Dating the first New Zealanders: the chronology of Wairau Bar.

Higham, Thomas ; Anderson, Atholl ; Jacomb, Chris 等

New Zealand was the last major landmass to be settled before the industrial age, but vigorous debate has ensued over the precise date of first colonization (Sutton 1987; 1994; Anderson 1991; 1994). Orthodox propositions based largely upon the comparison of material culture types between New Zealand and tropical East Polynesia took the earliest dates with which such assemblages have been associated to suggest colonization around 800 AD (Davidson 1984). However, when Kirch (1986) argued that typological and palaeoenvironmental evidence indicated settlement of tropical Polynesia as early as 2000 years ago, Sutton (1987) followed suit for New Zealand by suggesting that the presence of charcoal in selected natural sites might reflect colonization between 0 and 500 AD. His use of evidence has been criticized (Grant 1988; Enright & Osborne 1988; Wilmshurst 1997). The palaeoenvironmental approach in Polynesian prehistory has been the subject of subsequent dispute (Spriggs & Anderson 1991; Kirch & Ellison 1994; Anderson 1994; 1995). Similarly contested are radiocarbon dates of up to about 2000 BP on gelatin from the bones of Rattus exulans, a species introduced prehistorically to New Zealand (Anderson 1996; Holdaway 1996).

Less contentious evidence arises from radiocarbon ages on materials recovered from archaeological sites. Analyses of radiocarbon determinations from excavated New Zealand archaeological sites has suggested a more recent age for Polynesian colonization closer to the 12th and 13th centuries (Anderson 1991; McFadgen et al. 1994; Higham & Hogg 1997).

One archaeological site of great importance has been absent from these analyses so far. This is Wairau Bar, located in the north of the South Island [ILLUSTRATION FOR FIGURE 1 OMITTED]. It was found by Eyles in the 1930s and later excavated by Duff (1950; 1977), Wilkes (1964) and Trotter (1975a) amongst others [ILLUSTRATION FOR FIGURE 2 OMITTED]. The abundance, range and quality of material culture, especially of adzes and ornaments, produced by these investigations are unparalleled in New Zealand or elsewhere in East Polynesia. The Wairau Bar material effectively constitutes the type assemblage of the New Zealand Archaic Phase of East Polynesian Culture (Golson 1959). As such, there are close similarities with artefacts from the Cook, Society and Marquesas Islands, and elsewhere in New Zealand. Since Archaic East Polynesian Culture is held to represent the colonising population of the region (leaving aside Kirch's (1986) objections in relation to Hawaii especially), an understanding of the archaeology of Wairau Bar, and especially its chronology, is clearly crucial to defining the age and character of the prehistoric settlement of New Zealand.

Until now, there has been no reliable chronology for the site. We present a new series of radiocarbon determinations from material excavated previously from the site and outline some implications of the results for understanding the chronology of New Zealand.

Archaeological investigations

Wairau Bar is situated on a boulder bank at the mouth of the Wairau and Opawa Rivers [ILLUSTRATION FOR FIGURE 1 OMITTED]. Three discrete urupa (burial areas) and associated occupation areas have been located by archaeologists. In the northwestern area, [TABULAR DATA FOR TABLE 1 OMITTED] burials 1-7 disclosed a variety of funerary items including complete moa eggs, imitation sperm-whale tooth pendants, necklaces of whale ivory 'reels', adzes of characteristic early East Polynesian types rendered in local and imported New Zealand stone and joints of moa (Duff 1977). A second urupa contained four more burials (8-11) with fewer burial goods. They were trussed and appeared to have been disturbed post-depositionally (Duff 1977: 49). The third or southern urupa comprised the bulk of the interments and contained individuals with fewer mortuary items.

Duff's (1950) analysis of the skeletal remains and their burial goods suggested that a higher proportion of males were interred with grave items compared to females. Houghton's (1975) reanalysis suggested that some skeletons had been misidentified. Leach (1977) concluded on the basis of these data that there was no sexual division and sex was not statistically significant in determining grave good accumulations. Subsequently, the variety in mortuary items between different urupa has been interpreted as a reflection of temporal change (Anderson 1989). 'Early' phase burials (1-7) were associated with a higher proportion of Archaic artefact types and Anderson (1991) suggested that these might represent the first generation of Polynesian settlers in New Zealand.(1) Graves in other urupa contained crouched burials and some nephrite and shark teeth ornaments. Whale teeth and reel ornaments were scarce and moa remains absent. At Wairau Bar, only three of the graves excavated possessed artefacts fashioned from nephrite, a material noted to be more common in later prehistoric contexts.

