ROY F.

ELLEN

Non-domesticated resources in Nuaulu

ecological relations *

In this paper an attempt is made to calculate and then assess the significance
in energetic and broad ecological terms of the exploitation of non-domesticated
resources by the Nuaulu, a small group of some 500 hunters, collecters and
swidden cultivators of south central Seram, eastern Indonesia (see Figure 1)’.
It is hoped that it will demonstrate that consideration of such matters can
have important (sometimes unforeseen) consequences for analysing the eco-
nomies and social organization of what are ostensibly communities based
on swidden agriculture as their primary mode of subsistence. On the other
hand, it also indicates certain features which are either idiosyncratic of the
Nuaulu situation, or which stem from the particular kinds of resources exploi-
ted and the technical and organizational requirements necessary for their
appropriation.

Parameters of settlement

As must be evident to anyone who has seriously considered the matter, to

view a settlement pattern simply as a spatial entity results in only a partial
understanding of a complex situation. Nevertheless, for practical purposes
and as a point of departure, a particular area exploited by a specified population
locus may, in an approximate way, be defined cartographically. Thus, by
and large, the absolute limits of Nuaulu exploitation can be represented gra-
phically, although the boundaries are irregular and changing. Figure 3 shows
the entire area exploited by the Ruhuwa Nuaulu (with a population at the
time of my 1970 census of 180 persons) during the period in which fieldwork
took place -

some 900 square kilometres, broadly coinciding with the drainage

basins of three major watercourse systems: the Ruatan, Pia and Lata-Tevi. In
dynamic terms this represents the limit of the environment as it interacts with
a single specified population; in anthropocentric terms it represents the Ruhu-
wa-Nuaulu exploitive area, which biologically can be characterised as a single
ecosystem.
Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016
127-
128

3#f
(Z~
N

-
t3
z
5
Q
1
z
--
t3
.C,4
W
-5
Q
’0
fi
§
.S?
13
1:I
0
4~
#
sot
0
w
!I~

’Zs
ao
3
Q
-jz
-5
t3.!2
z
t4i
Q

Q
b
r,
~3
W

90
5i

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

 129

Figure 2. Nuaulu villages: their present location and immediate environs *

*
Shaded areas represent land above the 500 metre contour. Names in upper case lettering
indicate specific topographically-distinctive areas of forest distinguished and exploited by the
Nuaulu. Nuaulu villages are those in italic : 1. Nuelitetu; 2. Bunara; 3. Watane; 4. Aihisuru;
5. Hahuwalan; 6. Sepa; 7. Ruhuwa; 8. Tamilau.

And yet, having said all this, the differential intensity of exploitation for
various parts of this area as a whole indicate that the larger part is really of
very little direct ecological or economic relevance for Ruhuwa subsistence
or social organization. Field data on work scheduling for a test period of
524 man-days, illustrated in Table 1, shows that the exploitation of non-
domesticated resources occurred in 13 definable localities of varying size and
importance. However, the cumulative area of all 13 localities (214 square
Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016
130

5
IZL)

R!ql)
l~6)

5
k
q6)

;13
t3

14)
-Z
I’~,

l~3
.2
--z
ao
IZL)
Qo
14)

i~
cri
S
w,

an
E

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

 131

kilometres) does not match the entire area which the Nuaulu maintain they
theoretically exploit. Consequently, there is here a lack of congruity be-
tween ideology and statistics. However, this is emphatically not a distinction
between claimed territory and area actually exploited, since Nuaulu concepts
of territory, except in the narrow sense of land owned and cultivated, have
been largely only of historical importance since the occupation of the present
settlement localities 2.
Now, to some degree it must be expected that had the composition of my
field sample been different and had it covered a different period of time the
results would have accordingly differed. This I accept, but the important
point to grasp is precisely that there is this variability over time and that the
boundaries of the area exploited during one period are slightly different from
the boundaries for some other period. Furthermore, when the figures given
in Table I are themselves analysed, the differential intensity of exploitation
for each of the 13 separate localities is shown up dramatically, as is the varia-
bility from locality to locality in the type of activity undertaken. The data
considered in the table will be discussed further below, its relevance here has
simply been to illustrate the fact that attempts to define precise boundaries
for the exploitive environment are empirically impossible in the Nuaulu case
and that acceptance of such a boundary might result in inaccurate deductions
concerning their ecological relations. It is much more helpful to see the
threshold between the exploited and non-exploited parts of the environment
as a varying point along a spatial continuum, changing through time and
as a function of demand for various different non-domesticated resources.
Given the view that the boundaries of the area exploited are dynamic and
not static, it is far less likely that the arena in which the great majority of
Ruhuwa environmental interactions take place will be seen in isolation from
the wider socio-ecological context. I think that failure to appreciate this
has been responsible in the past for some concealment of the critical variables
affecting settlement generation. It becomes quite apparent, therefore, that
not only is a purely cartographic approach dangerous, but also that there
are some awkward definitional problems involved, which, if not treated, tend
to distort subsequent analysis 3.

