Paleoclimatic Reconstruction From Long Quaternary Pollen Records

There are now numerous palynological records that span the late Quaternary and, in some cases, extend back continuously into the Pliocene. Dating is, of course, problematic beyond the range of radiocarbon dating (~40,000 yr in most cases) and often the records are simply assumed to extend back to the last interglacial, and beyond, based on the character of the pollen record itself. In some cases, correlation with marine oxygen isotope stratigraphy, either directly (in marine sedimentary records) or indirectly (by comparisons between terrestrial and nearby marine records) can prove helpful in constructing a chronology. Here, a selected number of long palynological records from different areas of the world are presented to provide an overview of how such records, even if not well-dated or calibrated quantitatively, can reveal important paleoclimatic changes of large-scale significance. They provide important insights into changes in terrestrial climates on timescales comparable to the important ice core and marine sedimentary records, thereby completing the global picture of climatic variations over the Quaternary period.

9.7.1 Europe

Several long records extending >100 ka have been recorded from lakes and bogs in Europe (Fig. 9.17). Of these, La Grande Pile in the French Vosges mountains has been studied in the most detail, with over 20 cores recovered for pollen, plant macrofossils, and sedimentological and faunal (insect) analysis (Woillard, 1978; Woillard and Mook, 1982; Beaulieu and Reille, 1992; Seret et al, 1992; Pons et al, 1992; Guiot et al, 1992,1993; Ponel, 1995). The pollen record from this location is well-documented, enabling the overall sequence of vegetation changes over the last interglacial-glacial cycle to be established (Fig. 9.18). The penultimate glaciation (Riss) was characterized by an open grassland with few trees. The transition to peak interglacial conditions in the Eemian is marked by a clear sequence of taxa,31 first Juniperus, then Pinus and Betula, Ulmus and Quercus with Corylus, and, finally, Taxus, indicative of the interglacial climatic optimum. Subsequent climatic deterioration is marked by a rise in Abies and Carpinus, then Picea, Pinus, Betula, and Juniperus once again. This short cool episode is the first of two such periods (termed Melisey I and II) separated by more temperate conditions (St Germain I and II). This is clearly seen in the relative proportions of arboreal to nonarboreal pollen (AP/NAP) with the colder periods characterized by sharp increases in the NAP fraction of the pollen sum. The onset of full glacial conditions (the Pleniglacial) began ~70 ka B.P. when Artemisia and other cool steppe taxa increased in abundance. The following 50 ka was dominated by NAP, although occasional fluctuations in the abundance of arboreal taxa suggest that conditions were not completely static. Coldest and driest conditions occurred in the late glacial (Tardiglacial) as shown by lowest NAP values and highest levels of Artemisia (Fig. 9.18).

These interpretations have been quantified by Guiot et al. (1989, 1992) as discussed earlier (Fig. 9.15) based on calibration with modern pollen assemblages. However, the reconstructions are problematic because there are no good modern

31 For common names, see Table 9.1.

Last Glacial Extent Europe
FIGURE 9.17 Locations of the longest European pollen records, which span at least the last glacial-interglacial cycle (Guiot etal., 1993).

analogs (at least not in Europe) for pollen assemblages recorded in the coldest parts of the last glaciation. Indeed, the full glacial vegetation assemblage may have more in common with the cold, arid steppes of interior Asia and Tibet today than anywhere in Europe. Furthermore, periods of rapid change may not be adequately recorded in the pollen record because the overall inertia in vegetation assemblages makes pollen a poor indicator of short, abrupt climatic changes. These concerns have led to the incorporation of other climatic indicators, such as sedimentary characteristics (Seret et al., 1992) and insects (Ponel, 1995; Guiot et al., 1993) together with pollen in recent reconstructions. For example, the loess content of the sediment is highest during the coldest intervals of the Pleniglacial and organic matter content is lowest. Taking such factors into account results in significant differences of interpretation (up to 6 °C) with more variability of temperature during the full glacial period. Further studies have combined pollen with insect fauna to refine the reconstruction of mean annual temperature (Fig. 9.19). Insects respond quickly to climatic fluctuations (see Section 8.4) and so are especially useful in a period when climate is unstable (Ponel, 1995). Combining both

GRANDE PILE XX

GRANDE PILE XX

Pollen Diagram Grande Pile

FIGURE 9.18 Pollen diagram from La Grande Pile,Vosges, France representing changes from the penultimate glaciation (Riss) to the late glacial ("TardiWiirm" - see pollen zones indicated on the right). Numbers across bottom correspond to principal pollen types; the sequential change in vegetation during a complete glacial-interglacial cycle is clearly seen. I = Juniperus; 2 ~ Salix; 3 = Betula; 4 = Ulmus; 5 = Deciduous Quercus; 6 = Corylus; 7 = Fraxinus; 8 = Alnus; 9 = Taxus; 10 = Carpinus; 11 = Abies; 12 = Picea; 13 = Pinus; 14 = Poaceae; 15 = Artemisia; 16 = Heliophytes (various); 17 = Cyperaceae; 18 = isoetes.The thin line between 13 and 14 represents the overall arboreal/nonarboreal pollen ratio, with arboreal pollen increasing to the right (de Beaulieu and Reille, 1992).

