auckland university press kohika the archaeology of a late maori lake village sep 2004

List of tables Table 1.1 Frequency of sites by type in the Rangitaiki Plains and surrounding area Table 3.1 Radiocarbon dates, Kohika pollen site Square D17 Table 5.1 Chronometric data u

Ko Tupai, Ko Tutarakauika, Ko Te Rangihiiria, Ko Tuara nga taniwha

Ko Matataketake me Tiki nga mauri kohatu

Ko Te Awa o Te Atua te kotore

Ko Otamaroroa te papa whenua

Ko Tamarau te kaitiaki wairua

Ko Te Kaokaoroa te akua

Ko Waimea, Ko Waitepuru, Ko Awatarariki, Ko Awaitipaku, Ko Awaiti,

Ko Omehue, Ko Awakaponga, Ko Te Waikamihi, Ko Mangaone nga awa

Ko Te Otaramuturangi, Ko Tiepataua, Ko Te Awakaponga, Ko Te Ahikokoai,

Ko Awatarerehika, Ko Te Umuhika nga urupa

Ngati Awa pepeha

Whakapaukorero is the mountainTarawera is the riverNgati Awa is the tribeMataatua is the canoe

Te Kohika is the paTupai, Tutarakauika, Te Rangihiiria, and Tuara are the guardians

Matataketake and Tiki are the talismans

Te Awa o Te Atua is the estuaryOtamaroroa is the localityTamarau is the spirit guide

Te Kaokaoroa is the coastlineWaimea, Waitepuru, Awatarariki, Awaitipaku, Awaiti, Omehue, Awakaponga,

Te Waikamihi and Mangaone are the streams

Te Otaramuturangi, Tiepataua, Te Awakaponga, Te Ahikokoai, Awatarerehika

and Te Umuhika are the cemeteries

Ngati Awa proverbial saying

Kohika The archaeology of a late Maori lake village in the Ngati Awa rohe, Bay of Plenty, New Zealand

Edited by Geoffrey Irwin

Auckland University Press

First published 2004 Auckland University Press University of Auckland Private Bag 92019 Auckland, New Zealand www.auckland.ac.nz/aup

© the authors 2004 ISBN 1 86940 315 0 This book is Memoir 9 of the Whakatane and District Historical Society,

which has provided assistance with its publication.

National Library of New Zealand Cataloguing-in-Publication Data Kohika : the archaeology of a late Ma¯ori lake village in the Nga¯ti Awa rohe, Bay of Plenty, New Zealand / edited by Geoffrey Irwin.

Includes bibliographical references and index.

be reproduced by any process without the prior permission of the publisher.

Designed and typeset by Amy Tansell Printed by Printlink Ltd, Wellington

G.J Irwin and M.D Jones

R.T Wallace and G.J Irwin

R.T Wallace, G.J Irwin and R Neich

G.J Irwin, R.K Nichol, M.A Taylor, T.H Worthy and I.W.G Smith

13 Evidence for diet, parasites, pollen, phytoliths, diatoms and starch grains in

G.J Irwin, M Horrocks, L.J Williams, H.J Hall, M.S McGlone and S.L Nichol

G.J Irwin

R.T Wallace and G.J Irwin

List of tables

Table 1.1 Frequency of sites by type in the Rangitaiki Plains and surrounding area

Table 3.1 Radiocarbon dates, Kohika pollen site (Square D17)

Table 5.1 Chronometric data used in the current analysis

Table 5.2 Summary posterior distributions for Phase 2 start, Phase 2 end and Phase 2 durationTable 6.1 Characteristics of the bird spear fragments

Table 6.2 Comb dimensions (mm)

Table 7.1 Estimated dimensions of excavated houses and pataka

Table 8.1 Single spiral-wrapped bundles of harakeke (1SWB)

Table 8.2 Two-ply spiral-wrapped bundles of harakeke (2PSW)

Table 8.3 Three-ply cordage

Table 8.4 Three-ply braid variants

Table 8.5 Plaited twill with narrow strips

Table 8.6 Plaited check with broad strips

Table 8.7 Netting and component parts

Table 8.8 Other pieces: fragments whose technique is unclear

Table 8.9 Contents of boxes (KOH number and technique)

Table 9.1 Artefacts from Kohika

Table 10.1 Size, roundness and sphericity of Kohika obsidian pebbles

Table 10.2 XRF analyses of obsidian samples from Kohika and Maketu

Table 11.1 Complete flake mean dimensions (and standard deviation) by exterior scar direction

for all areas in the university excavation, Mayor Island obsidian

Table 11.2 Maximum dimension of cores by scar pattern for all areas in the university

excavation, Mayor Island obsidian

Table 11.3 Maximum dimension for proximal, medial and distal fragments by exterior scar

pattern, Mayor Island obsidian

Table 11.4 Complete flake mean dimensions (and standard deviation) from the Historical

Society assemblage by exterior scar pattern, Mayor Island obsidian

Table 11.5 Mean dimensions for complete platform rejuvenation flakes from all areas in the

university excavation, Mayor Island obsidian

Table 11.6 Mean dimensions for complete flakes of length greater than 23 mm, by exterior

scar direction from all areas in the university excavation, Mayor Island obsidianTable 11.7 Mean dimensions (and standard deviations) of complete tools compared with

complete flakes for all areas in the university excavations, Mayor Island obsidianTable 11.8 Mean dimensions (and standard deviation) for pieces with macroscopic edge

modification by type for all areas excavated by the university, Mayor Island obsidianTable 11.9 Mean dimensions (and standard deviation) for complete tools from the Historical

Society assemblage by edge modification type, Mayor Island obsidian

Table 11.10 Frequency of edge-modified pieces, flakes and cores by area, Mayor Island obsidianTable 11.11 Mean length (and standard deviation) of complete flakes with length > 23 mm by

exterior scar morphology and area, Mayor Island obsidian

Table 11.12 Frequency of complete and fragmented flakes of length > 23 mm by area, Mayor

Island obsidian

Table 11.13 Number and weight of flakes and flake fragments of maximum dimension less

than 10 mm by area, Mayor Island obsidian

Table 11.14 Complete flakes with length > 23 mm from the university excavations compared

with those excavated by the Whakatane Historical Society, Mayor Island obsidianTable 11.15 Frequency of complete and fragmented tools by type of edge modification and

Table 12.1 Minimum number of individual mammals

Table 12.2 Human bone by excavation area

Table 12.4 Taphonomic variables for identified dog bone by excavated area

Table 12.5 Estimated ages of dogs at death

Table 12.6 Avian taxa represented among identifiable elements in the Kohika assemblage with

data from all squares and layers amalgamated

Table 12.7 Frequencies of fish species, by area

Table 12.8 Fish species frequencies by layer, Area D

Table 12.9 Bone class frequencies for jack mackerel, Area D

Table 12.10 Shellfish from Kohika

Table 12.11 Shell samples from the White House, Area D

Table 12.12 Shell samples from the Yellow House, Area D

Table 12.13 Shell samples from the Bright Yellow floor, Area D

Table 13.1 Samples included in coprolite analyses

Table 13.2 Physical attributes of coprolites analysed

Table 13.3 Components of coprolites, weights and percentages

Table 13.4 Incidence of fish body parts in Kohika coprolites

Table 13.5 Seeds from Kohika coprolites

Table 13.6 Percentages of inorganic material

Table 13.7 Variation in coprolite no.19

List of figures

Figure 1.1 The former river courses of the Rangitaiki Plains and communication routes

recorded in early maps (Gibbons 1990, Hunia 1977)

Figure 1.2 Archaeological sites recorded in the area of the Rangitaiki Plains

Figure 2.1 The geomorphology of the Rangitaiki Plains

Figure 2.2 Former shorelines and river courses on the Rangitaiki Plains

Figure 2.3 Soils of the Rangitaiki Plains in the vicinity of Kohika (after Pullar 1985)Figure 3.1 The Bay of Plenty lowlands with pollen sites underlined

Figure 3.2 Pollen site stratigraphy: Kohika pollen site (excavation Square D17), Tunapahore

archaeological site complex and Thornton-Atkinson archaeological site

Figure 3.3a Kohika, percentage pollen diagram

Figure 3.3b Kohika, percentage pollen diagram

Figure 3.3c Kohika, percentage pollen diagram

Figure 3.3d Kohika, percentage pollen diagram

Figure 3.4a Thornton-Atkinson complex, percentage pollen diagram

Figure 3.4b Thornton-Atkinson complex, percentage pollen diagram

Figure 3.5 Tunapahore A, percentage pollen diagram

Figure 3.6a Tunapahore B, percentage pollen diagram

Figure 3.6b Tunapahore B, percentage pollen diagram

Figure 4.1 A contour map of Kohika showing the location of the excavations

Figure 4.2 Some representative section drawings from Area A

Figure 4.3 A plan of the excavated features in Area A

Figure 4.4 Square B1, north section

Figure 4.5 Square B1, south section

Figure 4.6 Square B1, plan of features at the base of the excavation

Figure 4.7 Square B3, east section

Figure 4.8 Square B4, west section

Figure 4.9 Square C1, south section

Figure 4.10 Square C1, west section

Figure 4.11 Square C7, south section

Figure 4.12 Square C10, north section

Figure 4.13 Area D, the layout of excavation units

Figure 4.14 Squares D1 and D2, all sections

Figure 4.15 Area D, the White House horizon

Figure 4.16 Area D, the Yellow House horizon (stakeholes less than 10 cm deep are not shown)Figure 4.17 Area D, the Bright Yellow horizon

Figure 4.18 Area D, Squares D12–15, trench

Figure 4.19 Historical Society investigations (with some of their notes)