The exploitive environment

I have been able to illustrate above that within the settlement area of Ruhuwa
as a whole, environmental exploitation is far from being uniform, for reasons
of the differential availability of various resources, or due to natural barriers
effectively limiting availability. This suggests that the ecological variation
within a given exploitive area may be quite considerable, determining the
breadth of choice available in the selection of particular locations for resi-
dence and exploitation of particular resources; but it is typical of tropical

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

132

Table I. Differential exploitation of the non-domesticated environment, Ruhuwa,

1970 (measured in man-days) *

*
For a discussion of the sample used, see Ellen (1973, pp. 467-474).
Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016
 133

forest cultivators that the actual latitude of choice is restricted very sub-
stantially by the monotypic nature of the ecology of the area 4.
Morphologically, the Nuaulu settlement pattern can be seen in terms of
a series of dual analytic components. First there is the distinction which
can be drawn between that part of the total environment which is exploited

by a single population isolate and the much more restricted area actually
inhabited by that population, emphasizing the separation of resources and
residence 6. Further, a valuable spatial distinction can be drawn between the
inner area of maximal exploitation, and the outer area in which man-environ-
ment interactions are much more diffuse. Finally, a distinction can be made
between the domesticated and non-domesticated environments, resting on the
nature of the exploitation of resources and therefore not necessarily implying
spatial separation 6. It should be noted that although these three kinds of
analytic distinction are parallel they are in no way identical.
Here I am not primarily concerned with the inner zone of concentrated
exploitation of the domesticated environment, an area approximately bounded
by the most distant gardens; rather the emphasis is on the outer zone of inter-
mittent exploitation, but also with the non-domesticated environment as
a whole. Since the greater part of this zone is occupied by mature and primary
forest it is this which much be considered first.

Primary forest

For the Nuaulu, the principal distinguishing features of wesie (largely unowned
and uncultivated forest land) and wasi (land that is owned and has been cul-
tivated) are the trees from which they are composed, in particular their size.
Thus, almost synonymous with the contrastive set wesie: wasi is the set aj
onata: aj ana wasi, &dquo;great trees: trees of the garden&dquo;. Not only is this a
simple dual classification of types of tree, of primary and secondary forest
associations, it also has far-reaching ecological and cultural implications.
Thus, the category aj onata has the following cultural associations:
1) the source of hini (the principal components of the house frame and the
village ritual house);
2) the habitat of many of the most malevolent spirits known to the Nuaulu;
3) the source of major game animals, and of birds for consumption and
decoration and ritual purposes the hombill (Aceros plicatus Forster),
-

cassowary (Casuarius casuarius L.), etc. ;

4) the source of forest products such as resin (Agathis alba Foxw.), ijane nuts
(Canarium commune L.), sago and many fruits; and,
5) the location of numerous natural objects with magical associations, and
of mythical events, stories and such like.
But aj onata removed from its natural forest context changes its character
entirely. It ceases to be NATURAL and becomes CULTURAL; the tran-
sition is cognitive, and frequently linguistic. To a certain extent, the properties
Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016
134

of natural species are thought to change as a result of this transformation;

attitudes towards uncooked (WILD) game animals become neutralized by
cooking and its presence in a DOMESTIC setting. For example, Python
reticulatus (Schneider) is ritually dangerous for some clans in all its manifes-
tations in a wild, live and natural state; once dead and cooked the only taboo
is on its consumption. Parallel dichotomous attitudes could be given for
many other natural products (floral and faunal) which are utilized culturally.
Thus, it is the context, the ecological zone in which it is found, that aj onata
exists and in which it manifests its significant characteristics.
Wesie has all the characteristics of aj onata and more, for the floristic compo-
nent of the Nuaulu environment consists of basically three categories of
vegetative type; aj, trees and woody shrubs; wane, lianas and trailers, and
rahue, herbs and shrubby plants. All of these are found in a particular
combination to give wesie. Each separate and specific component of the
flora has qualities by which the Nuaulu recognise it and utilize on appropriate
occasions, but the forest as a whole is an entity which divulges its own pecu-
liarities. It has noises; the shrieks of birds, falling trees and boughs, game
animals, the perpetual drone of cicadas and other insects, the resonance caused
by felling timber and the beating of huge buttress roots. But the reason
why such noises are so marked is because of the never-ending and all-pervading
background of silence on which they are only superimposed incidents. It
is said, &dquo;there is no wind in the forest, it is evil&dquo;.
Nuaulu venture into the forest frequently, but always with diffidence, taking
both practical and magical precautions. Travel is largely by established
minor and major paths, using well-known landmarks, such as standing stones
and prominent trees, together with artificial markers. The most important
of these thoroughfares are indicated by a broken line in Figure 3. At some
strategic points there are huts built as halting places for long-distance treks,
such as those to the north coast. These are said to afford little protection
against the spiritual as well as physical elements.

Distribution, structure and composition of wesie

The distribution of wesie at the present time in the vicinity of Ruhuwa is
illustrated in Figure 4. It will be noted that the observable distribution
falls into two categories: a), the bulk of forest stretching from a very minimum
of around 100 metres from the coast northwards and mountainwards, and
b), isolated patches and pseudopodia apart from the major block of forest
and forming barriers between cultivated areas. This latter residual distribu-
tion tends to be along ridges and around steep knolls, unsuitable for cultivation
under normal circumstances. As a result, these areas (together with the
margins of the main forest block) within easy access of the village, are a prin-
cipal source of construction timbers and certain other products, leading to
Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016
 135

gradual thinning and denudation. This subject is dealt with in further detail
below (see also ibid., 181-182).
Wesie consists of two basic types of forest association: the typical tropical
rain forest, which stretches from sea-level mountainwards, and montane
rain forest (including mossy forest areas), usually above 1 000 metres. The
lower tropical zone may include submontane rain forest, between 500 and
1 000 metres, but the critical heights given, based on calculations by van
Steenis (1935) (see Richards, 1952, p. 348), are only approximate, affected as
they are by the so-called Massenerhebung effect. This is usually defined as the
differential rate of change of climatic factors with height varying from place to
place and causing the actual altitudinal limits of the vegetation zones to differ
on different mountain ranges and different parts of the same mountain (ibid,

p. 347). But although the Nuaulu recognise zonal variation within wesie
areas, as it is of no practical utility to them, it is an unimportant aspect of
their environmental model. The only occasions on which montane zones
are traversed are during journeys to north Seram, or on the longer hunting

trips to the headwaters of the Nua, Ruatan, Kawa or Lata (Figure 3), or in
collecting resin from the conifer Agathis alba Foxw., a particularly prominent
feature of the higher forest zones.