FIGURE 9.18 Pollen diagram from La Grande Pile,Vosges, France representing changes from the penultimate glaciation (Riss) to the late glacial ("TardiWiirm" - see pollen zones indicated on the right). Numbers across bottom correspond to principal pollen types; the sequential change in vegetation during a complete glacial-interglacial cycle is clearly seen. I = Juniperus; 2 ~ Salix; 3 = Betula; 4 = Ulmus; 5 = Deciduous Quercus; 6 = Corylus; 7 = Fraxinus; 8 = Alnus; 9 = Taxus; 10 = Carpinus; 11 = Abies; 12 = Picea; 13 = Pinus; 14 = Poaceae; 15 = Artemisia; 16 = Heliophytes (various); 17 = Cyperaceae; 18 = isoetes.The thin line between 13 and 14 represents the overall arboreal/nonarboreal pollen ratio, with arboreal pollen increasing to the right (de Beaulieu and Reille, 1992).

insect and pollen data reveals many rapid changes in climate during the Pleniglacial, some of which correspond to Dansgaard-Oeschger oscillations seen in the Greenland ice cores, and possibly also to North Atlantic Heinrich events. This multivariate approach to paleoclimate reconstruction has much to offer as it is clear that no one variable can provide an accurate view of past climate in Europe during full glacial conditions. By pooling the information each proxy provides, a more reliable reconstruction can be obtained.

^----p

Dtleri only

r

\ A

V,

ñ

w

w

Ï2

j

V

V

Beetes*

u

ft**!

Ï

i

\ J

„ Ar

^ A

i\ N

V

V

I A

i

4

V

V/l

:::

(lr

I

V v

Ornanir mattnr +

Pelen

\

\

m

Û J

1

u

A

_ flf

JU

Time in ka BP

FIGURE 9.19 Mean annual temperature at La Grande Pile over the last 140,000 yr, reconstructed from pollen alone (top), and pollen constrained by the additional consideration of organic matter variations (lower diagram) or insects (Coleoptera) (middle) (cf. Fig. 9.15). Modern pollen analogs are not good descriptors of climatic conditions during the main glacial phase (-70-20 ka B.P.) or when climate changed abruptly. By considering insect fauna and sedimentological changes, additional constraints on paleotemperature estimates are introduced (Guiot et a/., 1993).

9.7.2 Sabana de Bogotá, Colombia

The longest continuous sedimentary record from South America comes from a mountain-rimmed basin in the Colombian Andes (Hooghiemstra, 1984). Known as the Sabana de Bogotá, it is a former lake basin, currently at -2550 m above sea level. The record (from near the village of Funza) extends back to late Miocene time and indicates that the region experienced -2 km of uplift between -5 and 3 Ma B.P. Vegetation in the area is zoned altitudinally, ranging from tropical forest below -1000 m, to Andean forest above -2300 m, to páramo above -3500 m. At the highest elevations, perennial snow is found. At times in the past, vegetation has migrated along the mountain slopes but maintained its characteristic floristic zonation as the climate has changed (Fig. 9.20) (Van der Hammen, 1974; Hooghiemstra and Ran, 1994). By grouping the pollen types characteristic of each major zone, it has been possible to reconstruct the altitudinal limits of these zones over time (Fig. 9.21). The chronological framework for the record has been provided by correlation with a marine

Altitudinal Zonation The Andes

• high plain of Bogota /if/xerophytic vegetation -Supper forest line

FIGURE 9.20 Altitudinal distribution of vegetation zones in the eastern Cordillera of Colombia at the present time and during the last glacial maximum (Van der Hammen, 1974, modified by Andriessen et al„ 1993).

• high plain of Bogota /if/xerophytic vegetation -Supper forest line

FIGURE 9.20 Altitudinal distribution of vegetation zones in the eastern Cordillera of Colombia at the present time and during the last glacial maximum (Van der Hammen, 1974, modified by Andriessen et al„ 1993).

isotope record from ODP site Gil (Hooghiemstra et ai, 1993) and by fission track dates on zircons from tephras in the sediments (Andriessen et al., 1993). It has been estimated that the altitudinal extent of the Andean forest can be correlated with threshold values of the overall arboreal pollen (AP) sum recorded at Funza; thus when total AP exceeded 75% (during the last -263 ka) the forest zone extended across the basin to above 3000 m in elevation. When the AP sum fell below 40% (in

100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500,

T 1I

o ce

Arboreal pollen (Funza I pollen record)

100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500,

T 1I

0 0

Post a comment