Hit-or-myth? Linking a 1259 AD acid spike with an Okataina eruption.

Lowe, David J. ; Higham, Thomas F.G.

Introduction: Bronze Age catastrophes and myth-making

In their recent provocative paper, Buckland et al. (1997) examined evidence for two Bronze Age 'catastrophes'. The first, the destruction of Bronze Age Thera (Santorini) by a cataclysmic volcanic eruption, was described as 'real and in need of a calendar date' (p. 581). The second, the apparent collapse of Middle Bronze Age settlement in upland Britain, was considered as a speculative event 'hypothesized on archaeological grounds and dated by a tenuous link through tree rings to an Icelandic volcano' (p. 581). The fundamental purpose of their critique was to demonstrate that great caution is required in the interpretation of interdisciplinary studies that attempt to link archaeological findings with those from other disciplines. This caution is essential because such age-based linkages may enter the literature as if proven fact because the limitations of the data are rarely communicated clearly, either unwittingly or otherwise. This is anathema to archaeology because the distinction between 'fact' and 'interpretation' may not always be obvious to its practitioners.

Whilst we agree in general with their conclusions, we think Buckland et al. have unintentionally violated one of their own tenets by constructing a linkage, based on the assumed correlation of radiometric and ice-core derived dates alone, between the 1259 AD acid spike in ice cores and an Okataina-derived volcanic eruption in New Zealand. Although only a minor part of their paper, the construction of this link by Buckland et al. nonetheless is viewed as the initial step in the mythicizing process they rightly wish to avoid. We demonstrate that such a link is untenable because the age data Buckland et al. applied to the Okataina eruption are flawed, and we suggest that the 'link' needs correcting before a new myth develops. This correction is particularly relevant to archaeological studies in the South Pacific because the Okataina-derived eruptive provides a valuable regional datum in dating New Zealand's exceptionally brief prehistory (Higham & Hogg 1997; Newnham et al. 1998).

The 1259 An acid spike and the Kaharoa eruption, Okataina

The 1259[+ or -]2 AD acid signal is one of the largest recorded in ice cores from Greenland and Antarctica for the past 2000 years (Hammer et el. 1980; Langway et al. 1988; Zielinski et al. 1994). Because the spike is common to ice-core records at both poles, Langway et al. (1988) suggested that the eruption must have been large and equatorial. El Chichon volcano (Mexico) is a possible source (Palais et al. 1992).

Buckland et el, however, attributed the 1259 AD acid spike to a mid-latitude eruption from Okataina volcano in North Island [ILLUSTRATION FOR FIGURE 1 OMITTED]. The 'Okataina' eruption is clearly the Kaharoa episode, the largest and most recent rhyolitic event in New Zealand, which resulted in extensive tephra fallout [ILLUSTRATION FOR FIGURE 1 OMITTED] (Lowe et al. in press a). The basis of Buckland et al.'s correlation with the 1259 AD ice-core acidity record is evidently the derivation of a date of 1259[+ or -]11 AD for the Kaharoa eruption via calibration of associated radiocarbon ages (Ramsey 1994). This calibrated date is based on the mean age of 770[+ or -]20 b.p. reported in Simkin & Siebert (1994) following Froggatt & Lowe (1990). We do not dispute the calibration process per se except to comment that the Southern Hemisphere offset correction (Vogel et al. 1993; McCormac et al. in press) does not seem to have been applied. based on Stuiver & Reimer (1993) and Stuiver & Becker (1993) and the intercepts method, 770[+ or -]20 b.p. corresponds to 1258-1283 AD without the offset, but with a -40-year offset correction the calibrated 1[Sigma] range is 1280-1291 AD, clearly incompatable with 1259[+ or -]11 AD at this level of significance. Of more importance to our discussion is the fact that the 770[+ or -]20 b.p. age has been revised since its original publication in 1990. Consequently it is no longer appropriate for obtaining a calibrated date, irrespective of application of the inter-hemispheric offset correction.