Although no detailed studies exist, in composition, lowland forest below

1 000 metres (mostly on low hill land) can be said to be broadly typical of
other afforested insular areas of southeast Asia and New Guinea, with a
structural complexity related to the great number of different tree species,
approximately 80-100. Over 250 distinguishable Nuaulu plant taxa are
apparently associated with wesie. These can be grouped into three broadly
identifiable vegetation layers: large trees, woody lianas and perching plants,
such as orchids and ferns (see Robbins, 1961) 1. The variety of species means
that the composition of tree communities vary considerably and that pure
stands are comparatively rare. Canopy height tends to be around 30-40
metres.
In Seram, the dominant of the more inland and highland areas,
typical
Agathis appears to be replaced towards the coast by other codominants,
such as Canarium commune L., species of Terminalia, Calophyllum, Myristica,
hardwoods such as Pterocarpus and Diospyros, in favourable localities the
giant Melaleuca leucodendron L., and in riverine areas, stands of Ficus and
Casuarina. These same lowland areas contain seasonal sago swamp, and
occasionally permanent swampland, of mixed species of some importance
in Nuaulu subsistence economy. Ground cover, generally thin, consists
largely of pteridophytes, principally ferns of various descriptions, but also
distinctive species of Selaginella, mosses, small herbs and tree seedlings (for
further details see Ellen, 1973, pp. 450-464).

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

 136

<

8
£
u
.fl
00
gf
gf
.
’@b
w

*
J3
’o
~3
33 at
to
>
&dquo;5
C4 a
’ty
li~, CM
o> ,o

IZSO 1 s
’s
0g
C
c
;13
tal
~3 w
X
14.) h
-S l#
li~~
b
w s
.5 3
1: §.
,a! 13%i
d
’ttE gas
-E, ’~ E ~ .~ o w >
’s ~
li l/X
0 .fl ~
S ’5S
u
ou
iz .i3
<c
bx
~-S
~ ~o
S6S

i-§
*P£
art
E.a.c~
~3 M
Hx
* ~0

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

 137

Exploitation of non-domesticated food resources

One fact that has not received the attention it deserves, and which can have
critically important implications for understanding the ecological and economic
organization of communities such as the Nuaulu, is the crucial role of hunting
and gathering of wild vegetable foods in societies nominally regarded as
subsisting by swidden agriculture (see Barrau, n.d., pp. 28, 30; Boserup,
1965, p. 54). The Nuaulu, in particular, devote a considerable proportion
of both their time and energy to these subsistence activities, and the food so
obtained contributes a relatively high proportion of their energy intake.
Indeed, although there is no Nuaulu concept of a primary and only dependable
single source of food, sago and meat, both essentially products of the non-
domesticated environment, rank at the top of lists of the most important
foods. Activities in which the Nuaulu interact with and exploit the envi-
ronment without cultivating it can be classified into the following six broad
but economically distinctive categories: hunting; fishing; the collection of
miscellaneous animal foods, such as insects and grubs; the cutting of sago;
the collection of tree-food products, plants and fungi, and the extraction of
non-edible forest products. The account given here is designed purely as
a short discussion of these particular subsistence activities as elements in
Nuaulu man-environment relations. Consequently, only minimal attention
is paid to the organizational and technological details involved.
Hunting provides the overwhelming proportion of animal protein consumed
by the Nuaulu, although the consumption of fish is nutritionally important,
almost certainly more so than in former times. In order of dietary impor-
tance, the major game animals are pig, cuscus, deer and cassowary. In this
respect the actual frequency of consumption matches well the value attached
to these animals as sources of food in Nuaulu thought. Theoretically, the
so-called Lord of the Land may impose ritual prohibitions, sasi, on the hun-
ting and collecting of certain species, ::s he can for the gathering of vegetable
products, though this seldom happens in practice 8. Table 2 condenses
information collected from two Ruhuwa households over the equivalent
of a four-month period in 1970 regarding the major sources of animal protein
and its consumption 9.
Game is hunted by the Nuaulu all the year round, though the supply is
less plentiful during the height of the rainy season, when success in hunting
tends to decrease. There exist basically two types of Nuaulu hunting orga-
nization : a), cooperative hunts between a large number of persons recruited
specifically for the purpose, sometimes on a clan basis, though not necessarily
along genealogical lines; and b), individual hunting, or hunting undertaken
in pairs, where the participants are usually closely related. In both types
domesticated dogs are used in the former case in large numbers, and to
-

considerable effect, in tracking, attacking and baiting. The most common

weapons of the hunt are bows and arrows and spears of various kinds. Fire-

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

138

<u
ja
.r
#
.-..
r/J
a~

5
R
0B
c~UJ
.~
f: &Scaron;’ o
3 M
(,)

-At .9
x
;3 s
1D .
§
0 .9~
r
.r
t.o)
~’7
~5 ~-
w
’- ......
:::: O °

0
0)
0.
~ 0.
§12
&dquo;0&dquo;&dquo;
l/5
lt3
a
.~ 0..2 c:
a U-
.9 ¡:¡j
C
,~1) 5 .5
Q
;>&dquo;0
a> u
3%
g
o
u 3
<u ~
*~ .....&dquo;0
S
§ #&copy; &copy; n # &copy;
M ’O
a a~
Q- M 3
r/J

t.o)
z &copy;
-0..
!L.
zs
<:) o
~&copy;
~-&Scaron;
10&dquo;
0 ..c:H8’- 0
=’
S
S S
&dquo;-e °~ &copy;
q6) ~-
OB
s
~ 0
110
I - n
c: ’E
a
.E V
c G&dquo;
>i
N -a,~
M 3
.2
zm
:ë flK4
Cd
f3 ~SG
O G
&dquo;0
1l6S
.9
S ’?
oq .DS
*
9
oj