The most important aspect of the site's stratigraphy is its shallow depth. The cultural remains average 30-40 cm below ground surface. Five discrete layers (with Layer I the lowest) have been identified in the site (Wilkes 1964). Two (3 and 4) are cultural. Layer 3 is a charcoal-stained soil into which prehistoric moa ovens were excavated. Layer 4 was the major occupation layer of the site and comprised shell and fishbone middens. A spatial analysis is a more profitable guide to the site because it is doubtful that the identified cultural layers are laterally continuous, although Wilkes (1964) was able to correlate layers 3 and 4 in his excavations [ILLUSTRATION FOR FIGURE 2 OMITTED].

Site chronology

There is a long history to radiocarbon dating at Wairau Bar, but both the reliability of the earlier determinations and what cultural activity they might date are uncertain (Challis 1991) (TABLE 1). Two 14C determinations of charcoal (NZ-50 and Y-204), thought to be from layer 4 (Anderson 1989; Challis 1991), were amongst the first dated materials in New Zealand. They are probably not accurate because they are of unidentified wood which may contain significant inbuilt age (McFadgen 1982; Anderson 1991). Because of this possibility, Trotter (1975a; 1975b) dated three further samples from the 1959 and 1964 excavations (NZ-1835, 1837 and 1838). One of the samples, human bone from burial 42, yielded a radiocarbon age of 780 [+ or -] 80 BP. A single moa bone measurement produced an age of 590 [+ or -] 60 BP, while a sample of the bivalve Paphies australis (pipi) yielded a reservoir-corrected age of c. 680 BP (TABLE 1). The precise provenance of these samples is not known, but Trotter (1975b) stated that they related to 'the main period of prehistoric occupation' which suggests that they came from Layer 4. Trotter (1975a) suggested that this implied settlement at about 600-700 years ago. Three additional bone determinations were obtained in 1977 (NZ-4442, 4443, 4444; TABLE 1) (Challis 1991) which disclosed greater variation. Nitrogen analyses of bone were undertaken by Houghton (see Institute of Geological and Nuclear Sciences (IGNS), Lower Hutt, 14C fossil record forms) in an attempt to develop a relative sequence. These failed to produce meaningful results (Trotter, letter to McGill, 23 January 1978, in fossil record forms, IGNS). Anderson (1989) and Challis (1991) concluded that the radiocarbon data appear to favour an extensive occupation covering centuries.

One question regarding these radiocarbon samples concerns the reliability of the bone determinations. Jansen (1984) has described the use of phosphoric acid as the routine method used to pretreat bone at IGNS at the time the Wairau Bar human and moa bones were dated. Stafford et al. (1987: 25) have suggested that the probability that this method will date even well-preserved bone accurately is very low. In addition, human bone may be affected by the ingestion of marine resources, particularly in coastal sites (Ambrose & Norr 1993). The [Delta](13)C values for the three human bone samples (NZ-4442, 4443, 4444), however, appear to be terrestrial values (-19-7, -18.7 & -23-1 respectively), although in the absence of [Delta](15)N measurements, these cannot be substantiated. Percent nitrogen values determined by Houghton (see IGNS fossil record R5433/3) for his investigations of a relative chronology suggest that burial 35 had a low nitrogen content, implying it was older. Given the unreliability of the nitrogen method as a dating tool, we think this suggests the bone was probably degraded. Our new series of results supports this, and implies the sample was contaminated with young carbon. The moa bone sample comes from a disturbed matrix (Wilkes 1964). The shell sample, NZ-1837, appears to be reliable.

Until recently, the problem of developing an accurate chronology for Wairau Bar appeared intractable, because there were no suitable samples available for dating and permission for new excavations was declined by Rangitaane tribal authorities. An alternative emerged when an investigation of the reliability of prehistoric moa eggshell established its validity as a material suited for radiocarbon dating (Higham 1994). In 1992, we proposed to radiocarbon date samples of moa eggshell found in mortuary contexts at the Wairau Bar site, stored in the Canterbury Museum. Eleven samples of eggshell were obtained (TABLE 2). In each burial, the eggshells are in primary context (Duff 1950) and therefore should provide accurate radiocarbon determinations. We also obtained two shell samples from excavations by Wilkes (1964).

