Changing Patterns of Firewood Use on the

Waim¯analo Plain

∗

Preprint to appear in Hawaiian Archaeology. Not for citation.

Thomas S. Dye Carl E. Sholin

February 9, 2013

Abstract

Wood charcoal identiﬁcations from 35 dated traditional Hawaiian ﬁre-pits on the

Waim¯analo Plain are analyzed for evidence of change over time and diﬀerence across

space. Plant taxa identiﬁed in the ﬁrewood are classiﬁed according to habit, origin,

and elevational distribution. Early in traditional Hawaiian times, ﬁrewood was

commonly brought to the plain from inland forests and ﬁres were made primarily

with native plants. Later, ﬁrewood was more likely to be collected locally, and it

typically included both Polynesian-introduced and native plants. This change in

behavior appears to have taken place in the ﬁfteenth century. It was likely associated

with a vegetational change in which the native lowland forest was replaced with a

variety of useful plants, especially near Puh¯a Stream.

1 Introduction

Wood charcoal is ubiquitous in the archaeological record of traditional Hawai‘i. It occurs

as pieces of various sizes in the general matrix of almost all cultural deposits, where

it is largely responsible for their diagnostic dark color. Most of the charcoal pieces

in the general matrix are too small to be conﬁdently isolated by eye or identiﬁed to

taxon microscopically, but most deposits contain some larger pieces of wood charcoal

in the general matrix that can be isolated and identiﬁed to taxon. This exercise is

not particularly useful, however, because the event that produced a piece of wood

charcoal deposited in the general matrix typically can’t be identiﬁed. More useful for

archaeological analysis is the wood charcoal recovered from features, such as ﬁre-pits, that

can be conﬁdently identiﬁed as archaeological events. Not only are certain behavioral

∗

The archaeological work at Bellows AFS reported here was made possible by federal archaeologists and

resource specialists Fred McGhee, Valerie Curtis, Jeﬀ Pantaleo, Kanalei Shun, Nhut Dao, and Craig

Gorsuch. We thank them for their long-term support of our research. Review comments from Mike

Desilets and Mara Mulrooney helped improve the argument. Any errors of fact or interpretation are

the authors’.

inferences relatively straightforward because the identiﬁed taxa are associated with a

particular event, rather than a long-term process, but the event itself can be located

precisely in space, and pieces identiﬁed as short-lived are suitable for dating with the

C method so the age of the archaeological event can typically be estimated with some

conﬁdence.

The analysis presented here is based on wood charcoal identiﬁcations from 35 ﬁre-pits

excavated over the last 15 years from four sites on the Waim¯analo Plain. The wood

charcoal in the ﬁre-pits was initially identiﬁed so that suitable dating material could be

selected (Dye and Pantaleo 2010). In addition to their use in dating, the wood charcoal

identiﬁcations from the ﬁre-pits are useful as a basis for inferences about changing

patterns of behavior in traditional Hawai‘i. This focus on changing behavior diﬀers

from much of the recent literature on what has come to be called anthracology, which is

primarily concerned with palaeoenvironmental investigation and using wood charcoal

to reconstruct vegetation patterns at diﬀerent times in the past (Scott and Damblon

2010; Allen and Murakami 1999; Huebert et al. 2010; Orliac 2000). The synchronic

palaeoenvironmental goal of much anthracological work has led researchers to develop

methods diﬀerent from those used on the Waim¯analo Plain (Th´ery-Parisot et al. 2010).

Where anthracologists often collect charcoal from the general matrix in the hope of

minimizing the eﬀect of cultural selection (e.g., Orliac 2000), the work reported here uses

only wood from short-term burning events because it emphasizes cultural selection at

a particular point in time. Where anthracologists worry about the combustion process

preserving woods diﬀerentially and distorting their representation in the assemblage (e.g.,

Huebert et al. 2010), the work reported here assumes that the combustion properties of

the diﬀerent kinds of ﬁrewood remained constant over the period of investigation, so that

changes over time in the representation of a ﬁrewood taxon is indicative of behavioral

changes rather than changes in the combustion qualities of wood.

The analysis of the wood charcoal from the ﬁre-pits is carried out within the framework

of a cognitive model for artifact analysis that divides the analytic process into three

stages, named acquisition, structuration, and reconstitution (Djindjian 2001).

2 Acquisition of Materials

In the ﬁrst stage of the analysis, the fundamental qualities inherent in the object of study

are acquired. These fundamental qualities are what they are, regardless of what or how

we think about them. They are the basic data, reported in the ﬁnest possible units of

identiﬁcation, location, and age so that others might ﬁnd them maximally useful. Here,

these fundamental qualities include the analyzed ﬁre-pits, the wood charcoal identiﬁed

from each of the ﬁre-pits, and the conventional

C ages of the ﬁre-pit use events.

The 35 dated ﬁre-pits for which detailed wood charcoal identiﬁcations have been carried

out were discovered over the last decade and a half during archaeological monitoring of

construction excavations, subsurface inventory survey with a backhoe, and controlled

archaeological data recovery excavations (table 1), all carried out in the context of

cultural resources management at Bellows Air Force Station. Fire-pits discovered during

monitoring and inventory survey with a backhoe are typically recovered by excavating

into the face of an open trench. In contrast, ﬁre-pits discovered during data recovery are

typically recognized at the base of a cultural deposit where their dark color contrasts

strongly with the light-colored calcareous sand that represents the basal deposit of sites

on the Waim¯analo Plain. Typically, the entire plan of the ﬁre-pit is exposed before

materials are collected from it. In both cases, where the ﬁre-pit is excavated out of the

side wall of a trench and where it is excavated in plan, only a portion of the ﬁre-pit ﬁll

is recovered. During archaeological monitoring and inventory survey with a backhoe

a portion of the ﬁre-pit is destroyed in the discovery process and in the case of data

recovery excavations the top of the ﬁre-pit is typically indistinguishable from the general

cultural deposit and material is only recovered from the base of the ﬁre-pit that extends

below the general cultural deposit.

The analyzed ﬁre-pits are all located on land used today for Bellows Air Force Station

at the northern end of the Waim¯analo Plain (ﬁg. 1). Nineteen of the ﬁre-pits were

discovered in Sites 50–80–15–4851 and –4853, which ﬂank Puh¯a (Waim¯analo) Stream.

The other 16 ﬁre-pits were discovered at Sites 50–80–15–4856 and –4857 at the very

northern end of the plain about 1.2 km from the stream. All of the analyzed ﬁre-pits

are located inland of the deepest and richest cultural deposits on the plain (see Dye and

Pantaleo 2010:ﬁg. 2). The reason for this is that the deep and rich coastal deposits were

typically reworked so thoroughly by traditional Hawaiian habitation activities that it

is diﬃcult to identify and isolate charcoal associated with particular ﬁre-pit use events

(McElroy et al. 2006). Thus, the 35 analyzed ﬁre-pits might yield an incomplete picture

of traditional Hawaiian settlement on the Waim¯analo Plain, one that focuses on events

removed a short distance inland from the coastal settlement core immediately inland of

the beach.

The wood charcoal from the ﬁre-pits was identiﬁed by Gail Murakami at the Interna-

tional Archaeological Research Institute’s Wood Identiﬁcation Laboratory. Murakami

identiﬁed wood charcoal pieces to taxon by comparing anatomical characteristics pre-

served in the charcoal with a reference collection of Hawaiian woods. Identiﬁed wood

charcoal from each of the ﬁre-pits is listed in the appendix. The sizes of the charcoal

samples from the ﬁre-pits varied considerably. The number of identiﬁed specimens ranges

from 19 at ﬁre-pit 900 2 to 3,049 at ﬁre-pit 900 16, with a median of 121. Weights of

identiﬁed specimens ranges from 0.29 g (gram) at ﬁre-pit 900 21 to 93.85 g at ﬁre-pit

900 16, with a median of 4.18 g. The richness of the samples varied, as well, from the

single taxon identiﬁed at ﬁre-pit 900 1 to the 24 taxa identiﬁed at ﬁre-pit 900 18. Eight

taxa were identiﬁed in the sample with median richness.

