Climate and Human Migrations

See allHide authors and affiliations

Science  25 Oct 2002:
Vol. 298, Issue 5594, pp. 764-765
DOI: 10.1126/science.1078163

Archaeological records are affected by a variety of natural and cultural processes at a variety of spatial and temporal scales (1). A given cultural phenomenon may appear across a broad range of environments, or may be limited to a narrow range of environments and time periods. Paleoecological studies can help to discriminate between these cases. But most reconstructions of early human ecosystems are based on the excavation and interpretation of individual archaeological sites. Paleoecological studies of long-term climatic change are also often limited in scope (2).

Integrative studies of multiple sites, multiple records, and larger areas over long time periods can dramatically change the interpretation (37). On page 821 of this issue, Núñez et al. (8) demonstrate the power of such a comprehensive approach. They closely integrate paleoecological and archaeological analysis to study the long-term interaction between hunter-gatherers and changing environments over the last 15,000 years in the Atacama desert of northern Chile.

The authors examine why initial human occupation occurred about 2000 years later in this hyperarid region than in more humid forested regions in south central Chile (9), and several centuries later than in less arid areas in the central and southern Andes. They also ask why a long “Silencio Arqueologico” (a cultural hiatus in the archaeological record) took place between 9500 and 4500 calendar years before the present (cal yr B.P.).

The possible reasons for these variations in human presence considered by Núñez et al. include migration lags, inhospitable late Pleistocene environments, biased survey and visibility, and rapid and long-term abandonment of the region. The study illustrates the importance of integrating local environmental and archaeological information in studying regional human ecosystems and in comparing the findings with other regions at a larger scale.

The authors assume that high-altitude ancient lakes (paleolakes), mid-altitude grasslands (puna), and low-altitude wetlands best indicate changes in habitat exploitation due to regional climatic change. Although they also examined open campsites, they concentrated their excavations on caves and rock shelters at different elevations in these three settings. Caves and rock shelters usually contain evidence of long-term cycles of human occupation and abandonment.

Archaeological and environmental zones (13) in northern Perú.

Patterns of flux in early archaeological records in this region are due primarily to climatic change but also to technological adaptability, resource availability, and social organization (14). People were always present in the area from at least 11,000 cal yr B.P., but during periods of increased aridity they primarily occupied spring-fed wetlands at different elevations between the coastal plains and foothills. Zones 2 and 6 were inhabited only between ∼11,000 and 9000 cal yr B.P. Zones 1, 4, and 5 reveal human activity between 11,000 and 5000 cal yr B.P. Zone 3 was occupied primarily between 11,000 and 7000 cal yr B.P. Since the late Holocene, all areas have been inhabited intermittently.

Núñez et al. show that the reason for human occupation fluctuations and the Silencios Arqueologicos was climatic change. When vegetation and animal life were abundant, mobile hunters moved into higher elevations between 11,800 and 10,500 cal yr B.P. They practiced seasonal migrations between the higher paleolakes and lower wetlands between 9900 and 8800 cal yr B.P. Around 9000 cal yr B.P., the paleolakes dried up and were abandoned until about 4500 cal yr B.P., when human reoccupation occurred along the shorelines of reduced lake levels. At this time, movement between paleolakes and wetlands was reestablished.

The findings lend support to the hypothesis that some of the first people in the Atacama region focused on favorable, humid habitats such as lakes, springs, and streams during periods of decreased aridity. During periods of increased aridity, people were absent at the higher elevations, especially during the long Silencio from 9000 to 4500 cal yr B.P. Núñez et al. point out that in more humid conditions such as central Chile (10) and the central highlands of Peru, the Silencio does not exist or is reduced in time and space to “micro-Silencios.”

Around 14,600 cal yr B.P., when people occupied the rainforests of south central Chile, there is no evidence of human occupation anywhere in the Atacama region. When favorable climatic conditions and biotic regimes were established between 10,000 and 9000 cal yr B.P., people colonized these previously inhospitable environments within a few hundred years, indicating that they lived nearby and perhaps occasionally probed and explored the edges of the region.

Late Pleistocene archaeological site.

In the foreground is a dense scatter of stone artifacts and hearths of a campsite dating around 10,300 cal yr B.P. The site is located 8 km from and 800 m above the fog-covered valley floor of the Zaña River in the background. Although no late Pleistocene sites appear on the valley floor, they are present near springs in the distant hills of zone 3.

Such exploratory behavior may account for the extremely low human activity observed by Núñez et al. in some environments during periods of extreme climatic stress. Although fluctuations in aridity levels appear to be the primary factor influencing continuities and hiatuses in the human presence, the authors also recognize that sampling bias and visibility may account for occupational fluxes in certain places at certain times.

The study raises other questions with important implications. It is clear that for climatic, social, and other reasons, people disappear from the archaeological record from time to time (see the figures). Patterns of presence and absence need to be factored in when making claims about early human migration and in extrapolating local site-specific findings from one region to another. If other early human ecosystems in the Americas varied across time and space as they did in the Atacama, then initial human colonization could not have been a blitzkrieg movement (11, 12) but was likely a stutter-step, characterized by hesitancy followed by rapid transience through or around inhospitable environments and slow migration through hospitable ones.

The report by Núñez et al. illustrates that our understanding of past ecosystems and early human migration and strategies of land use is enhanced by asking specific questions about specific habitat changes during reduced time periods. By going beyond the comparison of local records and performing integrative research, we can shed light on the relation between human societies and climate change.


View Abstract

Stay Connected to Science

Navigate This Article