Lowe & Hogg (1992) published four new radiocarbon ages for Kaharoa eruptives that gave a significantly younger mean age of 665[+ or -]17 b.p. An identical mean age of 665[+ or -]15 b.p. was obtained more recently using cluster analysis of 22 radiocarbon ages. This new age, derived from unscreened ages minus outliers, is supported by statistically identical ages obtained from three sets of screened ages selected to minimize the effects of inbuilt age or contamination (Lowe et al. in press a). The calibration curve in the vicinity of 665 b.p. is very wiggly and so even high precision radiocarbon ages translate into a relatively wide range of calendar dates [ILLUSTRATION FOR FIGURE 2 OMITTED]: 665[+ or -]15 b.p. corresponds to 1299-1327 AD and 1350-1391 AD at the 1[Sigma] level, and to 1291-1331 AD and 1343-1399 AD at the 2[Sigma] level (Lowe et al. in press a). These new calibrated dates show that the Kaharoa eruption must have taken place well after the 1259 AD acid spike was emplaced, and thus there can be no link between the two events. In reaching this conclusion we assume the 1259 AD acid signal date is accurate.

Conclusion: avoiding myth-making

That there is demonstrably no connection between the 1259 AD acid spike and the c. 1300-1400 AD Kaharoa eruption has little bearing on the main conclusions reached by Buckland et al., which are forcefully argued and timely (though disputed: Baillie 1998). Our point is that Buckland et al. have violated one of their own tenets - the need for caution in considering possible relationships between events connected by chronological matching alone - by implying that there is a link between the 1259 AD spike and a New Zealand eruption. The risk is that the 'link', unless corrected in the literature, will eventually move along the pathway from 'reasonable speculation' to 'proven fact'. Establishing an accurate date for the Kaharoa eruption (and discrediting a 'wrong' one) is essential for archaeology in New Zealand and potentially other parts of East Polynesia. This is because no prehistoric cultural remains are known to occur beneath the Kaharoa Tephra (Anderson 1991), and because of the critical role the tephra has in dating the earliest human-induced environmental impacts in New Zealand (Newnham et al. in press; Lowe et al. in press b).

The age data in FIGURE 2 imply that a sulphate spike (or other environmental effects) from the Kaharoa eruption may be represented in the ice-core and/or tree-ring records from c. 1290 to 1330 AD or c. 1340 to 1400 AD. The presence of volcanic glass in the same layer as a sulphate signal in an ice core provides the best way of directly identifying the source eruption, as shown for Santorini by Zielinski & Germani (1998). It also provides a potential means of linking the ice-core and tree-ring records (Baillie 1996). Thus, if glass from the compositionally distinctive Kaharoa eruption (Stokes et al. 1992) were identifiable in the ice sheets (e.g. Antarctica), it would provide an important late Holocene chronostratigraphic marker event for global ice-core and tree-ring studies and hence other disciplines including archaeology. In view of its importance, we are attempting to determine a more precise date for the Kaharoa eruption using dendrochronology (Lowe et al. in press a).

Acknowledgements. We are grateful to Greg Zielinski for discussions and a pre-print of Zielinski & Germani (1998}, and to anonymous reviewers for their comments.

References

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

BAILLIE, M.G.L. 1996. Extreme environmental events and the linking of the tree-ring and ice-core records, Radiocarbon 38: 703-11.

1998. Bronze Age myths expose archaeological shortcomings? a reply to Buckland et al. 1997, Antiquity 72: 4257.

BUCKLAND, P.C., A.J. DUGMORE & K.J. EDWARDS. 1997. Bronze Age myths? Volcanic activity and human responses in the Mediterranean and North Atlantic regions, Antiquity 71: 581-93.

FROGGATT, P.C. & D.J. LOWE.1990. A review of late Quaternary silicic and some other tephra formations from New Zealand: their stratigraphy, nomenclature, distribution, volume, and age, New Zealand Journal of Geology and Geophysics 33: 89-109.

HAMMER, C.U., H.B. CLAUSEN & W. DANSGAARD.1980. Greenland ice sheet evidence of post-glacial volcanism and its climatic impact, Nature 288: 230-35.

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.

LANGWAY, C.C., JR, H.B. CLAUSEN & C.U. HAMMER. 1988. An inter-hemispheric volcanic time-marker in ice cores from Greenland and Antarctica, Annals of Glaciology 10: 1028.

LOWE, D.J. & A.G. HOGG. 1992. Application of new technology liquid scintillation spectrometry to radiocarbon dating of tephra deposits, New Zealand, Quaternary International 13114: 135-42.