The number of specimens identiﬁed for each taxon ranges from 2 to 4,463. The most

commonly identiﬁed taxon, by far, is kukui, A. moluccana, which typically appears in the

ﬁre-pits as charred nutshells but also as wood charcoal. One reason for the large number

of identiﬁed specimens is that the nutshell is distinctive and easily identiﬁed. If it is

present in a sample, then it is certainly identiﬁed and counted. However, the main reason

for the large number of identiﬁed kukui specimens is ﬁre-pit 900 16 (table 1), a feature

that yielded an unusual charcoal collection in which 2,988 kukui nutshell pieces were

identiﬁed. Next in abundance are three native plants probably used as kindling, ‘ilima,

Table 1: Dated ﬁre-pits on the Waim¯analo Plain

Label Site Feature Reference

119 7 50–80–15–4857 Context 7 ﬁre-pit Dye and Dye (2009)

119 16 50–80–15–4857 Context 16 ﬁre-pit Dye and Dye (2009)

273 58 50–80–15–4856 Context 58 ﬁre-pit Sholin et al. (2012)

308 159 50–80–15–4856 Context 89 ﬁre-pit Dye et al. (2012)

308 163 50–80–15–4856 Context 99 ﬁre-pit Dye et al. (2012)

308 175 50–80–15–4856 Context 95 ﬁre-pit Dye et al. (2012)

308 180 50–80–15–4856 Context 93 ﬁre-pit Dye et al. (2012)

900 1 50–80–11–4856 Feature 12 Lebo et al. (2009)

900 2 50–80–11–4856 Feature 17 Lebo et al. (2009)

900 3 50–80–11–4856 Feature 10 Lebo et al. (2009)

900 4 50–80–11–4856 Feature 5 Lebo et al. (2009)

900 5 50–80–11–4856 Feature 9 Lebo et al. (2009)

900 6 50–80–11–4856 Feature 4 Lebo et al. (2009)

900 7 50–80–11–4856 Feature 23 Lebo et al. (2009)

900 8 50–80–11–4856 Feature 22 Lebo et al. (2009)

900 9 50–80–15–4853 Feature 1 Desilets and Dye (2002:111)

900 10 50–80–15–4853 Feature 5 Desilets and Dye (2002:162)

900 11 50–80–15–4853 Feature 9 Desilets and Dye (2002:166)

900 12 50–80–15–4853 Feature 13 Desilets and Dye (2002:133)

900 13 50–80–15–4853 Feature 15 Desilets and Dye (2002:137)

900 14 50–80–15–4853 Feature 16 Desilets and Dye (2002:138)

900 15 50–80–15–4853 Feature 17 Desilets and Dye (2002:140)

900 16 50–80–15–4853 Feature 18 Desilets and Dye (2002:142)

900 17 50–80–15–4853 Feature 19 Desilets and Dye (2002:144)

900 18 50–80–15–4853 Feature 20 Desilets and Dye (2002:122)

900 19 50–80–15–4853 Feature 24 Desilets and Dye (2002:99)

900 20 50–80–15–4853 Feature 25 Desilets and Dye (2002:101)

900 21 50–80–15–4853 Unit BT-5, Feature 6 Addison (1997)

900 22 50–80–15–4853 Unit BT-23, Feature 9 Addison (1997)

900 23 50–80–15–4853 Unit BT-23, Feature 10 Addison (1997)

900 25 50–80–15–4853 Trench 5, Feature 1 Dye (1998)

900 26 50–80–15–4851 Trench 4, Feature 3 Dye (1998)

900 27 50–80–15–4851 Trench 4, Feature 2 Dye (1998)

900 28 50–80–15–4851 Trench 4, Feature 1 Dye (1998)

900 29 50–80–11–4856 Feature 22 Putzi and Dye (2005)

‘akoko, and ‘¯ulei. The identiﬁed ki consists of wood charcoal and charred roots, the latter

of which might be remains of food prepared in the ﬁre-pit rather than a ﬁrewood. The

most common ﬁrewoods are ‘¯ohi‘a ‘ai, a Polynesian introduction, and ﬁve native trees:

hame, hao, lama, hau, and ‘¯ohi‘a lehua.

Table 2: Summary of identiﬁed wood charcoal taxa

Taxon Name Habit Origin Low range

∗

Count

Aleurites moluccana kukui tree

Polynesian

introduction

1 4463

unidentiﬁed ? ? 0 907

Continued on next page

Table 2: Summary of identiﬁed wood charcoal taxa

Taxon Name Habit Origin Low range

∗

Count

Sida fallax ‘ilima shrub native 1 708

Monocotyledonae ? ? nil 612

Chamaesyce sp. ‘akoko shrub-tree native 1 554

Osteomeles anthyllidifolia ‘¯ulei shrub native 2 258

Cordyline fruticosa ki shrub

Polynesian

introduction

5 234

Syzygium malaccense ‘¯ohi‘a ‘ai tree

Polynesian

introduction

200 229

Antidesma pulvinatum hame tree native 30 224

Rauvolﬁa sandwicensis hao tree native 100 204

Diospyros sandwicensis lama tree native 5 197

Chenopodium oahuense ‘¯aheahea shrub-tree native 1 195

Hibiscus tiliaceus hau shrub-tree native 1 176

Metrosideros polymorpha ‘¯ohi‘a lehua tree native 1 100

Abutilon sp. shrub native 1 98

Bobea sp. ‘ahakea tree native 100 85

Colubrina oppositifolia kauila tree native 240 80

Canthium odoratum alahe‘e shrub-tree native 10 73

Bidens sp. ko‘oko‘olau shrub native 1 70

Dodonaea viscosa ‘a‘ali‘i shrub-tree native 3 69

Poaceae grass ? nil 50

Artocarpus altilis ‘ulu tree

Polynesian

introduction

0 48

Hedyotis terminalis manono shrub-tree native 260 46

Saccharum oﬃcinarum k¯o grass

Polynesian

introduction

0 34

Nestegis sandwicensis olopua tree native 30 33

Acacia koa koa tree native 60 29

Nothocestrum latifolium ‘aiea tree native 460 27

latin{Gossypium tomentosum} ma‘o shrub native 1 22

Cocos nucifera niu tree

Polynesian

introduction

1 21

Gossypium tomentosum ma‘o shrub native 1 19

Pittosporum sp. h¯o‘awa tree native 150 15

Charpentiera sp. p¯apala tree native 110 12

Myoporum sandwicense naio tree native 1 12

Palm sp. tree ? nil 12

Nototrichium sp. kulu‘¯ı shrub-tree native 1 11

Psychotria sp. k¯opiko tree native 15 9

Pandanus tectorius hala tree native 1 8

Cheirodendron sp. ‘¯olapa tree native 310 5

Pteridophyta fern ? 0 5

Wikstroemia sp. ‘¯akia shrub-tree native 3 5

Scaevola sericea naupaka shrub native 1 5

Pinus sp. tree alien 0 5

Myrsine sp. k¯olea shrub-tree native 215 3

Ilex anomala k¯awa‘u tree native 50 3

Pteridophyta fern ? 0 2

Lagenaria siceraria vine

Polynesian

introduction

0 2

Continued on next page

Table 2: Summary of identiﬁed wood charcoal taxa

Taxon Name Habit Origin Low range

∗

Count

Senna sp. tree ? 5 2

Hibiscus sp. aloalo shrub native 70 2

∗

Elevation in meters above mean sea level.