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

 139

arms, in the form of breechloading rifles (usually obtained from the Police
and military) are occasionally used, especially in connection with ceremonial
and ritual group hunts (ibid., pp. 151-153). Trapping, as distinct from hunting,
is quite common but largely restricted to the immediate area of the gardens,
(the inner zone), often as a protective device against wild animals (ibid., pp. 303-
306). The most important of these are spear traps erected over known deer,
pig or cassowary trails, often in series of up to ten at a time. In addition,
weighted noose traps are used for cuscus, in series of up to twenty (see Ellen,
1972), spring noose traps are used for brush turkeys, civets, cassowaries (and
occasionally deer and pigs) and a further kind of noose trap, manually operated,
is used for a number of bird species, particularly hornbills, cackatuas and
lories. The pit trap is known to the Nuaulu but not used by them.
Sea fishing is alien to the traditional Nuaulu way of life and engaged in
only by a few men and boys, using baited lines of easily available trade nylon
with simple steel hooks, or multiple-headed hooks for squid. It is generally
restricted to littoral zones, although outrigger canoes are occasionally used.
Freshwater fishing, on the other hand (an activity of both sexes), is sanctioned
by tradition, and both fast-running streams and rivers and muddy beaches
are exploited by a number of methods; by poisoning with the root of Garcinia
dulcis Kurz.; with basket traps, particularly for freshwater prawns, when
they are set in weirs and rapids; damming, and with ordinary and multiple-
headed spears and arrows. The collection of shellfish for eating and the
preparation of lime is common from both marine and freshwater localities.
A wide variety of other minor fauna are gathered; as recognized secondary
sources of animal foods, such as grubs; as relishes and sweetmeats (e.g. honey-
comb), and as occasional snacks, as in the case of frogs and grass-hoppers.
Fauna are also the source of a variety of non-food products: bone and antler
for tools and other artifacts; shell for lime, implements and decoration; skins
for drums and hafting purposes; fur and feathers for ornamentation. Most
of these things are obtained as by-products from hunting game for consump-
tion as food. There are, however, some birds, for example Cacatua moluc-
censis Grn. and Tanygnathus megalorhynchus Boddaert, which are sought
solely for their plumage, for ritual reasons or to keep as pets.
Wild food plants and fruits collected are considerably more varied than
the fauna. Table 3 lists the most important of these and their contribution
to Nuaulu diet, as reflected in data collected during the four month survey
period 1°.
Clearly, if abundant forest food plants exist close to the village, individuals
are not going to go further than they have to to secure them, and generally

speaking (but with the important exception of sago) such roots and fruits
are relatively easily available, seldom requiring venturing far beyond the
limit of the furthest gardens. In fact, most gathering of wild Canarium nuts
(which are periodically consumed in large quantities in conjunction with
important festivals), bamboo shoots and such like, takes place in secondary
Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016
140

forest within the inner zone of exploitation, on the fringes of existing gardens.
The conclusion may therefore be drawn that despite the extent of the total
area of the non-domesticated environment, the intensity of its exploitation
increases in inverse proportion to the distance from residential areas, such
that the inner zone also represents the location of the most intensive exploi-
tation of non-domesticated resources. This appears to be the case, both
in terms of measurement of the number of man-days devoted to work in
particular specified localities (Table la), and in terms of the index derived
by dividing these figures by the approximate area the localities cover (Table
lb). Which of these measures is actually the more accurate indicator of
productivity is difficult to determine, on account of the large number of extra-
neous variables involved.

The collection of naturally propagated sago

Sago is seen by the Nuaulu as their most important food source, and no meal
is considered complete without it. It is attributed with strength-giving qua-
lities and regarded as an aphrodisiac. In Nuaulu terms, hunger is the condi-
tion which prevails in the absence of sago.
Most sago consumed by the Nuaulu is derived from wild palms which
vegetatively regenerate themselves by means of suckers u. During the period
from April to August 1970 it was calculated that 53.7 percent of all sago
consumed was from non-domesticated sources. On the basis of the figures
presented in Table 3 this would mean that about 1 056 cals per head per day
are acquired in this way, representing 76 percent of the total weight of non-
domesticated plant resources converted into available food energy. However,
since these data were obtained largely during the wet season, when conditions
are least favourable to the exploitation of wild sago, the real figures are quite

probably higher.
By far the greatest proportion of naturally propagated sago comes from
the rich, high-density, sago-palm forest areas towards the confluence of the
Nua and Ruatan rivers, known to the Nuaulu as Somau and to the adjacent
coastal peoples as Tehun (Figure 3, Table 1). Measured orthographically,
Somau is approximately 16 kilometres from Ruhuwa over hilly and riverine
afforested country. Under normal circumstances, it is considered a six hour
journey. In addition, wild sago palms are also cut at suitable localities,
usually where the land is permanently damp or swampy, on the river Pia,
at Lahati, at the mouth of the Lata, Pino and Satu, and as far eastwards as
the mouth of the Jala.
Trips to sago localities may be individual or group activities (involving
parties of up to 15 males), short or extended on the region of five days or
more, sometimes organised on a kinship basis but more often not. Extended
stays often allow for hunting to be combined with the cutting of sago, and
when particularly large quantities are required for important ceremonies
Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016
 141

0
0.-

0
3

-9
.sT
z
0
CZL~
&dquo;r4
z
t3

IZL)&dquo;
z
,~1)

,~3
0

11 Is,
Izu
L
3
11
BU

C).
&dquo;,t3
14.)
3
v
In
BU

’b
I::>
s:

0
0
’I~ c=
3
cri ,
6.

