CHAPTER 1
Archaeology, Paleoecosystems, and Ecological Restoration
Michael J. O'Brien
Archaeologists have long been interested in the role the physical environment played in structuring the activities of prehistoric human groups. More recently, research has begun to shift toward the role of humans as active participants in paleoecosystems. This move toward paleoecology has not signaled an abandonment of traditional research questions having to do with humans and their environment—why groups lived in certain locales and not others; why they selected certain foods over others; and why they chose particular raw materials for their clothing, tools, and weapons. It has, however, signaled an end to the centuries-old belief in environmental determinism.
Despite this trend, paleoecological research is still an undercurrent in terms of actual archaeological practice. Although modern archaeologists are more familiar with paleoecology than their predecessors were, today's "ecologically oriented" archaeological applications are often based on methods more than on anything else. With a few exceptions—for example, the earlier work of Kent Flannery (1968) on systems theory and, later, the work of human evolutionary ecologists (Bettinger 1991; Kelly 1995)—theoretical work on such topics as grain response (MacArthur and Pianka 1966; Pianka 1974), patchiness (Wiens 1976), and central-place foraging (Orians and Pearson 1979) have not had the kind of impact in archaeology that one might have hoped.
One reason archaeologists fail to connect their studies of paleoenvironments to paleoecology is because of the difficulty inherent in reconstructing historic settings. Archaeologists are aware that attempts to understand both the paleoenvironment of a specific locality or region and how that environment influenced the actions of organisms living in it are difficult endeavors because environments change, and the rate and magnitude of change are never constant. They realize that, in the end, the best we can hope for are environmental snapshots taken at different times. If the data permit, we can stack those snapshots to create a moving picture, but we have to be honest about what it is that we have created—a jerky composite consisting of a set of still photographs taken at arbitrary points along a continuum. We might speak colloquially about having "reconstructed" paleoenvironments, but we need to keep in mind that we haven't reconstructed anything. What we have done is to model paleoenvironments using available evidence, and, as we all know, models are merely educated guesses about how things might have been. On the positive side, the more lines of corroborating evidence there are, the more refined the model will be.
My goal in this chapter is threefold: (1) to present a brief overview of how archaeologists have adopted changing perspectives on human-environment interactions; (2) to discuss the sources of environmental information of which archaeologists routinely make use; and (3) to indicate some of the ways in which theoretical issues can be integrated into paleoecological research.
Human-Environment Interactions
Despite an early recognition of the importance of the environment and its effects on humans, archaeologists were for the most part slow to recognize the intricacies of human-environment interactions. Throughout the late nineteenth century and well into the twentieth century, most archaeologists were environmental determinists, viewing human actions simply as automatic responses to what the environment allowed (Wissler 1926). At an elementary level, the environment is deterministic—you can't, after all, grow corn at the North Pole—but to suggest that a particular environment triggers automatic, predictable responses on the part of any animal, particularly humans, is short-sighted. All organisms modify their environments—some in more dramatic fashion than others—and most enjoy a kind of phenotypic plasticity that allows them to escape sudden and automatic extinction in the face of environmental change.
By the 1950s, a significant portion of the discipline of archaeology had pushed aside the doctrine of environmental determinism of earlier decades. Archaeologists began leaning heavily on one of two sets of ideas: (1) environmental "possibilism"—a kind of indeterministic process in which the physical environment was seen not so much as determining cultural outcomes but as setting limits on cultural development; or (2) the cultural ecology of anthropologist Julian Steward (1949, 1955), which held that certain relational links between technology and environment, regardless of geographic locale, created similar cultural outcomes.
Included in the list of topics addressed in archaeological reports of the 1950s were the physiographic and geological setting of the locale, hydrologic resources, vegetation, climate, soils, and often the fauna native to the locality under investigation. However, those reports failed to consider both the changes that had taken place in the physical environment and any concomitant changes in human populations due to those changes. Lack of attention to those changes meant that the environment was really being viewed as a static backdrop against which human interactions were carried out. The lack of integration between archaeology and paleoecology was one of archaeologist Walter Taylor's major complaints in his 1948 summary of the state of the discipline. His (1957) report to the National Research Council's Committee on Archaeological Identification called for closer coordination of work between archaeologists and paleoenvironmentalists. The 1960s witnessed phenomenal growth in large-scale archaeological survey-and-excavation projects in the Near East, a significant number of which were directed by Robert J. Braidwood, Robert McCormick Adams, and their students (Adams 1962, 1965; Braidwood 1958, 1960; Flannery 1965; Hole 1966; Hole, Flannery, and Neely 1969; Kraelin and Adams 1960). These interdisciplinary archaeological studies, which focused on the physical environment and its role in the development of settled life, agriculture, and eventually urban areas, led to a series of conceptual changes in how archaeologists approached the topic of human-environment interactions. They spawned a whole new generation of similar studies in semiarid regions of the New World, such as Highland Mexico (Byers 1967; Flannery 1966, 1968; MacNeish 1964; MacNeish, Nelken-Terner, and Johnson 1967; MacNeish, Peterson, and Flannery 1970; MacNeish, Peterson, and Neely 1975). These studies became models both for later archaeological projects carried out in the United States (O'Brien, Warren, and Lewarch 1982) and for the currently emerging discipline of historical ecology (Crumley 1994; Balée 1998).