The conventional

C ages of the 35 ﬁre-pits are listed in table 3 and illustrated

graphically in ﬁgure 2, which displays a panel for ﬁre-pits at the northern end of the

Waim¯analo Plain near the boundary with Kailua Ahupua‘a and another panel for ﬁre-

pits near Puh¯a Stream. In the past, a failure to correct for the potential eﬀects of

in-built age by dating pieces of unidentiﬁed wood charcoal has yielded dates that are

not archaeologically interpretable (Dye and Pantaleo 2010; Dye 2000). In contrast, the

35 ﬁre-pit use events have been dated relatively precisely using identiﬁed, short-lived

taxa to minimize in-built age. The dates returned by the dating laboratory range from

60 at ﬁre-pit 900 21 to 720

4 at ﬁre-pit 900 22. Three ﬁre-pits returned median

dates: ﬁre-pit 900 14 with a date of 310

60, ﬁre-pit 900 26 with a date of 310

40, and

ﬁre-pit 308 175 with a date of 310 ± 30.

3 Structuration

The second stage of the analysis, called structuration, refers to an abstraction of a quality,

which is then used to classify and/or measure the acquired materials. In the present case,

identiﬁcation of wood charcoal to botanical taxon during the acquisition stage makes it

possible to classify the materials based on qualities of the taxa observed by botanists.

The particular qualities used in this analysis are habit, origin, and elevational distribution.

These qualities and the classiﬁcations based on them are described below, followed by a

series of maps and graphs that show their distribution over space and time.

3.1 Firewood Habit

Botanists use the term habit to describe the general appearance, growth form, or

architecture of a plant. The classiﬁcation used here assigns the identiﬁed taxa to one of

six habits: tree, shrub-tree, shrub, vine, grass, or fern. Trees are perennial woody plants

with a single main stem, or trunk, typically taller than 5–6 m (meter) at maturity. Shrubs

have multiple stems and are shorter than trees, typically under 5–6 m. Shrub-trees are

plants whose form can take on the characteristics of either a shrub or a tree. A vine

is a plant with a climbing stem or runners. Grasses are plants with narrow leaves that

grow from the base. Fern is not technically a habit, but refers instead to plants that have

stems, leaves, and roots, but have neither seeds nor ﬂowers. Ferns in Hawai‘i are typically

low plants the size of shrubs. The wood charcoal identiﬁed in the ﬁre-pits includes 23

taxa classiﬁed as trees, nine as shrub-trees, nine as shrubs, two as grasses, two as ferns

and one as a vine. Two taxa were identiﬁed to a general level not diagnostic of habit.

Figure 1: Locations of sites on the Waim¯analo Plain.

A plot of the proportion of tree wood charcoal against ﬁre-pit age shows a weak decline

over time (ﬁg. 3). This result is similar to the one reported by Dye (2010), which was

based on the weight of identiﬁed charcoal, rather than the number of identiﬁed specimens,

and included information from the 26 dated ﬁre-pits available at the time.

Maps showing the proportion of tree wood charcoal in the ﬁre-pits (ﬁg. 4) show that

tree wood charcoal is more common in ﬁre-pits near the stream than it is in ﬁre-pits at

the northern end of the plain. Most of the ﬁre-pits at the northern end of the plain were

fueled with shrubs and shrub-trees; ﬁre-pits with tree ﬁrewood are in the minority. In

contrast, near Puh¯a Stream most of the ﬁre-pits were fueled with some tree ﬁrewood and

ﬁre-pits that lack tree ﬁrewood are in the minority.

3.2 Locally-Collected Firewood

Botanists have also mapped the distributions of plants on the modern landscape, thereby

establishing elevational ranges for them (Wagner et al. 1990). Given an accurate topo-

Figure 2:

Conventional

C ages of ﬁre-pits based on dates from short-lived taxa. See

table 1 for the ﬁre-pit labels and table 3 for

C dates. Note that the

C age is

represented on a continuous gray scale, only the endpoints of which are shown in the

legend.

Table 3:

C age determinations

Label Laboratory #

C age

119 7 Beta-260904 580 ± 40

119 16 Beta-260905 400 ± 40

273 58 Beta-307650 370 ± 30

308 159 Beta-307651 220 ± 30

308 163 Beta-307654 470 ± 30

308 175 Beta-307653 310 ± 30

308 180 Beta-307652 470 ± 30

900 1 Beta-251242 200 ± 40

900 2 Beta-251246 240 ± 40

900 3 Beta-251244 250 ± 40

900 4 Beta-251245 260 ± 40

900 5 Beta-251243 350 ± 40

900 6 Beta-246786 380 ± 40

900 7 Beta-251248 390 ± 40

900 8 Beta-251247 450 ± 40

900 9 Beta-120317 140 ± 50

900 10 Beta-120318 150 ± 50

900 11 Beta-120319 350 ± 80

900 12 Beta-120320 230 ± 50

900 13 Beta-120321 110 ± 70

900 14 Beta-120322 310 ± 60

900 15 Beta-120323 170 ± 60

900 16 Beta-120324 250 ± 50

900 17 Beta-120325 270 ± 70

900 18 Beta-120326 330 ± 60

900 19 Beta-120327 400 ± 70

900 20 Beta-120328 220 ± 50

900 21 Beta-101869 30 ± 60

900 22 Beta-101871 720 ± 40

900 23 Beta-101872 680 ± 40

900 25 Beta-111022 150 ± 40

900 26 Beta-111023 310 ± 40

900 27 Beta-111024 140 ± 60

900 28 Beta-111025 540 ± 50

900 29 Beta-200230 550 ± 40

graphic model of O‘ahu Island, and assuming that the ﬁrewood identiﬁed in the wood

charcoal was not imported from another island, the elevational data can be used as the

basis for inferences about the ﬁrewood catchment area for each ﬁre-pit. The lower bound

Conventional radiocarbon age

Proportion tree firewood

0.0

0.2

0.4

0.6

0.8

●●

●

●●

●

100 200 300 400 500 600 700

Figure 3: Proportion of tree wood charcoal over time.

of the observed elevational range for each identiﬁed taxon is also shown in table 2. Most

of the plants identiﬁed in the ﬁre-pits can be found today at elevations similar to the

Waim¯analo Plain which ranges from sea level to about 10 m above sea level. Thirty-two

of the 48 identiﬁed taxa have lower elevational bounds in this range. Another ten taxa

grow down to 10–150 m and thus might have been found in the Keolu Hills adjacent to

the Waim¯analo Plain. Six taxa are found today only at elevations greater than 150 m

and these taxa must have come to the sites from more distant locations, the nearest

of which would be the slopes of the Ko‘olau range that mark the mauka boundary of

Waim¯analo Ahupua‘a.

When the proportion of identiﬁed taxa with lower elevational bounds less than 10 m is

plotted against the

C age of the ﬁre-pit a clear pattern emerges in which the proportion

of locally-collected ﬁrewood increases over time (ﬁg. 5). Before 400 bp, ﬁres made solely

with locally-collected ﬁrewood were relatively rare and ﬁrewoods collected from the Keolu

Hills and beyond were commonly brought to the plain. After 400 bp it became common to

fuel ﬁres using locally-collected ﬁrewood and both the frequency and amount of ﬁrewood

imported to the plain appears to have declined.

When the proportion of locally-collected ﬁrewood is plotted on a map of ﬁre-pit

locations (ﬁg. 6) it can be seen that there is no clear geographic pattern. Fire-pits fueled

with a high proportion of locally-collected ﬁrewood are found near Puh¯a Stream and at

the northern end of the plain. Similarly, ﬁre-pits that burned relatively high proportions

of non-local ﬁrewood are found in both places. These include ﬁre-pits 900 9, 900 14,

900 23, and 900 28 near Puh¯a Stream and ﬁre-pits 119 7, 273 58, 308 175, 308 180 and

900 209 at the northern end of the plain.

Perhaps the strongest evidence for a local origin of ﬁrewoods comes from the spatial

Figure 4:

Proportion of identiﬁed wood charcoal from trees. See table 1 for the ﬁre-pit

labels. Note that proportion is represented on a continuous gray scale, only three

values of which are shown in the legend.

