Quaternary North American Vegetational History: 1.6 Ma to the Present

The Quaternary Period encompasses the Pleistocene and the Holocene or Recent Epochs. The date used for the beginning of the Pleistocene depends upon which globally recognizable event is selected as representing a significant break with the preceding Pliocene Epoch. Candidates include the Gauss-Matuyama magnetopolarity boundary (~2.8 Ma; see Quaternary International, 1997); the initiation of widespread permafrost, a frigid Arctic Ocean, and rapid glaciation in the high northern latitudes (~2.4 Ma; Shackleton and Opdyke, 1977; Shackleton et al., 1984); or the African Olduvai paleomagnetic event between 1.87 and 1.67 Ma. The transition from hothouse to icehouse conditions was gradual, but the Pleistocene is typified at Vrica, Italy, as beginning at ~1.67 Ma (Aguirre and Pasini, 1985; Richmond and Fullerton, 1986; oxygen isotope stage 62), and that is the date used here. In the conterminous United States the Elk Creek till of Nebraska is 2.14 m.y. in age (Hallberg, 1986), and the onset of the full ice age is represented by the onset of repeated glaciations at ~850 Kya when glaciers extended down the Mississippi River Valley. Subsequently, glacial-interglacial conditions fluctuated until the latest retreat at ~11 Kya that began the Holocene or Recent Epoch. The chronology of ice age events began with the publication of Louis Agassiz’s (1840) Etudes surles Glaciers. In the absence of evidence to the contrary, a single glacial advance was envisioned as blanketing the high latitudes. In the 1940s Willard E Libby at the University of Chicago perfected the technique of radiocarbon dating, and Flint and Rubin (1955) applied this methodology of “isotopic clocks” to establishing the absolute chronology of drift deposits from the eastern and midwestern United States. Their radiocarbon dates showed evidence of two or more times of continental-scale glaciations; older organic material was “radiocarbon inert” and beyond the ~40-Ky range of the technique. A standard chronology eventually became established for North America that included four major glacial stages (Nebraskan, oldest; Kansan; Illinoian; and Wisconsin) separated by four interglacials (Aftonian, oldest; Yarmouth, Sangamon, and the present Holocene).

Cause and Effect Factors Influencing the Composition and Distribution of North American Plant Formations through Late Cretaceous and Cenozoic Time

The arrangement of vegetation over the landscape is a product of interactions between the environment, the ecological characteristics of individual organisms, barriers, dispersal potential, epidemic disease, anthropogenic influences, and the partially serendipitous factor of propagule availability. Within the complex of environmental factors, several are of special importance in tracing the history of North American plant communities. They include climate; plate tectonics as a mechanism for orogeny, volcanism, land bridges, and terranes; and catastrophes. Each have numerous interacting subcomponents, feedbacks, and amplifiers, and although constraints of format make it necessary to discuss these separately and sequentially, they are interconnected and pertubation of one affects the entire system. Diagrams summarizing these factors are presented at the end of the following sections. The diagrams are not intended as models for, indeed, the single factor of climate could be expanded into a component so vastly complex that it would be counterproductive to a general summary. Similarly, the hydrological cycle, which involves the largest movement of any substance on Earth, cannot be fully treated because a “systems” view of its role in influencing climate is not available (Chahine, 1992) and the roles of water vapor (a greenhouse gas) and cloud cover are just now being quantified (Cess et al., 1995; Ramanathan et al., 1995). Rather, the diagrams illustrate some of the factors and relationships discussed in the text and serve as a reminder of the complex interactive nature of physical and biotic events. Plants are limited in their ecological amplitude. Several important corollaries follow from this observation; one of the most fundamental is that changes in climate cause extinctions promote evolution, and alter the range and habitats of organisms. Because climate plays a central role in the arrangement of modern communities (Gates, 1993; Kareiva et al., 1993; Woodward, 1987) and in the development and distribution of past assemblages (Brenchley, 1984; Crowley and North, 1991; Hecht, 1985a), reference to some elements of general climatology is necessary for understanding the diversification, radiation, and reshuffling of North American paleocommunities during the Late Cretaceous and Cenozoic.