.2
s O
cd E
H u-
0

-a
0.)
3
..

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

142

the occasion is accompanied by great festivity and chanting. Such expeditions

often include women and children.
With the cutting of these naturally regenerating palms many individuals
may work a single stand until it is exhausted and then proceed to another.
However, separation of the flour from the pith is an individual or household-
oriented activity 12. Naturally set palms are always flowering, or about to
flower, when felled, when starch reserves are at their maximum. This is
not true for planted palms, where close proximity to the village encourages
cutting before a stand has fully matured. Most naturally reproducing sago
is consumed by the worker and his immediate kin and affines, though occasio-
nally small quantities are exchanged in the Muslim settlement of Sepa for
trade goods. On average it was found that 100 cubic centimetres of trunk
gave about 0.02 kilograms of moist flour, which in Nuaulu terms means that
a good palm can yield up to 20 casks (the equivalent of 106 x 104 cals).

Available sources of non-edible forest products

Mature forest, as well as the non-domesticated environment in general, is

as much valued as a source of a wide range of non-edible products as it is
for foods. The most important of these is timber for construction purposes
and (almost as important, on account of its ritual as well as economic signi-
ficance) kamane resin from the conifer Agathis. Most timber comes from
within a radius of three to four kilometres of the village, and usually from
land bordering the garden. Good kamane resin is much more difficult to
come by and accordingly Nuaulu may travel considerably further to obtain

it, sometimes 18 or 20 kilometres to the higher montane regions renowned

for its abundance and quality, such as the headwaters of the river Kawa, Satu
and Lasa (see Figure 3). Rattan of high quality is also regarded as worth
the effort of travelling substantial distances to obtain, and areas frequented
in this respect include Hatuhahu, Lahati and Sama (Figure 3). Other non-
edible forest products are usually obtained from within a much shorter radius,
rarely exceeding the boundaries of the most distant gardens. Such products
include firewood, usually dry dead bamboo of the otherwise unusable Schizos-
tachyum species, from interstitial and marginal thickets in and around gardens
and unowned areas of secondary growth. Collecting firewood is almost
solely an activity of females, though sometimes old men may assist them.
A wide range of types of bamboo are also in great demand as raw materials
for building, manufacture of utensils and basketry. The constant cutting
of bamboo from areas near the village tends to transform normally gregarious
bamboo thickets into more open areas, with a greater diversity of shrubs,
grasses, herbs and tree seedlings, despite an acknowledged fast growth rate.
On the steeper river banks the succession is typically to the various hydrophilic
herbs and ferns. Clearance of this kind can also result from a policy of
Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016
 143

marginal cultivation of fruit trees, coconut and sago palms, or be part of a

deliberate programme of clearing land for gardens.
Minor forest resources include bark for bow-strings and loin-cloths, varieties
of herbs, climbers and roots for dyes, poisons and the manufacture and deco-
ration of utensils and other objects. Apart from timber and bamboo, the
most important source of house-building materials is the sago palm, the
woody peticles of which are used for walling and the fronds of which give
the most superior thatch 13.
The obvious tendency to procure such products, in particular rattans,
timber, bamboo and Metroxylon products, as near to the focus of settlement
as possible inevitably leads to the depletion of stable primary associations
of forest the nearer one gets to the main loci of settlement. Consequently,
when primary forest is cut for gardens it has almost always been considerably
exploited already and contains plant associations more typical of regenerated
secondary forest (ibid., pp. 181-182). In this sense, it has already been trans-
formed to depleted primary growth, and, like open secondary associations,
may become subject to marginal cultivation as well as complete clearance:

primary forest, depleted forest, marginal cultivation,

--~ DENUDATION -> -->

bamboo thicket open secondary complete clearance

association

Input-output aspects of exploitation of non-domesticated environment

On the basis of the data provided above, and augmented by that supplied
elsewhere (ibid., pp. 446-464, 467-474) the contribution of non-domesticated
resources of the Nuaulu environment is clearly of some substance, in terms
of the variety offered, satisfaction of acknowledged requirements and as a
contribution to diet. Nutritionally, such sources provide in the region of
1 272 cals of energy per head per day compared with 1 813 cals derived from
all domesticated sources. It is striking that over forty percent of an estimated
daily adult intake of about 3 085 cals, a figure which is in itself quite high by
comparative standards 14, is derived from the natural component of the envi-
ronment among a group of sedentary swidden cultivators, since there is no
evidence that the area under cultivation is inadequate to meet the necessary
nutritional requirements in fact there are small surpluses. Non-domes-
-

ticated sources also provide 64 percent of the protein fraction of all consumed
food. Although I have only presented data here on carbohydrate and pro-
tein, it is clear that in terms of fat intake also non-domesticated resources
are of paramount importance. Pandanus, Canariwn, large and small game
alike all have a high fat content. while root tubers, fruits, greens and other
characteristic cultivated foods are notably low in fat, with the exception
of the groundnut.

