Geoscience Reference
In-Depth Information
and dispersal occurred quickly and that the observed shift
would be very small (less than 1 practical salinity unit or
psu), which would account for the lack of signal. However,
a distinct
500 m depth,
concurrent with the Agassiz discharge have been recognized
on Northeast Newfoundland Shelf [Miller et al., 2001,
2006], the sea surface impacts of the out
Although bottom water changes at 400
-
18 O minimum is recognized at this time on the
Laurentian Fan (Figure 1) and as far south as Cape Hatteras
[Keigwin et al., 2005]. Keigwin et al. [2005] proposed that
the meltwater might have
δ
ow along the
eastern Canadian margin have not been studied. Here we
investigate the out
s impact on sea surface conditions,
using new high-resolution palynological records from Notre
Dame Channel on the Northeast Newfoundland Shelf and
St. Anne
ow
'
flowed south as a coastal current,
hence the lack of isotopic signal along the Labrador Slope
or in the Labrador Sea, but this does not adequately explain
why they
s Basin on the northern Scotian Shelf. Basinal shelf
deposits have higher sedimentation rates and thus contain
expanded sequences allowing high-resolution analyses. We
selected the Notre Dame Channel core because it contains
DC beds deposited from discharge out of Hudson Strait
[Miller et al., 2006], one of which is coeval with the drainage
of Lake Agassiz. St. Anne
'
find a signal off the Scotian margin and as far
south as the Middle Atlantic Bight. The paucity of isotopic
evidence north of Grand Banks means that other avenues of
investigation are needed to determine the routing and the
impact of the outburst [Lewis and Miller,2005;Miller et al.,
2006; Lewis et al., 2009] on the surface waters along the
eastern Canadian margin; therefore we propose to use sed-
imentological evidence (e.g., detrital carbonate (DC)) and
palynological data to estimate the magnitude of changes in
sea surface conditions.
Sediment layers with enhanced DC contents are a proxy
for sediment transport out of Hudson Bay [Andrews and
Tedesco, 1992; Hillaire-Marcel et al., 2007]. These carbon-
ates originate from the entrainment and transport of particles
of glacial sediment derived from the limestone and dolomite
bedrock in Hudson Bay and Strait [Josenhans et al., 1986].
Holocene-age carbonate-rich sediment layers have been rec-
ognized in cores from several locations on the Labrador
Shelf and Newfoundland shelves and slopes [Piper et al.,
1978; Andrews et al., 1999; Miller et al., 2006; Hillaire-
Marcel et al., 2007; Tripsanas and Piper, 2008; Lewis et al.,
2009], and one has been found to be coeval with the
s Basin was selected because we
want to show that although most of the meltwater probably
was incorporated into the NAC, the impact of the meltwater
drainage was also felt farther downstream, although with
reduced amplitude of change in the sea surface conditions.
St. Anne ' s Basin provides a link between the DC proxy
evidence of the Agassiz discharge found off Labrador and
Newfoundland [Hillaire-Marcel et al., 2007; Lewis et al.,
2009] and the δ
'
18 O minimum observed off Cape Hatteras
[Keigwin et al., 2005].
1.2. Chronology for the Lake Agassiz Outburst Floods
oods through Hudson
Bay and Hudson Strait were radiocarbon dated to 8.47 ka
(7.7 14 C ka) with an error range of 8.16
The
final Lake Agassiz outburst
8.74 ka by Barber et
al. [1999] using marine biogenic carbonate (e.g., mollusc
shells) from sediment levels above and below a flood-gener-
ated red bed. This age was considered conservative (too old)
by Barber et al. [1999] because the 14 C dates were not
corrected (reduced) for the expected longer-than-present an-
nual sea ice cover during the early Holocene, which would
have limited the transfer of atmospheric 14 CinCO 2 to
oceanic surface waters [Bard et al., 1994; Mangerud et al.,
2006]. The effect of present seasonal sea ice cover is ac-
counted for in the normal practice of dating pre-A.D. 1950
museum specimens of mollusc shells to establish local 14 C
reservoir ages (and corrections for conventional 14 C dates).
These ages consist of a surface ocean model age (about 400
years in the Northern Hemisphere) plus a Δ R term specicto
the locality of interest [Stuiver and Braziunas, 1993]. The
-
nal
Lake Agassiz discharge [Lewis and Miller, 2005; Hillaire-
Marcel et al., 2007]. The DC was likely transported south-
ward via the Labrador Current (Figure 1) along the shelf and
upper slope [Lewis and Miller, 2005; Lewis et al., 2009].
Although carbonate bedrock also outcrops in the northern
Gulf of St. Lawrence region, it is unlikely that this bedrock
contributed early Holocene DC erosion products as the re-
gion was largely deglaciated by 10.85 ka (9.5 14 C ka) [Dyke
et al., 2003].
Based on the distribution of Agassiz DC beds, Lewis and
Miller [2005] and Lewis et al. [2009] proposed that the outer
portion of the meltwater plume would have dispersed into the
northern branch of the North Atlantic Current (NAC) along
the eastern Grand Banks slope, and the remainder would
have continued to
R
term accounts for local factors, which increase the reservoir
age; it is the difference between the model ocean age and the
known age of a museum specimen (elapsed time since col-
lection). The mean and standard deviation of present-day
Δ
flow southward inshore of the Gulf
Stream. Along Cape Hatteras, the temperature is warmer and
salinity higher, and hence, the contrast with the isotopic
composition of the meltwater in
ux would be pronounced
and could account for the almost 1 isotopic shift observed
by Keigwin et al. [2005].
R
values as determined by McNeely et al. [2006] are as follows:
southern Hudson Bay 275±40 years, Hudson Strait 125±40
Δ
Search WWH ::




Custom Search