Radiocarbon dating the prehistory of New Zealand

Abstract

The impact of radiocarbon dating upon chronological studies of New Zealand’s prehistory has been limited for several reasons. A long standing tradition of investigating archaeology as ethnology, based on the rich sources of ethnographic data has lessened the apparent need for chronological control. Second, there has been an element of consensus regarding the date of discovery by initial human settlement due to the establishment of a chronology based on tribal traditions. There has been a failure by some archaeologists adequately to implement radiocarbon dating programmes and scientific dating effectively as a research tool. The most important reason, however, is the short span of prehistoric occupation. This dissertation addresses a number of issues affecting accurate radiocarbon dating of New Zealand prehistoric material with particular reference to errors which restrict the application of high precision dating. Such errors come from three principal sources - during field collection of the samples, in the measurement of the radioactivity of the sample in the laboratory and in the variation in the amounts of ¹⁴C in the atmosphere and the measurement of the half-life. Field errors are common in New Zealand and often reflect post-depositional migration and erosion, and bioturbation and a failure to understand the site’s stratigraphy. Errors due to the misreporting of radiocarbon dates and failure to adequately interpret them have also been common. Accuracy, reproducibility and precision are analysed in the University of Waikato laboratory. Close statistical agreement is shown in the results obtained from the laboratory’s IAEA standard set of measurements. A laboratory error multiplier is calculated using 11 repeat dates of the IAEA C3 standard. Spectral shifts identified in certain standard and sample spectra is investigated and the external standard is found to be artificially influencing the position of the end point of successive spectra. The significance of this upon calculated radiocarbon dates is shown to be insignificant and reproducibility in count rate is maintained even with end point shifts of up to 7 channels. The three main sample types used in New Zealand archaeological are considered for their contribution to radiocarbon errors. All possess intrinsic problems. Moa eggshell carbonate samples from sites in Otago and Hawkes Bay are investigated for accuracy. These are the first moa eggshell samples dated from cultural contexts. The material is appropriate for the radiocarbon technique, dating the prehistoric event closely and resistant to contamination. The only drawback is its irregular presence in sites. Contamination in shell and charcoal samples is investigated. Charcoal from Shag Mouth, Pleasant River and Killermont sites is used in a series of physical and chemical pretreatments. It is largely free of contamination with the exception of the Pleasant River site, where adsorption by young carbonates has occurred. All samples possess errors due to rootlet intrusion, although these are negligible in their error. Shell dating is investigated extensively. Hard water effects are considered at North Otago sites. Dates of modem shell limestone regions at Shag Mouth and Pleasant River are compared with those from near Dunedin. No significant differences are identified. Differences are found, however, in prehistoric dates of shell of a variety of species. The major influencing variables are the immediate environment and the type of shellfish species selected for dating. Shellfish from estuaries which are regularly flushed with tidal waters are likely to give more accurate dates than those which have larger riverine inputs and barrier spits blocking tidal encroachment Shellfish obtain their carbon from a variety of sources. The proportion of carbon obtained from metabolic carbon and dissolved inorganic carbon utilised in their shell carbonate varies, so their radiocarbon uptake may vary. In an estuary, within which resources from a number of areas are collected, this variety may be manifested in different radiocarbon dates from different shellfish species. Models for accurate dating using charcoal and shell in the light of different environmental situations are described. Radiocarbon dates measured in the dissertation are considered in their contexts and sites and the initial colonisation of Murihiku is reviewed. A model of rapid colonisation and the short-term occupation of coastal, estuarine sites is presented on the basis of radiocarbon and archaeological variables. Further avenues of research are then examined.