When the Wisconsin ice sheet stood at its maximum position, tundra vegetation bordered the ice sheet. In the eastern United States, tundra extended at least 300 km due south of the ice border at 2700 ft (800 m) elevation on the Allegheny plateau. Spruce and jack (and/or red) pine forest grew at lower elevations in Virginia. On the coastal plain, and farther south, in the piedmont of northern Georgia, jack pine dominated the forest vegetation over a large region.
As the ice sheet receded, the vegetation underwent a series of changes. Coniferous forest was replaced by deciduous forest, beginning 13,600 B.P. in Georgia. The frequency of white pine began to increase in Virginia at about the same time, and the frequencies of deciduous trees, about 1000 yr later. On the Allegheny plateau, no change took place in the tundra vegetation until 12,700 B.P., when tundra was replaced by open, spruce woodland. Jack and/or red pine grew mixed with, or nearby, the spruce. Pollen from deciduous trees (mainly oak, ash, and hornbeam) reached the site in greater quantity than before. Possibly the increase indicates a change in prevailing wind direction.
On the Allegheny plateau, 10,500 years ago, the boreal woodland was replaced by a mixed coniferus-deciduous forest which included white pine. At about the same time (or perhaps a thousand years later), a similar change occurred in Connecticut. At lower elevations in the Shenandoah Valley, spruce forests including white pine were replaced by oak and other hardwoods.
In the early Holocene, at a time we unfortunately were not able to pinpoint by radiocarbon dating, deciduous forest began to grow on the Allegheny plateau. Later there was a series of changes in the composition of the forest. High frequencies of oak pollen occur throughout the sequence, with successive maxima of hemlock, beech, and finally, hickory. High percentages of chestnut pollen occur with a maximum approximately coincident with the maximum of beech. These changes are probably significant both from stratigraphic and paleoecologic points of view, and should be studied in greater detail at sites where radiocarbon dating will be possible. The early maximum of chestnut pollen is a major difference between the pollen sequence in the Alleghenies and southern and central New England, suggesting that this species was very slow to move northward, arriving in New England just 2000 B.P. as the result of migration, not climatic change.