Conventional radiocarbon age

Proportion local firewood

0.2

0.4

0.6

0.8

1.0

●

● ●● ●

●

●●● ●

●

100 200 300 400 500 600 700

Figure 5: Proportion of locally-collected ﬁrewood over time.

distribution of hau. The hau tree, which today grows thickly along the banks of Puh¯a

Stream, is found almost exclusively in ﬁre-pits near the stream (ﬁg. 7). Twelve of the 19

ﬁre-pits close to the stream yielded pieces of hau wood charcoal. In contrast, only one of

the 16 ﬁre-pits at the northern end of the Waim¯analo Plain, ﬁre-pit 308 175, yielded hau

wood charcoal. This evidence suggests that the catchment area for ﬁrewoods is typically

small, within a few minutes walk of the ﬁre-pit.

3.3 Polynesian Introductions Used as Firewood

The identiﬁed taxa can also be distinguished by the origin of the plant. The recovered

materials include thirty-four native plants that botanists believe to be endemic or

indigenous to the islands. Seven of the taxa are Polynesian introductions that on present

archaeological evidence were brought to the islands from the Eastern Polynesian homeland

over a period of three or four centuries (Dye 2011). The Polynesian introductions in the

ﬁre-wood charcoal include: i) the candlenut tree, kukui; ii) ti plant, ki; iii) breadfruit

tree, ‘ulu; iv) mountain apple tree, ‘¯ohi‘a ‘ai; v) coconut palm, niu; vi) sugar cane, k¯o;

and vii) bottle gourd, ‘ipu. Pine wood is the lone alien taxon identiﬁed in the material

from the ﬁre-pits. It likely represents driftwood from the northwest coast of America

(Strong and Skolmen 1963) that washed up on the beach near Site 50–80–15–4853, where

the pine wood charcoal was recovered.

A plot of the proportion of Polynesian-introduced taxa against the

C age of the

ﬁre-pit indicates that the use of Polynesian-introduced taxa as ﬁrewood increased over

time (ﬁg. 8). Here, the proportion of Polynesian-introduced taxa is based on specimens

identiﬁed as wood and excludes the nutshells of kukui. Before 400 bp, Polynesian-

Figure 6:

Proportion of identiﬁed wood charcoal collected locally near the ﬁre-pit. See

table 1 for the ﬁre-pit labels. Note that proportion is represented on a continuous gray

scale, only three values of which are shown in the legend.

Figure 7:

Distribution of hau, Hibiscus tiliaceous, wood charcoal. See table 1 on page 4

for the ﬁre-pit labels.

introduced taxa were relatively rare in the ﬁre-pits, typically less than 10 percent of the

identiﬁed specimens. Afer 400 bp, wood from Polynesian introductions frequently made

up the majority of identiﬁed ﬁrewood.

Conventional radiocarbon age

Proportion introduced taxa

0.0

0.2

0.4

0.6

0.8

●● ●● ●● ●

●

100 200 300 400 500 600 700

Figure 8: Proportion of Polynesian introductions over time.

When the proportion of Polynesian-introduced taxa is plotted on a map of ﬁre-pit

locations (ﬁg. 9) a clear geographic pattern emerges. Most of the ﬁre-pits near the stream

yield relatively high proportions of Polynesian-introduced taxa. In contrast, ﬁre-pits at

the northern end of the plain are typically fueled primarily with native taxa and the

incidence of Polynesian-introduced taxa is relatively rare.

4 Reconstitution

The ﬁnal stage of analysis, called reconstitution, refers to a reason behind the structuration

that inheres in the relationship between the results of acquisition and some phenomenon

outside the acquired materials. The outside phenomenon might be a scientiﬁc law, or

an historical explanation based on human intention and motivation. The goal of the

reconstitution presented here is to propose changing patterns of behavior to account for

the observed structure in the data and to situate these within the context of traditional

Hawaiian social organization.

In a recent article, Dye (2010) compared the ﬁrewoods identiﬁed in ﬁre-pits on the

Waim¯analo Plain with ﬁrewood identiﬁed from ﬁre-pits on the elite Maui Island sites

of Hale Ki‘i and Pihana Heiau (Kolb and Murakami 1994). Dye found that ﬁre-pits

on the Waim¯analo Plain relied on shrubs and shrub-trees for ﬁrewood more than did

the ﬁre-pits from the elite Maui Island sites, which were rich in tree wood charcoal. He

plotted the proportion of tree wood charcoal over time in the Waim¯analo Plain ﬁre-pits

Figure 9:

Proportion of identiﬁed wood charcoal from Polynesian-introduced taxa. See

table 1 for the ﬁre-pit labels. Note that proportion is represented on a continuous gray

scale, only three values of which are shown in the legend.

and found evidence for a decline. This was explained as a consequence of the breakdown

of traditional Polynesian lineage organization and a concomitant rise in the authority

of ali‘i, one result of which was a restriction in the ability of maka‘¯ainana to gain free

access to forests and tree ﬁrewood.

The decline over time in use of tree ﬁrewood is still evident with the augmented data

set of ﬁre-pits analyzed here. The hypothesis that this decline was due, at least in part,

to a decline in the amount of tree ﬁrewood imported to the sites from inland locations is

directly supported by the observed increase over time in locally-collected ﬁrewood. It

appears that access to inland forests by people looking to collect ﬁrewood with which to

build ﬁres near the coast declined over time.

The hypothesis that this decline was due to assertions by ali‘i of property rights over

ﬁrewood in forests, a practice documented in the nineteenth century historical record, is

diﬃcult to assess with the ﬁre-pit data. Coincident with the decline in access to inland

forests is a shift to the use of ﬁrewoods introduced to the islands by Polynesians and

presumably growing on the plain near the ﬁre-pits. This appears to be an instance of the

replacement of native lowland forest, perhaps one deﬁcient in good ﬁrewood trees, with a

landscape transported from the East Polynesian homeland of the Hawaiian people (Kirch

1984), relatively rich in good ﬁrewood. If this were the case, then the decline in access to

inland forests might have been because the wood from them was no longer needed during

the course of a normal day on the plain, rather than property rights claims of ali‘i.

Regardless of the cause, however, a shift in behavior is clearly indicated by the data

from the ﬁre-pits. During the Early Period, prior to 400 bp, ﬁrewood was frequently

carried onto the plain from inland forests, where it was burnt with some locally-collected

wood. During the Late Period, after 400 bp, most ﬁrewood was collected locally from

forests rich in introduced taxa that had been planted and maintained by maka‘¯ainana.

Bayesian calibration of the

C age determinations from Early Period ﬁre-pits and Late

Period ﬁre-pits using the BCal software package (Buck et al. 1999) yields an estimate of

the boundary between the two periods with a peak in the mid-ﬁfteenth century (ﬁg. 10).

5 Discussion and Conclusions

The wood charcoal identiﬁcations from 35 ﬁre-pits on the Waim¯analo Plain, which are

related to precisely located and dated archaeological events, reveal patterns of change over

time and space that can be associated with traditional Hawaiian ﬁre-making behavior

and local vegetation change.

The ability of the wood charcoal data to yield information on vegetation change

augments environmental data from coring projects in several ways. According to the

interpretation of the ﬁre-pit data developed here, establishment of a managed, anthro-

pogenic forest on the Waim¯analo Plain was accomplished by the mid ﬁfteenth century.

This is quite a bit later than the forest changes documented in the pollen record of envi-

ronmental cores, which appear to have taken place very shortly after Polynesian discovery

and colonization of the islands (Athens et al. 2002). In contrast to the environmental

changes documented in paleoenvironmental cores, which are often dated by interpolation

Calendar Year AD

Probability

0.000

0.005

0.010

0.015

0.020

Early/Late Period Boundary

1450 1500 1550 1600

Figure 10:

Age estimate of the boundary between Early and Late Periods. The 95%

highest posterior density region is AD 1429–1537. The 67% highest posterior density

region is AD 1440–1480.

without estimates of uncertainty (Athens et al. 2002:61), it is possible to estimate the

date of the environmental changes recorded in the ﬁre-pits relatively precisely with a

posterior probability distribution that carries information about the uncertainty of the

estimate. The temporal resolution made possible with the ﬁre-pit data thus might add

detail and precision to the broad patterns established by pollen analyses.