Figure 5.1 Summary of the calibrated distributions for the chronometric data given in Table 5.1Figure 5.2 Posterior distribution for Phase 2, start

Figure 5.3 Phase 2, end

Figure 5.4 Phase 2, duration

Figure 6.1 Bird spear point made from tree-fern trunk

Figure 6.2 Twelve digging sticks and one ko footrest

Figure 6.3 A weeder blade, a one-piece spade, and four handles of composite digging toolsFigure 6.4 Seven complete detachable digging-tool blades, one rough-out and one fragmentFigure 6.5 Part of a carving on a ceremonial ko

Figure 6.6 Seven broken shafts with terminal knobs, possibly handles from composite toolsFigure 6.7 Fifteen beaters and beater fragments

Figure 6.8 Four bowls and bowl fragments

Figure 6.9 A steering paddle rough-out

Figure 6.10 Canoe paddle and paddle fragments

Figure 6.11 Canoe hull pieces

Figure 6.12 Three canoe seats, one bulkhead and eight other fittings

Figure 6.13 Canoe bailers

Figure 6.14 Six heru or hair combs

Figure 6.15 Six darts or javelins

Figure 6.16 Seven potaka or spinning tops

Figure 6.17 An adze handle rough-out

Figure 6.18 A chisel handle and a chisel socket

Figure 6.19 A section of a putorino (flute)

Figure 6.20 Two net gauges

Figure 6.21 Two thread reels

Figure 6.22 Fibre-, net- and rope-working tools

Figure 6.23 Ladder

Figure 6.24 Wood-splitting wedges

Figure 6.25 Pegs

Figure 6.26 Items of unidentified function

Figure 7.1 A reconstruction of the carved house from the Historical Society (HS) AreaFigure 7.2 A reconstruction of the pole and thatch house from the Yellow House floor, Area DFigure 7.3 A reconstruction of the pataka from Area D

Figure 7.4 Types of lashing holes on house planks

Figure 7.5 Poupou and other vertical house elements from the HS Area

Figure 7.6 Door or window parts from the HS Area

Figure 7.7 Tumatahuki battens from the HS Area

Figure 7.8 Possible fragments of pataka from the HS Area

Figure 7.9 Dressed slabs split from pukatea tree trunks, HS Area and Area B

Figure 7.10 Timbers recovered from Area D

Figure 7.11 The three rafters from Area D

Figure 7.12 Detail of rafter tenon joints

Figure 7.13 Parameters used to estimate the width of a building in Area D, based on rafter

Figure 9.1 Bone hei tiki pendant, tooth pendant, bone toggle

Figure 9.2 Pounamu adze

Figure 10.1 Dimensions of obsidian pebbles from Kohika, Maketu and Otamarakau

Figure 10.2 Rb-Sr plots for analysed obsidian artefacts from Kohika (solid symbols) and source

samples from Maketu and Taupo

Figure 10.3 Zr-Rb plots for analysed obsidian artefacts from Kohika (solid symbols) and source

samples from Maketu and Taupo

Figure 10.4 Relative proportions of ‘grey pebble-type’ and ‘other grey’ obsidian from KohikaFigure 11.1 Terms used to describe flake fragments Proximal flakes include a platform, distal

flakes have a termination, and medial flakes lack a platform or a terminationFigure 11.2 Flakes with different exterior scar patterns The identification numbers are given

in brackets: a (1703), b (1637), uni-directional; c (1907), d (1735), bi-directional;

e (1893), sub-radial; f (1688), radial

Figure 11.3 Quadrants for assessing scar orientation The flake is orientated with the platform

at quadrant 1 (the figure is based on artefact 161)

Figure 11.4 Typology for edge modification: a (2617), b (2163), bifacial; c (2128), heavy;

d (1145), e (1687), f (2175), g (1850), light; h (2490), i and j (1822), k (1635),notch

Figure 11.5 Dimensions of a complete flake

Figure 11.6 Core shapes: a (1525), pebble; b (1747), c (2248), multiple platform; d (2875),

Figure 12.3 Size frequency distributions of pipi and tuatua, Area D

Figure 13.1 Components of Kohika coprolite samples, percentages by weight

Figure 13.2 Percentage pollen diagram for Kohika coprolite samples

Figure 13.3 Percentage phytolith diagram for Kohika coprolite samples

Figure 13.4 Percentage diatom diagram for Kohika coprolite samples

Figure 14.1 A schematic view northwards over Area D across the lake to the dunes and the

sea In the left foreground is a reconstruction of Area D during the Yellow Househorizon The houses, canoes, nets and racks on the right represent the artefactsand building timbers found in the Historical Society Area The palisade followsthe topography around the lake The lakeshore vegetation of raupo, flax andcabbage trees with patches of kahikatea and kanuka scrub is based on the pollenrecord The roofed pit and two small covered bins in the bottom right were actuallyfound in Area A

List of plates

Plate 1.1 In 1975, Kohika was an inconspicuous, low-lying grassed mound in an area of

agricultural swamp drainage

Plate 1.2 A whakanoa ceremony was conducted by Jack Fox, Romana Kingi, Mike Mason,

Harry Reneti and Albert Te Rere, of the Ratana, Anglican, Catholic, Ringatu andPresbyterian churches respectively The kaumatua placed the site and the artefactsinto the interim care of the University of Auckland

Plate 1.3 After the discovery of artefacts, the first investigations were undertaken by

mem-bers of the Whakatane and District Historical Society In this 1976 photographare (from left) Dave White, Ken Moore and the late Anton van der Wouden Thespoil in the background was removed from the drain by a digging machine.Plate 1.4 The University of Auckland excavations of Area D during the season of January,

1976

Plate 2.1 The new mouth of the Tarawera River The former course of the Rangitaiki

River is at the left of the entrance and the former Te Awa o te Atua estuary, whichcarried the combined waters of the two rivers, lay to the right of the entrance andflowed west to Matata The current road bridge is on the Kaharoa shoreline, and

a short distance upstream is the junction of the Tarawera and the Awaiti Stream.Canal 109 runs inland from the Awaiti past the remains of Lake Kohika

Plate 2.2 Looking seawards, a canal and stopbank now separate Lake Kohika from the

archaeological lake village, which formerly lay on its southwestern shore Thesite itself is located on a remnant of sand-dune that dates from the coastline of

2000 years ago The archaeological excavation can be seen at the end of thefarmer’s cattle-race

Plate 4.1 Kohika in January 1976 Work is in progress in Areas A, B and D The spoil

heaps are of different colour, which results from their varied composition.Plate 4.2 Excavations in Square A1 Extension

Plate 4.3 Part of a small bin surrounded by surface stakeholes in Square A3

Plate 4.4 A cross-section of a bin structure in Square A3 dug into the former sand-dune.Plate 4.5 An oval-ended pit in Square A1 Ext interrupted by a later rectangular pit lying at

right angles to it A charcoal sample for C14 dating was taken from underneaththe large pumice boulder found in the pit fill

Plate 4.6 Square B1 during excavation

Plate 4.7 Square B1 near the base of the excavation showing in situ posts

Plate 4.8 The drain section in Area B shows a flood deposit of reworked tephra alluvium

outside the site, where a later meander channel, visible in Square B3, has cut intoits surface

Plate 4.9 Wooden items from the peat below the flood deposit in Square B3 are triple-bagged

in plastic

Plate 4.10 A degraded pukatea board in the upper peat of Square B4

Plate 4.11 The pukatea board in Square B4 consolidated prior to removal

Plate 4.12 The defended edge of the site was sharply defined in Square B4 Palisade posts

were exposed in the side of the agricultural drain visible in the background andSquare B3 lies on the other side of it

Plate 4.13 The edge of the site reveals the effect of the flood in Square B4 Fine silt lies around

the palisade posts, reworked Kaharoa alluvium lies outside the site, and the twoare separated by the wave-lapped shore

Plate 4.14 Square B4, west section, showing a palisade post, pre-flood deposits that built up

during occupation, the flood alluvium of sand and silt, and post-flood sedimentsthat are culturally sterile

Plate 4.15 Squares C1 and C12 reveal a deposit of alluvium Above this, a late meander

channel in the Upper Peat can be seen in the baulk, while below the alluvium isthe culture-bearing Lower Peat This photograph shows the pump being primed

in 1975 at the start of a day’s work

Plate 4.16 Square C10 contained a complex succession of fire-pits and hangi with occupational

debris and fills composed of material quarried elsewhere

Plate 4.17 Excavating a swamp in wet weather can have its difficulties Area D

Plate 4.18 The site perimeter in Area D Inside the line of posts were artificial house floors;

outside many associated waterlogged artefacts were found preserved in peat.Plate 4.19 Square D2, with superimposed house floors visible in the south section

Plate 4.20 Square D2 A line of standing posts, with horizontal light wooden battens flexed

and pegged between them, marks the edge of an artificial floor of silt, packedwith water-rolled greywacke pebbles

Plate 4.21 Square D1, south section Also showing a test excavation of the lacustrine silt, the

Kaharoa Tephra and sedge peat below

Plate 4.22 Square D5, the White House floor

Plate 4.23 Square D4, firescoops in the White House horizon

Plate 4.24 Area D, exposing the extent of the Yellow House horizon, January 1976

Plate 4.26 Area D, Square DD during excavation of the Yellow House horizon Intruding

into the square is the corner of Square D2, dug previously, while the baulks arecomposed of backfilled spoil