Middle Miocene through Pliocene North American Vegetational History: 16.3-1.6 Ma

During the Middle Miocene through the Pliocene the Appalachian Mountains underwent continued erosion and approached modern elevations. The Rocky Mountains had undergone uplift to half or more of their present elevation during the Late Cretaceous to Middle Eocene Laramide Revolution; after a lull during the Middle Eocene through the Early Miocene, there was increased tectonic activity beginning ~12 Ma and especially between 7 and 4 Ma. Locally some highlands may have approached or attained modern elevations. The increasingly high mountains and plateaus of Asia and North America deflected the major air streams southward, bringing colder polar air into the middle latitudes of North America. An extensive Antarctic ice sheet further cooled ocean waters and contributed to the spread of seasonally dry climates. The elimination of most of the Asian exotics from the North American flora dates to the Late Miocene-Pliocene as a result of a decline in summer rainfall. The Sierra Nevada attained about two-thirds of their present elevation within the past 10 Ma. They were appreciably elevated at ~5 Ma, stood at ~2100 m at 3 Ma, and have risen ~950 m since 3 Ma (Huber, 1981). The California Coast Ranges and Cascade Mountains attained significant heights by 3 Ma, and there was a rapid rise of the Alaska Range at ~6 Ma. Temperatures increased between ~18 and 16 Ma. In the absence of major plate reorganization and intense volcanic activity and with increased erosion from continued replacement of the dense evergreen forest by deciduous forest and shrubland (increasing albedo), atmospheric CO2 concentration decreased and a sharp lowering of temperature occurred in the Middle Miocene between 15 and 10 Ma. Eolian dust deposits increased in the Late Cenozoic, suggesting greater aridity (Rea et al., 1985). This is supported by kaolinite records from North Atlantic deep sea sediments (Chamley, 1979). At ~4.8~4.9 Ma global cooling and a marine regression of ~40~50 m combined to isolate the Mediterranean Basin from the ocean and to concentrate large volumes of salt as water evaporated. The biota was destroyed, giving rise to the term Messinian salinity crisis.

Methods, Principles, Strengths, and Limitations

Methods of paleovegetation analysis can be grouped into two broad categories. Those that use plant microfossils for reconstructing terrestrial vegetation, past environments, migrations, and evolutionary histories constitute a part of paleopalynology that includes the study of pollen, spores, other acid-resistant microscopic structures, and phytoliths (distinctive, microscopic silicate particles produced by vascular plants). Those that use plant megafossils such as leaves, cuticles, cones, flowers, fruits, and seeds constitute paleobotany. Two important subdisciplines of paleobotany are dendrochronology (fossil woods) and analysis of packrat middens. The latter are sequences of nesting materials, constructed by packrats of the genus Neotoma, preserved in arid environments of the American southwest. The study of fossil fruits and seeds is a specialized field within paleobotany, and it is also used in studies on Quaternary vegetational history in the preparation of seed diagrams accompanying pollen and spore profiles from bog and lake sequences. In 1916 Swedish geologist Lennart von Post demonstrated that pollen grains and spores were abundantly preserved in Quaternary peat deposits and could be used to trace recent forest history and climatic change (Davis and Faegri, 1967). The term palynology was subsequently introduced by Hyde and Williams in 1944 to include all studies concerned with pollen and spores. Paleopalynology has come to denote the study of acid-resistant microfossils generally, while pollen analysis designates those investigations dealing specifically with the Quaternary. In the early 1950s researchers in the petroleum industry began to routinely apply paleopalynology to problems of stratigraphic correlation and the reconstruction of depositional environments in Tertiary and older strata (Hoffmeister, 1959). This added a practical dimension to a mostly academic pursuit and fostered interest in applied palynology and its use as a paleoecological research tool. This important development is reflected in the increased number of publications after about 1955. As the history of other innovations might predict, there was a period of exuberant claims, isolated specialization, and exaggerated charges of deficiency in the method; but for palynology this seemingly inevitable period was mercifully brief. The different terminology, principles, and techniques involved in megafossil paleobotany and paleopalynology still result in specialization, but this limitation is frequently overcome by coordinated or collaborative projects, and an increasing number of practitioners work in both disciplines.

