terra australis 37
Terra Australis reports the results of archaeological and related research within the south and east of Asia, though mainly
Australia, New Guinea and island Melanesia — lands that remained terra australis incognita to generations of prehistorians.
Its subject is the settlement of the diverse environments in this isolated quarter of the globe by peoples who have maintained
their discrete and traditional ways of life into the recent recorded or remembered past and at times into the observable present.
List of volumes in Terra Australis
Volume 1: Burrill Lake and Currarong: Coastal Sites in Southern New South Wales. R.J. Lampert (1971)
Volume 2: Ol Tumbuna: Archaeological Excavations in the Eastern Central Highlands, Papua New Guinea.
J.P. White (1972)
Volume 3: New Guinea Stone Age Trade: The Geography and Ecology of Traffic in the Interior. I. Hughes (1977)
Volume 4: Recent Prehistory in Southeast Papua. B. Egloff (1979)
Volume 5: The Great Kartan Mystery. R. Lampert (1981)
Volume 6: Early Man in North Queensland: Art and Archaeology in the Laura Area. A. Rosenfeld, D. Horton and
J. Winter (1981)
Volume 7: The Alligator Rivers: Prehistory and Ecology in Western Arnhem Land. C. Schrire (1982)
Volume 8: Hunter Hill, Hunter Island: Archaeological Investigations of a Prehistoric Tasmanian Site.
S. Bowdler (1984)
Volume 9: Coastal South-West Tasmania: The Prehistory of Louisa Bay and Maatsuyker Island. R. Vanderwal
and D. Horton (1984)
Volume 10: The Emergence of Mailu. G. Irwin (1985)
Volume 11: Archaeology in Eastern Timor, 1966–67. I. Glover (1986)
Volume 12: Early Tongan Prehistory: The Lapita Period on Tongatapu and its Relationships. J. Poulsen (1987)
Volume 13: Coobool Creek. P. Brown (1989)
Volume 14: 30,000 Years of Aboriginal Occupation: Kimberley, North-West Australia. S. O’Connor (1999)
Volume 15: Lapita Interaction. G. Summerhayes (2000)
Volume 16: The Prehistory of Buka: A Stepping Stone Island in the Northern Solomons. S. Wickler (2001)
Volume 17: The Archaeology of Lapita Dispersal in Oceania. G.R. Clark, A.J. Anderson and T. Vunidilo (2001)
Volume 18: An Archaeology of West Polynesian Prehistory. A. Smith (2002)
Volume 19: Phytolith and Starch Research in the Australian-Pacific-Asian Regions: The State of the Art.
D. Hart and L. Wallis (2003)
Volume 20: The Sea People: Late-Holocene Maritime Specialisation in the Whitsunday Islands, Central Queensland.
B. Barker (2004)
Volume 21: What’s Changing: Population Size or Land-Use Patterns? The Archaeology of Upper Mangrove Creek,
Sydney Basin. V. Attenbrow (2004)
Volume 22: The Archaeology of the Aru Islands, Eastern Indonesia. S. O’Connor, M. Spriggs and P. Veth (2005)
Volume 23: Pieces of the Vanuatu Puzzle: Archaeology of the North, South and Centre. S. Bedford (2006)
Volume 24: Coastal Themes: An Archaeology of the Southern Curtis Coast, Queensland. S. Ulm (2006)
Volume 25: Lithics in the Land of the Lightning Brothers: The Archaeology of Wardaman Country,
Northern Territory. C. Clarkson (2007)
Volume 26: Oceanic Explorations: Lapita and Western Pacific Settlement. S. Bedford, C. Sand and
S. P. Connaughton (2007)
Volume 27: Dreamtime Superhighway: Sydney Basin Rock Art and Prehistoric Information Exchange.
J. McDonald (2008)
Volume 28: New Directions in Archaeological Science. A. Fairbairn, S. O’Connor and B. Marwick (2008)
Volume 29: Islands of Inquiry: Colonisation, Seafaring and the Archaeology of Maritime Landscapes.
G. Clark, F. Leach and S. O’Connor (2008)
Volume 30: Archaeological Science Under a Microscope: Studies in Residue and Ancient DNA Analysis in
Honour of Thomas H. Loy. M. Haslam, G. Robertson, A. Crowther, S. Nugent and L. Kirkwood (2009)
Volume 31: The Early Prehistory of Fiji. G. Clark and A. Anderson (2009)
Volume 32: Altered Ecologies: Fire, Climate and Human Influence on Terrestrial Landscapes. S. Haberle,
J. Stevenson and M. Prebble (2010)
Volume 33: Man Bac: The Excavation of a Neolithic Site in Northern Vietnam: The Biology. M. Oxenham,
H. Matsumura and N. Kim Dung (2011)
Volume 34: Peopled Landscapes: Archaeological and Biogeographic Approaches to Landscapes. S. Haberle and
B. David. (2012)
Volume 35: Pacific Island Heritage Archaeology: Identity & Community. J. Liston, G. Clark and D. Alexander (2011)
Volume 36: Transcending the Culture-Nature Divide in Cultural Heritage: Views from the Asia-Pacific Region.
S. O’Connor, S. Blackwell and D. Byrne (2012)
terra australis 37
Taking the
High Ground
THE ARCHAEOLOGY OF RAPA,
A FORTIFIED ISLAND IN REMOTE
EAST POLYNESIA
Edited by Atholl Anderson and Douglas J. Kennett
© 2012 ANU E Press
Published by ANU E Press
The Australian National University
Canberra ACT 0200 Australia
Email: anuepress@anu.edu.au
Web: http://epress.anu.edu.au
National Library of Australia Cataloguing-in-Publication entry
Title:
Taking the high ground : the archaeology of Rapa, a fortified island in remote East Polynesia /
edited by Atholl Anderson and Douglas J. Kennett.
ISBN:
9781922144249 (pbk.) 9781922144256 (ebook)
Series:
Terra Australis ; Number 37.
Subjects:
Archaeology--Rapa Island.
Rapa Island--Antiquities.
Other Authors/Contributors:
Anderson, Atholl.
Kennett, Douglas J.
Dewey Number: 996.22
Copyright of the text remains with the contributors/authors, 2011. This book is copyright in all countries subscribing to the Berne
convention. Apart from any fair dealing for the purpose of private study, research, criticism or review, as permitted under the Copyright Act, no part may be reproduced by any process without written permission. Inquiries should be made to the publisher.
Series Editor: Sue O’Connor
Cover image: Tevaitau Fortification on Rapa Island. Photograph taken from south looking north, D. J. Kennett.
Back cover image: Tangarutu Shelter on Rapa Island. Photograph by Atholl Anderson.
Back cover map: Hollandia Nova. Thevenot 1663 by courtesy of the National Library of Australia.
Reprinted with permission of the National Library of Australia.
Terra Australis Editorial Board: Sue O’Connor, Jack Golson, Simon Haberle, Sally Brockwell, Geoffrey Clark
Contents
1
2
3
4
5
6
7
8
9
Archaeological research on Rapa Island, French Polynesia
Atholl Anderson, Douglas J. Kennett and Eric Conte
7
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of
Rapa Island, French Polynesia, AD 1791–1840
Atholl Anderson
25
Archaeology of the coastal sites on Rapa Island
Atholl Anderson
47
The archaeobotany of Rapan rockshelter deposits
Matiu Prebble and Atholl Anderson
77
Cordage from Rapan archaeological sites
Judith Cameron
97
Bird, reptile and mammal remains from archaeological sites on Rapa Island
Alan J. D. Tennyson and Atholl Anderson
105
Prehistoric fishing on Rapa Island
Yolanda Vogel and Atholl Anderson
115
The Tangarutu invertebrate fauna
Katherine Szabó and Atholl Anderson
135
Marine resource exploitation on Rapa Island: Archaeology, material culture
and ethnography
Katherine Szabó, Yolanda Vogel and Atholl Anderson
145
10 Palaeobotany and the early development of agriculture on Rapa Island
Matiu Prebble and Atholl Anderson
167
terra australis 37
11 A Bayesian AMS 14C chronology for the colonisation and fortification
of Rapa Island
Douglas J. Kennett, Brendan J. Culleton, Atholl Anderson and John Southon
189
12 The archaeology of Rapan fortifications
Douglas J. Kennett and Sarah B. McClure
203
13 Rapan agroecology and population estimates
Jacob Bartruff, Douglas J. Kennett and Bruce Winterhalder
235
14 The prehistory of Rapa Island
Atholl Anderson, Douglas J. Kennett and Eric Conte
247
Appendices
terra australis 37
257
1
Archaeological research on Rapa Island,
French Polynesia
Atholl Anderson
Department of Archaeology and Natural History, Research School of Asian and Pacific Studies, The Australian
National University, Canberra, Australia, atholl.anderson@anu.edu.au
Douglas J. Kennett
Department of Anthropology, The Pennsylvania State University
Eric Conte
Université du Polynésie Française
Introduction
This volume describes the results of archaeological and related research on Rapa Island, which
lies at the southern extremity of French Polynesia. Notable for its numerous fortified sites atop
the peaks of a spectacular volcanic landscape, Rapa has remained nonetheless an enigma in
Polynesian prehistory. It has been linked, on the one hand, with its more famous and nearnamesake Rapa Nui (Easter Island), in hypotheses of Amerindian migration and the dire
impacts of deforestation and societal isolation, and, on the other hand, with settlement and fort
construction in the similarly cool and remote nearest neighbour to the southwest, New Zealand.
Our project set out to construct a cultural sequence and palaeoenvironmental context so that
consideration of such issues might profit from the existence of a more diverse and comprehensive
database.
Rapa Island is located in a remote position at 27035S, 144020W, in the subtropical South
Pacific Ocean (Figure 1.1). In the form of a horseshoe, which reflects its origin as a breached
caldera, Rapa has 38 km2 of land area and, lacking any barrier or fringing reef, its outer coast
rises abruptly out of the sea towards jagged peaks, products of erosion around the caldera rim
(Figure 1.2), the highest of which is Mont Perau, at 650 m;
Where the steep sides of the jagged peaks reach the coast they form great cliffs falling vertically to the sea.
The coast is bold with deep caves worn in it by the sea. Except around the bays in the coast, which have
sandy beaches at their heads, the island is inaccessible. (Haslam 1982:83)
terra australis 37
8
Atholl Anderson, Douglas J. Kennett and Eric Conte
Figure 1.1. Rapa Island, showing the main topographical features and modern settlements.
Numbered squares indicate locations of previous research mentioned in this chapter: 1=Anapoiri Cave, 2=Tevaitau,
3=Moronga Uta, 4=Tangarutu Cave, 5=Kapitanga, 6=Potaketake, 7= Tapitanga. See Figures 3.1, 10.1 and 12.1 for the
sites examined during the 2002 field season. Inset, the location of Rapa in the South Pacific Ocean and Austral Island
group. Drafted by R. Van Rossman.
terra australis 37
Archaeological research on Rapa Island, French Polynesia
9
Figure 1.2. (a) The steep outer coast of Rapa near Akananue Bay. (b) Rapan cliffs on the northern coast.
Photographs A. Anderson.
Topographically, the redeeming feature of Rapa for human habitation is its large, protected
harbour with relatively extensive low and fertile ground at its head (Figure 1.3). The harbour
was called Aurai in the 1820s (Chapter 2) and it has been called Ha’urei or Ahurei since the
1940s at least (Naval Intelligence Division 1943:253). However, Stokes (n.d.) called it ‘Tairirau’
and on an early map published by Hall (1868) it is labelled ‘Boukakika’ (Figure 1.4). The two
main settlements of historical times, and today, are located opposite each other on Ha’urei
Harbour. They were written as Harea and Aruhei by Hall (Figure 1.5), and today are Area and
Ha’urei or Aurei respectively. We use Area for the smaller community on the north side of
Ha’urei Bay and Aurei for the larger community, located to the southwest, to distinguish it from
the bay name itself. It is worth noting here, as well, that the spelling of Rapan names is quite
varied because of the historical influence of Tahitian on the Rapan dialect. Thus we get Taga
and Tanga, Angairao and Agairao, Iri and Hiri and so on. Neither we, nor other writers, have
standardised the spelling.
About 83 km to the east-southeast of Rapa lie the bare, uninhabited Marotiri (Bass)
Islets. They have about 2 km2 of surface area, and bear some remains of stone-built structures
(Ferdon 1965b:71). Rapa and Marotiri are part of the Austral archipelago, named by the French
geographer Malte Brun (Ellis 1838:363), but they are sometimes separated out, and once were
known as the Bass group, in recognition of their relative remoteness from the other Austral
Islands (e.g. Naval Intelligence Division 1943; Kooijman 1972; Haslam 1982:82). The nearest
inhabitable island to Rapa is Raivavae, 537 km to the northwest, but the administrative centre
of the Australs is on Tubuai Island, 730 km from Rapa.
terra australis 37
10
Atholl Anderson, Douglas J. Kennett and Eric Conte
Figure 1.3. Ha’urei Bay, showing lowlands, taro plantations in the Tukou region and Tapui Island. Hilltop fortifications
are visible on several of the highest peaks. Photograph D.J. Kennett.
Figure 1.4. Map of Rapa published by Hall (1868:facing p. 80).
terra australis 37
Archaeological research on Rapa Island, French Polynesia
11
Figure 1.5. Plan of the harbour at Rapa, showing the main settlements (Hall 1868:facing p. 76).
The Rapan climate is relatively cool, cloudy and windy. Mean daily maximum temperatures
at the coast vary from 260C in February to 200C in July, about 30C below those elsewhere in the
Australs. Rainfall is high, at 2840 mm per annum, and 156 days a year have more than 70%
cloud cover. Rapa has mainly easterly conditions in the summer (60% of winds from an easterly
quarter in February), but in the winter there are numerous westerlies (32% in July) that create
cool, squally conditions (Haslam 1982:56). Wind directions change every few days and all
winds from the east-northeast through south to the north-northwest are described by Rapans as
‘cold and strong’; the only warm winds are from the north (Stokes n.d.:4).
Taken under French protection in 1867 and annexed in 1881, Rapa was once on the
main shipping route across the Pacific between New Zealand and the Panama Canal. A coaling
station was established in Ha’urei Harbour by the Panama–New Zealand & Australia Royal
Mail Company (PNZARMC) in 1867, but as steamers were replaced by motor vessels, Rapa
became more isolated again. It lacks suitable ground for an airfield, and is thus accessible only by
sea. In recent times, the population has varied around a total of 400, most of it in Aurei village
on the south side of Ha’urei Harbour (Figure 1.6). Cultivation of taro and fishing are the main
subsistence pursuits, and there is also hunting of feral goats and cattle, along with some fowling
along the cliffs, but the economy is subsidised heavily by the French Polynesian exchequer.
Discovering Rapan archaeology
The archaeological significance of Rapa has rested almost entirely upon its dramatic landscape
of sculpted peaks along the mountain rim that almost encircles Ha’urei Harbour. These were
observed as inhabited places during the first European visit in 1791, when they were described
terra australis 37
12
Atholl Anderson, Douglas J. Kennett and Eric Conte
Figure 1.6. Aurei village taken from the site of Ororangi just to the southeast and looking northwest up Ha’urei Bay.
Photograph D.J. Kennett.
as ‘fortified places resembling redoubts’ and likened to New Zealand Maori pa (Vancouver 1791
in Lamb 1984:374). Ellis (1838:364) noted that the fortifications:
are so constructed as to render them impregnable by any means which the assailants could bring against
the besieged. Wars have not been frequent among them, and, when they have existed, have been less
sanguinary than those among the islands to the northward.
By the late 19th century, the Rapan forts were being drawn into a popular discourse about
ancient Polynesian history and ethnology that sought to link the remote islands of Polynesia to
continental origins, either in Asia or South America, notably by reference to megalithic remains.
This focused, of course, on Easter Island, but the apparent linguistic, and implicitly ancestral,
connection of Rapa-nui (Big Rapa), as Easter Island was also known, to Rapa-iti (Little Rapa),
as Rapa was then commonly called, coupled with massive structural remains in both places,
encouraged speculation about ancient cultural connections between the two (see Smith 1910).
Captain John Vine Hall (1868:133) published a plan and drawings of two Rapan forts
(Figures 1.7 and 1.8). The source of this material is unclear. He was on Rapa for only two
days and did not visit any of the forts (Hall 1868:134). His information about them, and
indeed most of his information about Rapa generally, seems to have come from his colleague
in the PNZARMC, Captain McKellar, superintendent of the coaling station. Hall reported,
nevertheless, that:
On the summits of many of the steep hills are to be seen these square fortresses, some of very elaborate
construction. But what is very singular, they are mostly solid within. The stones are well squared, of very
large size, and well cemented. (Hall 1968:133)
terra australis 37
Archaeological research on Rapa Island, French Polynesia
13
Professor MacMillan Brown visited Rapa in 1917 (Best 1975:415), climbed up to the
highest fort and found it a disappointingly rudimentary structure of small lava slabs, ‘which had
by long infiltration of some adhesive element almost solidified into the appearance of a natural
rock’. Yet, distance lent enchantment to his view of others, for, like Captain Hall, he saw:
… away in the distance and across the harbour … the outlines of much more elaborate forts,
with terrace upon terrace, parapets and moats, and in almost every case there rose above all a
similar solid tower which was evidently meant as an outlook into the valleys beyond. These forts
were, in short, specimens of the rude megalithic masonry that I afterwards saw in the Society
islands, and still more in the Marquesas, consisting of great stones roughly squared to fit their
place without cement. (MacMillan Brown in Best 1975:415)
In 1921, the Routledges visited Rapa for 10 days while on a cruise of French Polynesia,
with the object of finding traces of culture analogous to that of Easter Island (Routledge and
Routledge 1921:438). They visited eight of the forts (known as pare, and recorded by them as
palé) and described their general construction, noting that, far from being structures of stone
blocks and cement, they were:
… in some cases little more than excavated terraces, of the nature of pure earthworks, but usually the
summit of the hill has been turned into a round tower or keep. In such cases, the rock is sometimes
utilized with no addition, or only a small amount of walling, but where this is not adequate the core has
been faced … with a dry masonry of basaltic fragments … in all cases [the towers] are solid, and never
contain apartments. (Routledge and Routledge 1921:454–455)
The pare each occupied one to three acres (0.4–1.2 ha) in extent and consisted of flights
of terraces, up to six in a series, formed by simple cutting or by cut and fill with stone facings.
Ditches restricted access into the fort. It was not clear to what extent the pare had been permanent
villages, for:
On the terraces of the forts there were never found buildings, nor foundations of buildings, nor even
subterranean shelters worthy of the name; the only depression in the ground had evidently been excavated
for purposes of cooking. Huts made of grass only would, no doubt, have disappeared, but would be
peculiarly unsuitable for so exposed a position. It seems most probable that the forts were designed only
to serve as a refuge for the inhabitants of the valley below in case of sudden raid or emergency. (Routledge
and Routledge 1921:455)
As John Stokes (Chapter 2) was already at work on Rapa, the Routledges abandoned their
plan to work there and moved to Mangareva. Stokes’ (n.d.) notes and drafts for his proposed
Bishop Museum monograph on Rapa show that he spent considerable time visiting the pare
and locating other archaeological sites. The salient details of his work are noted in Chapters 2,
3 and 12.
Using Stokes’ notes, Buck summarised the construction of pare thus:
A ridge with a peak was selected and the summit levelled off to form the topmost terrace. The sides
were cut down with digging implements of pointed wood and rude adzes of dyke basalt, until a second
terrace could be formed of sufficient width to accommodate houses. The military architects of the day
continued the plan of successive terraces, which necessitated high walls at the back. The razorback on the
ridge leading to the peak was levelled off and the sides were cut to increase the steepness against assault.
Deep ditches were cut across the main ridge on either side of the citadel to improve the defences. On the
secondary ridges leading up to the main fort, further terraces were dug to provide house accommodation
and outposts for defence. The back walls of the terraces, particularly near the citadel, were further
terra australis 37
14
Atholl Anderson, Douglas J. Kennett and Eric Conte
Figure 1.7. Drawing and plan of Markatea
pare (Hall 1868:facing p. 80).
Figure 1.8. Sketch of Poakoutakataka fort
(Hall 1868:facing p. 80).
terra australis 37
Archaeological research on Rapa Island, French Polynesia
15
reinforced with stone slabs carefully built in to protect the earth face against the detrition of wind and
rain. Projecting stones formed footholds by which the defenders could retreat from terrace to terrace …
On the topmost terrace of the citadel peak the high chief resided. (Buck 1954:183)
He went on to propose that, ‘the peculiar geographical formation’ of Rapa had induced the
development of fortifications and, alluding to the similar pa of New Zealand, which he knew well,
suggested that if he had managed to visit Rapa he might, ‘have sensed an affinity that personal
contact may convey with more subtlety than the written words of others’ (Buck 1954:184). In
later discussion, Buck (1954:288) implies a developmental sequence of fortifications in which
the Rapan structures occupy a middle ground between ditched enclosures, as in Tonga, and more
complex and permanently occupied Maori pa that combined terrace and ditch with palisades.
Earlier archaeological research
Although Stokes had cleared part of the pare Morongo Uta and put spade pits into a number of
sites, the first systematic archaeological research on Rapa was by the Norwegian Archaeological
Expedition to Easter Island and the East Pacific (1955–1956) organised by Thor Heyerdahl
(Heyerdahl and Ferdon 1961, 1965). Ferdon (1965:12–13) recorded stone structures above
the shore in Ha’urei Bay and surveyed the pare at Potaketake, Tapitanga and Kapitanga. Smith
(1965a) investigated stone-outlined features which he thought had been houses and, from an
exposed oven in a small terrace (R-16) at the head of Ha’urei Harbour, he took a charcoal
sample that was radiocarbon dated to ca. AD 1337 (M-707), the earliest date obtained on Rapa
at that time (Smith 1965a:83). Anapoiri, a burial cave, and an oven site were recorded in Mai’i’
Bay. Historical and modern burials were also investigated by Smith (1965b), with scant regard
to propriety or cultural sensitivity.
The main research was on the large pare, Morongo Uta (Mulloy 1965). This is a complex
structure occupying several ridges and their junction. It consists mainly of terraces covering an
area of about 5400 m2:
For the most part they were constructed by excavating horizontal floors into the steep slopes and shoring
up the front sides of these with dry masonry walls, or planing the ridge-tops and surrounding rectangular
areas with masonry. On most of these terraces presumably stood houses of perishable materials. (Mulloy
1965:23)
The entire site was cleared, overburden – generally from terraces higher up – was removed
and many of the displaced stones were used to repair damaged terraces. Mulloy (1965:25)
determined that the site had been built from its core outward, beginning with the tower, and
suggested that it was first constructed as a fort and then developed later into a fortified village.
Basalt prisms were used preferentially in walls and facings, but as these became locally scarce,
other stones were used as well. Low walls or parapets lined the outer edges of some terraces and
other low walls acted as terrace partitions. The terraces or enclosures, approximately 85 in all
(Mulloy 1965:53, cf. Stokes n.d., Chapter 2), are assumed to have been platforms for perishable
houses. Fourteen stone-kerbed fireplaces, complete with ash and charcoal, were found, but the
very wet weather that prevailed for much of the short field season precluded a systematic search
for post holes.
Recovered on the terraces were 60 each of adzes (whole and fragmentary) and pounders.
There were also some anvils and polishing stones, several probable sling stones and the charred
remains of a net or netting bag. Mulloy (1965:57) commented on the scarcity of evidence for
storage of food or water, and of deep midden deposits, suggesting that occupancy might have
terra australis 37
16
Atholl Anderson, Douglas J. Kennett and Eric Conte
been temporary. Two samples produced radiocarbon dates of 310 ± 300 bp and 210 ± 200 bp
(Mulloy 1965:59), which are essentially modern. They lend support to the Rapan belief that the
site was occupied until the European era.
In summary, Ferdon (1965) proposed that the pare, as all the fortified peaks can be called
(Ferdon (165:69) saw little in the distinction made by Stokes (n.d.) between fortified villages and
refuge villages), were all built quite rapidly by large groups of organised labour and according
to fairly simple architectural principles founded on the idea of a central fort surrounded by
defended house terraces with defensive ditches externally and internally, and outlying clusters
of undefended or lightly defended settlement (auga). Within the pare, small niches in the back
walls of terraces held two lines each of four to seven small vertical stones, and a few similar
features were recorded out on the open house terraces. These are almost certainly remains of
shrines (Ferdon 1965:74–75).
The Rapan results of the Norwegian expeditionary research were soon brought into
Heyerdahl’s more general thesis of Polynesian origins. The pare were seen initially as ‘mysterious
pyramids’ (Heyerdahl 1958:288) and, on excavation of Morongo Uta, as fortified villages,
once occupied by oval huts, ‘suspiciously reminiscent of Easter Island’ (Heyerdahl 1958:341).
Heyerdahl thought that the pare and other signs of habitation on the ridge-tops were evidence
that the Rapan population feared an external enemy, perhaps from Easter Island, and that having
‘built themselves curved reed houses and rectilinear stone ovens … instead of rectilinear houses
and round earth ovens as on all the islands in the neighbourhood’ (Heyerdahl 1958:343), they
had come from Easter Island as refugees at the time of the traditional wars. It was implicit in
this view that pare were built more or less at once by the colonising population, rather than later
in the settlement sequence.
In 1984, in the course of a French survey of the Rapan environment, a radiocarbon date
of 370 ± 60 bp was obtained from Tangarutu Cave, in Anarua Bay (Walczak 2001:250). In
1997, a French doctoral student, Jérôme Walczak (2001:297–303, 2003) undertook some small
excavations at the site (Chapter 3). He also surveyed the remains of a number of the abandoned
village sites and other structures on the low ground around Ha’urei Bay, in particular, which,
added to some similar work by the Norwegian expedition, has helped to balance the early
emphasis on forts.
However, the focus of Walczak’s (2001, 2003) research in Rapa remained the nature of
the pare: whether they were simply refuge forts necessitated by chiefly competition driven
by population growth, which was the prevailing model, as articulated by Hanson (1974), or
whether they had more complex functions, including ritual significance. Walczak discusses the
latter proposition:
Ces pare peuvent selon nous être assimilés à des villages d’altitude. Cet terrasses sont dominées par
un promontoire central, qui pourrait être un centre religieux comparable à ce quétaient les marae dan
d’autres îles de la Polynésie orientale. (Walczak 2003:30)
Assuming that the relatively young age of occupation at Tangarutu represents the era of
early settlement, then the age of pare construction cannot have been much later, in which case,
he concludes, it is unlikely to have been compelled by population pressure, and more probably
reflects the needs of ritual use.
Enlarging on this theme, Walczak emphasises the difference between public religious cults,
sustained largely by chiefs, and domestic religious observances, and proposes that it was not so
much depopulation per se, but chiefly rejection of the public cults that led to post-European
abandonment of those pare, or the higher levels of pare, that were associated with public ritual
terra australis 37
Archaeological research on Rapa Island, French Polynesia
17
functions, while some pare or parts of them, remained in use until the later 19th century by
the common people; Tevaitau being one (Walczak 2001:276). He concludes that pare were
essentially symbolic in function, perhaps modelled on the form of the burial island of Tapui in
Ha’urei Harbour. Consequently:
Grâce à cette hypothèse, nous pouvons expliquer leur unité de style architectural, leur situation surélevée,
leur configuration en réseau – voire en complexes – et l’absence de murallie ou de fossé protecteur.
(Walczak 2001:277)
There are, therefore, several competing hypotheses concerning the origins and functions
of fortified settlements in Rapa, as there are elsewhere, and an implication that they may have
occupied more complex roles in Rapan society than purely defence.
The 2002 expedition
Although previous research had produced substantial information about Rapan archaeology,
notably about the hill forts and associated material culture, there was little context in 2002 to
interpret these impressive sites. Indeed, the archaeology of Rapa was essentially a story about
castles in the air; coastal occupation and the concerns of life in the valleys had been barely
noticed, much less investigated. The age of Rapan colonisation was unknown, the age of fort
construction was largely unknown, the sequence of settlement pattern was based on a plausible
proposition, but one that remained untested, and the human impact on the island environment
and its possible consequences for settlement pattern and social development remained
unexplored. Each of these issues is important, not only to understanding the prehistory of
Rapa, but also in the wider contexts of East and South Polynesian prehistory. The project was
organised, consequently, into three research areas: island colonisation, the origins of fortification,
and trends in environmental change. This had the coincidental advantage of dividing fieldwork
into the three main zones, respectively, of the Rapan landscape: the bays and harbour, the high
ground, and the valleys and swamps.
Island colonisation
The key issue of East and South Polynesian colonisation, though far from the only one, was
the chronology. Earlier research (see summary by Kirch 1986; Kirch and Kahn 2007) had
reached a consensus that colonisation began in the early first millennium AD, if not earlier,
in tropical East Polynesia and there were arguments for exploration of New Zealand about the
same time (Sutton 1987; Holdaway 1999). Yet, renewed chronological research on samples
from the same provenances of key sites regarded as early, such as Haatuatua and Hane in the
Marquesas (Rolett and Conte 1995; Anderson and Sinoto 2002), Vaitootia-Faahia and Maupiti
in the Societies (Anderson et al. 1999; Anderson and Sinoto 2002), South Point in Hawaii
(Dye 1992), Anakena on Easter Island (Steadman et al. 1994) and Papatowai and Wairau Bar
in New Zealand (Anderson and Smith 1992; Higham et al. 1999), produced manifestly later
dates, suggesting no colonisation of East Polynesia before about AD 900–1000 (Anderson and
Sinoto 2002), nor of South Polynesia before AD 1200 (Anderson 2000, 2006), or possibly
later (Wilmshurst et al. 2011). As Rapa is both unusually remote and well south of central East
Polynesia, it was a moot point whether its colonisation chronology was closer to that of South
or that of East Polynesia, an issue that has interesting implications for early migration pace and
patterning.
In addition to chronology, we wanted to know about the behaviour of the early settlers.
Where did they live on Rapa and was there evidence of houses and other structures? Did
terra australis 37
18
Atholl Anderson, Douglas J. Kennett and Eric Conte
agriculture begin with colonisation or somewhat later? Was there evidence of faunal depletion or
extinction, of land birds especially? Were turtles, marine mammals, and large fish and shellfish
proportionately more common in early habitation levels? Were the rather distinctive Rapan
adzes an early development? Was there any evidence of distant contacts, in the form of obsidian
or other exotic materials or artefacts?
Based on research in other Pacific Island groups, we predicted that early colonists would
have preferred settling on the coast in relatively sheltered bays with easily accessible fishing,
reliable freshwater sources and near prime agricultural lands. We focused first on Ha’urei Bay,
which stands out as meeting those criteria and which is, of course, the preferred locus of modern
habitation and gardening. We travelled around the outside coast of Rapa in several expeditions
by open boat, and also by walking over the ridges from the harbour, in order to investigate each
bay for potential evidence of early settlement (Chapter 3). The results of this research, almost
entirely around the coast, are described in Chapters 3–9.
Origins of fortification
Competition for resources is an important driving force in the development of social and political
complexity (Kennett and Kennett 2000) and it has been argued repeatedly that it played a crucial
role in the development of Pacific Island chiefdoms (Kirch 1984; Rechtman 1992; Ladefoged
1993; Field 1998, 2004; Kuhlken 1999). Fortified hilltop villages in remote Oceania provide
the most obvious archaeological evidence for warfare prior to European contact and indicate
that inter-village conflict was an important component of social and political life (Green 1967;
Best 1993; Kirch 1994; Burley 1998), although recently a more nuanced understanding of the
complex roles of fortified sites has begun to emerge (Walczak 2001; Sutton et al. 2003).
The hyper-fortified nature of Rapa has often been used as an example of Polynesian
inter-village hostilities (Kirch 1984:212). But these fortifications had not been investigated in
sufficient depth to understand their age, origins or development. The archaeological studies
by the Norwegian expedition (Heyerdahl and Ferdon 1965) revealed the remains of house
platforms, hearths and habitation debris at the pare, but a basic chronological framework,
necessary to understand the development of these villages, was lacking. In addition, artefact
analysis was confined to large objects and no detailed midden constituent work was conducted.
The primary objective of our fieldwork was to recover organic material suitable for
constructing a radiocarbon chronology of fortification on Rapa. We mapped and tested (auger
probes and small sample test units) 14 hilltop fortifications to recover organic material for AMS
radiocarbon dating (see Chapters 11 and 12). We also excavated larger test units at four hilltop
village sites (Tevaitau, Ororangi, Potaketake and Tapitanga) to obtain additional information
about the distribution of artefacts and ecofacts (Chapter 12). IKONOS and Quickbird satellite
imagery and a laser transit were used to map the domestic and defensive features at these
locations. Based on previous excavations at Morongo Uta, it appeared that the deposits were
relatively shallow (ca. 1 m, Mulloy 1965). A standard bucket auger was used to probe subsurface
deposits for concentrations of organic material (shell or charcoal) necessary for AMS radiocarbon
dating. When concentrations of material were encountered, small test units (50 x 50 cm, larger
if necessary) were excavated to obtain samples from stratigraphically intact deposits.
Palaeoenvironmental change
Recovery and analysis of sedimentary cores for inorganic and organic materials indicative of
former environments is a long-established and well-proven technique in palaeoenvironmental
research. Debate about its application to Polynesian issues (e.g. Kirch and Ellison 1993; Parkes
terra australis 37
Archaeological research on Rapa Island, French Polynesia
19
1998; McGlone and Wilmshurst 1999; Kirch 2007 contra Anderson 1994, 1995; Prebble
2007) highlighted some problems in interpretation and dating which have been addressed in
more recent research. As Rapa was once covered by temperate rainforest but at some point in
the prehistoric past became almost completely denuded of forest, pollen analysis was the most
suitable approach to that sequence. Pollen analysis of shifts in plant communities provides a
basis for the detection of anthropogenic influence in deforestation and provides a proxy for the
history of agricultural expansion (Hope 1996; Hope and Pask 1998).
The general strategy of the Rapan project in this matter was to focus on the wetlands around
the head of Ha’urei Harbour, on the assumption that they, as the largest and most readily
accessible area of coastal flats, would have been the first choice for introduction of agriculture
and then for later development of intensive taro cultivation. Other bays where wetlands were
cored included Angatakuri, Iri, Anarua, Pariati, Agairao and Akatanui. The methods and results
of our field and analytical research are described in Chapter 10.
Limitations
The data reported here come from small samples, particularly the archaeological data that arose
from excavations, which, in other circumstances, might be regarded as hardly more than test
pits. Another field season, at least, would have been desirable in several respects: to extend
excavations in the main coastal site at Tangarutu, to investigate terraces and other features,
especially around the harbour, which appear to represent former villages and isolated houses,
to extend excavations on the extensive pare sites, and to undertake systematic research on the
agricultural systems. However, as the project occurred at a time when archaeological funding
for Polynesian projects was especially scarce and the attention of the principal investigators was
being drawn elsewhere, this was just not possible. This monograph, then, should be regarded as
an interim description of the prehistory of Rapa.
Acknowledgements
The field team (Figure 1.9), consisted of Atholl Anderson and Matiu Prebble (The Australian National
University), Douglas Kennett, Yann Doignan, Sarah McClure and Nathan Wilson (University of
Oregon), and Roti (Rosine) Oitokaia (Université de Polynésie Française). Eric Conte (Université de
Polynésie Française) was involved in the development of the project and organised the expedition travel
from and to Tahiti, but he was unable to participate in the fieldwork. Including several days in Papeete,
and also in Tubuai, the expedition was from July 1 to August 25, 2002, of which 43 days were spent
on Rapa.
The expedition was funded primarily by the National Geographic Society (Grant 7059–01),
to which our gratitude is due, and we also thank the universities represented for their supporting
subventions. For a research permit, and permission to remove material temporarily from French
Polynesia, we thank the Mission d’Aide Financière et de Coopération Régionale of the HautCommissariat de la République en Polynésie Française, and also Martine Rattinassamy (Le Chef de
Service) and Tamara Maric of the Service de la Culture et du Patrimoine in Tahiti.
On Rapa, we received permission to undertake our research from the Mayor and Council. Our
stay was made both profitable and pleasurable by our generous hosts, Annette and Freddy Riaria and
Faraire Cerdan. For assistance in particular matters, we thank Freddy Riaria for arranging travel by boat
around the island, and Teraura Oitokaia for information on traditional fishing, fowling and shellfishing.
terra australis 37
20
Atholl Anderson, Douglas J. Kennett and Eric Conte
For collaboration in post-fieldwork research we thank: Alan Tennyson (Te Papa, National Museum of
New Zealand), Yolanda Vogel (New Zealand Historic Places Trust and University of Otago), Katherine
Szabó (ANU and University of Wollongong), Judith Cameron (ANU), Jacob Bartruff (University of
Oregon), Virginia Butler (Portland State University), Bruce Winterhalder (University of California,
Davis) and John Southon (University of California, Irvine). We are also grateful for Rusty Van
Rossman’s assistance with drawings, graphics and maps, and for comment on the manuscript by Terry
Hunt (University of Hawaii at Manoa), and Richard Walter (University of Otago).
Figure 1.9. The 2002 Rapan field team on Tubuai Island before embarking for Rapa. Left to right, back row Atholl
Anderson, Yann Doignan, Douglas J. Kennett, Matiu Prebble; front row Nathan Wilson, Sarah B. McClure, Rosine Oitokaia.
terra australis 37
Archaeological research on Rapa Island, French Polynesia
21
References
Anderson, A.J. 1994. Palaeoenvironmental evidence of island colonization: a response. Antiquity
68:845–847.
Anderson, A.J. 1995. Current approaches in East Polynesian colonization research. Journal of the
Polynesian Society 104:110–132.
Anderson, A.J. 2000. The advent chronology of south Polynesia. In: Wallin, P. and Martinsson-Wallin,
H. (eds), Essays in honour of Arne Skjolsvold 75 years. Occasional Papers of the Kon-Tiki Museum
5:73–82.
Anderson, A.J. 2002. Faunal collapse, landscape change and settlement history in Remote Oceania
World Archaeology 33:375–390.
Anderson, A.J. 2006. Retrievable time: prehistoric colonization of South Polynesia from the outside in
and the inside out. In: Ballantyne, T. and Moloughney, B. (eds), Disputed Histories: imagining New
Zealand’s pasts, pp. 25–41. University of Otago Press, Dunedin.
Anderson, A.J. and Smith, I.W.G. 1992. The Papatowai site: new evidence and interpretations. Journal
of the Polynesian Society 101:129–158.
Anderson, A.J., Conte, E., Clark, G.R., Sinoto, Y. and Petchey, F.J. 1999. Renewed excavations at
Motu Paeao, Maupiti Island, French Polynesia: preliminary results. New Zealand Journal of
Archaeology 21:47–66.
Anderson, A.J. and Sinoto, Y.H. 2002. New radiocarbon ages of colonization sites in East Polynesia.
Asian Perspectives 41:242–257.
Best, E. 1975. The Pa Maori. Government Printer, Wellington.
Best, S. 1993. At the Halls of the Mountain Kings. Fijian and Samoan Fortifications: Comparison and
Analysis. The Journal of the Polynesian Society 102(4):385–447.
Buck, Sir P.H. 1954. Vikings of the Sunrise. Whitcombe and Tombs, Christchurch.
Burley, D.V. 1998. Tongan Archaeology and the Tongan Past, 2850–150 B.P. Journal of World Prehistory
12(3):337–392.
Dye, T. 1992. The South Point radiocarbon dates thirty years later. New Zealand Journal of Archaeology
14:89–97.
Ellis, W. 1838. Polynesian Researches, during a residence of nearly eight years in the Society and Sandwich
Islands. Volume III, Fisher, Son and Jackson, London.
Ferdon, E.N. 1965a. Report 2: A reconnaissance survey of three fortified hilltop villages. In: Heyerdahl,
T. and Ferdon, E.W. (eds), Reports of the Norwegian Archaeological Expedition to Easter Island and
the East Pacific, Volume 2 Miscellaneous Papers, pp. 9–21. Monographs of the School of American
Research and the Kon-Tiki Museum, 24 Pt. 2, Esselte AB, Stockholm.
Ferdon, E.N. 1965b. Report 4: A summary of Rapa Iti fortified villages. In: Heyerdahl, T. and Ferdon,
E.W. (eds), Reports of the Norwegian Archaeological Expedition to Easter Island and the East Pacific,
Volume 2 Miscellaneous Papers, pp. 69–87. Monographs of the School of American Research and
the Kon-Tiki Museum, 24 Pt. 2, Esselte AB, Stockholm.
Field, J.S. 1998. Natural and Constructed Defenses in Fijian Fortifications. Asian Perspectives 37:32–58.
Field, J.S. 2004. Environmental and climatic considerations: a hypothesis for conflict and the emergence
of social complexity in Fijian prehistory. Journal of Anthropological Archaeology 23:79–99.
Green, R.C. 1967. Fortification in Other Parts of Tropical Polynesia. New Zealand Archaeological
Association Newsletter 10:96–113.
Hall, J.V. 1868. On the island of Rapa. Transactions of the New Zealand Institute 1:128–134.
Hanson, F.A. 1970. Rapan Lifeways: society and history on a Polynesian island. Little and Brown, Boston.
Haslam, D.W. 1982. Pacific Islands Pilot, Volume III. Hydrographer of the Navy, Taunton.
Heyerdahl, T. 1958. Aku-Aku: The Secret of Easter Island. Allen and Unwin, London.
Heyerdahl, T. and Ferdon, E.W. (eds), 1965. Reports of the Norwegian Archaeological Expedition to
Easter Island and the East Pacific, Volume 2 Miscellaneous Papers. Monographs of the School of
American Research and the Kon-Tiki Museum, 24 Pt. 2, Esselte AB, Stockholm.
terra australis 37
22
Atholl Anderson, Douglas J. Kennett and Eric Conte
Higham, T.G.F., Anderson, A.J. and Jacomb, C. 1999. Dating the First New Zealanders: the
chronology of Wairau Bar. Antiquity 73:420–427.
Holdaway, R.N. 1999. A spatio-temporal model for the invasion of the New Zealand archipelago by
the Pacific rat Rattus exulans. Journal of the Royal Society of New Zealand 29:91–105.
Hope, G.S. 1996. Quaternary change and historical biogeography of Pacific Islands. In: Keast, A.
and Miller, S.E. (eds), The Origin and Evolution of Pacific Island Biotas, New Guinea to Eastern
Polynesia. SPB Publishing, Amsterdam.
Hope, G.S. and Pask, J. 1998. Tropical Vegetational Change in the late Pleistocene of New Caledonia.
Palaeogeography, Palaeoclimatology, Palaeoecology 142:1–21.
Kennett, D.J., Anderson, A.J., Prebble, M. and Conte, E. 2003. La colonization et les fortifications de
Rapa. In: Marchesi, H. (ed), Bilan de la recherché archeologique en Polynesie francaise, pp. 165–170.
Dossier d’Archéologie Polynésienne 2. Service de la Culture et du Patrimoine, Punaauia.
Kennett, D., Anderson, A.J., Prebble, M., Conte, E. and Southon, J. 2006. Prehistoric human impacts
on Rapa, French Polynesia. Antiquity 80:1–15
Kennett, D.J. and Kennett, J.P. 2000. Competitive and Cooperative Responses to Climatic Instability
in Coastal Southern California. American Antiquity 65:379–395.
Kirch, P.V. 1984. The Evolution of the Polynesian Chiefdoms. Cambridge University Press, Cambridge.
Kirch, P.V. 1986. Rethinking East Polynesian Prehistory. Journal of the Polynesian Society 95:9–40.
Kirch, P.V. and Ellison, J. 1994. Palaeoenvironmental evidence for human colonization of remote
Oceanic islands. Antiquity 68:310–321.
Kirch, P.V. and Kahn, J.G. 2007. Advances in Polynesian Prehistory: Assessment of the Past Decade
(1993–2004). Journal of Archaeological Research 15:191–238.
Kooijman, S. 1972. Tapa in Polynesia. B.P. Bishop Museum Bulletin 234, Honolulu.
Kuhlken, R. 1999. Warfare and Intensive Agriculture in Fiji. In: Gosden, C. and Hather, J. (eds), The
Prehistory of Food: Appetites for Change. Routledge: Taylor and Francis Group, New York.
Ladefoged, T.N. 1993. Evolutionary Process in an Oceanic Chiefdom: Intergroup Aggression and
Political Integration in Traditional Rotuman Society. Unpublished PhD Dissertation, University of
Hawaii.
Lamb, W.K. (ed), 1984. George Vancouver, A Voyage of Discovery to the North Pacific Ocean and Round
the World 1791–1795, Volume I. The Hakluyt Society, London.
McGlone, M.S. and Wilmshurst, J.M. 1999. Dating initial Maori environmental impact in New
Zealand Quaternary International 59:5–16.
Mulloy, W. 1965. Report 3: The fortified village of Morongo Uta. In: Heyerdahl, T. and Ferdon,
E.W. (eds), Reports of the Norwegian Archaeological Expedition to Easter Island and the East Pacific,
Volume 2 Miscellaneous Papers, pp. 23–68. Monographs of the School of American Research and
the Kon-Tiki Museum, 24 Pt. 2, Esselte AB, Stockholm.
Naval Intelligence Division 1943. Pacific Islands. Volume II Eastern Pacific. B.R. 519 B (Restricted)
Geographical Handbook Series. United States Navy, Washington DC.
Parkes, A. 1998. Environmental change and the impact of Polynesian colonization: sedimentary records
from central Polynesia. In: Kirch, P.V. and Hunt, T.L. (eds), Historical Ecology in the Pacific Islands:
prehistoric environmental and landscape change, pp. 166–199. Yale University Press, New Haven.
Paulding, H. 1831. Journal of a cruise of the United States Schooner Dolphin, among the islands of the
Pacific Ocean etc. G. & C. & H. Carvill, New York.
Prebble, M.J. 2005. Islands, Floras and History: an environmental history of plant introductions and
extinction on the Austral Islands, French Polynesia. Unpublished PhD dissertation, The Australian
National University, Canberra.
Rechtman, R.B. 1992. The Evolution of Sociopolitical Complexity in the Fiji Islands. Unpublished
PhD dissertation, University of California, Los Angeles.
Rolett, B.V. and Conte, E. 1995. Renewed investigation of the Haatuatua Dune (Nuku Hiva,
Marquesas Islands): a key site in Polynesian prehistory. Journal of the Polynesian Society 104:195–228.
terra australis 37
Archaeological research on Rapa Island, French Polynesia
23
Routledge, S. and Routledge, K. 1921. Notes on some archaeological remains in the Society and Austral
Islands. Journal of the Royal Anthropological Institute of Great Britain and Ireland 51:438–455.
Smith, C.S. 1965a. Report 5: Test excavations and surveys of miscellaneous sites on the island of Rapa
Iti. In: Heyerdahl, T. and Ferdon, E.W. (eds), Reports of the Norwegian Archaeological Expedition to
Easter Island and the East Pacific, Volume 2 Miscellaneous Papers, pp. 77–87. Monographs of the
School of American Research and the Kon-Tiki Museum, 24 Pt. 2, Esselte AB, Stockholm.
Smith, C.S. 1965b. Report 6: The burial complex on the island of Rapa Iti. In: Heyerdahl, T. and
Ferdon, E.W. (eds), Reports of the Norwegian Archaeological Expedition to Easter Island and the
East Pacific, Volume 2 Miscellaneous Papers, pp. 89–95. Monographs of the School of American
Research and the Kon-Tiki Museum, 24 Pt. 2, Esselte AB, Stockholm.
Smith, S.P. 1910. Easter Island (Rapa-Nui) and Rapa (Rapa-Iti) Island. Journal of the Polynesian Society
19:171–175.
Steadman, D., Vargas, P. and Cristino, C. 1994. Stratigraphy, chronology and cultural context of an
early faunal assemblage from Easter Island. Asian Perspectives 33:79–96.
Stokes, J.F.G. n.d. Ethnology of Rapa Island. Unpublished draft of B.P. Bishop Museum Bulletin of the
Bayard Dominick Expedition. Five unbound volumes with multiple pagination (page numbers
used here are generally the circled numbers). B.P. Bishop Museum, Honolulu.
Sutton, D.G. 1987. A paradigmatic shift in Polynesian prehistory: implications for New Zealand. New
Zealand Journal of Archaeology 9:135–155.
Sutton, D.G., Furey, L. and Marshall, Y. 2003. The Archaeology of Pouerua. Auckland University Press,
Auckland.
Walczak, J. 2001. Le peuplement de la Polynésie orientale. Une tentative d’approche historique par les
exemples de Tahiti et de Rapa (Polynésie française). Unpublished PhD dissertation, University of
Paris I – Pantheon Sorbonne, Paris.
Walczak, J. 2003. Presentation des données actuelles sur la préhistoire de Rapa Iti (archiple des
Australes-Polynésie Française). In: Orliac, C. (ed), Archéologie en océanie insulaire: peuplement,
sociétés et paysages, pp. 28–45. Editions Artcom, Paris.
Wilmshurst, J.M., Hunt, T.L., Lipo, C.P. and Anderson, A.J. 2011. High-precision radiocarbon dating
shows recent and rapid initial human colonization of East Polynesia. Proceedings of the National
Academy of Sciences 108(5):1815–20.
terra australis 37
terra australis 37
2
‘Dwelling carelessly, quiet and secure’
A brief ethnohistory of Rapa Island,
French Polynesia, AD 1791–1840
Atholl Anderson
Department of Archaeology and Natural History, Research School of Pacific and Asian Studies,
The Australian National University, Canberra, Australia, atholl.anderson@anu.edu.au
Introduction
In 1826, the first European missionary to Rapa, the Rev. John Davies, quoted Judges 18:7
in seeing the Rapans as ‘dwelling carelessly, quiet and secure, and having no business with
any man’ (in Stokes n.d.:28; an idiomatic rendering of the passage). It was to some extent,
possibly to a great extent, quite illusory. Rapa was certainly isolated by comparison with most of
East Polynesia, and it was small, mountainous and relatively cold, but even the first European
visitors found that Rapans exhibited evidence of contact with the outside world, and within
Rapan traditions, historical observations and ethnographic data which together form the stuff
of ethnohistory, the theme of contact and change is illustrated continually.
Rapan society was East Polynesian in ancestry and culture. Rapans spoke an East Polynesian
language, but its closest affinities were puzzling for a long time. The earliest historical contacts
with Rapans showed that they found both Hawaiian and Tahitian largely unintelligible and later
characterisation of Rapan by European scholars was confused because of the early introduction of
Tahitian by missionaries and, after 1863, of other Polynesian languages by Tongans, Tokelauans
and Cook Islanders, whose descendants came eventually to represent nearly half of the population
(Stokes 1955). Samuel Stutchbury had observed, presciently, in 1826 (in Richards 2004:5) that
the Rapan language was ‘something resembling the Marquesan’, but Horatio Hale (1968:141),
about 1840, ‘obtained at Tahiti, from a native of Rapa, a brief vocabulary of the language spoken
there, which turns out to be, with a few verbal exceptions, pure Rarotongan, and this in its
minute peculiarities’, while the missionaries William Ellis (1838) and M. Russell (1852:205)
thought that Rapan was closer to Maori than Tahitian. In the event, modern analysis shows that
Rapan is part of the group of languages that also includes Mangarevan, east Tuamotuan and
Easter Island, and probably once included languages on Pitcairn and Henderson islands, which
derive from ‘proto Southeastern Polynesian’. This group was probably the first to be differentiated
terra australis 37
26
Atholl Anderson
within proto East Polynesian (Fischer 2001), quite probably because of the relative remoteness
of the southeastern islands from the central archipelagos.
Rapan social organisation was divided historically (as described by Hanson 1970:19–22;
Hanson and Ghasarian 2007:59–60) among clans or ramages (kopu), each of which normally
occupied a single valley and adjacent fishing waters, the territory being commanded by a hill fort
or fortified village (pare maunga, pare tamaki or pa tamaki). The clans recognised membership
by unrestricted cognatic descent (i.e. through lineages of both parents) and access to land and
other resources by primary, active (arakaa) and secondary, latent (moekopu) rights. Primary
rights devolved both by descent and residence and held the clan together as a land-holding
and defending unit. Such groups of territorial definition and ambilateral descent are also called
‘ramages’, the term Hanson (1970) prefers, but ‘clan’ is used in the historical sources and
will be adopted here. The Rapan clan was headed by an ariki whose spiritual authority arose
primogeniturally from the clan ancestor, but whose practical authority in peace and war was
mediated through the consensus of family elders, the hui ragatira (Hanson 1970:22). The Rapan
traditions of pre-European events reflect the competitive development of the historical clans, of
which perhaps 18–22 existed by the arrival of Europeans (Hanson 1970:19).
Rapan traditions
During his stay on Rapa (April 1921–January 1922), as part of the Bayard Dominick Expedition
to the Austral Islands by the B.P. Bishop Museum, John Stokes (n.d.:Preface), accompanied by
his wife Margaret Stokes, collected historical traditions, mainly it seems from the official island
chief, Pataritari. In addition, Stokes collaborated with the linguist, Frank Stimson, who checked
Stokes’ material with Rapans living in Papeete. These included Teraau, son of the last main
‘king’ of Rapa, Terakau III, who died in 1887 (Queen Ruirau had succeeded but was deposed
after several months when Rapa was annexed to France in June 1887). Teraau, born 1862, traced
his descent primarily from the Gaitapana line, but he and his relatives, living in Tahiti, were
also descended from the clan Takatakatea, which Gaitapana had eclipsed. Others who assisted
Stokes were the Rapan men, Kareka (born 1867), Faaora (born 1865) and Gariki, plus Mato,
‘a native of Fakaho’ (Fakaofo in the Tokelaus) who arrived on Rapa in 1864 as part of a group
of Polynesians returning from Peru, and whom Stokes regarded as his most reliable informant.
Stokes’ account of Rapan traditions is the fullest available, but differs in its assignment of some
clan origins to multiple landfalls with that of Caillot (1932), who regarded all pre-European
Rapans as descendants of a single population (Hanson 1970:23 concludes that the data are
inadequate to decide the matter).
Accounts of initial colonisation obtained by Stokes (n.d.:1–28, 918–924, 951–954) said
that the first man to arrive on Rapa was Tiki (a common Polynesian mythological ancestor,
often regarded as the first man), who came from Avaiki (Hawaiki) in the Tuamotus and married
a Rapan woman. This is essentially a creation myth. They had two daughters, who also produced
a male and female child by Tiki, and these became the progenitors of all Rapans, but the boy,
Tamatiki, sailed off to Easter Island. There are other accounts of voyages to Easter Island or
New Zealand, by Temarago, and of Hotumatua, the Easter Island ancestor, as having been king
of Rapa.
The first settlers of Rapa included several peaceful peoples, such as the legendary Mana’une,
who were discovered by the East Polynesian ancestral figure, Hiro (or Iro). Hiro stayed briefly
in Rapa, and may have left a son there, while moving around the other Austral islands. By
genealogical reckoning, in which Stokes (n.d.) used the Rarotongan traditions accumulated by
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
27
J.B. Stair (1895) and S.P. Smith (1899), this Hiro is thought most likely to have been the cousin
of the equally well-travelled Tangiia, and is dated to about AD 1250, or possibly to about AD
1100. Buck (1954:119) has Tangiia searching for the son of Hiro, reputedly on Rapa, in the
mid-13th century. While Stokes (n.d.:26) accepted the Hiro story as well-documented by 19th
century sources, he doubted the historical validity of others, including those collected in his day
from informants who were ‘liable to include in the record what they think at the time and are
frequently unaware of the necessity of segregating ancient and modern information’.
Nevertheless, Stokes’ (n.d.:700–709) arrangement of accounts of the inter-clan history of
Rapa, although noting disputed stories by informants of opposing lineages at various points, is
probably as close to a chronicle of traditional, and putatively historical, events as we may get on
the surviving evidence. He says that some of Hiro’s followers remained on Rapa, intermarrying
with the indigenous people and forming the Mato clan, which lived in the northern lands
centred on Tupuaki Bay. About the same time, ancestors of the Takatakatea clan arrived at
Anarua or Iri on the west coast and, by intermarriage, expanded into the southern and eastern
areas, contesting with Ngate Mato the ownership of the flat land at the head of Ha’urei Harbour.
No pare (hill forts) were built at this time.
The early ascendant clan was Takatakatea. Sometime after it, ancestors of the Taukina clan
arrived on the western coast and settled among Takatakatea, but eventually disagreements arose
and warfare began. Takatakatea was driven out of the northwestern bays and constructed pare
at Karere, Paoreore, Motu and Rekie, then later at Mititipeiru, Napiri, west Pukutaketake,
Pukumaga and Nogoorupe. The expanding Taukina people, represented by their descendant
clans of Tipi and Okopou, continued to move southeast. Okopou occupied the Ha’urei lands
in the main harbour, and a strip across the main ridge to their earlier base at Iri Bay. The clan
wrested control of pare at Nogoorupe and Takitaga from Takatakatea, who then re-took them,
and, by managing to hold on to Morongo Uta and Pukutaketake, created an uneasy stalemate
in the west. The Tipi clan moved east and built pare at Ruatara and Vairu. At this point, Rapa
was divided into a Takatakatea territory in the south and southwest, with Tipi and its allies
in the north and east, and a small group of the Mato clan at Taratu, between them. The Tipi
a Manumanu (Tipi a Tepaiamarama) clan was the strongest and to it is attributed the first
historical high chief, Temarogo.
Several generations later, the Aureka people, who claimed descent from Tiki but probably
included new arrivals as well, formed a clan based on the eastern stronghold at Pukumia and
sought to bring the other clans under their control. Aureka built the pare at Tevaitau and from
it defeated Okopou in their nearby stronghold at Napiri, and then the Tipi clans at Anarua.
The Aureka clan, known later as Gaitapana, began its own kingly line with Aurariki I, dated
about 1600–1650, and that dynasty persisted until 1887 through, successively, Aurariki II–VII
and then Terakau I (AD 1775–1825), who lived through the period of European discovery and
arrival.
European discovery
The European discovery of Rapa began on December 22, 1791. Captain George Vancouver,
HMS Discovery, sailing a wide northeasterly track from New Zealand, saw three small islands
that, as he sailed closer, coalesced into Rapa Island (Figure 2.1). Vancouver was familiar with
Polynesians, having sailed on Cook’s second and third voyages (1772–75, 1776–79), and he had
no hesitation in seeking contact with the inhabitants. At 3 pm, about three nautical miles off the
western shore, three canoes arrived, their crews unwilling to board the ship but signalling the
terra australis 37
28
Atholl Anderson
ship’s crew towards land. Vancouver then sailed nearer to shore and several canoes paddled out,
from one of which a man boarded the Discovery, touched noses with Vancouver, and was given
an iron adze. Soon many men boarded and exchange began (Figure 2.2):
They all seemed perfectly well acquainted with the uses to which they could apply iron, and how to
estimate its value amongst themselves; as also in the manner in which it was regarded by Europeans. They
made no scruple, even with some force, to take articles of iron out of our hands; and in lieu of them
with great courtesy and address presented, in return, some few fish, fishing-hooks, lines, and other trifles,
which they seemed to wish should be accepted as presents, and not received in exchange. Looking-glasses,
beads and other trinkets of little importance, at first attracted their attention, and were gladly accepted;
but no sooner did they discover that articles made of iron were common amongst us, than they refused
all other presents, and wanted to barter every other gift for iron. I could not prevail on any of them to
accept a few medals.
Their visit seemed prompted only by curiosity, as they were completely unarmed, and brought with
them (except the few fish &c) neither articles of food nor manufacture. A few spears and a club or two,
were seen in one or two of the canoes only; two or three indifferent slings for stones were also noticed;
with which they parted without the least reluctance. (Vancouver in Lamb 1984:372–373)
The Rapan men attempted to carry off anchors, cannons and other large iron articles,
the weight of which surprised and frustrated them. They also stole what they could but had
little means of concealment. Some opened the shirts of sailors, evidently to see whether they
were women (Lamb 1984:372). Vancouver thought that the men most resembled Tongans.
An Hawaiian lad, Towereroo, on board Discovery was hardly able to understand the Rapan
language, but he had been in England since 1789. Vancouver attempted, with some difficulty
and eventual uncertainty, to obtain the name of the island; ‘at length I had reason to believe
the name of the island was Oparo; and that of their chief Korie’ (Lamb 1984:373). At 5 pm,
after only two hours of contact, Discovery sailed off to the north, the officers noticing part of the
north coast of the island as she departed. Vancouver summarised his impressions of Rapa thus:
Its principal character is a cluster of high craggy mountains, with perpendicular cliffs nearly from
their summits to the sea [high cliffs are especially prominent on the northwest coast] … the vacancies
between the mountains would more probably be termed chasms than vallies, in which there was no
great appearance of plenty, fertility or cultivation; they were chiefly clothed with shrubs and dwarf trees.
Neither the plantain nor other spontaneous vegetable productions common to the inhabited tropical
islands, presented themselves. The tops of six of the highest hills bore the appearance of fortified places,
resembling redoubts [George Hewett, Surgeon’s First Mate on the Discovery thought these were probably,
‘like the hippahs of New Zealand’ (Lamb 1984:374, fn 3)]; having a sort of block house, in the shape of
an English glasshouse, in the centre of each with rows of pallisadoes a considerable way down the sides
Figure 2.1. A sketch of Rapa drawn on the Discovery expedition 1791 (Lamb 1984 Volume 1:Plate 18 ‘The island of Oparo’). Published
with the permission of the Hakluyt Society and the State Library of New South Wales (item MRB/Q910.4/6 A/1).
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
29
Figure 2.2. A fishing device collected at
Rapa on the Vancouver expedition, 1791
(Beasley 1928:Plate LXXII). It consists of
a fibre strip line, a wooden spreader, and
two small shell bait hooks, which are
attached by cordage.
of the hills, nearly at equal distances. These, overhanging, seemed intended for advanced works, and
apparently capable of defending the citadel by a few against a numerous host of assailants. On all of them
we noticed people, as if on duty, constantly moving about … [the block houses] were sufficiently large to
lodge a considerable number of persons, and were the only habitations we saw. Yet from the number of
canoes that in so short a time assembled around us … [we could infer] … that the shores and not those
fortified hills which appeared to be in the center of the island [Vancouver was unaware of the central
harbour] would be preferred for their general residence. We saw about thirty double and single canoes,
though most of them were of the double sort: the single canoes were supported by an outrigger on one
side, and all built much after the fashion of the Society Islands, without having their very high sterns,
though the sterns of some of these were considerably elevated; and their bows were not without some
little ornament. They were very neatly constructed, although the narrowest canoes I ever saw. (Vancouver
in Lamb 1984:374)
Vancouver (Lamb 1984:374) remarked that the canoe builders ‘are nearly destitute of
iron, and possessed of very few implements of that valuable metal’. This comment, and his
earlier description of the behaviour of the men who came aboard Discovery, suggest that he saw
evidence to imply that iron had reached Rapa before his arrival. Of the canoes he wrote:
The island did not appear to afford any large timber; the broadest planks of which the canoes were made,
not exceeding twelve inches … Some of the stoutest double canoes accommodated from twenty-five to
thirty men, of whom, on a moderate computation, three hundred were supposed to have been seen near
the ship. These were all adults and apparently none exceeding middle-age; so that the total number of
inhabitants on the island can hardly be estimated at less than fifteen hundred. In this respect it must be
considered prolific, notwithstanding its uncultivated appearance. The natives, however, appeared to be
exceedingly well fed, of middle stature, extremely well-made. (Vancouver in Lamb 1984:375)
The plank-built construction of early Rapan canoes, evident in the various canoe planks found
in rockshelters by Stokes (n.d.:Preface) gave way later to dugout construction, as introduced
trees produced larger timbers (Haddon and Hornell 1975:147–151).
The Rapan men made every effort to get the Discovery crew to go ashore:
terra australis 37
30
Atholl Anderson
On their departure they took hold of the hand of everyone near them, with a view to get them into
their canoe. They all had their hair cut short; and, excepting a wreath made of a broad, long-leaved,
green plant, worn by some about the waist, they were intirely without clothing. Although the custom
of tatowing prevails so generally with all the islanders of this ocean, these people were destitute of any
such marks.
Independent of the protection their fortified retreats may afford, it did not appear that they were
subject to much hostility, as scarcely any scars from wounds or other marks of violence were observed
on their bodies. Their elevated fortified places … led some of us to conjecture, that they were frequently
annoyed by troublesome neighbours from some other islands not far distant. But, as the canoes we saw
were not even furnished with sails nor had any appearance of having been equipped for an expedition
beyond their own coast, it may reasonably be inferred, that they were not accustomed to voyages of any
length. Yet, on the other hand, when the small extent of their island is taken into consideration, it is hard
to reconcile that it is not the fear of foreign enemies, but the apprehension of domestic insurrection, that
has induced the laborious construction of their fortified retreats. (Vancouver in Lamb 1984:375)
Archibald Menzies, botanist and surgeon on Discovery, was another experienced eyewitness
of the first meeting, having earlier sailed around the northwest Pacific, including to Hawaii. He
adds some useful details to Vancouver’s narrative. His impression of the forts was that ‘each bore
some resemblance to a large block-house fenced round at a little distance with a high wall of
stone or turf ’ (Shineberg 1986:66). The Rapans on board Discovery were prevailed upon:
… to count their numerals to ten, which we found to agree exactly with those of Otaheite, & a few other
words which they repeated convinced us that they spoke a dialect of the same general language, but so
modified from their local situation that even Toowowero [above] could understand very little of what
they said. This being the case I think it is probable that Oparoo may not be the real name of the Island,
though it was often their answer to our interrogation on that head and therefore adopted … They suffer
their beards to grow long, but their hair which is naturally straight was croppd short round about the
nape of the neck & their ears were perforated though we saw them wear no ornaments in them excepting
the nails they got from us. (Menzies in Shineberg 1986:67)
No signs of tattooing were observed, and ‘the only cloathing they wore were a narrow strip
of cloth made from the bark of a tree which passed around their waist and between their legs’
(Shineberg 1986:67). Menzies thought the cloth was probably scarce because many men simply
wore bunches of Dracena leaves attached to a girdle (Prebble 2005:156 suggests the bark cloth
was more likely from Hibiscus tiliaceus than paper mulberry (below), and that Dracena was
probably Cordyline). Turning to the Rapan canoes, Menzies described them as:
… small & narrow but neatly formed, rising a little at each end to a sharp point with outriggers fitted
to them – similar to the generality of Canoes in this Ocean. They had also double canoes with Sails
constructed in the same manner, & though we observed no wood or Timber on the Island of a size
capable of making their canoes, yet they did not seem to be a scarce article among the Natives, for at one
time we counted no less than 30 canoes about the ship & between us & the Shore, eight or nine of them
were double canoes each of which had upwards of 20 men, & few of the single canoes had less than five
men, many of them had more, so that we estimated that the number of people that came off in these
Canoes from this Bay to be about 300, and as there were no women, children or any very old people seen
amongst them I think it may be safely inferrd that they were not one fifth of the Inhabitants of this little
Valley, which from whence would amount to upwards of 1500. But I would not conclude from this that
the Island is very numerously inhabited, perhaps the environs of this Bay may contain one-half of the
whole number. (Menzies in Shineberg 1986:68)
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
31
Both Vancouver and Menzies thought there might be other inhabited islands to the south
of Rapa, and these could explain both the need for fortification and the existence of so many
canoes in the apparent absence of substantial trees.
Menzies noted that ‘excepting a few small fish, none of these canoes brought off any kind
of refreshments – either Hogs Poultry or Vegetables, so that we remain entirely ignorant of the
produce of this Island’ (Shineberg 1986:68). Looking into the bay off which the Discovery lay,
he could see scattered bushes among which were ‘the habitations of the Natives & some little
signs of Cultivation’. Otherwise, the island appeared fairly open. On the south slopes of the bay
were some scrubby trees, especially in the hollows, and on the north side only grass. No coconut
palms were seen.
The Discovery observations
Although the result of barely two hours contact, the Discovery observations provide an
immensely useful picture of late 18th century Rapa. It seems that the Rapans had some iron but
their evident wonderment on boarding Discovery – ‘their attentions & curiosities absorbd with
everything they saw’ (Menzies in Shineberg 1986:67) so that it was very difficult to obtain their
compliance in answering questions – was such that it is very unlikely that any earlier European
voyagers had called in there. The iron must have come by travel to or from islands nearer to
earlier European contact. Don Thomas Gayanagos, Aguila, discovered Raivavae, the nearest
island to Rapa, in January 1775, and, in a brief encounter, traded some knives and nails with
the inhabitants (Corney 1915:179). Cook, accompanied by Omai, found Tubuai in 1777 but
was unable to induce anybody to step on board the Resolution. However, as the Societies had
experienced extensive contact with Europeans since discovery by Wallis in June 1767, it is quite
probable that iron had been traded into the Australs for 10–20 years before 1791. The Rapans,
at the end of the line, had much less than they wanted, and no idea initially, on boarding
Discovery, that they had chanced upon a fabulously rich source.
It is not certain just where the Discovery lay off the coast. As it sailed up from the southwest,
the six forts seen (five being visible at one time) were almost certainly six of the seven which
lie along the skyline of the west coast: south to north these are: Ungarere (Ororangi?), Ngapiri,
Tevaitau, Morongo Uta, Puketaketake, Noogurope and probably Kapitanga. Puketaketake and
Noogurope overlook Anarua Bay, which, as the only bay on the northwest coast, must have been
‘the small Bay on the Northwest side of the island’ (Menzies in Shineberg 1986:66) off which
Discovery lay. As the valley runs directly inland behind the sandy bay-head, the Discovery crew
would have had a good view of the inhabited landscape at a distance of about 1 km, and while
sailing up the coast, the forts would have been seen at 2.5 km to 4 km away, no great distances
for telescopic inspection.
European observations of the Rapans suggested that they were thought somewhat shorter
and darker and more robustly built than the generality of East Polynesians, and that their
language, although clearly East Polynesian, was a distinct dialect. It is difficult to understand
what the Rapans meant by referring repeatedly to ‘Oparo’ in response to enquiries about the
island, but it is possible that they interpreted those as questions about where they had come
from and replied that they were ‘Opare’, meaning inhabitants of the forts. More speculative still
is the possibility that ‘paro’ (stray or wander) and ‘korie’ or kore (absent, destroyed, annihilated),
if these meant the same in Rapan as Maori (Williams 1971), were references to the fate of the
people to whose place the Europeans were pointing. The name ‘Rapa’, it might be conjectured,
is an allusion to Ha’urei Harbour as a vagina, the means by which mediation between ordinary
and supernatural were negotiated with each passage (cf. Hanson and Hanson 1983:86–94).
terra australis 37
32
Atholl Anderson
Only men were seen, and there are no observations of differences in rank, either by deference
or appearance. The men had short, straight hair and probably no topknots, unlike at Raivavae,
for example (Corney 1915:179). There was little sign of weapons, there was no tattooing and
pierced ears held only nails. Clothing was restricted to breech-cloths and similar coverings, some
of them made from bark cloth, possibly from aute, the paper mulberry (Brousonettia papyrifera),
and bunches of ti (Cordyline sp.) leaves. The Discovery lay inshore long enough, and was visited
by sufficient canoes, that had there been local men of obviously substantial status then it is hard
to imagine they would not have made themselves known, as was the case elsewhere in East
Polynesia when Europeans arrived.
Other than the people who came aboard, the forts attracted most attention. The Discovery
observations clearly suggest that the six western skyline forts were in existence in 1791 and
perhaps that they were all inhabited. There is, at least, no comment to the effect that some
appeared abandoned or in disrepair. The description of these sites indicates that a large building
– possibly a communal house – had been erected in a central position, where it was surrounded
by a stone or turf wall and, at regular intervals away from the centre, by rows of outward-leaning
palisades. People were seen moving about within the defences.
Menzies saw structures that were possibly habitations, though they might have been garden
sheds, in the valley, but relatively little sign of cultivation, although the land was very largely
cleared. However, the European view of the valley floor of Anarua Bay, if that was the place,
would have been somewhat obscured by a dune system above the beach. Nevertheless, Menzies
saw no sign of pigs or poultry (or dogs, it can also be assumed, though he did not mention
them). Coconut palms and bananas were absent to view and neither vegetables nor fruits were
brought out to the ship.
Discussion of population size by the Discovery officers is particularly interesting. Vancouver,
evidently reporting observations other than his own, said that 300 men were seen about the
ship and the island population could hardly have been less than 1500, an estimate which has
been widely used in the subsequent literature on Rapa. However, Menzies is more specific, and
seemingly from direct observation. He describes the canoes, on which the population estimates
depended, in greater detail, noting that some double canoes had sails of a common Polynesian
type (as Vancouver seems not to have seen these, and as the Discovery lay close inshore, they
were perhaps furled around spars in the bilges of the canoes and may, thus, have been of the
form used by Maori), and the number of men in canoes of different type. Most importantly,
he says these canoes came out from the bay inshore of Discovery and so his estimate of a total
population of 1500 is for the bay alone. He was reluctant to extrapolate his data to the whole
island, suggesting it had perhaps 3000 people, but then he had seen only the west coast at close
range and knew nothing of Ha’urei Harbour or the large bays elsewhere on Rapa.
Early 19th century contact at sea
As European interest in the Pacific grew apace during the late 18th and early 19th centuries it
seems quite probable that Rapa was visited on various occasions during that period, especially
after the publication in 1798 of the first edition of Voyage of Discovery (Lamb 1984:267).
However, there are few confirmed data. The Sydney trader, Captain Roger Simpson probably
visited Rapa in 1802, Stephen Reynolds in the New Hazard sailed by it in 1813, and the Sydney
vessel Endeavour clearly made contact with Rapans in July 1815, when it found, ‘the Roppa
Islanders to be pilferers of anything they could lay their hands on on deck’ (Richards 2004:4.
This seems to be the earliest reference to ‘Rapa’). Captain John Powell of the Queen Charlotte
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
33
was becalmed near Rapa in 1815 or 1816 and a number of canoes came off but none of the
men would go onboard. Instead, about 50 of them jumped into the sea and, grabbing a loose
rope at the stern of the ship, began swimming towards land, ‘labouring and shouting with all
their might, as they supposed they were drawing the vessel towards the shore’. An ensuing tugof-war with the ship’s crew and the fortuitous arrival of a breeze ended the encounter (Ellis
1838:369–370).
The Queen Charlotte returned to Rapa with William Ellis of the London Missionary Society
(LMS) on January 27, 1817. While sailing slowly along the western coast, 30 canoes came
around the ship:
The men were not tataued, and wore only a girdle of yellow ti leaves round their waists. Their bodies,
neither spare nor corpulent, were finely shaped; their complexion a dark copper colour; their features
regularly formed; and their countenances often handsome, were shaded by long black straight or curling
hair. (Ellis 1838:365)
A crayfish, lying in the bilge of a canoe among some spears, was exchanged for fish hooks.
A man then came aboard, carefully sniffing Ellis’ hand as he did so. Many others then boarded
and began to take anything they could make off with, including a kitten. The crew eventually
produced long knives to drive off the remaining Rapans who, not understanding the nature of
the weapons, tried to seize them by the blade and were badly cut in their hands.
In the winter of 1820, Bellingshausen in command of the Vostok and Lazarev with the
Mirnyi reached Rapa on a passage from New Zealand. They came up from the southwest and
stood off the northern coast at a distance of at least 4.5 sea miles, in light and fickle winds. Fewer
canoes ventured out than had been seen by Vancouver: 15 containing 80 people on June 29
(although another report says 23 canoes with at least five men in each on the first day; Barratt
1988:211), and 22 with about 110 people on June 30. The canoes, fastened with twisted bark
cordage, and some of them 7.6 m long, showed no decoration (see Barratt 1988:Plate 26).
There were double canoes and some large canoes with double outriggers (Simonov in Barratt
1988:205). No sails were mentioned.
At noon on June 29, several canoes, each with five to seven men (Figure 2.3), came out
and greeted the Vostok with a speech of ‘much heat and volume’. Coming on board, the Rapans
presented ‘us with sea crabs and some sort of fermented dough’ (Barratt 1988:201). The dough
was almost certainly tioo (fermented taro) and the sea crabs appear from later description to
have been crayfish. One islander wore a bark sash, which he exchanged for a fish hook. Taro
roots were presented, and cooked and eaten by the Russians, who enjoyed them as much as
they abhorred the bitter taro dough. Shellfish, ‘a dried pumpkin which had nowhere been cut
through’ (Barratt 1988:211) and was very probably a gourd, and bast cords, probably of hau
bark, worn around the neck, were also in evidence.
The appearance of the Rapans (Figure 2.4) was much as it had seemed to Vancouver except
that by now there were signs of earlier sexual relations with Europeans, including a half-caste
youth (see Barratt 1988:Plate 27). On this occasion also a chief was recognised – indeed he was
the first to go aboard – and given appropriate privileges and presents. Rapan thievery remained
a problem but it was evident that the Rapans now understood the use of guns. The Russians
asked for fish, poultry and pigs, showing examples of the domestic animals to their visitors,
but none were produced. Looking at the island through telescopes, the Russians could see the
forts, ‘within which huts were visible’ (Barratt 1988:217) and narrow paths running up to
them. Yellowish-red and red patches on the hill slopes were surely signs of soil erosion (Barratt
1988:212, 216).
terra australis 37
34
Atholl Anderson
These later contacts at sea served to widen European knowledge of Rapans and, of course,
knowledge of Europeans by Rapans. They showed the first indications of varied status in Rapan
society and of a growing awareness of European things, including guns. Evidence arrived of
the prominent means of subsistence, notably of taro in several forms and perhaps the gourd,
but conversely, of the probable absence of domestic animals. A new form of canoe, the double
outrigger, was seen, or at least asserted (it seems improbable given the western Pacific distribution
of this type). Otherwise, Rapans looked and behaved as they had done in the late 18th century.
Early 19th century observations ashore
Between 1825 and 1830, Rapa was pulled into the European colonial world of industry, religion,
commerce and disease. The first extended contact began in July 1825 when a cutter, the Snapper,
owned by Tati, a Tahitian chief, but under the command of Captain Shout (Davies in Newbury
1961:279 has him as I. Shant), the first European to set foot on Rapa, arrived on passage from
the Tuamotus to Tahiti. Two Rapans were on board when the captain, feeling under threat
by many more canoes arriving, sailed off to Tahiti. There the two men, Paparua and Aitareru,
attended school and church under the supervision of the LMS missionary, John Davies and,
in October 1825, they were returned to Rapa, ‘loaded with presents and accompanied by two
Tahitians [Hota and Nene], who were sent to gain more accurate information relative to their
country, and the dispositions of its inhabitants … the captain of the cutter procured some tons
of sandal-wood and when he left, the Tahitians [also] returned’ (Ellis 1838:372).
The successful visit to Tahiti by the two Rapan men, ‘tho they did not understand much
Tahitian’ (Davies in Newbury 1961:280), and their congenial homecoming, during which Teraau
(also recorded as Teranga and Terakau), the ‘head chief of the island’, desired the Tahitians to
return (Davies in Newbury 1961:280), encouraged John Davies to begin a mission on Rapa.
The mission station was established ‘in Aurai [Ha’urei] harbour’ in January 1826. On the brig
Governor Macquarie Davies took Hota and Nene and their wives, plus Mahana (a schoolmaster)
and Pauo (a tradesman and boat-builder):
They carried with them not only spelling-books, and copies of the Tahitian translations of the scriptures,
but also a variety of useful tools, implements of husbandry, valuable seeds and plants, together with timber
for a chapel, and doors &c. for the teachers’ houses … Mr Davies … was pleased with his visit, and, upon
the whole, with the disposition of the people, although some appeared remarkably superstitious, and, as
might be expected, unwilling to embrace Christianity. This arose from an apprehension of the anger of
their gods, induced by the effects of a most destructive disease, with which they had been recently visited.
Figure 2.3. Drawing of a Rapan canoe with six crew produced from a sketch by Pavel Nikolaevich Mikhailov, June 29, 1820 (which shows
faint sketches of seven crew; Barratt 1988:Plate 26). The canoe shows similarities to Maori craft, including, on the sternpiece, a shape that
resembles the seated ancestral figure sometimes carved on the Maori sternpiece (taurapa) of a war canoe (Debenham 1945:Plate XVIII).
Published with the permission of the Hakluyt Society and the State Library of New South Wales (item MRB/Q98/32 A1).
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
35
Figure 2.4. Drawing of a young Rapan
man by Pavel Nikolaevich Mikhailov,
June 30, 1820 (Barratt 1988:Plate 27),
published with the permission of the
Hakluyt Society and the State Library of
New South Wales (item MRB/Q98/32 A1).
The man is probably wearing leaves in his
ear, as noted of some Rapan visitors, and it
is his only form of decoration, as common
among Rapans at that time.
The gods, they imagined, had thus punished them for their attention to the accounts from Tahiti. (Ellis
1838:373)
There had been ‘a great mortality amongst the people since they were visited by the Snapper’
and among the dead was Teraau, upon whose land the original chapel at Papara was re-erected
(Newbury 1961:280), presumably at the request of Teraau’s son, Koinikiko, who supported
the mission and gave it land and taro patches (Stokes n.d.:24). Davies was impressed with the
plantations of taro:
which are well laid out and display a degree of skill and contrivance, evinced in nothing else we saw. The
houses are miserable huts about five feet high, with holes to creep in, about two feet square. The natives
have neither maraes nor altars, but only a few rude stones placed in the ground, which denote a va’i tapu
or sacred place. These stones (they say) are mea mana nui, i.e. things of great power. There seems to be
no images worshipped. Everything exceedingly rude; their stone adzes &c. are similar to those found in
other islands, but more rude and uncouth. They do not tatau their bodies, and till they were visited by
the Snapper and Minerva were all without clothing … [Davies then says the two Rapans who had been
to Tahiti told their countrymen that nakedness was disliked and since then the Rapan chiefs, at least,
had worn clothes in the Tahitian fashion]… The vegetable productions of Rapa, with few exceptions,
terra australis 37
36
Atholl Anderson
resemble those of Tahiti and the Society Islands. There are found the various kinds of taro that are
cultivated at Tahiti, one kind of meia or banana, the umara or sweet potato, and a species of yam; the
ahia also, or red apple, but there is no bread-fruit nor cocoa-nut, neither the vi nor mountain-plantain.
One cocoa-nut tree, indeed, was found by Hota and Nene, when they were here before, growing on the
west side of the island; but this had been thrown up by the seas, and the natives did not then know what
it was. Abundance of wild celery grows here in the moist grounds similar to that of New Zealand. (Davies
in Stokes n.d.:25–26)
Davies comments usefully on the name of the island: ‘the natives call it Rapa only. The
name occurs in some of the old traditional songs of Tahiti’, and he wrote that the population, ‘is
supposed to be upwards of 2000, or nearly as many as those in Raivavae’ (Stokes n.d.:27). His
investigations of Rapan belief established that:
Puoero Poere is one of the chief gods of the Eriki, or chiefs, but their gods in general seem to be the spirits
of departed ancestors. I wrote down the names of about forty of them, with a collection of Rapa words,
but I has mislaid or lost the manuscript of them. That the natives of this island are of the same general
origin as their neighbours, their language, customs, and superstitions afford sufficient proof. They have,
however, many words not used at present in Tahiti, and retain the consonants k, g, ng, not used here [i.e.
in Tahiti]. On the other hand they reject both f and h. Their numerals are nearly the same. (Davies in
Stokes n.d.:28)
Davies reiterated his good impression of the state of taro cultivation, and observed that,
‘The planting, weeding, baking of it altogether the work of the women. In making canoes the
Rapa-men do not excel’ (Stokes n.d.:29).
