but it is conceivable that the freshwater
ﬂux included some Laurentide Ice Sheet melt, likely including
icebergs. Therefore, the volume estimate provided here most likely
includes both the LAO drainage and discharged icebergs. The relative
proportion of these two components, however, is difﬁcult to
A recent study in the Rhine–Meuse Delta inferred a sea-level jump
of ~3± 1.25 m at 8.54–8.2 ka (Hijma and Cohen, 2010). Although the
~3±1.25 m SLE is larger than our estimate of 0.8 to 2.2 m SLE, we
note that our estimate is exclusively associated with the ﬁnal stage of
LAO drainage, while the ~3 ±1.25 m SLE may well capture multiple
pulses of LAO drainage (Hijma and Cohen, 2010). Thus, the two
records can potentially be reconciled. Nevertheless, it is the ﬁnal pulse
of LAO drainage that triggered the widespread surface ocean
freshening and cooling (Ellison et al., 2006), corresponding to the
8.2 ka climate event as seen in most terrestrial records.
5.4. Implications for abrupt climate change
It has long been postulated that freshwater drainage can trigger
abrupt climate events, but large dating uncertainties have prevented
causal links from being convincingly established. Our new chronology
for the ﬁnal LAO drainage of 8.18 to 8.31 ka is indistinguishable from
the timing of the onset of the 8.2 ka event at 8.15 to 8.25 ka (Cheng
et al., 2009; Kobashi et al., 2007; Thomas et al., 2007)(Fig. 6b). This
allows for a near-instantaneous ocean–atmosphere response to
freshwater forcing, consistent with model predictions (LeGrande
et al., 2006; Wiersma and Renssen, 2006). Therefore, our study
provides independent chronologic evidence for the hypothesized
causal link between the ﬁnal LAO drainage and the 8.2 ka climate
event, and currently constitutes a rare ﬁrmly established example of a
major abrupt climate change that can be tied directly to a well-
identiﬁed source of freshwater forcing. The vigorous, ongoing debate
regarding such a causal link for other abrupt climate events such as
the Younger Dryas (e.g., Broecker et al., 2010; Carlson et al., 2007;
Firestone et al., 2007; Lowell et al., 2009) highlights the signiﬁcance of
independent age models for both cause and effect with century-scale
or better time resolution.
Finally, the new evidence presented here cannot only inform our
understanding of the sensitivity of the MOC to freshwater forcing, but
also help improve the accuracy of predictive climate models in the
context of future increased ice melt as a result of global warming.
Given that the freshwater volume that triggered the 8.2 ka climate
event likely amounted to more than 0.8 m of near-instantaneous
eustatic sea-level rise, our ﬁndings lend support to the notion (Meehl
et al., 2007) that abrupt cooling due to global warming in the next
century is relatively unlikely.
We present a high-resolution early Holocene sea-level record from
the Mississippi Delta that documents a distinct sea-level jump,
marked by a characteristic stratigraphic succession that is corrobo-
rated by paleoenvironmental reconstruction. The 0.20–0.56 m local
sea-level jump occurred within the 8.18 to 8.31 ka (2σ) time window
and is attributed to the ﬁnal drainage of proglacial Lake Agassiz–
Ojibway (LAO). Since the timing of the sea-level jump is indistin-
guishable from the onset of the 8.2 ka climate event, this study
provides compelling evidence for a nearly immediate ocean–atmo-
sphere response to the freshwater perturbation.
In addition, the total inferred eustatic sea-level rise at 8.2 ka (after
correction for gravitational effects) amounts to 0.8 to 2.2 m, consider-
ably higher than previous estimates for the ﬁnal stage of LAO drainage.
The new constraints on the timing and amount of ﬁnal LAO drainage
provide additional insight into the sensitivity of MOC to freshwater
perturbation, a crucial step toward understanding abrupt climate
change. For example, our ﬁndings support the notion that abrupt
cooling due to global warming in the next century is relatively unlikely.
Mike Blum (Louisiana State University) kindly made his Geoprobe
drilling system available for this study. Zhixiong Shen, Shiyong Yu,
Juan González, and Floyd DeMers are thanked for ﬁeld assistance and
land owners Debi Lauret and Antoine Luke for providing access to
their property. We are grateful to John Southon and his staff for
radiocarbon dating, to Brad Rosenheim for help with the stable carbon
isotope analysis, and to Hans Renssen, Marc Hijma, Sergio Fagherazzi,
Irv Mendelssohn, and George Flowers for elucidating discussion.
Comments by two referees signiﬁcantly improved the manuscript.
Funding for this study was provided by the Earth System History
program of the U.S. National Science Foundation (OCE-0601814) and
the McWilliams Fund of the Department of Earth and Environmental
Sciences, Tulane University.
Alley, R.B., Ágústsdóttir, A.M., 2005. The 8 k event: cause and consequences of a major
Holocene abrupt climate change. Quat. Sci. Rev. 24, 1123–1149.