Plate 4.27 Square DD, laid bracken-fern stems below an artificial house floor

Plate 4.28 Square D2, canoe bow in peat

Plate 4.29 Square D2, gourd shell

Plate 4.30 Square D2, adzed log and length of rope

Plate 4.31 Square D7, whale vertebra, wooden spear and coil of vine

Plate 4.32 Square D13, north section

Plate 4.33 Square D14, log at base of excavation

Plate 6.1 Thirteen sections of bird spear and one spear point

Plate 6.2 Two coils of rata vine

Plate 7.1 KOH14 Poupou base, Historical Society Area

Plate 7.2 KOH16 Poupou base, Historical Society Area

Plate 7.3 KOH17 Poupou base, Historical Society Area

Plate 7.4 KOH18 Poupou base, Historical Society Area

Plate 7.5 KOH44 Poutahuhu base, Historical Society Area

Plate 7.6 KOH53 Pare fragment, Historical Society Area

Plate 7.7 KOH1 Part of carving, Historical Society Area

Plate 7.8 KOH3 Fragment of carving, Historical Society Area

Plate 7.9 KOH345 Fragment of spiral from carving, Historical Society Area

Plate 7.10 KOH7 Poutokomanawa figure, Historical Society Area, and modern replica carved

by Paki Harrison, Dante Bonica and Wiremu Puke

Plate 7.11 KOH2 Fragment of elaborate carving, Area D

Plate 7.12 KOH6 Fragment of elaborate carving, Area D

Plate 7.13 KOH4 Fragment of elaborate carving, Area D

Plate 7.14 KOH174 Carved handle of bailer, Area D

Plate 8.1a KOH298.5 Fragment of single spiral-wrapped bunches of harakeke (1SWB)

resembling a handle

Plate 8.1b KOH300.3 Fragment of 1SWB resembling one half of a pair of two-ply

spiral-wrapped bunches (2PSW) of harakeke

Plate 8.1c KOH298.1 Short fragment of two-ply spiral-wrapped bundles of harakeke (2PSW).Plate 8.2 KOH303.8–11 Short braided fragments, showing two straight examples

(KOH303.8 and 9) and an X and a Y braid (KOH303.10 and 11)

Plate 8.3 KOH304.1 Fine twill close-up, showing condition of fibres

Plate 8.4 KOH305.1 A larger piece of fine twill, showing the curvature in the plaiting.Plate 8.5 KOH303.4 Broad checked plaiting, showing one folded-back strip (at lower part

of image)

Plate 8.6a KOH297 A Grommet with netting still attached

Plate 8.6b KOH297 Small fragment of mesh

Plate 9.1a Greenstone kuru pendant

Plate 9.1b Greenstone chisel pendant

Plate 9.2 Bone tiki pendant (human)

Plate 9.3a–c One-piece bone fishhooks (human)

Plate 9.4a Bone fishhook blank (human)

Plate 9.4b Fishhook point (dog tooth)

Plate 9.5a Bone needle (bird)

Plate 9.5b Bone needle (dog)

Plate 9.5c Bone awl (bird)

Plate 9.6a Bone awl (seal)

Plate 9.6b Bone chisel (dog)

Plate 9.7 Pounamu adze

Plate 9.8a Pounamu chisel

Plate 9.8b Pounamu adze flake

Plate 9.9a Sandstone file

Plate 9.9b Pumice pigment bowl

Plate 12.1 A sawn section of human cranium

Plate 12.2 Knife-cut marks on a human femur

Plate 12.3 Dog cranium with crushed parietal

Plate 12.4 Dog cranium with cut marks on nasal bone

Plate 12.5 Dog mandible with ventral margin removed

Plate 12.6 Dog-tooth marks on snapper bones and barracouta jaw

Plate 13.1 Examples of coprolites from Kohika

Plate 13.2 Fishbone extracted from coprolite no.24

Plate 13.3 Charcoal extracted from coprolite no.19

Plate 13.4 Egg of Toxocara canis.

Plate 13.5 Egg of Toxocara canis, showing characteristic pitting of the shell.

Plate 13.6 Egg of Capillaria hepatica.

For over 30 years this project has accumulated a huge debt to many individuals, some

of whom have now passed on, and to many institutions The whakanoa ceremony atthe site was conducted by Jack Fox, Romana Kingi, Mike Mason, Harry Reneti andAlbert Te Rere Professor Hirini Mead, Pouroto Ngaropo and Ngahuia Rawson of TeRunanga o Ngati Awa provided guidance for the return of the Kohika artefacts andfor their future

Members of the Whakatane and District Historical Society who first worked atKohika and then gave generous support to the University of Auckland team includedTiena Jordan, Jack Moller, Ken Moore, Dave White, Errol Westgate and Anton vander Wouden, who was also director of the Whakatane District Museum and Gallery

A later director, Warner Haldane, helped arrange the return of the artefacts toWhakatane

At Kohika, the farmer, Phil Jessop, and the sharemilkers, Les and Graeme Brownlee,could not have been more obliging Neighbouring farmers, especially Tony Pansier,helped to house the excavators and made gifts of food The Rangitaiki Plains DairyCompany provided us with showers and evening meals during one winter season.Institutional support and funding was provided by the Department of Anthropol-ogy, University of Auckland, the University of Auckland Research Committee, theLotteries Board of the Department of Internal Affairs, and the New Zealand HistoricPlaces Trust

All of the scholars who contributed to the analysis of archaeological material arenamed in the list of contents of the book Technical staff at the university made amagnificent contribution over the years: they include Karel Peters, Rod Wallace andDilys Johns for conservation; Joan Lawrence, Caroline Phillips and Seline McNameefor illustrations; and Tim Mackrell and Hamish MacDonald for photography A num-ber of anthropology students carried out preliminary study for research essays andtheses, as described in the book

Those friends, colleagues and students who took part in the excavations at Kohikainclude Harry Allen, Nola Arthur, Barry Baquie, Gary Barnett, Mark Bellingham, SimonBest, Steve Black, Joanna Boileau, Dorothy Brown, Ken Burnett, Linda Burnett, RichardCassels, Helen Charters, John Coster, Mark de Courcy, Marlene Deans, Sandra Dreifus,Clare Fawsett, Anne Geelen, Roger Green, Don Hanson, Terry Hunt, Joan Hunter, JillIrwin, Gabrielle Johnston, Garry Law, Ian Lawlor, Jennifer Leighton, Steve Mangan,Bernadine Naus, Tia Negerevich, Mary Newman, Reg Nichol, Peter Pearce, Karel Peters,Caroline Phillips, Michelle Phillips, Pamela Raspe, Pamela Russell, Peter Russell, BillShaw, David Stowe, Jan Stowe, Julie Stretton, Doug Sutton, Makiuti Tongia, AlanWalmsley, Graeme Ward, Tony Walton and Lynnette Williams Apologies to anyonewho has been overlooked

The quality of this book has been greatly improved by Janet Davidson, who carefully

read and commented on earlier drafts, and Andrew Mason who edited the text I amgrateful to Elizabeth Caffin and the staff of Auckland University Press including AnnieIrving, Katrina Duncan and Amy Tansell Diane Lowther compiled the index

The project would not have been possible without the support of the Irwin family.Jill made sure the excavation was fed, Sarah and Kate kept watch on the spoil heaps

to see that nothing of value was carelessly thrown away, and Tom took his first steps

at Kohika

Geoffrey Irwin

and archaeological context

G.J Irwin, R.G Law, I Lawlor and P Ngaropo

Archaeological sites in wetlands are unusually rich because they can preserve organicmaterials that rarely survive elsewhere These include all kinds of wooden artefacts,fabrics, food remains, and microfossils that indicate former environments In theirpublished Rhind Lectures for 1995, J and B Coles wrote (1996:133): ‘ the wet-lands of the world have continued to yield new and often surprising information aboutthe past wetlands have continued to produce archaeological evidence unobtainablefrom any other environment save the most extreme.’ However, such sites are increasinglyrare as wetlands are drained

For Polynesia, Kirch and Green (2001:199–200) estimate that about 20 per cent ofmaterial objects were archaeologically durable and the remaining 80 per cent perish-able, except in unusual circumstances This is based on ethnographic inventories ofmaterial culture and lexical reconstructions of the names for artefacts These cir-cumstances apply generally to New Zealand However, while wet sites offer hugeopportunities for research, considerable analytical sophistication and resources areneeded to investigate and preserve them It is in such a context that this 30-year arch-aeological wetland study reaches publication

At some time around AD 1700, a Maori lake village called Kohika in the Bay ofPlenty was abandoned after a flood and fortuitously preserved in peat swamp, togetherwith its contents It was rediscovered during agricultural drainage in 1974 The sitewas excavated during the late 1970s and produced a rare and comprehensive inventory

of waterlogged remains that are in close association with one another and, as such,represent a technological, economic and cultural entity The site is of fairly shortduration and provides an archaeological snapshot of Maori material culture and theway of life that had developed in the North Island prior to the arrival and influence ofEuropeans Although there have been previous excavations of wetland sites in NewZealand, none has produced as much rich and varied material as Kohika and had thebenefit of such a range of modern specialist analysis

Kohika is located in the west of the Rangitaiki Plains in what was formerly a greatswamp It is just a few hundred metres east of the Tarawera River and two kilometresinland from the sea It takes its name from the adjoining Lake Kohika, which lies inthe fork of the Tarawera River and Awaiti Stream (Fig 1.1) A little further downstreamthe Rangitaiki River formerly joined the Tarawera to form Te Awa o Te Atua, a riverestuary that ran three kilometres further westwards behind the coastal dunes to flowinto the sea near the present settlement of Matata The site had good access to theresources of swamp, floodplain and coast It was strategically located for coastalcommunication by canoe, and also by river and track to the interior of the NorthIsland; there is evidence for travel and trade in both directions

Kohika was a palisaded village on a small island in the swamp beside the lake,

2 Kohika

which was larger then than now Preserved in the swamp are organic remains that givedetailed information about former environments and diet The site has archaeologicalevidence for a broad range of domestic and social activities There were houses ofvaried construction, including the remains of what is currently the oldest-known carvedhouse in New Zealand There were raised pataka storehouses that have long beenelusive in prehistory, plus cooking shelters and semi-subterranean storage pits and bins

In addition to the remarkable inventory of wooden artefacts, there are fibre plaiting,cordage and netting, a large assemblage of flaked obsidian and some miscellaneousartefacts of other materials The remains reveal many aspects of life, including housing,canoe transport, food production, craft activities, defence and outside communica-tion There is evidence for music, play, personal status, art and religion

This book contains contributions from 20 scholars, and the various chapters dealwith geomorphology, vegetation history, excavation, chronology, wooden artefacts,houses and pataka, fibre work, artefacts of bone, pumice and pounamu, obsidiansourcing, obsidian flake technology, faunal remains and coprolite analysis, followed

by a general review of the evidence The rest of this chapter includes a brief account of

traditional and historical records of Kohika, and a review of archaeological sitedistribution in this part of the Bay of Plenty It concludes with a brief history of thearchaeological project since 1975, and negotiations that vested ownership of theartefacts with Ngati Awa.