[TABULAR DATA FOR TABLE 2 OMITTED]

Results

Each 14C determination comprises duplicate meaned AMS analyses, with the exception of AA-22557 and the shell determinations (TABLE 2). The calibrated results all suggest an occupation in the late 13th century AD. OxCal's combined probabilities method (Bronk Ramsey 1995) suggests that the calibrated ages of both estuarine shell and eggshell series are in close agreement. The eggshell determinations yield an age range of 1285-1300 cal AD (2[Sigma] [A.sub.overall] = 170.7%)(2) and the shell 1269-1324 AD (2[Sigma] - [A.sub.overall] = 131 [multiplied by]3%). Taken together, these combined series yield a calibrated range at 2[Sigma] of 1288-1300 cal AD ([A.sub.overall] = 115 [multiplied by] 5%). This suggests the radiocarbon determinations as a group are statistically indistinguishable and, in parallel with the shallow and uninterrupted stratigraphy at Wairau Bar implies a brief, late 12th-century occupation.

A comparison of our results with previous radiocarbon determinations of bone from this site demonstrates the unreliability of the burial 35 bone sample.

Discussion

Many moa-hunting sites which are essential to arguments about the age of human colonization in New Zealand are so large and stratigraphically complex that it is not possible to comprehend their chronologies on the basis of small samples of radiocarbon determinations. As Anderson (1991) and Anderson et al. (1996) have argued in relation to the Shag River Mouth site, only a large series will do. The new data from Wairau Bar go some way toward providing this. Material has been dated from both of the large urupa within the site and from areas of midden and other burial contexts. The results show that each urupa has a similar antiquity. In addition, the stratigraphy is shallow and there are no identifiable hiatuses. Taken together, these data appear to suggest a brief phase of occupation. Exactly how brief is a question difficult to answer. The calibrated data suggest a very short sequence, but based on the archaeological evidence we think they could represent an occupation of about 20 years or less. There were probably interments on Wairau Bar during later prehistory, as occurred occasionally in the historical era. Although some burials are undated, it can be assumed that within the three discrete urupa discussed here the burials were clustered in time. The radiocarbon results mean we can set aside with confidence previous interpretations of the site as a village which was occupied for some centuries (Duff 1977; Anderson 1989) and instead consider other alternatives.

Settlement patterns in the early, or Archaic phase in New Zealand are dominated by large sites located at the mouths of rivers with wide river valleys. At sites like Wairau Bar, we can envisage a Polynesian occupation at the dawn of settlement, when there were abundant resource opportunities. The optimal adaptation to exploiting such a novel environment may have been to operate from a permanently occupied, though short-term, base camp. Traditionally, it has been assumed that these sites were occupied for some centuries as permanent, or semi-permanent villages. Anderson (1991) has argued that these sites may represent less than a century of occupation from the beginning of the prehistoric period, when big game predominated in the subsistence strategy. The new radiocarbon determinations indicate Wairau Bar is a similar case.

Brevity of occupation has important implications for inferring the behaviour of the prehistoric inhabitants. It indicates a rapid local depletion of big-game resources, notably moa, indicating systematic and determined predation. It also invites reconsideration of Duff's (1977) suggestion that the concentration of grave goods in burials 1-7 represents status differentiation. Inference of social rank by mortuary practice of this kind is commonly argued in archaeology (e.g. Brown 1995). On the other hand, the relative scarcity of ornaments fashioned from moa bone in the southern burial areas at the site might reflect only the impact of predation on raw materials, and the beginnings of a shift to those which characterized high status in later phases of prehistory, notably nephrite. Since the urupa appear more or less contemporaneous by radiocarbon dating, that proposition and its alternative of ranking will need to be tested in future by further analysis of the burial attributes and by comparison with other cases.

Conclusions

A new series of radiocarbon determinations enables the critical site of Wairau Bar to be brought into discussion about the early colonization of New Zealand. Determinations on moa eggshell from grave contexts and estuarine shell from occupation layers show that the site was occupied towards the end of the 13th century AD. The brevity of occupation is consistent with similar early sites which also disclose rapid depletion of local big-game resources. On that ground, these sites appear to represent the earliest phase of human settlement in New Zealand. In terms of material culture, they contain both the widest range and the greatest abundance of types which belong to Archaic East Polynesian culture, regarded as the colonizing culture of New Zealand.

The similarity in age between Wairau Bar and other early settlements, together with their brevity of occupation and evidence of resource depletion, suggests that the first colonists engaged in a sustained assault upon the fragile populations of big-game taxa, shifting their settlements frequently as local reserves were depleted (Anderson & Smith 1996). That behaviour made the colonizing horizon more visible than it might otherwise have been. On a larger scale, the same seems to be true of the colonization of the southern Pacific generally. There is an emerging regional sphere of interaction between mainland New Zealand and other islands in the region evidenced by the similarity of the earliest radiocarbon determinations from archaeological contexts in the Kermadecs, Norfolk Island and New Zealand (Anderson 1991; Higham & Johnson 1996; Anderson n.d.) and by evidence of the transfer of obsidian between these groups (Anderson & McFadgen 1990; Anderson et al. 1997). Taken together, these data suggest that the New Zealand region was settled as part of a mobile, expansive phase of exploration from tropical East Polynesia, no earlier than about 750 years ago. The new dates from Wairau Bar remove the uncertainty that Archaic East Polynesian settlement might have occurred substantially earlier and opens intriguing new avenues of enquiry regarding the process of adaptation by human beings in a new land.