Although the early services, being in Tahitian, were ‘unintelligible’ to the Rapans, two
mission stations had been established by June 1826. By 1829, the LMS missionaries, Pritchard
and Simpson:
found that four chapels had been erected in different stations [Moturi, Tupuai, Iri, Aurai] at which, by
native Missionaries, religious instruction was statedly imparted. The inhabitants manifested a pleasing
attention during service, and their advancement in knowledge exceeded the expectations of their visitors.
(Ellis 1838:374)
They also found that ‘the various seeds and plants that had been taken to the island from
Tahiti did not thrive, the climate being much colder than that of Tahiti’ (Newbury 1961:281).
Their observations on Rapan belief amplified those of Davies. They found that Paparaa or
Paparua and Poere were the principal gods:
Paparaa was made of the coconut husk, neatly plaited in the form of a cask [Ellis 1838:364, reporting
similar information, says it was ‘a kind of cylinder, full in the centre, and smaller at the ends’]. Its length
is from 2 to 3 inches. To this god they prayed for victory in war, for the recovery of the sick, and for an
abundance of turtle in the harbour. Poere was a piece of stone from 12 to 15 inches long. This was planted
in the ground, and worshipped at the launching of a canoe, at the opening of a new house, praying that
in it there might ever be an abundance of food; and that they might have much water in the springs …
no sacrifice was offered except fish. (Stokes n.d.)
As for the mountaintop forts, Pritchard and Simpson heard that:
Their wars were not so frequent, nor so cruel, as in some of the neighbouring islands. There are still
remaining, several old castles, built on the tops of the highest mountains, to which the vanquished
repaired, and continued the siege for a long time. These fortifications appear to have been standing for
ages. It is now a long time since they have been used. Some of the oldest people say that they have no
recollection of war in their time. (Stokes n.d.:31)
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
37
From Tahiti, had been brought fowls, pigs, kumara, pumpkins, melons, papaw, cabbages,
onions, pineapples and potatoes. Breadfruit, mountain plantain (fei) and coconut, also
introduced, did not thrive, according to Pritchard and Simpson. Three Europeans who had
some ability in Polynesian languages, which they had used, reputedly, to recruit slaves, lived on
Rapa for a short period about this time, probably 1827–1829. They were occupied particularly
in distilling spirits from ti (Stokes n.d.:29).
Pritchard and Simpson baptised 251 people in 1829. Another missionary, Mr Darling,
visited in June 1831 on the Rarotonga, and another 140 people were baptised (75 men, 65
women; a later note by Davies (Newbury 1961:282) says it appears that Darling baptised 147
people (76 men, 71 women) and also 44 boys and 51 girls of baptised parents). At the latter
time, the names of the entire Rapan population were written down; there were 357 adults and
243 children, 600 in all (Davies in Newbury 1961:281). Among them was a Mangarevan man,
Mapuagua, who was the only survivor of seven Mangarevans whose raft had drifted to Rapa
many years earlier. Four had later set out in a strong westerly wind to attempt to sail back.
Mapuagua died in 1829 (Ellis 1838:374; Richards 2004:5).
The mission had arrived in Rapa in 1826 with a sandalwood expedition by Captain Samuel
Henry on the Snapper and Captain Thomas Abrill (or Ebrill) on the Minerva, which was owned
by the Tahitian ‘king’ Pomare III. Sandalwood (Santalum insularum) was in limited supply on
Rapa, but up to 20 tons were taken, and this encouraged the British Consul to send a Mr Young
with some Tuamotuan labourers to Rapa, on the brigantine Active (Captain Elley), to cut more
sandalwood and collect beche-de-mer. They were unsuccessful (Richards 2004:5), but Mr Young
remained on Rapa for at least six months, certainly until June 1826. The Pacific Pearl Company’s
ships, Sir George Osborne and Rolla, visited Rapa briefly in April and June 1826, unsuccessfully,
but the expedition leader, Samuel Stutchbury, made some interesting observations. He noted
that ‘the natives subsist on taro which is very abundant. They also have the Ti root … Rats
appeared innumerable and exceeding bold’ (in Richards 2004:5). From Mr Young, Stutchbury
heard that the Rapans were:
… very peaceable, living entirely upon fish and taro, and that the females [were] remarkably chaste, if
a man dies or is driven away from the island in his canoe which sometimes happen as they fish a great
distance out to sea, the widow will ascend one of the highest and most precipitous mountains and hurl
themselves down … When a person dies, they place the body in a rude kind of wicker coffin or basket
and carry it to the top of one of the mountains, the relatives and friends assemble, and lament and cut
themselves for many days continuance, after some weeks of exposure on the mountain, they will place the
body or bones in an old canoe, tow it a considerable way out to sea, then fix on a number of large stones,
and let it sink. If it is a chief, they will sometimes bury the body on shore. The generality of the natives
go about perfectly naked. (Stutchbury in Richards 2004:6)
In June 1826, the US Navy schooner, Dolphin, seeking yams and taro, visited several small
bays on the north and east coasts of Rapa, finding huts and taro patches in each, but was directed
by inhabitants to the main harbour, where Mr Young piloted it in, somewhat inexpertly. The
First Officer, Lt. Hiram Paulding, found that taro ‘was planted everywhere, in large patches,
where there was a small valley through which a stream of water [flowed] … the hills were green
with another species we had not before seen, called mountain tarrow. The latter is superior …
and will keep longer at sea’ (Paulding 1831:252; Prebble (2005:162) suggests ‘mountain tarrow’
was either the common taro, Colocasia esculenta, or Alocasia macrorrhiza). The Rapans permitted
the Dolphin crew members to dig as much taro as they wanted, perhaps because they were
unable to do so themselves. An epidemic disease seems to have been in progress and the people
‘had a sickly look, almost without an exception’ (Paulding 1831:253). They were dressed in:
terra australis 37
38
Atholl Anderson
… a heavy mat of grass, weighing from ten to fifteen pounds, which was thrown over their shoulders,
and another light mat, of the same material for their loins. Their deportment was modest and retiring,
and they evinced a disposition to have but little intercourse with us. A few of their houses were scattered
about on the hills. They were extremely miserable … long and narrow, about three or four feet high, so
that when one entered them, it was necessary to go down on hands and knees. (Paulding 1831:254)
Two whaleships visited Rapa in July 1826, and the British naturalist, Hugh Cuming, visited
on May 17, 1828. Cuming was ashore for only four days but he took good note of what he
observed of the Rapan people. Their dress consisted of:
Two pieces the size and shape of a very large door mat, thickly thrumbed. One of those they fasten over
the neck which reaches the loins. The other is made fast to the loins so that the upper one covers the
lower one a few inches. These species of cloaks are very heavy, are made of a large bush with the rind taken
off, woven by hand with the leaves of the Te [ti or Cordyline sp.] Plant split in fine threads … they can
withstand the rain for a month. (Cuming in Richards 2004:7)
Cuming implies, very plausibly, that this was the traditional winter dress of the Rapans.
Their summer dress was made in the same form, ‘only it is made entirely from the leaves of the
Te Plant finely drawn into threads and much longer than the winter dress, yet it is not above
half the weight’ (Richards 2004:7). There were also mortuary cloaks, some fragments of which
have been recovered from cave sites (Buck 1925).
The Rapans did not like the mission plan of houses set out in rows of 12 to 14, preferring
‘to live in their ancient sites and dwellings close to their taro grounds’ in dispersed dwellings
(Cuming in Richards 2004:7). Their houses were about 14 feet (4.3 m) long, 10 (3 m) broad
and 3 (1m) high, with an entrance at each end, two small windows and a fireplace at one end.
They were shaped like the top of a covered wagon, which presumably means rounded rather
than ridged, and thatched to a thickness of about 0.6 m (Cuming in Richards 2004:7). Cuming
could find nobody who knew much about the forts, except that they were places of refuge when
Rapa was divided between two warring chiefs:
The losing party then had to fly to these mountain fortresses for safety as every male that was taken in
the battle or afterwards was slain. The females and children was not molested and was permitted to carry
up to their relatives, food until they made peace with the conquerors which was soon effected by the
females. Most of those strongholds are square, the walls very thick about ten or eleven feet high with
some rugged stones by which they could get up on the platform. On the top was an immense heap of
large stones which they threw upon their pursuers. The mountains I ascended appeared to have steps cut
in the mount, winding several times around it … [the Rapan weapons, according to Cuming, were] …
a lance eighteen feet [5.5 m] long, very rudely made, and a short ugly unadorned club. Their fish hooks
were made of the roots of trees bent and hardened by fire. Their household furniture consist of a low stool
cut out of the solid wood [or possibly a wooden pillow], a stone knife and a few stones to keep the fire
together … Their food consisted of fish and tara. (Richards 2004:7)
The Belgian trader, Jacques-Antoine Moerenhout (1837) visited Rapa in 1834 to bring
back Rapan men whom he had employed as pearl divers. He remarked that, ‘fish and taro were
the only food formerly. Tiao [tioo] was made of taro, in place of breadfruit as in the islands
to the north’ (in Stokes n.d.:33). Tioo, or fermented taro, was a particularly important Rapan
food made by packing cleaned taro corms, or mashed raw taro, into pits lined with grass and
leaves, which were then covered with earth. It would keep for six months during which it
turned into a sour, pungent, but much esteemed paste (Stokes n.d.:24–26. Tioo storage pits
(rua tioo) of 2–4.5 m in diameter and more than 2 m deep were recorded around and on many
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
39
pare by Stokes n.d.:277). Moerenhout was offered provisions of cabbage, onions, taro, chickens
and pigs, an indication of the extent of change in agriculture. He heard that the trunk of the
candlenut or tiairi (Aleurites triloba), the largest type of tree on the island, was used for canoe
building, and its nuts for lighting, and he was struck by the extent that men were evidently tabu
and fed by their wives. Moerenhout thought that cultivation, cooking and other domestic tasks
and crafts were the responsibility of women, while men made nets, presumably fished, and built
canoes and houses.
European contact continued after the 1830s, of course, and with growing frequency, but
observations were becoming repetitive and they were, increasingly, of Rapan behaviour modified
by close European contact (e.g. Lucett 1851, who visited in 1843 and 1844; Hall 1868 in that
year), which is largely irrelevant to this volume. Most importantly, Rapan society had been
deeply transformed by a demographic catastrophe during the early 19th century.
Contact ashore had, predictably, a much greater impact on Rapan society than that at sea.
Much of this remains effectively invisible. For example, it is impossible to be sure of the nature
of the original Rapan pantheon. Later scholars remarked on the apparent absence of such major
East Polynesian gods as Rongo, Tane and Tangaroa (Buck 1954:179), but how much this owed
to the rapid destruction of the pagan belief system before it was recorded is unknown. Arrival
of Europeans ashore, and doubtless disputes over women, brought violence, as in 1826 between
Rapans and the crew of the Minerva, in which it seems some Rapan men were killed. Much
more lethal, however, was disease, but it is difficult to estimate its full impact without knowing
the approximate size of the Rapan population about AD 1790. In the light of discussion above,
it is apparent that the popular estimate of 1500 people was not intended by Menzies to represent
the entire population, which he put, extremely conservatively in the light of his method of
calculation, at only 3000, twice that estimated for a single bay, probably Anarua. On the other
hand, it is very improbable that the eight most habitable outer bays plus Ha’urei Harbour
each supported an average population of 1500, making a total of more than 13,000. Eugène
Caillot (1932:24), who visited Rapa in 1912, reports a Rapan estimate of a former population
of 6000.
Stokes (n.d.:35, 436–440) attempted an estimate for the late prehistoric population using
his detailed archaeological records of pare and coastal settlements. For each of 25 pare, he
established the number and size of habitation terraces and allowed an average of five people
for each (four to 10 according to size). For Morongo Uta this produced a population of 390,
for Ororangi 280, and 215 for Tapitaga, but most supported fewer than 100 people. The pare
total was 2377 people, and with 10 coastal settlements added on the same basis, the overall total
was 3027 people. Bartruff et al. (this volume) calculate the maximum sustainable population
according to an agricultural production model at around 2000, which is similar to the first
recorded estimate of ‘upwards of 2000’ made by Davies in 1826 (above), but he reported that
disease had already taken its toll.
It might be reasonable, therefore, to assume that the maximum population size in the early
1820s was close to 3000. It was nearer 2000 by 1826 and Darling’s first census in 1831 recorded
only 600 people. Stokes (n.d.:3) says that 85 per cent of the population disappeared in the eight
years following missionary arrival in 1826. In 1836, Darling’s census totalled 453 (more than
the 300 estimated by the French trader Armand Mauruc in 1834, Caillot 1932:25; MacArthur
1968:307). By the 1860s, especially after the landing in February 1863 of Polynesians sick with
smallpox and cholera, who were being returned from Peru, the population had plummeted to
100–150 people, from which it began a gradual climb in the 1880s to reach more than 450
today. As early as the mid-19th century, the demographic catastrophe had begun to transform
terra australis 37
40
Atholl Anderson
the social basis of landholding into an unrestricted cognatic system that, in time, conferred
virtually universal rights upon everybody (Hanson and Ghasarian 2007).
Additional evidence from Stokes’ informants
All of the informants upon whom Stokes relied had been born in the late 19th century, and not
all of them were Rapan (above), so it is necessary to treat their observations about life before
the arrival of Europeans with caution. On some subjects, Stokes found them unreliable. For
example:
the opinions of the natives are very diverse on the subject of the original native house, even to the point
of absurdity … the older natives are in agreement that the huts were small and low, but none of them
had seen one completed. The one point never disputed is that the thatch was of ti leaves. (Stokes n.d.:
333, 340)
There was some agreement that houses usually had a fireplace, although Stokes found this
to be rarely in evidence on the house terraces of the pare, where only a few small stone-kerbed
rectangular hearths were seen. Floor mats were evidently never made, the floors and sleeping areas
being covered only in dried grass (Stokes n.d.:341). Tapa cloth was always plain according to
Stokes’ informants, and only plain fragments have been recovered from rockshelters (n.d.:342).
Gourds were used to store water and for utensils; no wooden vessels were made because of the
absence of suitable timber, and stored food was normally wrapped in packages of ti or kiekie
leaves and suspended from the rafters (Stokes n.d.:342; Figure 2.5).
On religious structures, Stokes found little coherent information. Some areas of rudimentary
paving, including several raised and terraced rectangular areas, were described as marae, but
Figure 2.5. Packages of food, probably poi, suspended to dry. From a framed photo in the Mairie at Ha’urei, and with
the permission of the Maire. The provenance of this is unclear, but it might be from the Stokes expedition.
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
41
without particular details of ritual or deities beyond reference to cannibalism etc. There are
several monoliths around the shores of Ha’urei Harbour that appeared to Stokes (n.d.:901–
903) to have been boundary markers, although he thought they, and some cavities in the walls
of terraces, might also have had ritual significance. No putatively prehistoric stone images or
carvings were recorded. Routledge and Routledge (1921:454), who visited Rapa while Stokes
was there, said:
No marae or ceremonial structure of any kind was seen by us. That name is given to three monoliths
which stand on the main bay … [and] … which, it seems most probable, were boundary stones or
had reference to the fish ponds, which are numerous. We could gain information about one building
only, termed a marae, which was in one of the exterior bays we were unable to visit. It was said to be
an insignificant enclosure, perhaps 20 ft. square, surrounded by a low wall about 2 ft. 6 in. in height,
and having at one end a semi-circular platform of the same height. It is debatable whether the present
inhabitants of Rapa really know what constitutes a marae.
Stokes found that Rapans recalled three types of canoes. Two were used in fishing: a single
outrigger called vaka (or kami’a, a term otherwise found only in Mangareva), and the ordinary
double canoe, taurua. A special double canoe (taurua tamaki) that could carry up to 40 men
was used in war. His informants had no knowledge of former sails, nor did they use them in
sailing the canoes and whaleboats of the 1920s. Stokes recovered pieces of canoes from caves
on Rapa and deduced from them that the construction had consisted of strakes built up on
a round-bottomed [dugout] hull, the latter generally requiring several pieces to be butted
together because of the absence of large trees. By 1925, the Rapan outriggers had taken the
form customary in Tahiti (in Haddon and Hornell 1975:147–148).
Rapan subsistence economy and crafts are known only sketchily from historical observations,
but their scope can be amplified, with caution, from the records obtained by Stokes. In regard
to plant foods, the Stokes’ informants were generally agreed (Stokes n.d.:197 indicates minor
disagreement) that at least one variety of sweet potato and one of banana (tautau maori) were
grown prehistorically, along with taro, gourd (koali) and Cordyline (karokaro, of which the
edible root was known as ti). The paper mulberry (aute) was also cultivated for making tapa
cloth. Introduced in the missionary era were: sugarcane, arrowroot, manioc, the ufi form of
yam (although it seems not to have thrived in Rapa), and a range of bananas, sweet potato
and taro, among other Polynesian and European crops. The manufacture of poi or popoi from
crushed and pounded taro (Figure 2.6) was probably introduced from Tahiti, although Stokes
(n.d.:180–183) suggests some contrary evidence.
Cultivated plant foods were supplemented with native sources. The most important of
these (Stokes n.d.:204–210) were para (stem bases of an edible fern), aki (tree fern starch) and
‘omeka (a type of wild yam). Others were the Pandanus (kai’ara) from which the juicy bases of
the keys were eaten, but not the kernels, the magu (fruit of kiekie, Freycinettia sp.) matoe (a wild
cliff taro), poporo (Myoporum sp.), makiri (the aerial tubers of wild yams of Dioscorea spp.) and
paraira (or puparira, the stem bases of a sedge). Three kinds of edible seaweed were also collected.
The most commonly eaten shellfish according to Stokes’ observation of the middens on pare
was the pa’ua (Chama and possibly Spondylus spp.). Other species included pipi (Tellina rugosa),
mitata (Circe pectinata), ka’i (Asaphis spp.), tupere (Venus spp.), piu’u (Modiolus sp.), akaikai
(Arca sp.), pangi (Patella spp.), i’i (Nerita morio), pu (Triton sp.), and also Trochus, Cerithium,
and various sea urchins. At least six kinds of crabs were caught, also the crayfish (koura), and
river shrimps, and octopus and squid (Stokes n.d.:212–215).
Fishing from canoes with hook and line produced, primarily, the rari, described as a large-
terra australis 37
42
Atholl Anderson
Figure 2.6. Women sitting beside a
stream pounding and kneading taro to
make poi. From a framed photo in the
Mairie at Ha’urei, and with the permission
of the Maire. The provenance of this is
unclear, but it might be from the Stokes
expedition.
mouthed red fish like a rock cod, with large anal spines. Bagnis et al. (1974:105) identify rari
as Cephalopholis coatesi, a much-esteemed species caught on the outer slopes of coral reefs. It is
probably the six-spot grouper, C. sexmaculata, in Lieske and Myers (1994:25). Two carangids
(matu, maaki), which were probably trevallies (Caranx spp.), were also mentioned, and the
uhu (a generic term for parrotfish species), which was speared near the shore (Stokes n.d.:212).
Freshwater eel occurred commonly but was not eaten. In 2002, large freshwater eels were
abundant in the streams and taro ponds and still were not eaten. Local people have a legend
about a guardian spirit in the form of a blonde-haired woman who changes into an eel, which
accounts for it not being killed or eaten. The turtle (‘onu) was much sought after but not
common. Several informants told Stokes (n.d.:153) that several kinds of seals (kumi) had once
occurred abundantly in Rapa. In fact, they still occur, one being seen near Anarua while we were
on the island in 2002.
There were several methods of fishing. Large seines (rau) of suspended ti leaves were dragged
in the shallows of the main harbour, and fish were also driven into them by men in canoes,
splashing the water. The tough, heavy wood of the mairari shrub (Dodonea viscosa) was used
to make large and medium-sized bait hooks by training the branches to grow in the required
shape. Small hooks were made from candlenut shell (tuitui) by using coral files. They were
used to catch small fish called gaga and komokomo. A hook of mairari wood with a tuitui point
was used to catch albacore (ahi). Temporary hooks were tipped with rari anal spines (Stokes
n.d.:259–262). Pearl shell was absent at Rapa, but turtle shell and whale bone may have been
used in earlier days. Basket traps (‘inaki ika), generally made from kiekie, were also used for
fishing and crayfishing, and fish were speared in shallow water, especially parrot fish. Marine
eels were snared in the coral-reef shallows by women. They used two sticks, one of which held
a bait, and the other a slip noose. Stone fish traps (pa ika), regarded as of ancient origin, can be
seen along the margins of east coast bays (Stokes n.d.:265–280).
The native rat (kiore) was seldom eaten, but birds of all kinds were hunted. Noted by Stokes’
(n.d.:211) informants were (using identifications in Fontaine et al. 1999) a dove (turuturu,
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
43
possibly the Tahitian ground dove of that name), spotless crake (kotokoto), grey duck (mokora),
snipe or lesser golden plover (torea), blue-grey noddy (paraki), brown noddy (n’goi’o), common
fairy-tern (taketake), little shearwater (kakikaki), Kermedec petrel (ke’a) and red-tailed tropic
bird (tavake), which by legend brought fire to Rapa and was hunted on the cliffs. The latter
remains a common activity today, both birds and eggs being taken.
Stokes’ informants (n.d.:446–454) also added to the spare list of weapons in the historical
data. They described a form of wooden pike with a blade at each end (omore or komore), a
wooden dart (ie), of which a broken example has been recovered, and a fighting axe or adze (tapu
te toki), which was possibly of a form similar to the Maori patu. Such an implement, described
by Stokes (n.d.:448) variously as a ‘truncheon’, ‘two-edged cleaver’, or ‘hand mattock’ formed
from a thin broad prism of basalt with a pecked handle, was found at Angairo. The former
military use of pitfalls and nets (kupega tamaki) was asserted and also of the bow and arrow,
which Stokes doubts.
Conclusions
Rapa was small and isolated and its period of pre-European habitation was probably quite brief,
perhaps beginning in the 12th or 13th centuries, judging, very imprecisely, by the genealogical
reckoning of traditionalist scholars. Through the pre-European era, and especially in the past
few hundred years, inter-clan competition seems to have been endemic and fierce, even if the
level of death and injury was relatively slight. It may have been driven by population growth
which, on an island of few native resources and very little horticultural land, forced a change
in settlement patterns from coastal habitation to hilltop defended settlements, with outlying
groups of warriors on the lower hills above the plantations. Rapan social structure was similar
to that in New Zealand. In both cases, cognatic descent defined membership of the clan (kopu
in Rapa, hapu in New Zealand), and decisions taken mainly by family heads determined rights
of residence and resource access. Before European discovery, it seems that control over Rapa
had been consolidated into the dispensation of a single clan, in which a line of high chiefs had
begun.
European discovery in 1791 was restricted to a fleeting encounter off the coast of Rapa with
men that paddled out in outrigger and double-hulled canoes. These men were comparatively
dark and stocky, unadorned, without tattoos, apparently of homogeneous status, and aggressively
acquisitive of iron; their boats were well-made outriggers and double canoes, probably with sails
on some, and capable of carrying more than 20 crew. Hill forts were seen surrounded by rows
of palisades and had a large house at the centre. There was no sign of domestic animals and
relatively little to be seen of cultivation, although much of the land had been cleared of forest.
By the early 19th century, more shipboard contacts disclosed the existence of chiefs and taro
cultivation, and from 1825 when contact began ashore, the peculiarities of Rapan subsistence
became clear. Taro was by far the most important cultivated plant, its wet-field plantations
occupying nearly all of the coastal land; others were sweet potato, a banana, the gourd, paper
mulberry and Cordyline. These were supplemented with pandanus, wild celery, edible ferns and
other native plants. Strikingly absent were breadfruit, ufi yam, coconut, kava and most forms
of banana. The pig, dog and chicken were also absent, but the Pacific rat (Rattus exulans) was
abundant and tame. Fishing and shellfishing were important pursuits, the protein they provided
being augmented by the capture of seabirds and collection of their eggs.
The reduced subsistence diversity compared with more tropical Polynesian islands was
matched by a comparably narrower inventory of material goods and crafts, according to Stokes
(n.d.). Weapons seem to have been confined to club, spear, sling, and one or two additional
terra australis 37
44
Atholl Anderson
items. Along with an absence of tattooing, there were few ornamental items, no decoration of
tapa cloth, no kava drinking, no use of fish poisons (despite the presence of suitable plants), no
use of Pandanus in mat-making, and no featherwork. Despite the relatively late existence of high
chiefs or ‘kings’, and the construction of impressive hill forts, there was very little development
of constructed marae or shrines, or the maintenance of priests.
To what extent these features of Rapan society and culture can be ascribed to remoteness
from most other islands in East Polynesia, a relatively cool climate, a small population (even if
of relatively high density), scarcity of cultivable land, or some contingencies of history are issues
that are addressed in the light of archaeological investigations discussed in this volume.
References
Bagnis, R., Mazellier, P., Bennett, J. and Christian, E. 1974. Fishes of Polynesia. Landsdowne Press,
Melbourne.
Barratt, G. 1988. Southern and Eastern Polynesia. Volume 2 of Russia and the South Pacific 1696–1840.
University of British Columbia Press, Vancouver.
Beasley, H.G. 1928. Pacific Island Records Fish Hooks. London, Seeley, Service & Co.
Buck, Sir P.H. 1925. Maori clothing, Part V. Journal of the Polynesian Society 34:87–88.
Buck, Sir P.H. 1954. Vikings of the Sunrise. Whitcombe and Tombs, Christchurch.
Caillot, A-C.E. 1932. Histoire de l’Ile Oparo ou Rapa. Librairie Ernest Leroux, Paris.
Corney, B.G. (ed), 1915. The Quest and Occupation of Tahiti by Emissaries of Spain during the Years
1772–1776 etc. Volume II. Hakluyt Society, London.
Debenham, F. (ed), 1945. The Voyage of Captain Bellingshausen to the Antarctic Seas 1819–1821.
Translated from the Russian. The Hakluyt Society, London.
Ellis, W. 1838. Polynesian Researches, during a residence of nearly eight years in the Society and Sandwich
Islands. Volume III. Fisher, Son and Jackson, London.
Fischer, S. 2001. Mangarevan doublets: preliminary evidence for Proto Southeastern Polynesian. In:
Stevenson, C.M., Lee, G. and Morin, F.J. (eds), Pacific 2000: Proceedings of the Fifth International
Conference on Easter Island and the Pacific, pp. 417–424. Easter Island Foundation, Los Osos,
California.
Fontaine, P., Fossati, O., Fossati, J., Mu-Liepmann, V., Raust, P. and Vernaudon, Y. 1999. Manu: les
oiseaux de Polynésie. Manu: Société d’ornithologie de Polynésie française, Papeete.
Haddon, A.C. and Hornell, J. 1975. Canoes of Oceania. B.P. Bishop Museum Special Publications 27,
28 and 29. Bishop Museum Press, Honolulu.
Hale, H. 1968. United States Exploring Expedition during the Years 1838, 1839, 1840, 1841, 1842.
Under the Command of Charles Wilkes, U.S.N. Ethnography and Philology. The Gregg Press,
Ridgewood N.J.
Hall, J.V. 1868. On the island of Rapa. Transactions of the New Zealand Institute 1:128–134.
Hanson, F.A. 1970. Rapan Lifeways. Society and History on a Polynesian island. Little, Brown: Boston.
Hanson, F.A. and Ghasarian, C. 2007. ‘The land belongs to everyone’: the unstable dynamic of
unrestricted cognatic descent in Rapa, French Polynesia. Journal of the Polynesian Society 116:59–72.
Hanson, F.A. and Hanson, L. 1983. Counterpoint in Maori Culture. Routledge and Kegan Paul,
London.
Lamb, W.K. (ed), 1984. George Vancouver, A Voyage of Discovery to the North Pacific Ocean and Round
the World 1791–1795, Volume I. The Hakluyt Society, London.
Lieske, E. and Myers, R. 1994. Coral Reef Fishes: Indo-Pacific and Caribbean. HarperCollins, London.
Lucett, E. 1851. Rovings in the Pacific from 1837 to 1849. Two volumes. Longmans, London.
terra australis 37
‘Dwelling carelessly, quiet and secure’: A brief ethnohistory of Rapa Island, French Polynesia, AD 1791–1840
45
MacArthur, N. 1968. Island Populations of the Pacific. The Australian National University Press,
Canberra.
Moerenhout, J-A. 1837. Voyages aux îles du Grand Océan. 2 volumes. Bertrand, Paris.
Newbury, C.W. (ed), 1961. The History of the Tahitian Mission 1799–1830, written by John Davies
missionary to the South Sea Islands with supplementary papers from the correspondence of the
missionaries. The Hakluyt Society, Cambridge.
Paulding, H. 1831. Journal of a cruise of the United States Schooner Dolphin, among the islands of the
Pacific Ocean etc. G. & C. & H. Carvill, New York.
Prebble, M.J. 2005. Islands, Floras and History: an environmental history of plant introductions and
extinction on the Austral Islands, French Polynesia. Unpublished PhD dissertation, The Australian
National University, Canberra.
Richards, R. 2004. The earliest foreign visitors and their massive depopulation of Rapa-iti from 1824 to
1830. Journal de la Société des Océanistes 118:3–10.
Routledge, S. and Routledge, K. 1921. Notes on some archaeological remains in the Society and Austral
Islands. Journal of the Royal Anthropological Institute of Great Britain and Ireland 51:438–455.
Russell, M. 1852. Polynesia: a history of the South Sea Islands, including New Zealand etc. T. Nelson and
Sons, London.
Shineberg, D. (ed), 1986. Archibald Menzies’ account of the visit of the Discovery to Rapa and Tahiti,
22 December 1791–25 January 1792. Pacific Studies 9:59–102.
Smith, S.P. 1899. History and traditions of Rarotonga. Journal of the Polynesian Society 8:61–75,
171–175, 179–187, 242–249.
Stair, J.B. 1895. Flotsam and jetsam from the great ocean. Journal of the Polynesian Society 4:99–131.
Stokes, J.F.G. n.d. Ethnology of Rapa Island. Unpublished draft of B.P. Bishop Museum Bulletin of the
Bayard Dominick Expedition. Five unbound volumes with multiple pagination (page numbers
used here are generally the circled numbers). B.P. Bishop Museum, Honolulu.
Stokes, J.F.G. 1955. Language in Rapa. Journal of the Polynesian Society 64:315–340.
Williams, H.W. 1971. A Dictionary of the Maori Language. Government Printer, Wellington.
terra australis 37
terra australis 37
3
Archaeology of the coastal sites
on Rapa Island
Atholl Anderson
Department of Archaeology and Natural History, Research School of Pacific and Asian Studies, The Australian
National University, Canberra, Australia, atholl.anderson@anu.edu.au
Introduction
Archaeological research on the coastal landscape of Rapa was confined to remains of habitation
and associated resource exploitation, leaving aside various kinds of structural remains (below).
The objectives were to define the Rapan archaeological sequence, but primarily its beginnings
(Kennett et al. 2006), to describe the use of coastal resources, and to characterise broad variation
in coastal settlement patterns. On the basis of Pacific archaeological experience generally,
and according to previous archaeological and ethnographic data from Rapa, it was assumed
that pertinent evidence most probably would be found close to the shore. We focused on
an approximately 300 m wide coastal strip, searching for open and rockshelter locations of
habitation.
It should be conceded that this strategy leaves out some coastal sites that might eventually
prove significant to an understanding of the Rapan archaeological sequence. The most
immediately obvious are the valley systems of taro gardens. A few are currently under cultivation,
but substantial areas of them are historical and archaeological. Several particular aspects of these
features were examined in the palaeoenvironmental research (Chapters 3, 10), and considered
in relation to palaeodemography (Chapter 13), but there was no systematic research on the
archaeology of Rapan agriculture in our project. Associated with some former gardens are sets
of putative house terraces that can be ascribed to historical villages, as at Tokoroa, near Aurei,
and other more enigmatic structural remains, such as pit and terrace features, stone alignments
and standing stones. There are also stone-built fish traps along the shoreline. Some of these
diverse coastal remains, which are particularly abundant around Ha’urei Harbour, have been
surveyed and investigated (Smith 1965; Walczak 2001, 2003), but the thorough research that
they deserve will require a substantial archaeological project.
Our research began in the coastal margins of Ha’urei Harbour, then expanded into each
of the external bays, with the exception of Mai’i’ (inspected by the Norwegian Expedition) and
Takao, both of which are cut off by high cliffs and have very little coastal land. We observed these
terra australis 37
48
Atholl Anderson
southwest bays and the larger offshore islands (Karapoo and Tauturoo) from the sea, but were
unable to land. The fieldwork consisted of searching all caves, shelters and exposed stream and
wave-cut sections for early cultural deposits, using an auger and spade as needed. We also cored
back-beach deposits, especially dunes and loam flats behind them, spade-tested terraces, and
engaged in limited test-pitting by excavation. Forty-seven sites were recorded and in the 31 that
were tested (Figure 3.1), we sampled for available radiocarbon dating materials, these generally
Figure 3.1. Map of Rapa Island showing location of bays and sites mentioned in this chapter.
1=R2002-30 south entrance shelter, 2=R2002-31 Anatakuri shelters, 3=R2002-33 Anakere shelter, 4=R2002-29 Tangarutu
Cave, 5=R2002-34 Angairao shelter C, 6=R2002-36 Angairao shelter E, 7=R20002-38 Noogoriki shelter, 8=R200247 Taugatu shelter, 9=R2002-46 Autea shelter, 10=R2002-28 Akatanui Shelter 3, 11=R2002-27 Akatanui Shelter 2,
12=R2002-26 Akatanui Shelter 1, 13=R2002-44 Taga shelters, 14=R2002-20 Aitoke Oven 4, 15=R2002-25 Aitoke buried
soil, 16=R2002-18 Aitoke Site 2 terrace, 17=R2002-17 Aitoke Oven 1, 18=R2002-19 Aitoke Oven 3, 19=R2002-45 Probable
R16 terrace, 20=R2002-23 Tukou shelter, 21=R2002-24 Tukou Site 8 garden soil, 22=R2002-16 Tukou Site 7 terraces and
walls, 23=R2002-13 Tukou Oven 4, 24=R2002-14 Tukou Oven 5, 25= R2002-15 Tukou Oven 6, 26=R2002-11 Tukou Oven
2, 27=R2002-12 Tukou Oven 3, 28=R2002-10 Tukou Oven 1, 29=R2002-7 Maraia Oven, 30=R2002-3,4,5 Tokoroa ovens,
31=R2002-8,9 Tokoroa ovens.
terra australis 37
Archaeology of the coastal sites on Rapa Island
49
being restricted to charcoal because of the poor survivability of shell and bone in the damp
volcanic soils of Rapa. We gave each site that we recorded a code number of R2002-n. Most of
the sites recorded were small and evidently of single-phase usage, nearly all of them small earth
ovens, but more complex sites exhibiting deep stratigraphy were also found, and excavated. The
notes on archaeological sites collected by Stokes (n.d.) provided a valuable starting point.
Observations by Stokes
Stokes (n.d.:356) observed that flat land suitable for habitation was scarce on Rapa. The valleys
are swampy and there is very little dry land at the bay heads; that which does exist is often stony.
Nevertheless, he recorded a number of coastal habitation sites. He thought there might be older
occupational remains under the 2.5 ha site of the modern village of Aurei. Other possible village
sites mentioned by Stokes (n.d.:357–358) were at Angairao, Pairirao, Iri and Anarua. The last
of these was the most promising. There are relatively extensive shore flats and low dunes, among
which some patches of cobbles had the appearance of being arranged by hand and were possibly
indicative of house floors. In addition, Stokes (n.d.:359) refers to,
many shelter caves… along the shore and some are very large. They are now used for camping in mild
weather, either by fishing parties or by field workers [i.e. people working on the taro plantations etc], and
from native accounts they were also occupied in earlier days.
Stokes (n.d.:360–365) describes caves at Togorutu [Tangarutu], Tikaioe [Akatanui] and, in
Agairao Bay, at Noogoriki. Two tiny shelters in Angairao, suitable only for sleeping in, occurred
at Pukumauatoku and Gapitau. In addition, there is the highland shelter at Taga and the tunnel
caves in Mai’i’ Bay, the latter inspected by the Norwegian Expedition (Smith 1965).
In following up Stokes’ observations by augering at Ha’urei village, we found no evidence
of former sites, nor indeed at Area village, nor were there collections of artefacts from domestic
gardens to be seen. However, the village areas are relatively substantial and small or discontinuous
sites might have been missed. At Anarua, there are stone-built walls, terrace revetments of stone
and piles of cobbles, but nothing immediately indicative of houses. Patches of cobbles on the
shoreline showed some arrangement, possibly as fishing structures, but other patches of cobbles
and gravel lacked any cultural form or content that could be observed. Augering through the
dunes that lie about 100 m behind the beach produced no charcoal or any other evidence of
habitation.
In the bays of Pariati, Akatamiro, Tupuaki, Akananue, Angairao, Akatanui, Angatakuri and
Iri, coastal flats were augered in several places, and all existing stratigraphic exposures faced and
inspected. Occasional pieces of charcoal and basalt flakes were produced, but no evidence found
that was indicative of more than fleeting occupation. More intensive augering and digging of
test pits over large areas of the bay headlands might locate remains of villages that eluded our
initial investigation. In their apparent absence, attention concentrated on the cave sites recorded
by Stokes.
Anarua Bay: Tangarutu Cave
There are two large caves on the south side of Anarua Bay. One near the point is probably ‘Ogo
Cave’ referred to by Stokes (n.d.). It is 40 m wide, 10 m high at the entrance and extends 12 m
back. However, the entrance is almost at sea level and the cave floor is piled with water-rolled
cobbles and boulders. It is unlikely to contain any cultural deposits. About 300 m northeast is
Tangarutu Cave (R2002-29). This is a very large shelter, opening to the north and facing across
Anarua Bay and north-northwest out to sea (Figure 3.2). Reached by sea, the landing is feasible
terra australis 37
50
Atholl Anderson
Figure 3.2. Tangarutu Cave from
the sea (top) and from the east (left).
Photographs A. Anderson.
Figure 3.3. Plan of Tangarutu Cave in 2002.
terra australis 37
Archaeology of the coastal sites on Rapa Island
51
in light winds from the southeast through northeast, but it is exposed to the north and west.
By land, there is a steep climb from the inner bay around a bluff above cliffs. The cave is 80 m
across at the dripline and it extends up to 29 m south to the back wall (Figure 3.3). The dripline
is approximately 30 m from the water’s edge. Within the cave is about 1450 m2 of floor area to
the dripline, and about 1700 m2 to the maximum overhang.
In front of the cave are the remains of a sand dune, rising at the eastern edge to 5.3 m high,
which sloped down seaward on to a boulder beach, and landward into the cave where, by recent
times, a surface existed at up to about 4 m above mean sea level in the middle of the cave (Figure
3.4). Most of the dune is of light yellow-grey sand, which has been quarried away, leaving a sand
‘buttress’ at each side of the cave entrance. Sand also covers most of the floor of the cave, to a
depth of 1–2 m from the original surface, the latter being mostly quarried away inside the cave
but remaining around the inner sides and back (Figure 3.5). Within the remains of the dune
at the mouth of the cave can be seen substantial layers of roof-fall basalt and, in the sand, some
humus-stained layers, incorporating pebbles and clay, which are more strongly defined towards
the top of the dune (Figure 3.6).
Excavation in 1997
The first archaeological investigations in this site were by Walczak in 1997. He dug five test pits,
each 50 cm x 50 cm, in the back of the cave along a 12 m section near the southwestern wall of
the cave (Walczak 2001:Figure 57). The intact face has retreated south by 2–5 m since 1997 as
the result of continuing sand mining in the cave. Walczak’s excavations produced a total of 36
basalt flakes, and 930 g of fish bone. The general site stratigraphy was described as three levels
(Walczak 2001:298–299, Figure 56). Level 1, about 90–50 cm below the surface, contained
negligible cultural material near the top of an indurated brown-black sand. This overlay a unit
of brown, indurated sand finely lensed with loose white sand above a basal unit of white dune
Figure 3.4. Cross-section of Tangarutu Cave, and stratigraphy of section NS1.
terra australis 37
52
Atholl Anderson
Figure 3.5. (above) Looking
into the interior of Tangarutu
Cave, showing the quarried cave
floor and marginal remnants.
Photograph A. Anderson.
Figure 3.6. Section at the eastern buttress
of Tangarutu Cave.
terra australis 37
Archaeology of the coastal sites on Rapa Island
53
sand that lay on the basalt floor of the cave. Level 2, at 50–20 cm, consisted of loose sand in
which occurred shell, fish bones, bird bone, charcoal, oven stones and flakes of basalt. Level 3,
at 20–0 cm, was a layer of very indurated brown sand on a base of very coarse sand or pebbles,
up to 2 mm in diameter. The brown sand occurred as plates or lenses often showing whitish lines
indicative of carbonate precipitation.
Walczak (2001:299–302) interpreted the stratigraphy as indicating the influence of
climatic changes. His hypothesis is that as the sand dune in front of the cave increased, its back
slope steepened and spilled sand into the cave. During dry periods, the transported sediment
remained as loose sand, but during wetter periods, the sediments underwent some soil-forming
processes which formed layers indurated by the precipitation of iron pans (alios);
… les litages de sables bruns noirs qui nous font penser à des alios, c’est-à-dire a des concentrations
gréseuses, indiquent que des phases un peu plus humides sont régulièrement venues ponctuées ces
périodes ‘sèches’. L’alios correspond à une dégradation superficielle du sable qui a tendance à s’oxyder
sous l’impulsion d’une activité végétable ou animale. Le présence de carbonates et la consistence argileuse
de ces niveeaux indurés résulte vraisemblablement de ces processus érosifs. Le niveau archéologique
correspond à l’unité 2: it interviendrait donc durant une période plus sèche, qui aurait laissé plus de
possibilités aux vents d’apporter du sable. (Walczak 2001:301–202)
From two of the test pits, samples of charcoal were dated, to two sigma, as follows: from
Tangarutu I, 495 ± 40 bp (Ly-8577), and from Tangarutu II, 330 ± 45 bp (Ly-8578). Together
with the earlier date of 370 ± 60 bp (Chapter 1), these data suggested to Walczak that the site
had been occupied relatively late. He was inclined, nonetheless, to think that there must have
been earlier occupation on Rapa, perhaps extending to the 12th century.
Excavation in 2002
Bad weather and rough seas caused by southeasterly gales prevented access to Tangarutu until
July 18 and the field team of three people landed there had to leave after six days’ work when
supplies ran out. Conditions prevented returning until August 7 when a team of six began work
again, but left after two days in deteriorating weather. Nevertheless, nine days on site enabled
us to investigate it quite thoroughly and to obtain large samples for analysis. The cave floor and
entry were mapped by tape, line-level and measuring staff, and all sections keyed for relative
depth to E1 and V1 (Figure 3.7). There was no opportunity to get back to the site and survey it
by total station, so our levels have an estimated error of ± 15 cm over 25 m distance.
The archaeological sections were chosen to disclose variation in the stratigraphy and each
was cut down to a vertical surface 50 cm wide and generally 5–10 cm into the exposure. Augering
beneath each exposed section determined whether there was lower cultural stratigraphy and
defined the depth to rock. In most cases, a spade hole showed that this was the cave floor, but
where rock could be reached only by auger it was uncertain whether it was in situ cave floor
or simply rockfall. The broadly consistent level of the rock surface, however, suggested that it
was indeed the floor of the cave, and the fact that in nearly all cases the rock was covered with
clean sand indicates that the current basal layers of archaeological remains represent the earliest
habitation in the site (Figure 3.7). Transects of auger holes and spade holes, east-west and northsouth, across the disturbed central part of the site, showed a similar pattern, but disturbance by
sand quarrying was evident well down in the sand, as indicated by a fresh wooden stake at 35
cm depth, among midden, in Test pit NS1. Nevertheless, there, and elsewhere in the transects,
clean yellow-brown sand generally overlay rock at 1.1–1.3 m below the surface.
Inspection of the exposed stratigraphy in sections around the site margins showed that
terra australis 37
54
Atholl Anderson
while there might have been some intermittent and incipient soil formation occurring in the
sand, the overriding source of the stiff, brown, silty clays and clay-enriched sands, both often
containing numerous pebble-sized and larger clasts of basalt, lies outside the cave. The thickest
surface silty-clay layer, and the most frequent evidence of silty-clay lenses at depth, occured in
the western half of the cave where a continuous talus slope extends westward out of the cave
and into the adjacent hill slope (see Figure 3.2). The slope into the cave was marked by water
channels, recent clay flows and gravel spreads. In the eastern half of the cave, silty clay was
relatively infrequent, especially along the eastern margin of the cave.
The density distribution of archaeological materials was generally the inverse of the siltclay distribution, suggesting sand was preferred as a habitation surface by the cave occupants.
Consequently, the deepest surviving stratigraphy, with the most continuous and densest
archaeological remains, occurred along the northeast margin of the site (Figure 3.7). Section T1
illustrates the stratigraphy of this area (Figure 3.8). It shows a cleaned-down west-facing section
in sand, extending to bedrock at 2 m; cultural remains reached 1.6 m. In this northeastern
quadrant of the cave, the observable sequence, confirmed by three auger holes and excavation
E:1–3 (below), consisted of sterile yellow-brown sand and basalt shatter up to boulder size,
deeper towards the cave entrance (e.g. in T1), overlying an upper unit of yellow-grey to grey
dune sand in which there were conspicuous bands of desiccated leaf and fibre, especially higher
in the section, with intermixed gravel, midden, lenses of charcoal, and burnt and broken basalt.
A lower unit, below 100–110 cm in T1, had much less fibre in darker grey-brown sand, and
charcoal, midden and burnt stone were sparse in the lower unit. The lower levels of the section
were mostly sterile yellow sand, laminated with lenses of gravel and lines of light-brown or grey
sand in which some silty clay was apparent.
In the southeastern quadrant of the site, very little of the archaeological stratigraphy remains.
Section T3 (Figure 3.9) is within the distribution range of the last phase of silty-clay deposition,
which preceded a terminal drift of grey-brown dune sand in which some cobbles had been
Figure 3.7. Investigated sections in Tangarutu Cave set out in order and by approximate height above sea level.
terra australis 37
Archaeology of the coastal sites on Rapa Island
55
Figure 3.8. Stratigraphy at Section T1, Tangarutu Cave.
terra australis 37
56
Atholl Anderson
Figure 3.9. Stratigraphy at Section T3, Tangarutu Cave.
deposited. Beneath the silty clay, grey-brown sand contained abundant midden, especially of
fish bone, and in the lower part of the grey-brown sand, 55–100 cm, cobbles, burnt and broken
stone, charcoal and wood ash indicated the margin of an oven. Below 100 cm was clean, yellow
sand. Section T2 was similar (Figure 3.10), except that the upper-silt and clay was divided
into a number of fine bands separated by yellow sand, or, below 18 cm depth, by bands of fine
gravel. Midden was sparse below, and near the base of the grey-brown sand were remains of a
fire, with beach-rolled boulders alongside it. Section V1 (Figure 3.11) showed the same upper
stratigraphy as T2, but it was thicker, at 70 cm. The grey-brown sand beneath was relatively
thin, about 20 cm, but it contained abundant fish bone and lenses of charcoal.
Moving round to the western margins of the cave, it was apparent that fairly recent fluvial
erosion has cut down, in a stepped fashion, through the mainly silt and clay sediments, removing
the upper stratigraphy from some sections (S1, S2 and excavation R1), while retaining the
upper levels at S3 and S4. There was enough overlap exposed, or dug out, to fit S4 and S3 to
R1, and the latter to S1 and S2. Section S4 (Figure 3.12) was an exposure, entirely of moreor-less horizontally banded sand, gravel, silt and clay, forming the bank of a deeply incised
watercourse (100 cm deep), which had run through the silty clay and associated sand into the
middle of the site. The watercourse appeared quite fresh and it was very probably a consequence
of sand quarrying; the removal of sediments in the centre of the site rejuvenating the erosive
capability of temporary watercourses which would otherwise have dispersed across the silt-clay
terra australis 37
Archaeology of the coastal sites on Rapa Island
57
Figure 3.10. Stratigraphy at Section T2, Tangarutu Cave.
Figure 3.11. Stratigraphy at Section V1, Tangarutu Cave.
terra australis 37
58
Atholl Anderson
Figure 3.12. Stratigraphy at Section S4, Tangarutu Cave.
Figure 3.13. Stratigraphy at Section S3, Tangarutu Cave.
terra australis 37
Archaeology of the coastal sites on Rapa Island
59
floor of the western part of the cave. The upper 80 cm of Section S3 (Figure 3.13) comprised a
complexly lensed unit of brown silt-clay, finely laminated yellow sand and lenses of fine gravel,
in which there were a few charcoal flecks. Grey-brown sand, 80–120 cm, was laminated with
thin lenses of yellow sand and contained abundant charcoal and broken oven stones, intermixed
with midden and basalt flakes.
Section S1 (Figure 3.14) begins stratigraphically at the 100 cm point in the S4 sequence. It
contained similar units of sand and silty clay, but at 30 cm (1.3 m below the top of S4), there
was a 10 cm band of yellow sand containing part of an adze preform and some bone midden and
charcoal. Section S2 (Figure 3.15) begins stratigraphically at the 121 cm point in the Section S4
sequence. It had an upper layer of indurated grey-brown sand, silt and clay in which there was
abundant charcoal and sparse midden. Beneath it was an oven with burnt, cracked stone and
very abundant charcoal. A lens of brown sand through it seemed to divide successive episodes
of use. The sand beneath the oven had the pinkish-brown aspect of having been burnt. The
S4-S1 and S2-S4 sequences show that in the western part of the site the sediments were about
as deep as in the east, 1.8 m and 2.1 m respectively to bedrock, but composed more largely of
sterile sand, silt and clay. Along the south wall of the cave, the sand was shallower and bedrock
was reached at 1.1 m to 1.2 m depth. The original shape of the site surface when it had been
abandoned would, therefore, have been a shallow hollow, higher at each side and lower in the
middle.
Figure 3.14. Stratigraphy at Section S1, Tangarutu Cave.
terra australis 37
60
Atholl Anderson
Figure 3.15. Stratigraphy at Section S2, Tangarutu Cave.
Excavation E1-3
Two small excavations were carried out at Tangarutu in 2002. In the eastern area, a 3 m x 1
m excavation area was laid out, although in the event, the excavation was so time-consuming,
because of the depth and abundance of material and the time required for field sorting, that only
the first two squares (E1 and E2) were excavated, a total of 1.5 m2.
Square E1 was completed down to natural. In Square E2, only the adjoining or west half
of the square was excavated. When the 1 m x 0.5 m excavation had reached Spit 5, bad weather
was forecast, and with the boat due early the next day, the excavation was narrowed to a column
50 cm x 25 cm (Figure 3.18). It had the same stratigraphy as E1 and was excavated in 5 cm spits
with 3 mm sieving down to 90 cm. With the arrival of the boat, the remaining Spits 19–25
of cultural material were taken out as whole samples. In addition to whole samples by spit, we
also took out two large ‘slices’ in an attempt to preserve the microstratigraphy, especially the
structure of the fibres, which was breaking up in sieving. These were at 10–18 cm and 45–58 cm
(i.e. 8 cm and 13 cm thick respectively), each being 25 cm x 35 cm in area. They were packed
into plastic boxes.
The location of Tangarutu is such that only a relatively small quantity of material could be
taken out by backpack and the capacity of a small open boat was not much more, considering
the persistently heavy seas. In the absence of convincing layer boundaries, excavation was by 5
cm spits in the main. The material was sieved through a 3 mm screen, except at a few points
where the amount of gravel required a 4 mm screen; at those times, the residue was also sorted in
a 3 mm screen. Below 70 cm, there was an abundance of broken oven stone, gravel and charcoal,
but only sparse midden, mostly of large fish, and sieving switched to a 4 mm screen and 10 cm
spits. A 30 cm x 18 cm plastic bag of whole sample (about 1.5 litres) was taken from each 5
cm spit and two from each 10 cm spit, for later sorting at the ANU through a 2 mm sieve. The
residues were retained for future analysis.
terra australis 37
Archaeology of the coastal sites on Rapa Island
61
The stratigraphy of E:1-3 (Figure 3.16) consisted of an upper unit (Level III, 0–40 cm)
of slightly compact grey-brown sand and fine gravel packed with desiccated leaves, probably
Pandanus or Cordyline. These often appeared as if strewn or blown into the cave, and some
had become ‘felted’, but some also had been plaited. The leaves were very fragile and occurred
as patches rather than as continuous features. Among them was abundant midden of shell,
crayfish and crab exoskeleton, bone from small fish, occasional bird bone, pieces of gourd and
candlenut, Pandanus keys and chewed fibres that were thought to be from Cyathea tree fern pith,
along with flattened coprolites, pieces of cordage, charcoal, some basalt flakes and small fish
hooks of candlenut and shell. Similar material, but much sparser, occurred at 40–70 cm (Level
II), and below that (Level I, 70–150 cm) were oven remains of charcoal, oven stones and ash
going down through faint grey lenses into clean yellow sand (Figure 3.17).
The material retained for analysis in this and other coastal excavations consisted of: all
identifiable shell (whole shell and apertures, spires, columns etc, plus pieces of sea urchin, and
crab and crayfish exoskeleton); all fish bone except for ribs and spines; fish scales; all bird,
reptile and mammal bone; all wood and charcoal of >20 mm largest dimension and any pieces
with potentially identifiable features; all other plant tissue (much of it candlenut, but also
Pandanus keys and chewed pieces of possible Cyathea pith) and natural fibre; all woven, plaited
or otherwise modified fibre; all stone flakes; and all artefacts of any other kind. In various
places, samples of sand and silt were taken to enable a search for phytoliths, pollen and other
microscopic components.
Figure 3.16. Excavated Square E1, Tangarutu Cave.
Photograph A. Anderson.
terra australis 37
62
Atholl Anderson
Figure 3.17. (above) Stratigraphy at
Section E1, East baulk, Tangarutu Cave.
Figure 3.18. Excavation in progress
at Square E2, Tangarutu Cave.
Photograph A. Anderson.
terra australis 37
Archaeology of the coastal sites on Rapa Island
63
Excavation R-1
A 1.0 m x 1.0 m excavation was placed in this area to sample the cultural materials in the western
half of the cave (Figure 3.19). This excavation occurred at the same time as excavation of Square
E2, and was constrained by the same problems with weather and transport. Excavation using
the same protocols of 5 cm spits, 3 mm sieving and retention of whole samples and cultural
material proceeded at R1 down to 80 cm, at which time it switched to 10 cm spits and 4 mm
sieving, with two large whole samples (each about 20 litres) taken.
The cultural material, predominantly fish bone midden with scarce bird bone or shell,
but abundant charcoal, was found mainly in the top 30–40 cm (designated Level III), where
two units of grey sand were separated by thin lenses of yellow sand. The upper grey-sand unit
was finely laminated, indicating that it had been deposited without disturbance and the sparse
midden and charcoal that occurred in it had been presumably blown in. The lower grey-sand
unit (ca. 20–35 cm) had abundant charcoal but very sparse midden. Similar material in browngrey sand from 40 cm to 80 cm (Level II) contained patches of midden and charcoal. The lowest
material (80–105 cm, Level I) was essentially an oven full of broken oven stone, charcoal and
very sparse midden (Figure 3.20).
Figure 3.19. Excavated Square R1
at Tangarutu Cave. Photograph
A. Anderson.
Figure 3.20. Stratigraphy at Section R1, Tangarutu Cave.
terra australis 37
64
Atholl Anderson
Akatanui shelters
The Akatanui shelters consist of a main site (Shelter 3) and several small shelters at the west end.
Shelter 3 is a large, narrow, southeast-facing rockshelter, known ethnographically as Tikaioe
(Figure 3.21). It is 102 m long and up to 12 m across to the dripline, but mostly less than 5
m wide, and it is divided by rock buttresses into five alcoves containing archaeological deposits
(Figure 5.22). Most of the site is 2–2.5 m above sea level, and the shelter ceiling is 20 m high at
the dripline, so the site is exposed to wind-driven rain and spray. Two small shelters are <10 m
long and only one (Akatanui 1) has a cultural deposit.
Akatanui 1
A 10 cm thick deposit of midden was found over an area of 8 m2 in this site (R2002-26). A
1 m x 1 m excavation showed the midden to contain mainly broken gastropod shell, with some
fish bone, crayfish and crab shell. Among it were two shell buttons, an iron nail and a lens from
a pair of spectacles, all of early 20th century form. Akatanui 2 (R2002-27) was a smaller shelter
150 m to the east. It had a very thin scatter of midden similar to that in Akatanui 1.
Akatanui 3
A series of test pits along this large shelter (R2002-28), which begins 120 m east of Akatanui
2, showed that most archaeological deposits were thin and sparse. Test-pit A1 (50 cm x 100
cm) had a 15–20 cm deep deposit of ash and charcoal, within which were some fragments of
wood and fibre, and a few pieces of candlenut and gastropod shell. Test-pit B1 (50 cm x 80 cm)
revealed a 4 cm thick deposit of charcoal-stained, but otherwise sterile, sand. Test-pit E1 (50
cm x 100 cm) was a sand deposit 33 cm deep, from which a very small quantity of charcoal,
fish bone and gastropod shell was recovered, and Test-pit F1 (50 cm x 50 cm) was a 47 cm deep
deposit of sand, gravel and basalt shatter with no archaeological material (Figure 3.22).
The main excavation was of 2 m2 at C1 (Figure 3.23, 3.24). This part of the shelter is
Figure 3.21. Akatanui shelters looking northeast across Alcove B. Photograph A. Anderson.
terra australis 37
Archaeology of the coastal sites on Rapa Island
65
lower than elsewhere, about 1.2 m above sea level, but Akatanui Bay is relatively sheltered. A
broad coral reef breaks the swell offshore and a band of mangroves shelters the shoreline. The
stratigraphy (Figure 3.24) disclosed by excavation comprised a single 5–15 cm thick layer of
cultural deposit, which was overlain by yellow-brown carbonate sand and basalt shatter, plus
some beach-rolled cobbles (Unit 1 in Figure 3.24), and underlain by gravel and sand, going
down on to beach-rolled cobbles and boulders (Units 3 and 4). The cultural layer (Unit 2) was
of coarse sand and gravel containing beach-rolled cobbles and boulders, abundant oven-stone
fragments and charcoal, lenses of wood ash, some molluscan and crab shell, fish bone and
several patches of fibre, including of Pandanus.
Figure 3.22. Akatanui shelters showing the alcoves and excavations, and, below, the height of the shelter floor above
sea level.
Figure 3.23. Akatanui shelters showing Alcove C, and Excavation C1 in progress.
Photograph A. Anderson.
terra australis 37
66
Atholl Anderson
Figure 3.24. Akatanui shelters. Excavation
C1 plan, above, and section, below.
Angairao shelters
On the western side of Angairao Bay, a series of rockshelters occurs along the headland running
out to Point Komire, which separates Angairao from Akananue Bay (Figure 3.25). Shelters A
and B proved unpromising. Shelter D (R2002-35) is 21 m long and up to 3.5 m wide and is
filled with unsorted talus of clay and silt plus basalt clasts up to large boulder size. This material
slopes into the shelter and, as the dripline is inside the highest point, the shelter sediments
are damp. A spade pit reached 45 cm down to either bedrock or a very large boulder without
disclosing any cultural material. Shelter C (R2002-34) has a silt floor behind a low piled-rock
wall. A test pit (50 cm x 50 cm) disclosed an archaeological deposit of shell and bone midden
with charcoal and some patches of fibre or bark near the top of it (Figure 3.25). Excavation
continued down through yellow-brown silt and clay containing abundant basalt shatter to flat
rocks or bedrock at about 95 cm deep.
Shelter E rock art and excavation
The main fieldwork was in Shelter E (Figure 3.26), which is an annexe to the southeastern end
of Shelter D, where the talus slope falls away seaward, leaving a dry floor approximately 20 m
x 5 m. On the northwest face of this shelter (R2002-36), three figures had been pecked about
terra australis 37
Archaeology of the coastal sites on Rapa Island
67
Figure 3.25. Agairao Bay showing
distribution of shelters and plan and
section of excavation in Shelter C.
Figure 3.26. Agairao Shelter E (annexe of Shelter D). Photograph A. Anderson.
terra australis 37
68
Atholl Anderson
1.5 mm deep into the smooth basalt surface (Figure 3.27). These (R2002-37) were the only
examples of rock art that were seen on Rapa. One figure (30 cm in height) at the bottom of
the panel is rudimentary. At 33 cm above the highest incision of this begins a 61 cm tall figure
representative of a bird or possibly an eel. Diagonally above that is a 25 cm human figure. The
latter looked slightly fresher than the former two. Near the figures are indistinct black markings
that might once have formed letters, and across them a vertical streak of red ochre. The human
and the bird/fish figure are in typical prehistoric East Polynesian styles, and the possible lettering
and red mark are features of rock art in New Zealand (Trotter and McCulloch 1981), and
elsewhere in East Polynesia where writing partly supplanted drawing in the missionary era.
The Shelter E excavation was 1.2 m x 0.6 m and located in the back of the shelter (Figure
3.28), where the floor consisted of a 20 cm thick layer of cow dung. Beneath this the sediments
were mainly of gravel and silt, plus numerous pieces of burnt stone and large pieces of basalt
shatter, which were increasingly frequent with depth. Below the dung was a thin layer of
fireplace remains (called ‘Oven 1’) overlying orange-yellow silt, probable ash and gravel. In the
base of this unit was turtle bone and some fish bone midden. Below that unit was a 100 cm
thick undifferentiated layer of black silt and clay with lenses of ash, abundant oven stone and
charcoal. This was designated ‘Oven 2’. Midden was very sparse, consisting mainly of some
gastropod shell and fragments of fish bone. This unit was excavated in 10 cm spits and later
divided into three levels (Level III, 60–90 cm; Level II, 90–140 cm; Level I, 140–165 cm).
Below 90 cm the deposit was damp, which probably accounts in part for the scarcity of
organic remains, and below 140 cm it was wet, and charcoal was relatively scarce. At 160 cm
deep there was a 3 cm thick band of grey clay without charcoal or any other cultural material.
The natural layer beneath comprised red-brown, sticky clay and basalt clasts up to 40 cm in size.
This was excavated for 5 cm then augered down to 186 cm below the top of the excavation,
where rock was encountered all over, presumably the bedrock of the shelter.
Figure 3.27. Rock engravings on the wall of Agairao Shelter E,
above, and field sketch of them, left. Photograph A. Anderson.
terra australis 37
Archaeology of the coastal sites on Rapa Island
69
Figure 3.28. Agairao Shelter E excavation plan, above, and section, below.
Noogoriki shelter
On the eastern side of Angairao Bay, near Tematapu Point, is a large rockshelter, 25 m long and
about 3.5 m wide in the middle (R2002-38). It is 5 m above sea level, but right on the shore
(Figure 3.29) and exposed to westerly gales and high seas. Stokes (n.d.:360) reported that a
burial platform was found inside it. A 40 cm x 40 cm excavation encountered a 60 cm thick
cultural layer beneath 20 cm of modern organic material and silt (Figure 3.30). The cultural
layer consisted of black silt and gravel with abundant basalt shatter and oven-stone fragments.
The material was compact and damp and difficult to sieve even at 4 mm mesh. It contained very
little fish bone and shell, but plenty of charcoal. A large bulk sample was taken from 70–80 cm
depth.
terra australis 37
70
Atholl Anderson
Figure 3.29. Noogoriki Shelter in Agairao Bay. Photograph A. Anderson.
Figure 3.30. Plan and section of test pit in Noogoriki Shelter.
terra australis 37
Archaeology of the coastal sites on Rapa Island
71
Additional rockshelters
Ha’urei Harbour
Near the south entrance of the harbour there is a rockshelter (R2002-30) about 12 m above
sea level (Figure 3.1). It has a floor area of 10 m x 3 m behind the dripline. A 40 cm x 40 cm
spade hole in the centre of the floor revealed a 12 cm layer of modern organic material overlying
an 18 cm thick brown loam resting on bedrock. Within the loam were small lenses of ash and
charcoal containing gastropod shell and fish bone fragments. These were sampled. At the head
of the harbour is a small rockshelter (R2002-23), which contained a thin deposit of midden and
charcoal.
Angatakuri Bay
Five rockshelters can be seen in the middle of the north-facing part of the southern coastline in
this bay, and a further three are on the east-facing shore about 250 m northwest of the point (see
Figure 3.1). All the shelters are narrow and damp and most contain no suitable sediment. Shelter
A (R2002-31), about 5 m above sea level, was mostly clean down to bedrock, but contained
a 16 m2 area of sediment which was tested by a 40 cm x 40 cm spade pit. This showed 10–12
cm of brown sticky clay and roots over a 15 cm-deep deposit of wet, dark-grey clay containing
burnt stone and charcoal, plus one flaked piece of basalt. Beneath was a thin layer of brown clay
and gravel going down to bedrock. Shelter B (R2002-32) is about 8 m above sea level facing
north and has a floor 17 m x 3 m to the dripline. Rainwater runs back into the floor, which has
a row of taro planted in it. Two auger holes and a 40 cm x 40 cm spade pit disclosed 35 cm of
sticky brown clay over a wet deposit of dark-grey clay, silt and sand, containing broken rock and
charcoal, which extended to bedrock at 116–120 cm depth.
Iri Bay
The Anakere shelter (2002-33) was investigated. This is beside, and 1.5–2.0 m above, a small
stream about 12 m above sea level. Most of the shelter floor is enclosed within a rectangular
perimeter of placed stones, which probably had something to do with the reputed mortuary
function of the site (Stokes n.d.:258). Our investigations were restricted to augering in the outer
margins of the shelter (Figure 3.31). Cores 1 and 3 produced stratigraphic records. They showed
that the top 35 cm consisted of reddish-brown, very compact clay. Below that there was mid-
Figure 3.31. Plan of Anakere Shelter, Hiri Bay, showing location of augering.
terra australis 37
72
Atholl Anderson
brown clay containing water-rolled pebbles and some small pieces of charcoal down to 140 cm.
The deposit beneath was sloppy brown clay and gravel, with no apparent charcoal. It could not
be recovered by auger below 180 cm.
Tupuaki Bay
On the west side of this bay, in coral limestone, there are clefts and small shelters. In one of these
were four human skulls, placed in secondary burial, and partly stained with red ochre. These
were not touched.
Autea Bay
Near the southern point of Autea Bay, opposite Tarakoi Island is a 15 m long x 3 m deep
rockshelter, facing north (R2002-46). In its larger of two alcoves a 50 cm x 50 cm test pit
disclosed a 10–30 cm layer of wet black clay loam, containing charcoal, resting on the bedrock.
It was overlain by 25–30 cm of damp, red-brown clay loam.
Taugatu Cave
About 200 m to the south of R2002-46, on the open coast, and about 4 m above sea level is
a fairly substantial rockshelter (Figure 3.32), 20 m long by up to 4 m deep (R2002-47). A 50
cm x 70 cm test pit revealed, beneath 20 cm of modern soil and gravel, a grey-black clay loam
extending below 60 cm into a grey silty clay down to 85 cm. Charcoal was found sparsely
throughout and a large bulk sample was taken near the base to recover datable material. Beneath
85 cm depth was brown clay and abundant basalt shatter without charcoal (Figure 3.33).
Taga shelters
It is convenient to include the Taga shelters here (R2002-44), although they are located on high
ground below the Tanga fortified site. There is a lower stone shelf, 1.5–1.8 m wide, which is very
Figure 3.32. Taugatu Shelter from the sea. Photograph A. Anderson.
terra australis 37
Archaeology of the coastal sites on Rapa Island
73
Figure 3.33. Plan and section of Taugatu Shelter test pit.
exposed and had no remaining cultural deposit and, about 2.8 m higher and to the south, an
upper shelter about 30 m long in which an 8 m x 4 m area of cultural sediment was investigated
(Figure 3.34). A 40 cm x 40 cm test pit (A) showed a 20 cm deep deposit of very friable silt and
sand that was finely laminated in darker and lighter bands. That contained burnt stone, charcoal
and some shell fragments. A second test pit of the same size (B) disclosed the remnants of an
oven in section, containing abundant charcoal and burnt stone, and some fragments of marine
shell and fish bone.
Coastal open sites
In Ha’urei Harbour, small ovens were seen in plan or section, sometimes occurring as clusters,
in roading, stream and other erosion exposures, and basalt cores and flakes occur commonly
around the shoreline (see Figure 3.1). Details of stratigraphy are omitted here, but they are
available from the author. UTM coordinates for these sites are in Appendix D. In addition
to work at the sites enumerated below, spade pitting and augering occurred at many places in
which no cultural remains were disclosed. This was especially so to the east of Aurei village in
the coastal flats at Pararaki, where ovens and stone tools had been reported. At one particular
location, on the Faraire property, augering down to 2.25 m in several places showed only surface
(modern) ovens and small flecks of charcoal in the higher sediments beneath. Charcoal flecks
in coastal clays are common in Rapa, both in garden features and stream beds. These were
sometimes sampled for palaeoenvironmental research, as in Pariati Bay, where a flight of taro
terraces was systematically augered.
terra australis 37
74
Atholl Anderson
Figure 3.34. Plan, top, of Taga shelters, and plan and section, above, of Test-pit B.
Tokoroa
One main cluster of ovens was sampled along the south side of the harbour at Tokoroa, where
a historical village had stood on terraces built down a broad spur. Ovens were sampled near
the coastline (R2002-5), from a terrace section in which at least two events of oven use are
represented (R2002-3, R2002-4), and from hillside exposures in the vicinity of the terraces
(R2002-8, R2002-9). Another oven (R2002-7) was sampled in Maraia, the next small bay to
the west. These ovens contained abundant charcoal but lacked oven stones and flaked stone. On
the north side of the harbour, ovens were sampled in and near Aitoke (R2002-17, 18, 19, 20; of
which 20 also contained some fish bone).
Tukou
In the Tukou area at the head of the harbour where there are various clay roads, more ovens
were sampled in the banks or road beds. These were ovens containing both charcoal and oven
stones (R2002-10, 11, 12, 13, 14, 15, 22, 24, 25; and at 11 and 12 also flaked basalt). Charcoal
terra australis 37
Archaeology of the coastal sites on Rapa Island
75
samples were also taken from an exposed section (R2002-16), along the edge of a knoll, beneath
a stone-walled structure that formed part of a complex of features on the top of the knoll that
included enclosed burials marked with slab surrounds.
The terrace site designated R-16 by Smith (1965:82) was of particular interest because it
had provided a radiocarbon age of AD 1337 ± 200 (M-707). It was difficult to relocate, but
after close inspection of the area we are fairly confident that it is our site R2002-45, located on
the right bank of the small stream in Tukou that enters the bay near the prominent terrace site
of Manga Parahurahu (Smith 1965:81). The site R2002-45 is a stone-faced terrace sectioned
by stream erosion. Immediately beneath the level of the stone, facing and running back into
the terrace, was a layer of friable, more-or-less stone-free, silt loam containing some pieces of
charcoal, which were collected (Figure 3.35).
Conclusions
Coastal research provided several sets of data pertinent to the objectives. The site recording, though
limited in scope, showed that coastal settlement patterns were probably somewhat different
between Ha’urei Harbour and the outer bays. This was partly a function of the distribution of
cave and rockshelter sites, which are relatively more common and usually larger in the outer bays
and which had been chosen, consequently, as habitation sites. It was also, probably, a function
of the relatively greater area of agricultural land in Ha’urei Harbour, especially at its head. The
frequent occurrence there of possible house terraces and other structural features, plus numerous
dispersed ovens, contrasts with the typical pattern in the outer bays where all suitable land has
been converted into agricultural terraces and there is relatively little evidence of open habitation
or ovens (although it is also the case that roads and other features in which ovens are exposed
are more common in Ha’urei Harbour). One point that was widely apparent is that there has
been considerably greater mobilisation of sediments since the prehistoric era. This is especially
evident at Tangarutu, but in the coastal sites generally there are thick layers of hill slope and
other sediments deposited on the archaeological stratigraphy, and much less often within it.
Excavations and test pits at numerous sites, and collection of many radiocarbon-dating
samples provides a first overview of the chronology of the Rapan sequence. This shows that
coastal locations, especially in the outer bays, were used at the beginning of the Rapan sequence,
and continued to be used throughout it (Chapter 11). Tangarutu Cave, in particular, discloses a
long sequence that begins with the earliest archaeological radiocarbon dates from Rapa.
Figure 3.35. Section of R2002-45, which is probably R-16 of Smith (1965).
terra australis 37
76
Atholl Anderson
Analysis of excavated material (detailed in Chapters 4 to 9), most of it from the Tangarutu
excavations, discloses little evidence of variation in resource use over time, despite the loss of
a few bird species and the slightly greater use of larger fish taxa early on. To some extent,
the variation is taphonomic, notably at Tangarutu where the survival of fibre and plant-food
remains, in particular, is confined to upper levels. It is very likely that these resources were
exploited at least as commonly earlier. That aside, the overriding impression is of an early and
persisting focus on fishing for relatively small species and shellfishing for gastropods. Turtle was
uncommon and there were no pigs, chickens or dogs before the historical era.
References
Kennett, D., Anderson, A.J., Prebble, M., Conte, E. and Southon, J. 2006. Prehistoric human impacts
on Rapa, French Polynesia. Antiquity 80:1–15.
Smith, C.S. 1965. Report 5: Test excavations and surveys of miscellaneous sites on the island of Rapa
Iti. In: Heyerdahl, T. and Ferdon, E.W. (eds), Reports of the Norwegian Archaeological Expedition
to Easter Island and the East Pacific, Volume 2 Miscellaneous Papers. Monographs of the School of
American Research and the Kon-Tiki Museum, 24 Pt. 2, pp. 77–87. Esselte AB, Stockholm.
Stokes, J.F.G. n.d. Ethnology of Rapa Island. Unpublished draft of B.P. Bishop Museum Bulletin of the
Bayard Dominick Expedition. Five unbound volumes with multiple pagination (page numbers
used here are generally the circled numbers). B.P. Bishop Museum, Honolulu.
Trotter, M.M. and McCulloch, B.A. 1981. Prehistoric Rock Art in New Zealand. Longman Paul,
Auckland.
Walczak, J. 2001. Le peuplement de la Polynésie orientale. Une tentative d’approche historique par les
exemples de Tahiti et de Rapa (Polynésie française). Unpublished PhD dissertation, University of
Paris I – Pantheon Sorbonne, Paris.
Walczak, J. 2003. Presentation des données actuelles sur la préhistoire de Rapa Iti (archiple des
Australes-Polynésie Française). In: Orliac, C. (ed), Archéologie en océanie insulaire: peuplement,
sociétés et paysages, pp. 28–45. Editions Artcom, Paris.
terra australis 37
4
The archaeobotany of Rapan
rockshelter deposits
Matiu Prebble
Department of Archaeology and Natural History, Research School of Pacific and Asian Studies, The Australian
National University, and Nga Pae o te Maramatanga Trust, University of Auckland, New Zealand,
matthew.prebble@gmail.com
Atholl Anderson
Department of Archaeology and Natural History, The Australian National University
Introduction
Archaeobotanical records are becoming increasingly important in resolving several issues
in the archaeology of Remote Oceania. Robust chronologies for island colonisation have
been constructed for a number of islands through direct dating of plant materials with low
inbuilt radiocarbon ages (e.g. Allen and Wallace 2007; Wilmshurst et al. 2011). The nature of
biological introductions is better understood from abundant introduced plant remains found
in archaeobotanical records (e.g. Kirch et al. 1995; Weisler 1995; Orliac and Orliac 1998;
Burney et al. 2001; Allen and Wallace 2007). Increasingly, phytolith and starch grain analyses
are being used as a first measure of plant use and exploitation for archaeological deposits without
macrobotanical remains (e.g. Horrocks and Weisler 2007). Macrobotanical remains have been
excavated from numerous sites in Remote Oceania, but surprisingly few attempts have been
made to identify these plant remains.
Excavations of rockshelter deposits from Rapa (Figure 4.1) have produced abundant
archaeological remains (Chapter 3). Here we describe the plant remains excavated from all
the rockshelters, concentrating on the largest deposit at Tangarutu, Anarua Bay. Following the
methodology of Allen and Wallace (2007), we compare radiocarbon dates obtained from young
wood and seed remains as a means of excluding the potential inbuilt age effect from dating
unidentified wood charcoal (e.g. Anderson 1991). We then compare the archaeobotanical data
with other major excavations conducted in the Cook-Austral region. Finally, an ethnobotanical
synthesis for the main plant species in the assemblages is provided. As the use of the plants
identified within these deposits has changed since European colonisation, interpreting the
terra australis 37
78
Matiu Prebble and Atholl Anderson
significance of the remains is complex and is done here in reference to the ethnobotanical notes
of Stokes (n.d.) gathered from Rapan informants, 1921–1922.
Vegetation history and archaeobotany
The vegetation history of Rapa has been described by Prebble (In press) and is discussed here
in the context of interpreting the archaeobotanical record. Stokes (n.d.) described extensive
agricultural production along some of the embayments and surrounding hill slopes and
documented the abandonment of traditional land-use practices in many parts of the island
due to population decrease and the increased importation of goods and services. The numerous
dry-stone bund features that line many of the valley floors indicate formerly extensive Colocasia
esculenta (Araceae) agriculture. Small pockets of tree crops introduced before European contact,
such as Aleurites moluccana (Euphorbiaceae), Ficus tinctoria (Moraceae), Musa spp. (Musaceae)
and Cordyline fruticosa (Laxmanniaceae), also reflect the former importance of traditional land
use. The abandonment of horticulture and arboriculture complicates the interpretation of the
archaeobotanical record, given the lack of vegetation available for comparing and interpreting
modern and past plant use.
Environmental degradation of the island is most apparent in the valleys adjacent to the
coastal rockshelter excavations at Angairao and Akatanui and the upland rockshelter at Taga.
These valleys, denuded of local vegetation, show abundant signs of long-term abandonment of
Colocasia agriculture and are overrun with feral horses, cattle and goats. Taga rockshelter lies at
ca. 150 m elevation above the head of Ha’urei Bay and is surrounded by eroding embankments
and entirely exotic shrub and grass vegetation sustained by feral grazing animals and regular
human-lit fires. The present vegetation within each valley varies considerably, primarily as a
function of differing scales of human impacts across the island.
Tangarutu rockshelter is on the south side of Anarua Bay behind a boulder beach.
Indigenous trees including Pandanus tectorius (Pandanaceae), Fagraea berteriana (Loganiaceae),
Glochidion spp. (Euphorbiaceae), Myrsine sp. (Myrsinaceae) and Rubiaceae species are found
in small pockets along the coastline of Anarua Bay. Larger stands of indigenous plants are
restricted to the ridgelines or inaccessible gulleys at the head of the valley. Agricultural activity
has been abandoned in the bay and bund terrace features are now covered with exotic and
invasive vegetation (e.g. Syzygium jambos and Psidium gjuava in the Myrtaceae). The dominant
Figure 4.1. The Central Pacific, showing the location (circled) of the main islands mentioned in the text.
terra australis 37
The archaeobotany of Rapan rockshelter deposits
79
secondary vegetation at Anarua Bay, adjacent to Tangarutu rockshelter, is mostly of exotic tree,
shrub and grass species. Driftwood lines the boulder beaches after storms and no doubt provided
fuel to people inhabiting the coast.
Palaeoecological analysis of swamps in the Anarua, Akatanui, Iri and Angairao valleys
provides only limited information about vegetation change during the period of human
occupation, as the preservation and diversity of plant remains was limited in the swamp deposits
examined. By contrast, rich swamp deposits at Tukou, in Ha’urei Bay, reveal that lowland
vegetation was considerably more diverse before human colonisation and was quickly removed in
the process of establishing agricultural fields for Colocasia esculenta (Kennett et al. 2006; Prebble
In press). Based on the palaeobotanical evidence at Tukou, forest resources available to the
initial colonists would have varied little across each valley. On initial colonisation, people would
have encountered large Pandanus-dominant coastal swamp forest built up along riverbanks and
behind the boulder beaches. Along the inland margins of these swamp forests were lowland
forest trees and shrubs, including an extinct palm (Arecaceae), Celtis pacifica (Ulmaceae) and
Glochidion spp. Sedge and rush wetlands, dominated by Schoenoplectus subulatus var. subulatus
and Eleocharis spp. (both in Cyperaceae), would have persisted within the more permanently
swampy areas, particularly along the margins of the Pandanus forests.
The rockshelters and preservation of plant remains
With the exception of Tangarutu, the coastal rockshelters are less than 10 m in maximum
dimension and do not preserve large amounts of plant remains. The elevation of the coastal
rockshelters varies from site to site, with all the shelter floors lying at least 2 m above sea level.
Shelter entrances are usually behind large boulder beaches away from storm wave action or
tidal surges. There is no evidence of major disturbance from coastal influences at any of the
rockshelter sites. There is also no evidence to suggest that any archaeobotanical deposit is
derived from beach drift, other than in material brought into the site by humans, but some
remains of Pandanus or other plants growing near the shelter entrances might have entered the
sites naturally. All shelters consist of a basaltic platform with overlying calcareous beach or dune
sands, including coral pebbles and occasional boulders that have been accumulating, probably,
since the mid-Holocene sea-level high stand of up to 2 m above present levels (Dickinson 2001).
Most shelter floors (e.g. Tangarutu west section) show signs of recent alluvial and colluvial
sedimentation, with incised channels stretching along the front of the inner face of each shelter.
Reddish brown basaltic sediments associated with this channelling appear to be derived from the
talus hill slope above the buttress of each shelter. The dry beach sands are partly covered by an
overburden of these in-washed sediments. Despite this recent sedimentation, most deposits are
dry and preserve plant materials with little evidence of diagenesis through periodic inundation
and oxidation.
The excavations
Excavations on Rapa concentrated on coastal rockshelters, as they were expected to yield the
best evidence of initial habitation on the island, and therefore of initial plant use. Compared
with Tangarutu, excavations at Akatanui, Angairao and the upland shelters at Taga exposed
shallow stratigraphy with a low abundance and diversity of cultural material and plant remains.
Plant remains were recovered by hand picking during dry sieving (3 mm mesh) or as individual
finds from different layers, particularly of leafy material. The base of the Angairao rockshelter
stratigraphy dates to ca. AD 1350–1250 (Chapter 11), which is essentially the same as the
terra australis 37
80
Matiu Prebble and Atholl Anderson
oldest age at Akatanui, while the rockshelter at Taga suggests later occupation of this upland
site. The Akatanui sequence is more indicative of continuous occupation from ca. AD 1350–
1450 up to the period of European colonisation, marked by introduced materials (e.g. glass).
Apart from Tangarutu, archaeobotanical remains were richest at Akatanui, with abundant wood
charcoal fragments and Aleurites moluccana endocarps in cooking oven deposits. A summary
of the archaeobotanical remains found at Akatanui, Angairao and Taga, including charcoal
and endocarp abundance, is presented in Table 4.1, with radiocarbon ages from plant remains
presented in Table 4.2.
Table 4.1. Plant materials including unidentified charcoal and Aleurites moluccana endocarps obtained from the
Angairao rockshelter E, 2nd oven.