Kohika in history and tradition

According to Ngati Awa oral tradition, a man called Waitahaarikikore, who was cended from Toi and his wife Te Kuraimonoa, lived on Rarotonga He built a canoeusing the maihi (barge-boards) from his meeting house and called the waka Te Paepae

des-ki Rarotonga, which he sailed via the Kermadecs to the Bay of Plenty He lived for atime near the mouth of the Tarawera River at a place now called Te Otaramuturangi.Before moving on, he buried the canoe there and, looking southwards across theswamplands, said: ‘Te ko hika tera’ ‘Over there is the place I shall light my fires’ Bylighting his fires, Waitahaarikikore claimed mana whenua over the place now calledKohika (Anon 2000:2)

More recently, the wider Rangitaiki floodplain was the land of Ngati Awa and tions of Te Arawa and Tuwharetoa While the area is generally recognised as beingNgati Awa, the other tribes have ‘historical and land affiliations to the area ’ (Hunia

sec-1977:x) The location of Kohika, near Matata, is now in the rohe of Ngati Awa but

its traditional history is complex (Best 1925:690, Grace 1959:90–1, Gudgeon 1970:8,Lawlor 1979:17–22) The tribal history has fluid relationships and the missionary T.S.Grace refers in his diaries to connections between Maori he met in the RangitaikiSwamp and Lake Taupo (Grace 1928:132, Grace MS 1850–73, Feb 1867) The threerivers crossing the plain, the Rangitaiki, Tarawera and Whakatane, plus foot trails,provided access to Rotorua and the central North Island

Europeans came late to the Rangitaiki Plains and included traders, missionariesand soldiers (Lawlor 1979:22–8, see also London 1960) One description of condi-tions and communications in the swamp is by Cowan (1923:96):

This Rangitaiki Swamp was then accessible only by the tracks along the ward sandhills, or by canoe along the Tarawera River, the Awaiti-Paku, and theOrini River (connecting the Awa-a-te-Atua with Whakatane Harbour) and by thelabyrinth of reed-fringed waterways, navigable in small canoes, winding amongthe islets that rose above the water a few feet and made camping-grounds for eel-fishers and wild fowl hunters

sea-Cowan speaks of palisaded ‘island-like forts in the great swamp’ during the militaryoperations of the 1860s (Cowan 1923:96) Two historic pa close to Kohika were Oheuand Te Matapihi, on the west bank of the Tarawera, the latter being where the Rev.T.S Grace built a storehouse and landing-stage to service his inland mission to Taupo.The village of Matata was sketched in 1865 by H.G Robley (MS 1858–87:14, MS1898–1922) It was then on the east bank of the Tarawera at its junction with theRangitaiki River (Fig 1.1) but shifted to its present location after the land confiscations.Kohika was a prehistoric forerunner to these musket pa, not simply as a fort but as asite of substantial settlement

We have found only two specific references to Kohika One is in a note at the end

of Cowan’s chapter on military operations at Matata, where he quotes Mr F Burt,who for many years farmed the Matapihi Block at Matata: ‘The natives tell me whenthey were spearing eels near the Kohika Lake they came across the remains of an old

several years, and many settlers abandoned the area (Gibbons 1990:vii, 10–11) The

first drainage board was formed in 1894, the Rangitaiki was diverted to its new mouth

in 1914 and the Tarawera in 1924 (Pullar 1985:8) Once drained, the Rangitaiki Plainsbecame highly suited to dairy farming

Archaeological site distribution in the area of the Rangitaiki Plains

The Bay of Plenty enjoys a warm, moist climate and fertile soils ideal for horticulture.Together with its fisheries and forests, it offered ideal conditions for Maori settlement

The following account provides an archaeological context for discussing the location

of Kohika

Recording of Maori archaeological sites in the area began in the 1960s and wascarried on for the next two decades principally by Kawerau resident Ken Moore.Surveying in the Whakatane and Waimana river areas was done by the New ZealandHistoric Places Trust in the early 1980s Other contributions have come from forestrysurveys and research and heritage management work at Ohiwa, Moutohora and

Kawerau (Hayward et al 1987, Jones 1983, 1984, 1986, 1991, Lawlor 1983a, 1983b,

Moore 1973, 1974, 1980a, 1980b, Phillips 1996)

Figure 1.2 plots the recorded Maori sites in the area in the New Zealand logical Association Site Recording Scheme, as at February 2002 For this plot the siteswere extracted in a hierarchical manner from the Computerised Index of New Zea-land Archaeological Sites (CINZAS), using the site description field The site type inthis field is not necessarily the same as on the site record form but represents a moreconsistent approach to setting a site type than is often found on the form The number

Archaeo-of sites in Figure 1.2 is shown in Table 1.1 using the same system for classification.The site distribution in Figure 1.2 shows the broad patterns of Maori settlement

Pa are concentrated on higher ground along the coast, around Ohiwa Harbour and

on the ridges at the periphery of the Rangitaiki Plains and Whakatane River Kohika(V15/80) is the only site recorded as a swamp pa in the area, and the other pa shown

on the plains close by belong to the historic period

Terraces are the most common site type and are concentrated in the vicinity ofKawerau and Ohiwa, although this could be partly the result of surveys being morethorough there than elsewhere Pit sites may be underrepresented in the southwesternpart of Figure 1.2, which was blanketed by tephra from the 1886 Tarawera eruptionthat filled many pits on terrace sites (Lawlor 1983a:220)

Rua, while known from excavations, are rare as field recorded sites, unlike further

west in the Bay of Plenty However, in the Other category there are sites cautiously

described as ‘depressions’ which may be collapsed rua This low frequency may result

Table 1.1 Frequency of sites by type in the Rangitaiki Plains and surrounding area

• Pa are sites with that site description.

• Terrace sites are those with terraces in their description that are not pa They may also be sites with middens, pits, etc.

• Pit sites are those with pits in their description, but not pa or terraces They may also be sites with middens, etc.

• Rua are sites with cave pits in their description, but not the above.

• Middens do not include any of the categories above.

• Other sites are those coded as Maori An unusual site is a large grinding stone at Matata (Fulton 1921).

In general, the Rangitaiki Plains have a low site density compared with ing areas However, underrecording may be an issue, and the Tarawera Tephra andoutwash will have buried some sites Much of the seaward part of the plains wasswamp in the early 19th century and had clearly been so for centuries, as indicated bythe growth of peat Only raised areas in the swamp would have been attractive forMaori settlement, and Jones (1991) recorded cultivation sites on the beach ridges leftstranded by the prograding coastline, as well as on the better-drained river fans in thesouthern part of the plains A number of lowland sites have produced wooden artefacts(Mead 1984:200).

surround-In general, the site distribution shows the preference of Maori to live along thecoast, especially near harbours Away from the coast, sites are concentrated on soilsattractive to Maori horticulturalists and areas enjoying both defensible uplands andaccess to rivers Kohika is unusual in the context of the larger sample of sites, being aprehistoric lowland swamp pa in an area with a very low density of recorded sites

A brief history of the investigations since 1975

The site was discovered in November 1974 during drainage operations near LakeKohika at the swampy northern end of the farm of Mr P Jessop of Sutherlands Rd Adigging machine exposed a palisade of kanuka posts in the side of a drain on theeastern side of the site, and various wooden and other artefacts were thrown out withthe spoil During the summer of 1974–75, members of the Whakatane and DistrictHistorical Society investigated quite a large area of swamp on the northern side of thesite by probing They dug up many valuable wooden artefacts which were placed inwater for safekeeping

The New Zealand Historic Places Trust became involved, and the site was visited inearly 1975 by Auckland archaeologists It was agreed that G Irwin, who had previouslyworked at swamp pa in the Waikato excavated by staff of the University of Auckland(Bellwood 1978, Shawcross 1968), would direct an excavation as a University of Auck-land project with a grant from the Golden Kiwi Lottery Board (Irwin 1975, Moore1975) At this time there was little understanding of the site’s complexity, size orsignificance

A series of major excavations took place in May 1975, January 1976, December

1977 and April 1978, and further fieldwork to resolve particular issues was done in

1979 and 1981 (see Plates 1.1–4) A large-capacity pumping system was developedtogether with on-site conservation The fieldwork gradually came to terms with thecomplex geomorphology and site stratigraphy and, increasingly, the research designbecame interdisciplinary

Large quantities of archaeological material and other samples were taken to land (in a railway wagon after the 1976 season), but there was no conservationinfrastructure waiting to receive them Waterlogged artefacts were put into temporarytanks while a conservation laboratory was built and equipped in the new Human

Auck-Plate 1.2 A whakanoa ceremony was conducted by Jack Fox, Romana Kingi, Mike Mason, Harry Reneti and Albert Te Rere, of the Ratana, Anglican, Catholic, Ringatu and Presbyterian churches respectively The kaumatua placed the site and the artefacts into the interim care of the University of Auckland.