LOWE, D.J., B.G. MCFADGEN, T.F.G. HIGHAM, A.G. HOGG, P.C. FROGGATT & I.A. NAIRN. IN PRESS A. Radiocarbon age of the Kaharoa Tephra, a key marker for late Holocene stratigraphy and archaeology in New Zealand, The Holocene 8.

LOWE, D.J. R.M. NEWNHAM, T.F.G. HIGHAM, J.M. WILMSHURST, M.S. MCGLONE & A.G. HOGG. In press b. Dating earliest human impact and settlement in New Zealand, in D.G. Sutton (ed.), Origins of the first New Zealanders, 2nd edition. Auckland: Auckland University Press.

MCCORMAC, F.G., A.G. HOGG, T.F.G. HIGHAM, J. LYNCH-STIEGLITZ, W. BROECKER, M.G.L. BAILLIE, J.G. PALMER, L. XIONG, D. BROWN & S.T. HOPER. In press. Interhemispheric difference in 14C: an anthropogenic C[O.sub.2] effect, Geophysical Research Letters.

NEWNHAM, R.M., D.J. LOWE, M.S. MCGLONE, J.M. WILMSHURST & T.F.G. HIGHAM. 1998. The Kaharoa Tephra as a critical datum for earliest human impact in northern New Zealand, Journal of Archaeological Science 25.

PALAIS, J.M., M.S. GERMANI & G.A. ZIELINSKL 1992, Inter-hemispheric transport of volcanic ash from a 1259 AD volcanic eruption to the Greenland and Antarctic ice sheets, Geophysical Research Letters 19: 801-4.

PULLAR, W.A., B.P. KOHN & J.E. COX. 1977. Air-fall Kaharoa Ash and Taupo Pumice, and sea-rafted Loisels Pumice, Taupe Pumice, and Leigh Pumice in northern and eastern parts of the North Island, New Zealand, New Zealand Journal of Geology and Geophysics 20: 697-717.

RAMSEY, C.B.1994. OxCal.v. 2.14: radiocarbon calibration and statistical analysis program. Oxford: Research Laboratory for Archaeology.

SIMKIN, T. & L. SIEBERT.1994. Volcanoes of the world. 2nd edition. Tucson [AZ): Geoscience Press.

STOKES, S., D.J. LOWE & P.C. FROGGATT.1992. Discriminant function analysis and correlation of late Quaternary rhyolitic tephra deposits from Taupe and Okataina volcanoes, New Zealand, using glass shard major element composition, Quaternary International 13/14: 102-17.

STUIVER, M. & B. BECKER. 1993. High-precision decadal calibration of the radiocarbon time scale. AD 1950-6000 BC, Radiocarbon 35: 35-65.

STUIVER, M. & P.J. REIMER. 1993. Extended 14C data base and revised CALIB 3.0 14C age calibration programme, Radiocarbon 35: 215-31.

VOGEL, J.C., A. FULS, E: VISSER & B. BECKER. 1993. Pretoria calibration curve for short-lived samples, 1930-3350 BC, Radiocarbon 35: 73-85.

ZIELINSKI, G.A. & M.S. GERMANI. 1998. New ice core evidence challenges the 1620s BC age for the Santorini (Minoan) eruption, Journal of Archaeological Science 25: 279-89.

ZIELINSKI, G.A., P.A. MAYEWSKI, L.D. MEEKER, S. WHITLOW, M.S. TWICKLER, M. MORRISON, D. MEESE, R.B. ALLEY & A.J. GOW. 1994. Record of volcanism since 7000 BC from the GISP2 Greenland ice core and implications for the volcano-climate system, Science 264: 946-52.

Paul C. Buckland, Andy J. Dugmore & Kevin J. Edwards comment:

We have to admit that the suggestion of the link between the AD 1259 tephra in both Greenland and Antarctic ice cores and an eruption of Okataina, New Zealand, in table I of Buckland et al. (1997), reflected a very deliberate piece of coat-trailing to elicit response. Indeed, between the correction of final proofs and publication of the paper, Greg Zielinski (pers. comm.), on geochemical grounds, had already rendered the connection most unlikely. The eruption in 1259, however, provides one of the few isochrones which is potentially of global significance, and there is an urgent need to track it down to source and to refine tephrochronological techniques to the level of locating it, and many other eruptions, in sediments other than ice. We are particularly pleased to see Lowe & Higham's carefully reasoned response, updating the research and clarifying the situation for the Southern Hemisphere.