Spit/depth (cm)
Charcoal (g)
Aleurites moluccana endocarp (g)
Other (g)
Spit 3/20–30
4.18
–
23.31 (ash)
Spit 4/30–40
1.41
–
–
Spit 5/40–50
1.55
–
–
Spit 6/50–60
2.98
0.30
–
Spit 7/60–70
5.70
0.33
–
Spit 8/70–80
1.52
–
–
Spit 9/80–90
2.38
0.21
–
Spit 10/90–100
2.61
0.15
0.14 (leaf)
Spit 11/100–110
<10
Table 4.2. List of radiocarbon dates on unidentified charcoal obtained from the Angairao Rockshelter E, 2nd Oven.
Sample/depth (cm)
Weight (g)
Lab code
C14 age (uncal.)
2 s cal. AD
E1 Spit 10/90–100
20
ANU 11851**
500+/-50
1394–1618
E1 Spit 11/100–110
<5
UCI 14767*
375+/-15
1487–1627
E1 Spit 11/100–110
<5
UCI 14766*
220+/-20
1653–1803
Presented are the sample depths, radiocarbon laboratory (UCI=University of California Irvine; ANU=The Australian National University), weight of
charcoal dated/the amount of charcoal found in spit, laboratory sample codes, uncalibrated determinations (*AMS, **conventional) and calibrated ages
to 2s. Dates were calibrated using the program Calib 6.0. (Stuiver et al. 2005) using the Southern Hemisphere Calibration Curve (McCormac et al. 2004).
The Tangarutu sequence
The Tangarutu rockshelter at Anarua Bay is of such a notable size (80 m x 40 m) that it is
likely to have been used from the earliest period of settlement. Small test excavations on the
rockshelter beach sand at the southern end of the shelter by Walczak (2001, 2003) produced
calibrated radiocarbon dates between AD 1450 and 1700 (Ly-8577 and Ly-8578, see Chapter
11). Since Walczak’s excavations, a large amount of archaeological material has been removed in
the process of extracting sand for building projects. This may have resulted in the destruction
of most of the archaeological deposit. During the 2002 expedition, the remaining undisturbed
deposits towards the back of the shelter were sampled with a sand auger and about 4 m2 of the
deepest and richest deposits were excavated.
The east section of the shelter revealed the deepest archaeological deposit on the island,
with approximately 150 cm of continuous cultural strata. It included leaf and seed fragments,
terra australis 37
The archaeobotany of Rapan rockshelter deposits
81
abundant wood and charcoal remains of ovens, and other cultural material. A number of
coprolites consisting of densely matted organic matter, often with hair and small crushed bones,
were recovered during excavation. Plant remains were recovered by hand picking during dry
sieving (3 mm mesh) or as individual finds from different layers, particularly of leafy material.
This deposit was excavated in two 50 cm x 100 cm sections (East Section 1 and East Section
2) at 5 cm spit intervals for the first 70 cm and at 10 cm spit intervals for the remainder of the
deposit. Only East Section 1 (E1) is presented here.
Analysis
Seed and fruit analysis
Seed and fruit remains were identified on the basis of comparison with reference material held at
the Department of Archaeology and Natural History, ANU, macrobotanical reference collection.
Archaeological wood charcoal analysis
Charcoals were described by comparison with a wood charcoal reference collection composed
by M. Prebble from collections made on Rapa, and E. Dotte (unpublished data) from a number
of collections from New Caledonia. The reference collection is representative of woody taxa
regarded as fuel woods which were recorded by J.F.G. Stokes and E. Stokes between 1921 and
1922. The local plant names, botanical names and uses of these taxa are listed in Table 4.3.
Some taxa were not included in the reference collection due to their rarity, namely Metatrophis
margaretae (Urticaceae), known only from a single male specimen (Timothy Motley and Jean
Yves Meyer pers. comm.), and Santalum insulare var. margaretae (Santalaceae). Taxa included
in Stokes’ list that are known as post-European contact introductions were not included in the
reference collection (Albizia lebbeck, Citrus spp., Coffea arabica and Inga edulis).
Microscopy was completed using an Olympus BH-2 epi-illumination incident compound
microscope with 10x, 20x and 50x objectives. Given the low diversity of woody plant taxa on
Rapa (<70) and the low number of preferred fuel woods (Table 4.3), only up to 50 charcoal
pieces above 10 mm in widest dimension were randomly selected for analysis from each spit. As
the basal spit (Spit 21) revealed little charcoal, additional charcoals were examined from Spit 20.
Palynological analysis
Two coprolite samples were chosen for palynological analyses to assess the potential for identifying
dietary information about humans or dogs (Canis familiaris), the most likely producers
(although dogs have not been recorded for pre-European Rapa). One cubic centimetre (cm3)
volume sub-samples were taken from each coprolite. Pollen analysis of the coprolite samples
was conducted using the standard preparation techniques. Microscopic charcoal fragments were
also counted for each sample using the point count method. Pollen and spore identification was
assisted by reference material collected in the field and regional reference collections held in the
Department of Archaeology and Natural History, ANU.
Results
Archaeobotanical assemblages
The stratigraphic profile of the East Section 1 excavation is shown in Chapter 3. Associated
weights for each type of plant material identified from each excavation spit are presented in
stratigraphic order in Figure 4.2. The deposit can be divided into three units – beach sand with
almost no cultural remains, and basal and upper cultural units. In the basal cultural unit of
terra australis 37
82
Matiu Prebble and Atholl Anderson
Table 4.3. List of pre-European contact fuel woods used on Rapa, based on a botanical collection and ethnobotanical
survey conducted by Margaret and John Stokes between 1920 and 1921. ANU reference wood charcoals are also listed.
Rapa
vernacular
Stokes identification
Family: botanical name
Other uses apart
from fuel wood
Reference charcoal
A’i
Tree (sandalwood)
–
Santalum neocaledonicum*
Aito
Tree (Weinmannia rapensis)
Santalaceae: Santalum insulare var.
margaretae
Cunoniaceae: Weinmannia rapensis
Boat building
Weinmannia rapensis
Aki
Tree fern
Cyatheaceae: Cyathea spp.
Food, graters for oil
Cyathea medularis
Anei
Tree (Compositae)
Asteraceae: Fitchia rapensis
House posts
Fitchia rapensis
Ariki
Bush (Homolanthus sp.)
–
–
Gaio/Ngaio
Tree (Myoporum sp.)
Euphorbiaceae: Homolanthus stokesii,
H. nutans
Myoporaceae: Myoporum rapensis
Myoporum rapensis
Gatae/Ngatae/
Patai
Kaeka
Boat building,
medicinal
Canoes
Introduced tree (Erythrina indica) or Fabaceae: Erythrina variegata
possible Inga edulis
Tree (Eugenia sp.), probably
Myrtaceae: Syzygium malaccense
Food
introduced
Tree (Glochidion longfieldiae)
Euphorbiaceae: Glochidion longfieldiae –
Kaema /Kaima
Erythrina variegata
Syzygium malaccense*
–
Kai’ara
Tree (Pandanus sp.) natives state
‘indigenous’ and ‘introduced’
Pandanaceae: Pandanus tectorius
Food, medicinal,
Pandanus tectorius
house thatch (recent),
stringing, candles etc
Kakatua
Shrub (Piper spp.)
Medicinal
Macropiper puberulum
Karaka
–
Pouteria cinerea
Koe
Tree (Sapotaceae/Nesoluma
polynesicum)
Shrub/bamboo
Piperaceae: Macropiper sp., Peperomia
spp.
Sapotaceae: Nesoluma spp.
Poaceae: Schizostachyum glaucifolium
–
–
Kofe
Tree
Araliaceae: Meryta pauciflora
–
Schefflera spp.
Koi’ivai
Rubiaceae: Coprosma spp.
–
Fabaceae: Sophora tomentose
Mairirakau
Shrub (Rubiaceae/Coprosma
rapensis)
Tree (Sesbania sp./Sophora
tetraptera)
Shrub (Alyxia stellata)
Apocynaceae: Alyxia stellata
House posts, modern
tool handles
–
Alyxia stellata
Maroro
Bush (Metatrophis margaretae)
Urticaceae: Metatrophis margaretae
–
–
Mati
Introduced tree (Ficus tinctoria)
Moraceae: Ficus tinctoria
Cordage
Ficus tinctoria
Pakora/Nioi
Shrub (Plectronia kakenua var.
obiculata or possibly the tree
Cyclophyllum barbartum
Rubiaceae: Plectronia sp./Cyclophyllum –
barbartum
Po’oto
Tree (Eurya rapensis)
Theaceae: Eurya japonica var. Nitida
Purau
Introduced tree (Hibiscus sp.)
Malvaceae: Hibiscus tiliaceus syn
Talipariti tiliaceum
Pua
Tree (Fagraea berteriana)
Pūru/Purum
Rakau papa’a
Maiage
–
Sophora tomentosa*
Cyclophyllum barbartum*
Eurya japonica var. nitida
Tree or shrub (Sida rhombifolia)
House posts, canoes,
paddles, boats,
cordage, sandals,
medicinal
Loganiaceae: Fagraea berteriana
Houses, canoes,
medicinal, adornment
Euphorbiaceae: Claoxylon collenettei** Perfume
–
Introduced tree (Albizia lebbech)
Fabaceae: Albizia lebbeck
–
–
Raupata
Tree (two forms of Compositae)
Asteraceae
–
–
Rautea ***
Tree
Corokia collenettei and C. serrata
–
–
Tiairi or tuitui
Introduced tree (Aleurites triloba)
Euphorbiaceae: Aleurites moluccana
Canoes, lights, oil,
Aleurites moluccana
Tireuei
Tree (Melastoma sp./Celtis
paniculata var. rapensis
Ulmaceae: Celtis pacifica
–
Celtis pacifica, C. hypoleuca
Hibiscus tiliaceus syn.
Talipariti tiliaceum
Fagraea berteriana
*From Emilie Dotte (unpublished reference collection); all other woods are from Rapa and other islands of the Austral archipelago.
**Pūru given for Claoxylon collenettei by Papaafatu (pers. comm.), Rapa.
***Rautea/lautea recorded by Tim Motley (unpublished data).
terra australis 37
The archaeobotany of Rapan rockshelter deposits
83
terra australis 37
Figure 4.2. The material weights per excavated spit (the top 10 spits were excavated to a depth of 5 cm, with the lower seven spits to a depth of 10 cm) for the Tangarutu rockshelter from
the E1 section. The deposit is divided into three units, a sterile basal sand layer underlying a charcoal-rich oven feature with oven stones within a sandy matrix (basal cultural unit). From
Spit 7 (35-40 cm) to the surface, Aleurites moluccana and pieces of uncharred wood dominate a sandy matrix overlying a leafy layer at between 35 cm and 30 cm in depth (upper cultural unit).
Introduced taxa labelled in bold text.
84
Matiu Prebble and Atholl Anderson
the deposit, from 140 cm to 35 cm in depth, we found fragments of charcoal associated with
rounded basaltic stones indicative of an oven feature. Small amounts of Aleurites moluccana
(10–35 g per spit; see Figure 4.2) were located in the upper part of this unit from 120 cm to 40
cm. The archaeological charcoal from 110–130 cm (spits 20 and 21) and 15–20 cm (Spit 4) is
presented in Figure 4.3 and reveals distinct changes in fuel use. The basal spits are dominated by
introduced cf. Hibiscus tiliaceus syn. Talipariti tiliaceum (Malvaceae) and cf. Cordyline fruticosa,
but this changes in the uppermost spit to predominantly indigenous fuel and timber woods. The
upper spit is dominated by the endemic species cf. Myoporum rapensis (Myoporaceae) and cf.
Fitchia rapensis (Asteraceae), now restricted to small coastal populations and upland populations
respectively (Figure 4.4 shows modern reference and archaeological charcoals).
Above 35 cm, a 5–10 cm thick layer of leafy remains distinguishes the upper unit, and upon
sieving it produced pieces of Hibiscus tiliaceus syn. Talipariti tiliaceum cordage and fragments
of plaited Pandanus and Freycinetia (Pandanaceae) baskets (Chapter 5). The amount of A.
moluccana in each spit increased to weights between 70 g and 140 g per spit. Also in this unit
the amount of unburnt wood remains increased, whereas the amount of charcoal and oven
stones decreased. Fragments of the bottle gourd Lagenaria siceraria (Cucurbitaceae; Figure 4.4)
and Pandanus keys (individual drupes of syncarpous fruits) were also identified within this unit.
Other plant remains, including the abundant uncharred wood, have yet to be identified.
Two small, consolidated organic parcels located from the upper unit of E1, thought to be
coprolite samples on the basis of the presence of hairs and crushed bone, were processed for
pollen as a further means of identifying the presence of plant species in the Tangarutu deposit.
Percentage pollen diagrams of these samples are presented in Figure 4.5. The uppermost sample
found in Spit 1 (5 cm in depth) produced a range of pollen and spores, most known as winddispersed palynomorphs. The presence of Freycinetia and Euphorbiaceae pollen in these samples
suggests that material from these plants may have been brought into the site. In the lower sample,
found in Spit 2 (10 cm in depth), 75% of the palynomorphs identified are from Freycinetia (F.
arborea or F. rapensis). Other pollen types probably introduced rather than wind blown into the
site include Rubiaceae, cf. Liliaceae type and a high percentage of Cyperaceae pollen (ca. 18%).
The high percentage of Freycinetia pollen is indicative of the probable introduction of these large
flowers into the site, either brought in with leaves used for weaving and cordage, or with fruits
for consumption.
Chronology
The archaeological material yielding the oldest radiocarbon ages at Tangarutu consisted of wood
charcoal, unidentified to taxa at the time, immediately after excavation, when no means of
identification was available (the formation of a Rapan identified wood collection being one of
the objectives of the project, which took some time to reach fruition). Radiocarbon dates on
charcoal samples, unidentified to taxa, indicate that the base of Tangarutu dates to at least AD
1300, with the upper deposits beginning about AD 1600 and extending up to immediately
before European arrival (Chapter 11).
Later, three radiocarbon ages were taken from fruit and nut remains from East Section 1
(see Table 4.4), one from Aleurites moluccana endocarp, one from Lagenaria siceraria pericarp
and one from a Pandanus cf. tectorius key. Unlike unidentified charcoal, the inbuilt age of fruit
and nut remains is expected to be merely a year or so. Radiocarbon dates from A. moluccana
endocarp provide the most reliable chronology, because of their inherently low inbuilt age and
their availability through almost all of the stratigraphic units. The earliest date from A. moluccana
endocarp has an age range of AD 1049–1628 (ANU 12102), which brackets the age range
terra australis 37
The archaeobotany of Rapan rockshelter deposits
85
terra australis 37
Figure 4.3. The archaeological charcoal count data (by tentative family and species determinations) for the Tangarutu rockshelter from three spits of the E1 section. The basal spits (E1 spits 21
and 20) are from the base of a charcoal-rich oven feature with oven stones within a sandy matrix and represent the (basal cultural unit of the E1 section). The upper spit (E1 Spit 4) is from
a leaf (Freycinetia and Pandanus), candlenut (Aleurites moluccana endocarp) charcoal and wood-rich unit. Taxa are arranged firstly by first appearance, then by use (e.g. fibre etc). Introduced
taxa labelled in bold text.
86
Matiu Prebble and Atholl Anderson
Longitudinal section, modern Fitchia rapensis, 20x
Longitudinal section, Fitchia rapensis, E1 Spit 4, 20x
Cross section, modern Fitchia rapensis, 20x
Cross section, Fitchia rapensis, E1 Spit 4, 20x
Cross section, modern Myoporum rapensis, 20x
Cross section, Pandanus wood, E1 Spit 20, 10x
Cross section, Myoporum rapensis, E1 Spit 4, 20x
Cross section, Aleurites moluccana endocarp, E1 Spit 4, 20x
Figure 4.4. Digital images of modern reference and archaeological charcoals, Fitchia rapensis, Myoporum rapensis,
Pandanus and Aleurites moluccana endocarp.
terra australis 37
The archaeobotany of Rapan rockshelter deposits
87
terra australis 37
Figure 4.5. Percentage diagram of pollen and spore assemblages from two consolidated organic samples from the East Section 1 excavation (spits 1 and 2), Tangarutu rockshelter. Taxa with
>5% of total palynomorph sum are presented as bars; samples with proportions <5% are presented as triangle symbols; and total palynomorph counts. Introduced taxa labelled in bold text.
88
Matiu Prebble and Atholl Anderson
Table 4.4. List of radiocarbon dates obtained from the Tangarutu rockshelter East section Square E1 and for
comparison E2.
Sample/depth (cm)
Material dated
Weight (g)/amount in spit (g) Lab code
C14 age (uncal.) 2 s cal. AD
E1 Spit 4/20–25
Lagenaria siceraria pericarp
<2/6
UCI 14763*
320+/-15
1510–1648
E1 Spit 4/20–25
Aleurites moluccana endocarp
12/100
ANU 12101**
410+/-60
1442–1638
E1 Spit 11/55–60
Aleurites moluccana endocarp
5/5
UCI 14772*
345+/-20
1501–1642
E2 Spit 18/100–110
Aleurites moluccana endocarp
12/n.d.
ANU 12100
380+/-60
1449–1648
E1 Spit 19/110–120
Pandanus cf. tectorius key
2.55/20
ANU 12102*
660+/-150
1049–1628
E2 Spit 23–25/130–150 Unidentified charcoal
10/10
ANU 11848**
710+/-70
1228–1407
East Section T1/150
<10/20
UCI 14769**
905+/-20
1152–1222
Unidentified charcoal
Presented are the sample depths, radiocarbon laboratory (UCI=University of California Irvine; ANU=The Australian National University), material
dated, weight of material dated/the amount of material found in spit, laboratory sample codes, uncalibrated determinations (*AMS, **conventional)
and calibrated ages to 2s. Dates were calibrated using the program Calib 6.0. (Stuiver et al. 2005) using the Southern Hemisphere Calibration Curve
(McCormac et al. 2004).
from a piece of unidentified charcoal from the lowest unit of the excavation, AD 1128–1407
(ANU 11848). Unfortunately, the basal Pandanus key produced a large uncalibrated error due
to inadequate carbon concentration.
Archaeobotanical research in the Austral and Cook islands
Tangarutu rockshelter is notable for having well-preserved plant remains found in very high
concentrations that are comparable to other excavations from tropical and subtropical Remote
Oceania (e.g. Kirch et al. 1995; Weisler 1995; Burney et al. 2001; Allen and Wallace 2007).
Aside from charcoal, the recovery of identified plant remains from archaeological sites in the
Cook and Austral islands has been minimal. In tables 4.5 and 4.6, indigenous and introduced
plant remains excavated from Tangarutu are compared with those from excavations at Ureia,
Aitutaki (Allen and Wallace 2007), and Tangatatau, Mangaia (Kirch et al. 1995) in the Cook
Islands. Wood charcoal (unidentified to species) and endocarps of Aleurites moluccana are the
most common archaeobotanical remains found at each site.
Indigenous plant remains
Charred and uncharred keys and some wood charcoals of Pandanus cf. tectorius (Figure 4.4)
represent the indigenous food plant identified at Tangarutu, Tangatatau and Ureia (Table 4.3).
Apart from Pandanus, Hibiscus tiliaceus syn. Talipariti tiliaceum is the only wood found at Ureia
and Tangarutu, where it was probably used for both fibre and fuel (Table 4.3). All the remaining
taxa identified at Tangarutu are, from the ethnobotanical notes of Stokes (n.d.), primarily fuel
woods and mostly endemics, including species of Myoporum, Fitchia and Hernandia (Figure 4.3).
Pandanus tectorius Parkinson
It is possible that remains of Pandanus represent indigenous species other than P. tectorius,
but this cannot be established from the macrobotanical remains alone. Whistler (1991) has
questioned the status of Pandanus tectorius as an introduced species to Remote Oceania. He
suggested that P. tectorius of both Tonga and Samoa is represented by both indigenous and
introduced populations, but queried the indigenous status of Pandanus tectorius in the Cook,
Austral, Society, Marquesas and Hawaiian islands. Pandanus pollen has been identified from prehuman Holocene sediments on Rapa, Rimatara (Prebble and Wilmshurst 2009) and Mangaia.
terra australis 37
The archaeobotany of Rapan rockshelter deposits
89
Table 4.5. Comparison of indigenous plants identified from Ureia, Aitutaki (Allen and Wallace 2007), Tangatatau,
Mangaia (Kirch et al. 1995), and Tangarutu archaeobotanical deposits.
Indigenous taxa
Ureia
Tangatatau
Tangarutu
Argusia argentea syn. Tournefortia (Boraginaceae) wood
+
?
n
Caesalpinia major (Fabaceae) seed
?
+
n
Freycinetia sp. (Pandanaceae) leaf
n
n
+leaf, pollen
Guettarda speciosa (Rubiaceae) wood
+
?
n
+
-
Hernandia sp. (Hernandiaceae) seed
1
Hibiscus spp. cf. tiliaceus (Malvaceae) fibre, bark
+
?
+ fibre, charcoal?
Pandanus sp. (Pandanaceae) leaf, keys
+keys
+keys, leaf
+keys, leaf, charcoal
Pemphis acidula (Lythraceae) wood
+
?
n
Pouteria grayana syn. Plachonella (Sapotaceae) seed
+
?
n
Palmae (Arecaceae) wood
+
?
?
Terminalia (Combretaceae) wood
+
?
n
+
?
n
2
Thespesia populnea (Malvaceae) wood
+= identified in the deposit; ?= found on the island but has yet to be identified in the deposit; n=is currently not located on the island.
1 The indigenous or exotic status of Hibiscus cf. tiliaceus syn. Talipariti tiliaceum on the Cook and Austral islands is uncertain (Whistler 1991; Florence 2004).
2 The indigenous status of Thespesia populnea on Rapa is uncertain (Florence 2004).
Table 4.6. Comparison of introduced plants identified from Ureia, Aitutaki (Allen and Wallace 2007), Tangatatau,
Mangaia (Kirch et al. 1995), and Tangarutu archaeobotanical deposits.
Introduced taxa
Ureia
Tangatatau
Tangarutu
Aleurites moluccana (Euphorbiaceae) endocarp
+
+
+endocarp (Figure 4.4)
Artocarpus altilis (Moraceae) wood, fruit
+
?
?
Calophyllum inophyllum (Clusiaceae) wood
+
?
n
Cocos nucifera (Arecaceae) endocarp, husk, leaf, bracts, wood charcoal
+
+
?
Colocasia esculenta (Araceae) corm
?
+
+
Cordia subcordata2 (Boraginaceae) seed
+
?
n
Cordyline fruticosa (Laxmanniaceae) stem, leaf, charcoal
+
+
+leaf, charcoal?
Cyrtosperma merkusii (Araceae) corm
?
+
n
Ipomoea batatas (Convovulaceae)
?
+
?
Inocarpus fagifera (Fabaceae) wood charcoal
?
+
n
Lagenaria siceraria (Cucurbitaceae) pericarp
?
+
+
Musa sp. (Musaceae) leaf
?
+
?
Saccharum officinarum (Poaceae) stem
?
+
?
Schizostachyum glaucophylum (Poaceae) stem
?
+
?
1
+= identified in the deposit; ?= found on the island but has yet to be identified in the deposit; n=is not located on the island and is unlikely to be found
in the deposit.
1 Considered to be indigenous to the Cook and Austral islands, except on Rapa but has been repeatedly introduced by Polynesians, contributing to the
expansion of its natural range on islands (Athens 1997).
2 Considered to be introduced to the Cook Islands (Bill Sykes pers. comm. 2007).
terra australis 37
90
Matiu Prebble and Atholl Anderson
Ellison (1994) establishes the indigenous status of Pandanus on these islands and presumably
the other Cook and Austral islands. Further support for the indigenous status of Pandanus in
Remote Oceania comes from an unpublished report (in Athens 1997) of Pleistocene macrofossil
remains of Pandanus on the Hawaiian Islands, where it has a minimum age of 500,000 years.
The botanist Archibald Menzies (Shineberg 1986) on the initial European visit to Rapa in
1791 mentioned Dracaena leaves, which probably refers to Pandanus, in reference to a girdle
suspended around the waists of the local Polynesians. Indications from the pollen records from
Rapa suggest that these trees were probably common at first European contact. Whether the
Pandanus population had natural as well as cultural origins is unknown.
Pandanus is a dioecious genus with extreme morphological diversity that can be influenced
by cultivation practices. This has resulted in a number of inconsistent taxonomic determinations
of members of this genus (see Stone 1976, 1988; and St. John 1976, 1979). From his Pacific
Island plant collections from the 1920s, St. John (Fosberg and St. John 1934) identified 13
endemic species of Pandanus on Rapa alone. Most of these species have subsequently been
grouped into Pandanus tectorius, and one other species has yet to be described (Tim Motley
and Jacques Florence pers. comm. 2004). The morphological diversity in the genus Pandanus
that St. John identified on Rapa may be a result of genetic or environmental processes, but may
also reflect the introduction of cultivated varieties by Polynesians both pre- and post-European
contact.
The ethnobotanical importance of Pandanus on Rapa is confusing. One of Stokes’ (n.d. Box
7.1) main informants, Teraau, noted the indigenous status of kai’ara/fara (Pandanus) and magu
(Freycinetia spp.), but asserted that there was no mat-making, which is unlikely and contradicted
by abundant ethnographic evidence (Chapter 2) and some archaeological remains (Chapter 5).
Hibiscus L. and Thespesia populnea (L.) Solander ex Correa
The lack of early ethnographic evidence for fibre production involving the Polynesian miro
(Thespesia populnea) and purau (Hibiscus spp.) is intriguing. The indigenous status of T. populnea
on Rapa is uncertain and it is more often considered a Polynesian introduction (Florence 2004).
The few specimens that now remain on Rapa appear to survive only by cultivation, but the
overwhelming effect of feral grazing animals precludes any reasonable test of whether they
could survive otherwise. The most reliable and earliest radiocarbon age obtained directly from
archaeobotanical material in the Cook and Austral islands material comes from cf. T. populnea
wood at the Ureia site, Aitutaki, dated to AD 1240–1405 (Wk-18408; Allen and Wallace 2007).
No pre-human remains of T. populnea, pollen, wood or charcoals have been identified from the
Cook or Austral islands. The status of Hibiscus is more confusing given that Florence (2004)
has described two species (H. australensis and H. tiliaceus) and one subspecies (H. tiliaceus
subsp. tiliaceus). Like Thespesia, no pre-human remains of Hibiscus have been identified from
the Cook or Austral islands, and for this reason these genera should be considered Polynesian
introductions to Rapa and perhaps to the most southern islands of the Cook and Austral islands.
Introduced plants
Introduced taxa make up the greatest proportion of plant remains identified from the base of
the main E1 sequence. The basal cultural unit outlined is indicative of an oven and an associated
midden, initially used 1300–1400 AD. The dominance of remains from tropical and introduced
fibre plants (Cordyline fruticosa, Hibiscus tiliaceus syn. Talipariti tiliaceum and Thespesia populnea
and candlenut Aleurites moluccana) suggests that the early economic activity on this sub-tropical
island centred on processing plants imported from the tropics. Fibre plants used for cordage and
terra australis 37
The archaeobotany of Rapan rockshelter deposits
91
basketry are essential components of Polynesian material culture for clothing, binding for boats
and weapons, fishing equipment, lashing for building construction and numerous other tasks.
Apart from Pandanus tectorius and Freycinetia spp., there are few indigenous plants that could
provide fibre essential for sustaining Polynesian material culture.
In his ethnography of Rapa, Stokes (n.d. box 7.1) doubted the claims from his informants
that all the main fibre plants aside from Pandanus and Freycinetia – purau, orā, aute and mati
(See Table 4.3) – are indigenous. Even if these plants were not indigenous to Rapa, as on other
islands, they remain prime candidates for early introduction given their economic importance
to Polynesians. Almost all the plants identified in the assemblage, indigenous or introduced, are
now rare or absent from the vicinity of the rockshelter and from Anarua Bay.
Aleurites moluccana (L.) Willdenow (Euphorbiaceae)
Of the introduced plant taxa represented in rockshelter deposits on Rapa, A. moluccana endocarps
make up a large proportion of the archaeobotanical material identified, especially in the case of
the Tangarutu excavation. Kirch et al. (1995) did not present data on the amount of endocarp
identified from the Tangatatau excavation, but suggest it is abundant. In the Tangatatau and
Tangarutu excavations, A. moluccana endocarps are represented throughout each deposit.
Endocarps of A. moluccana are only represented in the basal Zone E deposit of Ureia (Allen and
Wallace 2007).
In Remote Oceania, endocarps of A. moluccana have been identified from several
archaeological sites in the Hawaiian Islands (e.g. Māhā’ulepū, Kaua’i, Burney et al. 2001; Mauna
Kea, Hawai’i, Allen 1981), Mangareva (Weisler 1995; Conte and Kirch 2005), Henderson
Island (Weisler 1995), Pitcairn Island (Weisler 1996), and more recently, New Caledonia (Emilie
Dotte pers. comm.). Athens and Ward (1997) located A. moluccana endocarp and wood from
Maunawili (Core 1) at 86–94 cm below the swamp surface, and provided an interpolated age
from a charcoal date of 688–655 BP (Beta-5490). The pre-European contact distribution and
introduction of A. moluccana may have extended as far as Raoul Island in the Kermadec Group
(New Zealand), although this could represent a historic introduction (Sykes and West 1996).
Despite the abundance of preserved A. moluccana endocarps in the archaeological record
from Rapa, this tree was not listed in any early accounts of the island until Jacques Moerenhout
(1837) visited in 1834. Stokes (n.d.) noted the use of A. moluccana wood for construction
of canoes and nuts for lighting. From visits to the island between the 1920s and 1930s, both
Stokes (n.d.) and Fosberg (in Mueller-Dombois and Fosberg 1998:403 respectively) considered
A. moluccana to be a major component of moist forests on Rapa. Despite its importance only
50 years ago, few trees currently exist on the island. Some trees are located precariously on the
margins of coastal plains or in a few localities around the major settlements of Ha’urei and Area,
where the nuts provide fodder for pigs.
Fosberg (1991:18) has questioned whether the tree has been introduced to Remote Oceania,
as the nuts of this tree are commonly found as beach-drift throughout the Pacific:
The theory that it is an introduction is favoured and, indeed, suggested by the unlikelihood of its
having climbed the steep mountain slopes [of the Hawaiian Islands] up to its present habitat without
human assistance. If it was brought by humans and carried up the hills by them, it spread to dominate
a whole zone on many islands, forming pure stands and shading out almost all other plants, and made
an important change. On some islands kukui forest occupies roughly the rainfall belt (between 1500
and 2250 mm of rainfall a year), with tongues of this forest running some distance down moist ravines
to lower elevations. The landscape was modified conspicuously by the introduction of A. moluccana,
whether by the hand of humans or dispersed naturally.
terra australis 37
92
Matiu Prebble and Atholl Anderson
Lagenaria siceraria (Mol.) Standlicher (Cucurbitaceae)
The most recent analyses of the distribution and ethnobotany of L. siceraria in Remote Oceania
come from Decker-Walters et al. (2001), Smith (2005) and Clarke et al. (2006). Archaeological
remains suggest that the domestication of L. siceraria may have been independently developed
in the Americas by 7000 BC (Smith 2005), South East Asia between 8000 and 4000 BC,
and Africa between 3000 and 2000 BC (Heiser 1979). Bellwood (1997) has observed that L.
siceraria was present in Taiwan at the time of the Austronesian expansion out of this area about
3000 BC, and Decker-Walters et al. (2001) assessed the diversity of landraces of L. siceraria from
each of the above areas, including New Guinea, using random amplified polymorphic DNA.
They found that the landraces of New Guinea could be distinguished from American specimens.
Based on linguistic and limited archaeobotanical evidence, Green (2000, following Ross
1996) suggests that L. siceraria was absent from Near Oceania at the time the closest islands in
Remote Oceania were first settled around 1500 BC to 1000 BC. Green supports this claim by
noting the absence of L. siceraria in the rich archaeobotanical assemblages from the waterlogged
sites at Dongan in the lower Ramu, Papua New Guinea (Swadling et al. 1991). Green (2000)
also cites the earliest interpolated date in Near Oceania for L. siceraria at 1230–1960 BC (ANU43) from the Manton site at Warrawau, upper Wahgi Valley, Central Highlands, Papua New
Guinea (Golson et al. 1967). Golson (2002), however, has now revised the Manton gourd rind
determination, and has tentatively assigned it to the wax gourd Benincasa hispida on the basis of
recent finds of this rind in an archaeological site at Kana (dated to 1005 BC to 20 AD; ANU9487), at a slightly higher altitude to Manton, also in the Wahgi Valley.
Fragments of L. siceraria pericarp have been identified from a number of archaeological sites
in the Hawaiian Islands, including the Ewa Plain and Kawainui Valley (Allen-Wheeler 1981)
on O’ahu, as well as Maha’ulepu Caves, Kaua’i (Burney et al. 2001). Horrocks et al. (2000)
identified L. siceraria pollen in an exposed section of two archaeological stone garden mounds
at Pouerua, Northland, New Zealand. L. siceraria has an entomophilous flowering biology in
which pollen is only deposited in the direct vicinity of the parent plant. This effect would be
intensified given the low stature of L. siceraria. Horrocks et al. (2002) have identified L. siceraria
pollen from dog or human coprolites of around AD 1300–1600 (NZA-12591) buried in a
beach dune on Great Barrier Island in the Hauraki Gulf, New Zealand. These data suggest direct
consumption of flowers, but might also reflect consumption of residues on young L. siceraria, or
even inadvertent ingestion during hand pollination (Horrocks 2004).
The introduction of L. siceraria on Rapa contradicts Green’s (2005) argument that L.
siceraria was brought with I. batatas after European contact. The 16th century age for bottle
gourd on Rapa suggests it was introduced pre-contact. However, the lack of evidence for I.
batatas from any excavation on Rapa and the late appearance of L. siceraria in the Tangarutu E1
sequence imply that neither plant was introduced during initial Polynesian colonisation.
Further support for the Polynesian introduction of L. siceraria to Rapa comes from the
description of ‘dried pumpkin’ by Faddei Von Bellingshausen, although this might otherwise
describe Ipomoea batatas tubers (Barratt 1988). Ethnographic information collected by Stokes
(n.d.) and others on Rapa suggests that L. siceraria was one of the few actively cultivated plants
on Rapa in the early 1900s, other than Colocasia esculenta, Dioscorea spp. and Ipomoea batatas.
Stokes (n.d. Group 2 Box 7.1) found that ‘Seeds of the gourd (koali) are set in the ground
with little preparation, and the plants are left to look after themselves’. Fosberg and St. John
(1934) only identified L. siceraria cultivations at Akatanui Bay, a site now largely abandoned for
cultivation.
terra australis 37
The archaeobotany of Rapan rockshelter deposits
93
Conclusion
The Tangarutu assemblage is exceptional in both the abundance and preservation quality of
plant materials. Aleurites moluccana endocarps, Lagenaria siceraria pericarp (bottle-gourd) and
leaf fibres of Pandanus spp. and Freycinetia spp. are well preserved. Archaeological charcoals in
the basal part of this deposit are derived primarily from introduced economic taxa, including
cf. Hibiscus tiliaceus syn. Talipariti tiliaceum, cf. Thespesia populnea and cf. Cordyline fruticosa,
suggesting that the establishment of introduced plants was rapid and an important part of
island colonisation strategy. Radiocarbon ages obtained from fruit and seed remains excavated
from the Tangarutu sequence are discussed in reference to the settlement sequence of Rapa.
The earliest ages obtained from A. moluccana endocarp and Cordyline wood, produced from
introduced and cultivated trees, are slightly later than the Bayesian-modelled age estimate for
initial colonisation of the island.
References
Allen, M.S. 1981. An analysis of the Mauna Kea adze quarry archaeobotanical assemblage. Unpublished
MA thesis, University of Hawaii.
Allen, M.S. and Wallace, R. 2007. New evidence from the East Polynesian gateway: substantive and
methodological results from Aitutaki, Southern Cook Islands. Radiocarbon 49:1163–1179.
Allen-Wheeler, J. 1981. Archaeological excavations in Kawainui Marsh, island of O’ahu. Report
prepared for Department of Planning and Economic Development. Department of Anthropology,
B.P. Bishop Museum, Honolulu.
Anderson, A. 1991. The chronology of colonization in New Zealand. Antiquity 65:767–795.
Athens, J.S. 1997. Hawaiian native lowland vegetation in Prehistory. In: Kirch, P.V. and Hunt, T.L.
(eds), Historical ecology in the Pacific Islands: prehistoric environmental and landscape change. Yale
University Press, New Haven, pp. 248–270.
Athens, J.S. and Ward, J.V. 1997. The Maunawili core: prehistoric inland expansion of settlement and
agriculture, O’ahu, Hawai’i. Hawaiian Archaeology 6:37–51.
Barratt, G. 1988. Southern and Eastern Polynesia, Russia and the South Pacific, 1696–1840. University of
British Columbia Press, Vancouver.
Bellwood, P. 1978. Archaeological Research in the Cook Islands. Pacific Anthropological Records 27, B.P.
Bishop Museum, Honolulu.
Bellwood, P.S. 1997. Prehistory of the Indo-Malaysian Archipelago. University of Hawai’i Press,
Honolulu, Hawaii.
Burney, D.A., James, H.F., Burney, L.P., Olson, S.L., Kikuchi, W., Wagner, W.L., Burney, M.,
McCloskey, D., Kikuchi, D., Grady, F.V., Gage, R. II and Nishek, R. 2001. Fossil evidence for
a diverse biota from Kaua’i and its transformation since human arrival. Ecological Monographs
71:615–641.
Clarke, A.C., Burtenshaw, M.K., McLenachan, P.A., Erickson, D.L. and Penny, D. 2006.
Reconstructing the origins and dispersal of the Polynesian bottle gourd (Lagenaria siceraria).
Molecular Biology and Evolution 23:893–900.
Conte, E. and Kirch, P.V. 2005. Archaeological investigations in the Mangareva Islands (Gambier
Archipelago) French Polynesia. Archaeological Research Facility, University of California, Berkeley.
Decker-Walters, D., Staub, J., Lopez-Sese, A. and Nakata, E. 2001. Diversity in landraces and cultivars
terra australis 37
94
Matiu Prebble and Atholl Anderson
of bottle gourd (Lagenaria siceraria; Cucurbitaceae) as assessed by random amplified polymorphic
DNA. Genetic Resources and Crop Evolution 48:369–380.
Dickinson, W.R. 2001. Paleoshoreline record of relative Holocene sea levels on Pacific islands. EarthScience Reviews 55:191–234.
Ellison, J. 1994. Paleo-lake and swamp stratigraphic records of Holocene vegetation and sea-level
changes, Mangaia, Cook Islands. Pacific Science 48:1–15.
Florence, J. 2004. Flore de la Polynésie française Volume II. IRD Éditions, Paris.
Fosberg, F.R. 1991. Polynesian plant environments. In: Cox, P.A. and Banack, S.A. (eds), Islands,
plants, and Polynesians. Dioscorides Press, Portland, pp.11–23.
Fosberg, F.R. and St. John, H. 1934. Check list and field notebook of the plants of Southeastern
Polynesia: Society Islands, Tuamotus, Austral Islands, Rapa. B.P. Bishop Museum Herbarium
Archive, Honolulu.
Golson, J., Lampert, R.J., Wheeler, J.M. and Ambrose, W.R. 1967. A Note on Carbon Dates for
Horticulture in the New Guinea Highlands. Journal of the Polynesian Society. 76:369–371.
Golson, J. 2002. Gourds in New Guinea, Asia and the Pacific. In: Bedford, S., Sand, C. and Burley,
D. (eds), Fifty years in the field: essays in honour and celebration of Richard Shutler Jr’s archaeological
career. New Zealand Archaeological Association, Auckland, pp. 69–78.
Green, R.C. 2000. A range of disciplines support a dual origin for the bottle gourd in the Pacific.
Journal of the Polynesian Society 109:191–197.
Green, R.C. 2005. Sweet potato transfers in Polynesian prehistory. In: Ballard, C., Brown, P., Bourke,
R.M. and Harwood, T. (eds), The sweet potato in Oceania: a reappraisal. Oceania Monograph.
Pandanus Books, Canberra, pp. 43–62.
Heiser, C.B. 1979. The gourd book. University of Oklahoma Press, Norman, Oklahoma.
Horrocks, M. 2004. Polynesian plant subsistence in prehistoric New Zealand: a summary of the
microfossil evidence. New Zealand Journal of Botany 42:321–334.
Horrocks, M. and Weisler, M.I. 2006. Analysis of Plant Microfossils in Archaeological Deposits from
Two Remote Archipelagos: The Marshall Islands, Eastern Micronesia, and the Pitcairn Group,
Southeast Polynesia. Pacific Science 60:261–280.
Horrocks, M., Jones, M.D., Carter J.A. and Sutton, D.G. 2000. Pollen and phytoliths in stone mounds
at Pouerua, Northland, New Zealand: implications for the study of Polynesian farming. Antiquity
74: 863–872.
Horrocks, M., Jones, M.D., Beever, R.E. and Sutton, D.G. 2002. Analysis of plant microfossils in
prehistoric coprolites from Harataonga Bay, Great Barrier Island, New Zealand. Journal of the
Royal Society of New Zealand 32:617–628.
Kennett, D., Anderson, A., Prebble, M., Conte, E. and Southon, J. 2006. Prehistoric human impacts
on Rapa, French Polynesia. Antiquity 80:340–354.
Kirch, P.V., Steadman, D.W., Butler, V.L., Hather, J. and Weisler, M.I. 1995. Prehistory and human
ecology in Eastern Polynesia: Excavations at Tangatatau Rockshelter, Mangaia, Cook Islands.
Archaeology in Oceania 30:47–65.
McCormac, F.G., Hogg, A.G., Blackwell, P.G., Buck, C.E., Higham, T.F.G. and Reimer, P.J. 2004.
SHCal04 Southern Hemisphere Calibration 0–11.0 cal kyr BP. Radiocarbon 46:1087–1092.
Moerenhaut, J.A. 1837. Voyages aux iles du Grand Ocean. Paris.
Mueller-Dombois, D. and Fosberg, F.R. 1998. Vegetation of the tropical Pacific islands. Springer, New York.
Orliac, C. and Orliac, M. 1998. The disappearance of Easter Island’s forest: overexploitation or climatic
catastrophe? In: Stevenson, C.M., Lee, G. and Morin, F.J. (eds), Easter Island in Pacific context,
South Seas Symposium. Proceedings of the Fourth International Conference on Easter Island and East
Polynesia. Easter Island Foundation, Los Osos, pp. 129–134.
Prebble, M. In press. The palaeobotanical record of Rapa (French Polynesia): phytogeographic
implications for the Austral Archipelago. In: Meyer, J-Y. (ed), Biodiversity of the Austral Islands.
Prebble, M. and Wilmshurst, J. 2009. Detecting the initial impact of humans and introduced species on
terra australis 37
The archaeobotany of Rapan rockshelter deposits
95
island environments in Remote Oceania using palaeoecology. Biological Invasions 11:1529–1556.
Ross, M. 1996. Reconstructing food plant terms and associated terminologies in Proto Oceanic. In:
Lynch, J. and Fa’afo, P. (eds), Oceanic studies: proceedings of the First International Conference
on Oceanic Linguistics. Pacific Linguistics Series C no. 133. The Australian National University
Canberra, Australia, pp. 163–221.
Shineberg, D. 1986. Archibald Menzies’ account of the visit of the Discovery to Rapa and Tahiti, 22
December 1791–25 January 1792. Pacific Studies 9:59–102.
Smith, B.D. 2005. Reassessing Coxcatlan Cave and the early history of domesticated plants in
Mesoamerica. Proceedings of the National Academy of Science, USA 102:9438–9445.
St. John, H. 1976. Revision of the Genus Pandanus Stickman. Part 40. The Fijian species of the section
Pandanus. Pacific Science 30:249–315.
St. John, H. 1979. Revision of the Genus Pandanus Stickman. Part 42 Pandanus tectorius Parkins. ex Z
and Pandanus odoratissimus L.f. Pacific Science 33:395–401.
Stokes, J.F.G. n.d. Ethnology of Rapa. Bernice P. Bishop Museum Archives, Honolulu.
Stone, B.C. 1976. The Pandanaceae of the New Hebrides, with an essay in intraspecific variation in
Pandanus tectorius. Kew Bulletin 31:47–70.
Stone, B.C. 1988. Notes on the genus Pandanus (Pandanaceae) in Tahiti. Botanical Journal of the
Linnean Society 97:33–48.
Stuiver, M., Reimer, P.J. and Reimer, R.W. 2005. CALIB 5.0. [WWW program and documentation].
Swadling, P., Araho, N. and Ivuyo, B. 1991. Settlements associated with the inland Sepik-Ramu sea.
Bulletin of the Indo-Pacific Prehistory Association 2:92–112.
Sykes, W.R. and West, C.J. 1996. New records and other information on the vascular flora of the
Kermadec Islands. New Zealand Journal of Botany 34:447–462.
Tuggle, H.D. and Spriggs, M. 2001. The age of the Bellows Dune site O18, O’ahu, Hawai’i, and the
antiquity of Hawaiian colonization. Asian Perspectives 39:165–188.
Walczak, J. 2001. Le peuplement de la Polynésie orientale: Une tentative d’approche historique par les
exemples de Tahiti et de Rapa (Polynésie française). Unpublished PhD thesis.
Walczak, J. 2003. Présentation des données actuelles sur la préhistoire de Rapa Iti (archiple des
Australes-Polynésie Française). In: Orliac, C. (ed), Archéologie en océanie insulaire: peuplement,
sociétés et paysages. Éditions Artcom, Paris, pp. 28–45.
Weisler, M.I. 1995. Henderson Island prehistory: colonization and extinction on a remote Polynesian
island. Biological Journal of the Linnean Society 56:377–404.
Weisler, M.I. 1996. Taking the mystery out of the Polynesian ‘mystery islands’: a case study from
Mangareva and the Pitcairn Group. In: Davidson, J.M., Irwin, G., Leach, B.F., Pawley, A. and
Brown, D. (ed), Oceanic culture history: essays in honour of Roger Green. New Zealand Journal of
Archaeology Special Publication, Dunedin, pp. 615–629.
Whistler, W.A. 1991. Polynesian plant introductions. In: Cox, P.A. and Banack, S.A. (eds), Islands,
plants, and Polynesians. Dioscorides Press, Portland, pp. 41–66.
Wilmshurst, J.M., Hunt, T.L, Lipo, C.P. and Anderson, A.J. 2011. High-precision radiocarbon dating
shows recent and rapid initial human colonization of East Polynesia. Proceedings of the National
Academy of Science, USA 108:1815–1820.
terra australis 37
terra australis 37
5
Cordage from Rapan archaeological sites
Judith Cameron
Department of Archaeology and Natural History, The Australian National University, Canberra, Australia,
judith.cameron@anu.edu.au
Introduction
More than 80 per cent of the items of material culture produced by traditional Polynesian
groups were made from plant fibres (Kirch and Green 2001:164–165), yet very little is known
about prehistoric fibre artefacts in French Polynesia. The first Europeans to discover Rapa in
1791 (Chapter 2) exchanged iron for Rapan artefacts, some of fibre. Among them was a fishing
line from the Vancouver Collection of the British Museum (Chapter 2). During his fieldwork
on Rapa (1920–21), Stokes (n.d.) found several fragments of archaeological cordage on Rapa,
but they were all surface finds of unknown age. An assemblage of securely provenanced fibre
artefacts was recovered during the 2002 excavations on Rapa. Here, the species of plants used
are identified, functional attributes of the specimens in the assemblage are described, function is
inferred, and the artefacts are compared with examples from other parts of Polynesia.
The assemblage
Although dampness and humidity destroy archaeological textiles (a generic term for cordage,
matting, basketry, textiles), the Rapan fragments were well-preserved for two reasons: relatively
cooler and drier conditions than occur in tropical Polynesian islands, and stratigraphic contexts
in dry sand, notably at Tangarutu. The assemblage consists of 19 fragments of worked fibres
(Figure 5.1), 17 from Tangarutu and two from Akatanui. The functional attributes of the
fragments (Table 5.1) include form, material composition, number and width of individual
elements, length and width of grouped elements, numbers and angles of crossing. The number
of crossings per 100 mm was documented as a measure of tightness and strength.
Form of cordage
The simplest cordage has two strands of the same material composition, diameter and twist.
Elaborations are generated by increasing the number of strands, varying the direction of
the twists (S and Z), employing elements of unequal diameter or altering the construction
process by including knots, loops, or other elements (Hurley 1979). The cordage in the Rapan
assemblage is of three different types: twisted, braided and knotted. Both twisting and braiding
are techniques that increase the tensile strength and durability of fibres, whereas knotting
terra australis 37
98
Judith Cameron
Figure 5.1. Assemblage of fibre artefacts from Tangarutu and Akatanui. See Table 5.1 for information about each item.
Table 5.1. Functional attributes of fibre artefacts from Tangarutu (RT) and Akatanui (RA).
Number Location
Form
Material
composition
Length
(mm)
Width
(mm)
Width of single
element (mm)
RT 1
E1 Spit 1
twisted cordage
Hibiscus sp.
RT 2
E1 Spit 2
braided cordage
RT 3
E1 Spit 2
RT 4
Number of Number of Crossings
Angle
elements crossings per 100 mm of crossing
50
6
1.2
4
n/a
n/a
50
Freycinetia sp.
230
4
1.2
3
n/a
n/a
60
braided cordage
Freycinetia sp.
110
4
1.1
3
6
5
60
E1 Spit 2
braided cordage
Freycinetia sp.
95
3
1.0
3
11
12
65
RT 5
E1 Spit 2
braided cordage
Freycinetia sp.
140
3
1.0
3
15
11
60
RT 6
E1 Spit 2
braided cordage
Freycinetia sp.
41
2
6.0
1
0.5
n/a
RT 7
E2 Spit 3
braided cordage
Freycinetia sp.
170
8
8.0
3
?
n/a
n/a
RT 8
E2 Spit 3
knotted fibre
B. papyrifera
70
19
19.0
1
2
3
n/a
RT 9
E2 Spit 3
braided cordage
Freycinetia sp.
25
4
1.1
3
3
12
65
RT11
E1 Spit 2
twisted cordage
Hibiscus sp.
44
9
4.0
3
2
4
50
RT 12
E1 Spit 1
knotted cordage
B. papyrifera
115
4
1.1
3
1
n/a
n/a
RT 13
E1 Spit 1
knotted cordage
B. papyrifera
14
10
4.5
2
1
n/a
60
RT14
E2 Spit 2
braided cordage
Freycinetia sp.
115
4
1.2
3
7
6
65
RA 15
C1 Spit 3
knotted fibre
B. papyrifera
35
20
6.0
?
n/a
n/a
n/a
RA 16
C1 Spit 2
knotted fibre
B. papyrifera
48
5
5.0
1
2
4
n/a
RT 17
E1 Spit 3
knotted fibre
B. papyrifera
35
20
4
12
n/a
RT 18
E1 Spit 3
knotted fibre
B. papyrifera
95
9
4.0
2
3
3
n/a
RT 19
E2 Spit 1
braided cordage
Freycinetia sp.
110
3
1.2
3
5
5
65
RT 20
E2 Spit 1
braided cordage
Freycinetia sp.
85
4
1.2
3
3
4
65
*Note: RT10 is not included. Microscopic analysis revealed it was actually an unworked piece of gourd. RT 19 and RT 20 are not shown in Figure 5.1.
terra australis 37
Cordage from Rapan archaeological sites
99
generally increases length. All specimens are small and fragmentary with no use-wear patterns,
dyes, or stains discernible.
The two twisted fragments (RT1 and RT11) consist of more than two strands of hard,
outer fibres. Although RT11 is defined as twisted cordage, strictly speaking it is simply wrapped.
The core element of RT11 is rigid, with untwisted fibre wrapped around the core, a twig. The
single elements of this three-stranded cordage measured 4 mm in diameter. The cord is not
wrapped at right angles, but in a helix spiralling from the lower left to the upper right, at regular
intervals. The cord would first have been attached to the twig, then wrapped continuously in
an oblique way until reaching the end of the twig, then reversed and wrapped in the opposite
direction, crisscrossing and overlapping the first layer cord six times, with greater density at one
end. The function of RT11 is not indicated by context, but this technique was often used in
west Polynesian coiled basketry. Coiling is a basketry technique where active, flexible elements
are wound around passive, stationary elements called the foundation. Successive circuits of the
foundation are bound together by wrapping. The same technique was also used to make other
artefacts such as the cordage used in the fishing device from the Vancouver Collection (Chapter
2) and some small fish hooks from Rapa that may also be from the Vancouver Collection (Figure
5.2), although that is not certain.
RT1 is the end piece of a fibre artefact showing an elaboration of the twisting technique
evidenced in RT11. RT1 was worked using four individual strands, each single element
measuring 1.2 mm in diameter. As with RT11, a stick forms the base, but instead of using a
single unspun, untwisted element for spiral wrapping, a number of semi-rigid fibre strands are
wrapped around the stick over the end, with four strands used to bind the cordage.
Ten of the fragments (RT2, RT3, RT4, RT5, RT6, RT7, RT9, RT14, RT19, RT20) in the
assemblage were flat, constructed by braiding. In the early contact literature, the term sennit or
sinnet was often used for braided coconut cordage and warp-woven tapes. Braiding is a process
derived from diagonal plaiting which produces cordage (and textiles) that is invariably greater
in length than width. Braiding can be either two-dimensional with odd or even sets of strands,
or three-dimensional forms (tubular and compact), including looped, interlooped, interlinked
and knotted structures (Emery 1980:60).
None of the individual elements of the braided cordage in the assemblage had been previously
twisted or spun. The commencement point of cordage is generally culturally determined, but this
feature could not be ascertained, as the archaeological pieces were fragmentary. The absence of
wooden stick fixtures at the site suggests that the braided cordage was probably worked with loose
ends. All fragments were worked from flexible fibres and shared the same basic configuration.
Six of the fragments were constructed using three strands, called three-ply braiding; one is fourply. The three-ply braids were plaited by passing the elements in the right-hand strand over those
in the middle strand, then by reversing the process. During construction, the terminal elements
were folded back on themselves at angles ranging from 60° to 65° and replaited into the body
of each band. The angle of crossing shows a high degree of standardisation. The number of
crossings per 100 mm indicates the tightness of the cordage: the greater the number of twists,
the stronger the cordage. Most of the braided cordage in the assemblage was tightly worked,
with measurements ranging from >10 crossings per 100 mm, with a few examples of loosely
plaited forms of <5 per 100 mm. Some fragments exhibit 2/1/2 and 2/4/2 shifts.
Seven fragments (RT8, RT12, RT13, RA15, RA16, RT17 and RT18) in the assemblage
were knotted. The knot in RT17 is a basic overhand knot, formed by pulling one end through a
simple loop. As the fragment is incomplete, its function is unclear. The longest knotted fragments
(RT16, RT18), measuring 4.4 cm in length, were found at Akatanui (Table 5.1). Other types
terra australis 37
100 Judith Cameron
Figure 5.2. Fishing hooks collected
from Rapa Island. In the British Museum
(Beasley 1928:Plate LXIII; 037 BM). It is not
clear whether these were in the Vancouver
Collection. All appear to have wood or
candlenut shanks, fully bound, but the
example at lower left seems to have a shell
or bone tip inserted into the shank.
of knotted cordage in the assemblage are more complex and give better insights into function.
RT16 is an example of a slightly more complicated knot, the sheet bend, which is still widely
used for fishing nets throughout Polynesia.
Material composition
Identifications by M. Prebble (pers. comm.) indicate that the knotted cordage was made from
Broussonetia papyrifera (paper mulberry), the braided cordage appears to have been worked from
the roots of Freycinetia spp. (kiekie vine) and the twisted cordage was worked from Hibiscus spp.
(purau or hau). While Cocos nucifera (coconut) fibre is not in the archaeological assemblage,
Stokes (n.d.) identified coconut fibre cordage, but this might have been from material of 19th
century origin, or identified wrongly. Altogether, 37 per cent of the cordage in the Rapan
assemblage was made from Broussonetia papyrifera, a fibre introduced by early Polynesians
primarily for tapa. The bast would have been removed from the outer bark and dried in the sun
before being cut into workable strips. All the B. papyrifera fibres were knotted, with individual
strips ranging from 1.1 mm to 20 mm in diameter. The largest was made from a single strip of
terra australis 37
Cordage from Rapan archaeological sites
101
the plant; most were multiple strips of one to three strands. In the Rapan assemblage, 10 per
cent of the cordage was made from Hibiscus spp. In the Society Islands, hau is the term for the
bast and more for the cordage from hibiscus. The beach hibiscus is pantropic in distribution,
deposited originally by seawater-dispersed seeds (Whistler 1991:65). More than half, 52 per
cent, of the cordage in the Rapan assemblage appears to be made of roots from Freycinetia spp.
(kiekie). All worked examples of this fibre were braided.
Archaeological comparisons
All of the cordage types in the Rapan assemblage have been recovered also in surface deposits
in caves or from stratified sequences in Hawaii (Kirch 1979; Burney and Kikuchi 2006) and
New Zealand (Goulding 1971; Lawrence 1989; Anderson et al. 1991; McAra 2004). The most
common cordage in the Pacific archaeological record is aha (Cocos nucifera). It comprises 49% of
the cordage in the B.P. Bishop Museum’s archaeological collections (Summers 1990). Of the 98
examples of twisted aha cordage in the Bishop Museum collection, 87 are two-ply of remarkably
uniform diameter (0.8–1.5 mm). Throughout Hawaii, the primary function of twisted aha
cordage was for nets (Summers 1990).
Braid is the dominant cordage type in the Bishop Museum’s archaeological collection.
On Rapa and other Polynesian islands, men were responsible for braiding to produce cordage,
whereas women were responsible for twisting cordage and for making bark cloth. Most
Hawaiian braid was untwisted and three-ply, but five-ply and seven-ply braided cords (B.P.
Bishop Museum:4755, 4756, 4749, 6870) were used to lash canoe hulls to outrigger booms
(‘iako). The width of the cordage in the Bishop Museum ranged from 4.8 mm to 5.4 mm,
whereas the three-ply cordage from Rapa measured between 3.0 mm and 4.0 mm. Four of the
seven hafted adzes in Cook expedition collections are lashed with braided three-ply aha cordage
(Kaeppler in Summers 1990:88).
On Rapa, Stokes (n.d.:294) found several pieces of nape kiekie (Freycinetia spp.), with a
piece of coir cordage attached to canoe boards in a burial cave at Akatanui. Stokes observed
that ka’a, the original Rapan name for this type of cordage, was slowly being replaced by nape,
the Tahitian term for braid. According to Stokes, nape from kiekie was said to be fairly durable
but did not last as long as coir cordage. Braid was also a component of the close-meshed kiekie
fishing baskets (inaki) woven on Rapa using the twining technique (Plate 8442). Kiekie baskets
from Rapa were in demand in Tahiti and Mangareva. Stokes (n.d.) also recorded wide-meshed
war nets, woven, probably, from kiekie.
Precisely the same technique was used to work a fragment of tightly braided three-ply uki uki
(Dianella sp.) cordage recovered from a burial site (designated 1985.669.013) in Kohala, Hawaii
(Summers 1990). Stokes (n.d.:Plates 8550, 8551) found decayed woven fragments (Plates 8550
and 8551), identified as mortuary cloaks, in some Rapan rockshelters. His descriptions of cloaks
indicate that warps were made of rigid fibres from the giant rush, Scirpus spp., placed in parallel
rows manipulated by the weaver’s feet. The soft, pliable wefts, made from twisted bast fibres,
such as more from Hibiscus tiliaceus, or raupo from Typhus spp., were worked at right angles. The
wefts created a soft pile regarded by Rapans as good for cold weather. (Stokes n.d.). The closedtwining technique was also used for ti (Cordyline spp.) kilts. While twined fish baskets were
widespread throughout Polynesia (Burrows 1970:26), twined kilts did not occur in western
Polynesia, being confined to the Tuamotus and Rapa. Short lengths of braided Freycinetia spp.
cordage were attached to cloaks and mortuary robes for fastening. Stokes compared Rapan
cloaks with New Zealand equivalents and found Rapan forms to be similar, but less refined.
Cloaks held considerable social and ritual significance throughout Polynesia and Ellis’ informant,
terra australis 37
102 Judith Cameron
Teraau, explained the colour symbolism of Rapan cloaks. Royal cloaks incorporated dark blue
feathers from kotokoto birds (species unknown, but possibly the Tuamotuan kingfisher, Halcyon
gambieri (SOPF 1999:41), the feathers of which were probably imported in the European era),
whereas those intended for commoners were made from black and white petrel feathers.
New Zealand excavations have also produced fragments of mortuary cloaks and cordage.
Lawrence (1989) recovered cordage during excavations of a historical site in the Waitakere
Ranges, and Lander (1992) similar material from Raupa. Anderson et al. (1991) found numerous
fragments of harakeke (Phormium tenax) cordage during excavations at Fiordland, and also a
piece of a woven harakeke and bird-feather cloak. Harakeke netting and cordage fragments were
also reported from waterlogged contexts at the late Maori village of Kohika (McAra in Irwin
2004) and from a cave on the Banks Peninsula (McAra in Irwin 2004).
Conclusions
The main sources of cordage on Rapa were the fibrous roots and vines of native kiekie and
the introduced paper mulberry, with a small proportion of samples from hibiscus. While
some coconut-fibre artefacts might have been collected by Stokes, no coconut fibre occurs in
the excavated samples, and this is consistent with the probable absence of coconut until the
historical era.
Given the multiple uses of cordage, it is difficult to determine the function of examples in
the Rapan assemblage. However, as most of the material came from Tangarutu cave, which had
functioned very largely as a fishing station, it is most likely that the fibre artefacts were associated
with that activity. The cord-wrapped stick from Tangarutu could either be the handle of a fishing
basket or part of a fishing line, with the softer bast elements wrapped around the rigid elements
to protect the user’s hands. The knotted aute cordage compares favourably with archaeological
aute from Hawaii identified as carrying bags and nets. The absence of twisting suggests the
Rapan samples are probably the remains of nets. The braided cordage in the assemblage seems
most likely to have been the remains of a longer length of cordage used for fishing lines or nets.
terra australis 37
Cordage from Rapan archaeological sites
103
References
Anderson, A., Goulding, J. and White, M. 1991. Bark and fibre artefacts. In: Anderson, A.J. and
McGovern-Wilson, R. (eds), Beech Forest Hunters: the archaeology of Maori rockshelter sites on
Lee Island, Lake Te Anau, in southern New Zealand, pp. 43–55. New Zealand Archaeological
Association, Auckland.
Beasley, H.G. 1928. Pacific Island Records: Fish Hooks. Seeley Service, London.
Burney, D. and Kikuchi, W.P. 2006. A Millennium of human activity at Makauwahi Cave,
Maha’ulepu, Kaua’i. Human Ecology 34:219–247.
Burrows, E.C. 1970. Western Polynesia: a study of cultural differentiation. University Book Shop, Dunedin.
Emery, I. 1980. The Primary Structures of Fabrics. The Textile Museum, Washington.
Goulding, J. 1971. Identification of archaeological and ethnographical specimens of fibre-plant
material used by the Maori. Records of the Auckland Institute and Museum 8:57–101.
Hurley, W.M. 1979. Prehistoric Cordage: identification of impressions on pottery. Taraxacum, Washington.
Irwin, G. 2004. Kohika: The Archaeology of a late Maori lake village in the Ngati Awa rohe, Bay of Plenty,
New Zealand. Auckland University Press, Auckland.
Kirch, P. 1979. Marine Exploitation in Prehistoric Hawaii. Pacific Anthropological Record 29. Bernice
P. Bishop Museum, Honolulu, Hawaii.
Kirch, P. and Green, R. 2001. Hawaiki: Ancestral Polynesia: An Essay in Historical Anthropology.
Cambridge University Press, Cambridge.
Lander, M. 1992. Fibre fragments from the Raupa site, Hauraki Plains. Records of the Auckland Institute
and Museum 29:7–23.
Lawrence, J. 1989. The Archaeology of the Waikakere Ranges. Unpublished MA thesis, University of
Auckland.
McAra, S. 2004. Kohika fibrework. In: Irwin, G. (ed), Kohika: The Archaeology of a late Maori lake
village in the Ngati Awa rohe, Bay of Plenty, New Zealand, pp. 149–159. Auckland University Press,
Auckland.
Stokes, J.F. n.d. Ethnography of Rapa. Manuscript in the Bernice P. Bishop Museum Collection,
Honolulu, Hawaii.
Summers, C. 1990. Hawaiian Cordage. Bernice P. Bishop Museum, Honolulu, Hawaii.
Whistler, W.A. 1991. Polynesian plant introductions. In: Cox, P.A. and Banack, S.A. (eds), Islands,
Plants and Polynesians: an introduction to Polynesian ethnobotany, pp. 41–66. Dioscorides Press,
Portland, Oregon.
terra australis 37
terra australis 37
6
Bird, reptile and mammal remains from
archaeological sites on Rapa Island
Alan J. D. Tennyson
Museum of New Zealand Te Papa Tongarewa, Wellington, New Zealand, AlanT@tepapa.govt.nz
Atholl Anderson
Department of Archaeology and Natural History, The Australian National University
Introduction
Rapa Island, French Polynesia, has a land vertebrate fauna typical of an isolated oceanic South
Pacific Island. Birds dominate the fauna and there are no native mammals or terrestrial reptiles.
Together with its smaller offshore islands, Rapa has a small, poorly known but important bird
fauna. Bird species recorded breeding or as migrant residents are: Christmas shearwater Puffinus
nativitatis, little shearwater Puffinus assimilis myrtae, Kermadec petrel Pterodroma neglecta
neglecta, Murphy’s petrel Pterodroma ultima, black-winged petrel Pterodroma nigripennis, whitebellied storm petrel Fregetta grallaria titan, Polynesian storm petrel Nesofregetta fuliginosa,
red-tailed tropicbird Phaethon rubricauda, grey duck Anas superciliosa, spotless crake Porzana
tabuensis, golden plover Pluvialis fulva, wandering tattler Tringa incana, brown noddy Anous
stolidus, grey noddy Procelsterna cerulea, white tern Gygis alba, Rapa fruit dove Ptilinopus huttoni
and long-tailed cuckoo Eudynamys taitensis (Pratt et al. 1987; Thibault and Varney 1991b). The
fruit dove and the subspecies of little shearwater and white-bellied storm petrel are endemic to
the island group.
The history of introduced mammals on Rapa is sketchy but Pacific rats (Rattus exulans)
arrived in prehistoric times and, apparently, are still the only rat species present today (Thibault
and Varney 1991a, 1991b). At times, rats have been ‘very numerous’ (Hall 1869). There is no
historical evidence suggesting the pre-European introduction of dogs (Canis familiaris), pigs
(Sus scrofa) or chickens (Gallus gallus). These were introduced in the 19th century, along with
sheep (Ovis aries), which ‘did not seem to thrive’ (Hall 1869:80), cattle (Bos taurus) and goats
(Capra hircus). The latter are now in large numbers and the vegetation and integrity of upland
soils is suffering considerably as a result. Cats (Felis catus) were introduced at an unknown date
but were established by 1921 (Holyoak and Thibault 1984).
Vertebrate remains recovered by archaeological excavations in the coastal sites were mainly
terra australis 37
106 Alan J. D. Tennyson and Atholl Anderson
of fish, but bird and rat bones were also quite common. Here, we report the identification and
quantification of bird, reptile and mammal bone.
Methods
Comparative collections in the Museum of New Zealand Te Papa Tongarewa (NMNZ) were
used by Tennyson to identify faunal remains. Specimens of closely related taxa were used when
comparative skeletons of species known from Rapa were not available. Numbers of individual
birds given represent only the minimum number of individuals based on the fragmentary
remains collected. When calculating minimum numbers of individuals, identifications to a
probable species were lumped with certain identifications of that species.
List of main comparative skeletons used (NMNZ unless otherwise noted)
Pink-footed shearwater Puffinus creatopus OR.27754; flesh-footed shearwater Puffinus carneipes
S.771, OR.26448; wedge-tailed shearwater Puffinus pacificus OR.16209, OR.27270, OR.27271;
Christmas shearwater OR.24682; little shearwater OR.23972, OR.24230; Audubon’s shearwater
Puffinus lherminieri OR.27467, OR.27551; black-winged petrel OR.24226; Kermadec petrel
OR.11423, OR.25216; Murphy’s petrel OR.24403; white-bellied storm petrel OR.16071,
OR.27781; Polynesian storm petrel OR.27482, OR.27504; white-tailed tropicbird OR.27835;
red-tailed tropicbird OR.16056; red-footed booby Sula sula OR.24594; brown booby Sula
leucogaster OR.16058; domestic chicken Gallus gallus OR.16485; banded rail Gallirallus
philippensis OR.22842; spotless crake OR.26455, OR.27985a; golden plover OR.22509;
wandering tattler OR.26437; sooty tern Sterna fuscata OR.22945; brown noddy OR.25348;
black noddy Anous minutus OR.23958; grey noddy OR.24377; white tern OR.23699; Cook
Islands fruit dove Ptilinopus rarotongensis OR.22463; extinct Fijian pigeon Ducula lakeba
S.39286; collared lory Phigys solitarius Fiji Museum 22; Horned parakeet Cyanoramphus cornutus
OR.25730; long-tailed cuckoo OR.18339; hawksbill turtle Eretmochelys imbricata RE.5307;
Norway rat Rattus norvegicus LM.1274; dog LM.1412; cat LM.807, LM.808, LM.1413,
LM.1414; pig LM.1410, Canterbury Museum FMa 4086; goat LM.539, ex D.S.I.R. collection
no.4494.
Identification of taxa
Spits within excavations are shown in brackets for identified taxa. S-numbers are NMNZ
accession codes. All rat remains were assumed to be Pacific rat, based on their small size.
Abbreviations. cf. = similar to and probably, juv = juvenile, prox = proximal, dist = distal, L = left,
R = right, pt = part, frag = fragment, cra = cranium, pmx = premaxilla, mand = mandible, quad
= quadrate, vert = vertebra, cor = coracoid, scap = scapula, ster = sternum, hum = humerus, uln
= ulna, rad = radius, cmc = carpometacarpus, alphal = alar phalange, pel = pelvis, fem = femur,
tbt = tibiotarsus, fib = fibula, tmt = tarsometatarsus, phal = phalanx.
Tangarutu Cave
Location T1
Location T1 in loose sand at base of main exposure directly below eroding pocket of bones
exposed at 155–160 cm depth. The loose bones were:
Kermadec petrel: 4 L 1 juv R 1 R hum, 3 R uln, 1 L rad, 2 R cmc; S.44406.
red-tailed tropicbird: 1 ster; S.44407.
Unidentified remains: S.44408.
terra australis 37
Bird, reptile and mammal remains from archaeological sites on Rapa Island
107
Additional material excavated from the face of T1 at 155–160 cm depth:
flesh-footed/pink-footed shearwater: 1 skull, 1 mand, 1 L 1 R hum, 1 L 1 R uln, 1 rad, 1 pel, 1
R fem, 1 R tbt, 1 toe; S.44391.
little shearwater: 1 dist R hum, 1 R uln, 1 R rad; S.44403.
Kermadec petrel: 1 frag mand, 1 furcula, 2 R cor, 7 dist L 6 dist R hum, 9 L 5 R uln, 7 L 7 R
rad, 6 L 6 R cmc, 6 L 3 R manus, 4 alphal; S.44404.
Unidentified remains: S.44405, S.44409.
Excavation E1
little shearwater: 1 shaft uln (1), 1 prox rad (1), 1 R cmc (14); S.44416, S.44451.
black-winged petrel: 1 pt cra (1), 1 mand (10), 2 R scap (14), 1 pt L 1 pt R 1 R cor (16), 2 pt
L 1 R hum (17), 1 L uln (18), 1 dist rad (19), 1 R cmc 2 (19), manus (19), 1 pel (16), 1 dist
L 2 L 1 R fem (16), 1 prox L 1 dist L 1 shaft L 1 prox R 1 dist R 1 R tbt (18), 2 R tmt (21);
S.44417, S.44434–35, S.44450, S.44459, S.44467, S.44475–76, S.44487, S.44489, S.44541,
S.44544, S.44553, S.44558.
Kermadec petrel: 3 pt mand (1), 2 R quad (7), 2 pt furcula (10), 2 L scap (13), 1 L 3 pt R cor
(14), 1 R scap (14), 1 frag ster (14), 1 prox L 1 L 2 prox R 1 frag R 2 R hum (16), 3 dist L 2 L 2
dist R 1 R uln (17), 2 dist R 1 prox R rad (18), 2 dist L 2 prox R 1 dist R 1 R cmc (18), 1 manus
(19), 3 alphal (14), 1 pt pel (14), 1 R fem (15), 2 prox L 2 dist L 4 dist R tbt (16), 1 fib (17), 3
pt L 2 R tmt (16), 5 toes (16); S.44412, 44429, S.44433, S.44445, S.44449, S.44458, S.44466,
S.44474, S.44484, S.44486, S.44491, S.44536, S.44539, S.44543, S.44547, S.44552.
red-tailed tropicbird: 1 pt mand (1), 1 prox L hum (14), 1 dist L hum 1 (14), shaft cmc (21);
S.44413, S.44452, S.44492, S.44537.
spotless crake: 1 dist L tbt (11); S.44596.
brown noddy: 1 R scap (3), 1 R rad (3), 1 dist R cmc (17), 2 dist R tbt (18), 1 dist R tmt (16);
S.44422, S.44468, S.44477, S.44545.
grey noddy: 1 cor (13); S.44446.
white tern: 1 pt pmx (11), 2 pt mand (15), 2 L cor (17), 1 L scap (18), 2 dist L hum (17), 1 L
cmc (17), 1 manus (17); S.44439, S.44456, S.44469, S.44478, S.44548, S.44556.
Rapa fruit dove: 1 pt L cor (13), 1 R cmc (18); S.44447, S.44480.
pigeon (cf. Ducula): 1 prox L fem (14), 1 toe (16); S.44453, S.44461.
parrot (cf. Cyanoramphus): 1 dist L tbt (18); S.44481.
Unidentified bird feathers: S.44418, S.44420, S.44423, S.44426–27.
rat: 1 cra, 9 mand, 2 teeth, 5 vert, 2 hum, 1 pt pel, 2 fem; S.44436, S.44438, S.44441, S.44443,
S.44460, S.44464, S.44470, S.44482, S.44528, S.44534, S.44538, S.44549.
cf. Goat: 1 pel; S.44415.
Unidentified mammal fur: S.44424.
turtle: 1 ?uln, 1 piece of plastron; S.44414, S.44483.
Unidentified remains: S.44419, S.44421, S.44425, S.44428, S.44430–31, S.44437, S.44440,
S.44442, S.44444, S.44448, S.44454–55, S.44457, S.44462–63, S.44465, S.44471–73,
S.44479, S.44485, S.44488, S.44490, S.44493, S.44529–33, S.44535, S.44540, S.44542,
S.44546, S.44550–51, S.44554–55, S.44557, S.44559.
Excavation E2
black-winged petrel: 1 dist R hum (10–13); S.44515.
Kermadec petrel: 1 pt L scap (4), 1 shaft L 1 L 1 R cor (8–9), 1 prox L 1 pt R hum (10–13),
1 dist R uln (14), 1 pt pel (16–17), 1 dist L tmt (23–25); S.44504, S.44510–11, S.44514,
terra australis 37
108 Alan J. D. Tennyson and Atholl Anderson
S.44518, S.44520–21, S.44564.
1 red-tailed tropicbird: 1 vert (9), 1 alphal (22), 1 pt pel (23–25); S.44512, S.44562, S.44565.
parrot (cf. Cyanoramphus): 1 prox R uln (10–13), 1 R rad (14); S.44516, S.44560.
Unidentified bird toes: S.44500, S.44526.
Unidentified feathers: S.44392–93, S.44494, S.44497, S.44501.
rat: 2 mand (1), 1 tooth (1), 2 pt pel (18), 1 tibia (14); S.44495, S.44498, S.44523, S.44561.
Unidentified remains: S.44496, S.44499, S.44502–03, S.44505–09, S.44513, S.44517,
S.44519, S.44522, S.44524–25, S.44563, S.44566.
Excavation R1
little shearwater: 1 L fem (4), 1 prox R fem (7); S.44574, S.44578.
black-winged petrel: 1 L cor (3), 1 dist L hum (4), 1 dist R hum (8); S.44571, S.44573, S.44580.
Kermadec petrel: 1 quad (1), 2 L 1 R cor (3), 1 L scap (4), 1 dist L 1 frag R hum (6), 1 dist
rad (7), 1 pel (8), 1 dist R fem (8), 1 toe (8); S.44567, S.44570, S.44572, S.44575, S.44577,
S.44579, S.44582.
Unidentified remains: S.44410–11, S.44568, S.44569, S.44576, S.44581.
Excavation S1
little shearwater: 1 ster (3); S.44362.
black-winged petrel: 1 L scap, 1 ster, 1 pel (all 3); S.44396.
Kermadec petrel: 3 frag cra, 1 L scap, 3 ster, 1 dist L uln, 1 dist L rad, 3 L cmc, 1 prox L tmt
toe (all 3); S.44395.
white-bellied storm petrel: 1 L cor (3); S.44397.
red-tailed tropicbird: 1 L cor (3); S.44398.
rail (cf. Gallirallus): 1 L tmt (3); S.44399.
pigeon: 1 frag ster (3); S.44400.
rat: 1 dist fem (3); S.44401.
Unidentified remains: S.44402.
Akatanui shelters
Shelter 1
white-bellied storm petrel: 1 dist hum (2); S.43679.
?chicken: 2 toes (2); S.43680.
rat: 2 teeth, 1 pel, 2 tibia (1); S.43676–77, S.43685.
cat: 1 pt maxilla (1); S.43674.
Unidentified mammal: 1 rib, 1 phal; S.43682.
Unidentified remains: S.43675, S.43678, S.43681, S.43683–84, S.43686.
Shelter 3 Square A1
rat: 1 tooth (1), 1 scap (1), 2 hum (1), 2 pel (2), 3 fem (2), 1 tibia (2); S.43688, S.43690,
S.43692, S.43694.
Unidentified bird feathers: S.43687, S.43691.
Unidentified remains: S.43689, S.43693, S.43695.
Shelter 3 Square B1
grey noddy: 1 dist hum (1); S.43696.
Unidentified bird feathers: S.43697, S.43703.
terra australis 37
Bird, reptile and mammal remains from archaeological sites on Rapa Island
109
rat: 3 mand (1), 1 tooth (1), 1 vert (1), 1 pel (2), 2 fem (2), 1 tibia (2); S.43699, S.43701,
S.43705, S.43708.
Unidentified mammal: 1 juv vert, fur; S.43704, S.43706.
Unidentified remains: S.43698, S.43700, S.43702, S.43707.
Shelter 3 Square C1
white-bellied storm petrel: 1 prox tmt (2); S.43713.
grey noddy: 1 prox L hum (1); S.43709.
rat: 1 skull (1), 1 mand (1), 3 scap (2), 1 hum (2), 2 fem (2), 3 tibia (3); S.43710, S.43715,
S.43719–20.
?pig: 1 tooth (1); S.43712.
Unidentified mammal: 1 juv hum, fur; S.43714, S.43716.
Unidentified remains: S.43711, S.43717–18, S.43721.
Shelter 3, Test-pit E1
little shearwater: 1 R fem (1); S.43723.
Kermadec petrel: 1 dist R cmc (2); S.43728.
red-tailed tropicbird: 1 shaft R uln (1); S.43724.
rat: 2 mand, 1 tooth, 1 prox fem, 1 dist tibia (all 1); S.43722, S.43726.
cf. goat: 2 vert, 2 rib (both 1); S.43725.
Unidentified remains: S.43727, S.43729.
Angairao Bay shelters
Shelter C
little shearwater: 1 pt cra, 1 dist L rad (both 2); S.44366, S.44368.
black-winged petrel: 1 dist L 1 dist R tbt (3); S.44372.
white-bellied storm petrel: 1 dist tbt (2); S.44367.
rat: 2 pt cra (2), 2 mand (2), 3 teeth (2), 4 vert (2), 1 hum (2), 1 pel (2), 3 fem (3), 4 tibia (3);
S.44364, S.44370, S.44373.
cf. goat: 1 vert (2); S.44375.
Unidentified remains: S.44365, S.44369, S.44371, S.44374.
Shelter E
rat: 1 vert (9); S.44386.
cf. goat (juv) (4); S.44376, S.44379.
turtle: 1 piece of plastron (4); S.44380.
Unidentified remains: S.44377–78, S.44381–85.
Noogoriki
little shearwater: 1 cmc (1); S.44388.
Unidentified bird feather: S.44389.
Unidentified remains: S.44387, S.44390.
Taugatu Shelter
Unidentified remains: S.44527.
terra australis 37
110 Alan J. D. Tennyson and Atholl Anderson
Discussion and conclusions
It is not certain that all of the bird bones recovered from the Rapa sites had been deposited
as midden rather than by natural accumulation. No bones showed clear signs of charring or
human butchery and elements from all skeletal components were preserved. However, the clear
cultural context of the sites and the presence of introduced mammal remains, particularly of rat
(cf. Rattus exulans), throughout the stratigraphy show that the deposits were at least laid down
after human settlement. Of domestic animals, the chicken was identified tentatively by two
toe bones recovered at a shallow depth (10–20 cm) in Akatanui Shelter 1, where there was also
material from the European era, including a spectacle lens, iron nails and a button. A cat maxilla
was also found (0–10 cm) at this site. A probable pig tooth occurred in the surface spit (0–10
cm) at Akatanui Shelter 3. Goat bones were the most common of introduced faunal material,
occurring in surface deposits at Akatanui, Test-pit E1, at Tangarutu E1 in surface sand and at
Angairao Shelter E, where the remains were at 30–40 cm depth below an oven. No dog bone
was recovered and the ethnographic evidence that Rapa had no introduced mammals other than
the Pacific rat before European arrival is sustained by the archaeological data. No seal bone was
recorded.
The reptile remains were exclusively from marine turtle. Plastron at Angairao Shelter E
occurred in conjunction with goat bone (above), but at Tangarutu excavation E1 it was recovered
from near the base of the site, in Spit 18. Turtles are not common around Rapa, but still the
scarcity of their remains in the sites is surprising.
Fifteen species of birds, represented by a total of 118 individuals, were identified by remains
from all sites combined (Table 6.1). Most bones identified were of species currently breeding
at Rapa (Pratt et al. 1987; Thibault and Varney 1991b), and most (90%) of the bird remains
identified were seabirds. The dominant species was Kermadec petrel, comprising almost half of
all avian remains. As this petrel is one of the more common seabirds breeding on Rapa today
(Thibault and Varney 1991b), its prevalence in the deposits is not surprising.
The little shearwater bones are presumed to be from the race endemic to Rapa. This is a
poorly known bird that currently only breeds on a few small islets off the main island (Thibault
and Varney 1991b). Black-winged petrels breed on some of the islets offshore from Rapa today
(Thibault and Varney 1991b). The species is undergoing a rapid expansion in its breeding
range (e.g. Jenkins and Cheshire 1982; Tennyson 1991), but the remains identified in this
study demonstrate that the Rapa population has been long-standing and therefore is not a new
colonisation, contrary to the conclusions of Thibault and Varney (1991b). Red-tailed tropic
birds, grey noddies, brown noddies and white terns are also common seabirds currently breeding
on Rapa (Thibault and Varney 1991b).