Plate 1.3 After the discovery of artefacts, the first investigations were undertaken by members of the Whakatane and District Historical Society.

In this 1976 photograph are (from left) Dave White, Ken Moore and the late Anton van der Wouden The spoil in the background was removed from the drain by a digging machine.

Plate 1.4 The University of Auckland excavations of Area D during the season of January 1976.

Sciences Building at the University of Auckland Conservation took many years Theanalysis (and re-analysis) of archaeological materials followed and many researcherswere involved, as described in the following chapters Several specialist studies werecarried out in the late 1990s, and this full report was written from 2000.

G Irwin returned the Kohika artefacts to the Bay of Plenty in May 1998,accompanied by P Ngaropo, representing a Ngati Awa hapu closely connected withKohika The taonga were first acknowledged at Te Umuhika Marae, Matata, beforebeing formally welcomed in Whakatane by representatives of the several iwi of theregion, the Bay of Plenty Regional Council, the Whakatane and District HistoricalSociety and the Whakatane and District Museum and Gallery The few remainingartefacts were returned to Whakatane in the care of P Ngaropo in 1999 The finalartefact, a pounamu adze, is to be presented to Ngati Awa together with a copy ofthis book

The Kohika Collection presented some complex ownership issues Initially thelandowner, P Jessop, gave to the Whakatane and District Historical Society, the arte-facts excavated by its members When the Department of Anthropology at the University

of Auckland took over direction of the project in 1975, the whole collection was placed

in its interim care at a whakanoa ceremony by kaumatua representing local iwi andreligious denominations After its return, ownership of the collection was negotiatedbetween Ngati Awa and the Whakatane and District Historical Society, who had beenresponsible for recovering the early part of the collection The issues were resolved bythe appointment of a board of trustees to manage the collection, comprising a majority

of Ngati Awa representatives as well as a representative each from Ngati Tuwharetoa

ki Kawerau and the Whakatane and District Historical Society, in recognition of theirassociations with the collection The trustees have placed the Kohika Collection onloan to the Whakatane District Museum and Gallery

In June and July 2000, an exhibition entitled Kohika: a glimpse of life in a wetland

pa near Matata, was held at Whakatane museum.

References

Anon., 2000 Te Kohika: a glimpse of life in a wetland pa near Matata during the 1600s AD.

Pamphlet prepared for the exhibition at Whakatane and District Museum and Gallery, TeWhare Taonga o te Rohe o Whakatane, 1 June–16 July, 2000

Bellwood, P.S., 1978 Archaeological research at Lake Mangakaware, Waikato, 1968–1970.

New Zealand Archaeological Association Monograph No.9

Best, E., 1925 Tuhoe, the children of the mist New Plymouth: The Polynesian Society Coles, J and B Coles, 1996 Enlarging the past: the contribution of wetland archaeology.

Society of Antiquaries of Scotland Monograph Series No.11 Exeter: Short Run Press

Cowan, J., 1923 The New Zealand Wars: a history of the Maori campaigns and the ing period Vol II Wellington: R.E Owen, Government Printer.

pioneer-Fulton, R., 1921 An account of a supposed Maori sharpening stone Transactions of the New Zealand Institute, 53:471–2.

Gibbons, W.H., 1990 The Rangitaiki, 1890–1990: settlement and drainage on the Rangitaiki.

Whakatane: Whakatane and District Historical Society

Grace, J Te H., 1959 Tuwharetoa: the history of the Maori people in the Taupo district.

Wellington: Reed

Grace, T.S., 1928 A pioneer missionary among the Maoris 1850–1879: being letters and journals

of Thomas Samuel Grace Edited by S.J Brittan, G.F., C.W and A.V Grace Palmerston

North: Bennett

Grace, T.S., MS Papers 191, Mission work Taupo and Tauranga, 1850–73 Alexander TurnbullLibrary

10 Kohika

Gudgeon, W.E., 1970 Te heke o Rangihouhiri Whakatane: Whakatane and District Historical

Society

Hayward, B.W., P.R Moore and P Bain, 1987 Prehistoric archaeological sites on Whale Island

(Motuhora), Bay of Plenty Tane, 32:73–86.

Hunia, L.Te A., 1977 Tangi Putauaki: a Maori history of the Rangitaiki Unpublished MAthesis, University of Auckland

Irwin, G.J., 1975 The Kohika site, Bay of Plenty Historical Review, 23:101–4.

Jones, K.L., 1983 Pa in two western segments of the Waiotahi and Whakatane Valleys, Bay

of Plenty New Zealand Archaeological Association Newsletter, 26:165–73.

Jones, K.L., 1984 Archaeological investigations in Waiotahi Valley, Bay of Plenty, November

1981 New Zealand Archaeological Association Newsletter, 27:109–18.

Jones, K.L., 1986 Polynesian settlement and horticulture in two river catchments of the

east-ern North Island, New Zealand New Zealand Journal of Archaeology, 8:5–31.

Jones, K.L., 1991 Maori settlement and horticulture on Rangitaiki Plains, Bay of Plenty,

New Zealand New Zealand Journal of Archaeology, 13:143–75.

Kirch, P.V and R.C Green, 2000 Hawaiki, Ancestral Polynesia: an essay in historical anthropology Cambridge: Cambridge University Press.

Lawlor, I., 1979 Palaeoenvironment analysis: an appraisal of the prehistoric environment ofthe Kohika swamp pa (N68/140), Bay of Plenty Unpublished MA thesis, University ofAuckland

Lawlor, I., 1983a Rua Kumara o Kawerau In S Bulmer, G Law and D Sutton (eds), A lot of spadework to be done: essays in honour of Lady Aileen Fox New Zealand Archaeological

Association Monograph No.14, pp.212–48

Lawlor, I., 1983b Maruka investigations, Kawerau, Bay of Plenty: Final Report for Stage IV.Department of Anthropology, University of Auckland

London, H.D., 1960 A field day at Matata Historical Review, 8:109–14.

Mead, S.M., 1984 Te Maori: Maori art from New Zealand collections Auckland: Heinemann Moore, K.W., 1973 Archaeology at Whakatane, N.Z Part 1 Historical Review, 21:113–22 Moore, K.W., 1974 Archaeology at Whakatane, N.Z Part 2 Historical Review, 22:50–63 Moore, K.W., 1975 Kohika site, N68/140, Bay of Plenty: a preliminary report Historical Review,

23:60–1

Moore, K.W., 1980a Te tangata whenua, the early people of Te Teko Historical Review,

28:63–7

Moore, K.W., 1980b Place names of the Te Teko district Historical Review, 28:68–71.

Phillips, K., 1996 The archaeology of the eastern Bay of Plenty Unpublished MA thesis,University of Auckland

Pullar, W.A., 1985 Soils and land use of the Rangitaiki Plains, North Island, New Zealand.

Lower Hutt: New Zealand Soil Bureau

Robley, H.G., MS Notebook, 1858–87 Alexander Turnbull Library

Robley, H.G., MS Letters Robley to Mair, sketches, etc., 1898–1922 Alexander TurnbullLibrary

Shawcross, F.W., 1965 Report on archaeological investigations at Thornton, Whakatane,

Bay of Plenty Historical Review, 13:186–92.

Shawcross, F.W., 1968 The Ngaroto site New Zealand Archaeological Association Newsletter,

11:2–29

of the Rangitaiki Plains

G.J Irwin

The Rangitaiki Plains are geologically active and the inhabitants of Kohika lived in

a landscape of frequent earthquake, volcanic eruption and flood

Whakatane Graben

The Rangitaiki Plains are a lowland some 340 square kilometres in area They stretch

22 kilometres along the Bay of Plenty coast between Matata and Whakatane and extendinland for a similar distance The plains are part of the Whakatane Graben, which islocated where the Taupo Volcanic Zone reaches the Bay of Plenty coast and intersectswith the north–south-trending North Island Shear Belt (Nairn and Beanland 1989).Within the graben, which is bounded by north–south fault lines (Fig 2.1), the plainhas subsided and the surrounding hills have risen The rate of subsidence is estimated

to be 2–3 mm per year over the last 5000 years As they have dropped, the plains havefilled with sediments from the three large river catchments of the Rangitaiki, Taraweraand Whakatane rivers, and also from the Taupo and Okataina volcanic centres to thesouth To put this in strict geological terms, Mesozoic basement rocks have down-faulted and the resulting basin infilled by Quaternary volcanics and sediments.Uplifting and tilting of the graben margins has accompanied subsidence of its floor(Nairn and Beanland 1989)