The very small catchment areas indicated by the wood charcoal analysis, and the

ability to identify likely imports in the ﬁrewood, make it possible to track environmental

change at an extremely ﬁne geographic scale. The catchment areas indicated by the

ﬁrewood data, which appear to be less than a kilometer, are orders of magnitude smaller

than the catchment areas for fossil pollen deposits, which include wind-dispersed grains

capable of traveling hundreds of miles from their sources. In the case of the Waim¯analo

Plain, eﬀorts to establish useful plants appear to have concentrated near Puh¯a Stream,

where fresh water was plentiful, and to have lagged a kilometer away at the northern

end of the plain. This is a spatial resolution that other kinds of paleoenvironmental

investigation in Hawai‘i have yet to achieve.

Finally, in the context of this special issue of Hawaiian Archaeology, it is appropriate

to emphasize that the data analyzed in this paper were collected over a period of 15

years by eight diﬀerent contract archaeology projects. The focus of these projects on

acquisition of wood charcoal identiﬁcations yielded a consistent set of data capable of

further analysis. This pattern of many small eﬀorts coalescing into a larger result ought

to be a common one in the cultural resource management industry, but it will not happen

regularly until the industry lays aside the competition of the marketplace and warms to

the camaraderie of pursuing shared goals.

A Charcoal identiﬁcations

Fire-pit Taxon Part Count Weight (g)

119 16 unidentiﬁed bark 16 0.68

119 16 Sida cf. fallax wood 16 0.24

119 16 unidentiﬁed wood 5 0.12

119 16 Chenopodium oahuense wood 3 0.04

119 7 cf. Canthium odoratum wood 28 8.41

119 7 cf. Chamaesyce sp. wood 13 2.09

119 7 cf. Osteomeles anthyllidifolia wood 4 0.18

119 7 unidentiﬁed wood 2 0.25

119 7 unidentiﬁed bark 2 0.27

273 58 cf. Canthium odoratum wood 14 0.6

273 58 Nototrichium sp. wood 8 0.08

273 58 Chenopodium oahuense wood 6 0.12

273 58 Osteomeles anthyllidifolia wood 6 0.07

273 58 Chamaesyce sp. wood 6 0.11

273 58 unidentiﬁed wood 4 0.05

273 58 Nestegis sandwicensis wood 2 0.07

273 58 Diospyros sandwicensis wood 2 0.03

273 58 Sida fallax wood 2 0.03

308 159 Osteomeles anthyllidifolia wood 28 1.43

308 159 Sida fallax wood 27 0.81

308 159 unidentiﬁed wood 26 1.14

308 159 unidentiﬁed wood 6 0.2

308 159 unidentiﬁed bark 6 0.36

308 159 Cordyline fruticosa wood 3 0.07

308 159 cf. Senna sp. wood 2 0.07

308 159 cf. Canthium odoratum wood 1 0.05

308 159 cf. Pteridophyta stem 1 0.03

308 159 cf. Chamaesyce sp. wood 1 0.02

308 163 Osteomeles anthyllidifolia wood 32 1.27

308 163 Chenopodium oahuense wood 26 1.31

308 163 Sida fallax wood 18 0.55

308 163 cf. Rauvolﬁa sandwicensis wood 17 0.64

308 163 Nestegis sandwicensis wood 15 0.58

308 163 Diospyros sandwicensis wood 14 0.84

308 163 cf. Dodonaea viscosa wood 10 0.4

308 163 Chamaesyce sp. wood 9 0.44

308 163 unidentiﬁed wood 6 0.2

308 163 Aleurites moluccana nutshell 6 0.44

308 163 unidentiﬁed wood 5 0.23

308 163 cf. Dodonaea viscosa wood 4 0.13

Continued on next page

Fire-pit Taxon Part Count Weight (g)

308 163 Cordyline fruticosa wood 3 0.23

308 163 Nototrichium sp. wood 2 0.05

308 163 unidentiﬁed wood 2 0.05

308 163 Nothocestrum latifolium wood 1 0.02

308 175 cf. Rauvolﬁa sandwicensis wood 34 1.22

308 175 Diospyros sandwicensis wood 23 1.68

308 175 cf. Dodonaea viscosa wood 22 1.69

308 175 Osteomeles anthyllidifolia wood 13 0.78

308 175 Chenopodium oahuense wood 12 0.63

308 175 Nestegis sandwicensis wood 7 1.07

308 175 Aleurites moluccana nutshell 6 0.37

308 175 unidentiﬁed wood 6 0.19

308 175 unidentiﬁed bark 4 0.15

308 175 Chamaesyce sp. wood 4 0.15

308 175 unidentiﬁed wood 3 0.07

308 175 Sida fallax wood 2 0.14

308 175 Hibiscus tiliaceus wood 1 0.08

308 175 cf. Pteridophyta stem 1 0.04

308 180 cf. Rauvolﬁa sandwicensis wood 21 1.17

308 180 Chenopodium oahuense wood 21 0.72

308 180 Sida fallax wood 14 0.5

308 180 Osteomeles anthyllidifolia wood 13 0.66

308 180 Diospyros sandwicensis wood 10 0.23

308 180 Chamaesyce sp. wood 5 0.09

308 180 unidentiﬁed wood 5 0.1

308 180 unidentiﬁed wood 4 0.14

308 180 unidentiﬁed wood 3 0.07

308 180 unidentiﬁed wood 1 0.02

308 180 cf. Canthium odoratum wood 1 0.07

308 180 Aleurites moluccana nutshell 1 0.06

308 180 unidentiﬁed wood 1 0.03

308 180 cf. Pandanus tectorius twig 1 0.04

900 1 Sida cf. fallax wood 112 5.13

900 10 Monocotyledonae root 80 1.01

900 10 Monocotyledonae root 63 0.68

900 10 Monocotyledonae stem 21 0.28

900 10 Monocotyledonae stem 17 0.16

900 10 Aleurites moluccana nutshell 10 0.37

900 10 Metrosideros polymorpha wood 10 0.25

900 10 cf. Sida fallax wood 9 0.09

900 10 cf. Sida fallax wood 8 0.09

900 10 Hibiscus tiliaceus wood 7 0.19

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 10 unidentiﬁed wood 6 0.11

900 10 Metrosideros polymorpha wood 5 0.06

900 10 Chenopodium oahuense wood 4 0.08

900 10 Chamaesyce sp. wood 4 0.08

900 10 Ilex anomala wood 3 0.11

900 10 Abutilon sp. wood 3 0.03

900 10 unidentiﬁed wood 3 0.07

900 10 Aleurites moluccana nutshell 3 0.09

900 10 unidentiﬁed wood 3 0.12

900 10 Pinus sp. wood 2 0.08

900 10 unidentiﬁed bark 2 0.03

900 10 unidentiﬁed bark 2 0.04

900 10 Chamaesyce sp. wood 2 0.03

900 10 unidentiﬁed parenchyma 1 0.02

900 11 cf. Aleurites moluccana wood 112 0.74

900 11 Monocotyledonae root 50 0.88

900 11 Chenopodium oahuense wood 42 1.83

900 11 Metrosideros polymorpha wood 41 4.19

900 11 Monocotyledonae root 33 1.5

900 11 Metrosideros polymorpha wood 23 1.73

900 11 Monocotyledonae stem 13 0.25

900 11 Monocotyledonae stem 12 0.17

900 11 Sida cf. fallax wood 12 0.53

900 11 Chenopodium oahuense wood 11 0.36

900 11 Bidens sp. wood 9 0.35

900 11 Bidens sp. wood 9 0.55

900 11 Abutilon sp. wood 9 0.58

900 11 cf. Syzygium malaccense wood 7 0.28

900 11 unidentiﬁed bark 6 0.1

900 11 Hibiscus tiliaceus wood 5 0.16

900 11 Aleurites moluccana nutshell 5 0.06

900 11 Metrosideros polymorpha wood 5 0.1

900 11 Aleurites moluccana nutshell 5 0.14

900 11 Sida cf. fallax wood 4 0.05

900 11 Diospyros sandwicensis wood 4 0.16

900 11 unidentiﬁed bark 4 0.09

900 11 cf. Aleurites moluccana wood 4 0.09

900 11 Abutilon sp. wood 4 0.07

900 11 Syzygium malaccense wood 3 0.46

900 11 cf. Dodonaea viscosa wood 3 0.38

900 11 unidentiﬁed wood 3 0.33

900 11 cf. Osteomeles anthyllidifolia wood 3 0.07

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 11 cf. Psychotria sp. wood 2 0.01

