exploitation strategy based upon stored foods and exchanges, including the payment of tribute to chiefs.

Chronological and Demographic Studies

In research on fortified pa, Groube (1964, 1970) plotted the regional distribution of chronological types based on their evolution from large, open-terraced pa to small pa that were heavily defended by multiple ditches, banks, and palisade posts. Initially, Groube explained this change in social terms as reflecting the breakdown of large tribal entities (and chieftainships?) into small, highly competitive descent groups. Chronology, demography, and resource stress were also included as factors. Groube argued that it would take 200 years or more for the population of a small group of colonists arriving in a.d. 1000 to reach about 25,000. After that, however, even moderate growth would take the population up to 100,000 by a.d. 1700. The thousands of fortified pa sites that dot the North Island landscape were the outcome of increasing competition for good-quality agricultural land within the zone of favorable climate. This was a form of resource stress through overpopulation that only had major effects during the final 200 years of New Zealand’s prehistory.

It was many years before the information on pa distribution and chronology would allow the testing of Groube’s hypotheses. Matt Schmidt (1996), using Anderson’s (1991) suggestions for chronometrical reliability, derived 72 robust dates from a series of 221 radiocarbon dates from ninety-six pa. These dates suggested, first, that pa sites appeared rather suddenly at about a.d. 1500 and, second, that Groube’s different classes of pa were spatially and chronologically coterminous. These conclusions repeated those of Geoffrey Irwin’s (1985) study of spatial and hierarchical relationships of the pa of Kaipara Harbor and elsewhere. Irwin found that there was evidence that the settlement system (of Pouto) went through a stress threshold during the late prehistoric period that resulted in a spate of pa building. Multiple small pa and a few massive ones indicative of a higher order of organization suggested that the pa building was a response to both internal division and external threat.

Anderson (1991), in an extensive review of the time of initial settlement of New Zealand based on radiocarbon dates, found little evidence for settlement earlier than a.d. 1200. He added that the possibility of a planned immigration by some hundreds of people could not be ruled out. Sutton (1987, 1994), by contrast, used indications of fire disturbance in New Zealand pollen cores to claim that initial settlement was substantially earlier than a.d. 800. Part of Sutton’s argument for an early date of colonization is based on life expectancy data and the implications of fertility and diet for population growth (Brewis, Malloy, and Sutton 1990).

Questions regarding the chronology of settlement and population growth in prehistoric New Zealand directly concern the time available for cultural processes. At present, the short chronology supports the idea of multiple settlement episodes and rapid and possibly intrusive changes in population, technology, and settlement patterns. The longer chronology supports arguments for the internal development of Pacific societies through evolutionary processes that led to the development of chiefly societies as a result of population increase, resource stress, agricultural intensification, or competition (Kirch and Green 1987).

Groube (1970) allied his model of population growth and resource stress with David Simmons’s attempt (Simmons 1976) to reinstate the status of Maori canoe traditions by shifting their location to voyages around and between the islands of New Zealand (see Orbell 1985). Linking the distribution of the highly defended “ring ditch” pa with the historic movements of the “Awa” peoples enabled Groube to claim that the exile and forced migration of this group followed a standard Polynesian response to overpopulation. Computer simulations (Levison, Ward, Webb 1973), experimental voyages (Finney 1977), and increased knowledge of Pacific archaeology and Polynesian sailing techniques (Irwin 1992) have led to a reassessment of Andrew Sharp’s pessimistic views. His model has now been replaced by the possibility of planned voyages, multiple episodes of discovery and colonization,