Paleoenvironmental Time Slice Reconstructions, Glaciated North America
Arthur S. Dyke
Geological Survey of Canada
Paleoenvironmental conditions in northern North America for the period from the Last
Glacial Maximum (18 000 radiocarbon years ago; 21 400 calendar years ago) to present
can be reconstructed from a variety of large data sets. These data sets comprise
radiocarbon and stratigraphically dated time series (e.g., pollen stratigraphic records),
fossil assemblages (e.g., marine molluscs and mammals), and environmental events (e.g.,
deglaciation, relative sea-level change, alpine glacier advances, deposition of eolian
sediment, or formation of ice wedges ). Paleoenvironmental databases are maintained by
the Geological Survey of Canada in order to document how past environments have
responded to climate change during the last major climatic hemicycle that was forced by
changes in solar radiation. The Table below lists the size and composition of these
databases.
Table 1. Paleoenvironmental Databases, Glaciated North America
Database
Size
Deglaciation
4042 dates
Sea Level Change
>10 000 dates; 535 RSL curves
Pollen
1170 sites
Plant Macrofossils
829 sites (ca 1000 dates)
Mammals
Wetland Inception
4555 dates (Terrestrial, 3571; Marine, 804;
Both, 175)
1741 dates
Eolian
1332 sites
Permafrost
672 sites
Marine Molluscs
4274 dated assemblages
Alpine Glaciers
639 14C dates; 1069 Dendro, etc ages
New deglaciation maps for North America were prepared in response to the need by
paleoclimate and isostatic adjustment modellers for updated ice margin sequences at 500
year time steps. These maps have much better chronological control than previous
versions, the net effect of which is to bring the deglacial history of North America into
better correlation with the climate history expressed in the North Atlantic marine
sediment records and in the Greenland ice cores. In particular, the Younger Dryas cooling
(see Fig. 1) is now recognized as a major driver of Laurentide end moraine construction
and the 8200 calendar year BP cold event is
firmly correlated with the flushing of giant glacial lakes Agassiz and Ojibway into the
Labrador Sea via Hudson Strait (see Fig 2).
These ice margin maps provide the input for reconstructions of ice surface elevation
through time, a critical boundary condition in paleoclimate modelling. The replication
targets in isostatic adjustment modelling are relative sea-level curves and isobase maps
derived from them and from the deformations of glacial lake strandlines. A new series of
isobase maps is currently under construction by Dyke, Shaw, Lewis, Thorleifson, and
James at GSC.
The history and future of wetlands is a major concern because paludification has
fundamentally altered the structure of Canadian forests, converting large areas into
‘muskeg’, and because peatlands are the largest terrestrial sink of carbon. Peatland
formation lagged deglaciation by 2000 to 3000 years, but appears to be proceeding apace
today (Fig. 3).
The major terrestrial biomes (vegetation and faunal assemblages) of northern North
America have been reconstructed for the interval from LGM to present from pollen, plant
macrofossil, and terrestrial mammal databases. These biome maps will soon be posted on
the GSC Quaternary Paleoenvironments Website. Biomes have not only shifted by
hundreds to thousands of kilometres during this interval but have also changed strongly
in composition. For example, the Herb Tundra in Beringia (unglaciated Alaska-Yukon)
of LGM to about 13 000 radiocarbon years BP bears little resemblance to herb tundras of
today. The ancestral Beringian herb tundra, despite colder LGM temperatures and drier
conditions, was much more productive and capable of supporting herds of large grazers
that could not be supported in any tundra area today. Dale Guthrie (2001; Quaternary
Science Reviews 20:549-574) argues that this productivity was due to persistent sunny
skies under the high atmospheric pressure at that time. Similarly, the Late Pleistocene
Boreal Forests were different in composition from those of today, as were other forest
belts. These differences are due to the fact that species, both plants and animals,
responded individualistically to climate change. Modern associations, which we tend to
think of as normal or natural, are thus not analogous to earlier associations. For the same
reason, future biomes, forced by anthropogenic climate change will not be
compositionally identical to present ones.
Marine biomes have shifted just as greatly as have terrestrial biomes. Indeed, shifts of
water mass boundaries during postglacial time were even greater. For example, for most
of postglacial time, the boundary between Arctic and Subarctic waters in Baffin Bay was
displaced 1000 km north of its present position, allowing warmth-demanding clams to
extend their ranges accordingly. Such biome shifts are poorly understood in terms of
exact forcings, but they have great implications for the potential fate of commercial fish
and clam species and for terrestrial climate.
The great challenge in terms of the impacts of future climate change is to understand past
biome changes sufficiently well that they may be used to forecast changes over the next
century to millennium. This will require biome models that can reproduce LGM to
present changes with convincing fidelity and accurate models of future climate linked to
biome models.
Figure 1. Paleogeography at 11 000 radiocarbon years ago, the beginning of the Younger
Dryas cold event.
Figure 2. Paleogeography at 7700 radiocarbon years ago (8400 calendar years ago), the
beginning of the 8200 BP Cold Event, caused by the flushing of Glacial lake AgassizOjibway into Hudson Strait along a subglacial channel indicated on the map.
Figure 3. The pace of wetland formation compared to the rate at which land became icefree (deglaciated) in North America since the Last Glacial Maximum.
Wetland Inception & Deglaciation
100
60
40
20
0
20000
15000
10000
5000
Radiocarbon years BP
% Deglaciated
% Wetlands formed
0
Percent
80