900 11 Diospyros sandwicensis wood 2 0.03

900 11 unidentiﬁed parenchyma 2 0.03

900 12 Aleurites moluccana nutshell 385 18.64

900 12 unidentiﬁed bark 126 2.95

900 12 Chamaesyce sp. wood 91 3.11

900 12 unidentiﬁed wood 74 2.14

900 12 cf. Antidesma pulvinatum wood 58 4.14

900 12 Bobea sp. wood 52 1.97

900 12 Aleurites moluccana nutshell 38 1.99

900 12 Diospyros sandwicensis wood 36 1.46

900 12 Sida cf. fallax wood 31 0.99

900 12 Bidens sp. wood 26 0.61

900 12 Monocotyledonae root 25 0.54

900 12 unidentiﬁed bark 17 0.44

900 12 Abutilon sp. wood 17 0.48

900 12 Hibiscus tiliaceus wood 16 0.3

900 12 Chamaesyce sp. wood 11 0.51

900 12 Syzygium malaccense wood 11 0.41

900 12 Osteomeles anthyllidifolia wood 10 0.31

900 12 Monocotyledonae stem 10 0.3

900 12 Monocotyledonae stem 9 0.15

900 12 cf. Bobea sp. wood 9 0.16

900 12 cf. Acacia koa wood 9 0.34

900 12 Hibiscus tiliaceus wood 8 0.14

900 12 Sida cf. fallax wood 7 0.33

900 12 cf. Antidesma pulvinatum wood 6 0.36

900 12 unidentiﬁed wood 5 0.19

900 12 Abutilon sp. wood 4 0.05

900 12 Monocotyledonae root 3 0.05

900 12 Osteomeles anthyllidifolia wood 3 0.05

900 12 Acacia koa wood 2 0.12

900 12 Cocos nucifera nutshell 1 0.09

900 12 Diospyros sandwicensis wood 1 0.03

900 12 cf. Aleurites moluccana seed embryo 1 0.06

900 12 Metrosideros polymorpha wood 1 0.14

900 13 Aleurites moluccana nutshell 72 2.42

900 13 Aleurites moluccana nutshell 63 2.92

900 13 unidentiﬁed bark 27 0.61

900 13 unidentiﬁed bark 23 0.67

900 13 Chamaesyce sp. wood 20 0.64

900 13 Sida cf. fallax wood 15 0.59

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 13 unidentiﬁed wood 13 0.68

900 13 cf. Antidesma pulvinatum wood 11 0.3

900 13 cf. Bidens sp. wood 10 0.02

900 13 Chamaesyce sp. wood 9 0.21

900 13 cf. Antidesma pulvinatum wood 9 0.26

900 13 Monocotyledonae root 7 0.08

900 13 Diospyros sandwicensis wood 7 0.32

900 13 Bidens sp. wood 6 0.27

900 13 cf. Bobea sp. wood 5 0.31

900 13 cf. Bobea sp. wood 4 0.08

900 13 Hibiscus tiliaceus wood 4 0.04

900 13 Monocotyledonae stem 3 0.04

900 13 Hibiscus tiliaceus wood 3 0.11

900 13 Sida cf. fallax wood 3 0.26

900 13 cf. Scaevola sericea wood 3 0.07

900 13 cf. Gossypium tomentosum wood 2 0.05

900 13 Monocotyledonae stem 2 0.04

900 13 unidentiﬁed wood 2 0.05

900 13 Osteomeles anthyllidifolia wood 1 0.01

900 13 Diospyros sandwicensis wood 1 0.05

900 14 Syzygium malaccense wood 111 5.82

900 14 Chamaesyce sp. wood 104 6.77

900 14 Aleurites moluccana nutshell 74 2.59

900 14 Syzygium malaccense wood 74 3.35

900 14 cf. Antidesma pulvinatum wood 47 3.44

900 14 unidentiﬁed bark 46 1.66

900 14 cf. Antidesma pulvinatum wood 43 3.2

900 14 unidentiﬁed bark 42 0.76

900 14 Chamaesyce sp. wood 27 1.08

900 14 unidentiﬁed wood 21 0.45

900 14 cf. Syzygium malaccense wood 17 1.19

900 14 unidentiﬁed wood 16 0.33

900 14 Aleurites moluccana nutshell 15 0.73

900 14 Sida cf. fallax wood 10 0.16

900 14 Sida cf. fallax wood 7 0.26

900 14 unidentiﬁed wood 6 0.32

900 14 cf. Bobea sp. wood 4 0.13

900 14 cf. Gossypium tomentosum wood 2 0.07

900 14 Monocotyledonae stem 2 0.03

900 14 cf. Acacia koa wood 2 0.03

900 14 Diospyros sandwicensis wood 1 0.08

900 14 Nototrichium sp. wood 1 0.01

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 14 cf. Scaevola sericea wood 1 0.01

900 14 unidentiﬁed wood 1 0.04

900 15 Aleurites moluccana nutshell 51 1.64

900 15 unidentiﬁed bark 24 0.2

900 15 Chamaesyce sp. wood 22 0.61

900 15 cf. Antidesma pulvinatum wood 8 0.33

900 15 Diospyros sandwicensis wood 5 0.22

900 15 Sida cf. fallax wood 4 0.05

900 15 Hibiscus tiliaceus wood 4 0.02

900 15 Monocotyledonae root 3 0.04

900 15 unidentiﬁed wood 3 0.03

900 15 cf. Scaevola sericea wood 1 0.01

900 15 Chenopodium oahuense wood 1 0.01

900 15 Osteomeles anthyllidifolia wood 1 0.01

900 16 Aleurites moluccana nutshell 2988 92.76

900 16 Sida cf. fallax wood 32 0.37

900 16 Cocos nucifera nutshell 9 0.47

900 16 unidentiﬁed wood 6 0.05

900 16 cf. Gossypium tomentosum wood 5 0.07

900 16 Hibiscus tiliaceus wood 4 0.04

900 16 Monocotyledonae stem 3 0.03

900 16 unidentiﬁed bark 2 0.06

900 17 Aleurites moluccana nutshell 90 2.9

900 17 Aleurites moluccana nutshell 74 2.52

900 17 Chamaesyce sp. wood 30 0.63

900 17 Chamaesyce sp. wood 28 0.83

900 17 Hibiscus tiliaceus wood 25 0.74

900 17 unidentiﬁed bark 25 0.58

900 17 unidentiﬁed wood 23 0.66

900 17 unidentiﬁed bark 22 0.27

900 17 cf. Antidesma pulvinatum wood 16 1.42

900 17 cf. Antidesma pulvinatum wood 13 0.41

900 17 unidentiﬁed wood 12 0.3

900 17 Monocotyledonae stem 9 0.14

900 17 Hibiscus tiliaceus wood 8 0.13

900 17 Sida cf. fallax wood 6 0.1

900 17 Sida cf. fallax wood 6 0.08

900 17 cf. Bobea sp. wood 6 0.26

900 17 unidentiﬁed wood 4 0.07

900 17 unidentiﬁed parenchyma 3 0.08

900 17 Syzygium malaccense wood 2 0.17

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 17 Syzygium malaccense wood 2 0.05