The four storm petrel bones identified are larger than the bones of the white-bellied storm
petrel subspecies breeding at the Kermadec and Juan Fernandez islands and smaller than bones
of the Polynesian storm petrel, so they are consistent with the large endemic white-bellied
subspecies breeding on islets off Rapa (Murphy and Snyder 1952; Thibault and Varney 1991b).
However, the flesh-footed or pink-footed shearwater specimen is a surprising find. Rapa is
thousands of kilometres outside the normal range of both these species, which breed to the
west and east of Rapa respectively (Marchant and Higgins 1990). Both species are migrants –
perhaps their migration route takes them nearer Rapa than previously realised. If so, the Rapa
specimen may be a vagrant. Alternatively, Rapa might be a former breeding location of one or
other species.
terra australis 37
Bird, reptile and mammal remains from archaeological sites on Rapa Island
111
Two rail bones were identified. The tarsometatarsus is similar to but smaller than that of a
banded rail. Therefore, it seems likely that an endemic form, in the widespread genus Gallirallus,
was once resident on Rapa. Estimates of the number of extinct rails in Oceania range from 200
to 3000 (Steadman 1995; Tennyson and Martinson 2006), so an extinct form on Rapa is to be
expected. The only rail known from Rapa today is the much smaller spotless crake (Pratt et al.
1987), which was represented by one distal tibiotarsus.
The smaller pigeon bones were assigned, on the basis of size alone, to the sole extant dove
on Rapa. In addition, at least one other larger species of pigeon, similar to a large Ducula species,
once occurred. Large Ducula species have been eliminated from several of the Pacific Islands
(e.g. Worthy 2001), so an extinct species on Rapa might be expected. The fragmentary pigeon
sternum of unknown relationships (S.44400 from Spit 3 of excavation S1 at Tangarutu) has a
low, rounded keel, indicating that this species may have had reduced powers of flight.
The three parrot bones are the first evidence that any parrots once lived in the Austral Island
group. Few parrot species occur in Eastern Polynesia and those that do (or did) belong to the
genera Vini (lorikeets) and Cyanoramphus (parakeets) (Pratt et al. 1987). The Rapa bones fit
Cyanoramphus well, so they probably represent an extinct endemic species and extend the range
of this genus further east than previously known. Both Cyanoramphus species recorded from
tropical Eastern Polynesia (C. zealandicus from Tahiti and C. ulietanus from Raiatea) are also
extinct (Pratt et al. 1987).
Overall, remarkably few of the resident bird species on Rapa today were missing from the
archaeological samples (the missing taxa are: Christmas shearwater, Murphy’s petrel, Polynesian
storm petrel, grey duck, golden plover, wandering tattler and long-tailed cuckoo). The small
size of the collection is likely to explain the absence of most of these species but they might also
have been relatively uncommon on Rapa. As the remains of large birds are more commonly
preserved in such deposits, the absence of some large taxa that are otherwise widespread in the
South Pacific (e.g. wedge-tailed shearwaters, boobies and frigatebirds) could suggest that these
may not have been present on Rapa.
It seems likely that further excavations will reveal more lost Rapa bird taxa because at least
three extinct species were detected in this small study. The remains of the fruit dove and local
subspecies of the little shearwater and white-bellied storm petrel are the first skeletal material of
these taxa to be collected. They provide a basis for future reference and taxonomic work. Further
work is required to determine the identity of the parrot, flesh-footed/pink-footed shearwater,
Gallirallus-type rail and large pigeon. It is highly likely that the bird extinctions detected were
the result of human impact and that seabird species now confined to nesting on small offshore
islands previously bred on main Rapa.
terra australis 37
112 Alan J. D. Tennyson and Atholl Anderson
Chicken
Parrot
Rapa fruit dove
Large pigeon
Brown noddy
Grey noddy
White tern
Spotless crake
Rail
Red-tailedtropic bird
White-bellied storm
petrel
Kermadec petrel
Black-winged petrel
Little shearwater
Flesh-footed/ Pinkfooted shearwater
Table 6.1. Distribution by site and excavation unit of identified bird bone.
Tangarutu
Location T1
1
1
14
1
3
1
Excavation E1
Spit 1
1
1
Spit 3
1
Spit 7
2
Spit 10
2
2
Spit 11
1
Spit 13
1
2
Spit 14
1
2
3
Spit 15
1
1
1
1
2
Spit 16
1
3
Spit 17
1
2
2
1
1
Spit18
1
3
Spit 19
1
2
Spit 21
2
1
1
11
38
4
MNI Total
1
3
1
6
1
2
1
4
2
1
1
2
1
Excavation E2
Spit 4
1
Spits 8-9
1
Spits 10-13
1
1
1
1
Spit 14
1
1
Spits 16-17
1
Spit 22
1
Spits 23-25
MNI Total
1
1
1
6
3
1
2
1
1
2
Excavation R1
Spit 1
1
Spit 3
Spit 4
1
Spit 6
Spit 7
1
1
Spit 8
MNI Total
2
1
1
1
3
7
Continued on next page
terra australis 37
Bird, reptile and mammal remains from archaeological sites on Rapa Island
113
Chicken
Parrot
Rapa fruit dove
Large pigeon
Brown noddy
Grey noddy
White tern
Spotless crake
Rail
Red-tailedtropic bird
White-bellied storm
petrel
Kermadec petrel
Black-winged petrel
Little shearwater
Flesh-footed/ Pinkfooted shearwater
Table 6.1. continued
Excavation S1
Spit 3
TOTAL
1
1
1
3
1
1
1
6
16
54
1
8
1
1
1
6
1
4
3
2
3
Akatanui
Shelter 1
Spit 2
1
1
Shelter 3
Spit 1
1
Spit 2
TOTAL
1
1
1
1
2
1
1
1
1
1
Angairao
Shelter C
Spit 2
1
Spit 3
1
1
Noogoriki
1
TOTAL
2
1
1
9
17
55
4
9
1
1
6
2
4
3
2
3
1
0.8
7.6
14.4
46.6
3.4
7.6
0.8
0.8
5.1
1.7
3.4
2.5
1.7
2.5
0.8
1
ALL SITES
TOTAL MNI
Per cent
terra australis 37
114 Alan J. D. Tennyson and Atholl Anderson
Acknowledgements
Thanks to Sunita Mahat for assistance with registration and databasing, Jean-Claude Stahl for help with
translating Holyoak and Thibault (1984) and Trevor Worthy for commenting on the manuscript.
References
Hall, J.V. 1869. On the island of Rapa. Transactions and Proceedings of the New Zealand Institute
1:128–134.
Holyoak, D.T. and Thibault, J-C. 1984. Contribution à L’étude des oiseaux de Polynésie Orientale.
Mémoires du Muséum National D’Histoire Naturelle Série A, Zoologie, 127:1–209.
Jenkins, J.A.F. and Cheshire, N.G. 1982. The black-winged petrel (Pterodroma nigripennis) in the
south-west Pacific and the Tasman Sea. Notornis 29:293–310.
Marchant, S. and Higgins, P.J. 1990 (co-ordinators). Handbook of Australian, New Zealand and
Antarctic Birds, Volume 1, Ratites to Ducks. Oxford University Press, Melbourne.
Murphy, R.C. and Snyder, J.P. 1952. The “Pealea” phenomenon and other notes on storm petrels.
American Museum Novitates 1596:1–16.
Pratt, H.D., Bruner, P.L. and Berrett, D.G. 1987. A Field Guide to the Birds of Hawaii and the Tropical
Pacific. Princeton University Press, New Jersey.
Steadman, D.W. 1995. Prehistoric extinctions of Pacific Island birds: biodiversity meets zooarchaeology.
Science 267:1123–1131.
Tennyson, A.J.D. 1991. The black-winged petrel on Mangere Island, Chatham Islands. Notornis
38:111–116.
Tennyson, A.J.D. and Martinson, P. 2006. Extinct Birds of New Zealand. Te Papa Press, Wellington.
Thibault, J-C. and Varney, A. 1991a. Numbers and habitat of the Rapa fruit-dove Ptilinopus huttoni.
Bird Conservation International 1:75–81.
Thibault, J-C. and Varney, A. 1991b. Breeding seabirds of Rapa (Polynesia): numbers and changes
during the 20th century. Bulletin of the British Ornithologists’ Club, 111(2):70–77.
Worthy, T.H. 2001. A giant flightless pigeon gen. et sp. nov. and a new species of Ducula (Aves:
Columbidae), from Quaternary deposits in Fiji. Journal of the Royal Society of New Zealand
31:763–794.
terra australis 37
7
Prehistoric fishing on Rapa Island
Yolanda Vogel
Department of Anthropology, University of Otago, Dunedin, New Zealand,
bones.stones@gmail.com
Atholl Anderson
Department of Archaeology and Natural History, The Australian National University
Introduction
This chapter presents the results of analysing fish-bone assemblages from Rapa, while Chapter
9 will discuss some points of interpretation. The material was recovered from a number of
rockshelter sites around Rapa (see Chapter 3, Figure 3.1 for site locations). Most of the fish
bone comes from the Tangarutu rockshelter, with another large assemblage from the Akatanui
3 rockshelter. Four other rockshelters (Akatanui 1, Angairao C, Angairao E and Noogoriki)
provided only small amounts of fish bone. The material was dry sieved in the field using
3 mm sieves, and bulk samples were also retained at most sites. Primary sorting occurred in
The Australian National University archaeological laboratory (Department of Archaeology and
Natural History), where bulk samples were sieved to 2 mm. Residues were retained. The fishbone component was sent to the Otago Archaeology Laboratory (OAL), in the Anthropology
Department, University of Otago, for further analysis by Vogel.
Methods
The fish bone is in a very good state of preservation. It was identified using the OAL Pacific
fish-bone reference collection, which contains 108 specimens in 33 families and 49 genera, with
seven specimens having no information below family level (see Appendix A for the full list).
Many studies of prehistoric Pacific fishing practices have relied on the protocol devised
by Anderson (1973, 1979) and Leach (1979, 1986), which involved identification of five
paired mouth bones (dentary, premaxilla, articular, maxilla and quadrate), plus various ‘special’
bones. However, recent work in Pacific fish-bone analysis at the OAL and other laboratories has
shown that different identification and quantification methods significantly influence measures
of relative abundance and the number of taxa represented, thereby affecting interpretations
(Weisler 1993; Walter 1998; Vogel 2005). During the rebuilding of the OAL Pacific fish-bone
collection in the early 1990s, it was decided to include a wider range of elements beyond the
five paired mouth bones and ‘special’ bones, in order to address a greater range of issues in
terra australis 37
116 Yolanda Vogel and Atholl Anderson
Pacific ichthyoarchaeology (Walter et al. 1996). Currently, the collection allows for the routine
identification of 17 elements from the cranium, pectoral girdle and pelvic girdle, plus otoliths
and some ‘special’ elements, and it is these that were used in this study.
Identification of the fish bone involved two steps – to element and to taxon. As much as
possible of the assemblage was identified to element, using reference specimens and diagrams
(e.g. Cannon 1987; Wheeler and Jones 1989; Rojo 1991). The remaining material consisted of
fragments with no landmark features, plus material that was unable to be identified to element
at this stage. Seventeen elements, plus dorsal spines for Acanthuridae and Balistidae, dermal
spines for Diodontidae, teeth and vertebra for Elasmobranchii, and fourth epibranchials for
Scaridae, were then identified to family (see Table 7.1). Where a family match did not exist in
the OAL reference collection, the element was recorded as Not in Collection (NIC), followed
by a number to distinguish between elements obviously from different families. Otoliths were
absent from the assemblages. It is possible they did not survive, but this does not seem likely,
given that otoliths have been shown to survive when other elements have not (Weisler 1993;
Weisler et al. 1998), and given the overall excellent preservation of the fish remains. Another
possibility, given the overall very small size of the fish represented in the assemblages, is that
even with the use of 3 mm sieves, the otoliths were so small they were not retained. Bulk
sample residues were retained, and microscopic investigation of these in the future may reveal
the presence of otoliths.
Considerable debate exists within archaeology about the best method of quantification
(Grayson 1984), and Pacific ichthyoarchaeology is no exception to this (Butler 1988; Walter
1998). Issues surrounding the use of NISP (Number of Identified Specimens) pertain to the
fragmentation level of the material analysed, differing numbers of elements identified for
different taxa, and taxa identified by ‘special’ bones, of which there are multiple numbers per
individual. The main issue with the use of MNI (Minimum Number of Individuals) has to
do with the units of aggregation used. In order to alleviate some of the problems inherent in
the different quantification methods, both NISP and MNI are reported here, along with the
raw data on which they are based, thus providing a much fuller account of the archaeological
assemblage than would be provided by MNI counts alone. The use of the two quantification
methods also allows for greater comparability between the Rapan assemblages and those from
other islands in the Pacific, as many reports contain results using only one or the other method.
Table 7.1. Elements identified to taxonomic level.
Elements
Vomer
Epihyal
Parasphenoid
Inferior pharyngeal cluster
Post temporal
Superior pharyngeal cluster
Articular
Cleithrum
Dentary
Scapula
Maxilla
Palatine
Premaxilla
Dorsal spine (Acanthuridae and Balistidae only)
Quadrate
Dermal spine (Diodontidae only)
Opercular
Tooth (Elasmobranchii only)
Preopercular
Vertebra (Elasmobranchii only)
Hyomandibular
4th epibranchial (Scaridae only)
Ceratohyal
terra australis 37
Prehistoric fishing on Rapa Island
117
NISP is given for all material, whereas MNIs were calculated using the 17 elements targeted for
family-level identification (see Tables 7.2, 7.3). MNI was calculated by taking the most frequent
sided element (or straight element count in the case of unpaired bones) for each family. No
attempts were made to account for size matching.
Results
Twenty-one families, one subclass (Elasmobranchii) and one order (Anguilliformes) were
identified in the assemblage, while at least six additional families represented in the archaeological
record were not in the OAL collection (elements designated as NIC). This latter number was
calculated based on the presence of six distinctive sets of hyomandibular bones throughout the
assemblages. The NIC material has been excluded from the remainder of this paper because
MNIs cannot be generated accurately; full details for the elements represented in this category
are given in Appendix A.
The top-ranked taxa are Scaridae (parrotfish), Muraenidae (moray eels), Chaetodontidae
(butterflyfish), Labridae (wrasses), Serranidae (grouper), Pomacentridae (damselfish) and
Congridae (conger eels). The presence of Muraenidae, Congridae, Chaetodontidae and
Pomacentridae in the highest-ranked taxa is somewhat anomalous, and will be discussed further
later. The remaining three taxa, Scaridae, Serranidae and Labridae, commonly rank highly in
tropical East Polynesian fish-bone assemblages. The following families were represented only in
small numbers: Lethrinidae, Acanthuridae, Mullidae, cf. Aulostomidae, Belonidae, Exocoetidae,
Holocentridae, Kyphosidae, Mugilidae, Siganidae, Bothidae, Carangidae, Diodontidae,
Elasmobranchii, Monacanthidae and Anguilliformes, and a number of specimens not in the
reference collection. The results for each site are presented below.
Tangarutu Cave
The material from Tangarutu forms the largest assemblage, with a total of 14,419 bones, and
it is made up of material from a number of sampling stations within the site (Appendix A).
Each of these will be considered as separate units, as the stratigraphic association between them,
although broadly apparent (Chapter 3), is uncertain in detail. The bulk of the material comes
from the E1 and E2 contiguous units, considered here as a single unit, and the R1 unit.
Excavation E1/E2
This was an excavation of approximately 1.5 m2, although, as explained in Chapter 3, only
part of the E2 area was taken down to natural. The stratigraphy was divided into three levels
(I (oldest) to III, details in Chapter 3) and results of bone identification are grouped into these
and shown in Figures 7.1–7.3. A total of 21 families were identified to taxa, although the exact
number and composition varies between levels. The assemblage from this excavation consists of
11,417 pieces of fish bone.
In the lowest level, the top five taxa, all representing more than 5% of the assemblage each,
make up 87% of the assemblage. The remaining 11 taxa in this layer all comprise less than 2%
of the assemblage each. Figure 7.1 shows a clear dominance by Scaridae, comprising 40% of the
assemblage. The other taxa in the top five, in order of abundance, are Muraenidae, Labridae,
Chaetodontidae and Serranidae.
The pattern seen in Level I changes in Level II (Figure 7.2). Scaridae drops to 17% of the
assemblage, and Muraenidae becomes the top-ranking taxon. Chaetodontidae also increases to
16%, Labridae increases slightly to 12%, while a second eel family, Congridae, takes the place of
Serranidae as fifth-ranked taxon. Serranidae drops to sixth in terms of rank order, representing
terra australis 37
118 Yolanda Vogel and Atholl Anderson
5
11
5
1
8
2
1
13
5
1
1
3
1
1
1
8
1
8
1
1
1
2
1
2
8
1
1
24
7
21
16
19
1
12
4
1
69
1
13
1
33
7
1
82
1
19
1
1
9
115
6
5
62
7
2
27
1
T1
Total
12
Siganidae
1
Serranidae
1
Scaridae
1
Pomacentridae
1
Muraenidae
I
1
Mullidae
R1
1
5
Mugilidae
1
II
15
Monocanthidae
III
R1
Lethrinidae
R1
1
Labridae
1
Kyphosidae
2
1
Holocentridae
1
I
Exocoetidae
II
Elasmobranchii
1
Diodontidae
2
Congridae
Belonidae
1
Chaetodontidae
Aulostomidae
E1/E2 III
Carangidae
Area Level
Tangarutu
Bothidae
Site
Acanthuridae
Table 7.2. MNI for all sites.
1
1
1
2
26
7
93
3
2
6
1
1
2
5
3
12
T3
Akatanui 3
C1
1
E1
1
1
1
1
2
8
1
1
1
1
1
8
1
9
23
A1
1
1
1
Angairao C
Angairao E
III
II
1
1
1
1
1
1
5
2
10
1
5
2
1
1
7
1
4
1
7
6 100
33 125
53
3
1
I
1
1
1
1
Total
8
4
4
1
3
50
30
5
5
2
5
6
51
8
1
2
Mugilidae
1
Monocanthidae
Noogoriki
2
1
Siganidae
1
Serranidae
Akatanui 1
1
5
31
2
2
6
11
1
4
56
3
1
27
1
1
1
2
42
1
5
1
5
97
15 180 230
712
67
3
75
57
1 337
52
4 303
57
7 505
3
56
2
73
55
291
2
26
32
13
89
2
1
T1
1
1
1
2
T3
Akatanui 3
0
C1
1
E1
2
1
2
2
5
33
1
13
2
2
3
31
4
31
A1
Akatanui 1
1
1
2
1
Angairao C
Angairao E
III
II
1
1
1
1
Noogoriki
9
12
2
terra australis 37
4 156 130
11
23
11
447
3
20
4
2
6
28
14
47
2
1
3
19
5
1
6
3
5
1
12
1
6
4
12
2
1
5
75
15
2
1
1
19
75 166
1
1
I
Total
Total
24
52
Scaridae
I
3
2
Pomacentridae
R1
1
1
8
6
Muraenidae
1
II
22
26
3
Mullidae
III
R1
43
Lethrinidae
R1
1
6
Labridae
5
4
Kyphosidae
3
6
28
Holocentridae
1
I
56
Exocoetidae
II
Elasmobranchii
4
Diodontidae
7
Congridae
Belonidae
6
Chaetodontidae
Aulostomidae
E1/E2 III
Carangidae
Area Level
Tangarutu
Bothidae
Site
Acanthuridae
Table 7.3. NISP for all elements identified to taxa.
12 219
15
1
2
17 341
1
99 907 518
6
29
9
Prehistoric fishing on Rapa Island
119
6% of the assemblage. A further eight taxa comprise 3% or less of the assemblage each.
There are also notable changes in rank order in Level III (Figure 7.3). Serranidae increases to
14% of the assemblage, and is ranked third. Pomacentridae now represents 6% of the assemblage,
up from 1%. Muraenidae remains the top-ranked taxon, but falls in terms of percentage of the
assemblage.
Figure 7.1. Relative abundance (MNI)
of taxa for Tangarutu E1/E2 Level I.
Figure 7.2. Relative abundance (MNI)
of taxa for Tangarutu E1/E2 Level II.
Figure 7.3. Relative abundance (MNI)
of taxa for Tangarutu E1/E2 Level III.
terra australis 37
120 Yolanda Vogel and Atholl Anderson
Excavation R1
This material comes from a 1 m2 square excavation and the assemblage consists of 2845 pieces
of fish bone, representing 14 families able to be identified to taxa. Overall, the top-ranking taxa
are the same as those seen in the E1/E2 area, although there are some differences in rank order
(Figures 7.4–7.5).
The lowest level of the R1 area contained very little fish bone (NISP=22), with only one bone
able to be identified to taxon. This belonged to the family Muraenidae. In Level II, the top four
taxa make up 78% of the assemblage, with a further six taxa contributing the remaining 22%.
Taxa in this latter group all have an MNI of one. Figure 7.4 shows a dominance by Muraenidae,
comprising 33% of the assemblage, followed by Scaridae, Congridae and Serranidae in order of
abundance. Together, Muraenidae and Scaridae make up more than half the Level II assemblage.
Level III shows a dramatic change in the pattern from that in Level II (Figure 7.5).
Muraenidae remains the top-ranking taxon, but there is a considerable increase in the number
of Chaetodontidae represented, increasing to 21% of the assemblage (Chaetodontidae is present
in Level II, but with an MNI of only one, falls into the category of ‘other’). Concurrently,
there is a marked decrease in Scaridae, now representing only 10% of the assemblage. Labridae
numbers also increase in Level III, contributing 13%, while Congridae falls slightly in terms
of percentage and Serranidae remains stable. Six taxa, each with an MNI of one, make up the
remaining 10% of the assemblage.
Figure 7.4. Relative abundance (MNI)
of taxa for Tangarutu R1 Level II.
Figure 7.5. Relative abundance (MNI)
of taxa for Tangarutu R1 Level III.
terra australis 37
Prehistoric fishing on Rapa Island
121
In light of the results for the R1 area, it was decided to re-examine the E1/E2 data for Level
II, to see whether the decrease in Scaridae and increase in Chaetodontidae could be pinpointed
more accurately. This revealed that Scaridae is still the dominant taxon at the base of Level II
(Figure 7.6), falling to fourth rank in the middle of Level II, where it comprises 13% of the
assemblage (Figure 7.7), and falling further to fifth rank and only 6% of the assemblage at the
top of Level II (Figure 7.8). Chaetodontidae represents only 7% of the assemblage at the base of
Level II, although it ranks third, but it increases significantly in the middle of Level II to become
the dominant taxa. At the top of Level II, Chaetodontidae and Muraenidae rank first equal.
Thus, the initial analysis of Level II masks some quite significant changes in the assemblage, and
the increasing importance of Chaetodontidae.
The E1/E2 data for Level III were also reanalysed, to see whether the changes in that level
were also more abrupt than the initial results suggest, particularly in terms of the Scaridae,
which clearly increase in numbers again from the low seen at the top of Level II. The data
for Level III were split into two subsets representing spits 1–4 and spits 5–8. The decision
about where to make the split was chiefly based on the appearance of fish hooks in this level
(see below). Muraenidae maintains its dominance in the bottom half of Level III (Figure 7.9),
but Chaetodontidae numbers drop off considerably, falling to third rank and only 12% of the
assemblage. In contrast, Serranidae numbers increase significantly, to 17%. Scaridae increases
to 12% of the assemblage, placing it fourth equal with Chaetodontidae, while Labridae and
Congridae both drop in numbers and rank.
Figure 7.6. Relative abundance (MNI) of
taxa for Tangarutu E1/E2 Level II spits 9–10.
Figure 7.7. Relative abundance (MNI)
of taxa for Tangarutu E1/E2 Level II spits
11–12.
terra australis 37
122 Yolanda Vogel and Atholl Anderson
Figure 7.8. Relative abundance (MNI) of
taxa for Tangarutu E1/E2 Level II spits 13–14.
In the top half of Level III, Scaridae becomes the dominant taxon once again, representing
20% of the assemblage (Figure 7.10). Muraenidae numbers drop, along with its fall to second
rank, as does Serranidae, slipping back to fourth rank and 13% of the assemblage. Chaetodontidae
remains fairly stable. It is in this upper part of Level III that Pomacentridae increases in numbers
also, now representing 8% of the assemblage. As with Level II, the initial analysis of Level III
masked some interesting changes in terms of taxa representation and rank order abundance.
Additional Tangarutu assemblages
Tangarutu T1 and T3 are both very small assemblages resulting from facing of sections (i.e.
cutting straight sections down exposed faces to inspect stratigraphy and content). The T1 area
yielded 49 pieces of fish bone, of which two could be identified to taxa, while another two
belonged to a taxon not represented in the OAL collection. This resulted in an MNI of one for
Serranidae and Holocentridae, and an MNI of two for the unknown taxon. The T3 area yielded
108 pieces of fish bone, of which none were identifiable to taxa. The majority of this assemblage
was made up of scales, with a NISP of 72.
Akatanui shelters
The Akatanui Bay assemblages come from two separate rockshelters: Akatanui 1 and Akatanui 3
(Appendix A). The Akatanui 3 excavations resulted in material from three areas: C1, which was
an excavation of 2 m2, and test pits at E1 and A1.
Akatanui 3 C1 excavation
This area yielded the greatest amount of fish bone for this shelter, with a total of 2490 pieces
recovered. The results of the taxonomic identification, based on MNI, are shown in Figure 7.11.
The taxa exploited are essentially the same as those at Tangarutu, but the rank ordering is quite
different. Scaridae and Pomacentridae together make up more than half of the assemblage.
While both eel families are still represented in the top six taxa, their contribution is far less than
at Tangarutu, particularly in the case of Muraenidae. Interestingly, Chaetodontidae, one of the
top-ranking taxa at Tangarutu, makes very little contribution to this assemblage.
Additional Akatanui assemblages
Akatanui E1 and A1 are both very small assemblages. The E1 material consists of 122 pieces of
bone. Three taxa were identified: Scaridae (MNI=3), Serranidae (MNI=2) and Acanthuridae
(MNI=1). The A1 assemblage contains only nine pieces of bone, but six of these were identified
terra australis 37
Prehistoric fishing on Rapa Island
123
Figure 7.9. Relative abundance (MNI) of
taxa for Tangarutu E1/E2 Level III spits 5–8.
Figure 7.10. Relative abundance (MNI) of
taxa for Tangarutu E1/E2 Level III spits 1–4.
Figure 7.11. Relative abundance (MNI)
of taxa for Akatanui 3 C1.
terra australis 37
124 Yolanda Vogel and Atholl Anderson
to taxa, resulting in an MNI of one each for Scaridae and Serranidae. Akatanui 1 yielded a
very small assemblage, consisting of 338 pieces of fish bone. Six taxa were identified: Scaridae
(MNI=5), Serranidae (MNI=3), Acanthuridae (MNI=1), Congridae (MNI=1), Elasmobranchii
(MNI=1) and Labridae (MNI=1).
Angairao Bay shelters
The material from Angairao Bay comes from three provenances (Appendix A). The most
productive of these was the excavation at Angairao E.
Angairao E excavation
This was a 1.2 m x 0.6 m excavation down to 70 cm and then 0.5 m x 0.6 m below that
(Chapter 3). In the top 70 cm, there were remains of an oven, but with no associated midden.
The midden was in an underlying unit (‘Oven 2’), at 70–165 cm depth, which was divided into
three levels (Chapter 5) for bone analysis.
The Angairao E assemblage yielded a total of 269 pieces of fish bone. While nine taxa
were identified in the assemblage, only three are present in all levels: Serranidae, Labridae and
Muraenidae. Scaridae, however, has the highest overall MNI. The MNI and NISP by level can
be seen in Tables 7.2 and 7.3.
Additional Angairao assemblages
The Angairao C assemblage consisted of 30 pieces of fish bone, from which four taxa were
identified: Labridae (MNI=2), Scaridae (MNI=1), Muraenidae (MNI=1) and Siganidae
(MNI=1). The Noogoriki assemblage, from the opposite side of Angairao Bay (Chapter 3),
consisted of 27 pieces of fish bone, with four taxa able to be identified: Scaridae (MNI=4),
Muraenidae (MNI=1), Serranidae (MNI=1) and Carangidae (MNI=1).
Discussion
The reconstruction of past fishing practices requires more than just archaeological faunal data.
Information on local ecology, ethnographic accounts of fishing, and information on material
culture relating to fishing (Walter and Weisler 2002) are also useful sources. Ethnographic
information on fishing for Rapa is discussed in Chapters 2 and 9, and these data can be combined
with analogies from other Pacific islands to provide a more rounded interpretation than afforded
by the faunal data alone.
Fishing strategies can also be reconstructed to some extent by considering the preferred
habitats and feeding behaviours of the taxa represented in the faunal assemblage (Butler 1994;
Walter 1998; Allen et al. 2001; Weisler 2001). This method is not without its problems, as
species within some families can inhabit different ecological zones (e.g. Serranidae). Using
feeding behaviour to infer capture techniques is also problematic, because some taxa can be
caught using a number of techniques, and feeding behaviour and capture techniques can also
vary between species within a family. Nevertheless, such data can help to delimit the possibilities
of fishing strategy.
Table 7.4 has been constructed by taking a broader view of tropical Polynesian ethnographic
information on fishing techniques and ecological variation. Walter (1998) divides the marine
exploitation zones into five categories (offshore pelagic, inshore pelagic, inshore benthic, reef
edge and reef flat), and Allen et al. (2001) include lagoon as one of their zones. Rapa differs from
many Pacific islands in that it has lagoon of any extent only in Ha’urei Harbour, while around
the open sea perimeter there is almost no intertidally exposed fringing reef. In this respect,
a better analogy in terms of marine environment may be provided by Easter Island. Rapa is
terra australis 37
Prehistoric fishing on Rapa Island
125
surrounded by a gently sloping platform, on which coral communities grow in the sheltered
bays, perhaps providing a zone similar to the reef flats of other islands. With this in mind, the
marine zones have been broken into only three categories: inshore, benthic and pelagic.
Another factor likely to be important to reconstructions of past fishing behaviour is the
dichotomy between the marine environment within the many bays and that outside the bays.
The bays are noted as being relatively calm, although Ha’urei Harbour is often choppy and
turbid, especially towards the head. The remaining coastline is characterised by steep cliffs and,
frequently, rough waters. While there are areas of rocky shore outside the bays, the predominance
of cliffs rising almost directly from the water’s edge ensured fishing outside the bays was usually
from a canoe.
Putting the information in Table 7.4 together with the results of fish-bone analysis, it is
possible to draw some conclusions about the nature of fishing at the various sites around Rapa.
Tangarutu
Fishing from this site seems to have concentrated almost exclusively on the inshore zone. The
few taxa that are associated more often with offshore zones rank very low, and they are probably
best interpreted as opportunistic or unintentional catches of fish that came into the bay and
Table 7.4. The ecology of taxa identified in the Rapan assemblages and possible methods of capture.
Family
Habitat/Zone
Activity
Feeding
Capture technique
Acanthuridae
Inshore benthic, reef edge, reef flat
Diurnal
Herbivore
Netting, trapping, spearing, angling
Aulostomidae
Outer reef, reef flat
Belonidae
Inshore
Trolling
Bothidae
Inshore
Netting
Carangidae
Reef flat, passes, outer reef
Various
Piscivore
Angling, netting, spearing
Chaetodontidae
Inshore
Diurnal
Invertebrates
Netting
Congridae
Reef
Nocturnal
Carnivore
Diodontidae
Reef flat, outer reef
Nocturnal
Invertebrates
Angling, trapping, netting
Elasmobranchii
Outer reef, pelagic
Omnivore/benthic
carnivore
Angling, spearing, netting
Exocoetidae
Pelagic
Various
Omnivore
Netting
Holocentridae
Reef flat, outer reef
Nocturnal
Invertebrates
Angling, netting
Kyphosidae
Reef flat, outer reef
Diurnal
Herbivore
Labridae
Reef flat, outer reef
Diurnal
Invertebrates
Angling, netting, trapping
Lethrinidae
Reef flat, outer reef
Nocturnal
Invertebrates
Angling, netting
Monocanthidae
Inshore
Omnivore
Mullidae
Reef flat
Omnivore
Netting, angling
Muraenidae
Reef flat
Nocturnal
Carnivore
Spearing, poisoning, trapping, snaring, netting
Pomacentridae
Inshore
Diurnal
Omnivore
Angling, netting, trapping
Scaridae
Inshore
Diurnal
Herbivore
Netting, spearing, trapping
Serranidae
Inshore
Piscivore
Angling, netting, trapping, spearing, poison
Siganidae
Inshore
Piscivore
Mugilidae
Netting, trapping
terra australis 37
126 Yolanda Vogel and Atholl Anderson
closer to shore than usual. This pattern is fairly typical of Pacific marine exploitation systems
(Anderson 1986:ix).
The oldest level in the E1/E2 area at Tangarutu shows a clear dominance of Scaridae, but
this gives way to a dominance by Chaetodontidae and Muraenidae in the middle level and
Muraenidae in the upper level. Interestingly, Scaridae numbers increase again in the upper level.
While the presence of Muraenidae in the assemblage is not unusual in itself, these eels usually
rank very low in terms of relative abundance at other sites throughout the Pacific.
Another interesting observation about this assemblage is the high incidence of
Chaetodontidae, particularly in Level II. This represents the first positive identification of
Chaetodontidae in an assemblage from the Pacific. This is probably related to the methodology
used for identification. However, it cannot be assumed that Chaetodontidae has merely been
overlooked in all other Pacific assemblages, nor that it would occur in such high numbers were
it to be identified in other assemblages. Given the dominance by Muraenidae in much of the
Tangarutu assemblage, it is safe to say that this assemblage differs considerably from many
others in the region in terms of some of the favoured, or available, taxa, and the presence of
Chaetodontidae is probably evidence of this.
Occupation at Tangarutu spanned about 500 years, and there is obvious change through
time in the assemblage. The first of these changes is an apparent decrease in Scaridae. One
possible explanation for this is simply differential survivability. Scaridae has some elements that
are particularly robust, and thus has an increased likelihood of survival through time (Bilton
2001). In this scenario, numbers of identifiable elements for other taxa are fewer in older layers
due to taphonomic factors, while Scaridae elements remain intact, but in younger layers greater
survival of elements from other taxa reduces the proportional representation of Scaridae. When
considering the percentage of unidentified fragments in each level, the amount for the oldest
level is significantly higher than for the other layers (32.5% in the oldest, versus 22% for the
middle level and 18.5% for the youngest), which does lend weight to this argument, although
no data were collected on the size of the fragments, so it might be only that the unidentified
fragments in the lowest layer are more fragmented. If so, that was not obvious during excavation;
indeed, the contrary seemed more apparent.
However, as the assemblage as a whole shows excellent preservation, it is unlikely that
differential survivability has skewed the numbers in favour of Scaridae to any great extent, and
therefore it cannot explain the marked change in abundance between the lowest and middle
levels. Also, Labridae is subject to the same positive bias as Scaridae in terms of survivability
(Bilton 2001), yet its relative abundance between the oldest and middle levels remains fairly
constant. Another possible explanation is that the earliest occupants focused their attention
heavily on Scaridae, thereby depleting the standing stock, which may have already been low
in comparison to other islands in the region due to Rapa’s subtropical location. Breeding rates
may have been lower for the same reason. However such an explanation is very difficult to
substantiate without further data on fish numbers, breeding rates and so on for the island.
Environmental changes may also provide an explanation for the changes in relative abundance
of taxa evident at Tangarutu, and the following chapter returns to this topic.
The other major temporal change that occurs in this assemblage is between the middle
level and the upper level. Here, Serranidae increases from 6% to 17% of the assemblage and
moves from sixth-ranked taxa in terms of relative abundance, to second in the bottom half of
the level, although its numbers drop off again slightly in the upper half. The upper level also sees
Pomacentridae enter the top-ranking taxa for the first time, comprising 8% of the assemblage in
the upper half of the level, and Lethrinidae also increases in numbers.
terra australis 37
Prehistoric fishing on Rapa Island
127
The data for the R1 area at Tangarutu are similar to those from the E1/E2 area in terms of
taxa composition. There are, however, some differences in relative abundance. There is a general
stratigraphic connection between the two areas (Chapter 3), but it is not certain that the levels
at R1 occupy the same periods as those at E1/E2.
The lowest level in the R1 area does not contain sufficient fish bone to make any statements
about it. In Level II, Muraenidae dominates, with Scaridae making a significant contribution also.
This changes significantly in Level III, where Chaetodontidae numbers increase substantially,
while Scaridae numbers drop. This level shows considerable similarity to the upper two-thirds
of the middle level in the eastern area, again indicating a predominance of netting or basket
trapping. Of particular note is the fact that the rapid decrease in the number of Scaridae is
evident in both areas.
Akatanui
The assemblage from Akatanui 3 shows some interesting differences from those at Tangarutu.
While three areas were excavated at Akatanui 3, only one of these provided sufficient fish bone to
make any meaningful comment on fishing strategies. Radiocarbon dates suggest that occupation
at this shelter began later than at Tangarutu, but that occupation at the two locations overlaps.
Akatanui is located on the eastern, and windward, side of the island, and although the shelters
investigated are located further into the bay than Tangarutu on the leeward side of the island,
the differences in ecology and environment may be expected to have had some impact on fishing
in terms of accessibility to some families of fish.
While the same species as in the Tangarutu assemblage are seen at Akatanui 3, there
are differences in ranking and relative abundances. The most startling of these is the large
proportion of Pomacentridae present, making up 25% of the assemblage and almost equal to
the proportion of Scaridae. Pomacentridae is rarely identified in Pacific fish-bone assemblages,
and when it is, it ranks very low. While this is probably due at least in part to the methodology
used for identification, the large number present in this assemblage is surprising. It is interesting
to note that Chaetodontidae ranks very low in the Akatanui 3 assemblage. The question of the
significance of both Pomacentridae and Chaetodontidae in the Rapan assemblages is taken up
below.
Eels continue to be important at Akatanui 3, as at Tangarutu, but in lower numbers.
Labridae and Serranidae also contribute to the top-ranking families, with roughly the same
numbers as Muraenidae and Congridae.
The differences seen between the Tangarutu assemblages and Akatanui 3 are probably
related to ecology, perhaps due to the leeward/windward distinction, rather than a choice to
focus on different fish at each site.
Unusual taxa
Returning to the large numbers of Chaetodontidae and Pomacentridae in the Rapan assemblages,
ethnographic data suggest that these fish should also be present in other Pacific Island fish-bone
assemblages, and that their absence in other assemblages is therefore due, at least in part, to
methodological issues. Akimichi (1986), Goto (1986, 1990), Kirch and Dye (1979) and Masse
(1986, 1989), in particular, note the presence of both these families in fish catches recorded
ethnographically for Satawal, Hawaii, Niuatoputapu and Palau respectively. Ayres (1986)
also reports Chaetodontidae and Pomacentridae as families known or suspected to have been
exploited on Easter Island, and provides a schematic drawing showing the zones in which they
were probably caught. Both Hawaii and Easter Island share similarities with Rapa in terms
of environment and marine ecology, with the cooler climates of these islands inhibiting coral
terra australis 37
128 Yolanda Vogel and Atholl Anderson
growth and their remoteness resulting in an impoverished range of marine taxa (Anderson 2001).
The coastal characteristics of Easter Island are analogous to those of Rapa in terms of the lack of
a fringing reef and the presence of deep embayments (Ayres 1986; Anderson 2001). Particularly
for the small number of Pacific Island assemblages that have yielded Pomacentridae mouth
parts, a reanalysis using the methodology used for Rapa may provide some very interesting
results, given that not a single bone from the standard five paired mouth bones normally used
in analyses of Pacific fish bone was identified in the Rapan assemblages.
The large number of marine eels present in the Rapan assemblages, particularly those from
Tangarutu, is also somewhat anomalous. This is unlikely to be due to methodology, as MNIs
for these were based on the five paired mouth parts for both Congridae and Muraenidae.
Rather, it seems likely that the environmental conditions on Rapa resulted in eels being more
easily accessible, or perhaps more plentiful, than those families usually more favoured by
Polynesians.
Fish size
A final observation about the Rapan assemblages is the very small size of the remains. While
no quantitative data on element size was collected (although it could be in another project on
the bone assemblages), casual inspection suggests the bones are much smaller than is typical
for tropical fish-bone assemblages. This is true for all families represented in the assemblage,
although the occasional large example does exist. In this respect, while Pomacentridae and
Chaetodontidae could possibly be considered of little economic importance in many fish catches
due to their size, for Rapa they are probably within the average size of the catch, although, of
course, this cannot be substantiated without quantitative data.
While Rapa may appear anomalous in terms of fish size, there are several other sites
throughout the Pacific where the archaeological fish-bone remains have been particularly small,
including Nan Madol, Leluh on Kosrae, Palau, and Cikobia in Fiji, and sites in New Caledonia
and the Loyalty Islands (Davidson et al. 2002). Leach et al. (1996) proposed various explanations
for the small size of the fish from Nan Madol: fishing was concentrated in areas where small fish
congregated (i.e. the canals formed by the artificial islets that make up Nan Madol); the fish
harvest was distributed according to rank (resulting in all the small-fish remains ending up in
one area); the selection of small specimens for ritual purposes; ecological impact (i.e. decreasing
size through time due to over-harvesting); the use of small mesh nets; and the use of fish poison.
Some of these explanations may be applicable to Rapa, although others can be ruled out.
Fishing in confined areas where small fish congregated seems an unlikely explanation for
Rapa, as areas such as those characteristic of Nan Madol were not present. It should be noted,
however, that stone-walled fish weirs were used historically and can be seen in some of the bays
around the island, especially in the main harbour. Over-harvesting also seems unlikely, given
that Level I in Tangarutu is thought to represent initial, or very early occupation of the island
(Kennett et al. 2006), and the fish remains are small throughout the entire span of occupation.
Distribution of fish according to rank also seems unlikely, given that the fish bone from all
excavated areas, in some cases several areas within one site, is equally small, and ritual uses can
be ruled out for the same reason. Of the possibilities listed by Leach et al. (above), this leaves
small meshed nets, the use of fish poison, and environmental constraints as the most likely
explanations for the small size of fish remains from Rapa.
A case has already been made for the use of netting to account for many of the top-ranking
fish families seen in the Rapan assemblages, and Masse (1986) discusses the use of fine-meshed
nets (derek) for capturing Chaetodontidae in Palau. However, this does not account for the small
terra australis 37
Prehistoric fishing on Rapa Island
129
size of the fish likely to have been caught using other methods, such as the Serranids and eels.
Davidson et al. (2002) propose an argument for fish poisoning at Nan Madol, but these
waterways are especially slow moving. Poisoning is a possible explanation, although it was
unknown ethnographically (Stokes n.d.). The presence of fish hooks at Tangarutu testifies to
the use of angling, and the very small size of some of them suggests that quite small fish were
being targeted by angling. It seems likely, therefore, that the small size of the fish in the Rapan
assemblages is best attributed to environmental factors relating to climate and accessibility. The
cooler waters surrounding Rapa might have resulted in slower growth rates and smaller mature
sizes for families better suited to tropical waters. Alternatively, as even in modern times fishing
is mostly confined to the calmer waters of the bays (Hanson 1970), larger-bodied offshore fish
might have been caught rather seldom.
Pacific comparisons
How does the archaeological evidence of Rapan fishing compare with that from other islands in
the Pacific? The analysis of archaeological fish bone from the Pacific has revealed an emphasis on
inshore exploitation, focused mainly on the reef but also including the benthic zone, with the
dominant families caught including Scaridae, Serranidae, Lethrinidae, Labridae, Diodontidae,
Holocentridae, Acanthuridae, Balistidae, Carangidae and marine eels (Butler 1988, 1994; Masse
1989; Allen 1992, 2002; Nagaoka 1993; Anderson 1997; Walter 1998; Leach and Davidson
2000; Allen et al. 2001; Weisler 2001; Walter and Anderson 2002). Exceptions are known;
some families that usually rank fairly low in assemblages dominate in a few cases, especially
where offshore fishing is more prominent, or it has been focussed on one family (Butler 1994;
Leach et al. 1997; Fraser 1988, 2001; Rolett 1998; Davidson et al. 1999; Leach and Davidson
2000; Walter and Anderson 2001). In terms of captured species, Pacific fish remains from
archaeological sites generally represent a wide range of species drawn from a limited number of
families.
In both the tropical Pacific and New Zealand selective predation of a particular family of
fish is suggested (Leach and Boocock 1993; Anderson 1997 (but see Weisler et al. 1998); Leach
et al. 1997; Fraser 1998, 2001; Walter and Anderson 2001). Scombridae (tuna) remains are
generally scarce or absent in Pacific fish-bone assemblages, but it appears that tuna fishing was
more common in some periods in the Marquesas and Society Islands during prehistory (Leach
et al. 1997; Fraser 1998, 2001; Rolett 1998; Davidson et al. 1999; Leach and Davidson 2000).
Walter and Anderson (2001) observed that Lethrinidae, a family of benthic-dwelling carnivores,
provided the dominant catch at Emily Bay, Norfolk Island, suggesting that temperate waters
resulted in a fishing strategy more akin to that of New Zealand than that of tropical Polynesia.
In New Zealand, fish-bone analysis implies an emphasis on barracouta (Thyrsites atun) in the
southern South Island, although there are exceptions, as at Kakanui. Whether these are formed
by different targeting or different analytical methodology remains open to debate (Weisler et al.
1998; Anderson and Smith 1999; Weisler 1999).
The Pacific fishing industry, derived from the Lapita cultural complex (Walter 1989),
displays a great deal of homogeneity across the region in terms of basic fishing strategies. But
there are important differences at both the intra- and inter-island levels. Butler (1994) reviewed
the evidence for Lapita fishing strategies and found that, contrary to earlier assertions, angling
was used by the Lapita peoples, as evidenced by the presence of carnivorous fish remains at some
sites. It was also found that eastern sites were dominated by herbivorous/omnivorous species,
while sites in western Melanesia tended to contain more carnivorous fish, implying a regional
difference in fishing strategies.
terra australis 37
130 Yolanda Vogel and Atholl Anderson
While the strategies employed across the region are much the same, the species targeted show
variations in relative abundance. Specialisations, such as in tuna fishing, are an extreme example
of this. The tuna fishing industry seems to be an example of cultural choice, but in some sites
the differences appear to be related to local ecology. Scaridae was a favoured fish family during
prehistory, yet its remains are scarce in the fish-bone assemblages from Ma’uke and Mangaia.
Walter (1998) attributes this to the inaccessibility of the outer reef face on these two islands,
highlighting the need for a knowledge of local fishing grounds, ecology and accessibility when
analysing fish remains.
While Rapa seems to fit the mould of an emphasis on inshore exploitation in general terms,
it has already been noted that the Rapan assemblages differ on a number of points, specifically
the high number of marine eels present, the presence of Chaetodontidae and Pomacentridae,
and the small size of the fish. But are there other islands where a similar pattern can be seen?
Comparisons between assemblages are difficult due to differences in identification and
quantification methodologies. Many early reports relied on a limited number of elements for
identification, and few have used more than the standard five paired mouth bones. Furthermore,
some researchers report only NISP values, while others give only MNI. NISP values may
be inflated for some families due to the use of a wider range of elements for identification
(e.g. Scaridae) or the use of elements for which each individual has multiple elements (e.g.
Diodontidae). The present study is obviously skewed in comparison with others by the use of
a wider range of elements for identification purposes. However, if these factors are taken into
account, some observations are able to be made.
Eels are generally considered unusual in prehistoric middens, yet they show up in a number
of middens throughout the region (e.g. Kirch and Yen 1982; Allen 1992; Nagaoka 1993; Leach
et al. 1996; Walter 1998; Butler 2001; Walter and Anderson 2001; Davidson et al. 2002; Walter
and Anderson 2002), although only occasionally in high-ranking positions. However, only on
one other island is the dominance as clear as at Rapa, and that is on Easter Island, with the other
dominant family in these sites being Labridae (Ayres 1986). The similarities between Rapa and
Easter Island in terms of ecology and geomorphology have already been discussed, and it is not
surprising to see similarities between the two in terms of fishing strategies. So, do the fish-bone
assemblages from Rapa and Easter Island represent what could be considered a subtropical
fishing strategy?
Only a small number of subtropical islands have been investigated archaeologically, and
reports on fishing are pending for some. Norfolk Island displays a fishing strategy more like that
of New Zealand, focusing on a single family (Walter and Anderson 2001) rather than the more
generalised strategy common throughout the tropical zone. However, Norfolk is further outside
the tropical zone than Rapa and Easter Island, and coral formation is extremely limited. Norfolk
Island also lacks the bays characteristic of both Rapa and Easter Island, which no doubt affected
accessibility to fishing grounds. This suggests that the pattern seen at Rapa and Easter Island is
just one way in which prehistoric Polynesians adapted to subtropical islands, and that as in the
tropical zone proper, local ecology played a large part in shaping fishing strategies.
The Rapan fish-bone assemblages show a strategy closely aligned to tropical Polynesia, with
some interesting adaptations to local ecology and its constraints. A generalised strategy focused
on several families is evident, dominated by inshore taxa. The large numbers of Pomacentridae
and Chaetodontidae in the assemblages are at this stage unique to Rapa, but the use of more
elements for the identification of assemblages from other localities should reveal whether or not
this is an anomaly. In particular, it is interesting that the scarcity or absence of these families from
archaeological assemblages in Hawaii and Easter Island has been queried, as both places have
terra australis 37
Prehistoric fishing on Rapa Island
131
some similarities with Rapa in terms of coastal geomorphology and ecology. The dominance of
marine eels, in particular Muraenidae, is paralleled in other parts of the Pacific, particularly in
Easter Island. What at first appears to be a strange set of assemblages may yet turn out to be not
so strange after all.
References
Akimichi, T. 1986. Conservation of the sea: Satawal, Micronesia. In: Anderson, A. (ed), Traditional
Fishing in the Pacific. Pacific Anthropological Records 37:15–33.
Allen, M. 1992. Dynamic Landscapes and Human Subsistence: archaeological investigations on
Aitutaki island, southern Cook Islands. PhD thesis, University of Washington.
Allen, M. 2002. Resolving long-term change in Polynesian marine fisheries. Asian Perspectives
41(2):195–212.
Allen, M, Ladefoged, T. and Wall, J. 2001. Traditional Rotuman fishing in temporal and regional
context. International Journal of Osteoarchaeology 11:56–71.
Anderson, A.J. 1973. Archaeology and Behaviour: prehistoric subsistence behaviour at Black Rocks
peninsula, Palliser Bay. MA thesis, Anthropology Department, University of Otago.
Anderson, A.J. 1979. Prehistoric exploitation of marine resources at Black Rocks Point, Palliser Bay.
In: Leach, B.F and H.M. (eds), Prehistoric Man in Palliser Bay. National Museum of New Zealand
Bulletin 21:49–63.
Anderson, A.J. 1986. Introduction. In: Anderson, A. (ed), Traditional Fishing in the Pacific. Pacific
Anthropological Records 37:ix–xi.
Anderson, A.J. 1997. Uniformity and regional variation in marine fish catches from prehistoric New
Zealand. Asian Perspectives 36(1):1–26.
Anderson, A.J. 2001. No meat on that beautiful shore: The prehistoric abandonment of subtropical
Polynesian islands. International Journal of Osteoarchaeology 11:14–23.
Anderson, A.J. and Smith, I.W.G. 1999. Letter to the Editor. Archaeology in New Zealand 42:182–184.
Ayres, W.S. 1986. Easter Island subsistence. Journal de la Société des Ocieanistes 80:103–124.
Bilton, M. 2001. Taphonomic Bias in Pacific Ichthyoarchaeological Assemblages: A Marshall Islands
Example. Unpublished MA thesis, Department of Anthropology, University of Otago.
Butler, V.L. 1988. Lapita fishing strategies: The faunal evidence. In: Kirch, P.V. and Hunt, T.L. (eds),
The Archaeology of the Lapita Cultural Complex: A Critical Review. Thomas Burke Memorial
Washington State Museum Research Report, No. 5, pp. 99–116, Seattle.
Butler, V.L. 1994. Fish feeding behaviour and fish capture: The case for variation in Lapita fishing
strategies. Archaeology in Oceania 29:81–90.
Butler, V.L. 2001. Changing fish use on Mangaia, Southern Cook Islands: Resource depression and the
prey choice model. International Journal of Osteoarchaeology 11:88–100.
Cannon, D.Y. 1987. Marine Fish Osteology: A Manual for Archaeologists. Simon Fraser University
Publication 18, Department of Archaeology.
Davidson, J., Fraser, K., Leach, B.F. and Sinoto, Y.H. 1999. Prehistoric fishing at Hane, Ua Huka,
Marquesas Islands, French Polynesia. New Zealand Journal of Archaeology 21:5–28.
Davidson, J., Leach, F. and Sand, C. 2002. Three thousand years of fishing in New Caledonia and the
Loyalty Islands. In: Bedford, S., Sand, C. and Burley, D. (eds), Fifty Years in the Field: Essays in
Honour and Celebration of Richard Shutler Jr’s Archaeological Career. New Zealand Archaeological
Association Monograph 25:153–164.
Fraser, K. 1988. Fishing for Tuna in Pacific Prehistory. Unpublished MA thesis, Department of
Anthropology, University of Otago.
terra australis 37
132 Yolanda Vogel and Atholl Anderson
Fraser, K. 2001. Variation in tuna fish catches in Pacific prehistory. International Journal of
Osteoarchaeology 11:127–135.
Goto, A. 1986. Prehistoric Ecology and Economy of Fishing in Hawai’i: An Ethnoarchaeological
Approach. Unpublished PhD thesis, University of Hawai’i.
Goto, A. 1990. Prehistoric Hawaiian fishing lore: An integrated approach. Man and Culture in Oceania
6:1–34.
Grayson, D.K. 1984. Quantitative Zooarchaeology: Topics in the Analysis of Archaeological Faunas.
Academic Press, San Diego.
Hanson, F.A. 1970. Rapan Lifeways: Society and History on a Polynesian Island. Little, Brown and
Company Ltd, Boston.
Kennett, D., Anderson, A., Prebble, M., Conte, E. and Southon, J. 2006. Prehistoric human impacts
on Rapa, French Polynesia. Antiquity 80:340–354.
Kirch, P.V. and Dye, T. 1979. Ethnoarchaeology and the development of Polynesian fishing strategies.
Journal of the Polynesian Society 88(1):53–76.
Kirch, P.V. and Yen, D.E. 1982. Tikopia: The prehistory and ecology of a Polynesian outlier. Bernice P.
Bishop Museum Bulletin 238, Honolulu.
Leach, B.F. 1979. Excavations in the Washpool Valley, Palliser Bay. In: Leach, B.F. and H.M. (eds),
Prehistoric Man in Palliser Bay. National Museum of New Zealand Bulletin 21:67–136.
Leach, B.F. 1986. A method for the analysis of Pacific island fish bone assemblages and an associated
database management system. Journal of Archaeological Science 13:147–159.
Leach, B.F. and Boocock, A. 1993. Prehistoric Fish Catches in New Zealand. Oxford: British
Archaeological Reports, International Series 584.
Leach, B.F., Davidson, J. and Athens, J.S. 1996. Mass harvesting of fish in the waterways on Nan
Madol, Pohnpei, Micronesia. In: Davidson, J., Irwin, G., Leach, F., Pawley, A. and Brown, D.
(eds), Oceanic Culture History: Essays in Honour of Roger Green, pp. 319–341. New Zealand Journal
of Archaeology Special Publication, Wellington.
Leach, B.F., Davidson, J., Horwood, M. and Ottino, P. 1997. The fishermen of Anapua Rock Shelter,
Ua Pou, Marquesas Islands. Asian Perspectives 36(1):51–66.
Leach, B.F. and Davidson, J. 2000. Fishing: A neglected aspect of Oceanic economy. In: Anderson, A.
and Murray, T. (eds), Australian Archaeologist: collected papers in honour of Jim Allen, pp. 412–426.
Coombs Academic Publishing, The Australian National University, Canberra.
Masse, W.B. 1986. A millennium of fishing in the Palau Islands, Micronesia. In: Anderson, A.
(ed), Traditional Fishing in the Pacific: Ethnographical and Archaeological Papers from the 15th
Pacific Science Congress. Pacific Anthropological Records, No. 37, pp. 85–119. Department of
Anthropology, Bernice P. Bishop Museum, Honolulu.
Masse, W.B. 1989. The Archaeology and Ecology of Fishing in the Belau Islands. Unpublished PhD thesis,
University of Southern Illinois.
Nagaoka, L. 1993. Faunal assemblages from the To’aga site. In: Kirch, P.V. and Hunt, T.L. (eds),
The To’aga Site: Three Millennia of Polynesian Occupation in the Manu’a Islands, American
Samoa. Contributions of the University of California Archaeological Research Facility No. 51,
pp. 189–216. Berkley, California.
Rojo, A.L. 1991. A Dictionary of Evolutionary Fish Osteology. CRC Press, Boca Raton.
Rolett, B. 1998. Fishing strategies: Ethnographic observations and analysis of the excavated fish
remains. In: Rolett, B. Hanamiai: Prehistoric Colonisation and Cultural Change in the Marquesas
Islands (East Polynesia), vol. 81, pp. 118–146. Department of Anthropology and the Peabody
Museum, Yale University, New Haven, Connecticut.
Stokes, John F.G. n.d. Ethnology of Rapa. Manuscript on file. Bernice P. Bishop Museum, Honolulu.
Vogel, Y. 2005. Ika. Unpublished MA thesis, Department of Anthropology, University of Otago.
Walter, R. 1998. Fish and fishing. In: Walter, R. (ed), Anai’o: The Archaeology of a Fourteenth Century
terra australis 37
Prehistoric fishing on Rapa Island
133
Polynesian Community in the Cook Islands. New Zealand Archaeological Association Monograph
22, pp. 64–73.
Walter, R. 1989. Lapita fishing strategies: A review of the archaeological and linguistic evidence. Pacific
Studies 13:127–149.
Walter, R. and Anderson, A. 2001. Fishbone from the Emily Bay settlement site, Norfolk Island. In:
Anderson, A. and White, P. (eds), The Prehistoric Archaeology of Norfolk Island, Southwest Pacific.
Records of the Australian Museum, Supplement 27, pp. 101–108. Australian Museum, Sydney.
Walter, R. and Anderson, A. 2002. Marine fishbone. In: Walter, R. and Anderson, A. (eds),
The Archaeology of Niue Island, West Polynesia. Bishop Museum Bulletin in Anthropology 10,
pp. 94–102. Bishop Museum Press, Honolulu.
Walter, R. and Weisler, M. 2002. Late prehistoric fishing adaptations at Kawakiu Nui, Moloka’i.
Hawaiian Archaeology 8:42–61.
Walter, R., Weisler, M. and Smith, I. 1996. The Pacific fish bone reference collection at the University
of Otago. Archaeology in New Zealand 39:200–212.
Weisler, M. 1993. The importance of fish otoliths in Pacific Island archaeofaunal analysis. New Zealand
Journal of Archaeology 15:131–159.
Weisler, M. 1999. Letter to the Editor. Archaeology in New Zealand 42:184–186.
Weisler, M. 2001. On the Margins of Sustainability: Prehistoric Settlement of Utrok Atoll, Northern
Marshall Islands. British Archaeological Reports International Series 967, Oxford.
Weisler, M., Lalas, C. and Rivett, P. 1998. New fish records from an Archaic midden, South Island.
Archaeology in New Zealand 42:37–43.
Wheeler, A. and Jones, A.K.G. 1989. Fishes. Cambridge University Press, Cambridge.
terra australis 37
terra australis 37
8
The Tangarutu invertebrate fauna
Katherine Szabó
Centre for Archaeological Science, University of Wollongong; Department of Archaeology and Natural
History, The Australian National University, Canberra, Australia, Kat.Szabo1@gmail.com
Atholl Anderson
Department of Archaeology and Natural History, The Australian National University
Introduction
The island of Rapa presents an interesting lens through which to investigate human decisionmaking and resource-use patterns in a marginal environment. In addition to being small and
isolated, Rapa is climatically marginal, being positioned on the southern fringe of the tropical
Indo–West Pacific marine province. Most obviously, this geographical situation translates to
restricted species diversity, with a great many common tropical taxa not able to survive the
conditions. However, as pointed out by Preece (1995:345), it would be a mistake to see the
marine fauna of marginal Polynesian islands as simply an impoverished subset of the tropical
Indo–West Pacific community. Indeed, Paulay and Spencer (1988, in Irving 1995:321) suggest
that there is a belt of islands in the central-eastern Pacific (including Easter Island, the Pitcairn
Islands, Kermadecs, Rapa and Lord Howe) that have a littoral fauna specifically adapted to
cooler conditions. Such species include a number of endemics, as well as species restricted solely
to the islands listed above. The dominant mollusc within the Tangarutu assemblage – Nerita
morio – is one such species (see below for discussion).
While this distinctive intermediate fauna may suggest a reasonable degree of faunal stability
through time, studies done at other islands within this zone (notably the Pitcairn Group) have
demonstrated otherwise. A comparison of Pleistocene subfossil assemblages and extant littoral
communities on Henderson Island showed considerable turnover (Paulay and Spencer in Irving
1995:321). Such turnovers can be explained in different ways, including by relatively small
population sizes, species existing at the limits of their environmental tolerances, and isolation
limiting recruitment of individuals from other islands. While only providing a small slice of
Rapa’s history, the Tangarutu marine invertebrate assemblage presents an opportunity to assess
human subsistence practices in an environment that is therefore not only impoverished, but
dynamic in its constituents.
The Tangarutu site is adjacent to Anarua Bay, which contains a variety of littoral invertebratebearing habitats. With tall cliffs at the northern and southern ends, the bay is west-facing and
terra australis 37
136 Katherine Szabó and Atholl Anderson
thus usually sheltered from the prevailing southeasterlies. The prevailing winter winds in Rapa
are westerlies (Irving 1995:321), meaning Anarua Bay would see high wave action for a period
each year. Within the bay itself are rocky intertidal platforms, sandy patches and basalt gravel
zones, providing a variety of niches for infaunal (within the substrate) and epifaunal (atop the
substrate) littoral marine life. The archaeological deposits excavated from Tangarutu contained
small but consistent quantities of invertebrate remains, representing a varied mix of mollusc,
crustacean and echinoid species. This chapter presents the results of analysis, and considerations
of subsistence, palaeoenvironment and gathering strategies.
Methodology
All shell, crustacean and echinoderm remains from the E1 and E2 squares of Tangarutu were
analysed, with all molluscan remains identified to the lowest possible taxonomic level – generally
species. Molluscan taxonomy follows Dharma (2005), with up-to-date taxonomy drawn from
the OBIS online database (http://clade.ansp.org/obis/). Crustacean remains included crab
exoskeleton fragments, as well as ‘goose barnacle’ (Lepas anatifera) plates. All crab remains were
quantified as a single taxonomic category without any attempt at identification, while Lepas
anatifera remains were quantified separately. Echinoderms were represented by teeth, spines and
test fragments of a single taxon, identified as Diadema cf. setosum (but see discussion of urchin
identification below).
Remains were highly fragmented at all levels, although otherwise very well preserved. While
both MNI and NISP values, along with weights, were recorded, the NISP values will be reported
here. This is not only due to the small sample sizes and consistent levels of breakage across taxa,
but also because the calculation of MNI values relies on detailed stratigraphic understanding
(Grayson 1984). As reported in Chapter 3, excavation at Tangarutu proceeded in 10 cm spits,
and while stratigraphic layers were also recorded, the excavators noted complexity that went
beyond the three recorded cultural layers (Vogel 2005:24). Given this, it was decided to report
the results using the recorded spits rather than layers.
Results
Forty molluscan taxa were identified, supplemented by crustacean remains in the form of
Lepas anatifera plates and abundant crab exoskeleton fragments. These classes of remains will
be discussed individually below, together with a discussion of spatial and temporal patterning
within the studied sample, insights into local ecology and gathering strategies, and the role of
various marine invertebrates within the overall Tangarutu assemblage.
Despite being adjoining squares, there are distinct differences between squares E1 and E2.
Firstly, there is a difference in gross sample size, with the E1 total (NISP) for molluscan remains
being 1280, and E2 874 (i.e. excluding crab, urchin, barnacle etc, shown in Table 8.1). However,
differences in species representation do not simply relate to the effect of variable sample sizes. A
number of species are more strongly represented in the smaller E2 sample, including the most
common mollusc Nerita morio (E1 n=376; E2 n=398), as well as Clypeomorus batillariaeformis
(E1 n=106; E2 n=123), and less common taxa such as Drupa morum (E1 n=0; E2 n=5). There
are also species clustered in E1 that are correspondingly underrepresented in E2, including the
limpet Cellana tahitensis1 (E1 n=169; E2 n=54), polyplacophoran (chiton) plates (E1 n=91; E2
n=24), and Trochus cf. radiatus (E1 n=81; E2 n=26). Taxa represented by 10 or more fragments
in E1 and E2 are graphed in figures 8.1 and 8.2 respectively. Table 8.1 shows overall NISP values
for all identified marine invertebrate fauna for both squares E1 and E2.
terra australis 37
The Tangarutu invertebrate fauna
137
What these clusters of different taxa appear to represent are episodes of dumping from a
specific series of gathering and/or processing events. Figure 8.3 clearly shows that these speciesspecific dumps are not directly correlated with vertical stratigraphy. For example, Nerita morio
fragments are entirely restricted to Square A2 in Spit 1, while there are much higher numbers
of Clypeomorus batillariaeformis within Square A2 over spits 5, 6 and 7. The complexity of the
Tangarutu stratigraphy has been commented on in Chapter 3, and these few examples from
the distributions of molluscan taxa amply demonstrate intricate horizontal as well as vertical
patterning.
Although the Tangarutu marine invertebrate sample was collected in arbitrary spits rather
than stratigraphic layers, considerable vertical patterning is evident. Most obviously, this
patterning can be seen in gross overall quantities of molluscan, crustacean and urchin remains.
While the pattern is slightly skewed by the conflation of spits 10–13, it can still be seen that
the collection of molluscs dominates over crustacean or urchin exploitation in the lower three
spits. Furthermore, this dumping of molluscan remains (see Figure 8.4) is restricted for the
Figure 8.1. Invertebrate taxa represented
by 10 or more fragments (NISP) in Square
E1, Tangarutu.
Figure 8.2. Invertebrate taxa represented by 10 or more fragments (NISP) in Square E2, Tangarutu.
terra australis 37
138 Katherine Szabó and Atholl Anderson
Table 8.1. Invertebrate taxa identified within the E1 and E2 samples from Tangarutu. NISP values given represent the
total number of fragments identified for each square. ^Indicates absence.
Taxon
E1
E2
Habitat
Crab
658
255
variable
Nerita morio
376
398
intertidal rocks
Urchin (Diadema sp.)
363
197
inter/sub-tidal
Cellana tahitensis
169
54
intertidal rocks
Gafrarium pectinatum
169
100
sandy intertidal
Clypeomorus batillariaeformis
106
123
intertidal rocks
Polyplacophora (“Chiton”)
91
24
intertidal rocks
Trochus cf. radiatus
81
26
intertidal rocks
Tellina palatum
38
26
sandy intertidal
Thais sp.
37
21
intertidal rocks
Cypraea caputserpentis
35
7
intertidal rocks
Cypraea spp.
33
4
variable
Modiolus sp.
21
10
Nerita spp. Opercula
18
15
Patella flexuosa
15
2
intertidal rocks
variable
intertidal rocks
Drupa spp.
15
7
intertidal rocks
Lepas anatifera
13
4
pelagic
Chama pacifica
12
1
inter/sub-tidal rocks
Tonna sp.
10
13
sandy inter/sub-tidal
Conus spp.
9
^
variable
Nerita plicata
8
1
upper intertidal rocks
Drupa ricinus
5
^
intertidal rocks
Pyrene sp.
5
^
intertidal sand/rubble
Drupa morum
^
5
intertidal rocks
Cardita variegata
4
1
intertidal rocks
Tellina scobinata
3
^
sandy intertidal
Antigona (Periglypta) reticulata
3
^
sandy intertidal
Drupa rubisidaea
^
3
intertidal rocks
Astraea rhodostoma
2
^
intertidal rocks
Phos sp.
2
2
intertidal rocks
Nassarius sp.
2
^
variable
Pinctada margaritifera
2
^
sub-tidal
Astraea sp.
1
^
intertidal rocks
Nerita costata
1
13
upper intertidal rocks
Planaxis sulcatus
1
^
upper intertidal rocks
Cypraea cribraria
1
^
sandy intertidal
Strombus sp.
1
^
sandy intertidal
Siphonaria sp.
1
15
intertidal rocks
Corbula sp.
1
^
sandy intertidal
Saccostrea cucullata
1
^
intertidal rocks
Codakia tigerina
1
^
sandy intertidal
Turbo cinereus
^
1
intertidal rocks
Hipponix conicus
^
1
on other shells
^
1
sandy intertidal
2335
1342
Fragum sp.
TOTAL
terra australis 37
The Tangarutu invertebrate fauna
139
Figure 8.3. The six highest-ranking (NISP) invertebrate taxa graphed by square and spit.
most part to Square E1. Indeed Spit 1 and Spit 6 represent the only two time brackets where
molluscan remains are primarily dumped in Square E2. Low levels of Diadema cf. setosum urchin
remains are present throughout the sequence, but urchin only predominates over molluscan and
crustacean remains in Spit 2, where it is strongly represented in both squares E1 and E2. There
are also notable levels of urchin remains in Spit 17, Square E1. Crustacean remains appear in
highly variable numbers, but are well represented towards the lower part of the sequence (spits
16 and 17, E1), in the mid-sequence (spits 10–13), and towards the upper part of the sequence
(Spit 4 in Square E1 and Spit 1 in Square E2).
Within the molluscan sample, the most obvious stratigraphic difference in species
representation can be seen in the large dump of Nerita morio in Spit 1 (largely Square E2) (see
Figure 8.3). Nerita morio is present in low levels throughout the sequence, but has a notably
strong presence in spits 16–18 (Square E1) at the lowest part of the Tangarutu sequence.
Conversely, the soft-shore bivalve Gafrarium pectinatum is represented most strongly in spits
3–6 where representation of Nerita morio is fairly weak. While this could feasibly be construed
as a shift away from a focus on the hard shore to the soft shore, the occurrence of other softshore species is scattered and diffuse, with a range of species contributing only a few individuals
(e.g. Tellina palatum, Antigona reticulata). As such, any pattern across soft-shore taxa is difficult
to pin down (see further discussion in Chapter 9).
Within the hard-shore intertidal category, there is clearly a number of complex relationships
between taxa, with both the limpet Cellana tahitensis and the horn shell Clypeomorus
batillariaeformis showing different patterns of occurrence to Nerita morio. Although virtually
restricted to Square E1, Cellana tahitensis is the dominant mollusc from Spit 7 to Spit 9, with
strong occurrences also in spits 10–13. Apart from a small accumulation in Spit 2 (Square
E2), Cellana tahitensis numbers are otherwise negligible. The small gastropod Clypeomorus
terra australis 37
140 Katherine Szabó and Atholl Anderson
Figure 8.4. Occurrences of molluscan, crustacean and urchin remains throughout the Tangarutu sequence (NISP
values).
batillariaeformis has two patches of fairly strong occurrence in the mid-sequence; the first in
spits 4–7 (largely in Square E2), and the second lower down, in spits 10–16, oscillating in
concentration between squares E1 and E2.
A gross environmental breakdown shows that molluscan taxa derive from two environmental
zones: the rocky/reef-flat intertidal niche and the sandy intertidal zone (see figures 8.5 and 8.6).
However, the foregoing discussion makes it clear that there are more complex environmental
and/or gathering issues at play than simply a focus on either hard- or soft-shore resources.
Nerita morio dominates the Tangarutu molluscan sample as a whole, but as mentioned
above, its importance within the overall invertebrate sample fluctuates markedly through time.
Nerita morio is one of a group of black nerites that inhabits the temperate and sub-tropical
fringes of the Indo–West Pacific, joined by the Australian Nerita atramentosa, the Australian,
Lord Howe, Kermadecs, Pitcairn Islands and northern New Zealand Nerita melanotragus, and
the Easter Island endemic Nerita lirellata (Spencer et al. 2007). Nerita morio can be found on
Easter and Pitcairn islands, as well as the Australs, and is fairly abundant within its restricted
geographical range (T. Eichhorst pers. comm.). An algal grazer, it lives under rocks and in tidal
pools of the intertidal zone (T. Eichhorst pers. comm.).
Little information is available on the limpet Cellana tahitensis, but it likely shares many
features with other species in this wide-ranging genus. Cellana spp. limpets are algal grazers,
and nearly all eschew limestone, being found primarily on intertidal basalt surfaces (Preece
1995:346). Clypeomorus batillariaeformis 2 tends to inhabit the upper mid-littoral zone among
terra australis 37
The Tangarutu invertebrate fauna
Figure 8.5. Graph to show the broad ecological niches from
which the E1 molluscan samples derive.
141
Figure 8.6. Graph to show the broad ecological niches from
which the E2 molluscan samples derive.
rocks or in crevices, being most active during nocturnal low tides (Ayal and Safriel 1983:40). C.
batillariaeformis can tolerate some degree of wave action and sand movement (Ayal and Safriel
1983:40), but personal observations of this species on Guam suggest its preference for calmer
waters (KS pers. obs. 2007–2008, Tapa’chang Beach, Guam).
The dominance of gastropods over bivalves in the Tangarutu sample is to be expected given
the isolation and ecology of Rapa. As Preece (1995:343) points out, insular marine molluscan
assemblages tend to be dominated by gastropods, and small islands appear to skew faunal
compositions even further in this direction. The presumed reason for this is the frequent paucity
of soft littoral substrates on small islands, as opposed the variety of hard-shore habitats that
accommodate a diversity of gastropods (Preece 1995:343). Dispersal mechanisms and food
availability may also play a part in this (Vermeij in Preece 1995:344).
All urchin remains recovered from the Tangarutu deposits were identified as belonging to
the species Diadema cf. setosum. The closely related Diadema savignyi has been identified for
Pitcairn Island (Paulay 1989), although greater caution in species attribution for Pitcairn is
shown by Irving (1995:314). It is possible that the Tangarutu specimens could belong to either
species, or indeed a mix of the two sympatric Diadema taxa. Identification keys are based on
discrete colour and anatomical differences (Muthiga 2003), neither of which could be utilised
in the analysis of the Tangarutu assemblage.
Diadema spp. urchins have a very fragile test, with long, hollow black spines (Miskelly
2002:22), but one of the most notable features of the Tangarutu Diadema spp. remains is
the low numbers of spine and test fragments (Figure 8.4). Instead, urchins were dominantly
represented by ‘teeth’ or other components of the feeding apparatus known as the ‘Aristotle’s
lantern’. This pattern would suggest that primary processing of urchin remains did not happen
at the rockshelter itself, but at or near the point of collection. Given the length of Diadema spp.
primary spines (up to 30cm), which are capable of causing a painful injury (Miskelly 2002:22),
there is good reason to remove the spines before transport. Thus, the Tangarutu urchin remains
would indicate that, generally, the flesh and mouthparts alone were transported back to the site.
Diadema setosum is an algal grazer, which prefers a substrate of coarse sand to rubble and
little in the way of macrophytes such as Sargassum spp. (Dumas et al. 2007:96). It has been
found to avoid finer sediments, such as fine sand or mud, and tends to shun more architecturally
intricate substrates such as areas of complex or branching coral cover (Dumas et al. 2007:96).
Observations by KS of Diadema setosum in eastern Indonesia (Pulau Kanawa, western Flores),
terra australis 37
142 Katherine Szabó and Atholl Anderson
suggest that its gregarious nature combined with its occurrence in shallow, coarse sand and
rubble areas makes it a highly visible target. Despite its nocturnal feeding patterns, KS has also
seen high levels of activity in the late afternoon. The clear abundance of this species in close
proximity to the site, coupled with the abundance of algal-grazing molluscs, indicates extensive
algal-bearing habitats in Anarua Bay.
A small number of Lepas anatifera barnacle plates were identified within both the E1 and
E2 samples. Goose barnacles attach themselves to floating objects such as driftwood by a flexible
stalk (Gunson 1983:44; Debelius 2001:310) and, as such, are only found in the intertidal zone
or on beaches when washed up with ocean debris. Thus, rather than representing a food source,
they indicate the presence of driftwood introduced to the site. While numbers are too small to
make any statistically meaningful statements, it is interesting to note that in both squares E1
and E2, occurrences of Lepas anatifera plates stratigraphically match spikes in the abundance of
Clypeomorus batillariaeformis. This may indicate periods of slightly greater exposure at Anarua
Bay or the introduction of material collected in the winter months.
Tangarutu invertebrate remains in context
When dealing with a culturally accumulated faunal assemblage, it is difficult to determine
whether patterning is due to human agency, the nature of available resources, or a complex
mixture of the two. This issue is surely compounded on Rapa, where already sensitive and
unstable ecological communities had to incorporate a new member – human colonists. Teasing
apart the direct and indirect effects of human intervention in an environment that is apparently
so precariously balanced without humans is unlikely to move past reasoned speculation.
That said, arguments for human impact through over-predation have been made for other
zones of marginal East Polynesia. Fluctuating frequencies of various molluscan taxa were also
detected by Weisler (1995) for Henderson Island in the Pitcairn Group.3 At Site HEN-10, Weisler
(1995:397, 399) noted that cerithid gastropods (identified mainly as Cerithium tuberculiferum)
completely disappear from the record after a period of heavy exploitation. Although Weisler
(1995:397, 399) cites this as an example of human over-predation, we would be reluctant to
apply such an interpretation to the fluctuating representations of various taxa at Tangarutu,
given the ecological complexities.
There can be little doubt that human predation impacted on the local littoral fauna on
Rapa, but whether humans directly caused the local extinction or extirpation of particular
species at Anarua Bay is more difficult to establish. In addition, while species such as
Clypeomorus batillariaeformis are completely absent from the uppermost spits of Tangarutu,
other gastropods such as Nassarius sp. and Tonna sp. are completely absent from the lower
sections of the stratigraphy. On present evidence, such patterns are just as likely to represent a
littoral community in flux as human gathering patterns and/or associated over-predation.
One of the major differences between Weisler’s (1995) Henderson Island sites and the
Tangarutu sample is the contribution made by urchins and crabs at the latter and their complete
absence within the former. Paulay (l989:4) observed in his 1987 survey of Henderson Island
invertebrate resources that Diadema sp. urchins occurred ‘spine-to-spine’. In his assessment
of coral reef biotopes on Henderson Island, Irving (1995:317) cites urchins as ‘the most
conspicuous invertebrates’. He goes on to state that molluscs and crustaceans were difficult to
locate on Henderson’s reefs, ‘with the presence of many only being revealed by their dry remains
on the strandline’ (Irving 1995:317). Whether urchins were as abundant around Henderson
Island in the past, or whether the archaeological pattern provides evidence of avoidance, is a
question that remains unanswered. Urchin exploitation was clearly important to the inhabitants
terra australis 37
The Tangarutu invertebrate fauna
143
of Tangarutu, and although the level of exploitation fluctuated, it was clearly never heavy enough
to wipe out local populations. It is more difficult to be categorical about the status of crab, as this
category represents a mixture of different species that may show different temporal patterning
with further analysis. Nevertheless, it is clear that decapod crustaceans were exploited at steady
levels throughout the utilisation of the Tangarutu shelter.
Looking at the overall faunal exploitation patterns in the Henderson Island and Tangarutu
assemblages, the differences appear greater than the similarities. Fishing is undoubtedly
important in both locales, and it was supplemented by consistent levels of mollusc exploitation.
However, there is greater variety of invertebrate exploitation at Tangarutu, with urchins and
crabs making a solid contribution to the overall subsistence picture through time. It is possible
that this diversification in littoral exploitation on Rapa was a response to the lack of pig and
turtle, which evidently were present on Henderson Island. A greater reliance on native resources
can also be seen in the use of Aleurites moluccana endocarp and human bone for fishhook
manufacture on Rapa (see Chapter 9), while more traditional shell resources were available (or
acquired) for fishhook (Pinctada sp.) and adze (Tridacna sp.) manufacture on Henderson Island.
There are obvious difficulties associated with living on small, isolated, impoverished and
climatically marginal islands. Given this, it is instructive to note the differences in littoral
exploitation patterns between Rapa and Henderson, islands which are so often grouped together
for the purposes of ecological discussion (e.g. Paulay 1989:12 and references therein). Both
the differences between human subsistence behaviour on the two islands and the fluctuating
importance of various marine invertebrate resources within the Tangarutu sample suggest that
flexibility and the ability to resource-switch are important factors for survival. Indeed, there can
be no ‘one-size-fits-all’ strategy in a fluid environment. The fact that domesticated animals were
not introduced to Rapa in prehistory and the fact that the prehistoric inhabitants of Henderson
Island were not capitalising on crustacean and echinoderm resources amply demonstrate that
even when resources are scarce, choices exist. A longer-range view of marine invertebrate
populations on Rapa, including palaeontological information and modern survey data, may
shed more light on fluctuations in community structure and bring into clearer focus the impact
of human predation on Rapa’s coastal resources.
Notes
1.
2.
3.
Cellana tahitensis is the current valid name for Cellana (=Patella) taitensis more commonly seen in
regional literature (OBIS database, record 35389).
Ayal and Safriel refer to this species by its junior synonym Clypeomorus moniliferum. Another
junior synonym commonly seen in the literature is Clypeomorus moniliferus.
It would appear from the molluscan species list that Henderson and Rapa are not directly
comparable, with Henderson having some ‘classic’ Indo–West Pacific taxa such as Turbo spp. and
Tridacna sp. which are largely absent from Rapa. The diminutive Turbo cinereus is the only species
of Turbo recorded for the Tangarutu sample.
Acknowledgements
Thank you to Tom Eichhorst for (once again) so generously sharing his vast knowledge of neritids, and
to Yolanda Vogel for unpublished information and discussion.
terra australis 37
144 Katherine Szabó and Atholl Anderson
References
Ayal, Y. and Safriel, Y.N. 1983. Does a suitable habitat guarantee successful colonisation? Journal of
Biogeography 10:37–46.
Debelius, H. 2001. Crustacea Guide of the World. Conchbooks, Hackenheim.
Dharma, B. 2005. Recent and Fossil Indonesian Shells. Conchbooks, Hackenheim.
Dumas, P., Kulbicki, M., Chifflet, S., Fichez, R. and Ferraris, J. 2007. Environmental factors
influencing urchin spatial distributions on disturbed coral reefs (New Caledonia, South Pacific).
Journal of Experimental Marine Biology and Ecology 344:88–100.
Grayson, D.K. 1984. Quantitative Zooarchaeology – Topics in the analysis of archaeological faunas.
Academic Press, Orlando, Fl, and London.
Gunson, D. 1983. Collins Guide to the New Zealand Seashore. Collins, Auckland.
Irving, R.A. 1995. Near-shore bathymetry and reef biotopes of Henderson Island, Pitcairn Group.
