The Kaharoa Tephra as a critical datum for earliest human impact in northern New Zealand

The debate over the timing of arrival and earliest environmental impacts of the first New Zealanders has intensified in recent years, in part fuelled by new evidence or reinterpretation of old evidence from other points along prehistoric Polynesian migratory routes. An examination of two radiocarbon-dated pollen records from northern New Zealand shows how divergent interpretations could be drawn from the same evidence to support both early and late colonization models. Radiocarbon-dated pollen records for deforestation at sites where the risk of contamination by older carbon may be high cannot be used to establish chronologies of settlement. Ombrogenous mires, where peat accumulation proceeds independently of groundwater or surface runoff, are likely to yield the most reliable records. Another solution applicable in northern New Zealand is to base chronologies on tephra layers with reliable age estimates from multiple determinations. Analysis of 11 pollen records currently known to contain the 665 ± 15 BP (c. 600 cal-BP) rhyolitic Kaharoa Tephra demonstrates the critical stratigraphic position that this isochronous surface occupies in New Zealand prehistory. The earliest inferred human impacts occurred at around the time of deposition of the tephra in eight of these sites (73%), and well after it as the remainder. These results are in agreement with a later colonization model and suggest that proximity to accessible food resources was more important than climate or latitude in determining early colonization sites. These findings will be tested fully when the known range of Kaharoa Tephra is extended beyond its present limited range through the application of micro-tephra analysis. In line with Anderson's (1995: 128) premise that the later colonization model is eminently falsifiable, we suggest that tephropalynological studies involving the Kaharoa Tephra may hold the key to resolving debates over the timing and spatial patterns of earliest human impact in northern New Zealand. In particular, these records permit the investigator to avoid the ambiguities that emerge frequently when pollen profiles indicating deforestation are dated by radiocarbon alone.