Alley, R.B., et al., 2003. Abrupt climate change. Science 299, 2005–2010. doi:10.1126/
Autin, W.J., et al., 1991. Quaternary geology of the Lower Mississippi Valley. In:
Morrison, R.B. (Ed.), Quaternary Nonglacial Geology: Conterminous U.S.: Geological
Society of America: The Geology of North America, K-2, pp. 547–582.
Barber, D.C., et al., 1999. Forcing of the cold event 8200 yr ago by catastrophic drainage
of laurentide lakes. Nature 400, 344–348.
Broecker, W.S., et al., 1989. Routing of meltwater from Laurentide Ice Sheet during the
Younger Dryas cold episode. Nature 314, 318–321.
Broecker, W.S., et al., 2010. Putting the Younger Dryas cold event into context. Quat. Sci.
Rev. 29, 1078–1081. doi:10.1016/j.quascirev.2010.02.019.
49Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50
Bronk Ramsey, C., 1995. Radiocarbon calibration and analysis of stratigraphy: the OxCal
program. Radiocarbon 37, 425–430.
Carlson, A.E., et al., 2007. Geochemical proxies of North American freshwater routing
during the Younger Dryas cold event. Proc. Natl. Acad. Sci. 104, 6556–6561.
Carlson, A.E., Clark, P.U., Haley, B.A., Klinkhammer, G.P., 2009. Routing of western
Canadian Plains runoff during the 8.2 ka cold event. Geophys. Res. Lett. 36, L14704.
Cheng, H., et al., 2009. Timing and structure of the 8.2 kyr event inferred from
records of stalagmites from China, Oman, and Brazil. Geology 37, 1007–1010.
Chmura, G.L., Aharon, P., Socki, R.A., Abernethy, R., 1987. An inventory of
abundances in coastal wetlands of Louisiana, USA: vegetation and sediments.
Oecologia 74, 264–271.
Clarke, G.K.C., Leverington, D.W., Teller, J.T., Dyke, A.S., 2004. Paleohydraulics of the last
outburst ﬂood from glacial Lake Agassiz and the 8200 BP cold event. Quat. Sci. Rev.
Clarke, G.K., Bush, A.B.G., Bush, J.W.M., 2009. Freshwater discharge, sediment transport,
and modeled climate impacts of the ﬁnal drainage of glacial lake Agassiz. J. Clim. 22,
Clement, A.C., Peterson, L.C., 2008. Mechanisms of abrupt climate change in the last
glacial period. Rev. Geophys. 46, RG4002 Paper number 2006RG000204.
Dominguez-Villar, D., et al., 2009. Oxygen isotope precipitation anomaly in the North
Atlantic region during the 8.2 ka event. Geology 37, 1095–1098.
Donnelly, J.P., Cleary, P., Newby, P., Ettinger, R., 2004. Coupling instrumental and
geological records of sea-level change: evidence from southern New England of an
increase in the rate of sea-level rise in the late 19th century. Geophys. Res. Lett. 31,
Ellison, C.R.W., Chapman, M.R., Hall, I.R., 2006. Surface and deep ocean interactions
during the cold climate event 8200 years ago. Science 312, 1929–1932.
Firestone, R.B., et al., 2007. Evidence for an extraterrestrial impact 12,900 years ago that
contributed to the megafaunal extinctions and the Younger Dryas cooling. Proc.
Natl. Acad. Sci. 104, 1616–1621.
Fisher, T.G., Yansa, C.H., Lowell, T.V., Lepper, K., Hajdas, I., Ashworth, A.C., 2008. The
chronology, climate, and confusion of the Moorhead Phase of Glacial Lake Agassiz:
new results from the Ojata Beach, North Dakota, U.S.A. Quat. Sci. Rev. 27,
González, J.L., Törnqvist, T.E., 2009. A new Holocene sea-level record from the
Mississippi Delta: evidence for a climate/sea level connection? Quat. Sci. Rev. 28,
Heinrich, H., 1988. Origin and consequences of cyclic ice rafting in the northeast
Atlantic Ocean during the past 130,000 years. Quat. Res. 29, 142–152.
Hijma, M.P., Cohen, K.M., 2010. Timing and magnitude of the sea-level jump preluding
the 8200 yr event. Geology 38, 275–278.
Jelgersma, S., 1961. Holocene sea level changes in the Netherlands. Meded. Geol. Sticht.
Ser. C 6 (7), 1–100.
Kane, H.E., 1967. Recent microfaunal biofacies in Sabine Lake and environs, Texas and
Louisiana. J. Paleontol. 41, 947–964.
Kendall, R.A., Mitrovica, J.X., Milne, G.A., Törnqvist, T.E., Li, Y.X., 2008. The sea-level
ﬁngerprint of the 8.2 ka climate event. Geology 36, 423–426.
Kleiven, H.F., et al., 2008. Reduced North Atlantic deep water coeval with the glacial
Lake Agassiz freshwater outburst. Science 319, 60–64.
Kobashi, T., Severinghaus, J.P., Brook, E.J., Barnola, J.-M., Grachev, A.M., 2007. Precise
timing and characterization of abrupt climate change 8200 years ago from air
trapped in polar ice. Quat. Sci. Rev. 26, 1212–1222.