Biological Journal of the Linnean Society 56:309–324.
Miskelly, A. 2002. Sea Urchins of Australia and the Indo-Pacific. Capricornia Publications, Lindfield,
NSW.
Muthiga, N.A. 2003. Coexistence and reproductive isolation of the sympatric echinoids Diadema
savignyi Michelin and Diadema setosum (Leske) on Kenyan coral reefs. Marine Biology 143:669–
677.
OBIS Indo-Pacific Molluscan Database. 2006. Cellana tahitensis species listing. URL: http://clade.ansp.
org/obis/search.php/35389 Accessed 14 June 2008.
Paulay, G. 1989. Marine invertebrates of the Pitcairn Islands: species composition and biogeography of
corals, molluscs and echinoderms. Atoll Research Bulletin 326:1–28.
Preece, R.C. 1995. The composition and relationships of the marine molluscan fauna of the Pitcairn
Islands. Biological Journal of the Linnean Society 56:339–358.
Spencer, H.G., Waters, J.M. and Eichhorst, T.E. 2007. Taxonomy and nomenclature of black nerites
(Gastropoda: Neritimorpha: Nerita) from the South Pacific. Invertebrate Systematics 21:229–237.
Vogel, Y. 2005. Ika. Unpublished MA thesis, University of Otago, Dunedin.
Weisler, M.I. 1995. Henderson Island prehistory: colonisation and extinction on a remote Polynesian
island. Biological Journal of the Linnean Society 56:377–404.
terra australis 37
9
Marine resource exploitation on
Rapa Island
Archaeology, material culture and ethnography
Katherine Szabó
Centre for Archaeological Science, University of Wollongong, NSW, Australia, Kat.Szabo1@gmail.com
Yolanda Vogel
Department of Anthropology, University of Otago
Atholl Anderson
Archaeology and Natural History, The Australian National University
Introduction
As Rapa lacked the usual suite of Polynesian domesticated animals, it is not surprising that
evidence for marine fishing and marine exploitation in general is strong. However, the discussion
of fishing techniques and broader aquatic resource exploitation must be placed within the
unique environmental context of the island; no straightforward transference of traditions or
interpretations in other parts of Polynesia will suffice to explain the patterns seen here. The
cultural adaptations that formed on Rapa are exemplified by a remarkable assemblage of very
small fish hooks produced in candlenut endocarp that was recovered from Tangarutu. In the
absence of tropical coral reef littoral environments, species of shell so important elsewhere for
fish-hook manufacture, such as the pearl oyster Pinctada margaritifera, were not locally available
on Rapa1 and creative new solutions had to be found. What follows here, then, is an investigation
into Rapan flexibility in modifying cultural techniques and practices to the limitations and
idiosyncrasies of the environment in which they lived.
After a description of the remarkably preserved Tangarutu fish hooks, the assemblage
will be considered within the context of both Rapan environments and generic Polynesian
fishing traditions. The Rapan fish hooks are argued to represent, among other things, the
creative confluence of traditional practice and material constraint. Neither cultural mores nor
environmental context is seen as a determining factor in its own right. Rather, both are seen to
inform each other in generating an inventive solution and, with it, a new trajectory in Polynesian
fishing technologies.
terra australis 37
146 Katherine Szabó, Yolanda Vogel and Atholl Anderson
The fish-hook assemblage
The fish-hook assemblage from Tangarutu consists of 15 complete hooks and hook fragments. Ten
of these have been manufactured from the tough endocarp of the candlenut (Aleurites moluccana),
with the remaining five produced in bone. Within the candlenut fish-hook assemblage, three
specimens are unfinished, thereby allowing additional insights into manufacturing procedures.
The candlenut fish hooks also preserve evidence of the Pandanus carrier(s) in which they were
wrapped, as well as the line fibres lashed below the protruding knobs at the heads of the hooks.
The candlenut and bone fish hooks will be discussed in turn.
Candlenut fish hooks
Candlenut seems an unlikely raw material in which to fashion fish hooks, although the use
of coconut (Cocos nucifera) endocarp is widespread (e.g. see Anell 1955:94–5, 98, 102). The
coconut palm grows poorly on Rapa, and perhaps grew not at all in the prehistoric past, while
Aleurites moluccana is common. The use of candlenut endocarp, however, restricts manufacture
to very small fish hooks. The average size of the finished hooks at Tangarutu, measuring from the
head to the bend, is 10.04 mm long, with a standard deviation of 0.43 mm. Full measurements
are given in Table 9.1. There is a standardised plan to the hook morphology, despite some
variation (see Figure 9.1), with all hooks, bar the one broken example, showing an incurving
point leg, a thickened bend area, and a protruding knob at the head to keep the line in place.
All candlenut hooks can be classified as ‘rotating’; largely by virtue of the abrupt incurve of
the terminal end of the point leg – similar to the fong hooks of Tobi described by Johannes
(1981:117–118). This hook morphology helps to keep the fish on the hook and/or hold the bait
on the hook (Johannes 1981:117).
Table 9.1. Maximum dimensions and provenance of all Tangarutu complete and fragmented fish hooks.
Fishhook description
finished candlenut hook
Length (mm)
Width (mm)
Provenance
1
10.04
7.25
E1, Spit 4
2
10.31
9.20
E1, Spit 4
3
9.21
6.85 (broken)
E2, Spit 2
4
9.98
7.55
E1, Spit 4
5
10.50
7.67
East E1, Spit 4
6
10.37
7.68
East E1, Spit 4
7
9.87
7.12
East E1, Spit 14
1
10.08
7.65
E1, Spit 3
2
13.17
9.09
East E2, Spit 1
3
6.39 (broken)
8.94
East E1, Spit 1
shank only
1
29.67
7.28
West R1, Spit 4
shank and bend
2
29.07
3.65
Unstratified
point leg
3
8.78
2.27
East E1, Spit 5
bend
4
22.05
7.07
E2, Spit 2
shank/partial bend
5
33.44
7.70
E1, Spit 4
candlenut hook preform
bone hook*
*None of the bone hooks are complete, so these values represent maximum fragment dimensions. Maximum widths represent maximum fragment
widths rather than hook widths.
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
147
Figure 9.1. The Tangarutu candlenut
finished fish-hook assemblage. Scale
in mm. Photograph D. Boyd.
The presence of the three preforms (see Figure 9.2) allows insight into how these hooks were
produced. The blank was obtained by cutting a number of straight edges to generate the general
outer form of the hook. A hole was then drilled through the blank to initiate the formation
of the inner shank, bend and point surfaces (seen most clearly in Figure 9.2c). This hole was
then widened through abrasion, and the outer periphery of the fish hook was abraded to the
desired finished shape. There are clear parallels here with the production of fish hooks in other
materials – most notably shell and bone – and this will be discussed further below. The preforms
indicate that the head was one of the last pieces of the hook to be shaped. The general abrasion
of finished hook surfaces, as well as the presence of the remnants of line lashing on all but one
finished specimen (see Figure 9.3), indicate that the hooks were used. Fragments of a woven
Pandanus case also adhere to all but one of the finished specimens, and while the hooks could
have been wrapped individually, it is conceivable they could have been wrapped together in a
woven fish-hook ‘kit’, as seen in Tahiti (see Anell 1955:Plate 3). In this respect, it is notable that
the one candlenut fish hook that does not have evidence of lashing or an outer wrapping was
recovered from Spit 14 of the Tangarutu excavations, whereas the other finished examples were
all recovered in close proximity to one another (refer to Table 9.1).
Bone fish hooks
All the bone fish hooks were manufactured from terrestrial mammal cortical bone, and given the
size, density and visibly homogenous texture, as well as the lack of domesticated mammals on
Rapa, it is most likely that they were produced from human bone (see Figure 9.4). There are no
terra australis 37
148 Katherine Szabó, Yolanda Vogel and Atholl Anderson
Figure 9.2. The Tangarutu candlenut
fish-hook preforms. Scale in mm.
Photograph D. Boyd.
Figure 9.3. Finished candlenut fish-hook
head and upper shank showing remnants
of cordage line attachment lashing; 20x
magnification. Photograph P. Piper.
complete examples and thus it cannot be determined whether the bone fish hooks were rotating
or jabbing types. Where detectable, features such as an incurving point leg and a projectingknob form of line attachment match those seen in the candlenut fish hooks. No remnants of
either lashing or casing are evident.
Discussion
The stratigraphic placement of the candlenut fish hooks indicates that these were deposited,
and probably made, late in the sequence represented at Tangarutu (ca. 1450–1600 AD, based
on associated, calibrated radiocarbon dates published in Kennett et al. 2006). This places their
manufacture considerably earlier than the first European contact with Rapans in 1792. There is,
however, evidence that they continued in production until after European contact.
The unpublished manuscript of Stokes (n.d.), produced in 1920–21, makes mention of
small fish hooks produced in candlenut (tuitui), although the discussion of them is based on
the memory of four older men, none of whom had actually seen such hooks being made. Two
forms of candlenut hook were stated as having been manufactured in the past, one from the
median section of the endocarp, and one cut ‘from the side of the nut shell’ (Stokes n.d.). The
examples recorded from Tangarutu fall into the latter category. Stokes further points out that the
hook produced from the side of the endocarp would have been much stronger than the ‘flimsy’
version produced from the cross-section (Stokes n.d.).
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
149
Figure 9.4. The Tangarutu bone
fish-hook assemblage. Scale in mm.
Photograph D. Boyd.
Stokes’ discussion of candlenut fish hooks demonstrates that this technological innovation
evidenced at Tangarutu was not simply a matter of shortlived experimentation, but a distinctively
Rapan solution that continued in use for some centuries. As pointed out by Stokes (n.d.),
the use of candlenut endocarp as a raw material for fish-hook manufacture is not recorded
elsewhere in Polynesia. However, the evidence contained in the preforms sheds light on how this
solution might have come about. The cutting of a blank out of a convex surface, the drilling and
subsequent filing of the inner hook form, the abrasion of the outer perimeter and the formation
of a knob-shaped head have clear parallels in fish-hook production in bone and shell across
Polynesia.
While shell fish-hook blanks were often chipped and filed into shape, rather than sawn
(e.g. Kirch and Yen 1982:239; Szabó 2007), bone fish-hook blanks were more commonly
generated through sawing (Emory, Bonk and Sinoto 1968:Plate 5). Modification of the inner
surfaces of shell fish hooks was initiated through either the drilling of a hole or the creation and
enlargement of a notch (Szabó 2007; Kirch and Yen 1982), while the drilling of a central hole
in the tab appears to be more standard in bone fish-hook production (e.g. see Emory, Bonk and
Sinoto 1968:Plates 4 and 5). Thus, while the use of candlenut endocarp as a raw material can be
seen as a technological innovation, the techniques applied are firmly rooted in other Polynesian
(and arguably Micronesian) fish-hook manufacturing traditions.
It is also noteworthy that the only other raw material used for the production of the Tangarutu
fish hooks is bone. There is no record of any of the standard Polynesian domesticates (pig, dog
or chicken) on Rapa, either within the archaeology or ethnographically (Buck 1954:181, 320)
and the bone hooks from Tangarutu are probably manufactured from human bone. Where
terra australis 37
150 Katherine Szabó, Yolanda Vogel and Atholl Anderson
identifiable, longbone has been used, rather than cranial pieces (see Skinner 1942:217). While
the use of human bone for fish-hook production has precedents across Polynesia, its use is
commonly bound up in symbolic aspects of war and revenge (e.g. see Best 1929:36), rather than
in serving a purely practical need. In the case of Rapa, it is unclear whether the use of human
bone for fish hooks had another level of meaning beyond a matter-of-fact answer to a shortage
of traditional raw materials. Stokes (n.d.), at least, does not record symbolic meanings beyond
practical application. He does, however, posit that materials such as turtle shell and sperm-whale
teeth would have been used in the past, despite local assertions to the contrary. The Tangarutu
fish-hook assemblage reinforces his Rapan informants’ views that candlenut endocarp and
human bone were the dominant materials used in the production of hooks (Stokes n.d.).
By virtue of their size and morphology, the candlenut hooks must have been produced to
target particular sorts of fish. In particular, these are likely to have been small-mouthed fish
capable of being taken effectively with a rotating hook. Stokes (n.d.) cites an informant who
stated that both forms of candlenut hook were used for catching ‘komokomo’. The informant
had been shown how to use the ring-section hooks to catch komokomo as a child, explaining
that three such hooks were attached to the line, with the hooks holding fish in place until all
three hooks were filled (Stokes n.d.). Stokes further notes that the candlenut hooks described
to him corresponded closely, in terms of form, to metal hooks used for catching komokomo at
the time of his observations (Stokes n.d.). Fishbase (2007) recognises the Rapan komokomo as
being Leptoscarus vaigiensis within the Scaridae. Stokes does not specify which species of fish we
should associate with komokomo in his unpublished manuscript, and later he (Stokes 1955:334)
translates komokomo simply as ‘a fish’. Randall and Sinoto (1978), however, state that komokomo
is L. vaigiensis, an algal and sea-grass grazer that inhabits sheltered bays, harbours and lagoons.
While it seems unlikely that such a fish would take a hook, both the Stokes reference and
information provided by a local fisherman to AA in 2002 confirm that the seagrass parrotfish is
indeed captured with a small candlenut hook, especially in the month of March. Stokes further
records that the tuitui fish hooks were also used to catch the damselfish Stegastes fasciolatus
(Pomacentridae). S. fasciolatus is an ‘algal farmer’ in shallow waters, and as such can be highly
territorial of a small patch on the reef (Cardona and Clayton 1999).
In the context of the Rapan assemblages, the appearance of these small fish hooks coincides
neatly with the increase in Scaridae and Pomacentridae in the latter half of the upper level of
the E1/E2 area. While serranid numbers, which could potentially be associated with angling
technologies, do increase in Level III, the increase occurs in the bottom half of the level, with
numbers dropping off somewhat in the upper half of the level, and thus does not directly
coincide with the appearance of the candlenut fish hooks. The combination of the ethnographic
data with those from the fishbone analysis suggests that the occupants of Tangarutu were indeed
using candlenut fish hooks to catch komokomo/L. vaigiensis and nganga/S. fasciolatus.
The uniqueness of the Tangarutu candlenut hooks makes direct analogy impossible, but
extremely small hooks are reported both archaeologically and ethnographically from other
locales in Polynesia and Micronesia. In his observations of fishing practices and archaeological
fishing gear from the southern Cook Island of Ma’uke, Walter (1988) reported the use of small
shell hooks in association with titomo fishing (see also Walter 1991). As described by Walter
(1988:222), the hooks were used on a short line for catching a type of mackerel known as
koperu. Unripe coconut is spread on the water’s surface to draw the fish, and after their arrival,
the fishermen enter the water and place their baited hooks in the group of feeding fish. Fish
are quickly jerked into the boat when caught. While archaeological examples of the hooks used
for this specialised form of fishing, identified for the archaeological site of Anai’o on Ma’uke
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
151
(Walter 1988), are roughly the same size as those in the Tangarutu assemblage, the titomo hooks
are jabbing hooks, whereas all those from Tangarutu are rotating.
The ethnographic fishing study undertaken by Johannes (1981) on the Micronesian island
of Tobi also discusses the specialised use of small hooks, and in particular, the small fongtype rotating fish hook called haufong. This is used by Tobians to target triggerfish, which are
notorious ‘bait-stealers’, only nipping at bait unless the hook is small enough for the fish to take
it in whole (Johannes 1981:117). Johannes (1981:118) comments that hooks similar to the
haufong were used in various locales across Oceania, but that in other examples the fong (i.e. the
recurved point) was longer and thicker, and generally served to hold the bait in place rather than
the fish on the hook. While such a generalised distinction is perhaps debatable, what is clear is
that the highly standardised though unique Tangarutu fish hooks were produced to a definite
format to serve a clear purpose. It should be noted that triggerfish (Balistidae) are altogether
absent from the Rapan fishbone assemblages.
Ethnoarchaeology and Rapan marine exploitation
The unpublished manuscript on Rapan lifeways by John F.G. Stokes (n.d.) provides a firsthand
account of Rapan fishing, as well as providing some details on the types of littoral invertebrate
fauna consumed. There are some clear disparities between Stokes’ information and the evidence
from the archaeological record, but the ethnographic window provided by his manuscript also
provides useful insights into how a living was made in this unique environment. Here, we
consider the information provided by Stokes’ unpublished manuscript, as well as his published
article ‘Language in Rapa’ (1955), supplemented by Randall and Sinoto (1978) and information
on fishing collected by AA during the course of the Rapan excavation season, and assess its
potential to inform the archaeological record. Detailed descriptions of fish-capture techniques
provide an opportunity to push beyond broad correlations between fish families and capture
technologies, and Stokes’ notations regarding invertebrate processing methods offer a possible
explanation for fragmentation patterns noted archaeologically. Perhaps most significantly, the
ethnographic record contains clear statements about cultural ideas surrounding ‘edibility’ and
which marine foods were or were not taken. This information allows us to assess change through
time in cultural attitudes towards aquatic food sources, both within the archaeology itself, and
with the archaeological and ethnographic records juxtaposed.
Fishing – the ethnographic record
Stokes’ (n.d.) manuscript provides ethnographic data on fishing and fishing technology,
including hook and line angling, the use of various forms of nets and basket traps, and snares for
eeling, as well as Rapan names for the fish commonly caught using each method. The names of
fish provided by Stokes (n.d.) and their capture techniques have been matched with the species
names given by Randall and Sinoto (1978), and this information is provided in Table 9.2.
Line fishing
The small candlenut hooks used for the capture of komokomo and nganga have been discussed
above, however Stokes (n.d.) also provides information on the use of hook and line for the
capture of other fish species. Fishing from canoes with hook and line produced, primarily, the
rari (a large-mouthed red fish like a rock cod, with large anal spines, possibly the flagtail grouper,
Cephalopholis urodeta (Lieske and Myers 1994:25). Rensch (1988:237) and Randall and Sinoto
(1978:298) also have rari as Epinephelus fasciatus, the blacktip grouper, which is another red fish
of similar form. While relatively small, around 27 cm in average length, it has the reputation
terra australis 37
152 Katherine Szabó, Yolanda Vogel and Atholl Anderson
Table 9.2. Aquatic vertebrate and invertebrate fauna covered in Stokes (n.d.) with local names and details of capture techniques
where specified.
Capture method
(from Stokes n.d.)
Present in
assemblage?
Muraenidae (moray eels)
snare
yes
Congridae (conger eels)
snare
yes
Family
Species
Rapan name
Pisces
Atherinidae (silversides)
Pranesus insularum
kiamu/kiamo
ngake (hand net)
no
Serranidae (groupers/basses)
Cephalopholis urodelus
Epinephelus fasciatus
Pseudocaranx cheilio
Serioloa lalandi
Parupeneus fraterculus
tumutumuraupoo
rari
matu
ma’aki
kature/katuri
hook and line
hook and line
rau, hook and line
rau, hook and line
‘inaki ika (basket trap)
yes
Chaetodon spp.
Chaetondon smithi
Stegastes fasciolatus
amuamu
vaiti
nganga
‘inaki ika
‘inaki ika
candlenut hook
yes
Labridae (wrasses)
Gomphosus varius
Hologymnous sp.
Pseudolabrus inscriptus
Thalassoma lutescens
pokou
pokou
kariva
pokou
‘inaki ika
‘inaki ika
‘inaki ika
‘inaki ika
yes
Scaridae (parrotfishes)
Leptoscarus vaigiensis
Scarus chlorodon
Scarus ghobban
Scarus globiceps
komokomo
pahoro
para
pahoro
yes
Acanthuridae (surgeonfishes)
Acanthurus leucopareius
mama/ma’a ma’a
spearing
candlenut hook, toto (oval
dip net)
‘inaki ika
‘inaki ika
‘inaki ika
toto
Carangidae (jacks)
Mullidae (goatfish)
Chaetodontidae (butterflyfishes)
Pomacentridae (damselfishes)
yes
yes
yes
yes
Echinoidea
Echinidae
‘Echinus sp.’
vana
No
Crustacea1
Palinuridae (marine crayfish)
probably Panulirus ascuensis
koura
Atyidae (river shrimps)
probably Caridina rapaensis
koura kotae
‘inaki ika (basket trap)
not noted
Bathysquillidae
Bathysquilla microps (Stokes: Squilla sp.)
pongaponga
not noted
Grapsidae
karami
yes
Carpiliidae
probably Leptograpsis variegatus (Stokes: ‘rock crab found at base
of cliffs’)
? Carpilius convexus (Stokes: ‘rock crab – common form’)
papaa
yes
Ocypodidae
Uca tetragonon (Stokes: ‘small mud crab with large red claw’)
tararoa
possibly
Portunidae (swimming crabs)
Thalamita cerasma (Stokes: ‘swimming crab’)
tumomi
possibly
?
species unknown (Stokes: ‘small crab found under stones’)
kōrao
?
?
species unknown – recently moulted crab? (Stokes: ‘small rock
crab with a soft shell’)
pakapakaraumia
?
yes
not noted
Mollusca
Nacellidae
‘a limpet’ – probably Cellana tahitensis
pangi
Trochidae
Trochus radiatus (observed by Stokes in midden deposits)
name unknown
yes
Neritidae
Nerita morio
’i’i
yes
Cerithiidae
name unknown
yes
Ranellidae?
Clypeomorus batillariaeformis
(observed by Stokes in midden deposits)
‘Triton sp.’
pu
no
Arcidae
Anadara (= Arca) sp.
akaikai
no
Mytilidae
Modiolus sp.
piuu
yes
Isognomonidae
Melina sp.
kotakota
no
yes
Chamidae
Chama sp.
pa’ua
Spondylidae
Spondylus sp.
pa’ua
no
Psammobiidae
Asaphis violascens
ka’i
no
Tellinidae
Tellina rugosa
Tellina sp. (‘small’)
Tellina scobinata
Gafrarium pectinatum
pipi
yes
tupere
mitata
yes
yes
Periglypta reticulata
tupere
yes
‘Venus sp.’
tupere
?
Veneridae
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
153
in Tahiti of never being poisonous (Bagnis et al. 1974:106), unlike many snappers and coral
trout. Two carangids (matu, maaki) were also caught by hook and line. Randall and Sinoto
(1978:299) equate these with Pseudocaranx cheilio and Seriola lalandi respectively. The tough
wood of the mairari shrub (Dodonea viscosa) was used to make large and medium-sized bait
hooks by training the branches to grow in the required shape. A hook of mairari wood with a
tuitui point was used to catch albacore (ahi). Temporary hooks were tipped with rari anal spines
(Stokes n.d.:259–262). Pearl shell was absent at Rapa, but turtle shell and whale bone may have
been used in earlier days.
Rau (leaf sweeps)
Stokes (n.d.) describes the occasional use of large leaf sweeps (rau) for fishing. These consisted
of two wings (rau) of stiff rope with suspended ti leaves attached to a central pohue, or pound,
and they were used to drive fish, as in using a seine net. Stokes (n.d.) observed one catch,
which included matu (Pseudocaranx cheilio) and maaki (Serioloa lalandi) and a large quantity of
unnamed smaller fish. Although deployed infrequently in modern times, the leaf sweep is said
to have been an ancient form of fishing on Rapa.
Netting
Stokes (n.d.) states that net fishing had been largely abandoned at the time of his visit, however
‘… in former days [it] was of the greatest importance’. Nets (kupenga) used included seines,
hand nets and dip nets. Seine nets were used either in shallow water or from canoes further
out, but no details of fish catches are provided. Seines were also used for the capture of turtles.
Hand nets (ngake) attached to two sticks were used to capture a small fish called kiamo, which
Randall and Sinoto (1978) identify as the silverside Pranesus insularum. Dip nets (toto) were said
to have been used for the capture of mama (Acanthurus leucopareius) and komokomo (Leptoscarus
vagiensis). These were oval in shape with a single handle and could be operated either by a single
person or with others driving the fish.
Basket traps
Baskets traps (‘inaki ika), generally made from kiekie, were also used for fishing and crayfishing.
These traps are used in the month of November to capture kature, which Randall and Sinoto
(1978) name as the goatfish Parupeneus fraterculus. In addition to the seasonal capture of kature,
‘inaki ika are also used to take pahoro (Scarus chlorodon, Scarus globiceps), para (Scarus ghobban),
keikei (taxon unknown), amuamu (Chaetodon spp.), vaiti (Chaetodon smithi), poko (Gomphosus
varius, Hologymnous sp., Thalassoma lutescens) and kariva (Pseudolabrus inscriptus) (Randall and
Sinoto 1978; Stokes n.d.).
Eel snaring
Marine eels were snared in the reef shallows by women (Stokes n.d.). The eel snare (ngati)
consisted of two sticks, one of which held a bait, and the other a slip noose. Following snaring,
the eel was dashed against the rocks to kill it.
Spearing
Fish were caught in shallow water by casting or stabbing a wooden spear. This was generally
done from shore or when wading, and spears were also carried in canoes. Fish caught by this
method were parrotfish and ‘other shore feeders’, with crayfish also taken using spears from
canoes (Stokes n.d.).
Stone weirs
Stone fish traps (pa ika), regarded as of ancient origin, can be seen along the margins of east
coast bays, although by the 1920s it seems they were no longer used. Stokes (n.d.) notes the use
terra australis 37
154 Katherine Szabó, Yolanda Vogel and Atholl Anderson
of one form that consisted of stone-built leaders and a ngake for the capture of kiamo (Pranesus
insularum), and also that spearing was used for taking fish from weirs.
Other ethnographic data on fishing
Enquiries of an experienced fisherman on Rapa in 2002 (Mr Teraura Oitokaia, interviewed by
his daughter, Roti Oitokaia) elicited the information that rari was still a popular fish, especially
when it was fattest, from March to May. Komokomo (above), apparently known also as tapio, is
taken by hook in March and by diving from December to January. Nanue and similar fish called
pakavai and karamami are caught from April to September. According to Bagnis et al. (1974),
nanue is herbivorous, a seaweed feeder, and a popular subtropical food fish. Randall and Sinoto
(1978) equate all three names with varying sizes of the rudderfish Kyphosus bigibbus. It is caught
by driving schools towards the shore, using seines of coconut fronds. If there were fish remains
from middens along the harbourside, these might well contain nanue. Other species caught
today include flying fish or marara (Cypselurus sp.), which are in best condition in April, and
the parapo’atu, which arrives at Rapa in March to eat the maamanga (identified by Stokes (n.d.)
as a type of seaweed). Randall and Sinoto (1978) identify parapo’atu as the rabbitfish Siganus
argenteus. Perhaps the most interesting aspect of this evidence is the well-defined seasonality of
Rapan fishing.
Gathering of aquatic invertebrates – the ethnographic record
Stokes (n.d. 1955) provides much less information about invertebrate collection and consumption
than he does about fish and fishing. Exploited crustaceans are listed and given their Rapan
names, and although virtually no information on capture is provided, Stokes does state whether
aquatic invertebrates were normally consumed cooked or raw. Similar information is provided
for urchins. For molluscs, Stokes (n.d.) generally provides both local and scientific names for
taxa gathered and consumed, as well as information about whether the animals were eaten
cooked or raw, with occasional information about processing techniques. Of particular interest
with regard to molluscs, Stokes (n.d.) casually investigated eroding shell-midden deposits in
various locations around Rapa. He recorded the major taxa present and reports this information,
together with the then-current ‘edibility’ status of these species as provided by contemporary
informants. Both the archaeological and ethnographic elements of this survey provide revealing
points of comparison with the Tangarutu record.
Stokes (n.d.) lists a variety of medium-large crustaceans consumed as food on Rapa at the time
of his fieldwork. The marine crayfish, or koura (probably the locally common Panulirus pascuensis),
is at the top of the list, and also listed as prominent among exchange items with early European
voyagers in Rapan waters (Barratt 1988:201). In a section of the Stokes manuscript detailing
fishing technologies, he mentions that crayfish were taken in basket traps (Stokes n.d.). Stokes
(n.d.) also names the river shrimp, koura kotae (probably Caridina rapaensis in the Atyidae family),
and a species of stomatopod termed pongaponga (probably Bathysquilla microps) among species of
exploited Crustacea. All were apparently consumed cooked. None of these species were obviously
recognisable within the archaeological samples from Tangarutu,2 but a lack of reference specimens
means that less-diagnostic fragments might have been overlooked. Even so, the majority of the
Tangarutu crustacean remains clearly derives from species of brachyuran crab.
Stokes (n.d.) lists seven types of crab, named only by local names, accompanied by
descriptions of varying usefulness. The tararoa, described as ‘a small mud crab with a red claw’,
is clearly Uca tetragonan, although an identification for the kōlao – ‘a small crab found under
stones’ – is more elusive. All named taxa bar the unidentified kōlao and pakapakaraumia (a crab
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
155
with a ‘soft shell’, perhaps a general term for newly moulted crabs) can be assigned to a species
with relative confidence with the aid of distributional checklists and databases cataloguing
French Polynesian crustaceans (Poupin 1996, 1998; also http://decapoda.free.fr/).
Of the crab taxa listed by Stokes (n.d.), karami (Leptograpsis variegatus) and papaa (Carpilius
convexus) are certainly represented in the Tangarutu assemblage, with possible additional
occurrences of tararoa (Uca tetragonan) and tumomi (Thalamita cerasma). Levels of fragmentation
within the assemblage are such that reference specimens would be required for full identification
and quantification. Nevertheless, with these provisional identifications in hand, it is clear that
relatively large and sometimes aggressive and very mobile crabs were being taken as food both
archaeologically and ethnographically.
Although Stokes (n.d.) gives no information on crab capture, it is possible that some crabs
were caught in basket traps that were laid for the capture of various fish and possibly also
crayfish (see above). Some information on crab capture on Rapa was provided to AA by Teraura
Oitokaia. Apparently, crabs are caught all year round and are captured by pressing down on the
carapace, flipping them over, and piercing the eyes with the thumb. Sticks are used to catch
those secreted in holes. The species, or perhaps group of species, captured in these ways is not
specified.
Urchins are mentioned only briefly by Stokes (n.d.), who provides the local name vana
for a species of Echinus. It is likely that the attribution to Echinus is incorrect and was simply a
shorthand reference to ‘urchins’ using a well-known European/Atlantic genus, in much the same
way as ‘Chiton sp.’ is sometimes used to refer to polyplacophorans generally rather than those in
the genus Chiton or family Chitonidae per se. There is no description given to guide scientific
identification, but clearly from Stokes’ (n.d.) record, urchins were collected, and he states they
were eaten raw. The collection of urchins for food was also mentioned by Teraura Oitokaia in
information passed on to AA. The species being discussed is unclear, but Mr Oitokaia indicated
a preference for collection before spawning.
Certainly if the urchin in question was Diadema setosum, hand collection is unlikely, given
the length and toxicity of the spines. This implies that some sort of tool-assisted capture method
was in use, and as with crabs it is possible that urchins were caught on occasion in basket
traps. As detailed in Chapter 8, the relatively low levels of spine and test fragments present in
the Tangarutu deposits relative to the number of mouth parts suggest that preliminary urchin
processing took place away from Tangarutu shelter.
With regard to molluscs, information about cephalopods, gastropods and bivalves is provided
by Stokes (n.d.). Three species of cephalopod (two octopuses and a squid) are mentioned as
being taken and eaten either cooked or raw, with the archaeology having, necessarily, nothing
to add to this information. Stokes (n.d.) goes on to list bivalves and gastropods eaten or known
to be edible to contemporary informants, and those he observed in eroding midden deposits.
Although he states that shelled molluscs were not widely collected or eaten on Rapa during
his stay, save by children, a fairly wide range of molluscs is listed, along with local names and
occasional additional information.
Stokes’ (n.d.) list of mollusc species named as food sources by local informants leans
heavily towards bivalves, with only a few gastropods named. Hard-shore taxa include Chama
sp., Spondylus sp., Modiulus sp. mussels and Isognomon (=Melina) sp., while soft-shore species
include Gafrarium pectinatum, various species within the Tellinidae, Asaphis violascens, Anadara
(=Arca) sp. and Periglypta reticulata. Stokes (n.d.) observed the colonial soft-shore bivalve
Gafrarium pectinatum to be the most commonly consumed mollusc, eaten either raw or cooked.
Only three gastropods are included in Stokes’ (n.d.) list, including a limpet identified as Patella
terra australis 37
156 Katherine Szabó, Yolanda Vogel and Atholl Anderson
(but more likely Cellana, which was until recently placed in the Patellidae), the endemic Nerita
morio, and a ‘triton’ species. A subset of those molluscs listed by Stokes (n.d.) was named by
Mr Oitokaia. He mentions only one gastropod, the pangi’i limpet. Mussels, Isognomon sp. and
Gafrarium pectinatum are the most regularly taken of the bivalves, with Tellina rugosa and T.
scobinata now stated as rare on the island.
In Stokes’ wanderings around the island, he came across shells deposited on ridgelines,
which he took – doubtless correctly – to represent midden refuse. He notes the presence in a
number of locations of Chama sp. and Spondylus sp., with the former being common enough for
Stokes to consider it one of the major molluscan foods of past times. Also noted as being present
on the hillslopes were pieces of the unidentified pu whelk and Trochus sp. Of the latter, Stokes
(n.d.) states that there was no local name at the time of his investigations. It is likely, based on
the Tangarutu results, that this was Trochus radiatus. Of particular interest is Stokes’ discussion of the diminutive horn snail Clypeomorus batillariaeformis. In a casual investigation of the
remnants of a hearth within a ‘cliff shelter’, he notes the presence of crabs’ claws and a number
of shells, including 165 Gafrarium pectinatum valves and 550 Clypeomorus batillariaeformis –
nearly all of which were broken. Although Stokes’ informants knew of no local name for C. batillariaeformis and did not consider it a food source, Stokes says it was the most common shell in
the ‘backwaters of the harbour’ and surmised that it was probably an ‘article of diet’ in the past.
This conclusion is confirmed by the presence of C. batillariaeformis in the Tangarutu midden.
Ethnographic information and the archaeological record
The early 20th century observations of Stokes, together with information provided to Anderson
by Mr Oitokaia, allow us to pinpoint changing subsistence practices through time relating to
both environmental changes and shifts in cultural frameworks.
The information on fishing techniques and the taxa exploited provided by Stokes (n.d.)
generally provides a good fit with the archaeological fishbone assemblages from Rapa. Although
identification of the archaeological material was only undertaken to family level, many of the
species discussed by Stokes are represented, generally strongly, within families present in the
assemblages.
Both the archaeological analysis and ethnographic data suggest that netting and the use
of basket traps are likely to have been important throughout the sequence at Tangarutu, with
angling likely increasing in importance in the latter stage of occupation. Stokes (n.d.) lists species
in the Scaridae, Chaetodontidae and Labridae as being caught using basket traps (Table 9.2), and
these taxa featuring prominently throughout the Tangarutu sequence. Eels can be caught using
a number of techniques, including netting, but the fact that they are nocturnal while the other
taxa present are diurnal suggests that perhaps another method was employed, unless nets were
left set overnight. Several authors report the use of basket traps to capture eels ethnographically
throughout the Pacific (Masse 1986; Goto 1990; Davidson et al. 2002), although Stokes
(n.d.) does not list eels among the taxa captured using this method. The fact that eel snaring is
documented for Rapa (Stokes n.d.) means that this also must be considered a possible method
of capture at Tangarutu. The data from Tangarutu suggests an increase in angling late in the
sequence, as evidenced by the concurrent increase in Scaridae and Pomacentridae stated by
Stokes to be taken with tuitui hooks, and the appearance of candlenut fish hooks as discussed
above. Species in the Serranidae and Lethrinidae, most commonly caught with a hook and line,
also increase in the upper portion of the site.
While the Stokes data and the information obtained in 2002 provide a good fit with the
high-ranking taxa from Tangarutu, there are several taxa mentioned in these accounts that are
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
157
either absent or rank very low in the assemblages. The first of these are the small silversides,
kiamo/Pranesus insularum. These are either wholly absent, or accounted for by some of the bones
that were not in the reference collection. Silversides are very small, slender fish, growing to an
average length of around 10 cm (Randall 2005), so their absence may be due to taphonomic
factors. The goatfish, kature/Parupeneus fraterculus, while present, is scarce at Tangarutu. Mullids
have very fragile bones (Bilton 2001), so there may again be a taphonomic explanation for this.
However, Stokes (n.d.) notes that these fish are caught seasonally in November, so it is also
possible that their absence reflects seasonality or sampling issues. Other fish mentioned in the
ethnographic records that do not feature prominently, but are present in the Rapa assemblages,
are the carangids matu/Pseudocaranx cheilio and maaki/Serioloa lalandi, the acanthurid mama/
Acanthurus leucopareius, the kyphosids Nanue, pakavai and karamami (K. bigibbus), the exocoetid
marara/Cypselurus sp., and the siganid parapo’atu/Siganus argenteus. Taphonomic factors may
account for the low numbers of exocoetids and siganids, and the seasonal nature of fishing
on Rapa may also provide an explanation for the low occurrence of many of these taxa in the
archaeological record. There are also several low-ranking taxa in the assemblages that are not
mentioned in the ethnographic information (see Table 9.3).
Stokes spends considerably less time discussing marine invertebrates, but comparison of
his information with the midden record from Tangarutu is revealing nevertheless. The species
of molluscs listed by Stokes (n.d.) are virtually all present in the Tangarutu midden sample,
outwardly signalling some sort of constancy through time. However, the relative importance
of various taxa seems to alter dramatically. At the most general level, there seems to be a clear
swing away from gastropods as the most frequently collected and consumed class of mollusc to
a reliance on bivalves by the time of Stokes’ observations. While most of the bivalve species that
Stokes lists are present in the Tangarutu sample, numbers are very low, with only occasional, and
sometimes solitary, specimens. Viewed another way, this shift from gastropods to bivalves can
equally be seen as a shift from hard-shore to soft-shore species. This may reflect changing coastal
conditions, with increased progradation in modern times, but more data would be required to
fully assess that possibility.
That things may have been more complex than a straightforward transition from rockier
to more sediment-rich littoral zones is suggested by the pattern seen in Chama and Spondylus.
Both of these genera cement the lower valve to a hard substrate, and as such are associated with
Table 9.3. Fish taxa not mentioned by Stokes (n.d.).
Family
Common/Rapan name if known
Belonidae
needlefishes
Exocoetidae
flyingfishes
Holocentridae
squirrelfishes
Aulostomidae
trumpetfishes
Lethrinidae
emperorfishes
Kyphosidae
rudderfishes
Mugilidae
mullets
Siganidae (Siganus argentus)
rabbitfishes (parapuata – small/moroa – large)
Bothidae
lefteye flounders
Monocanthidae
filefishes
Diodontidae
porcupinefishes
terra australis 37
158 Katherine Szabó, Yolanda Vogel and Atholl Anderson
hard shores. However, the numbers of Chama are low in the Tangarutu sample – despite a
preponderance of hard-shore taxa – and according to Stokes were important at the time of his
data collection. Chama pacifica is represented by only 12 fragments in E1 and a single fragment
in E2. Of these 13 fragments, three are heavily worn and cannot have entered the site with a live
mollusc inside. There are no Spondylus sp. remains in the Tangarutu sample at all.
At present, the same comparisons between Stokes’ information and the results of midden
analysis cannot be extended to the crustaceans or echinoderms. Species-level data would be
required on both sides. It is apparent that brachyuran crabs and urchins have been a fairly stable
part of the diet through time on Rapa, although the midden data at least would suggest that
their importance has fluctuated.
Dynamic environments and culture through time
Drawing together the archaeological and ethnographic data on marine exploitation on Rapa,
there is clear evidence of change through time, as on other East Polynesian islands, such as
Aitutaki (Allen 2002). Some of these changes are quite specific, such as those seen in fishing
technologies, whereas others concern the broad nature of marine resource use. Environmental
change is surely playing a part, and a palaeoecological investigation of the marine-derived faunas
through time demonstrates the subtlety of the dynamics of Anarua Bay. Following a discussion
of clear trajectories of change in the exploitation of marine resources, we assess, as a whole, the
various lines of evidence from both fish and invertebrate fauna to look at shifts in the nature of
Anarua Bay and human interaction with it through time.
Elements of change
The rising importance of angling
The archaeological data from Tangarutu point to a fishing strategy focused mainly on the use
of nets and basket traps, however the ethnographic data provided by Stokes (n.d.) indicates
that the use of hook and line had overtaken these methods in more modern times. While
Stokes’ information does provide details of netting and basket trapping, he points out that
these methods have either fallen into disuse, or are seldom practised. Indeed, he speaks of their
occasional use as a ‘fad’. The ethnographic data collected by Anderson in 2002 also emphasises
hook and line fishing. This would appear to be a continuation of the changes that can be seen
in the latter stage of occupation at Tangarutu, where a rise in angling is clearly evident and fish
hooks appear in the archaeological record.
The rise in importance of soft-shore niches
As detailed above, the rocky/hard-shore-dominated molluscan assemblage from Tangarutu
contrasts with the information on molluscan foods supplied by Stokes (n.d.). Stokes (n.d.) does
mention hard-shore species, and soft-shore bivalves are present in the Tangarutu assemblage, so
nothing as dramatic as a faunal turnover is evidenced. Even so, the transition from a hard-shore
gastropod-dominated focus to a soft-shore bivalve one speaks to a change in the nature of littoral
habitats and with it, approaches to the gathering of shellfish. Such shifts in species availability
are apparently not uncommon in the context of Anarua Bay, and will be discussed further below.
Shifts in the composition of fish catches
While there is a considerable amount of stability in the taxa commonly exploited at Tangarutu,
the changes in relative abundance of these taxa throughout the archaeological sequence are quite
pronounced. The implications of these changes in terms of ecology and fishing strategies will
be considered below. The archaeological data show a reliance on Scaridae at initial occupation,
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
159
lasting throughout Level I and extending into the base of Level II, although giving way somewhat
to Muraenidae at this stage. Above Spit 13, in the mid section of Level II, there is a dramatic
increase in the occurrence of Chaetodontidae, rising from 7% of the assemblage at the base of
that level to 23%, with Scaridae numbers dropping to only 6% of the assemblage. Muraenidae
ranks second in terms of relative abundance throughout Level II. The uppermost level of the
site sees further change, with Chaetodontidae numbers decreasing and an increase in Serranidae
in the bottom half of the level. This changes again in the top half of the level, with Scaridae
returning to the position of top-ranking taxon.
Shifts in the composition of molluscan assemblages beyond hard/soft shore
While the hard-shore/soft-shore dichotomy between the archaeological and ethnographic
records is distinctive, a number of more subtle changes can be detected within hard-shore
assemblages of the Tangarutu sample. Figure 9.5 shows the relative proportions of the four
major mollusc species, spit by spit, for squares E1 and E2. There are clear spatial differences
in deposition between E1 and E2, despite broadly similar changes. However, the strong
stratigraphic patterning in the deposits also validates interpretations of change through time.
Although sample sizes differ widely between spits, this percentage-based view of composition
still highlights some noteworthy patterns. Nerita morio is clearly dominant in the uppermost
spits, with Gafrarium pectinatum and then Clypeomorus batillariaeformis taking over in relative
importance from spits 3–6 in E1, and 3–7 in E2. The limpet Cellana tahitensis makes the
strongest contribution from spits 7 to 13 in E1, and in Spit 14 in E2. Below this, Nerita morio
is again the most important species in E1, with Clypeomorus batillariaeformis dominating the
much smaller samples towards the base of E2.
These differences in relative abundance of molluscan taxa in the Tangarutu midden hint at
environmental transformations considerably more subtle than swings between hard-shore and
soft-shore environments. Nerita morio, Cellana tahitensis and Clypeomorus batillariaeformis are
all microphagous algal grazers on hard surfaces. Indeed, Gafrarium pectinatum is the only major
mollusc species represented in the Tangarutu assemblage that is not an algal grazer. This hints
that the changes in community structure and composition may be related to shifts in the nature
and coverage of algae in Anarua Bay. Overlying the urchin and fish data adds increasing weight
to this argument.
Algae: The impact of putative changes in the lowest trophic level at Anarua Bay
The shell data from Tangarutu clearly show shifts in the local abundance and availability of
algal-grazing taxa through time. However, algae feeders also characterise the urchin sample and a
notable proportion of the fish assemblage from Tangarutu. As with the Tangarutu shell, changes
in the availability and exploitation of various taxa through time are conspicuous (see Figure 9.6).
The urchin Diadema setosum grazes on non-crustose filamentous algal turfs within and
below the intertidal zone. Studies have shown that Diadema spp. is one of the major grazers
and bioeroders in inshore contexts, with herbivorous fish, such as those in the Scaridae,
Acanthuridae and Pomacentridae, playing a relatively minor role in overall algae removal in
habitats conducive to urchins (Foster 1987). Indeed, following a mass die-off of the Caribbean
Diadema antillarum, algal cover in Jamaican inshore habitats increased markedly to a mean of
95% coverage of recorded substrates (Hughes et al. 1987; see also Liddell and Ohlhorst 1986).
Diadema density in inshore habitats appears to be influenced by a number of features. In terms
of the physical nature of preferred habitats, Diadema prefers coarser sediments, shunning muddy
or silty substrates, dislikes complex reef architecture, and avoids high-wave-energy areas (Foster
1987; Dumas et al. 2007). Diadema further prefers micro- over macroalgae, avoiding Sargassum
terra australis 37
160 Katherine Szabó, Yolanda Vogel and Atholl Anderson
Figure 9.5. Percentage graphs of the four major shell species in squares E1 (above) and E2 (below). Note that the E2 sample is
generally smaller and below spits 10–13 is negligible.
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
161
Spit #
Figure 9.6. Schematic representation of abundances of major marine taxa represented in Tangarutu squares E1 and
E2 by depth. Taxon picture placements indicate a high level of occurrence at that depth, with solid lines indicating a
sustained presence and dotted lines indicating lower levels of occurrence.
in particular (Shunula and Ndibalema 1986; Dumas et al. 2007). A number of studies have also
demonstrated that Diadema densities are considerably higher in impacted/fished inshore habitats
than in protected, unfished areas (McClanahan et al. 1994; Carreiro-Silva and McClanahan
2001). Indeed, Diadema setosum is one of the most conspicuous inhabitants of heavily altered
dynamite-fished coral-reef habitats (KS pers. obs.).
Diadema remains are present throughout most of the Tangarutu sequence, although there
are marked changes in frequency through time. In terms of the invertebrate record, the lowest
urchin densities in squares E1 and E2 coincide with the dominance of the bivalve Gafrarium
tumidum (see Figure 9.6), suggesting that Anarua Bay at these times hosted greater expanses
of fine-grained soft sediments than was ordinarily the case. The highest densities of Diadema
remains tend to occur with high levels of Nerita morio and/or Cellana tahitensis (see Figure 9.6),
all affirming the presence and productiveness of micro-algal turfs in Anarua Bay through many
parts of the early, middle and later Tangarutu sequence.
The strong showing of invertebrate taxa favouring algal environments is also reflected in the
fish taxa present. Given the low numbers of identified fish relative to invertebrates, the small
sample of fish taxa cannot be taken to present a comprehensive picture of inshore stocks or habitats.
Nevertheless, a presence-absence approach to the occurrence of certain taxa at different depths,
such as butterflyfishes and damselfishes, allows us to draw some conclusions regarding the nature
terra australis 37
162 Katherine Szabó, Yolanda Vogel and Atholl Anderson
of inshore niches at particular times in Anarua Bay. Algal-grazing fish taxa dominate the Tangarutu
assemblage, with grazing taxa including scarids, pomacentrids and chaetodontids, supplemented
by the fluctuating presence of serranids and the more consistent presence of eels in the Muraenidae
and Congridae. While there are 13 species in the Chaetodontidae known from Rapa, the only
species-level identification provided by Randall and Sinoto (1978) is vaiti/C. smithi, which is
known to aggregate in areas of algae-covered rocky reefs (Lieske and Myers 1994). As previously
discussed, both komokomo/L. vaigiensis and nganga/S. fasciolatus, associated ethnographically with
the candlenut fish hooks, also inhabit areas of abundant algal growth.
Collectively, the fish, urchin and shell data from Tangarutu suggest that Anarua Bay
supported extensive hard-substrate habitats, which in turn supported substantial non-crustose
micro-algal turfs. However, the oscillations in the frequencies of different algal-grazing mollusc
taxa, grazing Diadema urchins and fish taxa hint that parallel changes may have occurred in the
structure and extent of these local algal turfs. A closer look at the ecological literature on tropical
marine algal-grazing communities offers some insights into how community structure can be
altered in the face of rather subtle changes or disturbances – including by fishing.
As might be expected, various of these algivorous creatures compete with each other –
sometimes aggressively. Stegastes spp. are known collectively as ‘farmer damselfish’, as they carefully
maintain and ‘weed’ patches of non-crustose algae in a territory that they defend vigorously (Russ
1987). While microalgal-grazing molluscs may be tolerated in Stegastes territories, Diadema
urchins are not (Lieberman et al. 1984; Klumpp and Polunin 1989). Similarly, Stegastes may
patch-share with acanthurids, which are poorly represented in the Tangarutu sample, but will
ward off scarids. Given that in some zones of the Tangarutu deposits quantities of Diadema
and pomacentrid remains occupy the same stratigraphic position (e.g. the upper spits of E1), it
would appear that distinct, non-overlapping zones of the rocky reef are being exploited by the
local residents. But the ecological literature suggests that such discrete zones are likely to change
in extent and community composition through time.
Grazing strongly effects the distribution of algal species (Lieberman et al. 1984; McClanahan
et al. 1994). This is especially the case where grazers selectively target algal taxa that do not
dominate in biomass terms (Hatcher and Larkum 1983). Pomacentrids will selectively remove
non-favoured algae from their territories, thus increasing the biomass of favoured species
(Klumpp and Polunin 1989), while intensive urchin grazing promotes algal growth, keeping
the standing crop trimmed, thereby decreasing ‘self-shading’ (Klumpp and McKinnon 1989).
It stands to reason that the removal of algal grazers in any numbers, whether urchins, fish or
even numbers of grazing molluscs, will impact on the growth rate and taxonomic composition
of algal communities, and indeed such effects have been demonstrated.
The above discussion stresses natural processes and indirect human impacts that may be
contributing to changes in exploited marine resources through time at Tangarutu. However, we
also know that the forms of human impacts themselves were changing. The late-stratigraphic
association of the candlenut fish hooks demonstrates tangibly that fishing technologies
transformed through time, and this might be expected to have had an impact on the landed
assemblage.
The subtle oscillations through time in the community structure and balance of algalgrazing herbivorous marine taxa, both vertebrate and invertebrate, suggest that environmental
and habitat parameters were also constantly shifting. Some of this variability may relate to
strictly environmental factors such as variations in storm frequency, wave intensity and cycles of
erosion and progradation. However, the marine ecology literature would suggest that high levels
of occurrence of the urchin Diadema setosum are likely to be the product of at least moderate
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
163
human fishing pressure on marine communities within Anarua Bay (see McClanahan et al.
1994; Carreiro-Silva and McClanahan 2001). Although it is presumed that the rise in Diadema
densities in tandem with fishing pressure was largely due to the removal of urchin predators in
the system (e.g. McClanahan 1998), it is also likely that the greater availability of food following
the removal of herbivorous algal-grazing fish, particularly within the Scaridae, Acanthuridae and
Pomacentridae (Foster 1987), promoted growth in urchin populations.
In a similar vein to the relationship between fish and urchin population structure and
abundance, the relationship between levels of crayfish (Panulirus pascuensis) exploitation and
urchin abundance may turn out to be linked. A link between heavy exploitation of the urchin
predator Panulirus and a rise in urchin numbers has been established (Tegner and Levin 1983).
At present, this relationship cannot be assessed from the Tangarutu sample; species-level
identification of the crustacean remains would be required.
High levels of Diadema urchin grazing have been found to alter significantly the composition
of organisms that compete with algae (Sammarco 1980, 1982). Not only is coral recruitment
hindered by urchin grazing, but the diversity of fleshy algae, polychaete worms, encrusting
coralline algae, filamentous greens and foraminifers is reduced, with some groups being
virtually excluded (Sammarco 1980). This scenario has flow-on implications for the structure
and composition of herbivorous vertebrate and invertebrate communities, as well as for those
creatures that prey on them.
The consistent presence, and even more so the sporadically high abundance, of Diadema
setosum remains within the Tangarutu deposit indicates that Anarua Bay was not only subject
to fishing pressure, but that the various trophic levels and competing species within them were
constantly adjusting to intra-community and environmental pressures. The respective roles of
grazing gastropods, urchins, crustaceans, eels and various species of herbivorous and predatory
fish in the Anarua system varied as different species and classes jostled for survival and dominance.
Humans and marine resources at Anarua Bay
The archaeological and ethnographic records offer some insights into how humans impacted
this dynamic ecosystem through time. The relationship between densities of Diadema urchins
and fishing pressure has been discussed above, but an argument for depletion of the standing
stocks of Scaridae due to overfishing in the early part of the sequence could perhaps be advanced.
This family is present throughout the sequence, though it is probable that the species targeted
changed through time. The ethnographic data for differing capture methods for different species
of scarid provided by Stokes (n.d.) and the increase in Scaridae at the end of the sequence,
concurrent with the appearance of the candlenut fish hooks, would tend to support this.
Overall, the picture of marine resource use by the occupants of Tangarutu shelter through
time is one of flexibility. Resource-switching is frequent, probably reflecting concomitant
changes in inshore habitats and resources. Strategies employed for food capture seem to be
general enough that a wide range of taxa is accessible at any given time, yet flexible enough to
allow such resource-switching if local marine conditions and communities change. Both these
facets of subsistence behaviour are also in evidence in the ethnographic record. The fish and eel
capture technologies described by Stokes, as well as the roster of captured species, show that a
balance was maintained between more generalist capture strategies, such as the use of weirs, nets
and basket traps, and very targeted strategies, such as the use of candlenut hooks, to capture
Leptoscarus vaigiensis.
A closer comparison of the archaeological and ethnographic records also reveals that resourceswitching in a Rapan context was not always simply a matter of economic expediency – at least
terra australis 37
164 Katherine Szabó, Yolanda Vogel and Atholl Anderson
it is not always explained as such. As noted by Stokes, many of the species of mollusc seen by
locals as ‘inedible’ were abundant in midden deposits. Eels feature prominently in the assemblage
from Tangarutu throughout the sequence, and are also present in the assemblage from Akatanui
3. Sharks, while low ranking in terms of relative abundance, are also present in both assemblages.
The capture of both eels and sharks was still in evidence during the time Stokes (n.d.) collected
his ethnographic data. However, the information collected by AA during the 2002 field season
indicates that neither sharks nor eels are taken today. Rapans consider shark repugnant and
eels tapu. Thus, shifts in resource procurement and consumption also appear to have a cultural
dimension, with ideas about food sources changing markedly through time.
Conclusion
In the two-way traffic of human-environmental interactions seen at Anarua Bay there occurred
active modifications in traditional Polynesian ways of life. Many of the resources so important
elsewhere in Polynesia were not available to Rapans, and new strategies and solutions had to be
consciously developed. Such creative problem solving is clearly evidenced in the Tangarutu fishhook assemblage, where traditional working techniques are applied to novel materials. A further
potential way of solving the problem of the lack of availability of traditional raw materials would
have been to import them – a solution seen elsewhere in Polynesia (Weisler 1993). There is no
evidence that such action was ever pursued by Rapans.
Cultural innovation can typically be understood as a creative reconfiguration of traditional
practice; ideas do not emerge from nowhere, but from the knowledge base already present in
a given society (e.g. see Bijker 1987; Hickman 1995). As such, innovation often represents
incremental change rather than a complete disjuncture (Kroeber 1948:360). Thus, while the
candlenut fish hooks recovered from Tangarutu appear as a cultural novelty, they have strong
technological parallels with traditional Polynesian fish-hook manufacture in other materials.
Technological innovations do not follow trajectories, but rather create them (Bijker and Law
1992; Szabó 2005:93), and given the late provenance of the Tangarutu hooks it appears that
developing solutions to living in a Rapan environmental context was an ongoing and reciprocal
process between population and environment (Kennett et al. 2006) throughout prehistory.
Notes
1.
2.
Juvenile specimens of Pinctada margaritifera, with a maximum width of ca.12 mm, have been
recorded in the Tangarutu shell midden, but there is no indication that specimens of the size and
robusticity required for fish-hook manufacture were locally available.
Several crayfish mandibles were seen during excavation at E1 and E2, Tangarutu, but they were
not identified in the retained samples.
References
Allen, M. 2002. Resolving Long-Term Change in Polynesian Marine Fisheries. Asian Perspectives
41(2):195–212.
Anell, B. 1955. A Contribution to the History of Fishing in the Southern Seas. Studia Ethnographica
Upsaliensia IX, Uppsala.
Bagnis, R., Mazellier, P., Bennett, J. and Christian, E. 1974. Fishes of Polynesia. Landsdowne Press,
Melbourne.
Barratt, G. 1988. Southern and Eastern Polynesia. Volume 2 of Russia and the South Pacific 1696–1840.
University of British Columbia Press, Vancouver.
terra australis 37
Marine resource exploitation on Rapa Island: Archaeology, material culture and ethnography
165
Best, E. 1929. Fishing Methods and Devices of the Maori. Dominion Museum Bulletin 12, Wellington.
Bijker, W.E. 1987. The Social Construction of Bakelite: Toward a theory of invention. In: Bijker, W.E.,
Hughes, T.P. and Pinch, T.J. (eds), The Social Construction of Technological Systems, pp. 159–187.
MIT Press, Cambridge MA and London.
Bijker, W.E. and Law, J. 1992. Do Technologies Have Trajectories? In: Bijker, W.E. and Law, J. (eds),
Shaping Technology/Building Society: Studies in Sociotechnical Change, pp. 17–19. MIT Press,
Cambridge MA and London.
Bilton, M. 2001. Taphonomic Bias in Pacific Ichthyoarchaeological Assemblages: A Marshall Islands
Example. Unpublished MA thesis, Department of Anthropology, University of Otago.
Buck, P. 1954. Vikings of the Sunrise. Whitcombe and Tombs, Christchurch.
Cardona, M. and Clayton, W. 1999. The Algal Community of the Farmer Damselfish, Stegastes
fasciolatus, at Three Sites in Fiji and the Kingdom of Tonga. Bios 70(2):71–75.
Carreiro-Silva, M. and McClanahan, T.R. 2001. Echinoid bioerosion and herbivory on Kenyan coral
reefs: the role of protection from fishing. Journal of Experimental Marine Biology and Ecology
262:133–153.
Davidson, J., Leach, F. and Sand, C. 2002. Three thousand years of fishing in New Caledonia and the
Loyalty Islands. In: Bedford, S., Sand, C. and Burley, D. (eds), Fifty Years in the Field: Essays in
Honour and Celebration of Richard Shutler Jr’s Archaeological Career. New Zealand Archaeological
Association Monograph 25:153–164.
Dumas, P., Kulbicki, M., Chifflet, S., Fichez, R. and Ferraris, J. 2007. Environmental factors
influencing urchin spatial distributions on disturbed coral reefs (New Caledonia, South Pacific).
Journal of Experimental Marine Biology and Ecology 344:88–100.
Emory, K.P., Bonk, W.J. and Sinoto, Y.H. 1968. Fishhooks. Bernice P. Bishop Museum Special
Publication 47, Honolulu.
Fishbase 2007. Leptoscarus vaigiensis. http://filaman.ifmgeomar.de/Summary/SpeciesSummary. php? ID
=4360 &genusname =Leptoscarus&speciesname=vaigiensis
Foster, S.A. 1987. The relative impacts of grazing by Caribbean coral reef fishes and Diadema: effects of
habitat and surge. Journal of Experimental Marine Biology and Ecology 105:1–20.
Goto, A. 1990. Prehistoric Hawaiian fishing lore: An integrated approach. Man and Culture in Oceania
6:1–34.
Hatcher, B.G. and Larkum, A.W.D. 1983. An experimental analysis of factors controlling the standing
crop of the epilithic algal community on a coral reef. Journal of Experimental Marine Biology and
Ecology 69:61–84.
Hickman, L.A. 1995. Techniques of Discovery: Broad and Narrow Characterizations of Technology.
In: Pitt, J.C. (ed), New Directions in the Philosophy of Technology, pp. 207–218. Kluwer Academic
Publishers, Dordrecht, Boston and London.
Hughes, T.P., Reed, D.C. and Boyle, M-J. 1987. Herbivory on coral reefs: community structure following
mass mortalities of sea urchins. Journal of Experimental Marine Biology and Ecology 113:39–59.
Johannes, R.E. 1981. Words of the Lagoon: Fishing and Marine Lore in the Palau District of Micronesia.
University of California Press, Berkeley, Los Angeles and London.
Kennett, D., Anderson, A., Prebble, M., Conte, E. and Southon, J. 2006. Prehistoric Human Impacts
on Rapa, French Polynesia. Antiquity 80:340–354.
Kirch, P.V. and Yen, D.E. 1982. Tikopia – The Prehistory and Ecology of a Polynesian Outlier. Bernice P.
Bishop Museum Bulletin 238, Honolulu.
Klumpp, D.W. and McKinnon, A.D. 1989. Temporal and spatial patterns in primary production of a
coral-reef epilithic algal community. Journal of Experimental Marine Biology and Ecology 131:1–22.
Klumpp, D.W. and Polunin, N.V.C. 1989. Partitioning among grazers of food resources within damselfish territories on a coral reef. Journal of Experimental Marine Biology and Ecology 125:145–169.
Kroeber, A. 1948. Anthropology: Race, Language, Culture, Psychology, Prehistory. Second edition.
Harcourt, Brace and World Inc., New York and Burlinghame.
Liddell, W.D. and Ohlhorst, S.L. 1986. Changes in benthic community composition following the mass
terra australis 37
166 Katherine Szabó, Yolanda Vogel and Atholl Anderson
mortality of Diadema at Jamaica. Journal of Experimental Marine Biology and Ecology 95:271–278.
Lieberman, M., John, D.M. and Lieberman, D. 1984. Factors influencing algal species assemblages on
reef and cobble substrata off Ghana. Journal of Experimental Marine Biology and Ecology 75:129–143.
Lieske, E. and Myers, R. 1994. Coral Reef Fishes. Indo-Pacific & Caribbean including the Red Sea.
Princeton University Press, New Jersey.
Masse, W.B. 1986. A millennium of fishing in the Palau Islands, Micronesia. In: Anderson, A.
(ed), Traditional Fishing in the Pacific: Ethnographical and Archaeological Papers from the 15th
Pacific Science Congress. Pacific Anthropological Records, no. 37 pp. 85–119. Department of
Anthropology, Bernice P. Bishop Museum, Honolulu.
McClanahan, T.R. 1998. Predation and the distribution and abundance of tropical sea urchin
populations. Journal of Experimental Marine Biology and Ecology 221:231–255.
McClanahan, T.R., Nugues, M. and Mwachireya, S. 1994. Fish and sea urchin herbivory and
competition in Kenyan coral reef lagoons: the role of reef management. Journal of Experimental
Marine Biology and Ecology 184:237–254.
Poupin, J. 1996. Crustacea Decapoda of French Polynesia (Astacidea, Palinuridea, Anomura,
Brachyura). Atoll Research Bulletin, April 442:1–114.
Poupin, J. 1998. Crustacea Decapoda and Stomatopoda of French Polynesia (Dendrobranchiata,
Stenopodidea, Caridea, Thalassinidea and Stomatopoda, with additions to Astacidea, Palinuridea,
Anomura and Brachyura). Atoll Research Bulletin, September 451:1–62.
Randall, J.E. 2005. Reef and Shore Fishes of the South Pacific: New Caledonia to Tahiti and the Pitcairn
Islands. University of Hawai’i Press, Honolulu.
Randall, J.E. and Sinoto, Y.H. 1978. Rapan Fish Names. B.P. Bishop Museum – Occasional Papers
XXIV, 15:291–306.
Rensch, K.H. 1988. Fish Names of eastern Polynesia. Pacific Linguistics Series C – no. 106, ANU,
Canberra.
Russ, G.R. 1987. Is rate of removal of algae by grazers reduced inside territories of tropical
damselfishes? Journal of Experimental Marine Biology and Ecology 110:1–17.
Sammarco, P.W. 1980. Diadema and its relationship to coral spat mortality: grazing, competition, and
biological disturbance. Journal of Experimental Marine Biology and Ecology 45:245–272.
Sammarco, P.W. 1982. Effects of grazing by Diadema antillarum Philippi (Echinodermata: Echinoidea)
on algal diversity and community structure. Journal of Experimental Marine Biology and Ecology
65:83–105.
Shunula, J.P. and Ndibalema, V. 1986. Grazing preferences of Diadema setosum and Heliocidaris
erythrogramma (Echinoderms) on an assortment of marine algae. Aquatic Botany 25:91–95.
Skinner, H.D. 1942. A Classification of the Fish-hooks of Murihiku – with notes on allied forms from
other parts of Polynesia. Journal of the Polynesian Society 51:208–221.
Stokes, John F.G. n.d. Ethnology of Rapa. Manuscript on file. Bernice P. Bishop Museum, Honolulu.
Stokes, John F.G. 1955. Language in Rapa. Journal of the Polynesian Society 64:315–340.
Szabó, K. 2007. An assessment of shell fishhooks of the Lapita cultural complex. In: Anderson, A.J.,
Leach, B.F. and Green, K. (eds), Vastly Ingenious, pp. 227–241. Otago University Press, Dunedin.
Szabó, K. 2005. Technique and Practice: Shell-working in the Western Pacific and Island Southeast
Asia. Unpublished PhD thesis, The Australian National University, Canberra.
Tegner, M.J. and Levin, L.A. 1983. Spiny lobsters and sea urchins: analysis of a predator-prey
interaction. Journal of Experimental Marine Biology and Ecology 73:125–150.
Walter, R. 1991. Fishing on Ma’uke: An Archaeological and Ethnographic Study of Fishing Strategies
on a Makatea Island. New Zealand Journal of Archaeology 13:41–58.
Walter, R. 1988. Archaeology Beyond the Reef: Koperu Fishing on Ma’uke. Archaeology in New
Zealand 31(4):222–227.
Weisler, M. 1993. Long Distance Interactions in Prehistoric Polynesia: Three Case Studies.
Unpublished PhD thesis, University of California, Berkeley.
terra australis 37
10
Palaeobotany and the early development
of agriculture on Rapa Island
Matiu Prebble
Department of Archaeology and Natural History, The Australian National University, and Nga Pae o te
Maramatanga Trust, University of Auckland, New Zealand, matiu.prebble@anu.edu.au
Atholl Anderson
Department of Archaeology and Natural History, The Australian National University
Introduction
Palaeobotanical studies in tropical and subtropical Remote Oceania have established broad
records of vegetation and climate change covering the past 6000–7000 years. They mostly have
used sedimentary deposits from large, closed depositional basins such as high-elevation bogs or
volcanic caldera lakes in the tropical Hawaiian Islands (Selling 1948; Selling in Massey 1979;
Athens and Ward 1993) and subtropical Easter Island (Flenley 1979; Flenley and King 1984;
Flenley et al. 1991), and numerous bogs in northern New Zealand. While these records have
provided good regional pictures of vegetation change, including the effects of fire and increases
in seral pollen and charcoal particles that have been interpreted as indicating human activity,
rarely have they revealed more direct evidence of the introduction of agriculture. As a result,
many palaeoecological studies that argued the case for or against evidence of agriculture, and
its associated age, came under criticism for a lack of direct proxies of human activity, imprecise
chronology, and the sensitivity of the biological proxies to natural disturbance (e.g. Anderson
1994, 1995; McGlone and Wilmshurst 1999).
For these and other reasons, views on the significance of agriculture to Remote Oceanic
colonisation have tended to be polarised. Kirch and Green (2001), primarily on linguistic
grounds, argued that when humans first expanded into Remote Oceania around 1000 BC they
brought a complete roster of domestic animals (dog, pig and fowl), plus commensal rats and
other small vertebrates, and oceanic crops including taro (Colocasia esculenta), yam (Dioscorea
spp.), bananas (Musa spp.) and breadfruit (Artocarpus altilis). These were thought essential to the
colonisation process, particularly in the islands east of the Solomon Group, where the number
of plant genera and the natural distribution of economic plants declines dramatically, forming
terra australis 37
168 Matiu Prebble and Atholl Anderson
what Jones and Spriggs (2002) have described, somewhat extravagantly, as a ‘green desert’. The
necessity of agriculture to colonisation of Remote Oceania has been questioned (Davidson and
Leach 2001) and asserted (Addison 2008) on nutritional grounds.
The concept of ‘transported landscapes’ (sensu Anderson 1952) has been employed by
Kirch (1982, 2002) to emphasise the motivation behind ecological transformation of islands
in Remote Oceania, in which a newly occupied ecosystem was rapidly reshaped to operate in
much the same way as the homeland ecosystem. Anderson (2003, 2009a, b), conversely, has
argued that the distribution of oceanic crops and domestic animals is very patchy across Remote
Oceania, especially in the more remote islands, and that this is a function of the increasing
difficulty of transportation eastward and southward. As a result, some islands had no access, or
limited access, to agriculture, and others in which it was well established by the historical era
had probably created such landscapes by long-term accumulation of taxa rather than during
initial colonisation. Of course, some islands, largely by geographical circumstance, were able to
institute agricultural economies from almost the beginning of settlement.
Models aside, it remains the case that the importance and sources of vegetable carbohydrates
to early populations in Remote Oceania are poorly known and enigmatic, even in very wellstudied instances, such as in Vanuatu where stable isotope records from the Teouma burial site
on Efate indicate that around 800 BC the diet was dominated by terrestrial and marine protein
sources, with only a minor contribution from vegetable foods (Valentin et al. 2010). Whether
these were of introduced taxa is uncertain because the identification of starch grains, found on
stone tools and pots from archaeological deposits in Vanuatu (Horrocks and Bedford 2005;
Horrocks et al. 2009), is open to question on several methodological grounds (e.g. Hardy et al.
2009; Wilson et al. 2010).
Turning to the role of terrestrial palaeoecological records, it is necessary to acknowledge that,
in tropical regions especially, the nature of fossil proxies, the state of preservation, the degree of
chronological resolution and the completeness of the deposits can be highly variable. Prebble
and Wilmshurst (2009) argued that multiple proxies for human activities, primarily agricultural
production, are more often preserved in sites close to where such activities occurred and that
site selection is critical for addressing questions about agricultural expansion. In considering
active taro (Colocasia esculenta) gardens built on deep organically rich sedimentary deposits
underlying Maunutu Swamp on Rimatara (Austral Islands, French Polynesia), they found that
such deposits needed to be treated like archaeological profiles, as the various proxies accumulated
under different sets of conditions influenced by human activity. Artefacts, including garden
stakes and animal tethers, from the Rimatara deposits emphasised the inherent archaeological
character of the deposits. The physical mixing of sedimentary strata by digging, both in the
past and present, has increased the likelihood of obtaining unreliable radiocarbon chronologies.
Despite these potential problems, the array of fossil proxies available from these sites can reveal
important insights into the changing biodiversity of garden systems.
Rapa Island offers a case study of the initiation of agricultural activity in the early phase of
human colonisation. Kennett et al. (2006), Prebble and Dowe (2008) and work as yet unpublished
have identified multiple proxies for ecological change from Tukou Swamp, on Rapa, spanning
the past 7500 years. Preserved Pandanus tectorius syncarps and high concentrations of Pandanus
pollen characterise the pre-human swamp record. A coastal swamp forest developed on top of
estuarine sediments between 2000 BC and AD 1000. In the past 750 years, unprecedented
increases in charcoal particle concentration and seral fern spores, plus a rapid decline in Pandanus
pollen, indicate early forest clearance. A number of plant species became extinct, including a
palm tree (Prebble and Dowe 2008). The interpretation of cultural activity after about AD
terra australis 37
Palaeobotany and the early development of agriculture on Rapa Island
169
1200 is strengthened by the near-synchronous appearance of pollen from the introduced plant
cultigen Colocasia esculenta, and of weeds indicative of the expansion of agricultural activity.
Here we extend this analysis by outlining the availability of indigenous food plants and
historical evidence of agriculture, and then focus on the development of agriculture at Tukou
Swamp by examining the chronology of introduction and expansion of Colocasia across the
site. We highlight some of the complexities and the potential of palaeobotanical analyses for
understanding the role of agriculture in the colonisation of one of the more remote islands of
Remote Oceania.
Indigenous carbohydrate sources on Rapa
Understanding the extent of indigenous carbohydrate-rich plants (i.e. those plants that have not
been introduced) is one means of assessing the potential of indigenous food sources to sustain
growing human populations. The natural distribution of some of the main oceanic root and
tree crops is limited to the Western Pacific. Most of the remote Pacific Islands, including Rapa,
lacked bananas (Musa spp.), yams (Dioscorea spp.), taro (Colocasia esculenta), rhizome crops
including the gingers (e.g. Curcuma longa and Hedychium flavescens) and tree crops including
Syzygium malaccense, Burkella obovata, Barringtonia edulis and Inocarpus fagifera. What, then,
was available on Rapa?
Based on a summary of the pre-human fossil data from Tukou Marsh (Prebble unpublished
data) and survey data of the existing flora of Rapa (Meyer 2002), the variety of indigenous
carbohydrate, fat and sugar sources from vascular plants available at initial human arrival can be
defined (Table 10.1). Which species would have provided an adequate staple food or supplement
to an otherwise protein-rich diet at the advent of people cannot be deduced from this list, and
some plants that potentially served as important fat and carbohydrate sources are missing from
it. This is merely a partial guide to the potential sources.
Coconut trees are indigenous on many islands in Remote Oceania at least as far east as
Tubuai in the Austral Archipelago, but they are not indigenous to the Hawaiian Archipelago or
Rapa (Prebble and Dowe 2008). The distribution of coconut trees is partly controlled by rainfall,
as fruits remain unviable in areas or in years with less than 1000 mm annual precipitation
(Fosberg 1956). The distribution of pigs in the initial colonisation of Remote Oceania may have
been limited in part by the distribution of coconut, particularly the domesticated, large-fruited
trees. Most of the wild-type indigenous coconuts that were established on the more remote
Pacific Islands would have possessed very small fruits with minimal endosperm (Prebble and
Dowe 2008) and they may not have provided an adequate source of protein, vegetable fat and
carbohydrate to sustain pig husbandry and growing human populations (Kirch 2002). Dried
coconut meat (endosperm) from domesticated coconut consists of around 20% carbohydrate,
including soluble sugars, sucrose and starch (Jayalekshmy and Mathew 1990). Large servings
to satisfy daily requirements of carbohydrate would require domesticated large-fruit varieties.
Coconuts occurred rarely (Chapter 2), if at all, prehistorically on Rapa.
Another carbohydrate and fat source would have been provided by the seeds of Pandanus
tectorius, which, from fossil pollen, appears to have been very common across the remote
Pacific Islands including Rapa, particularly on islands that lacked mangroves in estuaries and
backswamps. Generally, Pandanus forms monospecific stands with a sedge/herb-dominated
understorey (Ash and Ash 1984), but no ‘true’ mangrove species occur east of the Cook Islands
(Ellison 1991). On Rapa, P. tectorius appears to prefer a mangrove habitat, forming dense
stands along the few margins of tidal flats where human activity has not been prominent. Other
Pandanus species may be indigenous to Rapa; for example, St. John (in Fosberg and St. John
terra australis 37
170 Matiu Prebble and Atholl Anderson
Table 10.1. Fossil taxa with potential food value identified from Tukou deposit in pre-human aged sediments older
than AD 900 are listed. Also listed are the possible representatives of the fossil taxa found in the modern flora of Rapa
and the biogeographic affinity of taxa. Extant indigenous taxa without fossil records that may provide some food value
are also listed (data sources from Meyer 2002).
Fossil taxa
Family
Indigenous representatives in
modern flora
Biogeographic affinity
Carbohydrate, sugar
and fat sources
Angiopteris
Angiopteridaceae
Angiopteris rapensis, A. longifolia
Indo-Pacific/endemic
pith and root
Apiaceae
Apiaceae
Apium australe
Indo-Pacific
root
Arecaceae: Arecoideae
type
Arecaceae, subfamily
Arecoideae, tribe
Iguanurinae
extinct
?
fruit and young shoots?
Celtis
Ulmaceae
Celtis pacifica
Indo-Pacific
fruit
Coprosma
Rubiaceae
Coprosma cookei, C. rapensis Coprosma Pacific/endemic
spp. Possible extinct species
fruit
Cyathea
Cyatheaceae
Cyathea affinis, C. medullaris, C.
rapensis, C. stokesii
Indo-Pacific/endemic
pith
Cyperaceae-
Cyperaceae
numerous species
pan-tropical, cosmopolitan root
Davallia/ Histiopteris
Davalliaceae/
Dennstaedtiaceae
numerous species
pan-tropical, cosmopolitan root
Dryopteridaceae
Dryopteridaceae
numerous species
pan-tropical
root
Freycinetia
Pandanaceae
Freycinetia arborea
Indo-Pacific
fruit and root
Hypolepis
Dennstaedtiaceae
Hypolepis punctata, H. tenuifolia
Indo-Pacific/Pan-tropical
root
Malvaceae
undifferentiated
Malvaceae
numerous species
?
shoots?
Monolete Psilate
?
numerous species
?
root
Monolete undiff.
?
numerous species
?
root
Myrtaceae undiff.
Myrtaceae
numerous species
?
fruits?
Pandanus cf. tectorius
Pandanus
Pandanus tectorius
Indo-Pacific/endemic?
fruits
Pteris
Pteridaceae
numerous species
?
root
Polypodiaceae
Polypodiaceae
numerous species
?
root
Rubiaceae
Rubiaceae
numerous species
?
fruit?
Trilete Psilate
?
numerous species
?
root
Trilete undiff.
?
numerous species
?
root
Urticaceae/ Moraceae
Urticaceae/Moraceae
numerous species
?
shoots?
Extant indigenous taxa without fossil records
Tetragonia tetragonioides
Aizoaceae
Tetragonia tetragonioides
temperate/sub-tropical
Pacific
leaves
Capparis sp.
Capparaceae
Capparis sp.
Indo-Pacific
fruit
Leptecophylla rapae
Ericaceae
Leptecophylla rapae
endemic
fruit
Vaccinium rapae
Ericaceae
Vaccinium rapae
endemic
fruit
Astelia rapensis
Liliaceae
Astelia rapensis
endemic
fruit
Macropiper puberulum
Piperaceae
Macropiper puberulum
indigenous
fruit
terra australis 37
Palaeobotany and the early development of agriculture on Rapa Island
171
1934) recorded 13 species of Pandanus in 1934, although Smith (1979) and Stone (1988)
suggest many of these are simply varieties of P. tectorius. In examining the genus on Tahiti, Stone
(1988) could not discount the human introduction of additional Pandanus species or P. tectorius
varieties and this also may have been the situation on Rapa.
Fern roots and pith could have served as an important substitute for tropical root crops
at the time of initial Polynesian arrival, especially in sub-tropical or temperate environments
(McGlone et al. 2005). Tree-fern pith (particularly of Cyathea medullaris) is still consumed on
Rapa (Freddy Riaria pers. comm. 2002) and it might have provided an important starch source,
given the likely former abundance of these trees at the time of human arrival on the island.
Other ferns with large piths include Marrattia spp. and Angiopteris spp., the latter having a
restricted distribution in Remote Oceania. It is unclear how abundant ground ferns including
Histiopteris incisa, Dryopteris sp. and Dicranopteris linearis were prior to human arrival, but with
cultural burning these plants would have become widespread along forest margins, providing an
abundant but rudimentary starch source.
Algae may have provided a source of carbohydrate to the first colonists of Remote Oceania.
Algal material is found in some archaeological middens in Remote Oceania (e.g. Kirch et al.
1995), but the extent of consumption is unclear (Conte and Payri 2006). Small fragments of
algae were found in the Tangarutu deposit on Rapa, but the possibility that they were blown in
cannot be excluded.
Historical observations of agriculture on Rapa
Although nobody landed from the Discovery in 1791, the vessel spent some time close inshore
and Vancouver thought Rapa conspicuous by the lack of Cocos nucifera (described as the ‘cocoanut’ palm). He did not see any Colocasia esculenta pondfield production systems, perhaps
because observation of the inland valleys was obscured by thick vegetation along the coastline.
Vancouver (in Lamb 1984:215) recorded that:
… they [the islands] were chiefly clothed with shrubs and dwarf trees. Neither the plantain
[Musa spp.] nor other spontaneous vegetable productions common to the inhabited tropical
islands, presented themselves.
Menzies (in Shineberg 1986:65–66) noted some plants that were used in the clothing of
Rapan men who came out to the ship:
The only cloathing they wore were a narrow slip of cloth made from the bark of a tree [prob. Talipariti
tiliaceum syn. Hibiscus tiliaceus but potentially Broussonetia papyrifera] which passed round their waist &
between their legs, this cloth appeard to be a very scarce article amongst them as many of them had not
sufficient of it to cover their nakedness, it was evident however that they generally wore something for
that purpose, as some of them had bunches of leaves of a species of Dracena [prob. Cordyline fruticosa]
suspended to a girdle round their middle for that intention.
Only three or four plant species may be inferred from the descriptions provided in both
accounts. More general comments were made regarding the status of the island’s vegetation,
including the impoverished state of the forests and absence of coconut palms, as summarised
here by Menzies (in Shineberg 1986:67–68):
The valley round the bottom of the Bay is tolerably pleasant when compard with other parts of the
Island being scattered over with Bushes among which we could perceive the habitations of the Natives
& some little signs of Cultivation, the hills behind & on the South Side of it appeard thinly coverd with
some verdure & here & there wooded with some scrubby Trees particularly in the hollow places between
terra australis 37
172 Matiu Prebble and Atholl Anderson
the hills, but they seemed of no great magnitude. Towards the North end the hills are not so rugged &
rocky but ascend with a smooth surface coverd with grass & destitute of Trees or bushes of any kind. We
observed no Cocoa Nut Trees [Cocos nucifera] anywhere on the Island.
In 1924, Stokes (m.s. Group 2 Box 7.1) spent most of the year compiling an ethnography of
the island but also recorded the recollections of the islanders and made many useful observations
of agricultural systems. He remarks that ‘in early days cultivation was confined almost entirely
to the wet-land taro’. He quotes Davies (1827), who in 1826 ‘inspected the plantations of taro,
which were all laid out, and display a degree of skill and contrivance evinced in nothing else we
saw’. He also quotes Davies who found that: ‘One coconut palm was reported as present from a
drift and was not recognized by the local natives.’ Stokes (m.s. Group 2 Box 7.2) also reported
in 1924 that:
There are now many coconut palms scattered through the island, which though vigorous and grown to
a fair height, according to report, drop their fruit before it matures. Other introductions at Tupuaki, on
the northern coast, are said to have borne fruit that was ‘killed by thunder’ perhaps to a cold storm. A
white man who has tried to grow coconuts for commerce at the northern end of the harbor apparently
abandoned the venture as a failure. It is evident that Rapa is too cold for the successful production of
coconuts, but that the palms will grow in many parts of the island, and will occasionally produce fruit.
The non-recognition of the coconut palm by the natives, reported by Davies, points very suggestively to
the fact that the palm was unknown to the ancestors of the Rapa people.