Faulting and earthquakes

Geological evidence of recent surface faulting has been largely obscured by the veryyoung sediments, mainly of volcanic origin, that form the modern ground surface.The 1987 Edgecumbe earthquake occurred along two known pre-existing faults thathad previously moved at some time after the AD 1350 Kaharoa eruption (Nairn andBeanland 1989) The 1987 earthquake also moved at three new surface traces thatwere presumed to overlie faults that had been concealed by sediments from both theTaupo eruption of c.AD 150 and the Kaharoa eruption It appears that faulting hasoccurred independently at Matata and Edgecumbe, that is on the western and easternsides of the graben respectively (Fig 2.1), which suggests that the Rangitaiki Plainsmay experience moderate to severe shaking more often than if the two graben mar-

gins were to move at the same time (Ota et al 1988) The study of a trench excavation

across a fault at Matata revealed that the most recent earthquake occurred there afterthe Kaharoa Tephra and possibly during the last 250 years, according to radiocarbon

dating (Ota et al 1988) In fact, it transpired that this earthquake occurred during

12 Kohika

the human occupation of Kohika, where its secondary faulting offsets could be moreclosely dated

Volcanic eruptions and coastal progradation

About 7000 years ago the shoreline probably lay along the cliffs at Whakatane, keri, Te Teko and Onepu (Pullar 1985) Since then it has prograded about 10 kilometres

Awa-as the lowland hAwa-as formed From time to time the Bay of Plenty hAwa-as been showered byvolcanic tephra, and the plains have been formed largely from outwash of tephra bythe Rangitaiki, Tarawera and Whakatane rivers (Pullar and Selby 1971) Thus, on theplains, tephra occurs both in airfall bands and as reworked alluvial sediments A series

of identified tephra has fallen on the plains over the last 5000 years or more and, asyounger beds overlap the older, Pullar and Selby (1971:419) have ingeniously foundthat ‘a succession of possible shorelines of decreasing age could be plotted at succes-sive points where each older ash disappears’ They regard the position on the groundwhere an airfall ash bed cuts out as a possible shoreline, whereas the position wheresea-rafted pumice boulders occur is a probable shoreline Coastal progradation evi-dently occurred in fits and starts and was rapid after each volcanic eruption

Figure 2.1

The

geomorph-ology of the

Rangitaiki Plains

Volcanic ashes

The main tephra are the Whakatane Ash of c.5500 years BP, the Taupo Pumice ofc.1850 years BP (Froggatt and Lowe 1990, Nairn and Beanland 1989), the Kaharoa

Tephra of cal 600 BP (Lowe et al 1998), and the Tarawera Tephra of AD 1886 The

three younger of these ashes have been identified at Kohika and the following scriptions of them as they occur on the plains are based on the extensive observations

de-of Alan Pullar (Pullar and Selby 1971:423) The Taupo Ash is 10–13 cm thick Thegrade is coarse ash and small vesicular lapilli In peat swamps the colour is pale olive(5Y 6/3) becoming brownish-yellow where the swamp is drained (10YR 6/6) Ondunes the bed is brown in colour (10YR 4/3) The Kaharoa Tephra is usually 10–15

cm thick In peat swamps the bed is white ash (5Y 8/1) and is finely shower-bedded

On dune ridges the ash bed is masked by black topsoil and is difficult to identify Indune swales the upper 5 cm is light yellowish-brown fine ash (2.5Y 6/4), and the lower

10 cm is darker (10YR 4/4) The Tarawera Tephra is often 2–7 cm thick but may be

up to 12 cm in dune swales In peat swamps the Tarawera is very dark greyish-brown(10YR 3/2); on dune ridges it is dark greyish-brown (2.5Y 4/2), becoming darker inthe swales The sea-rafted Taupo Pumice of c.AD 150 also occurs at Kohika and is ofunique appearance for Holocene times, ‘a light yellowish-brown colour [with] coarseirregular gas cavities It is easily broken and can be crushed in the hand’ (McFadgen1994:196)

Landforms of the Rangitaiki Plains

These comprise coastal and inland dunes, back-swamp lowlands and peat swamps,natural levee systems of rivers and streams, and floodplains of largely mixed pumi-ceous alluvium with minor greywacke alluvium

Dunes

The dunes are linear, and stranded ones lie generally parallel to the coast (Fig 2.2).The system has formed over some 7000 years as the coast has prograded Much ofthe dune sand came from the ash beds that still mantle the uplands drained by thethree rivers, while the dune surfaces have been created by long-shore drift from west

to east (Pullar and Selby 1971) The dune system is well developed only at theWhakatane end of the plains and near Matata in the west Near Kawerau, dunes reach

a height of about 30 m above sea level but elsewhere, due to subsidence, have beenfound buried under peat some 3 m below sea level The various dunes are dated bythe particular tephra that lies on them Those of the Taupo shoreline sometimes occurnear sea level or they may be exposed up to 3–4.5 m above sea level, except at Matatawhere they are higher Kohika was built on a remnant of the pre-Taupo shoreline thatdates to approximately 2000 BP Kaharoa and Tarawera shorelines lie closer to thesea than Kohika, parallel to the existing coast and higher in elevation The destruction

of vegetation following human settlement evidently caused the coastal dunes to beeroded and wind-blown during the 500 years bracketed between the Kaharoa andand Tarawera tephra (Pullar and Selby 1971)

Rivers, plains and swamp

Three major rivers cross the plains Several considerable streams join them as tributariesand these have formed fans where they emerge from the surrounding hills Over time,the rivers have changed their courses and these can be traced from palaeo-channelsacross the floodplains The rivers have also cut the stranded marine dunes that now

14 Kohika

sometimes survive as discontinuous strips parallel to the coast and were sometimesislands in prehistory The general pattern of rivers and streams that existed from thetime of the Kaharoa eruption c.AD 1350 until swamp drainage began early in the20th century is shown in Figure 2.2 The Rangitaiki had no mouth of its own and amajor distributary, the Orini Stream, flowed east to join the Whakatane River at itsmouth, while the Rangitaiki itself ran west behind the coastal sand dunes The AwaitiStream was another main distributary of the Rangitaiki; it flowed west towards Ma-tata and joined the Awaiti Paku, a distributary stream of the Tarawera, on the way.The combined waters of the Awaiti, Awaiti Paku, Rangitaiki and Tarawera werecalled Te Awa o te Atua, which entered the sea at what is now the western end ofMatata (Gibbons 1990) This common river mouth and estuary was influenced by atidal regime, but was altered in modern times when direct cuts were made to the sea(Plates 2.1 and 2.2)

The Rangitaiki and Tarawera rivers flow through floodplains built up by their owndeposits of silt and tephra In the past they frequently overflowed their banks andinundated surrounding land Two main deposits shown in Figure 2.1 are of TaupoPumice alluvium, closely associated with the Rangitaiki River and deposited after AD

150, and Kaharoa alluvium deposited after AD 1350 These are extensive areas overthe surface of the plains (Nairn and Beanland 1989) Further areas of Tarawera Ashalluvium have been deposited since AD 1886 In addition, over time there have beenfrequent flood episodes that have locally redeposited pumiceous alluvium around theswamp

In the back-swamp lowlands, peat horizons formed on Taupo Pumice over much ofthe plains and again, in places, on ash from the Kaharoa and Tarawera eruptions,resulting in complex interbedding The youngest deposits of the Rangitaiki Plains

Atua estuary, which carried the combined waters of the two rivers, lay to the right of the entrance and flowed west to Matata The current road bridge

is on the Kaharoa shoreline, and a short distance upstream is the junction of the Tarawera and the Awaiti Stream Canal 109 runs inland from the Awaiti past the remains of Lake Kohika.

Plate 2.2 Looking seawards, a canal and stopbank now separate Lake Kohika from the archaeological lake village, which formerly lay on its southwestern shore The site itself is located on a remnant of sand-dune that dates from the coastline of 2000 years ago The archaeological excavation can be seen at the end of the farmer’s cattle-race.

of the swamp in the early part of last century to create the rich agricultural landscape

of today

Soils and stratigraphy around the Kohika site

Dunes, back-swamps and floodplain deposits interfinger near the coast where riversand streams converge and shallow freshwater lakes have formed Figure 2.3 describesthe soils in the vicinity of Kohika as taxonomic and physiographic units, as mappedand described in detail by A Pullar on the basis of a large number of cores and examinedsections (Pullar 1985) Several of these units were encountered during archaeologicalexcavations at Kohika, and it is necessary to take account of the others to follow thegeomorphological history of the area immediately surrounding the site

Soils of the dunes

Pki (Pikowai) sand occurs as recent wind-blown sands on the current foredunes with

a very thin cover of Tarawera ash

Koe (Kopeopeo) soils are further inland on older stranded dunes and have Taupo and

Kaharoa tephra (both rhyolitic) on wind-blown sand, with a thin cover of Tarawera(basaltic) tephra The Kohika site has a natural core of Taupo-age dune

Soils of the former tidal flats

Muw (Muriwai) silt loams are weakly saline soils derived from mixed pumiceous and

greywacke alluvium on former tidal flats now cut off from river estuaries

Soils of the present floodplains

Ran (Rangitaiki) soils consist of rapidly accumulating pumiceous alluvium deposited

in former meander troughs of streams The area of Ran shown in Figure 2.3 was

deposited during floods in 1964

Ou (Opouriao) and Ori (Orini) fine sandy loams are from slowly accumulating

pumi-ceous alluvium and found on the levees of rivers and their distributaries

Soils of the former floodplains

Ats (Awaiti) sandy loam is a recent soil with a thin cover of Tarawera Tephra on

pumiceous alluvium derived from Kaharoa Tephra

Ome (Omehue) sandy loam and Omp (Omehue) sandy loam on peat are poorly drained

gley soils from fine pumiceous alluvium derived from Kaharoa Tephra with a thincover of Tarawera Tephra

