Palaeotsunamis and their influence on Polynesian settlement

Introduction

Recent tsunamis in the Pacific are a timely reminder of the devastating impact that such events can have on coastal settlements. The repercussions of the 2011 Tōhoku-oki tsunami have highlighted the necessity for coastal communities to heed the warnings of this event, and of the long history of previous events recorded by geology (Goto et al., 2011; Minoura et al., 2001; Normile, 2011). Like Japan, the islands of Polynesia are similarly exposed to tsunamis originating from the ‘Pacific Ring of Fire’, and other regional sources (Goff et al., 2011a).

In historical times, earthquakes on the Tonga-Kermadec trench (TKT) have generated multiple tsunamis, notably in ad 1865, ad 1917 and ad 2009, the last being a regional event that caused extensive destruction to coastal communities and ecosystems (Goff and Dominey-Howes, 2011; McAdoo et al., 2011; Reese et al., 2011; Richmond et al., 2011). These events indicate that tsunamis in the Pacific may, in human terms, be relatively frequent (Okal et al., 2004), and they provide a reason to more closely examine the record of palaeotsunamis in the region (Goff et al., 2011a, 2011b). The potential influence of palaeotsunamis on Polynesian prehistory has, however, received little attention from archaeologists. Considering that Polynesians were a largely coastal people, who relied on boats for travel (Irwin, 2010), this lack of attention represents a significant gap in our understanding of the region’s cultural past.

Polynesian settlement and palaeotsunamis

There are five key events in the Polynesian settlement chronology:

The two key events that palaeotsunamis may have had some influence on are the ‘long pause’ and the collapse of regional open-sea voyaging in the 15th century. For the latter event, the loss of cultural motivation and the onset of the ‘Little Ice Age’ (Nunn, 2000) have been invoked as an explanation, but these only offer a partial explanation, not a uniquely regional solution. While palaeotsunamis are a potential cause for both the ‘long pause’ in settlement of Eastern Polynesia and later cessation of voyaging, some researchers consider rapid-onset natural hazards such as earthquakes and tsunamis to be too localised in space and time to be important (Anderson, 2009). In their view, such events are recognised as adding a layer of complexity to Pacific human history, but not to have the almost ubiquitous and long-term impacts associated with climate change and anthropogenic modification (Anderson, 2009).

It is not as though palaeotsunamis are unknown in the archaeological record; they are recorded from Tahiti (Sinoto, 1979), New Zealand (McFadgen, 2007), and Hawaii (Burney et al., 2001). Furthermore, oral histories record events that warrant further study; e.g. New Zealand (McFadgen, 2007), the Cook Islands (Beaglehole and Beaglehole, 1938), and Hawaii (Carson and Athens, 2007). While we accept that archaeological and oral historical records are extremely sketchy with regard to palaeotsunamis, we point out that the signatures of palaeotsunamis in the stratigraphic record can be difficult to detect (refer to Goff et al. (2012) for details concerning the identification of palaeotsunami deposits), and oral traditions of such events are rarely clear and unambiguous. For instance, following a tsunami in the Cook Islands in April 2010, it was found that only five months later, stripped vegetation was already growing back, and without prior knowledge that the event had taken place, there was little easily identifiable evidence in the coastal stratigraphy to indicate that the area had been inundated by a tsunami (Goff, 2011). Equally, there were few in the local community who were aware of previous events that occurred as recently as 1997/1998 (Goff, 2011). The identification of older tsunamis in the Pacific is therefore not easy, a situation exacerbated by the fact that few archaeological projects have the assistance of appropriately skilled geoscientists in the field. Indeed, this may well be the primary reason that potential effects of tectonic-related events are so poorly represented in Pacific archaeology.

Given the region-wide impacts of historical tsunamis and the archaeological evidence of their effects, a revision of the view of tectonic hazards as an agent of change in the human history of Polynesia would be timely (Fitzpatrick, 2007; Goff and McFadgen, 2002; Satake and Atwater, 2007; Sinoto, 1979). The growing Pacific palaeotsunami data base will likely prove a rich record of locally, regionally, and distantly sourced events (Goff et al., 2011a). The most prominent records currently held are of significant regional events caused by probable TKT-sourced palaeotsunamis in c. 2800 BP, 1860–2000 BP and c. ad 1450; the last of these includes two events, one perhaps related to the ad 1452/1453 Kuwae eruption (Goff et al., 2011a) (Figure 1). The coincidence in timing of these palaeotsunamis with some of the key events in Polynesian settlement of the Pacific suggests a possible causal relationship in particular, with the long pause, and the cessation of voyaging in the 15th century. Even moderate tsunamis can be extraordinarily destructive, as historic events have recently shown. On the one hand would be the loss of physical assets – dwellings, food supplies, shellfish resources, and floatable artefacts especially canoes – and on the other hand would be the loss knowledge resulting from the deaths of elders, craftsmen, and navigators (McFadgen, 2007). Being largely a coastal people, Polynesian communities would have been extremely vulnerable. We suggest that the effects of tsunamis on the population and its culture, would have been sufficient to cause the cessation of sea voyaging seen in the archaeological record, until the people had recovered the skills and inclination to continue long distance voyaging.

OPEN IN VIEWER

Figure 1.

South and central Pacific Ocean showing estimated areas of 15th century palaeotsunami impact; the extent of the much smaller 2009 South Pacific Tsunami indicates the type of areal coverage that can be expected from larger events (N: Niue; Pp: Pukapuka; R: Rotuma; S: Samoa; T: Tonga; Tv: Tuvalu; W&F: Wallis and Futuna). Details concerning the magnitude of the 15th century palaeotsunamis are given in Goff et al. (2011a) – these range from prehistoric settlement destruction (W&F), deposits laid down up to 42 m a.s.l. (New Zealand) and the remobilisation of coastal dunes (French Polynesia).

Conclusion

We do not propose that all key events of Polynesian settlement are linked with tectonic-related activity, but we do suggest that palaeotsunamis, and the events that generated them, should be considered as factors potentially influencing the start of the ‘long pause’ c. 2800 BP and the 15th century collapse of long-distance voyaging. Whether the c. 1860–2000 BP event affected the duration of the ‘long pause’ is unclear. Contrary to current perceptions, rapid-onset natural hazards such as tsunamis may have had a significant region-wide influence on Polynesian settlement. Further work is needed to assess the veracity of this statement.