Of taro agriculture, Stokes provides an intriguing account of hand digging of pondfields,
commenting that ‘most taro cultivation was done with the bare hands a method still used
to a remarkable extent. Once the pond is prepared, the hand is really more suitable for taro
cultivation than any tool.’ He also suggests that many areas of the island ‘are dry and well-suited
for the cultivation of sweet potato, which though present in the early days, was evidently of very
little importance’. He goes on to comment on the cultivation of other crops:
Opinions differ as to the cultivation of bananas. Of other useful plants, there seem to have been
cultivated only koali (gourd) and aute (paper mulberry). Seeds of the gourd are set in the ground
with little preparation, and the plants are left to look after themselves. The growing of aute has been
given up, but apparently in former days it was definitely cultivated. A dry-looking hill slope bordering
the shore at Aupapa, Angairao Bay, is said to have been an aute field; the patches are small, narrow,
irregular terraces following the contour of the land. At the bottom of the slope the soil is retained by
a rough stone wall.
The benchmark for botanical observations of the island comes from an unpublished report
of the B.P. Bishop Museum Expedition of 1934 by Harold St. John and Raymond Fosberg, but
they made little comment about the varieties of plant cultigens or about the state of gardens
on the island. Additional information about the distribution and character of modern taro
pondfields is provided by Bartruff et al. in Chapter 13. From interviews with local gardeners
in 2002, Motley (unpublished data) recorded 25 varieties of Colocasia esculenta known on
Rapa. Of these, 20 were regarded as Polynesian introductions and the remaining five of recent
introduction. A number of varieties are known to have been lost as a result of disease and
probably neglect over the past decade, including 10 Polynesian varieties, as well as two recent
varieties. Around six varieties are currently grown for popoi, a type of cooked taro formed by
pounding and fermenting to create a thick paste that can be stored.
terra australis 37
Palaeobotany and the early development of agriculture on Rapa Island
173
Study site: Tukou Marsh, Ha’urei Harbour, Rapa
Ha’urei Harbour on Rapa represents the remnant of one of the original volcanic calderas formed
during island orogeny in the Pliocene. The shallow bathymetry of the harbour suggests the
entire area was exposed during glacial lowstand sea levels. The exposed reef at the harbour
entrance, backed by calcareous reef shoals along the interior, currently restricts wave action from
the sea, to the degree that the harbour could be described as a lagoon. At low tide, fine-grained
sediments of a mixture of alluvial sediments, shell and coral detritus are exposed to reveal three
prominent bird’s-foot river deltas and tidal flats extending out from the high-tide shoreline (see
Figure 10.1). Water depth surrounding the Tapuki (Tapu’i) basaltic islet in the centre of harbour
is no more than 1.5 m below low-tide level.
Shoreward of the exposed deltaic sediments is a series of levee-backed marshes positioned
at Tukou, Matataa and Aitoke (Figure 10.2). All of these sites have been heavily modified by
human activity and grazing by cattle, horses and goats. A map of Ha’urei Harbour sketched in
1864 by John Vine Hall (1869) shows the position of a number of coastal pondfield systems
Figure 10.1. Map of Rapa showing the main windward boulder-beach and leeward sandy-beach embayments. The
position of Tukou and the highest mountain peak (Perau, 650 m) are also indicated.
terra australis 37
174 Matiu Prebble and Atholl Anderson
that were used for Colocasia esculenta cultivation (Figure 10.2). Remnant dry-stonewall terrace
features are now interspersed within many of the levee-backed marshes, along the margins of
riverbeds, or along the present shoreline. At certain points, these marshlands extend more than
100 m inland and rise to an elevation of more than 2 m. On the most inland side, these marshes
are bound either by river levees or steeply rising embankments of the surrounding hill slopes that
appear to have been cut by a previous, probably mid-Holocene, sea-level high stand (Figure 10.3).
Tukou Marsh lies on the south side of the broadest river delta and associated tidal flats of
Ha’urei Harbour. The marsh is banked and divided by two smaller prograding river channels,
forming two waterlogged depressions. The central-most marsh depression was chosen for this
palaeoecological study, and sediment cores were taken in 2002 (June–August). The modern
marsh appears to have developed behind the southern levee of the main river, and the northern
levee of the smaller river bounded the marsh to the south. The marsh extends from the estuarine
shoreline, marked at high tide (ca. 0.4 m above mean sea level) by driftwood and estuarine
detritus, ca. 135 m to the embankment of the surrounding hill slope. Below the embankment
the marsh reaches a maximum elevation, based on laser transit measurements, of 3 m above the
high-tide mark. The remains of an agricultural terrace complex rise an additional 50 cm or more
above the marsh surface.
Overall, the climate on Rapa is subtropical to warm temperate. It receives around 2500–
3000 mm annual rainfall, based on measurements taken from Ha’urei village, also at sea level
(Rapa Meteo unpublished data). Tukou is fed by two permanent streams that drain the eastern
catchment of the Perau–Namuere range, with water flow gradients averaging above 30 degrees.
Hydrological data are unavailable for this catchment.
Figure 10.2. Section of a sketched map of the head of Ha’urei Harbour, Rapa, by John Vine Hall (1869a), showing the
location of the main marsh systems (‘Taro patches’), including Tukou.
terra australis 37
Palaeobotany and the early development of agriculture on Rapa Island
175
Figure 10.3. Approximate position of mid-Holocene high stand maximum marine incursion after sea-level
stabilisation at Tukou (black line). Up to ca. 3 m of organic rich sediment has accumulated at this location above
present high tide/spring tide line, marked by the interface between the marsh and the tidal flats.
The vegetation of the waterlogged marsh surface has been described by Prebble (In press).
The marsh is currently dominated by introduced agricultural grasses (e.g. Paspalum subjugatum)
and adventive herbs (e.g. Commelina diffusa and Ludwigia octovalvis), with some indigenous
sedges (e.g. Carex spp.) and rushes (e.g. Schoenoplectus subulatus subsp. subulatus; some authors
suggest that this may have been introduced to French Polynesia). The embankment on the
periphery of the marsh is largely open grass with some naturalised Syzgium jambos and Psidium
spp. trees (all in the Myrtaceae). Some indigenous trees are found at the site, including Talipariti
tiliaceum syn. Hibiscus tiliaceus (Malvaceae) and Metrosideros collina (Myrtaceae). Hall’s (1869)
sketch of the site indicates that stone-wall terraces lined the upper portion of the marsh and
were focused along the banks of the adjacent river channels. Some of these terrace features
remain, but in a degraded state. Only one feral C. esculenta specimen was located at Tukou,
with no other plant cultigens identified. The marsh is currently left for cattle and horse grazing.
Methods
Multiple profiling at Tukou Marsh
In order to determine the range of taphonomic processes affecting the pollen assemblages, and
also the initial establishment of Colocasia agriculture and its spatial expansion from palynological
and sedimentary proxies, a multiple profile approach was chosen. This allows for direct pollen
stratigraphic correlation (sensu Clark et al. 1986). Due to the complexities of palynological
analyses, multiple profiles are rarely analysed for palynological signatures. Instead, palynologists
terra australis 37
176 Matiu Prebble and Atholl Anderson
tend to either increase the temporal resolution of a single profile, or take single profiles in
order to compare between individual depositional settings within a region. The multiple profile
approach allows local or extra-local spatial variations in pollen representation to be examined in
more detail (Dumayne-Peaty and Barber 1998). If a reliable absolute chronology is obtained, it
also allows palynological signatures to be more precisely linked to stratigraphic processes, both
spatially and chronologically.
Six cores were taken at 20 m intervals along a transect running through the centre of the
marsh from the first core site (Core 1) at the shoreline to the embankment behind the marsh.
The five cores were all taken with a D-Section corer. An attempt was made with each coring
point to reach bedrock or a depth where compacted basal clays prevented further penetration.
The relative position to high-tide sea level was determined from laser transit measurements. The
main parameters of each core are outlined in Table 10.2.
Laboratory and numerical method
A composite stratigraphic diagram (Figure 10.4) was constructed using the program C2 Data
Analysis version 1.6 (Juggins 2010). In this diagram, each core profile is vertically aligned
according to relative sea level (at high tide) and its relative position from the original core site.
Calibrated radiocarbon age ranges and the Bayesian depositional age/depth models are presented
for each core in this diagram. Each model was processed using the program OxCal version 4.1
(Bronk Ramsey 2008). The palynological data for all of the Tukou Cores are presented in this
Table 10.2. Tukou Marsh multiple profile summary. Main features of each core are presented, including mean Bayesian
ages for the key listed events. All ages were modelled using the program OxCal 4.1 (Bronk-Ramsey 2008) of calibrated
radiocarbon determinations. Mid-point estimates refer to the average age of the mean Bayesian ages presented for
each event recorded at Tukou. For more details of the radiocarbon chronology and event stratigraphy see Prebble et al.
(In press).
Core features
Core 3
Core 6
Core 2
Core 5
Core 4
Core 1
Age
Depth of core (cm)
318
250
400
250
250
652
Basal clay reached
yes
no
yes
no
no
yes
Depth above high tide (cm)
300
243
196
161
111
20
Distance from shoreline (m)
105
85
65
45
25
5
number of AMS radiocarbon
determinations
3
4
5
2
4
6
30–32
70–72
60–62
48–50
48–50
–
after AD 1825
Colocasia end (AD)
1685
1920
1852
1816
1554
–
AD 1800
Colocasia onset (AD)
1170
1392
1024
1286
1098
–
AD 1210
Earliest unprecedented Microcharcoal
peak (AD)
1016
1117
1024
1073
1098
1193
AD 1105
241
585
-130 (BC)
1073
1002
1193
pre-human arrival
AD 1263
1016
1254
1510
1393
1376
1401
550
1117
1364
1180
1287
1297
Events
Depth (cm) of post-European plant
introduction
Microcharcoal onset (BC/AD)
Pandanus decline >30% from peak (AD)
Pandanus peak (AD)
terra australis 37
AD 957
Palaeobotany and the early development of agriculture on Rapa Island
177
same composite stratigraphic diagram. This examination of the Tukou core transect focuses
on radiocarbon ages, charcoal particle concentrations, and Pandanus cf. tectorius and Colocasia
esculenta pollen as key markers of forest clearance and agricultural expansion at the site. A more
detailed analysis of all of the fossil data recorded at Tukou, including the chronological data, is
in preparation.
Results
The palynological record of Tukou indicates that coastal Pandanus swamp forests expanded after
about 500 BC in response to changing sedimentary conditions following a mid-Holocene sealevel high stand (Prebble unpublished data). Within or around the period represented from the
basal ages of the Tangarutu archaeological sequence, certainly after AD 1000, between 50 cm and
120 cm of colluvial sediments accumulated at different rates across Tukou Marsh. This period
is represented by a rapid increase in sedimentation comparable with the accumulation rates
found from deposits in the other embayments (e.g. Akatanui, Anarua and Iri). The Pandanus
pollen records (Figure 10.4) indicate that swamp forest declined consistently across the site
within the past 1000 years. The stratigraphy and chronology of cores 1 to 6 from Tukou have
been described in the context of coastal swamp forest development (Prebble unpublished notes).
Here, we outline some of the key palynological indicators for human impact and attempt to
further define the stratigraphy and chronology for each core. Following this, we explore each of
the cores individually in order to define some of the key ecological responses to human impacts
at Tukou.
Summary of palynological analysis of human activity at Tukou
The cores are divided into four palynological zones on the basis of the major vegetation changes.
In the composite stratigraphic diagram (Figure 10.4) only the upper zone (IV) is marked. This
zone has been divided into two subzones (IVa and IVb) on the basis of palynological marker
taxa that represent the Polynesian (IVa, shaded in transparent light grey) and European periods
(IVb, shaded in transparent dark grey). These subzones and the lower three zones are discussed
in more detail in Prebble (unpublished notes).
Charcoal particles
The presence of high concentrations of charcoal particles in Zone IV of Core 1 is not directly tied
to sedimentation or preservation. As a working hypothesis, it reflects human-induced burning
at the site. In Zone IV of Core 1, counter to what might be expected, increasing charcoal
concentrations overlap with increasing Pandanus pollen concentrations. These Pandanus peaks
may be representative of one of the following: sediment compaction in the early part of the zone
allowing for pollen to accumulate and become more concentrated with depth; or the opening of
forest conditions allowing for more pollen deposition; or the possibility that Pandanus responds
to burning by increased flowering and pollen production. The decline in charcoal signals in the
later part of this zone also corresponds with a decrease in Pandanus pollen concentrations. This
might represent a decrease in the amount of fuel available at Tukou after an initial period of
burning.
The key problem identified from the analysis of the Core 1 data set is establishing the
chronology for deforestation represented by Pandanus pollen decline and increases in charcoal
particles alone. No radiocarbon dates were obtained from Zone IV sediments of Core 1, but the
Bayesian depositional model of the radiocarbon ages that were obtained suggests that the base
of Zone IV encompasses the radiocarbon chronology established from the archaeological record
(indicated by the vertical grey bar within the age model histogram in Figure 10.4).
terra australis 37
178 Matiu Prebble and Atholl Anderson
terra australis 37
Figure 10.4. Composite stratigraphic diagram for Tukou Marsh, showing the mean Bayesian age models (BC/AD) and bar histograms of charcoal particles, Pandanus cf. tectorius and Colocasia
esculenta pollen concentrations (per cm3). The lithology for each profile is shown, with four being coarse and one being fine sediment. An attempt was made to present the concentration data
proportionally, with each single tick mark representing 200k charcoal particles (per cm3), 100k Pandanus, 50k Cyperaceae and 1k Colocasia pollen grains (per cm3). The stratigraphic alignment
of each core is displayed, with the modern high-tide level marked at 25 cm relative to mean sea level (RSL). Only two subzones are distinguished on the basis of transitions in the palynological
record. Zone IVa is shaded in transparent dark grey and Zone IVb shaded in transparent light grey.
Palaeobotany and the early development of agriculture on Rapa Island
179
More radiocarbon ages obtained from the five additional Tukou cores (cores 2–6) better
establish the chronology of charcoal particle increases, particularly of microcharcoal. The
chronology for the onset of microcharcoal particle concentrations identified from pollen
concentrates has been modelled using radiocarbon interpolations based on the Bayesian paradigm
and generated using the program OxCal 4.1 (see Table 10.2). The earliest age obtained for the
initial appearance of microcharcoal particles comes from Core 2 in the centre of the marsh,
established from a direct AMS date at 130 BC, with the youngest age determined from Core
1, at AD 1193. The concentration of microcharcoal particles also varies considerably across
the site. The highest concentrations occur in Core 6 and concentrations overall are higher in
cores 2–6 than in Core 1. Differential fuel loads across the site could explain this variation in
microcharcoal concentrations, but the influence of tidal action might also have reduced the
amount of particle accumulation. The intensity of human activity at the site is higher towards
the inland part of the marsh where evidence for past agricultural activity is most apparent and
this may also explain the spatial variation in microcharcoal concentrations.
The mean Bayesian ages for the onset of microcharcoal concentrations from all of the Tukou
cores are consistently older than the earliest age determination obtained from the archaeological
record of Tangarutu. All of the mean Bayesian ages for peaks in microcharcoal concentrations
are recorded within the archaeological chronology established for Rapa. This is quite typical of
charcoal profiles in Pacific swamps and it is usually held to represent sporadic low-level charcoal
production resulting from infrequent natural fires, followed by abundant charcoal production
with a substantially increased rate of ignition on the arrival of people. Thunder and lightning
occur infrequently at any time of the year on Rapa and could ignite grass or scrub fires during
dry seasons, although such an event has not been recorded.
It is often very difficult to distinguish microcharcoal particles from blackened woody tissue
that develops in the reducing environment of estuarine water. The presence of charcoal particles
radiocarbon dated from the base of the cores could thus indicate either that there was no burning
(the black material not being charcoal), or that burning was part of the pre-human island
ecosystem, or, indeed, that charcoal had been produced from an early but transient human visit.
In the absence of other human indicators, or of the chemical determination of black particles as
charcoal, caution must be observed.
Initiation of agriculture: Colocasia esculenta pollen records
Apart from charcoal particle increases, no plant cultigen pollen or any unequivocal palynological
signatures of human presence were identified from Tukou Core 1. Pollen from the cultigen
Colocasia esculenta was identified from each of the remaining five cores from Tukou, all within
Zone IV. The presence of Colocasia pollen provides a more secure indication of human presence
as it could only have been derived from introduced populations. The distribution of C. esculenta
in Remote Oceania is restricted by oceanic barriers (Matthews 1995, 2004) and may also be
limited by the lack of suitable hydrological environments for it to establish without human
intervention. C. esculenta pollen has now been identified in numerous pollen records across the
Indo-Pacific, suggesting the records from Rapa are not anomalous.
The composite stratigraphic diagram (Figure 10.4) shows that, like Pandanus pollen and
microcharcoal particles, C. esculenta pollen concentrations are differentially represented across
the site. The highest concentrations are found in cores 2 and 4 in the central part of the marsh,
with the lowest concentrations found in the most inland core (Core 3). This trend could be
explained by Colocasia cultivation activity whereby plants are actively prevented from flowering
during cultivation, mainly by corm harvesting before flowering. C. esculenta is known to be a
terra australis 37
180 Matiu Prebble and Atholl Anderson
poor pollen producer, and it is relatively rare in the pollen record (e.g. Haberle 1994; Prebble
unpublished data). The high concentrations of C. esculenta pollen represented in some cores
may reflect a response to the increased seasonality of the more subtropical climate of Rapa.
Alternatively, high C. esculenta pollen concentrations in sediments may reflect the high local
presence of feral plant populations outside of immediate cultivation that may have been allowed
to set flowers. One hypothesis is that this may have been an integral strategy in the early
establishment of Colocasia crops, allowing for genetic exchange and the extension of cultivar
diversity in an ecosystem different from the homeland of the first Polynesians to arrive on the
island.
The mean Bayesian ages for the onset of C. esculenta pollen from cores 2–6 (see Table
10.2) all fall in an age bracket that overlaps the earliest age determination obtained from the
archaeological record of Tangarutu (Chapter 11). Most of these mean ages are older than
expected and this may reflect the loose chronological control for the first appearance of C.
esculenta pollen in three of the five cores (cores 3, 4 and 5). The onset of C. esculenta in Core 2 is
constrained by four AMS ages and may provide the most reliable age for the onset of cultivation
on Rapa at 1024 AD, but this still appears to be older than expected.
None of the mean Bayesian ages are more than 200 years older than the earliest archaeological
determination for the island. They could be correct. It is possible that the process of taro
cultivation resulted in the mixing of sediments. Modern digging of wet-fields usually involves
the turning of between 20 cm and 30 cm of topsoil and this process could have resulted in
older organic material being incorporated into the planting horizon. Yet, it is apparent that no
chronological inversions were noted among radiocarbon ages in any of the six profiles examined.
This suggests that the swamp was continually accumulating sediment and organic materials and
that if sediment mixing did occur, it was minimal.
With the exception of Core 4, all of the peaks in C. esculenta pollen occur in the latter part
of Zone IVa. With a mean Bayesian age of AD 1231, the sample showing first appearance of C.
esculenta pollen in Core 4 also has the highest concentrations. If this age is accepted, the early
presence of high pollen concentrations of C. esculenta in the most shoreward core site might
reflect the proposition, above, of high local presence of feral Colocasia populations outside areas
of immediate cultivation. Later in the sequence, certainly by Zone IVb across most of the cores,
Colocasia pollen drops out of the record, possibly representing a change in land-use practices at
the site. It is possible that harvesting of Colocasia intensified to the point where both cultivated
and feral plants were prevented from flowering. This could also represent the abandonment of
taro field cultivation at Tukou, which was apparent by the time Hall arrived on the island in
1864. Another alternative is that with the introduction of feral grazing animals, including goats,
cattle and horses, feral populations of Colocasia may have been browsed to the point where
flowering was prevented. Few feral specimens survive on the island today and no flowers have
been recorded or collected on the island by any visiting botanist (Jean-Yves Meyer pers. comm.).
Sedimentary deposits from other embayments on Rapa
Small sedimentary catchments are located at the base of windward boulder and leeward sandy
beach embayments of the island (Figure 10.5; see Figure 10.1 for locations). The more deeply
incised windward embayments are characterised by higher energy fluvial/alluvial systems,
forming in places of extensive swamp deposits adjacent to abandoned Colocasia esculenta
terraced pondfields. These swamps possess highly organic clay and silt deposits, preserving
botanical remains. The sedge Schoenoplectus subulatus var. subulatus dominates the vegetation
of most of these swamp sites. The swamp deposits were cored with the intent of reaching basal
terra australis 37
Palaeobotany and the early development of agriculture on Rapa Island
181
clays or bedrock from Akatanui (2 m deep Core 1; Figures 10.5 and 10.6), Iri (3 m deep Core
1; Figure 10.7) and Anarua (1.8 m deep Core 1; Figure 10.8). Magnetic susceptibility (K)
measurements from these cores provide some indication of the extent of organic accumulation
at each site. Radiocarbon ages were obtained from charcoal or black organic pieces found in
the basal sediments of these cores, revealing 2σ ages of AD 1420–1460 (UCIAMS 2194), AD
620–1280 (ANU 12154), and AD 1190–1950 (ANU 12155) respectively. The radiocarbon
date from Akatanui Core 1 corresponds to the archaeological chronology established from the
Tangarutu rockshelter (Chapter 11) and suggests that these sediments were deposited during the
early colonisation phase of the island. The age ranges of the two dates from Iri and Anarua are
insufficient to indicate the precise timing of sediment deposition. The sediments found at the
base of the 3 m core from Iri are lagoon sediments rich in marine diatoms (Jan Finn pers. comm.
2005), suggesting that the terrestrial organic sediments had accumulated at the site within the
past 1000 years. An additional 2.5 m core was obtained and radiocarbon dated from an upland
swamp deposit above Anatakuri Bay, at Otaikui (see Figure 10.9). The swamp appears to be
formed behind a landslip and little evidence remains of any agricultural features. The vegetation
of the swamp consists mostly of Schoenoplectus. A fragment of charcoal from the core gave a
modern determination (ANU 12157), suggesting these sediments were recently deposited.
Figure 10.5. Photograph of Akatanui, showing the position of two rockshelters, the extent of abandoned agricultural
terraces and the location of three coring locations (white dots), only one mentioned in the text. See Figure 10.1 for
location of Akatanui Bay. Photograph D.J. Kennett.
terra australis 37
182 Matiu Prebble and Atholl Anderson
Figure 10.6. Sediment lithology and magnetic susceptibility measurements (K) of the Akatanui Core 1 are shown.
*marks location of radiocarbon sample AMS ANU-12157 (charcoal).
Figure 10.7. Sediment lithology and magnetic susceptibility measurements (K) of the Iri Core 1 are shown (left). To the
right are photographs of Iri. A. shows the extent of Iri wetland and the position of Hiri Core 1 (white dot). B. shows the
author at the core site among a dense stand of Schoenoplectus subulatus var. subulatus. *marks location of radiocarbon
sample AMS ANU-12157 (charcoal). See Figure 10.1 for location of Iri Bay. Photographs D.J. Kennett and A. Anderson.
terra australis 37
Palaeobotany and the early development of agriculture on Rapa Island
183
Figure 10.8. Sediment lithology and magnetic susceptibility measurements (K) of the Anarua Core 1 are shown. To the
right are photographs of Anarua. A. shows the extent of Anarua wetland and embayment and the position of Anarua
Core 1 (white dot). B. shows the boulder beach at the head of Anarua backed by Pandanus tectorius trees. *marks
location of radiocarbon sample AMS ANU-12157 (charcoal). See Figure 10.1 for location of Anarua Bay. Photographs
D.J. Kennett and A. Anderson.
Figure 10.9. Sediment lithology and magnetic susceptibility measurements (K) of the Angatakuri Core 1 (Otaikui
Swamp) are shown. To the right is a photograph of Angatakuri that shows the extent of the wetland and the position of
the core (white dot). *marks location of radiocarbon sample AMS ANU-12157 (charcoal). See Figure 10.1 for location of
Angatakuri Bay. Photograph D.J. Kennett.
terra australis 37
184 Matiu Prebble and Atholl Anderson
Summary
The multiple profile approach revealed an uneven response of pollen deposition to sedimentary
influxes, whereby pollen zones could be identified but were represented in differing size units.
What is more striking are the direct indications for post-contact introduction of a range of seral
weed taxa (Prebble unpublished data). The initial presence of these taxa corresponds to distinctive
pollen zones that extend to the modern marshland surface. On Rapa, the earliest anthropogenic
pollen zone, described as a Polynesian phase, is characterised by C. esculenta pollen, a suite of
other potential introduced taxa (e.g. Aleurites moluccana and Erythrina variegata), the onset
of high charcoal particle concentrations, declines in arboreal taxa including the extirpation or
extinction of Arecaceae palm pollen, and an associated increase in fern, grass and other herb
taxa. The upper European phase palynofacies are characterised by a decline of C. esculenta pollen,
further local declines or extirpations of arboreal taxa (e.g. Pandanus tectorius) and the initial
presence and subsequent increase of a range of introduced weed taxa (e.g. Ludwigia octovalvis,
Commelina diffusa and Sonchus oleraceus). In addition, high charcoal particle concentrations
continue with a further increase in fern, grass and other herb taxa. Such a distinctive pattern of
pollen and spore deposition has only been found in a few other sites in island Remote Oceania
(e.g. Kawainui Marsh, O’ahu, Ward in Hammatt et al. 1990; Alenaio Swamp, Hawai’i, Wickler
and Ward 1992; Maunawili Swamp, O’ahu, Athens and Ward 1997).
Colocasia esculenta at initial colonisation?
We have noted a number of problems in the examination of agricultural swamp profiles from
back-beach deposits on Polynesian high islands. The mixing of swamp profiles during the
digging and planting process may have resulted in the mixing of organic materials, possibly
generating older than expected radiocarbon determinations. This issue could be resolved by
further high-resolution radiocarbon dating. Despite these problems, there is great potential for
understanding the timing of crop introduction and the rate of expansion using the multiple
profiling method employed in this study. It is by no means conclusive, but Colocasia may have
been planted at Tukou at the time of, or soon after, initial human arrival. The swamp appears to
have been planted from near the shoreline to the hillside embankment, covering an area of more
than 2 ha. Colocasia pollen is continuously represented up to the period of European contact,
when the swamp was either infrequently used or overrun by feral animals. Prebble (unpublished
data) has identified hydrological changes on the swamp from invertebrate and seed assemblages
that may represent the expansion of terraced pondfields.
The work at Tukou Swamp suggests that early deployment of agriculture within the
colonisation era in places where it was introduced at that time may not be too difficult to find,
contrary to the pessimism expressed by Addison (2008:151):
Finding archaeological evidence for the sequence of agricultural development … will be challenging.
The earliest use of natural wetlands for raised-bed cultivation is unlikely to have left any archaeological
signature. Early pondfield systems are likely to have been small, but they would have been built in
the easiest areas – areas that have probably seen relatively constant use since initial colonization. This
subsequent use will have obscured most traces of the earliest pondfields.
We agree that there are problems in untangling and dating evidence from sites that have
often been used repeatedly, perhaps continuously, but taking an archaeological approach to
sedimentary history coupled with a multiple profile coring method may help to resolve these.
Our work shows that the natural wetlands on Rapa were probably used for Colocasia production
terra australis 37
Palaeobotany and the early development of agriculture on Rapa Island
185
early, extensively and continuously and that the swamp profiles examined hold informative
signatures of agricultural activity. We cannot distinguish raised bed or pondfield systems from the
evidence presented, although hydrological proxies do suggest that pondfields were constructed
later in the sequence. We agree with Addison that Colocasia would have been cultivated in the
easiest areas. These were the Pandanus swamp forests that were easily and rapidly cleared by fire,
creating an immediate soil horizon suitable for crop production.
References
Addison, D.J. 2008. The changing role of irrigated Colocasia esculenta (taro) on Nuku Hiva, Marquesas
Islands: from an essential element of colonization to an important risk-reduction strategy. Asian
Perspectives 47:139–155.
Anderson, A.J. 1994. Paleoenvironmental evidence of island colonization: A response. Antiquity
68:845–847.
Anderson, A.J. 1995. Current approaches in East Polynesian colonization research. Journal of the
Polynesian Society 104:110–132.
Anderson, A.J. 2003. Initial human dispersal in Remote Oceania: pattern and explanation. In: Sand,
C. (ed), Pacific archaeology: assessments and prospects, 71–84. Noumea: Service des Musées et du
Patrimoine.
Anderson, A.J. 2009a. The rat and the octopus: initial human colonization and the prehistoric
introduction of domestic animals to Remote Oceania. Biological Invasions 11:1503–1519.
Anderson, A.J. 2009b. Changing archaeological perspectives upon historical ecology in the Pacific
islands. Pacific Science 63:747–757.
Anderson, E. 1952. Plants, man, and life. University of California Press.
Ash, J. and Ash, W. 1984. Freshwater wetland vegetation of Viti Levu, Fiji. New Zealand Journal of
Botany 22:377–391.
Athens, J.S. and Ward, J.V. 1993. Environmental change and prehistoric Polynesian settlement in
Hawai’i. Asian Perspectives 32:205–223.
Athens, J.S. and Ward, J.V. 1997. The Maunawili core: prehistoric inland expansion of settlement and
agriculture, O’ahu, Hawai’i. Hawaiian Archaeology 6:37–51.
Bronk Ramsey, C. 2008. Deposition models for chronological records. Quaternary Science Reviews
27:42–60.
Clark, J.S., Overpeck, J.T., Webb III, T. and Patterson III, W.A. 1986. Pollen stratigraphic correlation
and dating of barrier-beach peat sections. Review of Palaeobotany and Palynology 47:145–168.
Conte, E. and Payri, C. 2006. Present-day consumption of edible algae in French Polynesia: A study of
the survival of pre-European practices. Journal of the Polynesian Society 115:77–94.
Davidson, J. and Leach, F. 2001. The strandlooper concept and economic naivety. In: Clark, G.R.,
Anderson, A.J. and Vunidilo, T. (eds), The Archaeology of Lapita dispersal in Oceania. Terra
Australis 17:115–124.
Davies, J. Rev. 1827. Extracts from the journal of a visit to the islands of Rapa (or Oparo), Raivavai,
and Tupuai etc (July and October, 1927). Quarterly chronicle of transactions of the London
Missionary Society 3:323–332, 353–361.
Dumayne-Peaty, L. and Barber, K. 1998. Late Holocene vegetational history, human impact and pollen
representativity variations in northern Cumbria, England. Journal of Quaternary Science 13:147–164.
Ellison, J. 1991. The Pacific palaeogeography of Rhizophora mangle L. (Rhizophoraceae). Botanical
Journal of the Linnean Society 105:271–284.
Flenley, J.R. 1979. Stratigraphic evidence for environmental change on Easter Island. Asian Perspectives
22:33–40.
terra australis 37
186 Matiu Prebble and Atholl Anderson
Flenley, J.R. and King, S.M. 1984. Late Quaternary pollen records from Easter Island. Nature 307:47–49.
Flenley, J.R., King, S.M., Jackson, J. and Chew, C. 1991. The Late Quaternary vegetational and
climatic history of Easter Island. Journal of Quaternary Science 6:85–115.
Fosberg, F.R. 1956. Military geography of northern Marshall Islands. Tokyo: Intelligence Division Office
of the Engineer, Headquarters.
Fosberg, F.R. and St. John, H. 1934. Check list and field notebook of the plants of Southeastern
Polynesia: Society Islands, Tuamotus, Austral Islands, Rapa. Honolulu.
Haberle, S. 1994. Anthropogenic indicators in pollen diagrams: problems and prospects for late
Quaternary palynology in New Guinea. In: Hather, J.G. (ed), Tropical archaeobotany: Applications
and new developments, 172–201. Routledge, London.
Hall, J.V. 1869. On the island of Rapa. Transactions and Proceedings of the New Zealand Institute
1:128–134.
Hammatt, H.H., Shideler, D.W., Chiogioji, R. and Scoville, R. 1990. Sediment coring in Kawainui
Marsh, Kailua, O’ahu, Ko’olaupoko. Honolulu.
Hardy, K., Blakeney, T., Copeland, L., Kirkham, J., Wrangham, R. and Collins, M. 2009. Starch
granules, dental calculus and new perspectives on ancient diet. Journal of Archaeological Science
36:248–255.
Horrocks, M. and Bedford, S. 2005. Microfossil analysis of Lapita deposits in Vanuatu reveals
introduced Araceae (aroids). Archaeology in Oceania 40:67–74.
Horrocks, M., Bedford, S. and Spriggs, M. 2009. A short note on banana (Musa) phytoliths in Lapita,
immediately post-Lapita and modern period archaeological deposits from Vanuatu. Journal of
Archaeological Science 36:2048–2054.
Jayalekshmy, A. and Mathew, A.G. 1990. Changes in the carbohydrates and proteins of coconut during
roasting. Food Chemistry 37:123–134.
Jones, R. and Spriggs, M. 2002. Theatrum Oceani: Themes and Arguments concerning the Prehistory
of Australia and the Pacific. In: Cunliffe, B., Davies, H. and Renfrew, C. (eds), Archaeology: the
Widening Debate, 245–294. Oxford University Press for the British Academy, London.
Juggins, S. 2010. C2 Data Analysis. Newcastle.
Kennett, D.J., Anderson, A.J., Prebble, M., Conte, E. and Southon, J. 2006. Prehistoric human
impacts on Rapa, French Polynesia. Antiquity 80:340–354.
Kirch, P.V. 1982. Ecology and the adaption of Polynesian agricultural systems. Archaeology in Oceania
17:1–6.
Kirch, P.V. 2002. On the road of the winds. University of California Press, Berkeley.
Kirch, P.V., Steadman, D.W., Butler, V.L., Hather, J. and Weisler, M.I. 1995. Prehistory and human
ecology in Eastern Polynesia: Excavations at Tangatatau Rockshelter, Mangaia, Cook Islands.
Archaeology in Oceania 30:47–65.
Kirch, P.V. and Green, R.C. 2001. Hawaiki, ancestral Polynesia: an essay in historical anthropology.
Cambridge University Press.
Lamb, W.K. (ed), 1984. George Vancouver, 1803, A Voyage of Discovery to the North Pacific Ocean and
Round the World 1791–1795, Volume I. The Hakluyt Society, London.
Massey, J.E. 1979. The diatoms of contemporary and ancient sediments from Lake Waiau, Hawaii, and
their geochemical environment. Review of Palaeobotany and Palynology 27:77–83.
Matthews, P.J. 1995. Aroids and the Austronesians. Tropics 4:105–126.
Matthews, P.J. 2004. Genetic diversity in Taro and the preservation of culinary knowledge. Ethnobotany
Research and Applications 2:55–71.
McGlone, M.S. and Wilmshurst, J.M. 1999. Dating intial Maori environmental impact in New
Zealand. Quaternary International 59:5–16.
McGlone, M., Wilmshurst, J.M. and Leach, H. 2005. An ecological and historical review of bracken
(Pteridium esculentum) in New Zealand, and its cultural significance. New Zealand Journal of
Ecology 29:165–184.
terra australis 37
Palaeobotany and the early development of agriculture on Rapa Island
187
Meyer, J.-Y. 2002. Rapport de mission de l’expédition scientifique à Raivavae et Rapa (Australes) du 18
Novembre au 20 Décembre 2002. Papeete.
Prebble, M. In press. The palaeobotanical record of Rapa (French Polynesia): phytogeographic
implications for the Austral Archipelago. In: Meyer, J-Y. (ed), Biodiversity of the Austral Islands.
Prebble, M. and Dowe, J.L. 2008. The late Quaternary decline and extinction of palms on oceanic
Pacific islands. Quaternary Science Reviews 27:2546–2567.
Prebble, M. and Wilmshurst, J. 2009. Detecting the initial impact of humans and introduced species on
island environments in Remote Oceania using palaeoecology. Biological Invasions 11:1529–1556.
Prebble, M.P., Anderson, A. and Kennett, D.J. In press. Forest clearance and agricultural expansion on
Rapa, Austral Archipelago, French Polynesia. The Holocene.
Selling, O.H. 1948. Studies in Hawaiian pollen statistics, Part III: on the late Quaternary history of the
Hawaiian vegetation. Vol. 39. Honolulu.
Shineberg, D.E. 1986. Archibald Menzies’ account of the visit of the Discovery to Rapa and Tahiti, 22
December 1791–25 January 1792. Pacific Studies 9:59–102.
Smith, A.C. 1979. Flora Vitiensis Nova. Vol. 1. Pacific Tropical Botanic Garden, Lawai.
Stokes, J.F.G. n.d. Ethnology of Rapa. B.P. Bishop Museum, Honolulu.
Stone, B.C. 1988. Notes on the genus Pandanus (Pandanaceae) in Tahiti. Botanical Journal of the
Linnean Society 97:33–48.
Valentin, F., Buckley, H.R., Herrscher, E., Kinaston, R., Bedford, S., Spriggs, M., Hawkins, S. and
Neal, K. 2010. Lapita subsistence strategies and food consumption patterns in the community of
Teouma (Efate, Vanuatu). Journal of Archaeological Science 37:1820–1829.
Wickler, S. and Ward, J.V. 1992. Archaeological and paleoenvironmental investigations of Alenaio stream
flood control project, Hilo, Hawai’i Island. Honolulu.
Wilson, J., Hardy, K., Allen, R., Copeland, L., Wrangham, R. and Collins, M. 2010. Automated
classification of starch granules using supervised pattern recognition of morphological properties.
Journal of Archaeological Science 37:594–604.
terra australis 37
terra australis 37
11
A Bayesian AMS 14C chronology for
the colonisation and fortification of
Rapa Island
Douglas J. Kennett
Department of Anthropology, The Pennsylvania State University, University Park, PA, USA, djk23@psu.edu
Brendan J. Culleton
Department of Anthropology, The Pennsylvania State University
Atholl Anderson
Department of Archaeology and Natural History, The Australian National University
John Southon
Department of Earth Sciences, University of California, Irvine
Introduction
One of our primary objectives on Rapa was to establish a settlement chronology. Although we
encountered a range of archaeological sites and cultural features during our short time on the
island, our chronological work was focused on determining the likely age of colonisation, with
investigations of coastal rockshelters, and the age of hilltop fortifications and their proliferation
(Figure 11.1). Heyerdahl’s Norwegian expedition to Rapa occurred during the 1950s, shortly
after Libby’s breakthrough development of radiocarbon dating. Early applications of radiocarbon
dating in archaeology were often limited, and in this tradition, Mulloy (1965:59) acquired two
radiocarbon dates from the fortified site of Morongo Uta. One date of 310 ± 300 bp (bp =
uncalibrated result) came from a hearth in the middle of a centrally located terrace (Enclosure
2) and a second date of 210 ± 200 bp was obtained from the ‘floor level’ of another terrace on
the eastern side of the site (Enclosure 85). These dates were considered to be associated with
the last phase of occupation of this fortification, rather than the age of first construction, but
when the dates are calibrated they span AD 1260–1500 and AD 1400–1950, respectively. The
results place the known age of fortification on the island in the later stages of East Polynesian
prehistory, but they are of limited use due to the small sample size, lack of analytical detail and
low precision.
terra australis 37
190 Douglas J. Kennett, Brendan J. Culleton, Atholl Anderson and John Southon
Figure 11.1. Locations of archaeological sites with AMS radiocarbon dates mentioned in text. Drafted by R. Van
Rossman. 1 = Ororangi (R-20); 2 = Ngapiri (R2002-50); 3 = Tevaitau (R-18); 4 = Morongo Uta (R-1); 5 = Pukutaketake
(R2002-42); 6 = Tangarutu; 7 = R2002-49; 8 = Noogurupe (R2002-43); 9 = Kapitanga (R-5); 10 = Ruatara (R-17); 11
= Vairu (R-3); 12 = Angairao, Shelter E; 13 = Taua (R2002-40); 14 = Pukumia (R2002-39); 15 = Akatanui, Shelter 3; 16
= Pukutai (R-19); 17 = Tapitanga (R-4); 18 = Taga Rockshelter; 19 = Potaketake (R-2); 20 = R2002-48; 21 = R200211; 22 = R2002-16; 23 = R2002-20.
The Norwegian expedition focused on the enigmatic fortifications on Rapa and not the age
of colonisation. More recent work by Walczak (2001) identified the Tangarutu rockshelter as
a possible early site. From deposits exposed in the cave he obtained two radiocarbon dates, of
330 ± 45 bp (Ly-8578) and 495 ± 40 bp (Ly-8577), which calibrate to AD 1460–1670 and AD
1400–1500, respectively. These dates overlap in age with Mulloy’s dates from Morongo Uta and
are quite late in the East Polynesian sequence.
terra australis 37
A Bayesian AMS 14C chronology for the colonisation and fortification of Rapa Island
191
In this chapter, we describe the results of our 14C dating program. We AMS 14C dated
materials from stratigraphic excavations at various coastal rockshelters and fortifications; the
contextual information for these dates is described elsewhere in this volume. In this chapter,
we combine these data with additional AMS 14C dates on material collected opportunistically
from exposed stratigraphic deposits and test probes (augers and 25 cm circular sample test pits)
on fortifications, terraces and ovens. We use this AMS 14C survey technique in combination
with Bayesian statistical analysis to establish a chronology for broader patterns of landscape
use during the colonisation and fortification process. These observations will be valuable for
defining future work on Rapa.
Sample selection and analytical methods
We analysed 25 samples from four rockshelters, 30 samples from 10 different fortifications
and 10 additional samples from other types of contexts (e.g. ovens, domestic terraces). All of
the radiocarbon samples from the four coastal rockshelters come from excavations detailed in
Chapter 3. Datable materials were reasonably well preserved in these contexts and carbonised
plant remains were recovered from most stratigraphic units of interest. This was particularly
the case at Tangarutu rockshelter, where preservation was exceptional and artefacts made of
perishable materials were recovered (e.g. fish hooks, cordage). Carbonised wood fragments
were most commonly recovered, but we radiocarbon dated carbonised seeds and smaller twigs
preferentially if they were available (see discussion below regarding the old wood problem).
Obtaining datable materials from fortifications and other open-air sites was more
challenging. The volcanic soils on the island are highly acidic and do not favour the preservation
of organic materials. Hilltop fortifications are the most exposed site types and therefore
highly susceptible to water erosion and wind deflation, both substantial in Rapa’s subtropical
location. Compared with rockshelters, fortifications are also larger and more difficult to assess
with the limited excavations that our time allowed on Rapa. Regardless of these challenges,
we encountered pockets of organic rich sediment containing artefacts, faunal materials and
carbonised materials for radiocarbon dating on the terraces surrounding the central tower at
each large fortification. Some of the deepest and richest midden soils were encountered on the
flattened tops of the central towers themselves. We selected the richest areas for larger excavations
and the details of this work are provided in Chapter 12. We also used smaller test units (augers
and 25 cm circular sample test pits or STPs) to explore the distribution of midden soils across
these terraces. Carbonised plant materials were usually encountered and samples were selected
from the deepest intact deposits, which were often only 25 cm to 30 cm in depth. Analysis of
these samples was limited by funding, but we tried to analyse at least two samples from each
fortification. While visiting the hilltop fortifications, we encountered other domestic terraces
on ridges and agricultural terraces in valley bottoms and surrounding hillslopes. Although these
were not the focus of this project, we sampled them if exposures were present, in order to get an
initial sense of their age.
Most of the AMS radiocarbon dates in this study were prepared chemically at the
University of Oregon’s Archaeometry Facility (now moved to The Pennsylvania State University)
and analysed at the Keck Carbon Cycle Accelerator Mass Spectrometer at the University of
California, Irvine (KCCAMS; methods below). Our study also includes seven radiocarbon dates
run at The Australian National University and three run at the Waikato Radiocarbon Dating
Laboratory (New Zealand). At the ANU facility, all charcoal samples were physically cleaned to
remove adhering sediment, then washed in hot 10% Acid-Base-Acid (ABA), rinsed and dried.
At Waikato, charcoal samples were washed in hot 10% HCl, rinsed and treated with hot 0.5%
terra australis 37
192 Douglas J. Kennett, Brendan J. Culleton, Atholl Anderson and John Southon
NaOH. The NaOH insoluble fraction was treated with hot 10% HCl, filtered, rinsed and dried.
The ANU radiocarbon determinations were made at a time when the laboratory was experiencing
technical difficulties that resulted in several instances of demonstrable discrepancy in results.
As the extent and nature of the problems remains unknown, it would be prudent to treat the
ANU results cautiously. The results from Waikato and KCCAMS are considered reliable. We
also include in our analysis two AMS radiocarbon dates collected during previous excavation at
the Tangarutu rockshelter (Walczak 2001) and prepared at the radiocarbon laboratory in Lyon,
France, using comparable preparation techniques. We exclude Mulloy’s two radiocarbon dates
from Morongo Uta due to a variety of analytical uncertainties (laboratory used, half-life used,
etc) and the high error margins associated with each date.
Methods UO Archaeometry and UCI-AMS
The carbonised plant material analysed at KCCAMS was selected under a microscope to
minimise errors associated with old wood (e.g. by selecting twigs or seeds; Kennett et al. 2002).
The old-wood problem is thought to be minimal in this context, given the relative absence of
long-lived tree species (no older than 20-year inbuilt age) on the island during the interval of
interest (see Chapter 10). Sediments adhering to the samples were removed manually, then
soaked in a series of acid/base/acid (ABA) baths (1 N HCL and NaOH for 30 minutes each).
The HCL removes carbonate contamination and the NaOH extracts humic acid contamination
signalled by discolouration of the solution. Base washes were continued until the solution was
clear, indicating the near absence of contaminating humic acids.
A final acid wash removed secondary carbonates that could have formed during the base
treatment. Samples were then returned to neutral pH with two 15 minute baths in Nanopure
water at 70°C to remove chlorides, and dried on the heater block. Sample CO2 was produced by
combustion at 900°C for six hours in evacuated sealed quartz tubes using a CuO oxygen source
and Ag wire to remove sulfur and chlorine compounds. Sample CO2 was reduced to graphite at
550°C using H2 and a Fe catalyst, with reaction water drawn off with Mg(ClO4)2. Solid graphite
samples were pressed into AMS cathodes and loaded onto the target wheel with OX-1 (oxalic
acid) and other known-age standards, together with calcite and Queets wood blanks for AMS
analysis. Radiocarbon ages were δ13C-corrected for mass dependent fractionation with measured
values (Stuiver and Polach 1977).
Calibration and phase modelling
All AMS 14C dates were calibrated with OxCal 3.10 (Bronk Ramsey 1995, 2001, 2005), using
the Southern Hemisphere atmospheric curve for terrestrial samples (McCormac et al. 2004).
We assumed, based on ethnohistoric records, that each date preceded the Mission Period, so
each probability distribution was truncated using AD 1825 as a terminus ante quem model
in OxCal. The resulting 1-σ (68.2%) and 2-σ (95.4%) calibrated ranges are reported with
probability distributions given for discontinuous distributions. In a few cases, discontinuous
ranges containing small gaps (e.g. five to 10 years) were collapsed into a single interval for clarity.
The chronology of Rapa’s colonisation and settlement expansion underpins the studies
of resource use, landscape transformation and social change presented in this volume. This
complex settlement history unfolded within a relatively short period (ca. 800 years), during
a time of pronounced fluctuations in atmospheric 14C production, which leads to broad and
discontinuous calibrated age ranges (Blackwell et al. 2006). The timing and pace of key events on
Rapa (e.g. the initial establishment and subsequent proliferation of fortifications) are therefore
difficult to know precisely. To constrain these possible dates, we used Bayesian models provided
terra australis 37
A Bayesian AMS 14C chronology for the colonisation and fortification of Rapa Island
193
by OxCal to estimate the beginning and end of five overlapping phases in the settlement history:
Initial Colonisation; Expansion of Rockshelter Use; Initial Fortification; First Expansion of
Fortifications; and Final Expansion of Fortifications. The Boundary command essentially
proposes an event that has not been directly dated – the first fire built in a rockshelter, for
example – and estimates a probability distribution for its occurrence based on the known dates
included in the phase (Bronk Ramsey 2000, 2005). The resulting start and end dates for each
phase are presented with 1-σ and 2-σ ranges, similar to calibrated 14C dates.
Table 11.1. Radiocarbon dates from Rapa Island.
Lab #
Provenience
Site type
Material
UCIAMS-2197
Tangarutu, E2, 123 cm
rockshelter
ANU-11849
Tangarutu, NS1, base layer
UCIAMS-14771
14
C
Error
charcoal
465
25
rockshelter
charcoal
570
70
Tangarutu S3, 112 cmbs
rockshelter
charcoal
600
15
ANU-11847
Tangarutu, base
rockshelter
charcoal
650
100
ANU-11848
Tangarutu, E2, Spit 23–25
rockshelter
charcoal
710
70
UCIAMS-14769
Tangarutu T1, 150 cm
rockshelter
charcoal
905
20
ANU-11850
Tangarutu, Quarry Pit Core 1, 158–178
rockshelter
charcoal
1020
180
UCIAMS-2325
Tangarutu, E2, 10 cm
rockshelter
charcoal
380
25
ANU-11924
Tangarutu, E2, Spit 2
rockshelter
charcoal
440
60
UCIAMS-14726
Tangarutu, E1, Spit 4
rockshelter
gourd
320
15
UCIAMS-14768
Tangarutu E2, 23–25 cm
rockshelter
charcoal
475
20
UCIAMS-14772
Tangarutu, E1, Spit 11
rockshelter
Aleurites sp.
345
20
ANU-12100
Tangarutu, E2, Spit 18 (Lower Part of E2)
rockshelter
Aleurites moluccana endocarp.
380
60
ANU-12102
Tangarutu, E2, Spit 19 (Lower Part of E2)
rockshelter
Pandanus, cf tectorius key
660
150
UCIAMS-14770
Tangarutu V1, 90 cmbs
rockshelter
charcoal
350
15
ANU-12101
Tangarutu, E1, Spit 4 (upper part of E1)
rockshelter
Aleurites moluccana endocarp.
410
60
Ly-8577
Tangarutu, Unit 1, Walczak 2001
rockshelter
charcoal
495
40
Ly-8578
Tangarutu, Unit 2, Walczak 2001
rockshelter
charcoal
330
45
UCIAMS-14765
Akatanui, Shelter 3, C1, A1, Spit 2
rockshelter
Aleurites sp.
385
15
ANU-11925
Akatanui, base level, 30–40 cm
rockshelter
charcoal
480
70
UCIAMS-14763
Akatanui, C1, Spit 4, 30–40 cm
rockshelter
charcoal
610
15
ANU-11851
Angairao E, 2nd Oven, Spit 10, 140–150 cm
rockshelter
charcoal
500
50
UCIAMS-14767
Angairao, Shelter E, 2nd oven, Spit 11, 150–160 cm
rockshelter
charcoal
375
15
UCIAMS-14766
Angairao, Shelter E, 2nd oven, Spit 11, 150–160 cm
rockshelter
charcoal
220
20
ANU-11923
Taga, Test pit A, Spit 2, 20 cm (base of deposit)
rockshelter
charcoal
370
150
UCIAMS-14755
Morongo Uta (R-1), Exposure 4, 18 cm
fortification
charcoal
380
20
UCIAMS-2178
Morongo Uta (R-1), Exposure 2, 10 cm (Tower)
fortification
charcoal
350
20
UCIAMS-2177
Morongo Uta (R-1), Exposure 1, 20 cm
fortification
charcoal
145
20
UCIAMS-14773
Morongo Uta (R-1), Exposure 3, 10 cm
fortification
charcoal
130
20
UCIAMS-47240
Moronga Uta (R-1), 8 cmbs, Exposure 4
fortification
charcoal
210
15
Continued on next page
terra australis 37
194 Douglas J. Kennett, Brendan J. Culleton, Atholl Anderson and John Southon
Table 11.1. continued
Lab #
Provenience
Site type
Material
UCIAMS-14762
Ruatara (R-17), Exposure 1, 14 cmbs
fortification
UCIAMS-14774
Ruatara (R-17), STP#2, 15 cm, terrace below tower
UCIAMS-47243
14
C
Error
charcoal
345
15
fortification
charcoal
210
15
Ruatara (R-17) Exposure 3, 8 cmbs
fortification
charcoal
630
15
UCIAMS-47244
Ruatara (R-17), Exposure 2, 11 cmbs
fortification
charcoal
170
15
UCIAMS-2181
Potaketake (R-2), Unit 1, Feature 2, 10 cm
fortification
charcoal
210
25
UCIAMS-2188
Potaketake (R-2), Unit 1, Feature 3, 30 cm
fortification
charcoal
240
25
UCIAMS-2184
Potaketake (R-2), Unit 1, Feature 4, 51 cm
fortification
charcoal
240
20
UCIAMS-14757
Kapitanga (R-5), below tower, Exp. 3, 33 cmbs
fortification
charcoal
240
15
UCIAMS-14758
Kapitanga (R-5), upper terrace, Exp. 4, 35 cmbs
fortification
charcoal
195
15
UCIAMS-14760
Pukutaketake (R2002-42), STP#2, 19 cm
fortification
charcoal
145
15
UCIAMS-14759
Pukutaketake (R2002-42), STP#2, 35 cm
fortification
charcoal
235
15
UCIAMS-2190
Ororangi (R-20), Unit 1, Feature 1, RC-3, 12 cm
fortification
charcoal
200
25
UCIAMS-2182
Ororangi (R-20), Unit 1, RC-2, 60 cm
fortification
charcoal
185
20
UCIAMS-80840
Tevaitau (R-18), Unit 1, west side, 8 cmbs
fortification
single nut
240
20
UCIAMS-2186
Tevaitau (R-18), Unit 1, Feature 1, 20–30 cm
fortification
charcoal
195
20
UCIAMS-80841
Tevaitau (R-18), Unit 2, Feature 2, 20 cmbs
fortification
single charcoal
185
20
UCIAMS-2187
Tevaitau (R-18), Terrace E, Exp. 2, 22 cm
fortification
charcoal
140
30
UCIAMS-14725
Vairu (R-3), tower, Auger 7, 5–10 cm
fortification
charcoal
190
20
UCIAMS-14761
Vairu (R-3), Exp 1, 25 cm
fortification
charcoal
180
15
UCIAMS-2180
Tapitanga (R-4), Unit 1, Stratum I/II, S1, 20 cm
fortification
charcoal
145
25
UCIAMS-2179
Tapitanga (R-4), Unit 1, Feature 2, 35 cm
fortification
charcoal
140
20
UCIAMS-14756
Noogurupe (R2002-43), Exposure 1, 20 cm
fortification
charcoal
120
15
UCIAMS-47241
Noogurupe (R2002-43), STP#2, 25 cmbs
fortification
charcoal
615
15
UCIAMS-60739
Noogurupe (R2002-43), STP# 3, 15–25 cmbs
fortification
single charcoal
230
25
UCIAMS-60740
Noogurupe (R2002-43), STP#1, 10–19cmbs
fortification
single charcoal
140
25
UCIAMS-36955
Ngapiri (R2002-50), STP #1, 15cmbs, Rapa
small fortification charcoal
205
25
UCIAMS-36951
Pukumia (R2002-39), STP 3, 11cmbs, Rapa
small fortification charcoal
185
25
UCIAMS-36952
Pukutai (R-19), Auger 3, 18 cmbs, Rapa
small fortification charcoal
195
25
UCIAMS-46300
Taua (R2002-40), FS 114, Exposure 1, 25 cmbs
small fortification charcoal
250
25
UCIAMS-47242
R2002-49, STP#1, 18 cmbs
domestic terrace
charcoal
380
20
UCIAMS-36954
R2002-49, Exp. 1, E. End, 37 cmbs
domestic terrace
charcoal
485
25
Wk-14889
Ha’urei, R2002-20, Aitoke oven, road cut
oven
charcoal
673
41
Wk-14887
Ha’urei, R2002-11, Tukou oven, road cut
oven
charcoal
677
41
Wk-14888
Ha’urei, R2002-16, Tukou terrace, fill
agricultural
terrace fill
charcoal
338
61
UCIAMS-36953
Ha’urei, R2002-48 , Exp 1, 50 cmbs, base
domestic terrace
charcoal
200
25
terra australis 37
A Bayesian AMS 14C chronology for the colonisation and fortification of Rapa Island
195
Uncalibrated AMS radiocarbon results
Four coastal rockshelters produced cultural materials that were AMS radiocarbon dated between
1020 ± 180 and 220 ± 20 bp (Table 11.1). Tangarutu is located on the western side of the island
at Anarua Bay, and Akatanui and Angairao are positioned on the northern and eastern side of
the island, respectively, with Taga on the eastern slope of Ha’urei Bay just below the Tapitanga
fortification. Tangarutu was the largest rockshelter and was the most promising as an early
colonisation site given its size and stratigraphic complexity (see Chapter 3). Large sections of the
remaining deposits at the site were exposed by the more recent mining of sand from this locality
to produce concrete for roads, and Walczak (2001) reported two ages from these exposures at
the site; 330 ± 45 and 495 ± 40 bp.
Multiple test pits at Tangarutu were excavated and the stratigraphic sequence is complex,
with mixing occurring in some parts of the cave. Basal deposits date to between 1020 ± 180 and
465 ± 25 bp. The upper part of the sequence at Tangarutu produced ages between 320 ± 20 and
475 ± 20 bp and other ‘mid-sequence’ AMS radiocarbon ages range between 330 ± 40 and 660
± 40 bp. A majority of ages at Tangarutu falls between 300 and 500 bp.
The deposits at the remaining rockshelters were more ephemeral, indicating less extensive
use or a single occupation. Three dates from the Akatanui shelters range between 610 ± 15 and
385 ± 15 bp, and a similar range (500 ± 50 and 220 ± 20 bp) was identified in the Angairao
shelter. The thin deposits in the Taga shelter produced an age of 370 ± 150 bp.
We also obtained radiocarbon dates on 10 of the primary fortifications surrounding Ha’urei
Bay. These dates range between 630 ± 15 and 120 ± 15 bp. Based on these data, the two earliest
fortifications occupied were Ruatara and Noogurupe. Ruatara has a commanding view over the
rich agricultural lands at the head of Ha’urei Bay and three (Piriati, Akatamiro, Tupuaki) of the
five smaller bays along the north coast of the island. Noogurupe was identified by John Stokes
(Stokes n.d.) as on the western side of the island overlooking Ha’urei Bay to the east and Anarua
Bay to the west, the latter containing Tangarutu Cave. The earliest date at Ruatara is 630 ± 15
bp, and additional dates from domestic terraces at the site suggest habitation until 170 ± 15
bp. The earliest date at Noogurupe is 615 ± 15 bp, essentially equivalent to the earliest date at
Ruatara. Additional dates at Noogurupe similarly suggest persistent settlement through to 120 ±
15 bp. Relatively early and persistent settlement is evident at the Morongo Uta (350 ± 20 to 130
± 15 bp) fortification that overlooks Iri Bay to the west and Ha’urei Bay to the east. This is one
of the largest fortifications on the island and was the focus of the Norwegian expedition (Mulloy
1965). The remaining seven fortifications surrounding Ha’urei Bay date to between 250 and
150 bp. In fact, all 10 of the fortifications have 14C dates in this range and present challenges
with respect to calibration and chronology building on the island (see below).
Bayesian chronology
A total of 65 AMS 14C dates form the foundation of our provisional chronology for the
colonisation and fortification of Rapa. Much more archaeological work in the future will
be required to test and improve this chronology. We take a Bayesian statistical approach to
chronology building that takes into account knowledge about the context of AMS radiocarbon
dates (e.g. stratigraphic sequencing) and other a priori knowledge to establish the likely ages
(represented by probability distributions) of a series of developments on the island that include:
(1) initial colonisation, (2) coastal expansion, (3) initial fortification, (4) first expansion
of fortification, and (5) final expansion of fortification (Figure 11.2). These changes in the
distribution and character of settlement are modelled in a series of overlapping phases using the
program OxCal (see methods).
terra australis 37
196 Douglas J. Kennett, Brendan J. Culleton, Atholl Anderson and John Southon
Figure 11.2. Results of Bayesian analysis of radiocarbon dates from Rapa, showing the beginning and end of five phases,
starting with initial colonisation and ending with the final phase of fortification before the Mission Period at AD 1825.
The dates within phases were not ordered stratigraphically because these samples often come
from different archaeological sites or widely separated test units that do not allow correlation.
Stratigraphic relationships between dates are explored in the chapters focused on the excavation
of fortifications and coastal rockshelters. We used the historically recorded abandonment of
fortified settlements during the Mission Period (AD 1825) to constrain the distribution of the
final two fortification phases. We return to this complex question below, but adding this a priori
knowledge to the model helps resolve several issues related to the poorly behaved calibration
curve between AD 1600 and the present.
The initial colonisation phase is represented by a set of dates from Tangarutu rockshelter.
All of these dates come from the basal deposits at this site and these levels represent the earliest
known cultural materials on the island. The earliest of these dates comes from a small test probe
below the base of a modern sand mining pit in the middle of the cave. The large error margins
associated with this date from the ANU facility span 800 years between AD 600 and AD 1400
(Figure 11.3, unfilled outline). This is constrained by the other available dates for this phase,
between AD 1000 and AD 1400, with peak probability at ca. AD 1000 and AD 1400. It is
consistent with the earliest precise AMS 14C date (UCIAMS-14769) on comparable deposits,
of AD 1100 to AD 1200. That is consistent, in turn, with the palaeoecological study in which
erosion of likely anthropogenic origin has a similar age (Chapter 10; Kennett et al. 2006). The
phase boundary for initial colonisation is modelled between ca. AD 800 and AD 1300, with
peak probability between AD 1100 and AD 1200. The phase boundary ends between AD 1300
and AD 1600, and peaks between AD 1400 and AD 1500. The fairly wide age distributions for
these phase boundaries, compared with later phases, is related to the smaller number of dates
used to define the phase.
terra australis 37
A Bayesian AMS 14C chronology for the colonisation and fortification of Rapa Island
197
End
ANU-11850
UCIAMS-14769
ANU-11848
ANU-12102
ANU-11847
UCIAMS-14771
ANU-11849
Begin
500
1000
1500
Years (AD)
2000
Figure 11.3. Phase modelling results
showing probability distributions of
radiocarbon dates constraining the early
colonisation phase. Prior distributions
(routine calibration) are shown in outline,
and posterior distributions (modelled in a
phase) are solid.
The number of coastal rockshelters with evidence of human use increases between AD 1400
and AD 1600 (Figure 11.4). A majority of the dated components at the Tangarutu rockshelter
comes from this interval, indicating more intensive use of this location in conjunction with
expanded use of rockshelters elsewhere on the island. Three other rockshelters (Akatanui,
Angairao and Taga) show clear signs of human use during this interval and Akatanui was used
into the 1700s (UCIAMS-14766), and also later than that, as indicated by the incorporation
of European material in the upper stratigraphy of the site (Chapter 3). Taga is one of the few
small rockshelters positioned along the mid-slopes of Ha’urei Bay. The deposits were ephemeral
and we ran only one exploratory date. It has a high error margin (ANU-11923), but this has
been constrained in our model, based on the other dates from the rockshelters in this phase, to
between ca. AD 1400 and AD 1600. This is consistent with a handful of AMS dates on other
site types (e.g. ovens, terraces) dating to between AD 1200 and AD 1800 along the edge of
Ha’urei Bay (Figure 11.5).
Within the context of expanding coastal settlement, we see evidence for elevated settlements
(domestic terraces) on ridgelines and the first establishment of fortifications at Noogurupe and
Ruatara between ca. AD 1300 and AD 1400 (Figure 11.6), approximately 200 years after the
first evidence for island settlement. We suspect there is a historical connection between the
people who first occupied Tangarutu rockshelter, those who settled on the ridgeline overlooking
the pondfields of Anarua Valley, and those who later established the Noogurupe fortification
overlooking Anarua and Ha’urei bays. The contemporary establishment of Ruatara is likely to
have resulted from the fissioning of the founding community at Anarua. Our limited stay on
the island did not allow for a comprehensive inventory or examination of elevated domestic
terraces on ridgelines, although they appear to be fairly common, as noted by Walczak (2001).
More work is required to sort out the chronology of these types of sites, but the R2002-20 site
suggests their use at the same time as fortifications were first established.
At least two phases of expanded fortification occurred on the island before the Mission
Period. Persistent settlement is evident at Ruatara between AD 1400 and AD 1800, and this also
appears to be the case at Noogurupe, with a possible hiatus between AD 1400 and AD 1650.
Morongo Uta, which developed into the largest fortification on the island, was also established
terra australis 37
198 Douglas J. Kennett, Brendan J. Culleton, Atholl Anderson and John Southon
End
UCIAMS-14763
Akatanui
ANU-11851
Angairao
Tangarutu
Ly-8577
ANU-11925
Akatanui
Tangarutu
UCIAMS-14768
Tangarutu
UCIAMS-2197
ANU-11924
Tangarutu
Tangarutu
ANU-12101
Akatanui
UCIAMS-14765
Tangarutu
ANU-12100
Tangarutu
UCIAMS-2325
Angairao
Figure 11.4. (Left) Phase modelling
results showing probability distributions of
radiocarbon dates constraining the coastal
expansion phase. See Figure 11.3 for details
about how the figure is organised.
UCIAMS-14767
ANU-11923
Taga
Tangarutu
UCIAMS-14770
Tangarutu
UCIAMS-14772
Tangarutu
Ly-8578
UCIAMS-14726
Tangarutu
Begin
1000
1500
2000
Years (AD)
Wk-14887
Wk-14889
R2002-11, Oven
R2002-20, Oven
R2002-49, Domestic Terrace
UCIAMS-36954
R2002-49, Domestic Terrace
UCIAMS-47242
Wk-14888
R2002-16, Agricultural Terrace Fill
Taua, Small Satellite Fortification
Ngapiri, Refuge-Small Fortification
R2002-48, Domestic Terrace
Pukutai, Small Satellite Fortification
Pukumia, Refuge-Small Fortification
1000
1500
Years (AD)
terra australis 37
Figure 11.5. (Below) Calibrated AMS
radiocarbon dates for miscellaneous
archaeological sites and features on
Rapa that were not included in the five
modelled settlement phases. The posterior
distributions (solid) are constrained by the
terminus ante quem of AD 1825, based on
ethnohistoric records indicating the loss of
upland settlement by this time.
UCIAMS-46300
UCIAMS-36955
UCIAMS-36953
UCIAMS-36952
UCIAMS-36951
2000
A Bayesian AMS 14C chronology for the colonisation and fortification of Rapa Island
199
End
Tevaitau
UCIAMS-80840
Potaketake
UCIAMS-2184
Potaketake
Kapitanga
UCIAMS-2188
UCIAMS-14757
Pukutaketake
Noogurope
Final Fortification Expansion
UCIAMS-14759
UCIAMS-60739
Morongo Uta
UCIAMS-47240
Ruatara
UCIAMS-14774
Potaketake
Ororangi
Kapitanga
UCIAMS-2190
Tevaitau
UCIAMS-2186
UCIAMS-2181
UCIAMS-14758
UCIAMS-14725
Vairu
Tevaitau
Ororangi
UCIAMS-80841
UCIAMS-2182
Vairu
Ruatara
UCIAMS-14761
UCIAMS-47244
Morongo Uta
UCIAMS-2177
Pukutaketake
UCIAMS-14760
Tanga
Noogurope
UCIAMS-60740
Tevaitau
UCIAMS-2187
Tanga
Morongo Uta
UCIAMS-2179
UCIAMS-2180
UCIAMS-14773
Noogurope
UCIAMS-14756
Begin
End
First Fortification Expansion
Morongo Uta
Morongo Uta
UCIAMS-14755
UCIAMS-2178
Ruatara
UCIAMS-14762
Begin
End
Initial Fortification
Ruatara
Noogurope
UCIAMS-47241
UCIAMS-47241
Begin
800
1000
1200
1400
1600
1800
2000
Years (AD)
Figure 11.6. Phase modelling results showing the probability distributions of radiocarbon dates forming the basis of the first and
final fortification phases on Rapa. See Figure 11.3 for details about how the figure is organised.
terra australis 37
200 Douglas J. Kennett, Brendan J. Culleton, Atholl Anderson and John Southon
during the first fortification expansion phase between AD 1400 and AD 1650. Settlement at
Morongo Uta persisted until the early 1800s.
Details of the final fortification phase are marred by major fluctuations in the calibration
curve after about AD 1700 (Figure 11.7). All radiocarbon ages starting at about 240 +/- 15
bp are impacted by these fluctuations and cannot be distinguished from those on modern
materials. However, we know, based on mission records, that fortified villages were abandoned
by AD 1825 when the modern communities of Area and Ha’urei were established. This a priori
knowledge was built into our Bayesian model and it serves as an important calibration datum
(terminus ante quem) to help clean up the probability distributions. It truncates the probability
distributions at this date and shows a cluster of seven fortifications dating to the early 1700s and
at least 10 fortifications in the late 1700s, when Vancouver observed at least six forts in use on
the western side of the island (Chapter 2). Smaller satellite fortifications (e.g. refuges) that were
probably connected in some way with the larger fortifications nearby also date to the late 1700s
(see Figure 11.5, Tauo, Ngapiri, Pukutai and Pukumia). These sites are discussed more fully in
Chapter 12.
Figure 11.7. Southern Hemisphere atmospheric curve for terrestrial samples (McCormac et al. 2004) between AD
1400 and AD 2000, showing a major plateau and fluctuations between AD 1650 and AD 200 that make calibration
challenging.
terra australis 37
A Bayesian AMS 14C chronology for the colonisation and fortification of Rapa Island
201
Conclusions
The systematic use of radiocarbon dating as a survey technique provides a framework for
reconstructing a broad pattern of landscape use and how this changed through time. The
approach is becoming more commonly used in a variety of coastal and island contexts where
erosional processes expose deposits (Erlandson and Moss 1999; Kennett 2005) and it may be
particularly useful in East Polynesian contexts where ceramics are absent, along with other
temporally diagnostic artefacts used elsewhere to establish preliminary settlement chronologies.
This approach does not serve as a replacement for more intensive stratified excavations, and
intensive archaeological work is clearly required to establish more definitive site chronologies.
On Rapa, we used the survey technique to establish a broad chronology that can guide future
work on the island. Our data suggests that the island was most likely colonised between AD
1100 and AD 1200. This was followed by expanded use of coastal zones and construction of the
first elevated domestic terraces between AD 1200 and AD 1600. It is within this context that
the first two fortified hilltop settlements (Noogurupe and Ruatara) appear on the island between
AD 1300 and AD 1400. The number of fortifications expanded, slowly between AD 1400 and
AD 1650 and then more rapidly during the 1700s just before historic contact.
The need for further research, and quite soon, is underlined by the evidence that many of
the sites on Rapa are endangered by destructive processes. Hilltop fortifications are heavily wind
deflated, and animal grazing has removed much of the vegetation stabilising these sites, leading
to erosion, which is also destroying the rock terrace walls. In addition, the mining of sand for
road construction from the Tangarutu rockshelter presents a serious threat to these deposits.
References
Blackwell, P.G., Buck, C.E. and Reimer, P.J. 2006. Important features of the new radiocarbon
calibration curves. Quaternary Science Reviews 25:408–413.
Bronk Ramsey, C. 1995. Radiocarbon calibration and analysis of stratigraphy: the OxCal program.
Radiocarbon 37(2):461–74.
Bronk Ramsey, C. 2000. Comment on ‘The Use of Bayesian Statistics for 14C dates of chronologically
ordered samples: a critical analysis’. Radiocarbon 42(2):199–202.
Bronk Ramsey, C. 2001. Development of the radiocarbon program OxCal. Radiocarbon 43(2A):355–
363.
Bronk Ramsey, C. 2005. OxCal v. 3.10. http://c14.arch.ox.ac.uk/embed.php?File=oxcal.html.
Erlandson, J.M. and Moss, M. 1999. The systematic use of radiocarbon dating in archaeological
surveys in coastal and other erosional environments. American Antiquity 64(1):431–443.
Kennett, D.J. 2005. The Island Chumash: Behavioral Ecology of a Maritime Society. University of
California Press, Berkeley.
Kennett, D.J., Ingram, B.L., Southon, J.R. and Wise, K. 2002. Differences in 14C Age Between
Stratigraphically Associated Charcoal and Marine Shell from the Archaic Period Site of Kilometer
4, Southern Peru: Old Wood or Old Water? Radiocarbon 44(1):53–58.
Kennett, D.J., Anderson, A., Prebble, M., Conte, E. and Southon, J. 2006. Human Impacts on Rapa,
French Polynesia. Antiquity 80:340–354.
McCormac, F.G., Hogg, A.G., Blackwell, P.G., Buck, C.E., Higham, T.F.G. and Reimer, P.J. 2004.
SHCal04 Southern Hemisphere calibration, 0–11.0kyr BP. Radiocarbon 46(3):1087–92.
terra australis 37
202 Douglas J. Kennett, Brendan J. Culleton, Atholl Anderson and John Southon
Mulloy, W. 1965. The Fortified Village of Morongo Uta. In: Heyerdahl, T. and Ferdon, E.N. (eds),
Reports of the Norwegian Archaeological Expedition to Easter Island and the East Pacific, Volume 2,
miscellaneous papers. Monographs of the School of American Research.
Stokes, J.F.G. n.d. Ethnology of Rapa, Manuscript on file in the Bernice P. Bishop Museum, Honolulu.
Stuiver, M. and Polach, H.A. 1977. Discussion: reporting of 14C data. Radiocarbon 19:355–363.
Walczak, J. 2001. Le peuplement de la Polynésie orientale: Une tentative d’approche historique par les
exemples de Tahiti et de Rapa (Polynésie française). Unpublished Dissertation, University of Paris
(Panthéon Sorbonne).
terra australis 37
12
The archaeology of Rapan fortifications
Douglas J. Kennett
Department of Anthropology, The Pennsylvania State University, University Park, PA, USA, djk23@psu.edu
Sarah B. McClure
Department of Anthropology, The Pennsylvania State University
Introduction
Fortifications were common features in the East and South Polynesian sociopolitical landscape.
The initial appearance of defensive features coincides with the first evidence for political hierarchy
in West Polynesia (Fiji, Tonga and Samoa, Clark and Martinsson-Wallin 2007; Kennett and
Winterhalder 2008) and the competing political systems that they represent. This is roughly
coincident with the East Polynesian expansion, and competition for land in West Polynesia
may have been one contributing factor stimulating exploration and eventual settlement on
increasingly remote islands between AD 800 and 1200. This pulse of migration and settlement
to increasingly remote islands was an extension of the Austronesian expansion and provided
the agroeconomic context that promoted rapid increases in population, territoriality and
competition for limited lands and the periodic use of force to take land and surplus food. It is
within this context in Polynesia (and elsewhere) that people focus greater effort on constructing
defensive features. Cool and dry conditions in the equatorial Pacific during the Little Ice Age
between AD 1450 and 1850 also correlate with a proliferation of fortifications in several parts
of the Pacific (Field and Lape 2010).
Rapan fortifications have been used to exemplify defensive sites in Polynesia and have served
as an end-member of hyper-fortification and intergroup conflict (Kirch 1984). However, the
presence and function of hilltop fortifications on Rapa has been debated since Vancouver (Lamb
1984) reported people living in fortifications on the island in AD 1791. Ellis (1838) questioned
the mere existence of these fortifications and argued that the sculpted hilltops and terraces were
natural geological features. Stokes (n.d.) dispelled this idea with early ethnoarchaeological work
on the island that involved visiting these locations and talking to the Rapan people about their
history on the island and how the fortifications functioned. The Norwegian expedition provided
the first modern archaeological basis for work on these fortifications and the best descriptions of
terra australis 37
204 Douglas J. Kennett and Sarah B. McClure
these sites in existence. They clearly demonstrated that the largest hilltop sites were settlements
as Vancouver described and that a range of defensive features (ditches, etc.) was consistent with
the idea that they served a defensive purpose. Walczak (2001) has questioned this idea and
suggested that these sites were more ceremonial in nature. Our aim in this chapter is to pull
together the various strands of archaeological data available from previous studies and add our
own observations.
Archaeological observations
We conducted archaeological survey and small-scale excavation at 10 large and four small
fortifications on the island (Figure 12.1). All of the larger fortifications are strategically
positioned on the highest points along the main ridge surrounding Ha’urei Bay. The sites range
in size between 3040 m2 and 25,237 m2 and all of them have a central tower carved from the
soft basalt ridgelines. The smaller fortifications are often found in close proximity to the larger
fortifications and they are considered to be satellite communities or temporary refugia. Two of
the four smaller fortifications have central towers. We divided Rapa into southern, western and
northern sectors to facilitate our descriptions of the available archaeological data and our own
observations. We start with the southern sector because we used the Norwegian expedition’s
excavations at Morongo Uta to index other observations made on the island during the past
century.
Southern sector
Three hilltop fortifications are spaced equidistantly along the primary ridge in the island’s
southern sector (Figure 12.2). Morongo Uta (R-1, 258 m) overlooks the western side of Ha’urei
Bay to the east and west across the rich agricultural lands in the drainage associated with Iri Bay.
Tevaitau (R-18, 262 m) is located at a lower elevation along the same ridge. Today it overlooks
the modern community of Ha’urei positioned on what would have been highly productive
agricultural lowlands. Tevaitau is separated from Ororangi (R-20, 282 m) by a large drainage
and a series of high peaks along the main ridge system.
Morongo Uta is one of the largest and most architecturally complex fortifications on the
island (10,148 m2). It is also the best studied. Stokes (n.d.) worked at the site for three weeks
in 1921, clearing and recording terraces. Based on interviews with local people, he suggested
that the village was originally named Teruta and was occupied by a subdivision of a clan named
Kopogoiki that controlled territory between Tekoki and Pukumanga.
Work at Morongo Uta was a focal part of the 1956 Norwegian expedition (Mulloy 1965).
Using a large crew of local labour (between 12 and 53 people, see Heyerdahl 1958), the site was
cleared of vegetation and excavated between May 21 and June 18, 1956. Controlled excavations
were minimal, but the entire site was cleared to expose terrace floors, and formal artefacts
(e.g. adzes) were collected as the architecture was examined and mapped. Clearing of the site
involved the removal of talus slopes and overburden down to approximately 5–10 cm above
terrace surfaces. More careful excavation of terrace deposits was then carried out with smaller
hand tools. Excavated sediments were not screened for faunal materials or smaller artefacts
and persistent rain during this 28-day period limited identification of subtle features (e.g. post
moulds) and stratigraphic control.
Much of Morongo Uta was covered with vegetation during our visit, but all of the terrace
units mapped by the Norwegian expedition were visible (Figure 12.3). The site core consists of
a tall triangular tower and a surrounding triangular cluster of rectilinear terraces that also points
to the south. The tower was carved from the highest peak on the ridgeline and covered with a
terra australis 37
The archaeology of Rapan fortifications
205
Figure 12.1. Map of Rapa showing the locations of primary and secondary fortifications. The island is divided into
southern, western and northern sectors for data presentation purposes and discussion. Drafted R. Van Rossman. 1 =
Ororangi (R-20); 2 = Ngapiri (R2002-50); 3 = Tevaitau (R-18); 4 = Morongo Uta (R-1); 5 = Pukutaketake (R2002-42);
6 = Noogurope (R2002-43); 7 = Kapitanga (R-5); 8 = Ruatara (R-17); 9 = Vairu (R-3); 10 = Potaketake (R-2); 11 =
Tapitanga (R-4); 12 = Pukutai (R-19); 13 = Pukumia (R2002-39); 14 = Taua (R2002-40).
terra australis 37
206 Douglas J. Kennett and Sarah B. McClure
Morongo Uta (R-1)
Elevation: 258 m
2
Size: 10,148 m
0
UTM: 6942536.954; 761078.4296 E
50
100 m
Tevaitau (R-18)
Elevation: 262 m
2
Size: 3,666 m
0
UTM: 6941795.403 N; 761995.4893 E
50
100 m
Ororangi (R-20)
Elevation: 282 m
2
Size: 5,157 m
1
2
6
3 5
4
Unit 1
7 8
9
11
10
UTM: 6941628.882 N; 763452.8816 E
0
50
100 m
Figure 12.2. IKONOS satellite images of Morongo Uta (R-1), Tevaitau (R-18) and Ororangi (R-20) fortifications. Size
estimates, site boundaries and tower orientation are based on field observations and features (e.g. terraces) visible in
IKONOS imagery. UTM coordinates are from the centres of each tower. The small numbered triangles on the Ororangi
image are the locations of auger tests and the larger rectangle is a 1 m x 1 m test excavation. The locations of test
excavations at Morongo Uta and Tevaitau are shown in figures 12.3 and 12.7. Drafted J. Bartruff and D.J. Kennett.
terra australis 37
The archaeology of Rapan fortifications
207
dyke-stone masonry facade. It has a flattened upper surface measuring 10 m x 5.4 m. Stokes
excavated the upper flattened surface of the tower and documented a stone-lined oven or hearth
in its centre (n.d.:385). At the time, local informants told him that this was the chief ’s residence
and that his closest kin lived in houses positioned on the surrounding upper terrace units.
Mulloy (1965) found this implausible, but the upper surface is quite large (34 m2) and dark
midden soil containing charcoal and adze flakes is exposed on its northeastern edge (Exposure
2). A charcoal sample from this cleaned exposure was AMS radiocarbon dated to between AD
1500 and 1600 (UCIAMS-2178).
Mulloy (1965:61) mapped 89 terraces at the site, including the enclosure at the top of the
tower. The total surface area of these terraces was estimated to be 5406 m2. Large numbers of
terraces radiate out from the site core along the main ridge and down into the Iri Bay drainage.
Dyke-stone walls surround several of the ridgetop terraces and similar walls paralleling the ridge
were used as outer retaining walls which were backfilled to create flat surfaces. Many of these
walls were reinforced with double interlocking masonry walls filled with a rubble core. Some
of the longer terraces were divided into smaller units with walls. The organisation of terraces
becomes less formalised with distance from the site core and Mulloy (1965:23) suggested that
these were added as populations expanded and clans split into subclans. Stokes (n.d.) also
identified additional domestic terraces 300 m to the north of the site core on the same ridgeline
and argued that these satellite houses (or auga) were also part of the same community. Multiple
deep fosses and associated walls cut through the ridgeline at strategic locations point to a large
Figure 12.3. Map of Morongo Uta based
on the Norwegian expedition’s original
map, field observations and IKONOS
satellite imagery. Exposures were cleaned
to examine stratigraphy and collect
radiocarbon samples. Drafted R. Van
Rossman.
terra australis 37
208 Douglas J. Kennett and Sarah B. McClure
labour investment in defence that is consistent with Vancouver’s observations of multiple
wooden palisades surrounding these communities (Lamb 1984). Mulloy (1965) suggested that
some of these features were excavated after terrace units were in place and indicated heightened
aggression later in time.
The Norwegian expedition also documented a number of features and artefacts that provide
insight into the types of activities that took place in these communities. Most floors were
identified as concentrations of charcoal that were discontinuous across the terrace floor. Terrace
walls did not extend above the floor surface and were not designed to support a superstructure.
