Low coral atolls are the most precarious landscapes for human habitation since they have limited land area, humus-poor, undeveloped coralline soils and potable subterranean water restricted to the largest islets of an atoll. Since most atolls are scarcely more than 2 or 3 m above sea level and are frequently inundated by surge and seasonal typhoons, the long-term prehistoric occupation of atolls is all the more amazing (Figure 1, right).
Since 1993, a multi-disciplinary project has been examining the human colonization of these small, dispersed landmasses, the cultural and economic adaptations to these unique islands and the historical transformations of atoll societies.
)
Figure 1: The northern end of Maloelap Atoll. Extensive excavations were conducted at a large habitation site and horticultural system on Kaven, the largest islet of the atoll.
)
Figure 2: Maloelap Atoll, the seventh largest of the Marshall Islands, whose 71 islets have a combined land area of just 9·8 sq. km.
Situated just north of the equator and about 4000 km southwest of Hawai'i, the Marshall Islands consist of 29 atolls spread over nearly 1 million sq. km of ocean in a roughly linear arrangement traversing a rainfall gradient from 1500 mm in the dry north to more than 3000 mm in the wet south. Ecologists have demonstrated a significant relationship between modern population size and environment by examining atoll area and rainfall in the Marshall Islands. The archaeological investigation seeks to extend these modern observations into prehistory by examining the relationship of ancient habitation sites and horticultural systems to atoll land area and rainfall regime along the 1000-km long rainfall gradient. Four atolls were selected for study: Utrik Atoll in the dry north, Ujae and Maloelap atolls (Figure 2) in the middle of the archipelago and Ebon Atoll in the wet south. In total, more than 50 archaeological sites have been recorded, a dozen excavated (Figure 3; Weisler 1999a) and more than 100,000 faunal remains were predictably high in bones of inshore fish, with birds, turtles, rats, dogs, sea mammals and humans represented.
All of the nearly 1000 portable artefacts were fashioned from shell and bone, except for a few volcanic artefacts, the raw material of which floated to the atoll shores in the roots of drift logs. Some 56 radiocarbon age determinations anchor the culture-historical sequence which spans the past two millennia.
Although human colonists were initially drawn to the atolls for harvesting the untapped stocks of fish, molluscs and huge sea-bird colonies, the backbone of Marshallese terrestrial subsistence was based on the cultivation of Giant swamp taro (Cyrtosperma chamissonis) which was grown in pits with heavily mulched soils (Figure 4; Weisler 1999b). Detailed mapping and dating of these cultivation pits in reference to land boundaries suggest temporal changes in land tenure (Figure 5).
While a wealth of data now exists for addressing questions of atoll subsistence and economy, there is one fundamental question that needs resolution. Documenting clearly when various Pacific islands were first colonized by humans defines the speed and direction of the largest water-borne Neolithic expansion in human history. Since there is no general consensus of when any Pacific island was first settled, determining the chronology of island emergence provides an independent dataset that can be used to evaluate the dates for ancient habitation sites. In 2001, geological samples from the reef platform and solidified stranded beach features will be collected for dating island emergence. This will determine when the atolls were above sea level and large enough to sustain a human population. If there is a gap of several hundred years between substantial island emergence and the oldest documented dates for human habitation, then future archaeological excavations may locate even older cultural deposits. At first thought unworthy of serious study, atoll archaeology has much to offer regarding the long-term settlement of one of the most precarious landscapes.
)
Figure 3: This 2x4-m excavation unit revealed a thin cultural layer with post moulds, a stone alignment and larger pits for the preservation of fermented breadfruit paste, an important subsistence item on coral atolls.
)
Figure 4: Giant swamp taro (Cyrtosperma chamissonis) is still grown on some of the southern atolls that receive the highest rainfall in the Marshall Islands.
)
Figure 5: Portion of a settlement pattern map from Ebon Atoll showing the location of prehistoric (red) and historic (yellow) habitation areas in reference to the modern lagoon shore. The positioning of some Giant swamp taro pits (green) to traditional land boundaries suggests that changes in land tenure can be documented archaeologically.
Acknowledgements
Marshall Islands research is supported by grants from the University of Otago; the Historic Preservation Office, Republic of the Marshall Islands; the US National Park Service; the Australian South Pacific Cultures Fund; the United Nations, International Year of the World's Indigenous People Voluntary Fund; and Mobil Oil Corporation of Micronesia. A warm kommol tata to the dozens of Marshallese that have helped with this research. I thank Les O'Neill for preparing the figures.
References
- WEISLER M.I. 1999a. Atolls as settlement landscapes: Ujae, Marshall Islands. Washington (DC): Smithsonian Institution. Atoll Research Bulletin 460.
- ...1999b. The antiquity of aroid pit agriculture and significance of buried A horizons on Pacific atolls, Geoarchaeology 7: 621-54.