Onc (Omehue) coarse sandy loam is another poorly drained gley soil with a thin cover

of Tarawera Tephra over layered pumiceous Kaharoa alluvium (reworked tephra), on

a sub-surface of sand and gravels and with occasional thin layers of diatomaceousearth

Soils of the back-swamp

lowlands

Ag (Awakaponga) silt loam

and Agp (Awakaponga) silt

loam on peat are recent soils

from silty pumiceous

rhyo-litic alluvium

Pr (Paroa) silt loam and Prp

(Paroa) silt loam on peat

are poorly drained gley soils

from very fine pumiceous

alluvium

Soils of the peaty swamps

Awi (Awakeri) sandy loam

on peat is a poorly drained

gley soil with a clearly

layered profile of Tarawera

and Kaharoa tephra, peat,

silts, pumiceous alluvium

underlain by (dune) sand

derived from Taupo

Pum-ice There are sharp

discon-tinuities

Mtk (Matuku) silt loam is

a poorly drained gley soil

consisting of layered

mater-ials, including

diatom-aceous earth which has

formed in freshwater lakes,

peat and pumice alluvium,

with a very thin cover of

Tarawera Tephra

Aro (Awaroa) soils are

organic and formed from

peat with a thin cover of

basaltic ash

Soil profiles in the Rangitaiki Plain show that there was rapid infilling immediately

after the Kaharoa eruption and there are no buried soil horizons from this time

How-ever, intermittent infilling since about 400 years BP is indicated by buried palaeosols

The peat that has grown near Kohika is mainly sedge formed from Baumea spp There

are patches of diatomaceous earth, which can form during the open-water stage of

bog development Diatoms increase markedly following rhyolitic volcanic eruptions

such as the Kaharoa (Pullar 1985:36), and have been identified at Kohika Shallow

freshwater lakes were a feature of the floodplain around the site

Summary of the geological situation of Kohika

• The site is on a low island remnant of the 2000 BP shoreline dune that rises above

Figure 2.3

Soils of the Rangitaiki Plains

in the vicinity of Kohika (after Pullar 1985)

18 Kohika

lake level Archaeological correlations of stratigraphy between different areas ofexcavation around the edge of the mound are supported by the natural swampstratigraphy, which includes three distinctive tephra beds separated by silts, peatand diatomaceous earth, and in the upper deposits by palaeosols

• A band of cultural material occurs in the peat between the Kaharoa Tephra of cal

AD 1350 and the Tarawera Tephra of AD 1886 A greater depth of peat lay belowthe archaeological site than above it, indicating that occupation was in the laterpart of the deposit This will be supported by radiocarbon evidence in Chapter 5.Moreover, the peat above the cultural deposit was sterile, indicating the absence ofoccupation for a period prior to the Tarawera Tephra Further soils developed inthe swamp above the Tarawera

• Neither the Taupo nor the Kaharoa tephra remains intact on the mound itself,although each forms components of the soil there This is also true of the TaraweraAsh which lies like a continuous tidemark around the mound but which, every-where above this, has been mixed into the soil Evidently the dune at Kohika wasgardened both before the lake village was occupied and after it was abandoned

• Evidence for substantial occupation ends with the arrival of bands of pumiceousalluvium and silt, derived mainly from reworked Kaharoa Tephra, which represent

a local flood around parts of the site exposed to floodwaters from the TaraweraRiver Parts of the shallow lake edge suddenly filled and it was no longer easy toreach the island by canoe

• Core samples around Lake Kohika, both by A Pullar and the University of land, show that the lake was more extensive than today (Lawlor 1979, Figs 3.1 and3.2) However, this has to be seen as the result of modern drainage as well asprehistoric infilling

Auck-• Finally, there is evidence that an earthquake occurred while people were living at

Kohika and that this event relates to the Matata Fault nearby (Ota et al 1988).

This account has described mainly the natural stratigraphy The complex interplay

of natural and cultural stratigraphy will be unravelled in Chapter 4

Climate

The following information is compiled from Aldridge (1985), Jones (1991) and theNew Zealand Meteorological Service (n.d.) The Rangitaiki Plains are today noted fortheir sunny climate The duration of sunshine hours is among the highest in the NorthIsland and the mean daily maximum temperature in January is 25.5 degrees Celsius atKawerau and 25.3 degrees at Te Teko Plant growing days above 10 and 15 degreesCelsius show the area to be warmer than much of the North Island The mean annualrainfall is 1304 mm on the coast at Whakatane The number of days with rainfallequal to or greater than 1.0 mm averages ten per month from May to October andeight per month from November to April Heavy rainfall is associated with the pas-sage of cyclonic systems from the north and northeast in summer Other rainfall followsthe usual pattern of frontal passages Frost-free days (screen frosts) average 328 peryear While frosts are expected every year, between 1948 and 1975 the earliest frostday recorded at Whakatane was 5 May and the latest 10 September The average yearlyincidence of ground frost is 24 at Te Teko on the plains away from the RangitaikiRiver levee, 24 at Edgecumbe on the Rangitaiki levee, and 15 at Whakatane (Jones1991) Whatever differences may exist between the present and the time when peoplewere living at Kohika, the climate was certainly favourable

A brief historical observation about vegetation

According to Pullar (1985:6), at the time of European settlement swampland west

of the Rangitaiki River was densely covered with raupo and rushes Cabbage treesflourished on natural levees of rivers and streams with kahikatea, titoki, toetoe andflax on the back-swamp lowlands On the coastal dunes there was manuka, brackenfern and mingimingi while small teatree and cabbage trees grew on the inland dunes.Clearly, this landscape was transformed by Maori and influenced by volcanic erup-tion Buried stumps show that totara grew in the Omeheu locality (inland from Kohika)before the Kaharoa eruption and soil profiles on inland dunes suggest that podocarpforest flourished before the Taupo eruption Some waterlogged tree-trunks survive

in the dune at Kohika A detailed account of vegetation history follows in the nextchapter

and age New Zealand Journal of Geology and Geophysics, 33:89–109.

Gibbons, W.H., 1990 The Rangitaiki, 1890–1990: settlement and drainage on the Rangitaiki.

Whakatane: Whakatane and District Historical Society

Jones, K.L., 1991 Maori settlement and horticulture on the Rangitaiki Plains, Bay of Plenty,

New Zealand New Zealand Journal of Archaeology, 13:143–75.

Lawlor, I., 1979 Palaeoenvironment analysis: an appraisal of the prehistoric environment

of the Kohika swamp pa (N68/140), Bay of Plenty Unpublished MA thesis, University ofAuckland

Lowe, D.J., B.G McFadgen, T.F.G Higham, A.G Hogg, P.C Froggatt and I.A Nairn, 1998.Radiocarbon age of the Kaharoa Tephra, a key marker for late-Holocene stratigraphy and

archaeology in New Zealand The Holocene, 8:487–95.

McFadgen, B.G., 1994 Coastal stratigraphic evidence for human settlement In D.G Sutton

(ed.), The origins of the first New Zealanders Auckland: Auckland University Press,

pp.195–207

Nairn, I.A and S Beanland, 1989 Geological setting of the 1987 Edgecumbe earthquake,

New Zealand New Zealand Journal of Geology and Geophysics, 32:1–13.

New Zealand Meteorological Service, n.d., Summaries of climatological observations to 1980.

New Zealand Meteorological Service Miscellaneous Publication No.177

Ota, Y., S Beanland, K.R Berryman and I.A Nairn, 1988 The Matata Fault: active faulting at

the north-western margin of the Whakatane Graben, eastern Bay of Plenty New Zealand Geological Survey Record, 35:6–13.

Pullar, W.A., 1985 Soils and land use of the Rangitaiki Plains, North Island, New Zealand.

Lower Hutt: New Zealand Soil Bureau

Pullar, W.A and M.J Selby, 1971 Coastal progradation of Rangitaiki Plains, New Zealand

New Zealand Journal of Science, 14:419–34.

3 The impact of Polynesian settlement on the vegetation of the coastal Bay of Plenty

M.S McGlone and K.L Jones

The coastline of the Bay of Plenty forms a great bight, opening north and northeasttowards the central Pacific, creating a natural landfall for voyagers from the tropics(Fig 3.1) It enjoys a warm, mild, moist climate and has light, fertile soils, ideal forhorticulture Migrants from northern latitudes in prehistoric times are likely to haveestablished some of their first settlements in this region, as traditions about canoelanding-places at Whangaparaoa imply The Bay of Plenty coast is therefore central toour understanding of the settlement process This chapter looks in detail at the vege-tation history of the Kohika swamp, and some other sites in the surrounding region,

in order to reconstruct the original vegetation that would have confronted those firstsettlers, to date the beginning of human influences, and to chart the profound effects

of Maori settlement on the local environments

Intense human use has destroyed or disrupted nearly all of the original vegetationcover in the lowland coastal Bay of Plenty and similar areas throughout the country.Only where rugged hill country comes down to the coast does substantial undisturbedvegetation remain, as in the steep coastal country of the eastern Bay of Plenty (Nicholls

1971) While remnant stands of coastal vegetation give some picture of previousvegetation cover, they are often themselves modified (Jones and Moore 1985) andunlikely to represent the full range of original coastal communities.

Palaeoecological investigations are therefore an essential aspect of archaeologicalstudies in which reconstruction of original vegetation is important Organic sedimentsformed over thousands of years in wetlands preserve abundant pollen and spores which,when extracted and characterised, permit construction of detailed vegetation histor-ies Analysis of the pollen and spore content of peats and muds in wetland profilesfrom throughout New Zealand has given comprehensive insights into the unmodifiedenvironment, the timing of the Polynesian environmental impact, and the subsequentchanges to vegetation resulting from settlement (McGlone 1983a, McGlone and

Wilmshurst 1999, Newnham et al 1998).