Rectilinear terraces provided platforms for perishable structures made from wood, grass or
fronds. Postmoulds were identified, but wet conditions prohibited clear delineation of these
structures. Stone-lined cooking pits (ovens) and more formal hearths consisting of three or
four tabular stones were identified on many of the terraces (e.g. enclosures 2, 3, 8, 10, 12, 19,
24, 35, 45, 53, 55, 80, 82, Mulloy 1965, figures 7–11). Several smaller d-shaped terraces were
interpreted as possible storage features (p. 32) and cylindrical pits of different sizes may have
served to store water or taro (see discussion). A group of three prismatic stones driven upright
in Enclosure 18 were interpreted by Mulloy (p. 34) as a possible household shrine. Although
the Norwegian expedition did not screen and collect faunal and floral remains, Mulloy noted
that ‘the evidence of daily debris was plentiful’, and our observations confirm that dark midden
deposits containing charcoal, shell and bone are present at the site. They also collected artefacts
indicating a range of daily activities, including 60 whole or fragmented basalt adzes of different
types and 60 whole or fragmentary poi pounders (Figure 12.4). Poi anvils were identified and left
in place and other types of artefacts were also found in small numbers (circular disk, polishing
stones and other groundstone artefacts). This does not include the full range of perishable
artefacts that must have been used. Mulloy (1965:53) reports charred fragments of a fish net
or netted bag that were identified within a midden in Enclosure 6, but these remains were not
recovered intact. We pulled radiocarbon samples from four different midden exposures and
these dates suggest persistent use of this site between AD 1600 and 1830 (Figure 12.5).
Our work at Tevaitau was more extensive. The Norwegian expedition recorded this site
(R-18), but it does not feature prominently in its 1965 report. It is smaller than Morongo
Uta (3666 m2) and, given its overlapping age (AD 1700–1830) and close proximity, it may
have developed as a politically connected satellite community (Figure 12.6). Tevaitau has only
a distant view of Iri Bay and appears to be more connected visually with the southern part of
Ha’urei Bay and the agricultural lands that are now covered with the modern community of
Ha’urei. Use of the agricultural lands and access to marine resources via Iri Bay would only have
been possible in close consultation with Morongo Uta’s chiefly lineage.
Tevaitau is oriented north–south along the main ridge system and consists of a central tower
and 14 small surrounding domestic terraces. Most of the terraces were constructed along the
ridgeline, but terraces also drop down into the adjacent drainages to the west and east. There is
a steep drop into the Iri Bay drainage to the west and only one significant terrace below the core
group was identified. Several terraces extend down a steep ridge into the drainage to the east.
Morongo Uta is visible to the west and Ororangi is visible to the east. A nice example of dykestone masonry covers the tower and several masonry walls are evident within the core. Several
foss-like cuts occur between terraces to the north and south of the tower, and steep terrain to the
west and east makes the location highly defensible. A free-standing wall divides a long terrace
south of the tower. The site was relatively clear of vegetation and we took the opportunity to
map it carefully with a laser transit (figures 12.6 and 12.7).
terra australis 37
The archaeology of Rapan fortifications
209
Figure 12.4. Representative selection
of adzes and poi pounders collected at
Morongo Uta by the Norwegian expedition.
Redrawn R. Van Rossman.
UCIAMS-2190
Ororangi
UCIAMS-2182
Ngapiri
UCIAMS-36955
UCIAMS-80840
UCIAMS-2186
Tevaitau
UCIAMS-80841
UCIAMS-2187
UCIAMS-47240
UCIAMS-2177
UCIAMS-14773
Morongo Uta
UCIAMS-14755
UCIAMS-2178
1200
1400
1600
1800
2000
Years (AD)
Figure 12.5. Calibrated radiocarbon dates for fortifications in the southern sector based on phase modelling presented
in Chapter 11. Prior distributions (routine calibration) are shown in outline, and posterior distributions (modelled
within a phase) are solid.
terra australis 37
210 Douglas J. Kennett and Sarah B. McClure
Figure 12.6. Tevaitau (R-18) fortification
taken from the south looking north.
Photograph D.J. Kennett.
Figure 12.7. Map of the Tevaitau (R-18)
showing the locations of excavation units
and auger samples.
terra australis 37
The archaeology of Rapan fortifications
211
The site is disturbed by wind and water erosion and there are several looter’s pits present in
different terraces, the largest just to the south of the central tower. We also learned on the island
that Jerome Walczak (2001) had excavated several fortifications in the late 1990s and that he
had spent several days at Tevaitau. Documentation of all these excavations is limited (Endnote
1), but he reports excavating a large portion of the upper tower surface of Tevaitau (10 m2) and
this large open pit is visible today. He reports finding 922 basalt flakes during this excavation,
mostly blanks and adze flakes. These flakes were found in association with wood charcoal, shells
and bone, including a goat bone in the surficial deposits that is clearly intrusive given the late
introduction of this species. We augered (9 cm diameter) the remaining deposits around this test
excavation and identified dark midden deposits that extended from ca.10 cm to 30 cm below
the surface.
Similar midden deposits containing basalt flakes, shell, bone and charcoal were evident in
two natural exposures and in other auger tests across the site. These deposits range from 30 cm
to 80 cm in maximum depth. Based on the density of materials evident in natural exposures
and auger tests, we placed three 1 m x 1 m excavation units into undisturbed deposits on a
large terrace to the east of the central tower. The deposits are shallow in this location (10–25
cm) and the units bottomed out on the artificially flattened terrace below. Dark midden soils
containing charcoal and faunal materials occurred in all these units and there were several small
pits that were excavated into the flattened bedrock terrace below. Feature 1 in Unit 2 consisted
of two upright tabular stones arranged in a similar fashion to the hearths identified by the
Norwegian expedition at Morongo Uta (Figure 12.8). Dark midden soil containing charcoal
and faunal materials surrounded Feature 1 and one large concentration (ca. 20 cmbs) was AMS
Figure 12.8. Planview map of Unit 1
(20 cmbs) at Tevaitau (R-18).
Drafted R. Van Rossman.
terra australis 37
212 Douglas J. Kennett and Sarah B. McClure
radiocarbon dated (UCIAMS-2186) to between AD 1650 and 1750 (see Figure 12.5). A total
of 12 basalt adze flakes were recovered in these shallow excavations.
The sediments excavated in units 1–3 were screened in the field with 1/8th inch mesh.
Once we determined that sediments contained preserved faunal materials, we also took large
bulk samples (six litres in total) for flotation and complete recovery of all faunal and potential
floral material. In this way, we recovered the richest faunal assemblage recovered from an island
fortification, and associated candlenut fragments that were directly AMS dated (UCIAMS-80840)
to between AD 1700 and 1800. The faunal assemblage is composed of nearshore reef fishes
that include parrotfishes (Scaridae), damselfishes (Pomacentridae), wrasse (Labridae), pufferfish
(Tetraodontidae) and moray eel (Muraenidae). The parrotfishes are the best represented and this
may be due to their distinctive character and resistance to decomposition. Rat bones (Rattus sp.)
were also recovered and are consistent with the observations made by the Norwegian expedition
that these hilltop fortifications were also the primary nodes of residence (see Butler, Appendix
C). Three AMS radicocarbon dates from these units and one from Exposure 2 on the south end
of the site point to occupation between AD 1700 and 1830. Given the substantial overlap in
age with the later deposits at Morongo Uta, along with its close and unobstructed proximity, it
may have developed as a satellite community of this larger polity.
To the east of Tevaitau at about the same elevation (282 m) is the Ororangi fortification
(R-20; see Figure 12.1). It is geographically isolated from Morongo Uta and Tevaitau by a
valley that has a steep and impassible ridgeline in its headwaters. The fortification overlooks the
modern community of Ha’urei to the west and it is well connected to the exterior of the island
via Angatakuri Bay to the east. Tapitanga is visible across Ha’urei Bay and Stokes noted two
smaller fortifications in the vicinity (Mititipeiru and Karagarua) and one of these is visible from
Ororangi to the southwest.
Ororangi (5157 m2) is slightly larger than Tevaitau, but substantially smaller than Morongo
Uta. The architects at this site took advantage of a high point in the southwest-to-northwest
oriented ridge to construct an elongated tower. Two impressive upper terraces surround this
elongated tower and together they form the site’s core. A total of 43 terraces were documented
at the site, but vegetation covers large sections of it to the north and west. The largest number
of terraces occurs along the main ridge to the southwest and northeast of the site core. These
terraces are rectilinear along the flat portions of the southwestern terrace and more d-shaped
down a steeper section of the ridge to the northeast. A cluster of six small terraces (each 5 m x
5 m) occurs on the southwestern flank of the site core and small d-shaped terraces drop down
the steep ridges towards Ha’urei Bay. The largest domestic terraces drop down the more gradual
slopes towards Angatakuri Bay.
Ororangi does not have the same types of dyke-stone masonry architecture as Morongo
Uta and Tevaitau. Terraces were cut directly into the ridgelines and cutting and moving the
soft basalt bedrock created flattened surfaces. A sketch map of these terraces was produced and
11 auger samples were taken from across the site (see Figure 12.2). The upper ca. 10 cm of
these auger tests consisted of culturally sterile orange sediment that sometimes contained sparse
amounts of modern material. Below these upper deposits were prehistoric sediments between
10 cm and 40 cm thick. These sediments were darker and contained charcoal and the occasional
adze flake. At the base of these sections, the darker midden transitioned to the underlying soft
basalt bedrock that had been flattened to create the terrace. Midden deposits appear to be best
developed on the second and third terraces to the north of the tower (augers 5 and 6). Work on
the southwestern arm of Ororangi was limited by dense vegetation.
The tower at Ororangi is well preserved and the entire upper surface appears to be intact.
terra australis 37
The archaeology of Rapan fortifications
213
Sediments on top of the tower are dark brown and covered with a thick layer of grass. An auger
test near the centre of the tower suggested intact deposits containing charcoal were between 20
cm and 30 cm deep. A small test unit (Unit 1, 1 m x 1 m) was placed near the northwestern
side of the tower. Some modern debris was found in the upper 5 cm within sediments that
were otherwise culturally sterile. Dark midden soil containing charcoal, fire-cracked rock, adze
flakes and faunal material was encountered at ca. 10 cm below the surface. A small cluster of
fire-cracked rock was encountered in the eastern part of the unit at ca. 10 cm below the surface
(Feature 2) and may be the remnants of an oven. This feature occurred just above the basalt
surface of the terrace that had been flattened prehistorically. A much larger fire feature became
visible at ca. 10 cm in the northern and western portion of Unit 1 (Feature 1, Figure 12.9).
This feature was excavated into the underlying basalt to a maximum depth of ca. 50 cm. The
pit contained dark silty sediment with large quantities of charcoal and fire-cracked rock. This is
similar in character to the fire feature described by Stokes on top of the tower at Morongo Uta.
Most of these sediments were screened in the field (1/8th inch mesh), but a two-litre
flotation sample was taken and examined for floral and faunal remains. Six basalt adze flakes
were recovered, along with a small amount of faunal material, including 21 parrotfish (Scaridae)
bones and one Elasmobranch (shark or ray) bone (see Butler, Appendix C). A radiocarbon
sample taken at the top (UCIAMS-2190) and bottom (UCIAMS-2182) were nearly identical
and put Ororangi contemporary with Tevaitau and the later stages of occupation at Morongo
Uta (AD 1700 to 1830).
Western sector
The western sector of Rapa is rugged and contains the highest mountain peaks on the island
(Perau, 650 m). A ridge running to the south of Anarua Bay and its associated drainage forms
the southern boundary of this sector. Tupuai Bay and its associated drainage roughly delineate
its eastern boundary. Pukutaketake (R2002-42), Noogurope (R2002-43), Kapitanga (R-5) and
Ruatara (R-17) are the four primary fortifications in this area and they are all positioned on
Figure 12.9. Stratigraphy in Unit 1 at Ororangi. Drafted R. Van Rossman.
terra australis 37
214 Douglas J. Kennett and Sarah B. McClure
the main ridge system overlooking the prime agricultural lands at the head of Ha’urei Bay.
The Norwegian expedition mapped Kapitanga and visited Ruatara, but it is unclear from its
reports whether it examined or visited Pukutaketake and Noogurope. All of these fortifications
are clearly identified on Stokes’ map from the 1930s. Pukutaketake and Noogurope overlook
Anarua Bay to the west. Kapitanga and Ruatara overlook Pariati, Akatamiro and Tupuaki bays
along the northern shore. These sites range in size between 3000 m2 and 7000 m2. This was
the least studied part of the island, by us, but our preliminary work suggests that two of the
fortifications, Noogurope and Ruatara, were among the first to be constructed on Rapa. Future
work at Noogorupe and Ruatara should be a high priority.
Noogurope is the highest elevation fortification on the island (608 m), and Perau (650 m),
the tallest peak along the primary ridge system, provides a strong protective barrier blocking
easy passage to and from fortifications to the north and west (e.g. Kapitanga and Ruatara).
All of the other fortifications on the island are visible from this vantage point. Access from
Ha’urei Bay is also limited by a series of rock escarpments and cliffs blocking easy passage. The
fortification is small (3040 m2) and consists of 12 terraces surrounding a small circular tower.
Nicely formed dyke-stone walls were evident along the eastern and western sides of the upper
terrace and a fragmentary segment also occurs along its southern periphery. Unlike Kapitanga
to the northwest, the circular tower at Noogurope is not faced with dyke-stone. The longest
series of terraces extends to the northwest along the ridge leading towards Perau. Terraces also
extend down the subsidiary ridge system to the west towards Anarua Bay and there appear to be
several terraces to the east on the ridge dropping down to Ha’urei Bay. Farther down the ridge
to Ha’urei were several domestic terraces built with dyke-stone walls. These are associated with
a dyke-stone outcrop. A large talus slope is present on the northern side of the outcrop and the
large amounts of debris scattered across the surface are suggestive of quarrying activity, but more
work is needed to establish the extent and character of this site.
At Noogurope we excavated four STPs (test pits) on the uppermost terrace (Figure 12.10).
Charcoal-rich midden soils were encountered in all of these sample test pits and extended down
to between 10 cmbs and 25 cmbs. A charcoal sample collected at 25 cmbs in STP #2 returned
a date of 650 ± 15 bp (UCIAMS-47241). This is the earliest date at Noogurope and the earliest
date associated with a fortification on Rapa. Another date from STP #3 from the basal deposits
(15–25 cmbs) returned an age of 230 ± 25 bp (UCIAMS-60739). We also cleared off a natural
exposure (Exp #1) on the northeastern side of the second northern terrace. A charcoal-rich
midden deposit was defined between 10 cmbs and 37 cmbs, and a charcoal sample recovered
from 20 cmbs returned an age of 120 ± 15 bp (UCIAMS-14756). This is comparable in age
to STP #1 (10–19 cmbs, UCIAMS-60740) and suggests occupation until approximately AD
1830. Overall, the calibrated dates at this site range between AD 1300 and 1830 (Figure 12.11).
Preservation of faunal material was poor in these deposits and only a few adze flakes were
recovered during our small-scale excavations.
The imposing site of Ruatara sits on the tallest mountain peak (304 m elevation) on the
main ridge system between Vairu and Kapitanga. It is the largest (6940 m2) fortification in
the north-central sector and among the largest fortifications on the island. Due to its central
position, it has a commanding view of the prime agricultural lands at the head of Ha’urei
Bay and of three external drainages and associated bays to the north (Akatamiro, Tupuai and
Akanamu). The tower is one of the tallest on the island (5 m) and a step series of four narrow
terraces down its western flank adds to the prominence of the tower (Figure 12.12). Sixteen
terraces in total radiate from the tower along the primary ridge system and down subsidiary
terra australis 37
The archaeology of Rapan fortifications
Pukutaketake (R2002-42)
UTM: 6942940.457 N; 760071.2714 E
Elevation: 381 m
Size: 3975 m 2
A3
215
Kapitanga (R-5)
UTM: 694509.867 N; 760810.7515 E
Elevation: 288 m
Size: 3674 m 2
Exp. 1
A1
Exp. 2
A2
A4 A5
Exp. 3
Exp. 4
A6
0
50
100 m
Ruatara (R-17)
UTM: 6945695.104 N; 761746.9169 E
0
50
100 m
Noogurope (R2002-43)
UTM: 6944219.022 N; 759897.7827 E
Elevation: 304 m
Size: 6940 m 2
Exp. 1
Exp. 3
STP #3
Exp. 2
STP #2
STP #1
STP #1 STP #2
STP #3
STP #4
Unit 1
Elevation: 650 m
Size: 3040 m 2
Exp. 1
0
50
100 m
0
50
100 m
Figure 12.10. IKONOS satellite images of Pukutaketake (R2002-42), Kapitanga (R-5), Ruatara (R-17) and Noogurope
(R2002-43) fortifications. Size estimates, site boundaries and tower orientation are based on field observations and
features (e.g. terraces) visible in IKONOS imagery. UTM coordinates are from the centres of each tower. The small
numbered triangles are the locations of auger tests and sample test pits. A small 0.5 m x 0.5 m test unit was excavated
on top of the tower at Ruatara. Drafted by J. Bartruff and D.J. Kennett.
terra australis 37
216 Douglas J. Kennett and Sarah B. McClure
UCIAMS-14757
Kapitanga
UCIAMS-14758
UCIAMS-47244
UCIAMS-14774
Ruatara
UCIAMS-14762
UCIAMS-47241
UCIAMS-60739
UCIAMS-60740
Noogurope
UCIAMS-14756
UCIAMS-47241
800
1000
1200
1400
1600
1800
2000
Years (AD)
Figure 12.11. Calibrated radiocarbon dates for fortifications in the western sector based on phase modelling presented
in Chapter 11. Prior distributions (routine calibration) are shown in outline, and posterior distributions (modelled within
a phase) are solid.
ridges to the north and south. Compared with other sites on the island, the use of dyke-stone
construction was fairly limited.
One of the unique features of Ruatara is a series of wall-like linear earthen mounds that
separate the lower terraces at the site. A series of six well-defined terrace units, often divided
by these high earthen mounds, extends from the site core along the primary ridge to the west.
These terraces were heavily vegetated during our visit, but a cow trail cuts through and has
exposed midden deposits at several locations. Midden soils were shallow, but well developed
in both exposures 2 and 3. Exposure 3 occurs along the northwestern terrace wall (8 cmbs)
and contained charcoal that dates to 630 ± 15 bp (UCIAMS-47243), comparable in age to the
earliest deposits at Noogurope. A similar exposure on the western side of the next terrace to the
east produced a much more recent age (11 cmbs, 170 ± 15 bp, UCIAMS-47244). One of the
best examples of an earthern wall occurs between two of the most prominent terraces on the
southwestern arm of the site. Midden-like deposits were evident in a natural exposure on the
eastern side of this earthen mound and this was most likely recycled as construction fill from a
previous occupation. A charcoal sample collected from this exposure (14 cmbs) returned an age
of 345 ± 15 bp. Calibrated age ranges for the site are AD 1300 to 1830 (see Figure 12.11). Two
large pits similar to those found at other fortifications occur in the middle of the terrace just to
the west of this earthen mound feature. Future excavations should be undertaken to determine
the function of these pits, but they were probably used for the storage of taro, as proposed by
Stokes (n.d.).
The tower at Ruatara is impressive and it is unclear whether the thin terraces on its western
flank were defensive or domestic features. STP #1 was placed in the middle of the second terrace
west of the tower and excavated down to 40 cmbs without encountering midden soils. A second
test unit (STP #2) was excavated down to 35 cm and a thin midden soil containing charcoal
terra australis 37
The archaeology of Rapan fortifications
217
occurred at about 15 cm and was radiocarbon dated to 210 ± 15 bp (UCIAMS-14774). The
tower itself is tall (an estimated 5 m) and is squared off on its northern and western side.
Remnants of dyke-stone masonry are evident, but these stones were either recycled elsewhere
or the masonry was never as well developed as in other nearby locations (e.g. Kapitanga). The
planar upper surface of the tower is 8 m x 3 m in size and is oriented east–west. It was certainly
large enough to serve as a residence. Dark midden soil containing charcoal, shell, bone and adze
flakes was evident across the surface of the tower. A circular test pit (STP #3, 20 cm diameter)
was excavated in the centre of the tower and dark midden soil containing charcoal and adze
flakes extended down to 35 cm. Unit 1 (25 cm x 50 cm) was excavated 2 m from the tower’s
western edge and 50 cm from its southern edge. These deposits were excavated in 10 cm levels
down to 35 cm. Faunal materials were reasonably well preserved in these deposits and were
dominated by parrotfish (Scaridae, see Butler, Appendix C). Other fish (Serranidae, Labridae)
and rat (Rattus sp) bones were also identified in these deposits.
Kapitanga (288 m elevation, 3674 m2) is located between Noogurope and Ruatara in
the northern section of the island. Compared with the earlier fortifications in this sector, it is
highly consolidated with dyke stone-lined domestic terraces similar to Potaketake (R-2). The
Norwegian expedition was drawn to this location because of its hyper-fortified position and its
well-formed oval tower covered with a beautiful dyke-stone masonry facade. They considered
it the best example on the island. Defensively, this fort was enhanced relative to other locations
because of a steep rock precipice 100 m tall on the east side of the fortification. The western
Figure 12.12. Ruatara (R-17) fortification taken from the southwest looking northwest. Photograph D.J. Kennett.
terra australis 37
218 Douglas J. Kennett and Sarah B. McClure
side of the fortification is also well protected by a series of steep ridges. Eleven domestic terraces
occur in three compact tiers. Rubble-filled double retaining walls bound the upper terrace.
Multiple pit features similar to those at Potaketake were also identified and two fosses were
strategically placed on the western and southern flanks to cut off the only two viable entry
points into the fortification. Large basalt adze flakes were evident on the surface of the northern
exposure at the site.
Pukutaketake is positioned on a high point (381 m) overlooking the mountain pass
between Anarua and Ha’urei Bay. Noogurope is visible across the pass to the north and Morongo
Uta occurs on the same ridge system just to the southeast. Stokes referred to this site as west
Potaketake and people living on the island in the 1920s considered this to be one of the oldest
sites on the island. The site was heavily vegetated during our visit, but 23 terraces were identified
running along the main and subsidiary ridges. The overall size of the site is small (3975 m2)
compared with Ruatara. The site has a well-formed tower, but lacks the dyke-stone masonry
evident at other sites. Dyke-stone masonry is rare except for a few remnant sections along the
southern and east edge of the upper terrace.
We excavated seven exploratory STPs (each 20 cm in diameter) at the site, and most of
these (STPs 1–5) were in the terrace surrounding the central tower (see Figure 12.10). One of
these was excavated into the northeastern edge of the tower itself. Dark midden soil containing
high concentrations of charcoal was encountered in all of these units at approximately 10 cmbs.
Adze flakes were more commonly encountered in these deposits than in other locations on the
island. These midden deposits were between 20 cm and 30 cm thick, and charcoal appeared
to be more commonly concentrated on the outer edges of these terraces. Midden deposits
containing charcoal also occurred on the top of the tower, along with several fire-cracked rock
fragments, suggesting that a hearth or oven was located near the excavation unit. Two STPs were
also excavated at the end of a long terrace that extends east from the central tower. The deposits
were more ephemeral and probably severely wind deflated. The radiocarbon date from one test
pit (STP #2, UCIAMS-14760, -14759) places it in the final fortification phase between AD
1700 and 1830.
Northern sector
The northeastern sector of the island consists of land roughly delineated by the ridgeline
separating Tupuaki and Akananue bays (and associated drainages) and all of the land on the
eastern shore of Ha’urei Bay. Three of these drainages flow outwards to Akananue, Agairoa and
Akatanui Bays and one flows inwards to Ha’urei Bay. Sharp ridgelines delineate these drainages
and provide the easiest overland paths between fortifications and other upland sites. In certain
instances, ridges were flattened or modified (e.g. staircases) to promote movement between
locations, but ditches and fosses were also placed in strategic locations to inhibit travel. The
remains of three large fortifications occur along the ridge separating the inner and outer parts
of the island. Each fortification was strategically placed at the juncture between the main ridge
and one or two secondary ridges extending out to the coastline. Vairu (R-3) is positioned at the
top of the ridge (369 m) that separates Akananue and Angairao. Potaketake (R-2) overlooks
Agairao and Akatanui Bay at the ridge separating these two drainages, and Tapitanga (R-4, 268
m) overlooks Akatanui and Ha’urei bays and sits directly above the modern town of Area. The
available radiocarbon dates indicate that they were all established and occupied during the final
phase of fortification between AD 1700 and 1830 (Figure 12.13).
All the large fortifications were visited in this sector and limited excavations were carried
out at Tapitanga and Potaketake to build on previous work by the Norwegian expedition. Vairu
terra australis 37
The archaeology of Rapan fortifications
219
Figure 12.13. Calibrated radiocarbon dates for fortifications in the northern sector based on phase modelling
presented in Chapter 11. Prior distributions (routine calibration) are shown in outline, and posterior distributions
(modelled within a phase) are solid.
was visited briefly and is estimated to be 4213 m2 in size (Figure 12.14). As with the other
large fortifications, it has a well-formed central tower, and at least 11 domestic terraces extend
outwards from the central tower along the main ridge and down the secondary ridge separating
Akananue and Angairao. Auger tests on multiple terraces and on top of the flattened tower
exposed midden soils containing charcoal and the occasional basalt adze flake. Midden soils
were also exposed at the edge of the uppermost terrace on the western side of the site. None of
these midden soils were deeper than 30 cm and most were 10–20 cm thick. Two radiocarbon
dates, one from Auger #7 (tower) and the other from Exposure #1, suggest occupation between
AD 1700 and 1830.
Both Tapitanga and Potaketake sit above the headwaters of Akatanui Bay. AMS radiocarbon
work places both fortifications just prior to European contact, with Tapitanga dating earlier,
between 1700 and 1750, and Potaketake dating to between AD 1750 and 1830. The Norwegian
expedition mapped these two sites in the 1950s and we use its observations as a starting point
for describing our work at these sites. Tapitanga is the largest fortification on the island (25,237
m2), but the site is dispersed and clear evidence of terracing and use is isolated to a much smaller
portion of the entire area (Figure 12.15). A poorly developed tower carved from the highest
point of the ridge marks the centre of the fortification, and it is surrounded by a consolidated
cluster of smaller domestic terraces. Five more distant terrace groups, positioned away from
the central group, radiate out along the ridge systems surrounding Akatanui Bay and down the
slope towards Ha’urei Bay. Several pit features were visible on the surface of the site, as described
by Ferdon (1965), with the best examples in the central and eastern parts of the site.
Free-standing walls and stone-lined terraces are rare at this location and many of the
terra australis 37
220 Douglas J. Kennett and Sarah B. McClure
Potaketake (R-2)
Elevation: 217 m
Size: 5744 m 2
0
UTM: 6944304.046 N; 763395.9545 E
50
100 m
Vairu (R-3)
6
Elevation: 369 m
2
Size: 4213 m
5
4
3
Exp. 1
8
UTM: 6945589.853 N; 762849.2448 E
2 1
7
0
50
100 m
Tapitanga (R-4)
Elevation: 268 m
2
Size: 25,237 m
UTM: 6943851.502 N; 763427.8861 E
0
50
100 m
Figure 12.14. IKONOS satellite images of Potaketake (R-2), Vairu (R-3) and Tapitanga (R-4) fortifications. Size
estimates, site boundaries and tower orientation are based on field observations and features (e.g. terraces) visible in
IKONOS imagery. UTM coordinates are from the centers of each tower. The small numbered triangles are the locations of
auger tests and sample test pits. Drafted J. Bartruff and D.J. Kennett.
terra australis 37
The archaeology of Rapan fortifications
221
Figure 12.15. Map of Tapitanga showing locations of auger tests and excavation units. The map is based on the Norwegian
expedition’s original map, field observations and IKONOS satellite imagery. Drafted R. Van Rossman.
structures are heavily eroded. Ferdon (1965) noted more formal architecture, with a few exposed
masonry walls exposed in the central precinct and some of the southwestern terrace groups.
Terraces were cut directly into the ridge and the distribution of these terraces is skewed towards
the Ha’urei Bay side of the ridge, but the upper terrace and tower were clearly positioned to
overlook Akatanui Bay. Ferdon (1965) argued for an early occupation of this fortification
compared with other forts in the area, based on its dispersed form, its lack of formal walls and
defensive features (e.g. fosses), and the heavy erosion it has suffered. It does appear to have been
constructed earlier than the other fortifications in the northeastern sector.
Seven augers and one sample test unit (STP) were excavated in terraces on the western side
of the fortification and we also examined an exposed oven on the eastern end of the fortification
down the ridge towards Akatanui Bay. Dark midden soils containing charcoal and the occasional
adze flake were encountered at each location between 5 cm and 10 cm below the modern
ground surface. These midden soils were generally between 20 cm and 30 cm deep, but midden
soils extended to 60 cm below the surface in the centre of one well-formed rectilinear domestic
terrace.
We also excavated two units at this site. Unit 1 (1 m x 1 m) was established on the west side
of the tower along the northern wall of a domestic terrace. The unit was placed close to a large
basalt boulder that had been heavily battered at some point in the past. Surface collection in
the vicinity of the boulder and Unit 1 revealed a large number of primary basalt flakes (N=75).
Unit 1 was excavated in naturally occurring levels and all material was screened (1/4” mesh)
terra australis 37
222 Douglas J. Kennett and Sarah B. McClure
for maximum recovery. Stratum 1 varied in depth between 5 cm and 10 cm and was a welldeveloped dark-brown midden soil containing flecks of charcoal and basalt flakes. Removal of
this level exposed a stone terrace wall on the north side of the unit and several concentrations
of fragmented basalt in the east side of the unit. Stratum II comprised orange-brown soil that
contained flecks of charcoal. The cluster of broken basalt visible at the base of Stratum I was
revealed to be an oven containing high concentrations of fire-cracked rock, dark brown soil
and charcoal (Figure 12.16). An alignment of rocks occurs in the southern side of the unit
and was possibly the foundation of an interior wall. A hearth or oven originally excavated into
Stratum III was also identified at the base of Stratum II (30 cm) in the eastern side of the
unit. It contained large concentrations of charcoal and fire-cracked rock. Charcoal and basalt
flakes were recovered while screening (2 mm, 15 litres). Preservation was limited and no faunal
remains were recovered, but several burned candlenut (Aleurites moluccana) fragments were also
recovered from this stratum. Stratum III was an orange-brown soil with a high clay content
and contained no charcoal or cultural material. An auger probe in the northern side of the unit
indicated that this soil extended down 60 cm to bedrock. The only radiocarbon dates from this
site come from the interface between Stratum I and Stratum II (S1, UCIAMS-2180, 145 ± 25
bp) and from the hearth feature (UCIAMS-2179, 140 ± 20 bp).
Unit 2 (50 cm x 50 cm) was excavated on the southern side of the tower near a natural
exposure of midden soil containing bone. Midden soil extended to approximately 30 cm below
the surface and was excavated in 10 cm intervals. Faunal materials were recovered from all three
levels (2 mm mesh) and are dominated by parrotfish (Scaridae, NISP of 30, of which 15 are
teeth). Larger bones in the sequence are sea turtle, but the bone was poorly preserved and the
exact species could not be identified. A large number of unidentifiable vertebrate bones was also
recovered (Butler, Appendix C).
Figure 12.16. Stratigraphy in Unit 1 at Tapitanga. Drafted R. Van Rossman.
terra australis 37
The archaeology of Rapan fortifications
223
Potaketake (R-2) is located at the junction between the ridge lines that form the Akatanui,
Angairao and Ha’urei watersheds, at 217 m elevation. From this vantage point there is a
commanding view of agricultural lands in two external drainages, Akatanui and Angairao, and
the upper portions of the Ha’urei watershed. The hilltop village of Tapitanga (R-4) is located
only 0.5 km to the south across a saddle in the ridge at the head of Akatanui Bay. A series of
rockshelters with occupation ages between AD 1300 and 1500 are present along the northern
shore of the bay, along with several stone fish traps that occur along the edge of the bay.
The heavily wind-blown and eroded remnants of a tower form the centre of Potaketake
(Figure 12.17). Well-formed domestic terraces surround this tower in two distinct tiers along
the north–south axis of the site. Defensive ditches were cut into three ridges surrounding the
residential core. Two of these occur within the residential compound interpreted either as an
expansion of the core settlement or a two-part defensive system. Potaketake was surveyed and
mapped by the Norwegian expedition in 1956 (Ferdon 1965:13). At that time the tower was
intact and composed of a conical bedrock and earthern fill spire reinforced by a masonry wall.
Free-standing masonry walls lined the edges of the primary domestic terraces surrounding the
central tower. Ferdon noted several large pits within the confines of the site that he interpreted
as defensive features. However, local people referred to these as taro pits and the position of
some of these features suggests that they were water cisterns or storage pits. Stone slab hearths,
most likely for heating room blocks, were also identified on the surface of the northern sector of
Figure 12.17. Map of Potaketake
showing locations of auger tests and
excavation units. The map is based on
the Norwegian expedition’s original map,
field observations and IKONOS satellite
imagery. Drafted R. Van Rossman.
terra australis 37
224 Douglas J. Kennett and Sarah B. McClure
the site. An eroded face on the northern sector of the site revealed a dark midden soil and bellshaped poi pounders similar to those found during the excavations at Morongo Uta.
Potaketake (217 m) was covered with low grass during our visit and most of the architecture
was visible on the surface (Figure 12.18). The terrace walls and stone-lined hearths described
by Ferdon (1965) were clearly visible, along with the dark midden soil eroding from the edges
of the stone-lined rectilinear terraces. Double stone walls filled with soil surround most of the
terraces in the central precinct. Much of the tower is now destroyed and only a short dyke-stone
wall remains surrounding the eroded basalt core. The largest eroded exposure at the site occurs
on the northern edge of the central group of domestic terraces. Large flecks of charcoal and
fragments of shell were visible in the exposure, including the inner whorl of a large gastropod.
The remnants of an oven are also visible in this exposure (fire-cracked rock, burned soil and high
concentrations of charcoal). One of our workmen showed us an adze that he had collected from
the site (Figure 12.19). Unit 1 (1 m x 1 m) was placed near this erosional face and excavated in
naturally occurring levels (Figure 12.20). The uppermost stratum (10–15 cm) consisted of very
dark greyish brown (10 YR 3/2) silty soil with sparse amounts of charcoal. The sediment in this
stratum is well sorted and contains very few pebbles or small rock inclusions. Soil from this level
was screened through a 2 mm sieve and larger chunks of charcoal and eight basalt flakes were
recovered. Underlying this stratum is a very dark brown (10 YR 2/2) silty clay containing higher
concentrations of cultural material. Charcoal concentrations remain high in this stratum, mixed
with sub-angular pebbles and stones.
A series of superimposed oven features (features 2–4) was encountered during the excavation
of this stratum and it is visible in the southern and eastern walls of the excavation unit (Figure
12.20). One of these ovens (Feature 4) was excavated into the underlying dark, yellow-brown,
12.18a
terra australis 37
The archaeology of Rapan fortifications
225
12.18b
12.18c
Figure 12.18. Photographs of Potaketake: a) fortification from the south looking north; b) looking north from tower
over domestic terraces; c) remnant of dyke-stone facade on tower. Photographs D.J. Kennett.
terra australis 37
226 Douglas J. Kennett and Sarah B. McClure
Figure 12.19. Adze fragment found
on the surface of Potaketake.
Illustration R. Van Rossman.
Figure 12.20. Stratigraphy in Unit 1 at Potaketake. Drafted R. Van Rossman.
terra australis 37
The archaeology of Rapan fortifications
227
Figure 12.21. Planview of hearth
excavated in Unit 2 at Potaketake.
Drafted R. Van Rossman.
sterile, parent soil/rock and it contained eroding basalt. A bifacially flaked basalt tool was
found during the excavation of Feature 4, along with four basalt flakes recovered in the 2 mm
screens. The three oven features were radiocarbon dated, with the two deepest features (3 and
4 respectively) dating to 240 ± 25 bp (UCIAMS-2188) and 240 ± 20 bp (UCIAMS-2184)
respectively, and the uppermost feature dating to 210 ± 25 bp. The calibrated ranges for these
dates all fall between AD 1700 and 1800 (see Figure 12.13). Unit 2 (1 m x 1 m) was placed on
top of a hearth previously identified by Ferdon (1965) in the northwestern corner of the terrace
just north of the tower (Terrace 8, Figure 12.21). Dark yellowish-brown soil and charcoal was
found in its centre at 16 cm, but as yet no radiocarbon dates are available for this feature.
Secondary fortifications and refugia
Most of our work focused on the larger fortifications on the main ridge surrounding Ha’urei
Bay, but we also visited and sampled several smaller hilltop sites. Additional fortified locations
were also observed from a distance, but were not visited. Those at the highest elevation are
consistent with what Stokes (n.d.) described as refuges. Others positioned at lower elevations on
secondary ridgelines are usually near larger fortified settlements and they were probably the sites
of satellite communities that were linked to the larger sociopolitical systems.
Four smaller hilltop sites were investigated on the island during the 2002 field season (Figure
12.22). One of these smaller fortifications, Ngapiri (R2002-50), is located in the southern sector
of the island near Tevaitau and Morongo Uta. The other three smaller upland sites are located in
the northern sector of the island near Tapitanga and Potaketake. As discussed previously, Tevaitau
may have served as a large satellite community of Morongo Uta, given its close proximity and
terra australis 37
228 Douglas J. Kennett and Sarah B. McClure
temporal overlap, between AD 1700 and 1830. The site of Ngapiri is smaller than Tevaitau
and located higher (320 m) on the same ridge in a more defensible location surrounded by
steep-faced boulders. The site consists of eight terraces and some of these occur on a razor-thin
ridgeline. Dyke-stone walls are apparent along the edges of two terraces and one of these is close
to 1 m high. We excavated three sample test pits and dark midden soil containing charcoal and
adze flakes were identified in each. The maximum depth of these midden deposits was 25 cm,
but most were about 10 cm thick. One charcoal sample from the uppermost terrace produced
a calibrated radiocarbon age similar to those found at Tevaitau and the later dates at Morongo
Uta (see Figure 12.5). The site does not have a central tower and, with its elevated position, it
was probably a refugium as described by Stokes. Given its age and close proximity to Tevaitau, it
was either a satellite community or it may have served periodically as a fortified highland refuge
for people living at Tevaitau or Morongo Uta.
The other three small hilltop sites investigated occur in the north sector of the island near
Akatanui and Angairao Bay. Two of these overlook Akatanui Bay and are linked by ridgelines to
the larger fortifications of Tapitanga and Potaketake. Both of these fortifications were constructed
during the final fortification phase between AD 1700 and 1825 (see Figure 12.13), but they
were probably not occupied simultaneously given their close proximity, their architectural
differences, and the available radiocarbon dates. Tapitanga was more likely occupied in the
earlier stages of the final fortification phase, whereas Potaketake was probably occupied after
Taua (R2002-40)
Ngapiri (R2002-50)
Elevation: 280 m
Elevation: 320 m
Size: 1190 m
STP#2
Size:1405 m 2
2
STP#1
Exp.1
0
50
0
100 m
UTM: 6941622.067 N; 762001.338 E
50
100 m
UTM: 6945846.76 N; 764947.418 E
Pukumia (R2002-39)
Pukutai (R-19)
Elevation: 143 m
Elevation: 417m
Size:1104 m
2
Size: 2195 m 2
A3
A1
STP#3
A2
STP#1
STP#2
0
UTM: 6944769.457 N; 764576.546 E
50
100 m
0
50
100 m
UTM: 6943757.846 N; 764180.756 E
Figure 12.22. IKONOS satellite images of secondary fortifications Ngapiri (R2002-50), Taua (R2002-40), Pukumia
(R2002-39) and Pukutai (R2002-19). Size estimates, site boundaries and tower orientation are based on field
observations and features visible in IKONOS imagery. UTM coordinates are from the centres of each site. The small
numbered triangles are the locations of auger tests, sample test pits or exposures sampled. Drafted J. Bartruff and
D.J. Kennett.
terra australis 37
The archaeology of Rapan fortifications
229
this time and into the mission period. The two smaller fortifications date to the latter part of the
final fortification phase and therefore appear to be more contemporary with Potaketake.
One of these smaller sites (R-19) was visited and recorded by the Norwegian expedition,
but the expedition appears not at have worked at the site. It sits at an intermediate elevation
(143 m) equidistant between Point Tekogoteemo and Tapitanga. Today people refer to this
location at Pukutai. It overlooks the northern shore of Ha’urei Bay and the flat agricultural lands
where the modern town of Area is now located. Compared with Tapitanga and Potaketake, this
site is small. Only a few terraces are now visible on its surface, but a tower was purposely carved
from the hilltop and remnants of dyke-stone facing are still evident. Three augers were placed on
the uppermost terrace surrounding the tower. Organically rich midden soils containing charcoal
began at ca. 10 cm and extended down to ca. 20 cm. One radiocarbon sample from Auger 3
(18 cmbs) produced a calibrated radiocarbon age in the later stages of the fortification process
between AD 1750 and 1830 (UCIAMS-36952). Pukutai may have served as a moderately
fortified satellite community of Potaketake. Access to freshwater, agricultural lands and coastal
resources was greater from this location than from the larger fortresses.
The upland site of Pukumia (R2002-39) is quite different from Pukutai. It is located 1 km
to the northeast of Pukutai across the drainage associated with Akatanui Bay. Pukumia sits at
an elevation of 417 m and has a commanding view of the entire northern part of the island. It
is surrounded by steep terrain and several sheer cliffs block access to the site from the south and
east. Nine terraces were identified at this upland location. It does not have a central tower and
dyke-stone masonry is rare. Stokes appears to have visited this site and classified it as a highland
refuge. The small terraces were cut expediently from the ridgeline and were not reinforced with
rubble-filled masonry walls. Vegetation was sparse across this site and many of the terraces were
exposed and wind deflated. Three test pits were excavated, one on each arm of the refuge. Thin
midden soils occurred at all three locations, but were best developed on the northern arm of
the refuge. These terraces are larger in this area and there are three extending out towards Point
Teruapake to the north. A central hearth-like feature was excavated in the first of these. Large
quantities of charcoal were evident at the base of this feature and they produced a date in the
later stages of fortification expansion between AD 1750 and 1830 (UCIAMS-36951, STP #3,
11 cmbs).
Taua (R2002-40, 280 m) was the last small fortification site that we examined in the northern
sector. It is positioned near Point Teruapake on the ridge extending north from Pukumia. Agairao
Bay is visible to the west and the site overlooks Piritua to the east. It is reminiscent of R-19 in
that it has a miniature central tower and a series of rectilinear terraces, four to the south and
one to the north of the tower. The tower was cut from a natural hillock. Remnant dyke-stone
walls line some of these small terraces, but not extensively. A deep fosse was cut into the ridge
on the southern end of the site. The location is well protected on the east and north by steep
sea cliffs. The terraces are separated by slight differences in elevation and midden deposits were
exposed between the two southernmost terraces. Dark midden containing charcoal and adze
flakes is best developed between 20 cmbs and 30 cmbs. A radiocarbon age indicates occupation
in the beginning of the final fortification phase (AD 1700–1750), suggesting that Taua is more
contemporary with Tapitanga than Potaketake.
Discussion and conclusions
Our work at hilltop fortifications builds on previous work by Stokes (n.d.), Ferdon (1965),
Mulloy (1965) and Walczak (2001). We examined 10 previously identified large fortifications
on the main ridge surrounding Ha’urei Bay. At the core of each fortification is a central tower,
terra australis 37
230 Douglas J. Kennett and Sarah B. McClure
carved from a high point in the ridge, which is surrounded by one to two large terraces. Terraces
extend out from the site cores along ridges or down into surrounding drainages. The sites range
in size from 3040 m2 (Noogorupe) to 25,237 m2 (Tapitanga) (Figure 12.23).
The number of terrace units varies at these sites and is not directly related to site size. Some
locations are more consolidated than others. Potaketake is 5744 m2 and has ca. 30 terraces.
Tapitanga has a similar number of terraces (ca. 42) dispersed over a much larger area (25,237
m2). There is a tendency for terraces along flat ridgelines to be rectilinear in shape and those on
steeper terrain to be d-shaped, a difference due primarily to the natural topography. Towers are
commonly covered with dyke-stone masonry facades and rectilinear terraces are often reinforced
by dyke-stone masonry walls. Sometimes, parallel series of dyke-stone blocks with a rubble or
sediment-filled interior were used to reinforce these platforms. These walls were used to contain
Figure 12.23. Size and configuration of
the primary Rapan fortifications.
Drafted J. Bartruff and D.J. Kennett.
terra australis 37
The archaeology of Rapan fortifications
231
sediments cut from the ridgeline to create a large level living surface. None of these walls reach
above the level of the terrace and they were not constructed to be weight-bearing. Structures
made from perishable materials were built in the centres of these platforms, as indicated by
the position of hearths (Mulloy 1965). Tapitanga and Ororangi stand out relative to the other
fortifications on the island due to the general absence of visible dyke-stone masonry. The
most extensive natural dyke-stone deposits are in the vicinity of Noogorupe and there is some
evidence that these deposits were quarried. Determining the extent and character of dyke-stone
deposits will provide key insights into dyke stone’s varied use in fortification construction across
the island.
One of our main goals was to build a radiocarbon chronology for the primary fortifications on the island (Kennett et al. 2006). We collected radiocarbon samples from either natural exposures or test excavations at these sites. Thirty AMS radiocarbon dates place the initial
fortification at Noogurope and Ruatara between AD 1300 and 1400. Morongo Uta appears to
have been established between AD 1500 and 1600. This is followed by the proliferation of fortifications between AD 1700 and 1830. Most of the fortifications on the island were established
during this interval and there was a general increase in the number of fortifications on Rapa
after AD 1750. Work at Tapitanga and Potaketake (both overlooking Akatanui Bay) suggests
that dyke-stone masonry developed after AD 1750 at Potaketake. Most sites with dyke-stone
masonry have components that date after this time. Ororangi appears to be the only site dating
to between AD 1750 and 1830 where masonry architecture was not used. Mulloy (1965) argued
that terrace units were added as populations expanded at each location. Our budget did not allow us to test this hypothesis directly, but the dates at Morongo Uta conform to this expectation.
More work will be needed to determine how these sites expanded. Three of the four smaller fortifications tested date to between 1750 and 1830 and support the hypothesis that populations
were at a maximum when Vancouver encountered the island in AD 1791. Overall, these data
support the hypothesis that populations on the island generally increased through time.
Mulloy (1965) argued that the hilltop sites were residential and not simply used as periodic
refuges during times of war. This was based on large-scale cleaning and excavation at Morongo
Uta, where he identified house platforms (terraces) with domestic features (hearths and ovens),
residential debris (shell, bone, charcoal), and artefacts (e.g. adzes, poi pounders), indicating a
range of daily activities. Our work at multiple locations is consistent with the hypothesis that
these sites were residential. Hearths and ovens were encountered in most of our excavation
units and our auger and sample test-pit excavations indicate that dark midden soils containing
cultural materials cover these sites. Preservation in these deposits is poor, and wind deflation and
water erosion are diminishing their integrity. Basalt adze flakes were commonly found during
our excavations, but formal artefacts were rarely encountered in our units or on the exposed
surfaces of these sites. Faunal materials were recovered from most of the sites excavated, but
the assemblages are small and poorly preserved (see below). Mulloy (1965:53) estimated the
population at Morongo Uta to be ca. 425 people, based on the number of terraces and hearths
at the site, along with a conservative estimate of five people per household. This assumes that
all 85 terraces were occupied at the time of its maximum extent between AD 1750 and 1830.
Future work on the chronologies of individual sites will be required, along with excavations
at disassociated terraces on ridges and along valley walls, to determine what percentage of the
population lived in these hilltop settlements and whether this changed through time.
We infer from paleoecological data (Prebble and Anderson, Chapter 10) and the remnants
of prehistoric pondfields (Bartruff et al., Chapter 13) that the inhabitants of these hilltop
settlements had invested heavily in, and were reliant on, wet taro agriculture in lowland
terra australis 37
232 Douglas J. Kennett and Sarah B. McClure
pondfields. The limited distribution of lowland environments suitable for this highly productive
agricultural practice may be one reason they placed their settlements on ridgetops and ridgelines.
However, the cost of transporting taro and other resources (marine foods and water) to upland
locations was high and suitable water and food storage within these hilltop settlements would
have been essential. Smaller terraces within these complexes may have served as platforms for
above-ground storage facilities (Mulloy 1965). Large and small pits are also evident at most of
these fortifications. Stokes (n.d.) observed large pits ranging in size between 1.8 m and 4.5 m.
Some were as deep as 2.8 m. The best examples that we observed occur at Potaketake, Ruatara
and Kapitanga, and in all cases these pits were excavated into terraces on the outer flanks of
the radiating terrace units. Based on ethnographic data, Stokes (n.d.) argued that these pits
were lined with banana leaves, filled with crushed taro, sprinkled with water, and covered with
ti leaves and a layer of sediment (pp. 174–175). Future excavations should target these pits to
determine their character and collect sediment samples for microbotanical analysis.
In East Polynesia, agricultural produce was commonly combined with meat from introduced
domesticated animals (e.g. chickens, pigs) or marine foods (shellfish and fish). Faunal remains
were not well preserved in the acidic soils of these highly exposed sites. The use of 2 mm
screens to process sediments in the field helped us identify sediments with reasonable faunal
preservation, which were then taken in bulk for flotation. A total of 400 bones representing 10
different vertebrate taxa were identified using this method (see Butler, Appendix C). The results
from five different fortifications indicate the greatest reliance on near-shore reef fishes, with
parrotfishes (Scaridae) the most represented species. This is consistent with the use of stone fish
traps that are common along the shores of the island. The best-recorded example comes from
the northern edge of Ha’urei Bay (Ferdon 1965:13). Other carnivorous fish species were also
present in small numbers (groupers, snappers, moray eels) and suggest the use of hook and line
fishing. The only larger animal identified in these assemblages was an unidentified species of sea
turtle in the tower deposits at Tapitanga. Domestic animals were noticeably absent and rats were
the only introduced animal species identified in these assemblages. Overall, the faunal data is
suggestive of near-shore fishes supplying the primary source of protein on the island.
Political hierarchies and status rivalries were common throughout East Polynesia. Hereditary
chiefs controlled access to land and competed with rival chiefs to expand territorial boundaries
and improve food security for their kin (Vayda 1976; Kirch 1984). The notion of political and
social hierarchy was certainly carried to Rapa, but it only becomes apparent archaeologically
with the establishment of the Noogurope and Ruatara fortifications between AD 1300 and
1400. The appearance of two fortified villages signals a status rivalry between two competing
polities with chiefly lineages at their core. The structural organisation of these fortifications
also reflects this hierarchy. Stokes (n.d.) argued that the flattened tops of each tower served as
the platform for the chief ’s residence and that his closest kin lived in the upper surrounding
terraces. We identified dark midden soils on the tops of the intact towers that were similar to
those found on lower domestic terraces (e.g. charcoal, adze flakes). Faunal materials were also
identified on the tops of Ruatara, Ororangi and Tapitanga. An oven similar to the one identified
by Stokes (n.d.) on the tower at Morongo Uta was identified on the northwestern portion of
Ororangi’s tower. These data are consistent with the idea that the flattened tower tops served
as at least one residential component of the chief ’s quarters. The presence of sea-turtle bones
on the tower at Tapitanga is consistent with this interpretation, given their role as chiefly food
elsewhere in Polynesia (Anderson pers. comm.). When combined with site distributions and the
radiocarbon ages, these data suggest that there were 10 competing chiefly lineages on the island
in the century before European contact. Our work on secondary fortifications is also suggestive
terra australis 37
The archaeology of Rapan fortifications
233
of subordinate chiefly lineages living in satellite communities that were integrated into larger
political systems during the final phase of fortification.
Walczak (2001) has questioned the idea that the hilltop settlements on Rapa were fortifications
constructed in the context of inter-tribal warfare between competing chiefdoms. Instead, he
argued that these structures were primarily ritual in nature. Certainly, household shrines at
Morongo Uta point to the important role of ritual in Rapan society, and the very nature of East
Polynesian hierarchies is imbued with a ritual dimension that served to legitimise established
social and political orders. But establishing a functional dichotomy between ritual and residence
obscures the primary observation that these sites were first and foremost both inhabited and
defended. Chiefs were often accomplished warriors (Kirch 1984:196) and struggles over power
and land often resulted in war (Williamson 1937). Stokes’ (n.d.) ethnographic data from the
1920s, especially his record of Rapan traditions (Chapter 2), make it clear that war for land and
food stores was an important part of Rapan life. The strategic position of these sites, combined
with the presence of defensive features (e.g. fosses), is consistent with this interpretation. Sites
dating to the century before missionisation in 1830 tend to have more defensive features (e.g.
Morongo Uta, Potaketake, Kapitanga), and some of these features appear to have been cut
through existing architecture, suggesting that the incidence or severity of war was increasing.
The higher-elevation refugia sites (Ngapiri and Pukumia) also suggest increased hostilities late in
the Rapan sequence. This is consistent with Vancouver’s observations that people on Rapa were
living in hyper-fortified hilltop communities surrounded by multiple rows of palisades.
Our research was necessarily limited to a few important objectives, notably of chronology
and general site function. We have argued that the fortified sites need to be seen primarily in
that role, as defended villages, and that the broad history of their construction is consistent with
rising levels of population, inter-group competition and hostility, probably most particularly over
access to agricultural lands. This situation continued, and possibly was becoming increasingly
tense, into the early European era.
Note
1.
Walczak (2001) visited and mapped Ngapiri, Namuere, Pukutaketake, Tevaitau, Ungarere,
Ororangi and Ruatara in the late 1990s. He also excavated Tevaitau (10 units, 50 cm2), Ngapiri (2
units, 1 m2), Ruatara (1 unit, 2 m2) and Ororangi (1 unit, 2 m2). Only the excavations at Tevaitau
are described in detail.
References
Clark, G. and Martinsson-Wallin, H. 2007. Monumental architecture in West Polynesia: Origins,
chiefs and archaeological approaches. Archaeology in Oceania 42:28–40.
Ellis, W. 1838. Polynesian Researches, during a residence of nearly eight years in the Society and Sandwich
Islands. Volume III, Fisher, Son and Jackson, London.
Ferdon, E.N. 1965. Report 2: A reconnaissance survey of three fortified hilltop villages. In: Heyerdahl,
T. and Ferdon, E.W. (eds), Reports of the Norwegian Archaeological Expedition to Easter Island and
the East Pacific, Volume 2 Miscellaneous Papers. Monographs of the School of American Research
and the Kon-Tiki Museum, 24 Pt. 2. Esselte AB, Stockholm. pp. 9–21.
terra australis 37
234 Douglas J. Kennett and Sarah B. McClure
Field, J.S. and Lape, P.V. 2010. Paleoclimates and the emergence of fortifications in the tropical Pacific
Islands. Journal of Anthropological Archaeology 29:113–124.
Heyerdahl, T. 1958. Aku-Aku: The Secret of Easter Island. Allen and Unwin, London.
Kennett, D.J., Anderson, A., Prebble, M., Conte, E. and Southon, J. 2006. Human Impacts on Rapa,
French Polynesia. Antiquity 80:340–354.
Kennett, D.J. and Winterhalder, B. 2008. Demographic expansion, despotism, and the colonisation
of East and South Polynesia. In: Clark, G., Leach, F. and O’Connor, S. (eds), Islands of Inquiry:
Colonisation, seafaring and the archaeology of maritime landscapes (Terra Australis 29), pp. 87–96.
Australia National University Press, Canberra.
Kirch, P.V. 1984. The Evolution of the Polynesian Chiefdoms. Cambridge University Press, Cambridge.
Lamb, W.K. (ed), 1984. George Vancouver, A Voyage of Discovery to the North Pacific Ocean and Round
the World 1791–1795, Volume I. The Hakluyt Society, London.
Mulloy, W. 1965. The Fortified Village of Morongo Uta. In: Heyerdahl, T. and Ferdon, E.N. (eds),
Reports of the Norwegian Archaeological Expedition to Easter Island and the East Pacific, Volume 2,
miscellaneous papers. Monographs of the School of American Research.
Stokes, J.F.G. n.d. Ethnology of Rapa. Manuscript on file in the Bernice P. Bishop Museum, Honolulu.
Vayda, A.P. 1976. War in Ecological Perspective. Plenum, New York.
Walczak, J. 2001. Le peuplement de la Polynésie orientale: Une tentative d’approche historique par les
exemples de Tahiti et de Rapa (Polynésie française). Unpublished Dissertation, University of Paris
(Panthéon Sorbonne).
Williamson, R.W. 1937. Religion and Social Organization in Central Polynesia. Cambridge University
Press, Cambridge.
terra australis 37
13
Rapan agroecology and population
estimates
Jacob Bartruff
Department of Geography, University of Oregon, Eugene, USA, jbartruf@uoregon.edu
Douglas J. Kennett
Department of Anthropology, The Pennsylvania State University
Bruce Winterhalder
Department of Anthropology, University of California, Davis
Introduction
Colocasia taro (C. esculenta) grown in pondfield irrigation systems is the staple of Rapa’s
subsistence economy. Pondfield irrigation systems are well known in Oceania. Their development
usually coincides with the prehistoric expansion of island populations and the associated need to
increase crop yields from limited amounts of land (e.g. with intensification, Kirch 2000:317).
The primary goal of this type of irrigation system is to create the most favourable growing
conditions for Colocasia taro, pools of slow-moving water 2.5 cm to 5 cm deep. This involves
the construction of artificial terraces, berms and irrigation canals that require a significant
investment of time and labour for a pay-off of increased crop yields in the future. Augmentation
of the land increases its value and one of the social manifestations of this investment is greater
territoriality (Dyson-Hudson and Smith 1978). Hereditary ownership and the need to defend
these territories from interested neighbours often develop within this context.
Colocasia taro is not native to Rapa, but was introduced by the early colonists of the island,
along with several other commensal species. Wild variants of this plant can be found throughout
the Indo-Pacific region from east India to Formosa and south into the Sahul region (Cable
1984). Wild-taro-starch residue identified on stone tools from the Solomon Islands extends its
use back to 28,000 years ago (Loy et al. 1992). The traditional view is that domesticated taro was
introduced to New Guinea by intrusive Austronesian populations from island Southeast Asia
(Bellwood 1975), but recent evidence points to the possibility of an independent domestication
9000 years ago (Golson 1990; Yen 1991). Colocasia taro appears in west Polynesia during the
late Lapita Period (ca. 3000 BP) and was carried in canoes to more remote islands of east
terra australis 37
236 Jacob Bartruff, Douglas J. Kennett and Bruce Winterhalder
Polynesia, including Rapa, starting 1500 to 2200 years later (Kennett et al. 2006a).
Oral traditions suggest that Colocasia taro was the staple food on Rapa from the time of
initial settlement (Hanson 1970; Stokes n.d.), now estimated to be between ca. AD 1100 and
1200 (Kennett et al. 2006b; Chapter 11). Pollen from well-dated stratified lowland sediments
indicates its use by at least AD 1200 (Chapter 10). Pondfields dot the alluvial lowlands of Rapa
today. Remnant dry stonewall terrace features covering many of the alluvial valley bottoms are
indicative of much greater former extent of these agricultural systems. Many of the abandoned
or fallow terrace systems remain waterlogged and are now dominated by introduced agricultural
grasses (e.g. Paspalum subjugatum), sedges (e.g. Carex spp.), rushes (e.g. Schoenoplectus subulatus
subsp. subulatus) and adventitious herbs (e.g. Commelina diffusa). These terraces are visible in 1
m resolution IKONOS satellite imagery, and the purpose of this paper is to define the former
extent of taro production based on this imagery. We also explore prehistoric Rapan population
levels based on calculations of potential agricultural yield and maximum sustainable population
sizes under different regimes of diverted production.
Colocasia taro production on Rapa
Active and remnant pondfield terraces are visible in all the lowland valleys of Rapa (Figure 13.1).
This irrigation method is one of the four types of true irrigation (e.g. diversion of water to fields)
in the Pacific, as defined by Spriggs (1981, 1988). Ferdon (1965) conducted a preliminary
survey of these terraces and found that almost all of the alluvial valley bottoms on the island
were terraced at one time or another, including the land beneath the modern town of Ha’urei.
Some of these terraces are still in use today but they represent only a small proportion of the
available arable land once under production.
Figure 13.1. A) Modern and remnant pondfields in Akatanui Valley. B) Inset showing Colocasia taro growing in a pondfield.
Photograph D.J. Kennett.
terra australis 37
Rapan agroecology and population estimates
237
Information regarding Rapan irrigation techniques comes from Hanson (1970). Rapan
terraces are rectangular and about ‘forty to fifty feet long [12–15 m] by twenty to twentyfive feet [6 to 7.5 m] wide’ (Hanson 1970:70). The terraces are irrigated by narrow ditches.
Water from dammed streams runs through notches cut into the curbs surrounding the terraces,
descending downhill from the upper to lower levels into the still-water bays surrounding the
island. Ideally, each terrace thus is covered with slow-moving water about 5 cm deep.
Taro can be planted at any time during the year but grows best if planted when conditions
are warm, especially January through April. Preparations for taro planting generally begin in
September. If there are plans for new terraces or repairs on existing ones the work is done by the
men of the household during the last months of the year. To prepare an area for taro irrigation,
vegetation is cleared and the surface is levelled. The earth is turned to a depth of about a third
of a metre and dirt curbs are built around the terrace, with openings for the ingress and egress
of water. The irrigation ditch is excavated next and the terrace is filled with water until the soil is
muddy. According to Hanson, ‘if the soil is especially good, a terrace can produce a high quality
of taro for up to thirty years of continuous cultivation. Rapans feel they have a poor return on
labor invested if yields are high only for five or six years’ (Hanson 1970:71).
In preparation for the planting season, small peripheral plants are harvested from existing
terraces between October and December. Instead of being discarded after the edible corm is
broken off, the stalk of the plant is saved. The leaves are removed to reduce the chance that wind
will uproot the plant, and the stalk is placed in the terraces. After a month or two, these stalks
put down roots and start to grow. In January, these stalks are transplanted in the terraces using
the correct spacing for optimal production.
The pondfields are planted by men and women after the New Year, between January and
April. Rapans say taro grows best if it is planted during a full or new moon. Taro planting is
avoided during the first two weeks of March due to the belief that the corm will mature with a
hole running through its centre and is likely to develop corm rot (Hanson 1970:72–73).
Corms and cormels (small, younger corms attached to a fully developed corm) of Colocasia
taro are ready for harvest eight months after planting on Rapa. When planted at the correct time
of the year, they can be left to grow for up to two years without rotting. Once removed from the
pondfield, though, taro spoils in about five days. Both men and women harvest taro throughout
the year. Harvesting and transporting taro is considered the hardest part of the process because
of the weight burdens and other travel costs associated with retrieving these bulky tubers. Field
processing at the harvest site consists of removing the corm from the stalk (see Metcalfe and
Barlow 1992; Bettinger et al. 1997). There is little to discard to reduce the weight of these corms
for transport. Individuals may have to trek up and down steep ridges carrying 80 lbs (ca. 36 kg)
or more of taro for household consumption. Occasionally, boats are used to reach fields in more
distant watersheds.
Today, women prepare taro for household consumption. Taro corms and cormels must be
cooked because in raw form they contain inedible calcium oxalate crystals (Hanson 1970:70).
Taro is consumed in a variety of ways once it has been boiled, but usually, it is prepared as a
fermented paste called popoi. The preparation of this paste is labour intensive and usually done
by younger women in a household. Corms are first boiled and then mashed on the surface of a
flat rock until a rounded mass of paste is created (45–60 cm in diameter). Water and old popoi
is added to aid fermentation and then strenuously kneaded to the desired consistency. The popoi
is wrapped in leaf bundles and fermented for one to two days before being consumed by the
household (Hanson 1970:73–76) without further preparation.
terra australis 37
238 Jacob Bartruff, Douglas J. Kennett and Bruce Winterhalder
Remote sensing of pondfield agricultural systems
To explore pre- and post-contact ecology and population levels on Rapa, we estimate the
maximum extent of pondfield Colocasia taro agriculture using high-resolution IKONOS
satellite imagery (1 m). Modern pondfields are clearly visible in IKONOS imagery and appear
as distinctive gray to black areas in the panchromatic image, and dark green in the true colour
image (Figure 13.2). Multi-spectral data, specifically false colour infrared, were also used to
highlight and define older and more subtle agricultural features (e.g. inactive terraces, wall
alignments and old canals). The lateral distribution of these features was used to define the
maximum extent of ancient pondfield agricultural systems in all viable lowland environments.
Due to erosion, sedimentation, modern development and vegetation obscuring these features,
we consider this to be a conservative estimate of the maximum prehistoric agricultural extent.
Modern pondfields were only identified in five of the 11 watersheds that contained evidence
for more ancient systems, and much more land (83.75 ha) was under production in the past
compared with the present (3.73 ha; see discussion and results below). The spatial distribution
of modern and prehistoric pondfield agricultural systems varies and is discussed in relation to
our efforts to model pre-contact population levels described in the following section.
Figure 13.2. Geographic extent of modern and pre-contact pondfield irrigations systems in Angatakuri watershed.
terra australis 37
Rapan agroecology and population estimates
239
Modelling population size
We employ a model that estimates Colocasia taro crop yields and maximum sustainable
population sizes based on these estimates, given different levels of diverted production (BaylissSmith 1978, 1980). Based on its successful application in other parts of Oceania (Spriggs 1981;
Kirch and Sahlins 1992), we implement a simplified version of this model. Model parameters
include: 1) total taro production (in kcals); 2) annual consumption of taro per person (in kcals);
3) annual labor costs; and 4) percentage of the population involved in agricultural production.
Population estimates based on the extent of these agricultural systems are determined from
this information, given that a portion of the crop may have been diverted away from direct
consumption into trade or tribute. Our analysis allows for varying degrees of diversion from 0%
to 70%. The amount of labour (hours per week) associated with different diversion amounts is
also determined.
Parameter 1: Total taro production
We calculate total taro production on Rapa by estimating: 1) the maximum extent of pondfield
irrigation systems (in ha); 2) the raw yield of this land (mt/ha/yr); 3) the percentage of the
raw yield that is consumable by humans; and 4) the caloric energy provided by taro corms
(in kcals). Pondfield agricultural systems were defined in the Rapan lowlands based on the
identification of terrace wall alignments and irrigation canals visible in pan-sharpened satellite
imagery (See Figure 13.2). Active and inactive pondfields are visible in this imagery and are
generally consistent with our observations in the field and Ferdon’s independent observations
in 1956 (Ferdon 1965). We employ an estimate based on a comparable high-yielding Colocasia
taro pondfield agricultural system from the Cook Islands to estimate the annual productivity
of 1 ha of land (26.67 mt/ha/year; Manarangi 1984). Geographically and environmentally, this
is the closest analogous agricultural system and is considered a better productivity measure on
Rapa than the more distant and environmentally different island of Maewo (Vanuate; 35.71 and
58.10 mt/ha/year; Spriggs 1981, 1984; Kirch and Sahlins 1992). In determining crop yields, we
assume that harvests consist of 34% large corms and 66% small cormels. Approximately 15% of
large corms and 40% of cormels consist of inedible peelings (Bayliss-Smith 1980:71–72; Kirch
and Sahlins 1992:158). We use 106 kcals to 100 g of edible taro to convert crop yields, based
on previous nutritional studies (Bayliss-Smith 1980:72).
Parameter 2: Annual consumption of taro
The total energy requirement for humans used in our calculations is 800,000 kcals/person/
year (or ca. 2190 kcal/day). This is based on estimates from elsewhere in Oceania (BaylissSmith 1980; Spriggs 1981; Kirch and Sahlins 1992) and is a simplified number that does not
take into account variation by age, size, activity level, or the source of the energy. Based on
previous research by Stokes (n.d.) and Hanson (1970), along with the limited range of resources
available on Rapa (e.g. the island is an inhospitable environment for tropical plants such as
coconut, see Chapter 10), we assume that taro comprised 80% of the pre-contact diet. Based
on our own estimates of dietary composition (based on anecdotal evidence collected during
archaeological fieldwork between June and August 2002), we argue that taro comprises about
20% of the contemporary diet. This is due to the importation of a range of supplemental foods
from Tahiti. Given these assumptions, we estimate the pre-contact consumption of taro to be
640,000 kcals per person annually (ca. 1750 kcal/day); 160,000 kcals (ca. 435 kcal/day) is
estimated for modern populations.
terra australis 37
240 Jacob Bartruff, Douglas J. Kennett and Bruce Winterhalder
Parameter 3: Labour costs
Labour costs are based on analogous modern pondfield agricultural systems in Maewo (Spriggs
1981). The labour costs associated with initial preparation, planting, weeding and harvesting
Colocasia taro growing in pondfield irrigation systems is estimated at 4947 person hours/ha/year
using stone tools, and 3296 person hours/ha/year for steel tools. Canal excavation and terrace
construction is a large initial investment and Rapans expect to get more than seven years of
continuous production from a pondfield, and sometimes up to 30 years. To explore changing
labour costs associated with continued field use, we use labour estimates for one and 10-year
field usage.
Parameter 4: Labour force
The percentage of the population involved with taro production is more difficult to estimate. On
Rapa, men generally prepare pondfields for use and both men and women take part in planting,
weeding and harvesting (Hanson 1970). However, we do not know what percentage of men and
women took part in these activities in the past and how labour was divided during pre-contact
period. Following Kirch and Sahlins (1992), we use a range of estimates for the percentage of
people engaged in agricultural activities (80, 60, 40, and 30% of total population).
Population modelling results
Population estimates using the Bayliss-Smith model are presented in Table 13.1. These estimates
assume Colocasia taro yields of 26.67 tonnes (‘mt’ in Table 1) each year (per hectare) and it is
assumed that the entire system was in use prehistorically. Prehistoric population estimates range
from 760 to 2534, depending on the amount of taro diverted away from direct consumption.
Given our estimates that the current dietary contribution of taro (20%) and the area of active
cultivation taken from the IKONOS imagery would support 135 to 450 persons depending on
the degree of diverted production, this range overlaps with the current population on the island.
Also shown in Table 13.1 are the person hours required each week given different assumptions
about the fraction of the population engaged in taro production (30–80%). Individual labour
requirements decrease as the percentage of the population involved increases. Bayliss-Smith
(1980:80) argued that subsistence activities in the Pacific consume between 10 and 20 hours
a week. Estimates above 20 hours per week are identified in parentheses and are thought to be
unrealistic.
Two sets of labour estimates are shown in Table 13.1, the first indicating the necessary
hours/person/week if pondfields were used for one year and the second showing use amortised
over a 10-year period. These data indicate that individual labour requirements are reduced by
more than 50% with multi-year pondfield use. This is because initial field preparation is a large
part of production costs. As an example, the model predicts a prehistoric population of 1774
with a 30% crop diversion. If 60% of the population was involved in taro production, then the
average producer could expect to work around 7.49 hours each week if fields were used for one
year. As a long-term average, weekly labour investment is substantially lower if fields were used
continuously over a 10-year period (3.18 hours per week).
Table 13.2 lists the extent of modern and remnant pondfield agricultural systems in each of
the 11 watersheds on Rapa, along with population estimates associated with each based on the
Bayliss-Smith model. The spatial distribution of these field systems is shown in Figure 13.3. It
is immediately evident that even at their maximum extent, these productive agricultural systems
comprised only a small portion of the island’s total landmass (2.21%) and that these lands were
highly circumscribed due to the island’s rugged terrain. Modern pondfield agricultural systems
terra australis 37
Rapan agroecology and population estimates
241
Table 13.1. Carrying capacity model for Rapa (using 26.67 mt/ha/yr as yield).
Surplus (%)
Population
Labor hours/week (one-year field use)
Labor hours/week (10-year field use)
Productive % of population
Productive % of population
80%
60%
40%
30%
80%
60%
40%
30%
Pre-contact taro fields (83.75 ha)
0
2534
3.93
5.24
7.86
10.48
1.67
2.22
3.34
4.45
5
2407
4.14
5.52
8.27
11.03
1.76
2.34
3.51
4.68
10
2280
4.37
5.82
8.73
11.65
1.85
2.47
3.71
4.94
20
2027
4.91
6.55
9.83
13.10
2.09
2.78
4.17
5.56
30
1774
5.62
7.49
11.23
14.97
2.38
3.18
4.77
6.36
40
1520
6.55
8.73
13.10
17.47
2.78
3.71
5.56
7.42
50
1267
7.86
10.48
15.72
(20.96)
3.34
4.45
6.67
8.90
60
1014
9.83
13.10
19.65
(26.20)
4.17
5.56
8.34
11.12
70
760
13.10
17.47
(26.20)
(34.94)
5.56
7.42
11.12
14.83
Modern taro fields (3.73 ha)
0
450
0.65
0.87
1.31
1.75
0.39
0.52
0.77
1.03
5
428
0.69
0.92
1.38
1.84
0.41
0.54
0.81
1.08
10
405
0.73
0.97
1.45
1.94
0.43
0.57
0.86
1.14
20
360
0.82
1.09
1.64
2.18
0.48
0.64
0.97
1.29
30
315
0.94
1.25
1.87
2.49
0.55
0.74
1.10
1.47
40
270
1.09
1.45
2.18
2.91
0.64
0.86
1.29
1.72
50
225
1.31
1.75
2.62
3.49
0.77
1.03
1.55
2.06
60
180
1.64
2.18
3.27
4.36
0.97
1.29
1.93
2.58
70
135
2.18
2.91
4.36
5.82
1.29
1.72
2.58
3.43
are even more restricted, occurring in only seven of the 11 major drainages on the island.
The extensive lowlands surrounding Ha’urei Bay were divided into three areas for analytical
purposes, but in total comprise 60.56% (2.25 ha) of the lands currently under production.
Large pondfield agricultural systems are also in use in Angatakuri/Angatakuri Naku (26.02%)
and Iri (13.26%) drainages, both located in the southern half of the island, with smaller systems
positioned in the Pariati drainage (0.16%) to the north. The model-derived population estimate
based on the modern extent of pondfield agricultural systems is 450 individuals. This estimate is
based on the assumption that there is little taro diverted away from direct human consumption
due to heavy subsidies from the French Polynesian government.
Pondfield agricultural systems were more extensive in the past. Remnant pondfield systems
were identified in all of the major drainages examined (n=11) and their overall extent (83.75
ha) was ca. 20 times larger than the modern systems. These abandoned and now fragmentary
systems were most extensive around the periphery of Ha’urei Bay (41.12 ha, 49.10%) and were
particularly concentrated in the lowlands along its northwestern shore (30.94%; Figure 13.4).
Large pondfield systems were also located in Iri (11.35 ha) and Angatakuri (5.38 ha) drainages
to the south, and Akatanui (6.48 ha), Agaira (4.85 ha) and Pariati (4.44 ha) drainages to the
north (see Figure 13.3). Smaller systems were also identified in several drainages in the north
terra australis 37
242 Jacob Bartruff, Douglas J. Kennett and Bruce Winterhalder
(Akananu, Akatamiro, Anarua and Tupuai). These data point to much higher and more widely
distributed populations in the past. We conservatively estimate an island population of 2027
based on the maximum identifiable extent of pondfield systems on the island (Table 13.1).
This estimate is based on an 80% dietary contribution of taro from pondfield systems and does
not include other dry farming systems (e.g. terraces) of unknown extent and importance. This
estimate includes a diversion of 20% of the production, which we argue to be reasonable given
the need to protect against drought, lower crop yields on a year to year basis, and food losses
due to war. Taro production is spatially distributed almost evenly throughout the island, with
the highest pre-contact production systems occurring in the central to northwest section of the
island, followed by the southern section of the island, with the east to northeastern areas of the
island a close third.
Table 13.2. Carrying capacity by watershed.
Watershed
Pondfield area (h)
Percentage of total
pondfield area
Surplus (%)
Population
Pre-contact taro extents
Agaira
4.85
5.79%
20
117
Akananu
2.95
3.52%
20
71
Akatamiro
1.07
1.28%
20
26
Akatanui
6.48
7.74%
20
157
Anarua
1.20
1.44%
20
29
Angatakuri
5.38
6.42%
20
130
Angatakuri Nako
2.18
2.60%
20
53
Ha’urei East
6.67
7.96%
20
162
Ha’urei South
8.53
10.19%
20
206
Ha’urei West
25.92
30.94%
20
627
Iri
11.35
13.55%
20
275
Pariati
4.44
5.30%
20
108
Tupuai
2.73
3.25%
20
66
Total
83.75
100.00%
Angatakuri
0.79
21.25%
0
96
Angatakuri Nako
0.18
4.77%
0
23
Ha’urei East
0.78
21.04%
0
94
Ha’urei South
0.36
9.66%
0
44
Ha’urei West
1.11
29.86%
0
134
Iri
0.49
13.26%
0
59
Pariati
0.01
0.16%
0
1
Total
3.73
100.00%
2027
Modern taro fields
terra australis 37
450
Rapan agroecology and population estimates
243
Figure 13.3. Spatial extent of precontact pondfield systems for each
watershed (+ denotes watersheds with
modern pondfield systems).
Figure 13.4. Inactive remnant pondfields on the north shore of Ha’urei Bay. Note the remnants of the prehistoric
fortified communities of Morongo Uta and Pukutaketake. Photograph D.J. Kennett.
terra australis 37
244 Jacob Bartruff, Douglas J. Kennett and Bruce Winterhalder
Discussion and conclusion
Since European contact, the population of Rapa has plummeted from an estimated high of
2000 in 1826 to a low of 120 in 1867 (McArthur 1967:307). Within the past 140 years, the
population has slowly increased. It stands at 497, according to the latest official census taken in
2002 (http://www.polynesie-francaise.gouv.fr/hc/iles-australes/dossiers.asp#rapa). Our model
prediction of 450 persons – under the assumptions that 3.72 hectares are in production and
that taro is 20% of the diet – is quite close to the contemporary census result. Today, Rapan
people rely on supplies from Tahiti, imports from abroad and introduced domesticated animals,
and this has decreased the importance of taro and taro-based products in the diet. For these
reasons, in the present it is more difficult to determine population solely from taro production.
Inhabitants of Rapa do not have to depend on taro agriculture for a majority of their diet and
the time that would be involved in more intensive agriculture is spent on introduced cash crops
such as coffee. Further research on the modern populations is needed to account for all the
energy sources that are available and their relative contribution to the diet.
Many of these complexities were absent in pre-contact time due to the heavy subsistence
dependence on taro and fish and the lack of imported bulk commodities. This simplification
of the pre-contact diet makes it easier to calculate the maximum sustainable population sizes
based on taro production. With a diversion of 20%, the model results in a maximum sustainable
population of 2027 individuals. This result is supported by early estimates from European
sources within the first 30 years of contact, which put the population between 1500 and
2000 individuals. When Vancouver first encountered the island of Rapa in 1791, his ship was
confronted at sea by “about thirty” canoes that contained “on a moderate computation, three
hundred men, all adult and apparently none exceeding a middle-age; so that the total number
of inhabitants on the island can hardly be estimated at less than fifteen hundred” (Vancouver
1801). Stokes (n.d.) thought this number too low, reasoning that Vancouver was probably only
met by inhabitants from the north and west of the island. In 1826, John Davies visited Rapa
for a second time and estimated the population to be around 2000, even though there had
been ‘much sickness and death in the island’ since his first visit earlier in the year (McArthur
1967:307).
More recently, Stokes (n.d.) estimated the pre-contact population by surveying the fortified
and unfortified terraced mountain sites and calculated that if all the sites were occupied at
the same time they could house about 3000 inhabitants (see also Anderson, Chapter 2). If
Stokes’ estimate is closer to the number of people living on Rapa then – with respect to our
model estimates – production diversion must have been lower, yield per hectare must have been
higher, and/or more land not evident in our satellite imagery was dedicated to taro production.
Ferdon (1965) mentions taro terraces further up the valleys of several watersheds. These terraces
may have been for dry taro agriculture, but they cannot be seen on the satellite imagery and
were not added to the maximum pre-contact extent calculations. It is not known for certain
whether these were domestic terraces, similar to the ones Stokes surveyed, dry taro terraces, or
a combination of both.
Through the use of the Bayliss-Smith model, we have shown that within the circumscribed
environment of Rapa and the limited amount of arable land, the island could have supported
population levels in line with or above early European estimates. The use of satellite imagery
and the Bayliss-Smith model (Bayliss-Smith 1978, 1980) allows for a cost-effective approach to
reconstructing agricultural systems and population levels on Rapa and could readily be used on
other islands in the Pacific. Additional ethnographic fieldwork is also needed to help constrain
terra australis 37
Rapan agroecology and population estimates
245
model parameters, and archaeological work to establish the extent of remnant pondfields and
other agricultural features (e.g. dry terraces, irrigation canals).
The population estimates presented in this paper are intriguing given the known
development and expansion of fortified hilltop communities on the island (Kennett et al.
2006). These communities were strategically located on the highest peaks along the ridge system
dividing the primary watersheds on the island. The distribution of these communities suggests
that they demarcated territories. The earliest fortifications (Morongo Uta and Ruatara) were
established 100 years after the island was colonised (ca. AD 1300–1400, see Chapter 11) and
were positioned near the most suitable lands for pondfield agriculture. The number of fortified
communities proliferated after AD 1700, just before the first recorded European contact in
AD 1791. Population expansion and community fissioning to fortified locations parallels
the expansion of pondfield agricultural systems on the island. We suggest that the expansion
of pondfield agricultural systems increased the value of land that was highly circumscribed
and defensible. This stimulated territorial disputes that were likely one of several interacting
contextual variables favouring increased inter-group aggression and greater investment in
defence, as has been argued for the Northern Channel Islands of southern California (Kennett
et al. 2009; Winterhalder et al. 2010).
References
Bayliss-Smith, T.P. 1978. Maximum populations and standard populations: The carrying capacity
question. In: Green, D. Haselgrove, C. and Spriggs M. (eds), Social Organisation and Settlement,
pp. 129–51. Part 1. BAR International Series (supplementary) 47. Oxford.
Bayliss-Smith, T.P. 1980. Population pressure, resources, and welfare: Towards a more realistic measure
of carrying capacity. In: Brookfield, H.C. (ed), Population-environment Relations in Tropical Islands:
The Case of Eastern Fiji, pp. 61–93. M.A.B. Technical Notes 13. Paris.
Bellwood, P.S. 1975. Man’s Conquest of the Pacific. Collins, London.
Bettinger, R.L., Malhi, R. and McCarthy, H. 1997. Central place models of acorn and mussel
processing. Journal of Archaeological Science 24:887–899.
Cable, W.J. 1984. The spread of taro (Colocasia sp.) in the Pacific. In: Chandra S. (ed), Edible Aroids,
pp. 28–33. Clarendon Press, Oxford.
Dyson-Hudson, R. and Smith, E.A. 1978. Human territoriality: An ecological reassessment. American
Anthropologist 80:21–41.
Ferdon, E.N. Jr. 1965. A reconnaissance survey of three fortified hilltop villages. In: Heyerdahl, T. and
Ferdon, E.N. (eds), Reports of the Norwegian Archaeological Expedition to Easter Island and the East
Pacific, Volume 2, miscellaneous papers. Monographs of the School of American Research.
Golson, J. 1990. Kuk and the development of agriculture in New Guinea: Retrospection and
introspection. In: Yen, D.E. and Mummery, J.M.J. (eds), Pacific Production Systems: Approaches to
Economic Prehistory pp. 139–47. Australia National University, Canberra.
Hanson, F.A. 1970. Rapan Lifeways: Society and History on a Polynesian Island. Little, Brown and
Company Ltd, Boston.