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

144

This food value may be obtained from a maximum of in the order of 120
Linnaean species of fauna (about 146 named and edible Nuaulu terminal
taxa) and around 48 Linnaean species of plants (about 92 named and edible
terminal taxa). But although the total number of non-domesticated natural
species exploited in this fashion is extensive, over 80 percent by weight of
plant-derived food comes from a single genus, Metroxylon. The fact that
carbohydrate input is as high as it is is amost solely due to the importance of
the exploitation of naturally-reproducing sago in Nuaulu economy. Similarly,
almost 90 percent by weight of animal-derived foods come from sources
covered by only 4 Nuaulu primary taxa: pig, fish, cuscus and deer. Conse-
quently, despite apparent variety, the principal sources of food energy and
protein, which these two groups respectively represent, are extremely specific.
Nuaulu male labour input in terms of both time and energy resources,
regarding the exploitation of the non-domesticated environment (Table 4),
is more than the contribution to working domesticated resources, at least
56 percent of the total energy expenditure devoted to productive activities
during the test period as opposed to 44 percent (ibid., pp. 347.351). Again, the
importance of the non-domesticated component in Nuaulu man-environment
relations (an ostensibly horticultural people) is effectively demonstrated.

Table 4. Energy expenditure of males involved in exploitation of non-

domesticated resources, Ruhuwa, 1970

*
Figures in parenthesis indicate actual number of man-days. For notes on the method-
ology employed in these calculations see Ellen (1973, pp. 467-474), and Norgan, Ferro-
Luzzi and Durnin (1974, pp. 339-341).
**
Source: Hipsley and Kirk, 1965, p. 43.

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

 145

In adjacent Muslim communities (even taking into consideration the impor-

tance of fish as a source of fats and protein) energy expenditure in the exploi-
tation of the non-domesticated environment and the calorific contribution
it makes to their diet is much less, while, conversely, much more time is spent
in the exploitation of domesticated resources. It is not surprising, then,
that the Nuaulu exploitive environment should be considerably greater spa-
tially (as it is), despite the relative smallness of its population. This illustrates
the subtle way in which the overall shape of the exploitive environments and
settlement patterns of the respective communities diverge. The implications
of the minimization of territorial overlap in terms of intergroup competition
and conflict are considerable, leading to the conclusion that the manner in
which the Nuaulu manage their relations with the non-domesticated envi-
ronment helps to regulate relations between them and adjacent communities,
by relieving the pressure on garden land and other resources (cf. Barth 1956).
Space does not permit an examination of the detailed mechanics of this process
here, but it is hoped that a full account will be published elsewhere.
Finally (as an afterword), I wish to return to the question of the differential
importance of various localities for the exploitation of non-domesticated
resources, and consider once again the data presented in Table 1. Above
and in Table 4, I have illustrated the relative importance of different activities
in terms of energy expenditure, but I have not related this to the locality in
which the activity is undertaken. There are two important reasons for paying
some attention to this question ; the first is that the distance from the village
will affect actual energy expenditure, and the second the significance of the
differential intensity of exploitation for an understanding of the settlement
pattern.
In the computation of energy expenditure in Table 4, I was forced to rely
on rates of expenditure for the same or similar activities on the work of others
in comparable areas (e.g. Hipsley and Kirk, 1965; Morgan, Ferro-Luzzi and
Durnin, 1974). For reasons which I have given elsewhere (Ellen, 1973,
pp. 467-474), I believe that the inaccuracies incurred by so doing are insuffi-
cient to affect significantly the results obtained and the conclusions I have
drawn from them. Had this been a specifically nutritional survey precise
measurements would have been essential. The point ought to be made,
however, that the time and energy devoted to travel in extracting wild sago,
is substantially greater than, say, fishing or even hunting (see Table 1).
Table la shows the number of man-days for various activities undertaken
in a number of different localities. It indicates that more time was spent in
two localities, the Upa river area and Somau, than all other localities put
together (some 62 percent of the total number of man-days). Together,
these areas occupy about 62 square kilometres. Compared with the total
area exploited by the Nuaulu, say in the course of one year, this is very small,
and it might be concluded that its productivity is therefore high. In a sense
this is correct, but if productivity is measured by a man-day/area index (Table

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

146

lb) picture that emerges is somewhat different. While, as I indicated

the
above, real productivity by locality proves a difficult index to compute accu-
rately, the numerical data does emphasize the most important fact to bear
in mind; that exploitation patterns do vary quite considerably, from area
to area, according to vegetative cover, faunal composition, topography and
relative accessibility.

Conclusion

From the data presented in this brief paper, I think four main points have
emerged:
1) that the area and intensity of exploitation of non-domesticated resources
varies time, and that the intensity of exploitation varies from locality
over
to locality, together with the particular kind of activity involved;
2) that in terms of both energy input and output the importance of exploi-
tation of non-domesticated resources in Nuaulu ecological relations and
economy is considerable;
3) that animal flesh obtained through hunting, trapping and fishing is the
most important source of protein and fat in Nuaulu diet; and

4) that to a very large degree the emphasis on collecting and those activities
in general in which the Nuaulu interact with the non-domesticated environ-
ment is determined by the prominent role of extraction of flour from naturally-
reproducing sago-palms.
Despite the apparent elementary nature of these conclusions, it is important
to remember that they are not immediately obvious and that it has only been
possible to demonstrate their real significance through careful and systematic
investigation. In particular it should make us wary of, a) the bias introduced
by purely geographical approachs to settlement patterns which tend to empha-
size only those readily observable cartographic features of human manipu-
lation of the environment, and b), parallel bias hitherto associated with ethno-
graphic accounts of communities comparable to the Nuaulu. After all,
they are unlikely to be totally exceptional in these respects ls.

Roy F. Ellen (born 1947) is Lecturer in Social Anthropology at Keynes College, University
of Kent at Canterbury, England. He is engaged in research on the "Ecology and ethno-
biology of the Nuaulu, Seram, Eastern Indonesia"
.