Sources of Environmental Information
One of the obvious problems facing archaeologists interested in understanding how aspects of the physical environment have changed through time is locating the data necessary to pinpoint the changes. The basic mechanics of landscape evolution became well known by the middle of the nineteenth century, at which time it became evident that the North American continent had undergone successive climatic changes, including periodic "ice ages." One of the burning issues of nineteenth-century prehistory was whether Ice Age humans had been present to hunt the megafauna, especially mastodons and mammoths, that amateur prehistorians had recovered from mid-continental bogs (O'Brien 1996). By the late nineteenth century the issue had been expanded as archaeologists attempted to determine whether there had been a North American Paleolithic period (Meltzer 1983, 1985) similar to that in Europe.
Answers to many of these questions came with the development of radiocarbon dating in the late 1940s (Marlowe 1999). We often think of radiocarbon dating as being solely the province of archaeologists, but its development was also a boon to paleoclimatologists, who could then date their pollen sequences and, like archaeologists, begin to focus on finer and finer units of time. It no longer was enough to place a terminal date on the last glacial episode and then to lump the last ten thousand or so years into a single unit. Through the use of radiocarbon dating, it became clear that the Holocene had witnessed considerable climatic variation and that if the proper data were available, the period could be subdivided into finer units of time.
Archaeologists quite naturally were interested in such research because the data would allow them to understand something of the past environments in which their subjects had lived. Some of the earliest work on paleoenvironments of the terminal Pleistocene and early Holocene (ca. 30,000-7,000 B.C.) was done in the western United States and the Mississippi River alluvial valley. Research done in various parts of the West, including the Rocky Mountains (Matthes 1951), the Great Basin (Antevs 1948; Heizer 1951), the Southwest (Bryan 1950; Sayles and Antevs 1941), and the Great Plains (Antevs 1950; Moss 1951; Schultz, Lueninghoener, and Frankforter 1951), incorporated a variety of data such as rates of varve buildup (Antevs 1950), cross-correlation of terrace sequences (Moss 1951), sediment analysis (Heizer 1951; Sayles and Antevs 1941; Schultz, Lueninghoener, and Frankforter 1951), the positioning of glacial moraines (Antevs 1950; Matthes 1951), and floral analysis (Schultz, Lueninghoener, and Frankforter 1951). Many of the paleoclimatic interpretations that grew out of this early research are now outdated, but what is important is that by 1950 archaeologists had begun to team up with geomorphologists and paleoclimatologists in an effort to document the myriad landscape and climatic changes that various localities had witnessed during the terminal Pleistocene and the Holocene.
Farther east, most of the work carried out in the Mississippi River alluvial valley—that portion of the valley from near Cairo, Illinois, to the Gulf of Mexico—was by Harold Fisk (1944) and those influenced by him (Saucier 1964, 1968). Fisk realized that the Mississippi Valley contained a chronicle of geomorphological responses to changing Tertiary and Quaternary climatic episodes that had affected the North American continent. His goal was to reconstruct the history of landscape development, especially during the terminal Pleistocene and Holocene epochs.
Archaeologists (Ford, Phillips, and Haag 1955; Phillips, Ford, and Griffin 1951) viewed Fisk's development of a relative chronology as a boon to their own work because it gave them another way to date their study sites. Although subsequent work, primarily by Roger Saucier (1974, 1968, 1981) and P. D. Royall and his colleagues, Paul and Hazel Delcourt (1991), has shown Fisk's chronology to be grossly inaccurate and flawed in terms of his postulations about the magnitude of certain processes, the importance of his work is unquestioned. He was the first researcher to tie the landscape evolution of the alluvial valley to climatic change and to the corresponding rates of water and sediment discharge in the upper reaches of the Mississippi system.