Lajeunesse, P., St-Onge, G., 2008. The subglacial origin of the Lake Agassiz–Ojibway ﬁnal
outburst ﬂood. Nat. Geosci. 1, 184–188. doi:10.1038/ngeo130.
LaSalle, M.W., de la Cruz, A.A., 1985. Species proﬁles: life histories and environmental
requirements of coastal ﬁshes and invertebrates (Gulf of Mexico): common rangia.
US Fish Wildl. Serv. Biol. Rep. 82 (11.31) US Army Corps of Engineers, TR EL-82-4:
LeGrande, A.N., et al., 2006. Consistent simulations of multiple proxy responses to an
abrupt climate change event. Proc. Natl. Acad. Sci. 103, 837–842.
Leverington, D.W., Mann, J.D., Teller, J.T., 2002. Changes in the bathymetry and volume
of glacial Lake Agassiz between 9200 and 7700
C yr B.P. Quat. Res. 57, 244–252.
Lowell, T.V., et al., 2009. Radiocarbon deglaciation chronology of the Thunder Bay,Ontario
area and implications for ice sheet retreat patterns. Quat. Sci. Rev. 28, 1597–1607.
Meehl, G.A., et al., 2007. Global Climate Projections. In: Solomon, S., Qin, D., Manning,
M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Climate Change
2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA.
Muscheler, R., Beer, J., Vonmoos, M., 2004. Causes and timing of the 8200 yr BP event
inferred from the comparison of the GRIP
Be and the tree ring Δ
C record. Quat.
Sci. Rev. 23, 2101–2111.
Parker, R.H., 1959. Macro-invertebrate assemblages of central Texas coastal bays and
Laguna Madre. Am. Assoc. Pet. Geol. Bull. 43, 2100–2166.
Phleger, F.B., 1965. Patterns of Marsh Foraminifera, Galveston Bay, Texas. Limnol.
Oceanogr. 10, R169–R184 (Suppl.).
Reimer, P.J., Baillie, M.G.L., Bard, E., et al., 2009. IntCal09 and Marine09 radiocarbon age
calibration curves, 0–50,000 years cal BP. Radiocarbon 51, 1111–1150.
Rohling, E.J., Pälike, H., 2005. Centennial-scale climate cooling with a sudden cold event
around 8,200 years ago. Nature 434, 975–979.
Sasser, C.E. 1977. Distribution of vegetation in Louisiana coastal marshes as response to
tidal ﬂooding. M.S. thesis, Louisiana State University.
Scott, D.B., Suter, J.R., Kosters, E.C., 1991. Marsh Foraminifera and arcellaceans of the
lower Mississippi Delta: Controls on spatial distributions. Micropaleontology 37,
Silvestri, S., Deﬁna, A., Marani, M., 2005. Tidal regime, salinity and salt marsh plant
zonation. Estuarine Coastal Shelf Sci. 62, 119–130.
Thomas, E.R., et al., 2007. The 8.2 ka event from Greenland ice cores. Quat. Sci. Rev. 26,
Thornalley, R.J.R., Elderﬁeld, H., McCave, I.N., 2009. Holocene oscillations in temper-
ature and salinity of the surface subpolar North Atlantic. Nature 457, 711–714.
Törnqvist, T.E., Bick, S.J., González, J.L., Van der Borg, K., De Jong, A.F.M., 2004a. Tracking
the sea-level signature of the 8.2 ka cooling event: new constraints from the
Mississippi Delta. Geophys. Res. Lett. 31, L23309. doi:10.1029/2004GL021429.
Törnqvist, T.E., et al., 2004b. Deciphering Holocene sea-level history on the U.S. Gulf
Coast: a high resolution record from the Mississippi Delta. Geol. Soc. Am. Bull. 116,
1026–1039. doi:10.1130/B2525 478.1.
Törnqvist, T.E., Bick, S.J., Van der Borg, K., De Jong, A.F.M., 2006. How stable is the
Mississippi Delta? Geology 34, 697–700.
Van de Plassche, O., 1982. Sea-level change and water-level movements in the
Netherlands during the Holocene. Meded. Rijks Geol. Dienst. 36, 1–93.
Van de Plassche, O., 1986. Introduction. In: Van de Plassche, O. (Ed.), Sea-Level
Research: a Manual for the Collection and Evaluation of Data. Geo Books, Norwich,
Wiersma, A.P., Jongma, J.I., 2009. A role for icebergs in the 8.2 ka climate event. Clim.
Dyn. 35, 535–549. doi:10.1007/s00382-009-0645-1.
Wiersma, A.P., Renssen, H., 2006. Model-data comparison for the 8.2 ka BP event:
conﬁrmation of a forcing mechanism by catastrophic drainage of Laurentide Lakes.
Quat. Sci. Rev. 25, 63–88.
50 Y.-X. Li et al. / Earth and Planetary Science Letters 315–316 (2012) 41–50