New Zealand is at the southern limits of the human-extended geographic range of the Pacific rat Rattus exulans On the two main islands, radiocarbon dates on rat bones from natural sites show that rat populations were established more than 1000 years before permanent human settlement, presumably during transient visits by Polynesian voyagers Both main islands were colonised after these first contacts, but offshore and outlying islands were not reached by rats until after Polynesian settlement about 700 years B P Chatham Island was not colonised by Pacific rats until about 650 years B P I present a model that relates the time of first appearance of rats in the fossil record and the exploitation of native fauna to the pattern of spread of the rat through the archipelago I hypothesise that the stepwise spread of the rat through the archipelago is mirrored by the pattern of reduction and extinction of indigenous fauna vulnerable to rat predation The 1000 year delay between the arrival of rats and permanent human settlement suggests that the New Zealand biota was already stressed by an introduced predator before humans added habitat destruction and over-hunting Keywords Rattus exulans. Pacific rat, colonisation, New Zealand, radiocarbon dating, Polynesian voyaging
INTRODUCTIONThe geographical range of the Pacific rat Rattus exulans has been vastly extended by human vectoring From its onginal distribution in South East Asia, the rat has spread into the north, east, and southern Pacific (Wodzicki & Taylor 1984, Roberts 1991) The present southern limit of its range is on a few of the small islets off the southwest coast of Stewart Island, New Zealand (Atkinson & Moller 1990) The source or sources of the New Zealand populations of Pacific rats are not known with certainty, but mitochondnal DNA (mtDNA) studies are providing important new data on possible source populations (Matisoo-Smith 1994) Roberts (1991) summansed the earliest radiocarbon dates for human presence m the Pacific as a proxy for the earliest dates for Rattus exulans Roberts's model is based on the assumptions that (1) rats and people travelled together, so first dates of colonisation by the two species should coincide, and (2) the island groups were settled, and Rattus exulans populations established, m order of accessibility of island groups to human colonisation, as proposed by Irwin (1989) Rats reached possible source areas at least 2000 years B P For New Zealand, Roberts's model implies first colonisation by people and rats about 1000 years B P , following Irwin et al (1990) Nonarchaeological occurrences of the rat were not presented by Roberts (1991), so the data she presented for human arrivals could not show whether rats were already present in each island group It is axiomatic that dates from an archaeological site cannot provide information on