In this chapter, we discuss climatic events that marked the onset of the recent ice age, their impact on biotas, and their likely causes. Our focus is on the North Atlantic region because this part of the world, which became bordered in the north by major ice caps, was the scene of more severe environmental and biotic changes than occurred in other areas of the globe.
The detailed chronology for major glacial events of the past 3 million years (m.y.) or so has come primarily from deep-sea deposits. Many deep-sea cores provide relatively continuous records, for which magnetic reversals and biostratigraphic data yield key dates. Also especially useful are changes in the isotopic composition of microfossils; the proportion of 18O increased in these forms at times of glacial maxima, both because isotopic partitioning during skeletal secretion varies with temperature and because 16O is preferentially evaporated from the oceans, transported in water vapor, and sequestered in glacial ice.
Fossils in shallow marine and terrestrial sequences also record climatic changes of the recent ice age. Some biotic changes represent clear evidence of climatic transitions, but others, which will be discussed separately, can only be interpreted a posteriori as reflecting these environmental changes.
Historically, a great variety of hypotheses have been invoked to explain the onset of Plio-Pleistocene glaciation in the Northern Hemisphere. Although the issue is complex, forcing factors resulting from tectonic events were probably responsible. We also review ways in which ice sheets have been influenced by periodicity in the Earth's orbital motion and have themselves behaved as climatic forcing factors in a complex feedback system.
The climatic changes that have occurred in the North Atlantic region since the ice age began about 2.5 Ma can be understood only in the context of events that were under way millions of years earlier. The climatic changes of the past 2.5 m.y. have been cyclical, relatively rapid, and associated with orbital-scale variations in ice volume in the Northern Hemisphere. In contrast, the major changes prior to 2.5 Ma represented net trends that were relatively gradual. For the most part, they resulted from tectonic events that are discussed in the final section of this chapter.
In the present section, we first evaluate climatic events prior to the Late Pliocene, which began at 3.4 Ma; next, events at 3.2-3.1 Ma that preceded the start of the ice age; and finally, events that marked the development of vast ice sheets close to 2.5 Ma.
Climatic responses to tectonic events of Miocene and Early Pliocene time were regionally complex. Although in general these climatic trends developed gradually over many millions of years, they were at times interrupted by more dramatic ''steps," or brief intervals when critical thresholds in the system were exceeded and large-amplitude responses were triggered. Both the slow climatic drift and the steps altered the distributions of plants and animals, and may also have affected their evolution. Some marine records appear to be relatively continuous, but the discontinuous nature of most continental sedimentation precludes fully adequate resolution of many long-term climatic trends. The basic patterns that can be detected involve widespread cooling, especially at high and middle latitudes, and a mosaic of more regional trends toward wetter and drier climates.
The cooling trend is particularly evident at higher latitudes and elevations. Thick ice appeared on Antarctica early in the Cenozoic (Barron et al., 1989). Deposits from small mountain glaciers are first recorded between 10 and 5 Ma in the Coast Range of Alaska (Denton and Armstrong, 1969) and the Andes of South America (Mercer, 1983). Pollen data from high-altitude sites in Iceland indicate significant cooling by 10 Ma (Mudie and Helgason, 1983). Traces of ice-rafted sand in Norwegian Sea sediments by 4 Ma (Henrich et al., 1989) suggest at least the sporadic presence of mountain glaciers along the east coast of Greenland or the west coast of Scandinavia at this time.
Another major late Cenozoic trend was a progression toward more highly differentiated regional extremes of wet and dry climate. During the past 15 m.y., deserts have formed or expanded into new terrain in Asia (Wolfe, 1979), North Africa (Tiedemann et al., 1989), and North America (Axelrod, 1950), while monsoonal climates have persisted or intensified in the Indo-Asian subtropics and South American tropics.
Several regions of west-central North America became markedly drier during the late Cenozoic. On the Northern Plains, prairie savanna gave way to grasses and herbs after 15 Ma (Thomasson, 1979). During the same interval, vegetation adapted to summer drought gradually came to dominate the California coast (Axelrod, 1966).
Most of the profound climatic changes in the Northern Hemisphere during the Pliocene Epoch prior to about 3.1 Ma were regional in scale. Forest yielded to scrub vegetation in the rain shadow of the Cascades around 4 Ma, for example (Leopold and Denton, 1987), and desert vegetation expanded in the Great Basin near 4 to 3 Ma (Axelrod,