900 17 Diospyros sandwicensis wood 2 0.02

900 17 Abutilon sp. wood 1 0.02

900 17 Cocos nucifera nutshell 1 0.03

900 17 cf. Acacia koa wood 1 0.01

900 17 Osteomeles anthyllidifolia wood 1 0.04

900 18 Chamaesyce sp. wood 63 1.94

900 18 unidentiﬁed bark 46 1.09

900 18 Aleurites moluccana nutshell 43 1.12

900 18 Sida cf. fallax wood 38 1.32

900 18 Hibiscus tiliaceus wood 31 0.65

900 18 Aleurites moluccana nutshell 30 0.78

900 18 Aleurites moluccana nutshell 29 0.48

900 18 Diospyros sandwicensis wood 26 1.27

900 18 Dodonaea viscosa wood 22 0.68

900 18 unidentiﬁed wood 22 0.47

900 18 unidentiﬁed bark 21 0.47

900 18 Sida cf. fallax wood 20 0.31

900 18 Gossypium tomentosum wood 18 0.39

900 18 unidentiﬁed bark 18 0.3

900 18 Chamaesyce sp. wood 17 0.49

900 18 Chamaesyce sp. wood 15 0.48

900 18 Hibiscus tiliaceus wood 14 0.18

900 18 Abutilon sp. wood 14 0.32

900 18 Abutilon sp. wood 14 0.26

900 18 Diospyros sandwicensis wood 14 0.33

900 18 unidentiﬁed wood 11 0.2

900 18 unidentiﬁed wood 10 0.14

900 18 Metrosideros polymorpha wood 10 0.24

900 18 Aleurites moluccana wood 9 0.13

900 18 Artocarpus altilis wood 9 0.12

900 18 cf. Psychotria sp. wood 7 0.23

900 18 Sida cf. fallax wood 7 0.13

900 18 cf. Pandanus tectorius wood 7 0.2

900 18 cf. Antidesma pulvinatum wood 6 0.07

900 18 Diospyros sandwicensis wood 6 0.25

900 18 Aleurites moluccana wood 6 0.07

900 18 cf. Dodonaea viscosa wood 6 0.22

900 18 unidentiﬁed wood 6 0.08

900 18 cf. Acacia koa wood 5 0.2

900 18 Abutilon sp. wood 5 0.25

900 18 Cocos nucifera nutshell 4 0.07

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 18 unidentiﬁed stem/root 4 0.09

900 18 Cocos nucifera nutshell 4 0.15

900 18 cf. Acacia koa wood 3 0.07

900 18 Metrosideros polymorpha wood 3 0.04

900 18 cf. Acacia koa wood 3 0.08

900 18 Osteomeles anthyllidifolia wood 3 0.06

900 18 Pteridophyta stem 3 0.04

900 18 Bobea sp. wood 2 0.04

900 18 Dodonaea viscosa wood 2 0.23

900 18 Chenopodium oahuense wood 2 0.02

900 18 Pteridophyta stem 2 0.01

900 18 cf. Syzygium malaccense wood 2 0.04

900 18 Hibiscus tiliaceus wood 2 0.01

900 18 cf. Bobea sp. wood 2 0.05

900 18 unidentiﬁed parenchyma 1 0.01

900 18 Osteomeles anthyllidifolia wood 1 0.11

900 18 unidentiﬁed wood 1 0.01

900 18 Artocarpus altilis wood 1 0.1

900 18 Bobea sp. wood 1 0.05

900 18 Chenopodium oahuense wood 1 0.3

900 18 Gossypium tomentosum wood 1 0.06

900 19 Aleurites moluccana nutshell 156 2.92

900 19 Aleurites moluccana nutshell 30 0.66

900 19 Sida cf. fallax wood 24 0.3

900 19 Sida cf. fallax wood 18 0.44

900 19 Aleurites moluccana nutshell 14 0.41

900 19 Osteomeles anthyllidifolia wood 14 0.52

900 19 Osteomeles anthyllidifolia wood 13 0.35

900 19 Sida cf. fallax wood 12 0.19

900 19 Osteomeles anthyllidifolia wood 9 0.09

900 19 unidentiﬁed bark 7 0.06

900 19 unidentiﬁed bark 5 0.1

900 19 unidentiﬁed bark 3 0.04

900 19 unidentiﬁed parenchyma 2 0.01

900 19 Hibiscus tiliaceus wood 2 0.03

900 19 cf. Bidens sp. wood 2 0.02

900 19 unidentiﬁed parenchyma 1 0.02

900 19 cf. Chamaesyce sp. wood 1 0.01

900 19 unidentiﬁed wood 1 0.01

900 19 Hibiscus tiliaceus wood 1 0.01

900 19 Chenopodium oahuense wood 1 0.02

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 2 Sida cf. fallax wood 7 0.09