Origins of North American Biogeographic Affinities

An aspect of plant distribution that has intrigued biogeographers for over 200 years is the occurrence of similar biotas in widely separated regions. The North American flora has affinities with several such areas: the Mediterranean, the dry regions of South America, eastern Asia, and eastern Mexico. The origin of some patterns is relatively clear, while for others hypotheses are just now being formulated. During times when the dogma of permanence of continents and ocean basins held sway, explanations for these disjunctions required imaginative thinking that often bordered on the bizarre. The pendulum or schwingpolen hypothesis was offered to explain the perceived bipolar distribution of several taxa (Gnetum, Magnolia, Pinus section Taeda; Simroth, 1914). By this view, the Earth swings in space like a pendulum, creating regular fluctuations in environments and often causing the symmetrical placement of taxa at two points on opposite sides of the Earth. Other disjunctions were explained by casually placing geophysically impossible land bridges at any point in time between any two sites where the presence of similar communities seemed to call for land connections (see review in Simpson, 1943). The presence of teeth of Hipparion, an ungulate related to the horse, in Europe and South Carolina-Florida prompted French geologist Leonce Joleaud to propose a land bridge extending from Florida through the Antilles to North Africa and Spain. Subsequently, to accommodate eight new passengers, it was broadened to encompass the entire region from Maryland and Brazil across to France and Morocco and its life was prolonged to include virtually all of the Tertiary. With the later discovery that there were periodicities in similarity between Old World and New World Cenozoic faunas, the continents were envisioned as moving back and forth like an accordion. George Gaylord Simpson, who favored the North Atlantic land bridge to connect North America and Europe, was beside himself with these theories and characterized Joelaud’s as “the climax of all drift theories.” The bridge became well established in the literature even though it never existed in the Atlantic Ocean (Marvin, 1973). Udvardy (1969) plotted all the Cretaceous and Tertiary land bridges postulated for the South Pacific up to 1913.

Context

The interaction between vegetation and the environment over time is one of the most complex of the Earth’s integrated systems. In addition to the direct methodologies of paleopalynology and paleobotany, there are other techniques that provide independent sources of information for interpreting this interaction. These include paleotemperature analysis, sea-level changes, and faunal history. The first two are also forcing mechanisms as discussed in Chapter 2, but for this survey the summary curves can also serve as convenient context information. Each is a vast subject with an extensive literature, and all are presently generating considerable discussion. For paleotemperature analysis, unsettled issues include the extent of temperature change in equatorial waters during the Early and Middle Tertiary, which would affect the poleward transport of heat by conveyer-belt mechanisms. Estimates range from surface waters as warm or warmer than the present to considerably cooler. For the Neogene, CLIMAP estimates based on the ecology of coccolithophores, diatoms, radiolarians, and especially foraminifera are that temperatures in the tropics did not cool significantly; modeling results, terrestrial paleontological evidence, and new Barbados coral data suggest they cooled by ~5°C. There is uncertainty as to when glaciations began on Antarctica; recent estimates range from the Early Eocene to late Middle Eocene to Middle Oligocene (45-35 Ma; Birkenmajer, 1990; Leg 119 Shipboard Scientific Party, 1988). This affects interpretation of 18O values during the Paleogene because they could reflect temperature alone or could be due to ocean water temperature and ice volume changes. Another challenge is to unravel the extent to which benthic temperature records track insolation-induced changes in water temperature versus new thresholds in ocean bottom-water circulation. Discussions of sea-level fluctuation are presently focused on their causes during the preglacial Early Cenozoic. In faunal history the timing of the North American Land Mammal Ages (NALMAs) or provincial ages are being revised. For vegetational history much of the older literature describes events in terms of geofloras, but this conceptual context, at minimum, requires substantial renovation, and the boreotropical hypothesis is emerging as an alternative for envisioning biotic events in the high northern latitudes.

Setting the Goal: Modern Vegetation of North America Composition and Arrangement of Principal Plant Formations

Vegetation is the plant cover of a region, which usually refers to the potential natural vegetation prior to any intensive human disturbance. The description of vegetation for an extensive area involves the recognition and characterization of units called formations, which are named with reference to composition (e.g., coniferous), aspect of habit (deciduous), distribution (western North America), and climate, either directly (tropical) or indirectly (tundra). Further subdivisions are termed associations or series, such as the beech-maple association or series within the deciduous forest formation. Formations and associations constitute a convenient organizational framework for considering the development of vegetation through Late Cretaceous and Cenozoic time. For this purpose seven extant plant formations are recognized for North America: (1) tundra, (2) coniferous forest, (3) deciduous forest, (4) grassland, (5) shrubland/chaparral- woodland- savanna, (6) desert, and (7) elements of a tropical formation. Several summaries are available for the modern vegetation of North America, including Barbour and Billings (1988), Barbour and Christensen, Kuchler (1964), and Vankat (1979). The following discussions are based primarily on these surveys. Tundra (Fig. 1.2) is a treeless vegetation dominated by shrubs and herbs, and it is characteristic of the cold climates of polar regions (Arctic tundra) and high-altitude regions (alpine tundra). In the Arctic tundra a few isolated trees or small stands may occur locally, such as Picea glauca (white spruce), but these are always in protected habitats. The Arctic region experiences nearly continuous darkness in midwinter, and nearly continuous daylight in midsummer. There is a short growing season of only 6-24 weeks; this accounts, in part, for the fact that 98% of all Arctic tundra plants are perennials (Vankat, 1979). Strong winds are another feature of the Arctic landscape, often exceeding 65 km/h for 24 h or more. They likely account for the frequency of rosettes, persistent dead leaves, and the cushion growth form, in the center of which wind velocities may be reduced by 90%. The harsh growing conditions also result in leaves of the microphyllous size class being comparable to those of desert plants. Vegetative reproduction and self-pollination is common, and phenotypic plasticity is high among Arctic tundra plants.