To a great extent, Baillie's response is understandable for seeking to bolster the dendrochronological dates for Santorini and Hekla 3, but we, trained as environmental scientists, refute any accusation of protecting archaeologists from environmental determinism. Yet having spent many years researching the relationships been human populations and environments in the North Atlantic region and beyond (cf. Dugmore & Buckland 1991), it is evident that rigid socio-economic constraints can push systems into collapse in the face of relatively minor environmental changes - the case of Norse Greenland, lost without any intercession by volcanoes in the late medieval period [Buckland et al. 1996), is particularly apposite.

We were also aware of the supposed Exodus connection [Bruins & van der Plicht 1996), yet to mix Hebraic with Greek myth only serves further to bolster the modern myth. The most telling comment upon the dating of Santorini still comes from Bietak's (1996) work at Avaris and the Egyptian chronology; a 130-year revision in this well established chronology in order to fit in with a speculative correlation between tree rings, ice cores and eruptions is unlikely, although this does not mean that we should seek to reject this if proven on sound archaeological evidence. Kuniholm et al. (1996) have identified anomalous growth in the drought-stressed trees of Anatolia in the 17th century BC, but the correlation to the Santorini eruption is not direct, and still requires a connecting climatic impact; the mechanism by which this could be achieved is still not clear.

A central point of our argument is that coincidence between tree-ring anomalies and vulcanism does not necessarily prove causal relationships, especially when major eruptions significantly outnumber frost-ring dates. 75% of frost-ring dates shown in Baillie's (1998) table 1 may indeed be associated with large historical eruptions, but, as shown in our table 1 (Buckland et al. 1997), in the period since AD 550 there have been at least 83 eruptions [greater than or equal to] VEI4, where only 31 frost ring events have been identified (a 37% presumed association). This estimate of the number of large eruptions is conservative, and is likely to increase as more tephrochronological work is undertaken in poorly studied volcanic regions. A key point is that the mechanisms that might link eruptions and impacts upon the tree-ring record are not well understood. The type of eruption, the scale, location relative to global circulation systems, the timing and coincidence with other eruptions may all be critical variables affecting the impact of an eruption on climate, even before the ecological response of trees to environmental change is considered.

We certainly have no intent to 'rubbish' the provision of hypotheses, but we are only too aware, as Lowe & Higham have rightly pointed out, that hypotheses rapidly become 'fact' in archaeology. There maybe connections between particular volcanic eruptions, tree rings and acidity spikes in ice cores in prehistory, but we would still hold that, as in Scottish law, most remain 'not proven'.

References

BAILLIE, M. 1998. Bronze Age myths expose archaeological shortcomings? a reply to Buckland et al. 1997, Antiquity 72: 425-7.

BIETAK, M. 1996. Avaris. Capital of the Hyksos. London: British Museum Press.

BRUINS, H.J. & J. VAN DER PLICHT. 1996. The Exodus enigma, Nature 382: 213-14.

BUCKLAND, P.C., A.J. DUGMORE & K.J. EDWARDS. 1997. Bronze Age myths? Volcanic activity and human response in the Mediterranean and North Atlantic region, Antiquity 71: 581-93.

BUCKLAND, P.C., T. AMOROSI, L. BARLOW, A.J. DUGMORE, P. MAYEWSKI, T.H. MCGOVERN, A. OGILVIE, J.P. SADLER & P. SKIDMORE. 1996. Bioarchaeological and climatological evidence for the fate of Norse farmers in medieval Greenland, Antiquity 70: 88-96.

DUGMORE, A.J. & P.C. BUCKLAND. 1991. Tephrochronology and Late Holocene soil erosion in south Iceland, in J.K. Maizels & C. Caseldine (ed.), Environmental change in Iceland:past and present. 147-60. Dordtrecht: Kluwer Academic Press.

LOWE, D.J. & T.F.G. HIGHAM. 1998. Hit-or-myth? Linking a 1259 AD acid spike with an Okataina eruption, Antiquity: 427-31.