Abrupt Climate Changes During the Holocene Across North America From Pollen and Paleolimnological Records - Abrupt Climate Change: Mechanisms, Patterns, and Impacts - page 172

accumulated pollen data. The North American Pollen Data-
base (NAPD) [Grimm, 2010] has incorporated pollen dia-
grams from across North America, with a relatively high
density of sites in the northeastern region (Figure 1). Avail-
able at the National Climate Data Center (www.ncdc.noaa.
gov/paleo), the data are stored in a uniform and accessible
format, so queries can extract all the data needed for a
particular project. The database is taxonomically complex,
so usage is not straightforward; some examples are reviewed
in the work of Gajewski [2008].
For quantitative paleoclimate reconstructions, a spatial
array of modern pollen samples, along with associated cli-
mate data are necessary. Recently, a continental database of
modern pollen data has been made publically available in
easily accessible form [Whitmore et al., 2005]. This database
consists of roughly 4600 sites from across North America,
and new data are being periodically added (Figure 1). Anal-
ysis of these data illustrates the climatic constraints on
the major pollen taxa used in paleoclimate reconstructions
[Williams et al., 2006].

The large numbers of sites provides opportunities but also
challenges. A dense network of sites across an entire conti-
nent provides opportunities for regional and continental
maps depicting climates at various times in the past [Wright
et al., 1993; Viau et al., 2006]. However, when synthesizing
the large numbers of pollen diagrams, each slightly different
due to local site conditions or disturbance histories, it is easy
to get lost in all of the details. Diagrams from different
regions of North America can contain dissimilar pollen taxa,
so to determine which changes are caused by climate varia-
tions and which by some local factor requires knowledge of
the local ecological and biogeographic conditions (see be-
low). Many of these diagrams have few radiocarbon dates
and low temporal resolution of samples. The spatial replica-
tion of the results nevertheless lends con dence to our con-

clusions, as inspection and multivariate analysis of many
pollen diagrams from a region show similarities between
sites, suggesting a climate cause to many of the vegetation
transitions [Gajewski, 1987, 1993]. Studies of the transitions
show preferred times of transition in pollen data across North
America [Viau et al., 2002] and that these are coherent with
changes in Europe [Gajewski et al., 2006].

2.2. Data Suitability and Biogeography

At a descriptive level, paleoecologists have been able to
make general conclusions about past climate changes using
pollen and plant macrofossil data. Indeed, many of the terms
used in discussing Holocene climate variability, such as
Younger Dryas, Bolling, and Ållerød were originally seen in
peat and lake sediments from Scandinavia [Birks and Birks,
1980]. The overall Milankovitch-scale climate variability has
been successfully reconstructed using pollen data from sev-
eral regions of the world and compared to climate model
output [Wright et al.,1993;Sawada et al., 2004]. Many
issues remain, and research is continuing to develop the data
needed for this research effort. There are still large areas with
few available fossil or modern data, for example, in desert
regions, in very cold regions, and in mountain areas.
One reason for the success in using vegetation and pollen
records for paleoclimate work is that the ecological toler-
ances and biogeography of the major tree taxa is well known.
Forestry and ecological data and knowledge, accumulated
over a century, have resulted in not only an understanding of
the site conditions associated with optimal growth for a
species but also in maps depicting the distribution in space
of most vascular plants. In Europe and eastern North Amer-
ica, the postglacial migration of the major tree taxa has been
mapped [Williams et al., 2004; Huntley and Webb, 1989].
This basic biological knowledge means that the reasonable-
ness of resulting climate reconstructions can be assessed.

Figure 1. Location of pollen cores (open circles) in the Global
Pollen Database [Grimm, 2010]. Closed circles are location of
modern pollen data, used for calibration of paleoclimate reconstruc-
tions [Whitmore et al., 2005].