Middle Eocene through Early Miocene North American Vegetational History: 50-16.3 Ma

During the Middle Eocene through the Early Miocene, erosion of the Appalachian Mountains exceeded uplift and there was a net reduction in elevation. In the Rocky Mountains uplift continued through the Middle Eocene (end of the Laramide orogeny), waned in the Middle Tertiary, and then increased beginning at about 10 Ma. Earlier reconstructions placed paleoelevations in the Rocky Mountains during the Middle Eocene through the Early Miocene at approximately half the present relief. The maximum elevation in the Front Ranges during the latest Eocene was estimated at ~2500 m (~8000 ft; MacGinitie, 1953). Recent approximations are for nearly modern elevations in several areas by the Eocene-Oligocene. Extensive Eocene volcanism deposited ash and blocked drainage systems, augmenting uplift and facilitating the preservation of extensive fossil floras and faunas. In the far west the beginning of Tertiary volcanism in the Sierra Nevada is dated at ~ 33 Ma near the Eocene-Oligocene boundary. A drying trend becomes evident in the Middle Eocene and reduced moisture, along with the waning of volcanic activity in the Oligocene, restricted conditions favorable to fossilization. The number of Oligocene floras in the northern Rocky Mountains is considerably fewer than in younger deposits to the west. In the absence of extensive plate reorganization and orogeny, CO2 concentration decreased, which contributed to a temperature decline that continued through the Cenozoic and intensified in the Late Tertiary. Recall from Chapter 2 (sections on orogeny and volcanism) that uplift plays a role in determining long-term climate by creating rainshadows, altering atmospheric circulation patterns, and increasing the erosion of silicate rocks that causes a drawdown of CO2. This allows heat to escape from the troposphere and results in lower temperatures. Marine benthic temperatures were ~10°C in the early Late Eocene and ~2°C near the Eocene-Oligocene boundary, assuming an essentially ice-free Earth during that time, and increased to ~5-6°Cnear the end of the Early Miocene. Temperatures over land in the midnorthern latitudes are estimated to have dropped by ~12°C between the Late Eocene and Early Oligocene (Wolfe, 1992a).

Late Cretaceous through Early Eocene North American Vegetational History: 70-50 Ma

At the end of the Cretaceous the Appalachian Mountains had undergone 180 m.y. of erosion since their principal uplift in the Middle Pennsylvanian through the Late Permian (300-250 Ma), but they were higher and more rugged than at present and provided a somewhat more diverse vertically zoned array of habitats. In contrast, the Rocky Mountains were only ~1 km above sea level at 65 Ma; the Coast Ranges, Sierra Nevada, and Cascade Mountains would not attain substantial heights until late in the Tertiary. Computer models in the NM mode, which simulate conditions in western North America in the Late Cretaceous, show a nearly continuous westerly jet stream with relatively small amplitude between the troughs (lowpressure systems) and ridges (high-pressure systems). The present north-south seasonal meandering of the jet stream was also less. Thus, in the models precipitation and temperatures were more uniform throughout the year and there was less regional differentiation in climate. CO2 concentrations during the Cretaceous are estimated to have been 4-8 times to 10-12 times higher than at present. With a 2-5°C warming for each doubling of CO2, this provides part of the explanation for the higher MAT documented for the Late Cretaceous and Early Tertiary. High CO2 concentrations near the end of the Cretaceous may have been a holdover from earlier intense volcanism in the South Pacific that began to subside at ~ 100 Ma. The term epeirogeny refers to vertical motions of the Earth’s crust, and these movements affect the ocean floor, as well as the continents. There was a 50% increase in the production of ocean crust in the Middle Cretaceous compared to earlier times, as represented by the early Aptian Ontong Java Plateau, the Earth’s largest oceanic plateau, now submerged over 2 km off the Solomon Islands. A sense of the magnitude of this structure can be gained by comparing its volume with that of the surface-exposed Deccan Traps of India (66 Ma). The latter are ~1 km thick and have a volume of 1.5 x 106 km3. The Ontong Java Plateau is ~36 km thick and has a volume of 50 x 106 km3.