The region in which Fisk was working—an active floodplain setting in humid, mid-continental North America—was not comparable to western North America. In fact, many of the problems that Fisk encountered in his attempt to document the evolution of the Mississippi alluvial valley would have been foreign to researchers such as Ernst Antevs (1948, 1950), who was doing the same kind of work but in more stable environments. Thus it is quite understandable that the majority of paleoenvironmental research conducted in the 1940s and 1950s took place in the West (Bryan 1950; Heizer 1951; Matthes 1951; Moss 1951; Schultz, Lueninghoener, and Frankforter 1951).
Sustained interest in the Holocene climate and its relation to other aspects of the midwestern physical environment dates at least to the 1950s. Archaeologists viewed the upper Midwest, in particular, as an excellent laboratory for the study of environmental change because of its unique vegetational composition—a mixture of forest and prairie—with the specific makeup depending on location within the region. Stretching eastward from the Great Plains to central Indiana is a complex mosaic of tallgrass prairie and deciduous forest known as the Prairie Peninsula (Transeau 1935; Kuchler 1964, 1972) (figure 1.1). The peninsula of interdigitating fingers of prairie upland and forested river valleys formed early in the Holocene and throughout its history has been vulnerable to climatic change. This vulnerability shows up in the pollen record and in such things as soil composition, valley-fill sequences, and archaeological site locations.
In 1955, W. H. Horr completed the first boreal pollen record in the mid-continent at Muscotah Marsh in northeastern Kansas (see also Wells [1970], who identified the pollen as being from spruce). The zone containing the boreal pollen was subsequently dated to 13,000 ± 1500 B.C. Here was evidence that during the close of the Pleistocene, spruce forests had grown where now there were tallgrass prairies. By the 1960s, pollen studies from other parts of the Midwest supported the long-held proposition that boreal forests had been far south of their current latitudes during the terminal Pleistocene and that, as these forests retreated northward with a rise in temperature, they were replaced in some areas by hardwood forests and in other areas by prairies. What happened next—that is, how climate changed throughout the Holocene—was open to question.
In 1960, James B. Griffin advanced the notion that climatic change contributed to the growth and decline of some northern prehistoric cultures through its effects on their crop cycles (Griffin 1960). The importance of that paper was not whether Griffin was correct (we now know that the situation was much more complex than he suspected) but that he tied the disappearance of a major Native American culture, termed Hopewellian by archaeologists and dating roughly A.D. 1-200, to climatic change.
This examination of the nature of the relation between climate change and human settlement and subsistence practices was the basis for a long-term study of late-ceramic-period cultures in the upper Midwest. It was titled "Climate, Ecology, and the Oneota Culture" and was done under the direction of archaeologist David Baerreis of the University of Wisconsin-Madison. That National Science Foundation-supported program involved intensive study of archaeological materials from Oneota and Mill Creek culture sites in Iowa, Missouri, Minnesota, extreme southeastern Nebraska, and western Wisconsin. It also involved studies of paleoclimatic indicators such as plant and animal remains, snails, and pollen. Baerreis's collaboration with climatologist Reid Bryson of the University of Wisconsin-Madison produced a series of reports and monographs (Baerreis and Bryson 1965, 1967; Bryson 1966; Bryson, Baerreis, and Wendland 1970; Henning 1970) that modeled Holocene climate in the upper Midwest and examined the archaeological record, especially the late portion, in terms of the documented changes in climate (see also Bryson [1966] and Bryson and Wendland [1967]). The site-specific analyses built on work that was more regionally extensive (Borchert 1950; Deevey and Flint 1957; Wright 1968) and produced a concise set of data that would repeatedly be incorporated into later work in the Midwest (Baerreis, Bryson, and Kutzbach 1976; Webb and Bryson 1972; Wendland 1978; Wendland and Bryson 1974).
As a result of numerous palynological and sedimentological analyses, paleoclimatologists (Bryson, Baerreis, and Wendland 1970; Bryson and Wendland 1967; Wendland 1978, 1995; Wendland and Bryson 1974) isolated ten postglacial climatic episodes, each of which affected portions of the Midwest, though the effects often differed substantially from area to area. Four episodes are of concern here: the pre-Boreal, dating between about 8050 B.C. and 7550 B.C., the Boreal (7550—6550 B.C.), the Atlantic (6550-3050 B.C.), and the sub-Boreal (3050—850 B.C.). These are shown in figure 1.2 along with generalized climatic conditions, vegetational changes, and cultural periods. Some of the general climatic trends that characterized these episodes can be extrapolated to regions outside the Midwest, but the magnitude of the trends is uncertain.
