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 145

palynomorphs in the sediments [Stockmarr, 1971]. Palyno-
morphs were examined, identified, and counted under a light
transmission microscope (Nikon Eclipse) at a magni

were evident where total carbonate values rise to between
15% and 40% (Figure 2). Similar DC beds are found at other
locations along the eastern Canadian margin [Lewis et al.,
2009]. In core 19, the four DC beds occur at 265

cation
of 400 times. At least 300 pollen grains and 300 dinocysts
were counted.
Zonation (for pollen as well as dino

285,
295 - 380, 405 - 440, and 820 - 860 cm depth with the thickest
bed at 295

-

agellate cyst data) is
based on cluster analysis of the data, using the PSIMPOLL
and PSCOMB software [Bennett, 1996] and visual inspec-
tion. The ecological af

ed as coeval with the Agassiz
drainage (Figure 2). A sympathetic variation of dolomite
(CaMg(CO 3 ) 2 ) (not shown) and total carbonate (Figure 2)
indicates these beds are mainly of detrital origin and are not
derived from biogenic sources, which would produce mainly
calcite or aragonite (CaCO 3 ). Nearby core 07 (50°53.2 ′ N 53°
18.0

- 380 cm identi

agellate cysts species
were also considered when determining the zonations (see
regional overview).
Sea surface conditions were reconstructed by the best
analog technique, using the GEOTOP dino

nities of dino

W, 457 m water depth) also has radiocarbon chronological
control [Miller, 1999; Miller et al., 2006; Lewis et al.,2009].
This control was transferred to core 19 using the correlation of
DC layers between the two cores (Figure 2). In addition to the
marine radiocarbon dates, terrestrial radiocarbon ages in a
radiocarbon-dated pollen record from a Northeastern New-
foundland coastal site (Compass Pond) [Dyer, 1986] were
correlated to core 19 using similar features and zones in their
pollen records. More details are given in the pollen zonation
section below. The same procedure was done for core 12.
Core 12 is composed mainly of olive to olive gray clay or
mud [Freeman, 1986]. Carbonate analysis was not done on
core 12. Two radiocarbon ages were obtained on foraminif-
era from piston core 12 (Table 2). As for core 19, we
supplemented our 14 C radiocarbon chronology by comparing
the pollen record from core 12 to onshore pollen records
from Cape Breton and mainland Nova Scotia [Livingstone,
1968; Green, 1987; Ogden, 1987].
The DC beds were found to contain reduced concentra-
tions of palynomorphs and foraminifera (see section 5)

′

agellate cyst
database [de Vernal et al., 2001; Guiot and de Vernal,
2007]. Owing to the strong regionalism in the distribution
of the dino

k and de Vernal, 2008;
Marret and Zonneveld, 2003], we selected sites from the
North Atlantic Ocean and adjacent basins (Gulf of St. Law-
rence, Gulf of Maine, Hudson Bay, Labrador Sea, Norwegian
Sea) and the Arctic Ocean. Surface dinoflagellate cyst assem-
blages from another 58 sites from the Scotian Shelf and Slope
were added to provide additional analogs for core 12 [Levac,
2002]. A number of unpublished surface sediment samples
from the Northern Newfoundland Shelf were added to

agellate cyst taxa [Taou

ll a
void in that area of the database [see Levac, 2002, 2003]. In
total, 601 reference sites were used.
For the reconstructions, each fossil assemblage is compared
with all the modern spectra in the reference database, and the
10 most similar spectra are selected, using the same weight for
each taxon. The degree of similarity (the weight) between the
fossil assemblage and each of the modern analogs is deter-
mined by the Euclidian distance between them. The set of
analogs provides the paleoenvironmental data from which a
weighted average is calculated for each sea surface parameter
(SST, SSS, seasonal duration of sea ice cover). The minimum
and maximum environmental values in the set of analogs are
used to de

dence interval [de Vernal et al., 2001;
Guiot and de Vernal, 2007]. The accuracy of the method, with
the data set used, is de

ne the con

ned as the standard deviation of
reconstructed values. For August SST, the accuracy is
±1.35°C, for SSS, it is ±2 ‰ , and for the seasonal duration of
the sea ice cover, it is ±1.28 months year 1 .

4. CORE STRATIGRAPHY

Core 19 is composed of clay or silty clay, with colors
ranging from gray, to olive gray, or dark gray (Figure 3).
Some intervals were initially observed to be calcareous,
which is confirmed by carbonate analysis (Figure 2).
Background values in total carbonate are low (less than
10%) in core 19. However, four intervals enriched in DC

Figure 3. Sedimentology of core HU87033-19, 0 - 898 cm, Notre
Dame Channel, on the Northeast Newfoundland Shelf. TWC means
trigger weigh core; PC is piston core.