Acknowledgements. We acknowledge Rangitaane iwi for their permission to undertake this work. Dr F. Petchey (University of Waikato) commented on the bone dates from Wairau Bar.

1 Wilkes (1964) and Anderson (1989) have also suggested that burials 1-7 predate the rest of the Wairau mortuary remains because they were interred from graves cut from a lower stratigraphic horizon than the other burials.

2 The agreement index ([A.sub.overall]) indicates the extent to which the final or posterior calibration age distribution (meaned value) overlaps the original distribution (the individual calibrated range). The value for [A.sub.overall] should not fall below 60%. If it does, the reliability of the calibrated age range could be called into question. In this instance, the agreement index is highly acceptable because the value exceeds 100% (see Bronk Ramsey 1995).

References

AMBROSE, S. H & L. NORR. 1993. Experimental evidence for the relationship of the carbon isotope ratios of whole diet and dietary protein to those of bone collagen and carbonate, in J. B. Malbert & G. Grupe (ed.), Prehistoric human bone - archaeology at the molecular level: 1-38. Berlin: Springer-Verlag.

ANDERSON, A. J. 1989. Prodigious birds. Moas and moa-hunting in prehistoric New Zealand. Cambridge: Cambridge University Press.

1991. The chronology of colonization in New Zealand, Antiquity 65: 767-95.

1994. Palaeoenvironmental evidence of island colonisation: a response, Antiquity 68: 845-7.

1995. Current approaches in East Polynesian colonisation research, Journal of the Polynesian Society 104: 110-32.

1996. Was Rattus exulans in New Zealand 2000 years ago? AMS radiocarbon ages from Shag River Mouth, Archaeology in Oceania 31: 178-84.

N.d. Implications of obsidian transfer in southern Polynesia. Unpublished manuscript, Australian National University, Canberra.

ANDERSON, A. J., W. AMBROSE, B. F. LEACH & M. WEISLER. 1997. Material sources of basalt and obsidian artefacts from a prehistoric settlement site on Norfolk Island, South Pacific, Archaeology in Oceania 32: 39-46.

ANDERSON, A. J. & B. G. MCFADGEN. 1990. Prehistoric two-way voyaging between New Zealand and East Polynesia: Mayor Island obsidian on Raoul Island and possible Raoul Island obsidian in New Zealand, Archaeology in Oceania 25: 24-37.

ANDERSON, A. J. & I.W.G. SMITH. 1996. The transient village in southern New Zealand, World Archaeology 27: 359-71.

ANDERSON, A. J., I.W.G SMITH & T.F.G. HIGHAM. 1996. Radiocarbon chronology, in A. J. Anderson, B. J. Allingham & I.W.G. Smith (ed.), Shag River Mouth: the archaeology of an early southern Maori village: 60-69. Canberra: ANH Publications. Australian National University Research Papers in Archaeology and Natural History 27.

BRONK RAMSEY, C. 1995. Radiocarbon calibration and analysis of stratigraphy: The OxCal program, Radiocarbon 37: 425-30.

BROWN, J. 1995. On mortuary analysis, with special reference to the Saxe-Binford research programme, in L.A. Beck (ed.), Regional approaches to mortuary analysis: 2-26. New York (NY): Plenum.

CHALLIS, A. J. 1991. The Nelson-Marlborough region: An archaeological synthesis, New Zealand Journal of Archaeology 13: 101-42.

DAVIDSON, J. 1984. The prehistory of New Zealand. Auckland: Longman Paul.

DUFF, R. S. 1950. The Moa-hunter period of Maori culture. Wellington: Government Printer.

1977. The Moa-hunter period of Maori culture. 3rd edition. Wellington: Government Printer.

ENRIGHT, N. J. & N. M. OSBORNE. 1988. Comments on D.G. Sutton's paper 'A paradigmatic shift in Polynesian prehistory: Implications for New Zealand', New Zealand Journal of Archaeology 10: 139-46.

GOLSON, J. 1959. Culture change in prehistoric New Zealand, in J.D. Freeman & W.R Geddes (ed.), Anthropology in the South Seas: Essays presented to H.D Skinner. 29-74. New Plymouth: Thomas Avery & Sons.

