Palaeoenvironmental evidence of island colonization: a response.

Anderson, Atholl

More on whether evidence of prehistoric environment on the Pacific island of Mangaia does or does not demonstrate an early human presence there.

Kirch & Ellison (1994) argue in the June ANTIQUITY that new palaeoenvironmental research on Mangaia (Cook Islands) seriously challenges the short chronology of East Polynesian pre-history proposed by Spriggs & Anderson (1993). It does not, for reasons shall discuss following clarification of several points.

Spriggs & Anderson (1993) do not hold the view that a standstill of colonization in West Polynesia was followed by resumed colonization eastward driven by explosive population expansion (Kirch & Ellison 1994: 319). We said the opposite, that 'East Polynesian colonists presumably outran the push of population growth' (Spriggs & Anderson 1993: 211, my emphasis), and have argued that case repeatedly. We are not looking for the 'first colonization site' as a 'Holy Grail', nor do we take a negative position with regard to palynological evidence (Kirch & Ellison 1994: 318). Actually, we are both currently engaged in palynological projects concerning Pacific prehistory. Lastly, Kirch & Ellison (1994: 318) say there is now evidence of anthropogenic palaeoenvironmental disturbance dated at 1400 b.p. in New Zealand. However, Striewski et al. (1994: 22) claim only that the determination suggests disturbance by about 1000 b.p.

The Mangaian case

The Mangaian case is based on analysis of three, out of 24, sediment cores. Two cores (VT6, TM7) show a rapid decline in tree pollen (although Pandanus, generally a disturbance indicator, disappears at this time in VT6), an increase in fern, and the first appearance of charcoal at about 2500 b.p. The third core (TIR-1), from the same basin as VT6 and initially published (Kirch et al. 1991) as disclosing similar changes from about 1600 b.p., has now been re-interpreted as showing the forest decline also beginning at about 2400 b.p. (Kirch & Ellison 1994). These data are held to demonstrate the presence of people on Mangaia by 2500 b.p., which is 1500 years earlier than the first known archaeological evidence.

My first objection to this proposition is that the reinterpretation of TIR-1 is too convenient. Without benefit of any new evidence different palynologists have perceived the beginning of deforestation at points 900-1100 years apart.

Secondly, while much is made of the initial appearance of charcoal at about 2500 b.p. in VT6 and TM7, it is not acknowledged by Kirch & Ellison (1994) that charcoal was also present by about 7000 b.p. in TIR-1 (Kirch et al. 1992: 177) and from similarly low down in TIR-2 (Kirch et al. 1991: 322). There are no charcoal counts for these cores, but quantification is misleading in any case. How much charcoal ends up in a sampled basin, and its pattern of distribution in a core, are dependent on various factors: size, type and proximity of fires, efficiency and modes of charcoal transportation, intervening trapping mechanisms, etc. It cannot be simply correlated with a presumed scale and origin of burning events.

Thirdly, since the early charcoal in TIR-1 occurs at about the time of a substantial forest decline, the argument that similar phenomena mark human colonization in the late Holocene should apply also to events around 7000-6000 b.p. (interpreted by Kirch et al. (1992: 177) as resulting from a drier climate at that time). Kirch & Ellison (1994) implicitly refute this point by arguing that it is the replacement of peat by clay infill, from extensive erosion, which confirms the anthropogenic status of the later events as against those earlier. But the chronology of clay infill is not at all clear. They note (1994: 314), that a transition from peat to clay infill is dated in four cores between about 3000 and 1600 b.p. They do not report that exactly the same phenomenon is also dated at 6260 [+ or -] 80 b.p. in core KA3 (but 2980 [+ or -] 80 b.p., in core KA4 -- the difference attributed to infill rates), and at 610 [+ or -] 100 b.p. in core VT6, that bands of clay are dated as early as 6480 [+ or -] 100 b.p. in core TM7, and that clay infilling is entirely absent in core VT4 (Ellison 1994: figure 2). Furthermore, considering the mechanisms by which disconformities can occur in sedimentary sequences, it is by no means certain that the peat dates refer in any meaningful way to the age of the clay deposits above them, for which no dates are available.

It is apparent, therefore, that deforestation and clay and charcoal influxes are not confined to the late Holocene. Neither is there a clear-relationship between declining tree pollen and massive clay influx, events occurring from 7000-6000 b.p. onward and separated by about 2000 years in VT6 and a substantial period, undated, in TM7. The pattern perceived by Kirch & Ellison (1994), disappears with consideration of the full data-set.

Turning to the radiocarbon dates on peat from TIR-1, Spriggs & Anderson (1993: 211) thought of potential contamination by old carbon. Kirch & Ellison (1994: 315) argue that since Mangaian streams flow over basalt there could be no contamination by ancient limestone carbon. However, the inner edge of the coralline makatea drains into the sampled basins, and ancient carbon could also have entered Lake Tiriara, the basin of the crucial TIR and VT core series, through sea water. The lake is brackish and was possibly more saline in the past (Ellison 1994). Aquatic plants can take up old carbon from water in ways not easily alleviated by sample pre-treatment, or perceived in the fractionation data (Marcenko et al. 1989). Steadman & Kirch (1990: 9607) also suggest that there might be old soil charcoal in the Lake Tiriara cores, which is quite probable given mobilization of the inland soils with extensive erosion. This, too, might affect the dating results.

