The Impact of the Final Lake Agassiz Flood Recorded in Northeast Newfoundland and Northern Scotian Shelves Based on Century-Scale Palynological Data - Abrupt Climate Change: Mechanisms, Patterns, and Impacts - page 154

normal ranges, and shelf faunas reappeared [Miller, 1999;
Miller et al., 2006].
On the Scotian Shelf, the impact of the meltwater plume
might have been restricted to the upper part of the water
column, as benthic foraminiferal evidence of the influx is
absent in Emerald Basin (core DA77002-20 [Scott et al.,
1984]), Canso Basin (core DA80004-33 [Scott et al.,
1984]), and St. Anne ' s Basin (core 12 [Freeman, 1986]).
There is no evidence of a freshwater excursion in benthic
foraminiferal isotopic records from Emerald and Canso ba-
sins [Scott et al., 1989].

The discontinuous geographical distribution of micropa-
leontological evidence for a freshwater presence could stem
from the monospeci agellate

cysts as well as planktic foraminifera). Extreme conditions
can lead to monospecific or low-diversity assemblages, mak-
ing it dif

c nature of assemblages (dino

cult to interpret palaeoenvironmental changes. An-
other explanation is that because the Labrador Shelf region
remained under a meltwater “ regime ” until the end of the
melting [Levac, 2001, 2002], the Labrador Shelf was already
affected by relatively frequent freshwater pulses, so that mi-
crofossil assemblages were relatively insensitive to yet an-
other (even if much larger) meltwater event. Lowered SSS
characterized the North Atlantic margins in the early Holo-
cene [de Vernal and Hillaire-Marcel, 2000]. Analysis of early
Holocene summer-melt layers in ice caps from the Canadian
Arctic [Koerner and Fisher, 1990] also suggest that pulses of
meltwater from the Arctic might have continued to affect the
Labrador Shelf episodically in the early Holocene. This idea
is supported by the fact that dino

6.3. Geographical Distribution of Paleoecological
and Isotopic Evidence for Meltwater Presence

The Agassiz outburst event is recorded on the Northeast
Newfoundland Shelf, Scotian Shelf, Laurentian Fan, and off
Cape Hatteras, with a latitudinal decrease in the amplitude of
the SSS changes associated with the floods. We suggest that
the Agassiz fresh water traveled southward via the Labrador
Current and gradually started to disperse into the NAC along
the eastern margin of the Grand Banks. A smaller portion
continued

agellate cyst assemblages
similar to the modern ones are found in most cores from the
eastern Canadian margin sometime after about 7 ka, once the
in

uence from the melting ice sheet had decreased suf cient-

flowing south toward the Scotian Shelf, and along
the Scotian margin, and some of this water may have mixed
with the Gulf Stream. This is consistent with Keigwin et al.
[2005] who proposed that the meltwater could have initially
followed a narrow path along the coast and then dispersed
into the shelf/upper slope water farther south.
While the DC proxy for

ly [Levac, 2001, 2002; Levac et al., 2001].

7. CONCLUSIONS

A DC layer in Notre Dame Channel core 19 (the Northeast
Newfoundland Shelf ) has been radiocarbon dated to the age
of the Lake Agassiz floods out of Hudson Strait is

distinctive and continuous along the Labrador and New-
foundland shelves and upper slopes [Lewis et al., 2009], the
event is not recorded by foraminiferal proxies in records
closer to the meltwater source or in Hudson Strait and the
Labrador Sea [Andrews et al., 1999; Keigwin et al., 2005;
Hillaire-Marcel et al., 2007]. The lack of a

final drainage [Lewis et al., 2009] and

contains dino agellate cyst assemblages indicative of lower

SST, SSS, and of higher water stratification. Reconstructions
reveal two pulses of lower SSS within the Agassiz DC bed
(Figure 12). Core 12 from St. Anne

s Basin (Scotian Shelf )
records a smaller-amplitude drop in SSS at that same time
(Figure 13). The event is also recorded at other sites on the
Scotian Shelf (e.g., La Have Basin, Emerald Basin) [Levac,
2001, 2003; Scott et al., 1984].
Based on this new evidence, we suggest that the Lake
Agassiz meltwater

'

18 O signal in
the Labrador Sea is not surprising if, as Keigwin et al.
[2005] and Hillaire-Marcel et al. [2007] suggested, the
freshwater transport was along the shelf with some over-

δ

flow downslope, a conclusion also reached by Lewis et al.
[2009] after analyses at additional sites along the margin.
The water that would have been transported farther offshore
would have dispersed quickly, too quickly to leave an
isotopic signature. The absence of the signal in Hudson
Strait and along the margin can be explained by the small
difference between the isotopic composition of Lake Agassiz
waters and the present-day freshwater discharge in the NW
North Atlantic, a difference of only about 4

flux traveled southward over the North-
east Newfoundland Shelf, and some was dispersed in the
NAC along the eastern Grand Banks slope. Some water
continued to

flow over the Scotian Shelf/upper slope as
evidenced from the signal found in the Laurentian Fan
[Keigwin et al., 2005] and on the Scotian Shelf [Levac,
2001; this chapter]. Some of this water may have then
dispersed into the Gulf Stream. The remainder flowed south,
inshore of the Gulf Stream and was recorded off Cape
Hatteras [Keigwin et al., 2005].
The absence of micropaleontological evidence for the
event in cores from the northern Labrador Shelf could be
explained by the low diversity of the microfossil assemblages

[Hillaire-
Marcel et al., 2007, 2008], which corresponds to less than
a0.1 ‰ shift in isotopic composition in Neogloboquadrina
pachyderma within the salinity range of its habitat [Hillaire-
Marcel et al., 2007, 2008].

‰