The Bay of Plenty lies in a major zone of active volcanism Three tephra layers – theTarawera, Kaharoa and Taupo – are commonly found in the upper sediments of de-posits throughout the region, thus providing an excellent chronology for palynology.These tephra falls have also had a profound impact on the vegetation

In this chapter we present the full pollen analysis of the Kohika pa site (abbreviatedversions have been published previously), new analyses of two other lowland coastalBay of Plenty wetland sites (Thornton and Tunapahore), and discuss the environmen-tal history of the Bay of Plenty coastal zone in relation to Maori settlement

Methods

Most pollen samples were taken directly from cleaned faces of excavation units,although the lower section of the Kohika site was sampled with a Hiller corer Standard

palynological preparation techniques were used (Moore et al 1991): disaggregation

in potassium hydroxide, wet sieving, digestion in 40 per cent hydrofluoric acid,acetolysis, bleach where needed to remove resistant lignin fragments, and mounting inglycerine jelly Pollen results are presented as relative percentages of varying pollensums A terrestrial pollen sum is used for all dryland plants, which excludes tree fernsand all ground ferns (except for bracken, which is treated as a shrub because of itsecological behaviour) Tree ferns are expressed as a percentage of a terrestrial pollensum that includes tree ferns Wetland and aquatic taxa are expressed as a percentage

of total pollen and spores Note that Leptospermum type (Leptospermum scoparium and Kunzea ericoides) is not included as a terrestrial pollen type, as it has poorly dispersed pollen and is likely to have been derived mainly from manuka (Leptospermum

scoparium) growing on peaty wetland soils Nothofagus subgenus Fuscospora includes

all Nothofagus spp with the exception of N menziesii Charcoal counts were made

by a grid point technique (Clark 1982) in which eleven points per field of view arescored for presence or absence of charcoal at the same time as the pollen count Countsare expressed as percentages of the terrestrial pollen sum

Site locations, stratigraphy and chronology

Rangitaiki Plains: Kohika Pa archaeological site (Figs 3.1 and 3.2)

As the Rangitaiki Plains are intensively farmed, no unmodified natural vegetationremains At the time of first European settlement, swampland to the west of theRangitaiki River was densely covered with raupo and rushes (Pullar 1985) Sufficient

flax (Phormium tenax) was present in this area to support a flax mill at Matata Ti

Several small lagoons ringed by raupo (Typha orientalis) and willow (Salix spp.) persist

close to Kohika

The location of the pollen site (Square D17) is shown in Figure 4.13 (below) Thepeat was sampled by digging a hole 1.5 m square to a depth of 2.2 m and then coringwith a Hiller peat corer a further 1.3 m to the underlying sand dune Total sedimentdepth from the base of the Tarawera Ash (datum for all measurements) to the surface

of the buried sand dune is 3.3 m

The pollen site stratigraphy (Fig 3.2) relies for its chronology on the three includedtephra layers and seven radiocarbon dates (Table 3.1) The site is capped by theTarawera Tephra of AD 1886 Two radiocarbon dates in the underlying peat of 353 ±

57 BP and 535 ± 57 BP indicate typical swamp accumulation rates of 1–2 mm peryear The three radiocarbon dates from this site bracketing the Kaharoa Tephra have areversed stratigraphy NZ4804 (656 ± 57 BP) is from organic muds immediately belowthe Kaharoa, and NZ4803 (678 ± 75 BP) on a diatomaceous organic silt immediatelyabove it NZ4802 (729 ± 58 BP) is from a grey, largely inorganic fine silt, 22 cm

above the Kaharoa Tephra The two dates bracketing the Kaharoa Tephra are tically indistinguishable NZ4802 is 63 years older than NZ4804, but still within onestandard error NZ4804 is no more than 10 radiocarbon years different from theestablished age of the Kaharoa Tephra of 665 ± 17 BP, or c.AD 1350 when calibrated,based on 22 screened dates over its entire distribution (Lowe and Hogg 1992) and isalmost certainly correct However, swamps are particularly prone to contamination

statis-by old organic material washed in to the lakes after forest clearance (McGlone andWilmshurst 1999), and it is clear from NZ4801 that there was very rapid accumulation(c.5 mm per year) of silt-rich sediments after the deposition of the Kaharoa Tephra.Therefore, as there is a high probability that NZ4803 and NZ4802 are based onsediments contaminated with old organic material incorporated during deposition,they will be ignored in subsequent discussion On the basis of NZ4805 and NZ4806,organic sedimentation was a moderate 1–2 mm per year- in the years leading up to theKaharoa Tephra eruption, but rapid in the silt-rich sediments following the TaupoTephra (c.5 mm year)

Table 3.1 Radiocarbon dates, Kohika pollen site (Square D17)

Rangitaiki Plains: Thornton-Atkinson archaeological site complex

The Thornton-Atkinson archaeological site (W15/121) is on the inland side of a Taupo sand dune on the Rangitaiki flood plain, some ten kilometres west of Whakatane(Fig 3.1) The crests of these dunes are mantled with ash soils that are easily disturbed,mobilising loose sand that can drift down the dune faces The site consisted of a surfacescatter of shell and blackened soils underlain by pits cut through the ash soils of thecrest of the dune A C14 date for shell in one of the pits has a date of 595 ± 50 yrs BP(NZ7543) with a marine calibration of cal AD 1553–1860 at the 2 sigma limit (Jones1991:153–9) The pre-European settlement vegetation at this site was similar to that

pre-at Kohika

The pollen profile is in a section in a drainage ditch south of the dune crest andclose to the Rangitaiki River cut The sediment profile (Fig 3.2) is undated, but post-dates the Taupo Tephra Tephra within the profile is lensed with sand and has thereforebeen reworked from tephra deposits within the dune sand

Hawai Bay: Tunapahore archaeological site complex

Hawai Bay lies 20 kilometres east of Opotoki (Fig 3.1) The coastal lowlands areonly 800 m wide and consist of partly dissected tephra-covered uplifted marine terracesbacked by steep hills Immediately west of the two pollen profiles is an area of terraces.The area was heavily used by Maori in pre-European times and there are at least ten

pa around the bay

24 Kohika

Two sites were sampled for pollen (Fig 3.2) Tunapahore A is from a drainageditch through a swampy area immediately inland of the beach ridge and 150 m fromhigh-water mark Tunapahore B is at the toe of the slope of a terrace and is close to anow-destroyed swamp pa (X15/105) It is at the extreme inland margin of the samegleyed silt and peat deposit as Tunapahore A, 200 m distant

Tunapahore A contains airfall Kaharoa Tephra but is otherwise undated because ofthe unsuitable weakly organic clay silts that make up most of the section Tunapahore

B has water-laid Taupo Tephra (absent in Tunapahore A) at the base overlying ayellow-brown pumice-rich soil (Whakatane Hill Soil) A date of 1252 ± 87 BP(NZA3485) on a grey pumice-rich silt immediately above the tephra confirms the iden-tification From the 65 cm level in the profile, macroscopic charcoal fragments occur

in a grey silt matrix and probably derive from reworking from the slopes of the nearbyterraces A discrete lens of grey tephra at the 60 cm level surrounded by charcoal-richsilt dating to 781 ± 69 yrs BP (NZA3479) gives a maximum age for the start of charcoalinflux The base of a mottled dark grey silt at 40 cm yielded a date of 1824 ± 142 yrs

BP (NZA3484) that is nearly identical with the standard date of 1850 ± 20 BP for theTaupo Tephra (Froggatt and Lowe 1990) It is apparent that charcoal associatedwith the deposition of Taupo Tephra has been reworked into the profile at a muchlater date

Pollen stratigraphy

The pollen diagrams are zoned on the basis of changes in pollen and spore occurrences,for convenience of description and discussion

Kohika Pa (Figs 3.3a–d)

Zone KO-1: immediately before Taupo Tephra; 1850 yrs BP

This zone is dominated by local swamp shrubs and trees, most notably Leptospermum

cf scoparium, Coprosma (probably small-leaved shrubs such as C propinqua), Myrsine

cf divaricata and Elaeocarpus cf hookerianus (pokaka) Wood fragments within the

sandy peat at this level indicate that shrubs were growing directly on the site Thereare indications of well-drained sandy soils within the developing swamp/dune complex:

Paesia scaberula (hard fern) characterises open dryland sites and is abundant in this

zone; Phyllocladus cf trichomanoides (probably tanekaha) and increasing Metrosideros

(rata type) point to the presence of a coastal forest

Zone KO-2: 1850–1600 yrs BP

Interpretation of this zone is complicated by the eruption of the Taupo Tephra andsubsequent in-wash of pumice from the Rangitaiki River system The Rangitaiki River

headwaters drain the beech-clad (Nothofagus) axial ranges of the east central North

Island, an area partly covered with Taupo ignimbrite deposits and totally blanketedwith thick volcanic tephra fall (Wilson and Walker 1985) Two discrete lenses of silt

at this location above the Taupo Tephra are accompanied by high levels of microscopic

charcoal fragments of Fuscospora (Nothofagus subgenus Fuscospora, including all but

N menziesii of the New Zealand species) and Cyathea smithii-type tree-fern spores.

The first lens of silt has much higher levels of these elements than the second Brackenspores increase during the first silt episode, peak between the silts, and decline abruptly

above the second silt Grass and tutu (Coriaria spp.) percentages rise between the two

silts, both peaking in the second silt, with grass continuing on in significant amounts

to the end of the zone The most likely interpretation is that the Fuscospora charcoal

Figure 3.3a