900 2 unidentiﬁed bark 5 0.09

900 2 Aleurites moluccana nutshell 2 0.1

900 2 Chenopodium oahuense wood 2 0.06

900 2 Aleurites moluccana nutshell 2 0.04

900 2 cf. Acacia koa wood 1 0.01

900 20 Aleurites moluccana nutshell 23 0.8

900 20 Sida cf. fallax wood 19 0.23

900 20 Aleurites moluccana nutshell 18 0.56

900 20 Osteomeles anthyllidifolia wood 14 0.48

900 20 Chenopodium oahuense wood 11 0.26

900 20 Sida cf. fallax wood 10 0.25

900 20 Osteomeles anthyllidifolia wood 8 0.15

900 20 unidentiﬁed bark 5 0.03

900 20 Chenopodium oahuense wood 4 0.12

900 20 unidentiﬁed bark 3 0.03

900 20 Chamaesyce sp. wood 3 0.22

900 20 cf. Bidens sp. wood 1 0.01

900 20 unidentiﬁed parenchyma 1 0.01

900 20 Chamaesyce sp. wood 1 0.01

900 21 Palm sp. wood 12 0.15

900 21 cf. Saccharum oﬃcinarum stem 12 0.01

900 21 Chamaesyce sp. wood 6 0.07

900 21 Monocotyledonae stem 3 0.04

900 21 Sida cf. fallax wood 1 0.01

900 21 Cordyline fruticosa wood 1 0.01

900 22 cf. Osteomeles anthyllidifolia wood 24 0.65

900 22 Monocotyledonae stem 1 0.01

900 22 Cordyline fruticosa wood 1 0.02

900 22 unidentiﬁed bark 1 0.01

900 23 Colubrina oppositifolia wood 80 3.68

900 23 Nothocestrum latifolium wood 26 0.29

900 23 cf. Osteomeles anthyllidifolia wood 11 0.26

900 23 cf. Antidesma pulvinatum wood 7 0.34

900 25 Sida cf. fallax wood 106 6.22

900 25 Cordyline fruticosa root 96 3.58

900 25 Rauvolﬁa sandwicensis wood 82 5.18

900 25 Chamaesyce sp. wood 35 1.54

900 25 Abutilon sp. wood 24 1.22

900 25 Chenopodium oahuense wood 23 1.71

900 25 Hibiscus tiliaceus wood 23 1.32

900 25 cf. Saccharum oﬃcinarum stem 19 0.44

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 25 Bidens sp. wood 7 0.19

900 25 unidentiﬁed wood 3 0.24

900 25 Metrosideros polymorpha wood 2 0.08

900 25 Aleurites moluccana nutshell 2 0.16

900 25 unidentiﬁed tuber 2 0.12

900 25 Diospyros sandwicensis wood 1 0.03

900 25 cf. Aleurites moluccana kernel 1 0.01

900 25 Lagenaria siceraria rind 1 0.1

900 25 unidentiﬁed wood 1 0.09

900 26 Cordyline fruticosa root 35 0.79

900 26 Hibiscus tiliaceus wood 15 1.35

900 26 Poaceae stem 12 0.11

900 26 Cordyline fruticosa root 10 0.22

900 26 Poaceae stem 4 0.03

900 26 Cordyline fruticosa root 3 0.04

900 26 Rauvolﬁa sandwicensis wood 3 0.34

900 26 cf. Gossypium tomentosum wood 2 0.07

900 26 cf. Saccharum oﬃcinarum stem 2 0.01

900 26 cf. Saccharum oﬃcinarum petiole 1 0.01

900 27 cf. Cordyline fruticosa root 52 1.73

900 27 Poaceae stem 18 0.41

900 27 Cordyline fruticosa root 15 0.43

900 27 Poaceae stem 12 0.12

900 27 cf. Gossypium tomentosum wood 11 0.37

900 27 Cordyline fruticosa root 10 0.13

900 27 Sida cf. fallax wood 10 0.42

900 27 Sida cf. fallax wood 8 0.25

900 27 cf. Canthium odoratum wood 7 0.13

900 27 Rauvolﬁa sandwicensis wood 6 0.7

900 27 unidentiﬁed wood 5 0.4

900 27 Poaceae stem 4 0.08

900 27 Sida cf. fallax wood 3 0.03

900 27 Myrsine sp. wood 3 0.37

900 27 Acacia koa wood 3 0.17

900 27 Chenopodium oahuense wood 2 0.07

900 27 cf. Canthium odoratum wood 2 0.16

900 27 unidentiﬁed wood 1 0.01

900 27 unidentiﬁed twig 1 0.01

900 27 unidentiﬁed wood 1 0.05

900 27 Rauvolﬁa sandwicensis wood 1 0.19

900 27 Aleurites moluccana nutshell 1 0.04

900 28 cf. Hedyotis terminalis wood 46 1.75

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 28 Rauvolﬁa sandwicensis wood 40 1.9

900 28 Artocarpus cf. altilis wood 37 2.61

900 28 Canthium odoratum wood 20 0.96

900 28 unidentiﬁed wood 12 0.66

900 28 Nestegis sandwicensis wood 8 1.37

900 28 unidentiﬁed wood 7 0.09

900 28 Diospyros sandwicensis wood 5 0.21

900 28 cf. Cheirodendron sp. wood 5 0.12

900 28 Sida cf. fallax wood 2 0.31

900 28 unidentiﬁed wood 2 0.12

900 28 Hibiscus sp. wood 2 0.16

900 28 Lagenaria siceraria rind 1 0.04

900 28 Cocos nucifera nutshell 1 0.02

900 28 unidentiﬁed wood 1 0.01

900 29 Chamaesyce sp. wood 23 5.8

900 29 Pittosporum sp. wood 15 5.87

900 29 Myoporum sandwicense wood 10 1.64

900 3 Sida cf. fallax wood 24 0.98

900 3 Aleurites moluccana nutshell 5 0.15

900 3 Chenopodium oahuense wood 3 0.02

900 3 Sida cf. fallax wood 3 0.05

900 3 cf. Chamaesyce sp. wood 2 0.02

900 3 cf. Aleurites moluccana cf. kernel 2 0.005

900 3 Chenopodium oahuense wood 2 0.06

900 3 cf. Artocarpus altilis wood 1 0.01

900 4 cf. Osteomeles anthyllidifolia wood 11 0.4

900 4 Chenopodium oahuense wood 11 0.14

900 4 Wikstroemia sp. wood 5 0.1

900 4 Osteomeles anthyllidifolia wood 5 0.09

900 4 Aleurites moluccana nutshell 4 0.17

900 4 Chenopodium oahuense wood 4 0.08

900 4 cf. Osteomeles anthyllidifolia wood 3 0.14

900 4 unidentiﬁed parenchyma 1 0.005

900 4 cf. Nestegis sandwicensis wood 1 0.01

900 4 unidentiﬁed wood 1 0.02

900 4 Chamaesyce sp. wood 1 0.01

900 5 Aleurites moluccana nutshell 23 1.02

900 5 Aleurites moluccana nutshell 12 0.65

900 5 Sida cf. fallax wood 2 0.05

900 5 Sida cf. fallax wood 2 0.07

900 5 Sida cf. fallax wood 1 0.06

900 5 Chamaesyce sp. wood 1 0.03

Continued on next page

Fire-pit Taxon Part Count Weight (g)

900 5 Chenopodium oahuense wood 1 0.02

900 5 unidentiﬁed wood 1 0.04

900 5 Chenopodium oahuense wood 1 0.06

900 6 Aleurites moluccana nutshell 37 2.32

900 6 Sida cf. fallax wood 30 1.09

900 6 unidentiﬁed wood 6 0.14

900 6 Aleurites moluccana cf. kernel 2 0.03

900 6 Chenopodium oahuense wood 1 0.03

900 6 cf. Cocos nucifera wood 1 0.03

900 7 Diospyros sandwicensis wood 11 0.54

900 7 Aleurites moluccana nutshell 8 0.31

900 7 Osteomeles anthyllidifolia wood 5 0.1

900 7 Osteomeles anthyllidifolia wood 4 0.18

900 7 Aleurites moluccana nutshell 1 0.05

900 8 Diospyros sandwicensis wood 26 1.55

900 8 cf. Osteomeles anthyllidifolia wood 18 0.64

900 8 Cordyline fruticosa wood 5 0.12

900 8 unidentiﬁed wood 4 0.16

900 8 Myoporum sandwicense wood 2 0.49

900 9 Monocotyledonae root 89 1.64

900 9 Monocotyledonae root 72 0.67

900 9 Monocotyledonae stem 51 0.58

900 9 Monocotyledonae stem 22 0.28

900 9 Sida cf. fallax wood 12 0.21

900 9 Charpentiera sp. wood 9 0.09

900 9 Sida cf. fallax wood 4 0.04

900 9 Abutilon sp. stem 3 0.1

900 9 Charpentiera sp. wood 3 0.23

900 9 Pinus sp. wood 3 0.21

900 9 Hibiscus tiliaceus wood 2 0.03

900 9 unidentiﬁed bark 2 0.02

Glossary

anthropogenic Of, relating to, or involving the impact of man on nature.

coconut The palm, Cocos nucifera.

context A unit of stratiﬁcation associated with a natural or cultural process or event.

detritus

Material produced by the disintegration and weathering of rocks that has been

moved from its site of origin, or a deposit of such material.

diachronic

Of, or relating to, or dealing with phenomena as they occur or change over a

period of time. See also synchronic.

ﬁll

Any sediment deposited by any agent so as to ﬁll or partly ﬁll a valley, sink, or other

depression.

ﬁre-pit

A pit of varying depth, often bowl shaped at the base, usually identiﬁed by

a concentration of charcoal and/or burned material in the ﬁll, especially at the

feature interface.

fossil

An object of natural origin, such as a pollen grain, found in a geological or

archaeological context.

habit

A botanical term used to describe the general appearance, growth form, or archi-

tecture of a plant.

in-built age

The age of a material when it was incorporated into the archaeological

record.

mountain apple

The tree, Eugenia malaccensis, a forest tree to 50 ft. high, that bears

fruit resembling an apple. Traditionally the fruits were eaten both raw and pickled.

richness The number of classes in a collection or population, a measure of diversity.

sand

Detrital material ranging in size from 0.5 mm to 2 mm in diameter. See also

detritus.

shrub

A plant with multiple stems and shorter than trees, typically under 5–6 m. See

also habit.

shrub-tree

Plants whose form can take on the characteristics of either a shrub or a tree.