Haut-commissariat de la Polynésie française: Subdivision administrative des les Australes: Communes.
http://www.polynesie-francaise.gouv.fr/hc/iles-australes/dossiers.asp. Accessed on 20 November
2006.
Kennett, D.J., Anderson, A.J., Prebble, M., Conte, E. and Southon, J. 2006a. The Ideal Free
Distribution, Food Production, and the Colonization of Oceania. In: Kennett, D.J. and
terra australis 37
246 Jacob Bartruff, Douglas J. Kennett and Bruce Winterhalder
Winterhalder, B. (eds), Behavioral Ecology and the Transition to Agriculture, pp. 265–288.
University of California Press, Berkeley.
Kennett, D.J., Anderson, A. and Winterhalder, B. 2006b. Prehistoric human impacts on Rapa, French
Polynesia. Antiquity 80:340–54.
Kennett, D.J., Winterhalder, B., Bartruff, J. and Erlandson, J.M. 2009. An ecological model for
the emergence of institutionalized social hierarchies on California’s Northern Channel islands.
In: Shennan, S. (ed), Pattern and Process in Cultural Evolution, pp. 297–314. University of
California Press, Berkeley.
Kirch, P.V. 2000. On the road of the winds: an archaeological history of the Pacific Islands before European
contact. University of California Press, Berkeley.
Kirch, P.V. and Sahlins, M. 1992. Anahulu: The Anthropology of History in the Kingdom of Hawaii,
Volume Two: The Archaeology of History. University of Chicago Press, Chicago.
Loy, T., Spriggs, M. and Wickler, S. 1992. Direct evidence for human use of plants 28,000 years ago:
starch residues on stone artefacts from northern Solomon Islands. Antiquity 66:898–912.
Manarangi, A. 1984. Important taro varieties in the Cook Islands. In: Chandra, S. (ed), Edible Aroids,
pp. 24–27. Clarendon Press, Oxford.
McArthur, N. 1967. Island Populations of the Pacific. The Australian National University Press,
Canberra.
Metcalfe, M.D. and Barlow, K.R. 1992. A model for exploring the optimal trade-off between field
processing and transport. American Anthropologist 94(2):340–356.
Spriggs, M. 1981. Vegetable Kingdoms: Taro Irrigation and Pacific Prehistory. Unpublished PhD
dissertation. The Australian National University, Canberra.
Spriggs, M. 1984. Taro irrigation methods in the Pacific. In: Chandra, S. (ed), Edible Aroids, pp.
123–35. Clarendon Press, Oxford.
Spriggs, M. 1988. The Past, Present and Future of Traditional Taro Irrigation in the Pacific: An
Example of Traditional Ecological Knowledge. SPREP Occasional Paper Series 3. South Pacific
Commission, Noumea.
Stokes, J.F.G. n.d. Ethnology of Rapa. Manuscript on file in the Bernice P. Bishop Museum, Honolulu.
Vancouver, G. and Vancouver, J. 1801. A voyage of discovery to the North Pacific ocean, and round the
world. J. Stockdale, London.
Winterhalder, B., Kennett, D.J., Grote, M.N. and Bartruff, J. 2010. Ideal Free Settlement of
California’s Northern Channel Islands. Journal of Anthropological Archaeology 29:469–490.
Yen, D.E. 1991. Polynesian cultigens and cultivars: The question of origin. In: Cox, P.A. and Banack,
S.A. (eds), Islands, Plants, and Polynesians, pp. 67–98. Dioscorides Press, Portland, Oregon.
terra australis 37
14
The prehistory of Rapa Island
Atholl Anderson
Department of Archaeology and Natural History, Research School of Pacific and Asian Studies, The Australian
National University, Canberra, Australia, atholl.anderson@anu.edu.au
Douglas J. Kennett
Department of Anthropology, The Pennsylvania State University
Eric Conte
Université du Polynésie Française
Introduction
Rapa is a small, high and substantially isolated volcanic island lying in East Polynesia at the
temperate edge of the subtropical South Pacific. Discovered by Europeans in 1791, Rapa has
been the subject of lengthy ethnological and anthropological fieldwork but relatively little
archaeology and that directed almost entirely at the hilltop fortifications that dominate the
landscape. Our fieldwork, during six weeks in the winter of 2002, has produced sufficient variety
and abundance of material from archaeological sites (Appendix D) and palaeoenvironmental
research to enable the sketching of a first prehistory of Rapa.
Before turning to that, it is worth noting that we saw no evidence during fieldwork or
subsequent analyses of the proposed connection to Easter Island that infused 19th century
speculation about Rapan archaeology and drove the only substantial investigations that had
occurred previously on Rapa, by the Norwegian Archaeological Expedition (Chapter 1). To the
contrary, there is no historical support for the name ‘Rapa-iti’ as a derived companion to ‘Rapanui’ with colonisation implications; the Rapan language is part of the same southeastern group
that includes Mangarevan and Rapanuian, but it is not closer to the latter (Chapter 2); none
of the classic Easter Island architectural forms were reproduced on Rapa, nor did Rapa obtain
the chicken, sweet potato or most of the other cultivable plants that existed on Easter Island. It
is possible that migrants from Rapa reached other islands, including New Zealand, as has been
conjectured in relation to the forts, but the evidence is slight, because fortification occurred
widely in Polynesia, and New Zealand’s typical ditch-and-bank constructions are very different
from the stone-faced terraces of Rapa.
This is not to assert that there was no prehistoric movement to or from Rapa. The Rapan
traditions suggest some additional arrivals after initial colonisation, and the evident familiarity
terra australis 37
248 Atholl Anderson, Douglas J. Kennett and Eric Conte
of Rapans with iron at first European contact also implies external contact (Chapter 2), but
evidence of connections is otherwise scarce. Archaeologically, there is no obsidian, which might
have come from Pitcairn Island or Easter Island, and no pearl shell, which could have been
brought from the more tropical Austral, Society, Tuamotu or Gambier islands. While further
research might show evidence indicative of more regular contact, our working hypothesis has
been that the development of archaeological features on Rapa through the pre-European era
reflects events and contingencies of colonisation followed by adaptation and innovation that
occurred very largely in isolation.
Our project focused on questions about the chronology of human colonisation, the nature
of settlement and subsistence, landscape change and the rise of fortifications.
Chronology of colonisation
The chronology of colonisation was investigated through sedimentary coring and archaeological
research. The palynological record (Chapter 10) indicates that coastal swamp forests expanded
on the island after about 500 BC in response to changing sedimentary conditions following
a mid-Holocene high-stand in sea level. After about AD 1000, there was rapid sediment
accumulation in the embayments around Rapa, and at Tukou in Ha’urei Harbour. At Tukou,
sedimentation was associated with a decline in swamp forest, a rise in Pandanus and an onset
of high concentration in charcoal particles. Radiocarbon ages for the onset of microcharcoal
concentrations in the Tukou cores vary from 130 BC to about AD 1200 and are consistently
older than the earliest age determinations for human colonisation evident in the archaeological
record, although all radiocarbon ages for peaks in microcharcoal concentration occur within the
span of archaeological chronology (after AD 1000). This is typical of many charcoal profiles in
Pacific swamps and it is thought to represent infrequent natural fires followed by the substantially
increased rates of cultural ignition. Radiocarbon ages on charcoal for the beginnings of plant
cultivation, as represented by the onset of pollen deposition from taro (Colocasia esculenta),
generally overlap the beginning of the archaeological chronology. However, the strongest age
estimate, from Core 2 at Tukou, is about AD 1020, more than 200 years older than the earliest
archaeological determination for the island. This is possibly due to the mixing of sediment by
early agricultural activities in these wetland contexts (Chapter 10).
The initial phase of human colonisation is represented archaeologically by a set of dates
from Tangarutu rockshelter (chapters 3 and 11), which is by far the largest and most habitable
of the rockshelters on Rapa. All the relevant dates are from basal cultural deposits underlain by
clean carbonate sand that goes down to the basalt boulders and in situ floor of the rockshelter
(Chapter 3). The earliest date came from a small test probe below the base of a modern sand
mining pit in the middle of the rockshelter floor. The sand was too loose to allow an excavation
and even the core sample was difficult to obtain. The date is from the ANU radiocarbon dating
laboratory, and as with many later results from that laboratory, it has very large standard errors,
extending in this case from AD 600 to 1400. In our Bayesian analysis, this anomalously early
date is constrained to AD 1000–1400 based on the other early dates for the island, with peak
probability between AD 1000 and 1400. It is, at least, consistent with the earliest precise AMS 14C
date (UCIAMS-14769) of AD 1100 to 1200 on comparable deposits. The phase boundary for
initial colonisation has been modelled as about AD 800 to 1300, with peak probability between
AD 1100 and 1200. The phase boundary ends between AD 1300 and 1600 with peaks between
AD 1400 and 1500. The fairly wide age distributions for these phase boundaries, compared
with later phases, is related to the smaller number of dates used to define the colonisation phase
(Chapter 11).
terra australis 37
The prehistory of Rapa Island
249
On these two data sets, the age of onset of human colonisation still remains less precise than
was expected. The archaeological data suggest reasonably clearly that the 12th century was the
most probable period of human arrival (Chapter 11). The data from sedimentary cores are more
problematic, with some evidence indicative of cultural influence in the 11th century and even
considerably earlier than that, although the data are few. There are, however, plausible reasons
for thinking that the record of microcharcoal concentration in the samples from sedimentary
cores includes non-cultural evidence. Some radiocarbon dating samples that had the appearance
of charcoal could actually have been, or included, wood tissue blackened in the reducing
environment of swamp water, and some microcharcoal concentrations might represent natural
firing. As the dates referring to taro pollen were on charcoal samples, the same uncertainty exists
(Chapter 10). The span of colonisation could probably be specified more precisely and narrowed
further through more radiocarbon dating, but our chronological hypothesis at present is that
there was no human colonisation of Rapa before the 12th century AD. Colonisation at that
time is consistent with initial human dispersal throughout central East Polynesia (Wilmshurst
et al. 2011) and shows that even quite remote islands were reached in this phase.
Coastal settlement and subsistence
Our radiocarbon chronology shows that settlement, measured by occupation of rockshelters,
expanded in coastal areas of Rapa between AD 1400 and 1600. Most of the Tangarutu
occupational sequence dates to this interval. The second and third largest rockshelter complexes
on Rapa, at Akatanui and Angairao, were used first in this period, as were the small Taga
rockshelters on the mid-slope of Ha’urei Harbour (Chapter 11). It is very probable that there
were settlements on open ground, for example in the coastal flats at the head of the harbour and
in most large bays, in the period AD 1400–1600. Earth ovens and terraces around the shores
of Ha’urei Harbour, sectioned by road cutting and erosion, or investigated by coring or spade
pits, produced radiocarbon dates extending back to about AD 1200. As the terraces are about
the size suitable for a domestic unit, often around 8 m by 12 m, and sometimes have remains of
ovens or graves, they probably supported individual households. However, neither postholes in
section, nor cultural deposits across the floors were found in our preliminary investigations of
the Ha’urei sites. In addition, exploratory coring of sand dunes and other sedimentary features
in the outer bays did not pick up any evidence of buried occupational levels (Chapter 3).
Judging by excavations in rockshelters (and in hilltop forts, below), subsistence from the
beginning of settlement on Rapa was reliant on fishing, foraging and the cultivation of taro.
Coastal fowling, fishing and shellfishing
The list of animal resources available on Rapa was short. Pacific rat bone occurred throughout
archaeological deposits, but no dog, pig or chicken bone was found in any context that was
plausibly pre-European. No seal bone was recorded and the only reptile remains were from
marine turtle. Of 118 birds, identified to 15 species, 90% were seabirds. About half of all bones
were from the Kermadec petrel, and that species plus the little shearwater, red-tailed tropic
bird, grey noddy, brown noddy, white-bellied storm petrel and white tern, which were also
represented in the archaeological material, are still breeding on Rapa. Among landbirds, remains
of a Gallirallus indicate that Rapa, like most Pacific islands, once had more species of rail. Bones
of a large Ducula pigeon species and of a Cyanoramphus parakeet indicate other extinct taxa
(Chapter 6).
Fishing was clearly of primary importance in Rapan subsistence, and fishing patterns were
rather unusual. At Tangarutu, the oldest level shows a dominance of Scaridae, which gave way
terra australis 37
250 Atholl Anderson, Douglas J. Kennett and Eric Conte
to a dominance by Chaetodontidae and Muraenidae in the middle level, and Muraenidae, plus
Scaridae and Serranidae, in the upper level. Muraenidae and Congridae eels are generally very
scarce archaeologically in Pacific Island archaeological assemblages, and the high incidence
of Chaetodontidae also represents the first positive identification of this family in a Pacific
assemblage. Labridae, Serranidae, Muraenidae and Congridae are the other main families
represented (Chapter 7). The inter-site differences probably reflect variation between leeward
and windward inshore ecologies, but intra-site temporal changes are more difficult to interpret.
The rocky coasts of Rapa may have supported comparatively large and accessible populations
of marine eels, and large freshwater eels (Anguillidae) are common in streams and taro ponds.
The latter were avoided, as they are today, but marine eels are represented in the Rapan middens
to an extent matched only on Easter Island. As Rapa and Easter Island have similar inshore
ecologies and also are both isolated, the catch in each case may have been shaped by resolutely
inshore fishing of taxa in which the major species were those locally common (Chapter 7).
The abundance of Pomacentridae and Chaetodontidae needs to be seen in the context of
a generally small size of fish in all families represented in the Rapan assemblages. In the lowest
level at Tangarutu there are some remains of larger fish, but as most are small, fish size in the
Rapan sites is probably not an effect of over-exploitation. Sampling effects caused by our very
small excavations, or social division of catches, cannot be ruled out, but the small fish size
occurred both at Tangarutu and Akatanui, and in every other assemblage (Chapter 7; Butler,
Appendix C). We suspect that fishing technology was the main factor involved. It is apparent
from the absence of trolling lures, of large hooks or points from large composite hooks, and of
remains from large benthic or pelagic taxa that open-sea and deep-sea fishing was not common,
perhaps because of the danger of being lost in such an isolated location. In turn, this possibly
resulted from the absence of a sailing technology capable of bringing canoes back to the island in
varied wind conditions. The very small size of most of the Rapan hooks found archaeologically,
and known ethnographically, suggests that small fish were being targeted in this method, and it
is probable that fine-meshed nets and traps were also employed routinely. Relatively cool ocean
temperatures might also have produced smaller mature sizes of fish in families that exhibited
larger mean sizes in tropical waters (Chapter 7).
The shellfishing data from Tangarutu show evidence of a transition from hard-shore
towards soft-shore collecting within the context of a dominance of algal grazing taxa (Chapter
8). The urchin Diadema setosum, in particular, is a prominent algal grazer. As algal-grazing fish
such as scarids, pomacentrids and chaetodontids dominate the Tangarutu assemblage, the fish,
urchin and shell data from Tangarutu imply, together, that Anarua Bay supported extensive
hard substrates on which grew micro-algal turf. Oscillations in frequencies of different algalgrazing molluscs, urchins and fish taxa in the catch data suggest complementary changes among
vertebrate and invertebrate algal feeders. Thus, sporadic high levels of Diadema occurrence
probably resulted from fishing pressure on competing algal-grazing fish; strong initial pressure
on Scaridae might have promoted the relative abundance of other algal-grazing taxa and later
the pressure on those could have promoted a return to greater numbers of scarids, for which the
small candlenut hooks were developed late in the sequence (Chapter 9). Overall, variations in
catches within the marine exploitation data suggest a rise in the employment of angling relative
to nets and traps, and movement of shellfishing into the soft-shore bayheads, within a pattern
of resource use that was changing continually according to complex inter-relationships between
algal-grazing taxa (Chapter 9).
terra australis 37
The prehistory of Rapa Island
251
Plant foods and fibres
Just as Rapa lacked all of the domestic animals of Remote Oceania, so it lacked key crops,
notably coconut, breadfruit, yam and sweet potato, as well as most, if not all, bananas. Pandanus
(Pandanus tectorius and probably other species) produced edible fruits. Pith of the tree fern
(Cyathea medullaris) may have been consumed, as it is today, and pith from Marattia spp. and
Angiopteris spp. was also potentially available, but these and the rhizomes of ground ferns such
as Histiopteris incisa, Dryopteris sp. and Dicranopteris linearis were rudimentary starch sources at
best. Species introduced prehistorically that had food value, as well as other properties of fibre or
fruits in some cases were: candlenut (Aleurites moluccana), bottle gourd (Lagenaria siceraria), taro
(Colocasia esculenta) and tii (Cordyline cf. fruticosa). At least two timber trees, the miro (Thespesia
populnea) and hau or hibiscus (Talipariti tiliaceum syn. Hibiscus tiliaceus), were probably
introduced, and also the coral wood (Erythrina variegata), which produced ornamental flowers
– along with hibiscus flowers, an essential component of social life in Polynesia. Archaeological
charcoals show that tii and hau had been used as firewood in the lowest level at Tangarutu, and
candlenuts occurred throughout the sequence and in other rockshelter deposits, being notably
abundant at Akatanui (Chapter 4). The existence of these introduced materials in the earliest
recorded archaeological levels implies that initial colonisation had occurred earlier still, but
perhaps not much earlier.
In the upper level at Tangarutu, lenses of leafy remains produced fragments of plaited
Pandanus and Freycinetia (Pandanaceae) baskets or mats. There were also fragments of bottlegourd rind, dated earliest to about AD 1600, and many Pandanus keys. Knotted cordage
made from Broussonetia papyrifera (paper mulberry), braided cordage worked from the roots
of Freycinetia spp. (kiekie vine) and twisted cordage from Hibiscus also occurred in this level
(Chapter 5).
Agriculture
Pollen profiles show that the natural wetlands on Rapa were being used for taro production early
and extensively. They seem to have been used continuously throughout Rapan prehistory (starting
sometime between AD 1100 and 1200), along with constructed pondfields on higher ground.
Remnant pondfield systems can be seen in each of the major bays and the main harbour, and
overall they occupy an area (about 84 ha) some 20 times greater than the extent of the modern
pondfields (Chapter 13). A general absence of observable charcoal in the pondfield structures
and insufficient time or resources to embark on a systematic subsurface investigation prevented
our obtaining any information about the development and use of these systems through the
prehistoric era. Their distribution through all available bays and the plausible linkage of each
system to nearby hilltop forts, or to apparent remnant hamlets at lower levels, suggest that
at one stage, if only in late prehistory, they were all in use at once. On that assumption, and
given an 80% dietary contribution of taro from pondfield systems, including 20% diverted
away from direct consumption (waste, trade, tribute, etc), an island population of around 2000
people could have been sustained (Chapter 13). Lower estimates of the dietary contribution
of taro would produce larger population sizes, and some historical data suggest, though very
imprecisely, that there might have been around 3000 people living on Rapa before the impact
of European disease (Chapter 2).
Chronology, characteristics and purpose of forts
No aspect of the landscape or history has been so widely the subject of commentary as the
forts (pare) that stand out dramatically along the Rapan skyline. They beg the question of
terra australis 37
252 Atholl Anderson, Douglas J. Kennett and Eric Conte
what circumstances drove such an investment in fortification and our research has attempted
to answer that by investigating the chronology of construction and material evidence of the
purposes served by these structures.
Chronology of construction
Thirty AMS radiocarbon dates, analysed by Bayesian methods, help to put fort construction
into a general sequence. As most of the construction occurred after about AD 1700, when
the radiocarbon calibration curve has strong fluctuations that obscure precise determination
of age, we used the evidence from mission records that fortified villages were abandoned by
AD 1825 to constrain the probability distributions (Chapter 11). Radiocarbon dating places
initial fortification at Noogorupe and Ruatara between AD 1300 and 1400, approximately
200 years after initial island settlement. There may have been historical connections between
the occupants of Tangarutu rockshelter in Anarua Bay, seemingly a colonisation settlement,
and those who used the ridgeline above the bay and built the Noogorupe fort, which overlooks
Anarua and Ha’urei Harbour. Similarly, the contemporary establishment of Ruatara possibly
resulted from fissioning of the founding community at Anarua.
Settlement continued at Ruatara up to about AD 1800, and also at Noogorupe, with a
possible hiatus between AD 1400 and 1650. Morongo Uta appears to have been established
between AD 1500 and 1600 and it continued until about AD 1800. Most other forts were
constructed or occupied between AD 1700 and 1830, with a general increase in construction
after AD 1750. Seven forts date to the early 1700s and at least 10 forts to the late 1700s. Smaller
forts also date to the late 1700s – e.g. Tauo, Ngapiri, Pukutai and Pukumia (chapters 11 and
12). Investigations at Potaketake suggest that dyke-stone masonry developed after AD 1750
and most sites with dyke-stone masonry have components that date after this time. Ororangi
appears to be the only site dating to between AD 1750 and 1830 where masonry architecture
was not used (Chapter 12).
Characteristics of forts
Archaeological survey and small-scale excavation occurred at 10 large and four small forts
(Chapter 12). All of the larger forts are strategically positioned on the highest points along the
main ridge surrounding Ha’urei Bay. The sites range in size between 3040 m2 (Noogorupe) and
25,237 m2 (Tapitanga). At the core of each fort is a central tower carved from a high point in the
weathered basalt ridge and surrounded by one to two large terraces. Terraces extend out from
those site cores along ridges or down the slopes at the head of surrounding valleys. The smaller
forts are often found in close proximity to the larger forts and they are considered to be satellite
communities or temporary refugia. Two of the four smaller forts have central towers.
The number of terraces per fort varies considerably and is not clearly related to site size.
Potaketake was 5744 m2 and has ca. 30 terraces. Tapitanga has a similar number of terraces (ca.
42) dispersed over a much larger area (25,237 m2). There is a tendency for terraces along flat
ridgelines to be rectilinear in shape and for those on steeper terrain to be d-shaped. Natural
slabs of dyke stone were used widely in masonry and facing. The most extensive natural dykestone deposits are in the vicinity of Noogorupe and there is some evidence that these deposits
were quarried. Towers were commonly covered with dyke-stone masonry facades and rectilinear
terraces formed by cut-and-fill were usually reinforced with dyke stone laid in parallel series
as masonry walls. None of the walls reach above the level of the terrace and they were not
constructed to be weight-bearing. Tapitanga and Ororangi stand out as exhibiting very little
evidence of dyke-stone masonry (Chapter 12).
terra australis 37
The prehistory of Rapa Island
253
Structures made from perishable materials, most probably houses or huts, were constructed
on these platforms, their positions indicated by remains of hearths. Around them are scatters
of midden comprising charcoal, bone and burnt or flaked stone. The midden is mostly of fish
bone, and among it, Scaridae is dominant (Chapter 12; Butler, Appendix C), as at Tevaitau,
Orotangi, Ruatara and Tapitanga. Additional taxa are also similar to those found in the coastal
sites: Pomacentridae, Labridae, Muraenidae etc. Rat bones (Rattus sp.) were also recovered.
The inhabitants of these hilltop settlements were probably also reliant on the wet taro
agriculture of lowland pondfields, and the limited distribution of agricultural land may be one
reason why settlements moved increasingly to ridgelines and hilltops. These localities were also
suitable for storage and protection of harvested crops, probably in the large and small pits that
are evident at most of the forts (Chapter 12).
The purpose of forts
Several hypotheses have been advanced to account for the construction of forts on Rapa. Through
the telescopes of Vancouver and his officers in 1791 (Lamb 1984) they appeared as fortified
habitations with people clustered in them, but in the 19th century, when the structures were
manifestly uninhabited, the view prevailed that they were simply fortresses or redoubts used as
places of refuge and defence during times of war. In the early 20th century, Peter Buck (1954),
with New Zealand experience in mind, thought them rather like Maori pa, which were often
inhabited, though not necessarily continuously. The Norwegian archaeological investigations
(Heyerdahl and Ferdon 1965), especially the huge effort of exposure and excavation on Morongo
Uta, concluded that they were, indeed, fortified villages (Chapter 1).
More recently, however, Jerome Walczak (2003) has argued that the structures served
primarily ritual functions. Given that East Polynesian hierarchies had a powerful ritual dimension
that served to legitimise established social and political order, it can hardly be doubted that pare
were localities of ritual function. The same, of course, would have been true of settlements
around the coast. If the argument is made in archaeological terms, then slight evidence of ritual
activity in the pare, such as the small alcoves with upright stones set into walls at Morongo
Uta, needs to set against substantial evidence of fortification, habitation platforms, ovens and
middens. Our research at multiple locations is consistent with the hypothesis that these sites
were, substantially, defended residential complexes.
We argue that as political and social hierarchy is embedded in East Polynesian culture, it
was certainly carried to Rapa. It first becomes apparent archaeologically with the establishment
of the Noogorupe and Ruatara fortifications between AD 1300 and 1400, the new need for
defensive architecture implying the beginning of stronger status rivalry between competing
polities than existed at the time of initial colonisation or developed soon afterward. If it is
accepted, as ethnographic data suggest, that the flat-topped towers at the centres of large forts
were places of chiefly habitation, then 10 competing chiefly polities existed on Rapa by the 18th
century. Pare at that time tended to have more defensive features (e.g. Morongo Uta, Potaketake,
Kapitanga), some of which cut through existing architecture, suggesting improvements to
defensive structure. The higher elevation refugia sites (Ngapiri and Pukumia) also suggest
increased warfare late in the Rapan sequence. The overall trend in fort construction, from two in
the 14th century, gradual increases into the 17th century and an accelerated burst through the
18th century, suggests that conflict and the threat of war increased through the sequence. The
most likely reasons for this increase were either direct population growth or indirect population
pressure on resources, such as agricultural land.
terra australis 37
254 Atholl Anderson, Douglas J. Kennett and Eric Conte
The subtropical depriment
Rapa is one of a series of Polynesian islands that lie in the subtropical zone below the Tropic
of Capricorn. Others are: the Norfolk Islands, the Kermadec Islands, Raivavae, the Pitcairn
Group (including Pitcairn, Henderson, Oeno and Ducie) and Easter Island. Leaving aside
Raivavae, which has an extensive encircling lagoon, and the two small atolls (Oeno and Ducie),
the subtropical islands are distinguished by their virtual absence of coral lagoons.1 The loss of
diversity and biomass in marine resources that this represents is a significant deficit for long-term
settlement relative to tropical Polynesia. In addition, the absence for climatic reasons of some
tropical Polynesian native food plants and the difficulty or impossibility of growing coconut
or breadfruit reduced the value of the terrestrial resource array. Equally, the low number of
breeding sea birds (and the virtual absence of seals) was a significant deficit relative to temperate
South Polynesia (Anderson 1996). In other words, the subtropical islands lay unfavourably
between the lagoonal and agricultural landscapes to the north and the resource landscapes to
the south rich in marine birds and mammals (Anderson 2001, 2002). They suffered from what
can be called ‘the subtropical depriment’, i.e. the depression in resource opportunity for human
settlement that resulted from subtropical geography and climate.
It was further exacerbated by the fact that most subtropical islands were small (Pitcairn
5 km2, Norfolk, Raoul, Rapa, Henderson 29–38 km2), lacking bays or harbours, and notably
remote. Pitcairn is 524 km from Mangareva, Rapa 537 km from Raivavae, Norfolk 733 km and
Raoul 980 km respectively from New Zealand, and Easter Island 1915 km from Henderson
Island. Only the relatively large Easter Island (164 km2) and Rapa, by the advantage of its large
harbour and associated coral reefs in creating a lagoonal surrogate, were inhabited continuously
into the European era. Small size meant that population growth soon reached territorial and
resource limitations. Remoteness reduced both the range of resources that arrived during
colonisation and those that could be readily obtained later. There is no evidence that food crops
reached either Norfolk or the Kermadecs in the prehistoric era (although banana was growing
on the former by 1788), and they were relatively limited on the other subtropical islands: Rapa
had taro, ti, gourd and candlenut, possibly banana; Henderson had coconut, ti, swamp taro and
candlenut. Easter Island had sweet potato, taro, yam, banana and sugarcane, a range sufficiently
large to suggest that initial colonisation might have involved a number of canoes or some twoway voyaging. Other than the almost ubiquitous Polynesian rat (Rattus exulans), the dispersal of
animals was slight: possible dog on Norfolk and Easter Island, dog on Raoul, chicken on Easter
and pig on Henderson (Anderson 1981; Weisler 1996; Anderson and White 2001; Mieth and
Bork 2004).
Overall, then, the subtropical islands were prone to occupational stress caused by a variety
of factors and it is difficult to say that the histories of settlement were influenced more by one
factor than another. The impact of isolation, either inherent in remoteness or by the decline
of early interaction networks (Weisler 1996), was doubtless influential in the prehistoric
abandonment of most of the subtropical islands (Norfolk, Kermadecs, Pitcairn, Henderson),
but Rapa, and especially Easter Island, were also isolated. Survival on those probably depended
on the greater resource availability of a much longer coastline (Easter) or a large harbour (Rapa),
coupled with the development of agricultural systems absent or little in evidence elsewhere in
the subtropical islands.
Even so, it can be hypothesised that levels of competition for resources had reached quite
extreme levels by late prehistory. On Rapa, there may have been 10 competing polities, each
the proud lords of, on average, fewer than 4 km2 of land, nearly all of it steep and largely
terra australis 37
The prehistory of Rapa Island
255
barren. Perhaps only such a level of competition can have made the labour and inconvenience
of building and living in the skyline forts seem at all worthwhile.
Further research
We are well aware of the preliminary nature of our investigations on Rapa and the probability
that our results and conclusions will need significant modification in future research. The need
for more research is urgent. Increasing development of roads, housing and other facilities has
already exposed numerous sites around Ha’urei Harbour; sand deposits needed for concrete
are scarce and the largest of them underlies the oldest site on Rapa, at Tangarutu, which is
now about 85% destroyed by sand mining; major works, including the development of an
airfield, are envisaged. In addition, grazing cattle and, especially, a large population of feral
goats have created massive hillslope erosion in many bays. Even without these mechanisms of
accelerating site damage, the forts are decaying rapidly. The substrates are mostly saprolitic clays
and weathered basalts, which erode quickly in the absence of forest cover and soon bring down
masonry and other structures. There is, we suggest, a need for prompt and serious attention to
the archaeology of Rapa well beyond that which our resources could meet.
Among those aspects of Rapan archaeology which demand attention are these:
1. Establishing the sequence of cultural landscape development in much more detail. We
used AMS 14C dating as a survey technique (Chapter 11) to establish the broad pattern
of landscape use, but that is a relatively crude, if initially useful, approach, which, ideally,
would be superseded by high-density dating to determine temporal relationships between
site clusters, types, localities (such as the individual valley systems), and individual sites
and stratigraphy. Some of this research needs to be directed at the major colonisation sites,
especially Tangarutu.
2. One particular set of archaeological features, critical to understanding the development
of Rapan prehistory, and which has yet to be investigated systematically, consists of the
pondfields (Chapter 13) and associated structures. The history of these offers a means
of testing conjecture about the trajectory and chronology of population growth and the
development of pressure on resources that underlies an economic proposition for the rise of
warfare and fortifications.
3. Another set of features, mostly located on low ground behind the coast, or on low hills and
ridges, and consisting mainly of terraces, but also ovens, hearths, burial areas defined by
stone kerbing, and the ethnographically recorded remains of rudimentary marae, may refer
to what is currently a poorly resolved middle phase in Rapan prehistory. We, and others
before us, have barely touched on these features.
4. Much more excavation is needed, preferably in large open areas, of the surviving pare.
Too little is known of how habitation on these sites was organised, whether there were
houses of different sizes and locations, to what extent different economic and technological
functions were carried out at them, how long settlement lasted at different sites, whether
agricultural storage was indeed a major purpose, and whether conjecture about sociopolitical relationships between major and minor pare can be sustained.
terra australis 37
256 Atholl Anderson, Douglas J. Kennett and Eric Conte
Note
1.
There are very small lagoons on Norfolk Island.
References
Anderson, A.J. 1981. The archaeology of Raoul Island and its place in the settlement history of
Polynesia. Archaeology and Physical Anthropology in Oceania 15:131–141.
Anderson, A.J. 1996. Origins of Procellariidae Hunting in the Southwest Pacific. International Journal
of Osteoarchaeology 6:1–8.
Anderson, A.J. 2001. No meat on that beautiful shore: the prehistoric abandonment of subtropical
Polynesian islands. In: Anderson, A.J. and Leach, B.F. (eds), Zooarchaeology of Oceanic Coasts and
Islands: Papers from the 8th International Congress of the International Council of Archaeozoology,
23–29 August 1998, Victoria B.C., Canada. Special Issue of the International Journal of
Osteoarchaeology 11:14–23.
Anderson, A.J. 2002. Faunal collapse, landscape change and settlement history in Remote Oceania.
World Archaeology 33:375–390.
Anderson, A.J. and White, J.P. (eds), 2001. The Prehistoric Archaeology of Norfolk Island, Southwest
Pacific. Records of the Australian Museum, Supplement 27, Sydney.
Buck, Sir P.H. 1954. Vikings of the Sunrise. Whitcombe and Tombs, Christchurch.
Heyerdahl, T. and Ferdon, E.W. (eds), 1965. Reports of the Norwegian Archaeological Expedition to
Easter Island and the East Pacific, Volume 2 Miscellaneous Papers. Monographs of the School of
American Research and the Kon-Tiki Museum, 24 Pt. 2, Esselte AB, Stockholm.
Lamb, W.K. (ed), 1984. George Vancouver, A Voyage of Discovery to the North Pacific Ocean and Round
the World 1791–1795, Volume I. The Hakluyt Society, London.
Mieth, A. and Bork, H-R. 2004. Easter Island – Rapa Nui: scientific pathways to secrets of the past.
Man and Environment 1. Christian-Albrechts-Universität zu Kiel, Kiel.
Walczak, J. 2003. Presentation des données actuelles sur la préhistoire de Rapa Iti (archiple des
Australes-Polynésie Française). In: Orliac, C. (ed), Archéologie en Océanie Insulaire: Peuplement,
sociétés et paysages, pp. 28–45. Editions Artcom, Paris.
Weisler, M.I. 1996. Taking the mystery out of the Polynesian ‘mystery’ islands: a case study from
Mangareva and the Pitcairn group. In: Davidson, J., Irwin, G., Leach, F., Pawley, A. and Brown,
D. (eds), Oceanic Culture History: essays in honour of Roger Green. New Zealand Journal of
Archaeology, Special Publication, pp. 615–629. Dunedin.
Wilmshurst, J., Hunt, T., Lipo, C. and Anderson, A. 2011. High-precision radiocarbon dating shows
recent and rapid initial human colonization of East Polynesia. Proceedings of the National Academy
of Sciences 105:7676–7680.
terra australis 37
Appendices
Appendix A: Reference and identification data OAL Reference Collection
Family
Elements not held in collection
Acanthuridae
Anguillidae
Ceratohyal, cleithrum, epihyal, hyomandibular, maxilla, opercular, palatine, parasphenoid, pharyngeal plate, post
temporal, preopercular, scapula, vomer
Balistidae
Belonidae
Ceratohyal, epihyal, palatine, post temporal, vomer
Bothidae
Parasphenoid, pharyngeal plate, scapula, vomer
Carangidae
Chaetodontidae
Articular, ceratohyal, epihyal, hyomandibular, maxilla, palatine, parasphenoid, pharyngeal plate, post temporal, scapula,
vomer
Chanidae
Pharyngeal plate, scapula, vomer
Cirrhitidae
Ceratohyal, cleithrum, epihyal, hyomandibular, opercular, palatine, parasphenoid, pharyngeal plate, post temporal,
preopercular, scapula, vomer
Congridae
Cleithrum, maxilla, palatine, pharyngeal plate, post temporal, preopercular, scapula
Coryphaenidae
Diodontidae
Ceratohyal, cleithrum, epihyal, hyomandibular, opercular, palatine, parasphenoid, post temporal, preopercular, scapula,
vomer
Exocoetidae
Post temporal, vomer
Fistularidae
*reference specimen provided by Dr Marshall Weisler
Gempylidae
Ceratohyal, cleithrum, epihyal, hyomandibular, opercular, palatine, parasphenoid, pharyngeal plate, post temporal,
preopercular, scapula, vomer
Holocentridae
Kyphosidae
Labridae
Lethrinidae
Lutjanidae
Monocanthidae
Palatine, pharyngeal plate, post temporal
Mugilidae
Mullidae
Muraenidae
Ceratohyal, cleithrum, epihyal, maxilla, opercular, palatine, pharyngeal plate, post temporal, preopercular, scapula
Pempheridae
Pomacanthidae
Palatine, pharyngeal plate, vomer
Pomacentirdae
Parasphenoid, vomer
Pricanthidae
Pharyngeal plate
Continued on next page
terra australis 37
258 Appendices
Family
Elements not held in collection
Scaridae
Scombridae
Scorpanidae
Serranidae
Siganidae
Ceratohyal, epihyal, palatine, pharyngeal plate, post temporal, vomer
Syphyraenidae
Appendix B: Element counts and taxonomic identification of fish
Tangarutu E1/E2 Level III
Family
Element
Acanthuridae
Palatine
Premaxilla
1
Preopercular
1
1
2
1
1
4
6
2
Ceratohyal
1
Epihyal
1
Opercular
2
1
2
3
2
4
4
4
4
4
4
12
Articular
1
1
Dentary
1
1
Belonidae Total
1
1
1
1
3
1
4
Articular
1
1
2
Preopercular
1
1
2
Quadrate
1
2
3
3
4
7
Cleithrum
9
9
18
Opercular
10
6
16
Preopercular
15
7
22
34
22
56
Articular
2
2
4
Dentary
5
4
9
Hyomandibular
2
1
3
Premaxilla
3
5
8
Quadrate
2
1
3
cf. Aulostomidae Total
Chaetodontidae Total
Vomer
Congridae Total
1
Scapula
Preopercular
Congridae
1
1
Opercular
Chaetodontidae
Total
1
Apode Total
cf. Aulostomidae
Right
1
Unidentified element
Belonidae
N/A
Quadrate
Acanthuridae Total
Apode
Left
1
14
1
1
13
28
Continued on next page
terra australis 37
Appendices
259
Tangarutu E1/E2 Level III continued
Family
Element
Diodontidae
Dermal spine
Left
Diodontidae Total
Elasmobranchii
Tooth
Elasmobranchii Total
Exocoetidae
N/A
4
4
4
6
6
6
6
1
1
Scapula
Exocoetidae Total
1
Articular
1
Cleithrum
2
3
1
2
1
1
1
1
3
1
1
3
6
Articular
1
1
Post temporal
1
1
2
2
Articular
1
1
Cleithrum
2
1
3
Dentary
4
1
5
Hyomandibular
2
2
4
4
Maxilla
2
1
3
Opercular
1
4
5
Premaxilla
12
9
21
Preopercular
3
1
4
Quadrate
1
1
2
2
2
20
52
Superior pharyngeal plate
Labridae Total
28
Articular
1
Dentary
5
4
1
1
6
Hyomandibular
1
1
Premaxilla
1
1
Preopercular
1
1
Quadrate
1
1
5
11
Articular
1
1
Dentary
1
1
Opercular
1
1
Lethrinidae Total
6
Premaxilla
1
Scapula
Mullidae Total
1
1
Inferior pharyngeal plate
Mullidae
1
1
Kyphosidae Total
Lethrinidae
1
Preopercular
Holocentridae Total
Labridae
1
Hyomandibular
Scapula
Kyphosidae
Total
4
Hyomandibular
Inferior pharyngeal plate
Holocentridae
Right
1
1
1
1
4
5
Continued on next page
terra australis 37
260 Appendices
Tangarutu E1/E2 Level III continued
Family
Muraenidae
Element
Left
N/A
Articular
8
7
15
20
24
44
8
3
11
2
2
4
14
Premaxilla
Quadrate
10
Vomer
Muraenidae Total
11
46
11
Opercular
NIC 10 Total
NIC 11
Opercular
NIC 11 Total
NIC 12
Opercular
NIC 12 Total
NIC 13
Opercular
NIC 13 Total
NIC 14
1
2
2
2
2
Palatine
Dentary
Scapula
Scapula
Inferior pharyngeal plate
Superior pharyngeal plate
Vomer
Hyomandibular
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
Cleithrum
7
1
8
Opercular
2
1
3
2
2
1
1
2
10
5
15
3
3
Post temporal
Scapula
Pomacentridae Total
Scaridae
1
3
NIC 6 Total
Pomacentridae
1
1
NIC 45 Total
NIC 6
1
2
NIC 34 Total
NIC 45
1
3
NIC 33 Total
NIC 34
1
1
NIC 23 Total
NIC 33
97
1
NIC 22 Total
NIC 23
40
1
NIC 19 Total
NIC 22
11
2
NIC 14 Total
NIC 19
Total
Dentary
Hyomandibular
NIC 10
Right
4th Epibranchial
Articular
2
5
7
Ceratohyal
2
1
3
Cleithrum
2
Dentary
7
3
10
Epihyal
1
1
2
2
Continued on next page
terra australis 37
Appendices
261
Tangarutu E1/E2 Level III continued
Family
Element
Left
Total
4
10
6
Inferior pharyngeal plate
1
Maxilla
1
1
2
Opercular
5
9
14
Palatine
2
6
8
Premaxilla
13
14
4
4
14
10
24
Preopercular
7
6
13
Quadrate
5
1
6
Scapula
6
5
11
19
41
Superior pharyngeal plate
21
Toothed element
Scaridae Total
1
6
82
24
6
74
180
Articular
4
11
15
Ceratohyal
9
3
12
Cleithrum
5
5
10
Dentary
11
16
27
Epihyal
6
3
9
Hyomandibular
2
3
5
Maxilla
14
11
25
Opercular
12
16
28
Palatine
4
3
7
Parasphenoid
Post temporal
5
5
3
5
8
Premaxilla
12
11
23
Preopercular
15
16
31
Quadrate
11
7
18
Scapula
5
Vomer
Serranidae Total
Unidentified
Right
Hyomandibular
Parasphenoid
Serranidae
N/A
2
113
Basioccipital
Basipterygium
5
1
Coracoid
9
Dorsal spine
Dorsal spine/pterygiophore
Fragments
Hyomandibular
110
5
1
230
2
1
Cleithrum
Epihyal
7
2
2
Ceratohyal
Ectopterygoid
5
10
1
2
3
7
17
222
222
57
57
8
6
1
14
1
885
885
1
1
Continued on next page
terra australis 37
262 Appendices
Tangarutu E1/E2 Level III continued
Family
Element
Left
Hypural
Hypural 1 and 2
Hypural 3 and 4
2
Hypural 5
Identifiable
Interopercular
4
Maxilla
N/A
Right
Total
4
4
13
13
10
12
4
4
209
209
29
33
2
2
Mesopterygoid
1
1
Metapterygoid
6
6
Misc spines and rays
Opercular
1
Parasphenoid
Preopercular
4
5
1
Scale
Supracleithrum
291
3
Pterygiophore
Subopercular
291
1
5
8
114
114
1360
1360
7
2
2
11
10
2
4
16
Toothed element
13
13
Urohyal
4
4
Vertebra
728
728
Unidentified Total
Grand Total
57
3959
32
4048
412
4023
353
4788
Left
N/A
Right
Total
Tangarutu E1/E2 Level II
Family
Element
Acanthuridae
Quadrate
Acanthuridae Total
Apode
1
1
1
1
Basihyal
1
Ceratohyal
2
1
3
Epihyal
2
5
7
Opercular
4
5
9
Premaxilla
1
Unidentified element
Apode Total
Belonidae
9
Articular
Dentary
Chaetodontidae Total
1
5
Belonidae Total
Chaetodontidae
1
6
5
11
26
1
1
2
2
3
3
Cleithrum
11
10
21
Opercular
7
9
16
Preopercular
2
4
6
20
23
43
Continued on next page
terra australis 37
Appendices
263
Tangarutu E1/E2 Level II continued
Family
Element
Left
Congridae
Articular
Dentary
Right
Total
3
1
4
5
2
7
Hyomandibular
1
1
2
Premaxilla
2
3
5
Quadrate
1
1
2
Vomer
Congridae Total
Diodontidae
2
12
Dermal spine
Diodontidae Total
Holocentridae
N/A
2
1
1
Cleithrum
Kyphosidae Total
Labridae
1
1
1
1
1
1
1
2
4
Cleithrum
8
3
11
Dentary
2
2
4
1
1
2
4
2
Inferior pharyngeal plate
4
Maxilla
4
1
1
1
2
Palatine
1
Post temporal
1
Premaxilla
3
8
11
Preopercular
2
6
8
Quadrate
2
4
6
Scapula
1
1
2
Superior pharyngeal plate
1
Labridae Total
23
1
1
29
56
1
2
3
1
2
3
Articular
3
3
6
Dentary
9
19
28
Hyomandibular
5
2
7
Dentary
Lethrinidae Total
Parasphenoid
4
2
2
Premaxilla
2
5
7
Quadrate
2
2
4
Toothed element
1
Vomer
Muraenidae Total
1
2
Hyomandibular
Muraenidae
1
1
Epihyal
Lethrinidae
1
1
Post temporal
Quadrate
22
1
Preopercular
Maxilla
8
1
Holocentridae Total
Kyphosidae
2
1
12
21
15
12
31
67
Continued on next page
terra australis 37
264 Appendices
Tangarutu E1/E2 Level II continued
Family
Element
Left
NIC 13
Opercular
1
1
1
1
NIC 13 Total
NIC 15
Palatine
NIC 15 Total
NIC 16
N/A
1
1
1
Premaxilla
Premaxilla
NIC 17 Total
NIC 24
Articular
NIC 24 Total
NIC 30
Ceratohyal
NIC 30 Total
NIC 4
1
1
1
1
1
1
1
1
1
1
Scute
2
2
2
2
1
1
1
Cleithrum
Pomacentridae Total
4th epibranchial
1
1
1
1
1
Scapula
1
1
1
2
3
3
3
6
1
1
Articular
Cleithrum
1
Dentary
3
4
7
Epihyal
1
2
3
Hyomandibular
1
3
4
Inferior pharyngeal plate
1
4
4
Maxilla
1
2
3
Opercular
1
2
3
Palatine
2
Parasphenoid
2
2
2
Post temporal
2
1
3
Premaxilla
5
1
6
Preopercular
1
3
4
Quadrate
2
1
3
6
18
Scapula
Superior pharyngeal plate
2
Scaridae Total
3
37
Articular
9
3
29
2
Ceratohyal
Cleithrum
2
12
Toothed element
Serranidae
1
1
Hyomandibular
Preopercular
Scaridae
1
1
NIC 41 Total
Pomacentridae
1
1
NIC 4 Total
NIC 41
Total
1
NIC 16 Total
NIC 17
Right
2
75
2
1
1
3
5
Continued on next page
terra australis 37
Appendices
265
Tangarutu E1/E2 Level II continued
Family
Element
Left
3
6
Epihyal
1
3
4
Hyomandibular
4
2
6
Maxilla
2
1
3
Opercular
2
1
3
Palatine
1
1
2
2
2
3
5
Premaxilla
3
2
5
Preopercular
4
2
6
Quadrate
3
1
4
Scapula
2
1
3
24
57
1
1
1
1
31
2
Preopercular
Basipterygium
1
2
Ceratohyal
Coracoid
3
Dorsal spine
Dorsal spine/pterygiophore
Ectopterygoid
2
Fragments
3
1
1
3
6
158
158
35
35
1
2
5
670
670
Hypural 1 and 2
1
1
Hypural 3 and 4
2
2
Hypural 5
1
1
134
134
12
12
Identifiable
Interopercular
Maxilla
1
Misc spines and rays
Parasphenoid
1
2
340
340
2
2
Preopercular
1
1
Pterygiophore
109
109
Scale
709
709
1
1
Scapula
Grand Total
2
Post temporal
Siganidae Total
Unidentified Total
Total
3
Serranidae Total
Unidentified
Right
Dentary
Parasphenoid
Siganidae
N/A
Subopercular
5
Supracleithrum
9
1
6
11
7
17
Toothed element
4
4
Urohyal
3
3
Vertebra
478
478
21
2663
21
2705
183
2704
190
3077
terra australis 37
266 Appendices
Tangarutu E1/E2 Level I
Family
Element
Left
Acanthuridae
Cleithrum
1
Dorsal spine
N/A
1
Acanthuridae Total
Apode
1
1
2
2
2
4
1
1
3
3
Hyomandibular
1
Metapterygoid
1
Preopercular
1
Quadrate
1
3
3
4
10
1
1
1
2
1
4
Scapula
1
5
1
1
1
1
Cleithrum
6
8
14
Opercular
7
2
9
3
3
13
26
Articular
2
2
Dentary
2
2
Chaetodontidae Total
13
Hyomandibular
1
Parasphenoid
Premaxilla
Congridae Total
1
2
2
1
2
Dermal spine
Diodontidae Total
1
2
4
8
3
3
3
3
Preopercular
Kyphosidae Total
Labridae
6
2
Preopercular
Kyphosidae
2
Opercular
Carangidae Total
Diodontidae
2
2
Belonidae Total
Congridae
1
1
Apode Total
Chaetodontidae
1
1
Unidentified element
Carangidae
1
Epihyal
Premaxilla
Belonidae
1
1
Pterygiophore
Total
1
Opercular
Scapula
Right
1
1
1
1
Articular
1
Cleithrum
1
1
2
Hyomandibular
1
4
5
Inferior pharyngeal plate
1
2
Maxilla
2
1
1
8
10
Preopercular
1
1
Quadrate
1
1
2
3
Premaxilla
2
Scapula
1
Superior pharyngeal plate
1
1
Continued on next page
terra australis 37
Appendices
267
Tangarutu E1/E2 Level I continued
Family
Element
Labridae Total
Monocanthidae
Cleithrum
Monocanthidae Total
Mugilidae
Premaxilla
Mugilidae Total
Mullidae
N/A
Right
Total
7
2
18
27
1
1
1
1
2
2
2
2
Dentary
Hyomandibular
Mullidae Total
Muraenidae
Left
1
1
1
1
1
2
Articular
2
2
4
Dentary
13
9
22
3
2
5
Hyomandibular
Parasphenoid
1
3
1
4
Quadrate
3
2
5
Toothed element
2
2
Vomer
9
9
24
12
Opercular
NIC 12 Total
NIC 18
Premaxilla
NIC 18 Total
NIC 21
Scapula
NIC 21 Total
NIC 25
Articular
NIC 25 Total
NIC 27
Epihyal
NIC 27 Total
NIC 28
NIC 4
NIC 43 Total
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Hyomandibular
NIC 42 Total
NIC 43
1
1
1
Superior pharyngeal plate
Hyomandibular
2
1
NIC 4 Total
NIC 42
2
1
Ceratohyal
NIC 38 Total
2
1
Epihyal
Quadrate
2
1
NIC 35 Total
NIC 38
52
1
NIC 31 Total
NIC 35
16
1
NIC 28 Total
NIC 31
1
Premaxilla
Muraenidae Total
NIC 12
1
2
2
2
2
1
1
1
Opercular
1
1
1
1
1
Continued on next page
terra australis 37
268 Appendices
Tangarutu E1/E2 Level I continued
Family
Element
NIC 44
Vomer
Left
NIC 44 Total
NIC 45
Vomer
NIC 45 Total
NIC 9
N/A
1
1
1
1
1
1
1
Preopercular
Cleithrum
Inferior pharyngeal plate
Preopercular
Scapula
Pomacentridae Total
Scaridae
4th epibranchial
Articular
1
1
1
1
1
1
1
1
1
1
1
1
4
4
8
12
12
9
21
1
1
Ceratohyal
Cleithrum
5
2
7
Dentary
19
18
37
9
10
19
Hyomandibular
Inferior pharyngeal plate
Maxilla
29
29
3
4
7
Opercular
10
8
18
Palatine
6
5
11
1
1
20
13
33
Preopercular
3
7
10
Quadrate
9
9
18
Scapula
5
5
10
33
30
63
Post temporal
Premaxilla
Superior pharyngeal plate
Toothed element
4
4
Vomer
2
2
Scaridae Total
Serranidae
1
1
2
Total
1
NIC 9 Total
Pomacentridae
Right
130
303
Articular
138
3
35
3
6
Ceratohyal
2
1
3
Cleithrum
3
2
5
Dentary
3
3
Epihyal
2
2
Hyomandibular
2
3
5
Maxilla
1
7
8
Opercular
1
1
2
Parasphenoid
1
Post temporal
2
Premaxilla
5
Preopercular
1
3
5
5
1
1
Continued on next page
terra australis 37
Appendices
269
Tangarutu E1/E2 Level I continued
Family
Element
Left
Quadrate
6
Scapula
1
Vomer
Serranidae Total
Siganidae
8
3
1
2
28
57
1
2
1
1
1
2
1
Preopercular
1
1
Scapula
1
1
4
Basioccipital
Basipterygium
3
7
4
13
1
9
Ceratohyal
1
4
Cleithrum
1
Coracoid
2
4
1
2
4
Dorsal spine
92
92
Dorsal spine/pterygiophore
25
25
Ectopterygoid
2
Epihyal
Fragments
2
4
1
1
1155
1155
Hypural 1 and 2
6
6
Hypural 3 and 4
11
11
1
1
163
163
21
21
1
1
279
279
Hypural 5
Identifiable
Interopercular
Maxilla
Misc spines and rays
Opercular
6
Parasphenoid
2
Preopercular
Quadrate
1
Superior pharyngeal plate
Supracleithrum
1
2
2
547
1
4
13
2
123
547
2
7
2
123
Scale
Subopercular
1
2
Pterygiophore
5
4
10
23
Toothed element
2
2
Urohyal
1
1
Ventral hypohyal
Vertebra
Grand Total
2
Premaxilla
Siganidae Total
Unidentified Total
Total
1
Maxilla
Unidentified
Right
2
26
Hyomandibular
N/A
1
1
519
519
30
2966
24
3020
264
3032
256
3552
terra australis 37
270 Appendices
Tangarutu R1 Level III
Family
Element
Left
Acanthuridae
Cleithrum
1
1
1
1
Acanthuridae Total
Apode
Ceratohyal
1
2
1
1
Dentary
1
1
Epihyal
2
2
Opercular
1
1
4
1
16
Opercular
2
2
4
Preopercular
1
3
4
6
18
24
Articular
3
5
8
Dentary
4
3
7
Hyomandibular
2
2
4
Premaxilla
5
1
6
Quadrate
2
1
3
3
16
Dermal spine
Diodontidae Total
Vertebra
Elasmobranchii Total
Ceratohyal
Holocentridae Total
Labridae
3
12
31
2
2
2
1
1
1
1
1
1
1
1
Cleithrum
2
Dentary
2
3
5
2
1
Inferior pharyngeal plate
7
1
7
Maxilla
3
2
5
Premaxilla
5
8
13
Preopercular
3
2
5
Quadrate
3
Vomer
Labridae Total
3
1
18
Dentary
Post temporal
8
1
16
42
1
1
1
Premaxilla
1
1
1
1
2
3
Articular
4
3
7
Dentary
19
8
27
Mullidae Total
Muraenidae
3
2
Hyomandibular
Mullidae
11
13
Congridae Total
Holocentridae
6
3
Vomer
Elasmobranchii
4
4
Cleithrum
Chaetodontidae Total
Diodontidae
Total
1
Apode Total
Congridae
Right
Ceratohyal
Unidentified element
Chaetodontidae
N/A
Continued on next page
terra australis 37
Appendices
271
Tangarutu R1 Level III continued
Family
Element
Hyomandibular
Left
N/A
2
Parasphenoid
2
Toothed element
NIC 13
NIC 13 Total
NIC 17
8
NIC 37 Total
NIC 39
Quadrate
NIC 39 Total
NIC 4
NIC 40 Total
Pomacentridae
Cleithrum
1
5
5
1
4
5
1
1
1
1
1
1
2
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
2
4th epibranchial
3
1
4
Articular
4
3
7
Ceratohyal
2
2
Cleithrum
1
1
2
5
Dentary
3
Epihyal
1
Hyomandibular
1
Inferior pharyngeal plate
1
1
4
2
4
Maxilla
2
4
6
Opercular
2
1
3
Palatine
2
Parasphenoid
Scaridae Total
4
1
Pomacentridae Total
Scaridae
2
4
Hyomandibular
Quadrate
4
1
NIC 4 Total
NIC 40
4
56
Premaxilla
Quadrate
2
21
NIC 17 Total
NIC 37
6
5
27
Opercular
4
1
Vomer
Muraenidae Total
Total
2
Premaxilla
Quadrate
Right
2
1
1
Post temporal
2
Post temporal
1
Premaxilla
3
6
9
Preopercular
3
3
6
Quadrate
3
1
4
Scapula
1
2
3
Superior pharyngeal plate
4
5
9
33
73
35
1
3
1
5
Continued on next page
terra australis 37
272 Appendices
Tangarutu R1 Level III continued
Family
Element
Left
Serranidae
Articular
Right
Total
3
2
5
Cleithrum
5
4
9
Dentary
5
4
9
Epihyal
2
3
5
Hyomandibular
2
2
4
Maxilla
5
5
Opercular
1
1
Palatine
2
2
Parasphenoid
1
1
Post temporal
1
1
2
Premaxilla
2
2
4
Preopercular
2
1
3
Scapula
1
Vomer
Serranidae Total
Unidentified
N/A
1
1
28
Basioccipital
2
1
22
3
Basipterygium
2
Cleithrum
5
Coracoid
3
52
3
1
3
5
4
7
Dermal spine
1
1
Dorsal spine
73
73
Dorsal spine/pterygiophore
Ectopterygoid
31
1
Fragments
3
31
2
6
953
953
Hyomandibular
1
1
Hypural
1
1
Hypural 1 and 2
5
5
Hypural 3 and 4
Identifiable
Interopercular
Misc spines and rays
Opercular
1
Parasphenoid
Post temporal
Preopercular
5
5
46
46
6
6
138
138
1
2
2
2
1
1
1
Pterygiophore
Scale
1
32
32
202
202
Subopercular
3
2
5
Supracleithrum
3
4
7
Toothed element
1
1
Urohyal
1
1
Vertebra
233
233
Unidentified Total
Grand Total
terra australis 37
19
1739
13
1771
158
1772
151
2081
Appendices
273
Tangarutu R1 Level II
Family
Element
Apode
Basihyal
Ceratohyal
Left
N/A
1
Apode Total
cf. Aulostomidae
1
1
1
1
2
2
5
1
1
1
1
1
1
2
1
1
2
1
2
1
Opercular
cf. Aulostomidae Total
Chaetodontidae
Cleithrum
Chaetodontidae Total
Congridae
Parasphenoid
Premaxilla
2
Kyphosidae
1
1
3
3
5
6
Articular
Kyphosidae Total
Labridae
1
1
2
1
1
Premaxilla
1
1
Preopercular
1
1
4
Dentary
Articular
1
Family
Element
Left
Premaxilla
Mullidae Total
1
5
1
1
1
1
1
N/A
Right
Total
1
1
2
2
Articular
2
2
9
15
Hyomandibular
3
3
Premaxilla
3
3
Quadrate
1
1
Dentary
6
Vomer
Muraenidae Total
2
6
2
Palatine
NIC 15 Total
Scaridae
1
Opercular
Mullidae
NIC 15
1
1
Lethrinidae Total
Muraenidae
1
Cleithrum
Labridae Total
Lethrinidae
2
1
Vomer
Congridae Total
Total
1
1
Epihyal
Opercular
Right
2
18
26
1
1
1
1
Ceratohyal
1
Dentary
1
1
2
Hyomandibular
1
2
3
Inferior pharyngeal plate
Maxilla
7
7
1
Opercular
Palatine
1
1
3
1
3
1
Continued on next page
terra australis 37
274 Appendices
Tangarutu R1 Level II continued
Family
Element
Premaxilla
Left
N/A
3
Quadrate
2
5
2
2
1
1
2
Superior pharyngeal plate
2
3
5
14
32
1
1
1
2
11
7
Articular
Cleithrum
1
Dentary
2
2
Hyomandibular
Opercular
1
2
1
Parasphenoid
Premaxilla
Serranidae Total
1
2
7
Coracoid
Dorsal spine/pterygiophore
Ectopterygoid
Identifiable
Pterygiophore
Scale
5
13
3
5
5
5
1
4
398
3
3
32
32
1
57
57
8
8
55
55
1
Vertebra
Grand Total
3
19
Subopercular
Unidentified Total
1
1
Misc spines and rays
Supracleithrum
1
398
Hypural 3 and 4
1
19
3
Fragments
Maxilla
1
2
Dorsal spine
1
2
Palatine
Unidentified
Total
Scapula
Scaridae Total
Serranidae
Right
2
2
1
2
53
53
7
630
7
644
41
646
52
739
Left
N/A
Right
Total
Tangarutu T1
Family
Element
Holocentridae
Dentary
Holocentridae Total
NIC 36
1
1
1
1
Quadrate
NIC 36 Total
Serranidae
Maxilla
Serranidae Total
Unidentified
Dorsal spine
1
Fragments
15
2
2
2
2
1
1
1
1
1
15
Continued on next page
terra australis 37
Appendices
275
Tangarutu T1 continued
Family
Element
Left
N/A
Right
Total
Misc spines and rays
1
1
Pterygiophore
1
1
22
22
5
5
Scale
Vertebra
Unidentified Total
45
Grand Total
45
1
45
3
49
Left
N/A
Right
Total
Tangarutu T3
Family
Element
Unidentified
Dorsal spine
3
3
Dorsal spine/pterygiophore
2
2
Fragments
13
13
Identifiable
1
1
Interopercular
1
1
Misc spines and rays
3
3
Pterygiophore
2
2
72
72
Scale
Supracleithrum
1
Vertebra
1
9
2
9
Unidentified Total
1
106
1
108
Grand Total
1
106
1
108
Left
N/A
Right
Total
Akatanui 1
Family
Element
Acanthuridae
Parasphenoid
Acanthuridae Total
Congridae
Parasphenoid
Congridae Total
Elasmobranchii
Vertebra
Elasmobranchii Total
Labridae
Inferior pharyngeal plate
Labridae Total
NIC 17
Premaxilla
NIC 17 Total
NIC 20
NIC 5 Total
1
1
1
1
1
1
2
2
2
2
1
1
1
1
1
1
1
Scapula
1
1
1
1
Ceratohyal
1
1
Epihyal
1
1
2
2
NIC 26 Total
NIC 5
1
1
NIC 20 Total
NIC 26
1
Hyomandibular
2
2
2
2
Continued on next page
terra australis 37
276 Appendices
Akatanui 1 continued
Family
Element
NIC 7
Hyomandibular
1
1
1
1
1
4th epibranchial
1
1
Dentary
1
Hyomandibular
Maxilla
3
1
1
1
Parasphenoid
Premaxilla
4
1
1
Opercular
1
4
Quadrate
1
1
5
9
1
1
Scapula
1
1
2
Superior pharyngeal plate
1
3
4
Toothed element
Scaridae Total
3
9
5
3
14
28
Articular
1
1
Ceratohyal
3
3
Dentary
1
1
Opercular
1
1
Palatine
1
1
Post temporal
2
1
3
Premaxilla
1
1
Preopercular
2
2
Quadrate
Serranidae Total
Unidentified
Total
1
Inferior pharyngeal plate
Serranidae
Right
1
NIC 8 Total
Scaridae
N/A
1
NIC 7 Total
NIC 8
Left
1
1
5
Basioccipital
Basipterygium
9
1
1
2
2
Coracoid
1
Dorsal spine
14
1
1
1
116
116
Hypural 1 and 2
3
3
Hypural 3 and 4
3
3
25
25
2
2
24
24
1
1
12
12
33
33
Fragments
Identifiable
Interopercular
Misc spines and rays
Opercular
Pterygiophore
Scale
Subopercular
1
Supracleithrum
2
1
1
3
Continued on next page
terra australis 37
Appendices
277
Akatanui 1 continued
Family
Element
Left
N/A
Right
Total
Urohyal
1
1
Vertebra
54
54
Unidentified Total
Grand Total
5
276
2
283
24
286
28
338
Left
N/A
Right
Total
Akatanui 3 C1
Family
Element
Acanthuridae
Dorsal spine
1
Acanthuridae Total
Apode
1
Epihyal
1
Opercular
2
Unidentified element
Apode Total
Bothidae
4
3
4
8
Ceratohyal
1
1
Epihyal
1
1
2
2
Articular
1
1
1
Articular
Cleithrum
2
2
Opercular
2
1
3
4
1
5
Articular
5
4
9
Dentary
4
8
12
Hyomandibular
2
1
3
3
Premaxilla
3
Quadrate
1
Congridae Total
15
Dermal spine
Vertebra
Elasmobranchii Total
Inferior pharyngeal plate
Quadrate
Exocoetidae Total
Holocentridae Total
3
3
1
1
Diodontidae Total
Holocentridae
2
1
Vomer
Exocoetidae
1
1
Parasphenoid
Elasmobranchii
1
1
Chaetodontidae Total
Diodontidae
1
1
cf. Aulostomidae Total
Congridae
4
1
Carangidae Total
Chaetodontidae
2
2
Opercular
cf. Aulostomidae
1
1
Bothidae Total
Carangidae
1
4
1
14
33
1
1
1
1
13
13
13
13
1
1
1
2
1
1
2
1
Cleithrum
1
1
Hyomandibular
1
1
2
2
Continued on next page
terra australis 37
278 Appendices
Akatanui 3 C1 continued
Family
Element
Kyphosidae
Ceratohyal
1
1
2
3
2
2
3
5
Dentary
3
3
Hyomandibular
1
1
1
Articular
2
2
Premaxilla
8
2
10
Preopercular
1
1
Quadrate
1
1
Superior pharyngeal plate
1
1
14
31
1
2
Opercular
1
1
Palatine
1
1
1
3
4
Articular
1
5
6
Dentary
6
9
15
12
Hyomandibular
5
1
Parasphenoid
2
2
Toothed element
1
1
Vomer
7
7
7
Opercular
NIC 13 Total
Ceratohyal
NIC 32 Total
10
1
1
1
1
1
1
Hyomandibular
Vomer
1
1
1
1
1
1
1
1
Cleithrum
23
19
42
Opercular
11
5
16
2
2
Post temporal
Preopercular
1
4
5
Scapula
4
6
10
Pomacentridae Total
Scaridae
31
1
NIC 45 Total
Pomacentridae
14
1
NIC 4 Total
NIC 45
5
2
Muraenidae Total
NIC 4
5
Opercular
Mullidae Total
NIC 32
1
Quadrate
Labridae Total
NIC 13
Total
1
Inferior pharyngeal plate
Muraenidae
Right
1
Cleithrum
Mullidae
N/A
1
Cleithrum
Kyphosidae Total
Labridae
Left
39
36
75
4th epibranchial
1
2
3
Articular
1
1
2
1
1
Ceratohyal
Continued on next page
terra australis 37
Appendices
279
Akatanui 3 C1 continued
Family
Element
Left
Right
Total
Cleithrum
4
3
7
Dentary
11
4
15
4
5
9
Hyomandibular
Inferior pharyngeal plate
N/A
26
Opercular
9
8
17
Palatine
7
6
13
Parasphenoid
1
Post temporal
Premaxilla
Preopercular
2
2
2
5
7
15
6
21
6
2
8
Scapula
1
Scaridae Total
1
23
84
27
10
33
55
166
Articular
4
1
5
Ceratohyal
4
1
5
Cleithrum
1
6
7
Dentary
7
5
12
Epihyal
3
1
4
Hyomandibular
2
2
4
Opercular
1
Palatine
2
Parasphenoid
1
3
1
Post temporal
5
1
2
2
Premaxilla
5
4
9
Preopercular
2
4
6
Quadrate
5
5
10
2
2
36
75
3
8
1
1
Scapula
Vomer
Serranidae Total
Unidentified
1
Quadrate
Superior pharyngeal plate
Serranidae
26
2
36
Basipterygium
3
5
Ceratohyal
2
Cleithrum
1
3
4
Coracoid
4
8
12
Dorsal spine
Dorsal spine/pterygiophore
Ectopterygoid
Fragments
117
117
57
57
3
1
388
Hyomandibular
Hypural
4
388
2
2
3
3
Hypural 1 and 2
10
10
Hypural 3 and 4
10
10
Continued on next page
terra australis 37
280 Appendices
Akatanui 3 C1 continued
Family
Element
Left
Hypural 5
N/A
Right
Total
1
1
166
166
Interopercular
32
32
Metapterygoid
1
Identifiable
Misc spines and rays
1
2
178
178
Opercular
2
2
Parasphenoid
2
2
Post temporal
Preopercular
2
1
Pterygiophore
Scale
2
1
69
69
792
792
Subopercular
4
2
5
11
Supracleithrum
3
2
4
9
Toothed element
2
2
Urohyal
2
2
Vertebra
144
144
Unidentified Total
Grand Total
21
1980
30
2031
230
2050
210
2490
Left
N/A
Right
Total
1
1
Akatanui 3 E1
Family
Element
Acanthuridae
Opercular
Parasphenoid
1
Acanthuridae Total
Scaridae
1
Dentary
1
Hyomandibular
1
1
2
3
3
Parasphenoid
1
1
2
Quadrate
Superior pharyngeal plate
Scaridae Total
2
6
4
1
1
1
3
1
1
1
3
5
15
Hyomandibular
2
2
Preopercular
1
1
3
3
Serranidae Total
terra australis 37
2
Inferior pharyngeal plate
Premaxilla
Unidentified
1
1
Post temporal
Serranidae
1
Dorsal spine
1
1
Fragments
59
59
Hypural
1
1
Identifiable
5
5
Interopercular
4
4
Misc spines and rays
3
3
Continued on next page
Appendices
281
Akatanui 3 E1 continued
Family
Element
Left
N/A
Right
Total
Parasphenoid
1
1
Pterygiophore
5
5
13
13
Scale
Supracleithrum
1
Vertebra
9
Unidentified Total
Grand Total
1
9
101
1
102
9
106
7
122
Left
N/A
Right
Total
Akatanui 3 A1
Family
Element
Scaridae
Articular
1
1
Palatine
1
1
Premaxilla
1
1
Preopercular
Scaridae Total
Serranidae
1
1
1
3
4
Hyomandibular
1
1
Preopercular
1
1
2
2
Serranidae Total
Unidentified
Fragments
2
2
Identifiable
1
1
Unidentified Total
3
Grand Total
3
1
3
5
9
Left
N/A
Right
Total
Angairao C
Family
Element
Labridae
Premaxilla
Labridae Total
Muraenidae
2
1
1
2
1
1
2
Inferior pharyngeal plate
Scaridae Total
Siganidae
2
2
Dentary
Muraenidae Total
Scaridae
2
1
1
1
1
Preopercular
Siganidae Total
Unidentified
1
1
1
Dorsal spine
2
2
Fragments
17
17
Identifiable
1
1
Misc spines and rays
1
1
Pterygiophore
1
1
Vertebra
2
2
24
24
Unidentified Total
Grand Total
1
3
25
2
30
terra australis 37
282 Appendices
Angairao E Level III
Family
Element
Kyphosidae
Ceratohyal
Left
N/A
Kyphosidae Total
Labridae
Cleithrum
Labridae Total
Muraenidae
Muraenidae Total
Scaridae
Cleithrum
1
1
1
1
1
2
1
2
3
1
1
1
5
1
2
Preopercular
3
1
1
2
Superior pharyngeal plate
3
1
4
6
19
Cleithrum
1
1
Hyomandibular
1
1
8
5
2
2
Quadrate
2
Basipterygium
Dorsal spine
1
1
3
5
1
1
2
2
Ectopterygoid
1
1
Entopterygoid
1
1
Fragments
67
67
Hypural
1
1
Hypural 1 and 2
1
1
Identifiable
1
1
15
15
Pterygiophore
1
1
Scale
3
3
Metapterygoid
1
Misc spines and rays
Supracleithrum
1
1
Vertebra
terra australis 37
1
1
Serranidae Total
Grand Total
1
1
Preopercular
Unidentified Total
1
Quadrate
Scaridae Total
Unidentified
1
5
1
Palatine
Serranidae
1
1
1
Inferior pharyngeal plate
Premaxilla
1
1
1
Dentary
Opercular
Total
1
Articular
Dentary
Right
2
9
3
9
2
100
5
107
14
105
17
136
Appendices
283
Angairao E Level II
Family
Element
Acanthuridae
Scapula
Left
N/A
Acanthuridae Total
Congridae
Articular
Congridae Total
Labridae
Quadrate
Labridae Total
Muraenidae
Dentary
Serranidae
Serranidae Total
Unidentified
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Epihyal
Total
1
Vomer
Muraenidae Total
Right
1
1
1
2
1
1
1
1
Dorsal spine
2
2
Fragments
23
23
Interopercular
1
1
Misc spines and rays
3
3
Scale
10
10
Vertebra
5
5
44
44
Unidentified Total
Grand Total
3
45
2
50
Angairao E Level I
Family
Element
Elasmobranchii
Vertebra
Left
Elasmobranchii Total
Labridae
Parasphenoid
Labridae Total
Mullidae
N/A
1
1
1
1
1
1
1
Premaxilla
1
Dentary
2
2
3
3
Epihyal
Inferior pharyngeal plate
1
1
1
1
1
1
Superior pharyngeal plate
1
2
Post temporal
1
1
Preopercular
Scaridae Total
Serranidae Total
1
1
Opercular
Serranidae
1
1
NIC 29 Total
Scaridae
1
Articular
Muraenidae Total
NIC 29
Total
1
Mullidae Total
Muraenidae
Right
1
1
1
1
2
2
5
1
1
1
1
Continued on next page
terra australis 37
284 Appendices
Angairao E Level I continued
Unidentified
Dorsal spine
2
2
Dorsal spine/pterygiophore
1
1
Ectopterygoid
1
1
Fragments
42
42
Misc spines and rays
12
12
4
4
Scale
Supracleithrum
3
1
4
Toothed element
1
1
Vertebra
3
3
Unidentified Total
Grand Total
4
65
1
70
10
68
5
83
Left
N/A
Right
Total
Angairao N (Noogoriki)
Family
Element
Carangidae
Dentary
Carangidae Total
Muraenidae
Dentary
Muraenidae Total
Scaridae
1
1
1
1
1
1
1
Hyomandibular
1
1
Preopercular
1
1
3
1
4
3
3
6
Superior pharyngeal plate
Scaridae Total
Serranidae
1
Ceratohyal
1
1
Cleithrum
1
1
Hyomandibular
1
1
Opercular
Serranidae Total
Unidentified
3
terra australis 37
1
1
4
Dorsal spine
1
1
Dorsal spine/pterygiophore
1
1
Fragments
3
3
Misc spines and rays
3
3
Scale
7
7
15
15
Unidentified Total
Grand Total
1
8
15
4
27
Appendices
285
Appendix C: Analysis of faunal remains from upland fortifications sites
Virginia L. Butler
Department of Anthropology, Portland State University
This report summarises the faunal remains recovered during excavation at six upland fortification
sites tested in July/August 2002 on Rapa, French Polynesia (Table A). Remains were recovered
during field screening (1/8" mesh) and flotation of bulk samples. Materials were identified
to the finest taxon possible, using reference materials in Butler’s possession at Portland State
University. The “probable turtle” specimens were eroded, tabular fragments with porous
structure that did not resemble the texture of other large to medium-size vertebrates. These
specimens could be from any of the three marine turtles known for the Pacific (leatherback
Dermochelys coriacea, green turtle Chelonia mydas, hawksbill Eretmochelys imbricata). All of the
bones and teeth collected during field screening were documented and tallied. Unidentified
vertebrate specimens – nonfish, include materials which were clearly not from fish, but which
could be from other vertebrates, including turtle, marine mammal, pig, dog, or possibly human.
None of the remains appear to be from bird, based on texture and robusticity. These remains
were extremely fragmentary and poorly preserved. The flotation sample specimens tended to be
extremely small (many less than 2 mm in linear dimension) and many could not be distinguished
as fish or other vertebrate class. Thus, I only recorded those specimens from flotation samples
that could be identified below vertebrate class (e.g. family). Number of Identified Specimens
(NISP, Grayson 1984) was used to quantify the specimens.
Results
A total of 400 specimens were documented from the sites. Ten different vertebrate taxa were
identified altogether (Table A), including eight fish taxa, rat and probable turtle. Parrotfish far
and away dominates the fish fauna. This species was recovered at most sites and has the highest
frequency in individual sites. The prominence is partially due to bias in identification, given that
parrotfish teeth are more distinctive than those from other taxa. When teeth are excluded from
the tallies, the overall frequency (and ubiquity) of parrotfish is reduced (Table A), though it still
ranks highest of the fishes.
Method of sample collection needs to be considered in interpreting results. Overall,
the flotation samples provided many more identified specimens, especially of fish, than field
screening. At Tevaitau (R-18), where both field screening and flotation samples were used, the
flot samples generated 36 NISP (21 with teeth excluded); whereas the field screening provided
only four NISP (three when teeth were excluded). Fish faunal recovery was especially affected
by collection method. At R-18, only two fish taxa were identified in the field screening, whereas
seven were identified in the flot samples.
All of the fish could have been taken nearshore. They represent a variety of feeding ecologies
and would have likely been taken using different methods (e.g. Leach and Intoh 1984; Leach et
al. 1984; Allen 1992; Butler 1994, 2001). Thus the carnivores (groupers, snappers, moray eels)
were more apt to be taken with hook and line and the herbivores (parrotfishes, damselfishes,
puffers) were more likely taken with nets or traps.
Faunal representation greatly differs across sites. Turtle was only identified at Tapitanga (R4). Fish is also very scarce at this site, although the scarcity may be due to the fact that flotation
samples were not taken from this site. Tevaitau has the richest assemblage, with eight taxa
present. This may be an artefact of sampling, however, given the relatively large sample size at
the site and the commonly found association between sample size and richness (Grayson 1984).
terra australis 37
286 Appendices
Table A. Frequency of taxa by site and recovery method. Values in parenthesis are with teeth excluded.
Taxon
R-2
R-4
flot
1/8
flot
R-18
1/8
flot
Carangidae (k, pompano)
R-20
1/8
flot
flot
flot
1/8
1/8
1 (0)
Labridae (wrasse)
2
Muraenidae (moray eel)
2
Pomacentridae (damselfishes)
2
5 (0)
30 (15)
1
21 (6)
1 (0)
Serranidae (grouper, sea bass)
2
2
Tetraodontidae (pufferfishes)
1
Unid Fish
15 (0)
13 (4)
1
48
Rattus sp. (rat)
5
Probable Turtle
1
2
46
Unidentified Vertebrate (nonfish)
Total NISP (excludes unid fish,
unid vertebrate)
R-17
1
Elasmobranch (shark, ray)
Scaridae (parrotfishes)
1/8
R2002-49
196
5
76
1
36
4
1
16
1
17
References
Allen, M.S. 1992. Temporal variation in Polynesian fishing strategies: the southern Cooks in regional
perspective. Asian Perspectives 31:183–204.
Butler, V.L. 1994. Fish feeding behaviour and fish capture: the case for variation in Lapita fishing
strategies. Archaeology in Oceania 29:81–90.
Butler, V.L. 2001. Changing fish use on Mangaia, Southern Cook Islands: resource depression and the
prey choice model. International Journal of Osteoarchaeology 11:88–100.
Grayson, D.K. 1984. Quantitative Zooarchaeology, New York: Academic.
Leach, B.F. and Intoh, M. 1984. An archaeological fishbone assemblage from the Vitaria Site, Rurutu,
Austral Islands. Journal de la Societe des Oceanistes 78:75–77.
Leach, B.F., Intoh, M. and Smith, I.W.G. 1984. Fishing, turtle hunting, and mammal exploitation at
Fa’ahia, Huahine, French Polynesia. Journal de la Societe des Oceanistes 79:183–197.
terra australis 37
Appendices
287
Appendix D: UTM locations of archaeological sites investigated
in the current project
ID
Description
Northing
Easting
R-1 (Morongo Uta)
Fortification
6942536.95
761078.43
R-2 (Potaketake)
Fortification
6944304.05
763395.96
R-3 (Vairu)
Fortification
6945589.85
762849.25
R-4 (Tapitanga)
Fortification
6943851.50
763427.89
R-5 (Kapitangi)
Fortification
6945400.86
760810.75
R-17 (Ruatara)
Fortification
6945695.10
761746.92
R-18 (Tevaitau)
Fortification
6941795.40
761995.49
R-19 (Pukutai)
Secondary fortification
6943757.85
764180.76
R-20 (Ororangi)
Fortification
6941628.88
763452.88
R2002-1
Lithic scatter
6944363.89
761417.44
R2002-2
Domestic terrace
6944922.56
761844.26
R2002-3
Oven
6942355.87
762266.74
R2002-4
Garden soil
6942357.15
762265.81
R2002-5
Oven
6942428.79
762301.56
R2002-6
Garden soil
6943206.33
761630.33
R2002-7
Oven
6943084.84
761772.44
R2002-8
Oven
6942388.64
762239.62
R2002-9
Oven
6942405.36
762231.16
R2002-10
Oven
6944033.31
761177.57
R2002-11
Oven
6944201.57
761149.69
R2002-12
Oven
6944203.04
761197.14
R2002-13
Oven
6944416.30
761145.78
R2002-14
Oven
6944398.69
761180.51
R2002-15
Oven
6944370.32
761299.86
R2002-16
Sectioned terrace
6944429.01
761594.31
R2002-17
Oven
6944478.21
762623.15
R2002-18
Sectioned terrace
6944467.49
762626.94
R2002-19
Oven
6944275.28
762616.22
R2002-20
Oven
6943802.31
762648.68
R2002-21
Domestic terrace
6944360.04
761604.74
R2002-22
Oven
6944565.47
761832.53
R2002-23
Rockshelter
6944563.61
761829.85
R2002-24
Garden soil
6944549.62
761751.07
R2002-25
Garden soil
6943792.39
762655.15
R2002-26
Rockshelter
6944255.20
764831.75
R2002-27
Rockshelter
6944308.95
764914.98
R2002-28
Rockshelter
6944394.56
765075.74
R2002-29
Rockshelter
6943591.20
758682.53
R2002-30
Rockshelter
6942754.43
764938.23
Continued on next page
terra australis 37
288 Appendices
ID
Description
Northing
Easting
R2002-31
Rockshelter
6940906.57
765096.11
R2002-32
Rockshelter
6940902.14
765124.29
R2002-33
Rockshelter
6941880.04
760167.66
R2002-34
Rockshelter
6947459.35
763650.47
R2002-35
Rockshelter
6947407.65
763668.50
R2002-36
Rockshelter
6946949.59
763755.07
R2002-37
Rockart
6946947.42
763754.02
R2002-38
Rockshelter
6946870.82
764780.06
R2002-39 (Pukumia)
Secondary Fortification
6944769.45
764576.54
R2002-40 (Taua)
Secondary Fortification
6945846.76
764947.42
R2002-41
Lithic Scatter
6943973.67
761677.78
R2002-42 (Pukutaketake)
Fortification
6942940.46
760071.27
R2002-43 (Noogurope)
Fortification
6944219.02
759897.79
R2002-44
Rockshelter
6943923.78
763326.62
R2002-45
Domestic Terrace
6944702.88
761837.02
R2002-46
Rockshelter
6945162.28
765532.68
R2002-47 (Taugatu Cave)
Rockshelter
6944950.65
765466.75
R2002-48
Agricultural Terrace
6943577.40
761368.96
R2002-49
Domestic Terrace
6944068.24
759446.83
R2002-50 (Ngapiri)
Secondary Fortification
6941622.07
762001.34
terra australis 37