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

 147

Notes

*
Paper delivered at the Colloquium &dquo;tcologie et Soci6t6 en M61an6sie&dquo; (Ecology and
Society in Melanesia) held at the Fondation Maison des Sciences de I’Homme, Paris, May
1973.
1. The data discussed here were collected during an 18 month fieldwork period, between
1969 and 1971. I should like to express my gratitude to Lembaga Ilmu Pengetahuan Indone-
sia (The Indonesian Academy of Sciences) in Jakarta, under whose auspices I worked; and,
for support through grants in aid, to the Social Science Research Council, the London-Cornell
Project for East and Southeast Asian Studies, the Central Research Fund of the University
of London and the Galton Foundation.
I should like to make a general point here. Because the relevant detailed support studies
in botanical ecology, dietary composition and energy expenditure do not yet exist for Seram,
I have been compelled to rely heavily on comparable research undertaken in these fields in
Papua New Guinea. I have felt justified in doing this on account of the close ecological,
subsistence economic and population biological similarities between the two areas. Further-
more, my concern in this paper is not so much with precise nutritional and input-output
calculations (although in the circumstances I have aimed for a high degree of accuracy), but
rather with ratios between domesticated and non-domesticated resources for different food
categories, and with the ratio between the labour inputs involved in exploiting them. The
practical difficulties involved in measuring energy flow for human communities under
field (as opposed to laboratory) conditions are considerable and often underestimated in
studies of this kind, but I am satisfied that those errors which remain in my own measurements
do not substantially affect the criterial ratios I discuss.
2. The Nuaulu are not traditionally a coastal people. The present pattern of village dis-
tribution is largely a result of Dutch colonial attempts to pacify and bring under greater
governmental control the mountain peoples of central Seram, prior to and around the begin-
ning of the present century. Tichelman (1925, pp. 659-661) reports that by 1925 no perma-
nent mountain village existed in the Amahai district (in which Ruhuwa lies), although some
clans periodically occupied highland villages in a clandestine fashion.
3. For a discussion of some aspects of the ethnographic significance of mapping, the inter-
pretation of culturally-structured landscapes and the importance of setting socio-ecological
problems in a regional sphere of reference see Conklin (1968, pp. 111-120).
4. The sources and effects of such ecological variation as there is (see Conklin, 1961, p. 30)
have been discussed in some detail in relation to the selection of sites for gardens, village loca-
tion and layout in an unpublished analysis of Nuaulu settlement patterns (Ellen, 1973).
5. For many communities this distinction between village area and exploited area will be
something of an abstraction. Unfortunately it is an abstraction which is commonly perpe-
tuated and reflected in the ethnographic literature, largely due to the neglect of one or other
of these parts. For example, although there exists an impressive collection of details on the
village-plans of some Amazon tribes (e.g. James, 1949; L&eacute;vi-Strauss, 1956), there is precious
little data on the arrangements of their cultivation plots and their relation to the village.
Needless to say, a settlement pattern constitutes the sum total of such arrangements.
6. The terms domesticated and non-domesticated used here are adopted from the usage of
Rappaport (1968). As far as this writer is aware, this particular phraseology has not been
used elsewhere, but it does (I think) have clear advantages.
7. In alluvial areas there is a tendency for this arrangement to be replaced by five distin-
guishable vegetation layers and a general higher diversity index.
8. Sasi almost always relate only to crops, though in traditional Nuaulu territories they
could apparently be imposed with respect to fishing rights in certain rivers. Another, rather
exceptional, example was given to me of a sasi imposed by the Dutch in the thirties on the

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

148

hunting of deer for two years, as shrinkage in their numbers had apparently given rise to
some concern by the district authority. This was the only case informants could give of a
sasi relating to game. See also Ellen, 1973, pp. 352-354.
9. It has not been possible to enter the proportion of food given away to other households.
The figures given here relate to the actual consumption as recorded in daily dietary records
and not to the total yields of food for each household. It is estimated that for meats in the
pene category (pig, deer and cassowary) the weight consumed by a single household rarely
exceeds 50 percent by weight of the orig inal carcase. For other meat the amount redistributed
tends to be negligible. This is also true for other foods from non-domesticated sources,
with the exception of sago, where the amount retained by the household is frequently below
60 percent of that originally worked.
10. Some idea of the total number of different animal species exploited by the Nuaulu for
food has been given in a more comprehensive report on Nuaulu ecology and settlement
(Ellen, 1973, pp. 446-449). A more complete list of wild food plants and fruits utilized is
also given in the same volume (pp. 450-464).
11. The importance of the distinction between naturally reproducing and cultivated sago
palms in Nuaulu economy is reflected in their thought, for it demonstrates a single but never-
theless striking instance of the dichotomy between the natural and the cultural, and the impli-
cations of this in terms of cognition and behaviour (ibid., 159-160). The same phenotypic
varieties, of which some ten are recognized (ibid., Table 15), appear in both cultivated and wild
types, being genetically indistinguishable and interchangeable, suckers from the wild palm
often being collected for planting. This, however, does not appear to lessen the cognitive
importance attached to the distinction. It was generally agreed though that spinous varie-
ties were more typical of the wild type, while smooth varieties were more frequently planted.
This is despite the fact that informants were perfectly aware of the fact that spineless varieties
can give rise to spinous ones and vice versa, and that several gradations of spinosity are dis-

tinguished taxonomically, tending to support the proposition that spinescence is of little or