Inferences of human activity from the cores cannot be reconciled with those from the archaeological record. In particular, the rapid extinction of eight species of land birds about 1000 b.p., as recorded in the earliest archaeological site on Mangaia, Tangatatau (MAN-44), looks very much like a human colonization event (Steadman & Kirch 1990: 9607), particularly since the remains are concentrated in the lower occupation levels above sediments which are culturally sterile. It is argued those species survived in the makatea during 1500 years of human habitation elsewhere on the island until that resource zone -- difficult of human access -- was exploited (Kirch et al. 1991: 324; Kirch & Ellison 1994). This is implausible. At least half the species are highly susceptible to predation by introduced rats which had little difficulty getting about the makatea, not to mention also pigs and fowls. The argument for 1500 years of survival by eight vulnerable species in one part of a small island, and the attempt to generalize this point (Kirch & Ellison 1994: 317), contradicts the common experience in Polynesia Where 'the major effect on island faunas often appears at the earlier end of an occupation sequence' (Kirch 1984: 146, cf. Steadman 1989: 200). Consequently, it is more likely that Tangatatau and other early sites on the island, including Vairorongo, a coastal settlement containing similar Archaic East Polynesian material (Walter pers. comm.), are exactly what they seem to be -- sites representing the initial coastal and inland colonization of Mangaia about 1000 b.p.

Conclusions

The Mangaian case does not demonstrate human settlement thereby 2500 b.p. Indeed, it is hard to see below palaeoenvironmental data could, unless of a very unusual kind (e.g. pollen of cultigens which required human transportation). At best, these data suggest useful hypotheses which must then be tested archaeologically, not substituted for archaeology. For example, there needs to be some attempt to locate the early sites suspected by the authors to lie under clay deposits around the swamps.

At present, then, we have three hypotheses worth considering:

Firstly, that Kirch & Ellison (1994) are correct.

Secondly, that some or all of the radiocarbon dates from the cores are too old, and that the environmental events in question actually occurred later. This might be partially tested, as I recommended to Kirch last year, by AMS dates on the charcoal fragments -- although problems of old soil charcoal would remain.

Thirdly, that the dates are generally correct but the events were caused by natural processes, as is conceded for similar phenomena from the early Holocene section of the cores. One salient point in this respect is that a significantly drier climate seems to have been general in the south Pacific around 3000-2000 years ago. Evidence of this, including declining tree pollen and increased charcoal frequency, is documented from southeast Australia and throughout New Zealand, including the humid sub-tropical region of Northland, at the same time as the events in question on Mangaia. It is attributed to increased frequency of drought and natural forest fires (e.g. Dodson et al. 1988; Newnham et al. 1989).

Whichever of the hypotheses is more probable, Kirch & Ellison (1994) have yet to establish a case from Mangaia for overturning the Spriggs & Anderson (1993) conclusions about the chronology of East Polynesian prehistory.

Acknowledgements. For assistance on particular points I thank Joanna Ellison, Geoff Hope, Matthew Spriggs, Douglas Sutton and Richard Walter.

References

DODSON, J.R., N.J. ENRIGHT & R.F. MCLEAN. 1988. A late Quaternary vegetation history for far northern New Zealand, Journal of Biogeography 15: 647-56.

ELLISON, J. 1994. Paleo-lake and swamp stratigraphic records of Holocene vegetation and sea-level changes, Mangaia, Cook Islands, Pacific Science 48: 1-15.

KIRCH, P.V. 1984 The evolution of the Polynesian chiefdoms. Cambridge: Cambridge University Press.

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

KIRCH, P.V., J.R. FLENLEY & D.W. STEADMAN. 1991. A radio-carbon chronology for human-induced environmental change on Mangaia, Cook Islands, Radiocarbon 33: 317-28.

KIRCH, P.V., J.R. FLENLEY, D.W. STEADMAN, F. LAMONT & S. DAWSON. 1992. Ancient environmental degradation: prehistoric human impacts to an island ecosystem: Mangaia, Central Polynesia, National Geographic Research and Exploration 8: 166-79.

MARCENKO, E., D. SRDOC, S. GOLUBIC, J. PEZDIC & M.J. HEAD. 1989. Carbon uptake in aquatic plants deduced from their natural 13C and 14C content, Radiocarbon 31: 785-94.

NEWNHAM, R.M., D.J. LOWE & J.D. GREEN. 1989. Palynology, vegetation and climate of the Waikato lowlands, North Island, New Zealand, since c. 18,000 years ago, Journal of the Royal Society of New Zealand 19: 127-50.

SPRIGGS, M. & ANDERSON 1993. Late colonization of East Polynesia, Antiquity 67:200-17.

STEADMAN, D.W. 1989. Extinction of birds in eastern Polynesia: a review of the record and comparisons with other island groups, Journal of Archaeological Science 16: 177-205.

STEADMAN, D.W. & P.V. KIRCH 1990. Prehistoric extinction of birds on Mangaia, Cook Islands, Polynesia, Proceedings of the National Academy of Sciences 87: 9605-9.

STRIEWSKI,B., M.B. ELLIOT, J.R. FLENLEY & D.G. SUTTON, 1994. Date of colonization of Northland, Part 1, Year 2. Unpublished report, Massey University.