GRANT, P. J. 1988. Interpretation of evidence for the early prehistory of New Zealand: reply to Sutton, New Zealand Journal of Archaeology 10: 109-28.

GUMBLEY, W. 1996. Conservation plan for Wairau Bar moa hunter site. Wellington: New Zealand Historic Places Trust.

HIGHAM, T.F.G. 1994. Radiocarbon dating New Zealand prehistory with moa eggshell: Some preliminary results, Quaternary Geochronology (Quaternary Science Reviews) 13: 163-9.

HIGHAM, T.F.G. & A.G. HOGG. 1997. Evidence for late Polynesian colonisation of New Zealand: University of Waikato Radiocarbon measurements, Radiocarbon 39: 149-92.

HIGHAM, T.F.G. & L. JOHNSON. 1996. The prehistoric chronology of Raoul Island, the Kermadec Group, Archaeology in Oceania 31: 207-13.

HOLDAWAY, R.N. 1996. Arrival of rats in New Zealand, Nature 384: 225-6.

HOUGHTON, P. 1975. The people of Wairau Bar, Records of the Canterbury Museum 9: 231-46.

JANSEN, H. 1984. Radiocarbon dating for contributors. Institute of Nuclear Sciences, Department of Scientific and Industrial Research 328.

KIRCH, P.V. 1986. Rethinking East Polynesian prehistory, Journal of the Polynesian Society 95: 9-40.

KIRCH, P. V. & J. ELLISON. 1994. Palaeoenvironmental evidence for human colonisation of remote Oceanic islands, Antiquity 68: 310-21.

LEACH, B. F. 1977. Sex and funeral offerings at Wairau Bar: A re-evaluation, New Zealand Archaeological Association Newsletter 20: 107-13.

MCCORMAC F. G., A.G. HOGG, T.F.G. HIGHAM, M.G.L. BAILLIE, J.G. PALMER, L. XIONG, J.R. PILCHER, D. BROWN & S.T. HOPER. 1998. Variations of radiocarbon in tree-rings: Southern hemisphere offset preliminary results, Radiocarbon 40: 1153-9.

MCFADGEN, B.G. 1982. Dating New Zealand archaeology by radiocarbon, New Zealand Journal of Science 25: 37992.

MCFADGEN, B.G., F.B. KNOX & T.R.L. COLE. 1994. Radiocarbon calibration curve variations and their implications for the interpretation of New Zealand prehistory, Radiocarbon 36: 221-36.

MILLER, G. H., J.W. MAGEE, B.J. JOHNSON, M.L. FOGEL, N.A. SPOONER, M.T. MCCULLOCH & L.K. AYLIFFE. 1999. Pleistocene extinction of Genyornis newtoni: Human impact on Australian megafauna, Science 283: 205-8.

SPRIGGS, M. & A. J. ANDERSON. 1993. Late colonisation of East Polynesia, Antiquity 67: 200-217.

STAFFORD, T.W., JR., A.J.T. JULL, K. BRENDEL, R.C. DUHAMEL & D. DONAHUE. 1987. Study of bone radiocarbon dating accuracy at the University of Arizona NSF accelerator facility for radioisotope analysis, Radiocarbon 29: 24-44.

STUIVER, M., P.]. REIMER, E. BARD, J.W. BECK, G.S. BURR, K.A. HUGHEN, B. KROMER, E.G. MCCORMAC, J. VAN DER PLICHT & M. SPURK. 1998. INTCAL98 Radiocarbon age calibration, 24,000-0 cal AD, Radiocarbon 40: 1041-83.

STUIVER, M., P.J. REIMER & S. BRAZIUNAS. 1998. High precision radiocarbon age calibration for terrestrial and marine samples, Radiocarbon 40: 1127-51.

SUTTON, D.G. 1987. A paradigmatic shift in Polynesian prehistory: implications for New Zealand, New Zealand Journal of Archaeology 9: 135-57.

1994. Conclusion: origins, in D.G. Sutton (ed.), The origins of the first New Zealanders: 243-58. Auckland: Auckland University Press.

TROTTER, M.M. 1975a. Further excavations at Wairau Bar, New Zealand, Asian Perspectives 19: 75-80.

1975b. Radiocarbon dates for Wairau Bar and Wakanui, South Island, New Zealand Archaeological Association Newsletter 18: 90-91.

WILKES, O. 1964. Further work at Wairau Bar. Canterbury: Canterbury Museum.

WILMSHURST, J. M. 1997. The impact of human settlement on vegetation and soil stability in Hawke's Bay, New Zealand, New Zealand Journal of Botany 35: 97-111.