no value as a diagnostic characteristic (Barrau, 1959, p. 153, 1958, p. 37; Lea, 1964, p. 120).
12. The actual method of extraction of sago-flour has been discussed by other writers for
southeast Asian and New Guinea peoples (e.g. Morris, 1953; Lea, 1964, pp. 120-122) and
Nuaulu techniques are largely similar (see Ellen, 1973, pp. 157-162). The details are there-
fore not dealt with here.
13. An extensive inventory of these products which can be collected in nearby mature and
secondary forest, and even within cultivated areas (e.g. sago frond thatch), is given elsewhere
(ibid., 450-464).
14. The average daily intake, irrespective of age, in those households for which consump-
tion was measured comes to 2 314 cals, compared with the 3 085 cals estimated daily adult
intake (ibid., pp. 467-474). Both these values are high when compared with the figures for
somerecent nutritional studies of similarly organized communities in New Guinea:
1. Chimbu Venkatchalam, 1962 1 930
2. Tsembaga Rappaport, 1968, p. 283 2 015
3. Pari* Hipsley and Kirk, 1965 2 360&plusmn;185
* 4. Bomagai-Angoiang Clarke, 1971, p. 178 2 650
* 5. Siane Salisbury, 1962, pp. 80-81 2927
* Adult males only.
Earlier work, such as that undertaken by Hipsley and Clements (1947), appears to have rather
underestimated the actual per capita consumption (see Rappaport, 1968, pp. 283-284).
15. Some further data on the overall consumption pattern in Ruhuwa is given in Ellen
(1973, pp. 467-474).
16. This would place the Nuaulu between type 1 and type 2 of Watters’ classification of
swidden agriculturalists (1960, p. 63); that is, they are almost equally dependent on hunter/
gathering and gardening. However, they could subsist with little difficulty on non-domesti-
cated resources alone.

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

 149

References

Barrau, J.
1958 Subsistence agriculture in Melanesia. Honolulu, Bernice P. Bishop Museum
(Bulletin 219).
1959 "The sago palms and other food plants of marsh dwellers in the South Pacific
Islands", Economic botany 13: 151-162.
n.d. An ethnobotanical guide for anthropological research in Malayo-Oceania. Jakarta,
Unesco Science Cooperation Office for Southeast Asia. (Preliminary draft.)
Barth, F.
1956 "Ecological relationships of ethnic groups in Swat, North Pakistan", American
anthropologist 58: 1079-1089 (reprinted in: A.P. Vayda (ed.). Environment and
cultural behavior : Ecological studies in cultural anthropology. New York, Natural
History Press, 1969).
Boserup, E.
1965 The conditions of agricultural growth : The economics of agrarian change under
population pressure. London, Allen and Unwin.
Clarke, W.C.
1971 Place and people: An ecology of a New Guinean community. Berkeley/Los
Angeles, Calif., University of California Press.
Conklin, H.C.
1961 "The study of shifting cultivation", Current anthropology 2: 27-61, (also published
in: Pan American Union Studies and Monographs VI, Washington, DC).
1968 "Some aspects of ethnographic research in Ifugao", Transactions of the New
York Academy of Science 30: 99-121.
Ellen, R.F.
1972 "The marsupial in Nuaulu ritual behaviour", Man 7: 223-238.
1973 Nuaulu settlement and ecology : An approach to the environmental relations of an
Eastern Indonesian community. (Thesis submitted for the degree of Doctor of
Philosophy in the University of London, unpublished.)
Hipsley, E.H.; Clements, F.W.
1947 Report of the New Guinea nutrition survey expedition. Canberra, Department of
External Territories.
Hipsley, E.H.; Clements, F.W.; Kirk, N.
1965 Studies of dietary intake and expenditure of energy by New Guineans. Noumea,
South Pacific Commission (South Pacific Commission Technical Paper 147).
James, A.G.
1949 Village arrangement and social organization among some Amazon tribes (Thesis
submitted for the degree of Doctor of Philosophy, New York, Columbia Univer-
sity, published).
Lea, D.A.M.
1964 Abelam land and sustenance Swidden horticulture in an area of high population
density, Maprik, New Guinea. (Thesis submitted for the degree of Doctor of
Philosophy in the Australian National University, Canberra, mimeographed.)
L&eacute;vi-Strauss, C.
1956 "Les organisations dualistes existent-elles?", Bijdragen tot de taal-Landen volken-
kunde 112: 99-128. (Reprinted in: C. L&eacute;vi-Strauss, Structural anthropology,
New York/London, Basic Books.)
Morris, H.S.
1953 Report on a Melanau sago producing community in Sarawak. London, Her Majes-
ty’s Stationary Office (Colonial Office Research Studies 9).
Norgan, N.G.; Ferro-Luzzi, A.; Durnin, J.V.E.A.
1974 "The energy and nutrient intake and the energy expenditure of 204 New Guinean",
Philosophical Transactions of the Royal Society of London, B 268 : 309-348.

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016

150

Platt, B.S.
1962 Tables of representative values of foods commonly used in tropical countries. Lon-
don, Her Majesty’s Stationary Office (Medical Research Council Special Report
Series, 302).
Rappaport, R.A.
1968 Pigs for the ancestors : Ritual in the ecology of a New Guinea people. New Haven,
Conn./London, Yale University Press.
Richards, P.W.
1952 The tropical rain forest, an ecological study ? Cambridge, Mass., University Press.
Robbins, R.G.
1961 "The vegetation of New Guinea", Australian territories 1: 21-32.
Steenis, C.G., van
1935 "Maleische vegetatieschetsen", Tijdschrift v.d. Nederlandsche Aardijkskundig
Genootschap: 25-67, 171-203, 363-398.
Tichelman, G.L.
1925 "De onderafdeeling Amahei, Seram", Tijdschrift van het Koninklijk Nederlandsch
Aardrijksundig Genootschap 42: 653-724.
Venkatachalam, P.S.
1962 A study of diet, nutrition and health of the people of the Chimbu area. Territory
of Papua and New Guinea, Department of Public Health (monograph 4).
Watters, R.F.
1960 "Some forms of shifting cultivation in the Southwest Pacific", Journal of tropical
geography 14: 35-50.

Downloaded from ssi.sagepub.com at PENNSYLVANIA STATE UNIV on March 4, 2016