The colonisation of New Zealand by the Polynesians

Abstract

Individual radiocarbon dates of events within the last thousand years, such as the colonisation of New Zealand by the Polynesians, are difficult to interpret because of statistical and technical errors of dating. Charcoal dates on the one hand, and moa-bone collagen and marine shell dates on the other, track that episode with a mean disparity of about 240 years. Analysis of moa-bone collagen and marine shell dates of moa-hunter sites indicates that about 45% of withinsite variance of dates is statistical error, and about 43% is technical error; only about 12% reflects the interval of deposition. That calculation returns an estimate of about a century for the average period over which moas and Polynesians coexisted in a district. Models of the colonisation of New Zealand by Polynesians are tested against 55 non-charcoal radiocarbon dates of moa-hunter sites. The model favoured by Green, by Bellwood, and by Davidson, of an explosive coastwise colonisation, does not match the data. Duff's notion of a measured cadence of dispersal from a landing in the Bay of Plenty is similarly at odds with the geographic trend of those dates. However the dates lend support to Trotter's and McCulloch's model, limited to the South Island, of a wave of colonisation of constant velocity from north to south down its east coast. The fit of dates to that model is enhanced by shifting the point of landing in the South Island to the vicinity of the Kaikoura Peninsula and by allowing an acceleration in the velocity of spread. This “Kaikoura model” implies that the initial landing in New Zealand was on the north-east coast of the South Island and that the Polynesians spread from there both north and south. Regression of dates on distance outward suggests an initially slow rate of spread, below 1/2 km y^-1. until AD 1200 (collagen chronology), pushed by a population growinqat 3% y^-1. By AD 1300 the rate of radial expansion had increased to 1 km y^-1 and by AD 1400, shortly before the colonisation of New Zealand was completed, the rate had reached its final value of about 10 km y^-1. The pattern of colonisation so deduced — an initially slow rate of spread accelerating to an approximately constant rate of radial expansion — is consistent with the scant data describing the colonisation of Australia by the Aborigines more than 30,000 years previously and reconciles the apparent differences between the patterns of colonisation of New Zealand and of the Americas.