Content uploaded by Wiebke Kirleis
Author content
All content in this area was uploaded by Wiebke Kirleis on Sep 20, 2014
Content may be subject to copyright.
1
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
Chapter 16
The Megalithic Landscape of Central Sulawesi, Indonesia:
Combining Archaeological and Palynological Investigations
Wiebke Kirleis, Johannes Müller, Corinna Kortemeier,
Hermann Behling and Santoso Soeghondo
Abstract
Central Sulawesi is characterized as both a hot spot of biodiversity as well as a megalithic landscape
with a clustered distribution of megalithic sites. One of the main goals of this preliminary evaluation is
to untangle the development of societies with monumental constructions in relation to environmental
change and diversity. Initial results presented in this paper highlight the research potential of a combined
investigation of ecological data and archaeological surveys. The conclusions are based both on pollen data
as well as data regarding the spatial distribution and architecture of megalithic “kalambas” (jars), combined
with the interpretation of the few available excavation reports. The palynological investigations concentrate
on a comparison of on-site and off-site cores, to identify both the general vegetation dynamics as well as
local anthropogenic influences that correspond with the environment. The archaeological considerations
are based on spatial analyses and typological comparison of the architectural features of 15 sites (including
93 megaliths) in the Besoa valley and particularly at the Pokekea site. As a result of this work, we assume
the existence of one central prehistoric domestic site with ritual activities and an associated spread of
smaller hamlets, reflected in the distribution of megaliths for each basin. However, the chronological
framework of the development is still uncertain. Anyhow, we offer a terminus ante quem for the erection
of kalambas of around AD 830. Hypothetically, an initial opening of the landscape around 50 BC, as
indicated by pollen analyses, may be linked to a possible early construction phase of the monuments.
However, the combination of both environmental and preliminary archaeological analyses results in the
first interpretations of ecological and monumental developments in the central Sulawesian region.
Introduction
The Besoa valley (60km²) study area is situated in a UNESCO biosphere reserve surrounding the Lore
Lindu National Park in central Sulawesi, covers 218,000ha and holds the largest remaining mountainous
rainforest of Sulawesi. Vegetation gradients range from lowland rainforests below 1000m to upper
montane and elfin forests above 2000m (Kirleis et al. 2011: 166). The temperature is constant over the
whole year. Lowland temperatures vary between 26° and 32°C during the day, whereas they decline about
6°C for every 1000m elevation towards the highlands (Guide Lore Lindu 2001: 3).
The prominent archaeological features in the Besoa valley are monuments, the megalithic kalambas,
that indicate long-term human impact on the landscape. The kalambas, with their elaborate decorated
disc-shaped lids, are unique megalithic features of central Sulawesi [Fig. 16.2]. They are supplemented
by cup-marked stones and ornamented stelae of animal- and human-shaped figures in relief (Kaudern
1938; Dwi Yani Yuniawati 2001: 20). Sparse remains of, potentially later disturbed, kalamba infillings
indicate that the kalambas were used for burial purposes (Haris Sukendar 1980a). Urn burials excavated
in the vicinity of kalambas, typological considerations of retained artifacts, stylistic parallels, and first
results of radiometrically-dated pollen analysis from two kalambas point to construction times before
1
Crossing Borders 16.indd 1 6/11/2012 3:48:33 PM
2
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
Fig. 16.1: Locations of Bada, Besoa and Napu valleys and two further important palaeo-ecological sites in Sulawesi,
Indonesia, and distribution of archaeological and palaeo-ecological sites in the Besoa valley (Map: W. Kirleis / I. Reese).
1. Pokekea: 60 features (21 kalambas, eight lids, one stela, four boulder complexes, seven cup-marked stones, 19 blocks); 2.
Pongkore: three kalambas; 3. Bangkeloeha: one kalamba; 4. Hanggira: one block; 5. Entovera: three kalambas; 6. Lempe:
one block; 7. Lempe 2: two kalambas; 8. Padang Hadoa: five kalambas; 9. Pengkoa: two boulder complexes; 10. Doda-
South: one kalamba; 11. Doda-Besoa: two kalambas; 12. Doda Village: one block; 13. Tadulako and surroundings: seven
megaliths (one stela, six kalambas); 14. Padantanga: one block; 15. Masora: three features (two blocks, one kalamba).
A Pollen profile Bariri Tower, BT [Fig.16.7], B Pollen profile Pokekea (POK) [Fig.16.8], C and D Pollen profiles MP-2a
and MP-4a [Figs. 16.7 and 16.8].
Crossing Borders 16.indd 2 6/11/2012 3:48:37 PM
3
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
AD 1210 and before AD 830 respectively (Haris Sukendar 1980a, 1980b; Dwi Yani Yuniawati 2001: 28;
Kirleis et al. 2011: 174). Megaliths and urn burials appear in several central Sulawesi valleys (Kaudern
1938; Bintarti 2000: 73), such as the Besoa, the Bada and the Napu valleys [Fig. 16.1]. Both previous
and recent sondages revealed domestic waste in between the megaliths (Haris Sukendar 1980a; Dwi Yani
Yuniawati 2001: 27; Dwi Yani Yuniawati et al. 2004). Therefore, most of the kalambas were probably
constructed within domestic sites.
All archaeological data (location, size of sites and monuments, architectural types and decoration
patterns) was combined in a dataset of 15 sites consisting of 93 megaliths [Fig. 16.1]. A quantification and
qualification of the data provided first insight into possible interpretations of the archaeological record
[Table 16.1]. A spatial analysis based on Thiessen polygon structures was conducted to identify aspects
Fig. 16.2: Kalambas and lids at Pokekea, Besoa valley (Photos: W. Kirleis). Kalambas sampled for pollen analyses
[kalamba 74 at the very left: see Figure 16.7, kalamba 27 at the very right: see Figure 8] (Source: W. Kirleis).
Crossing Borders 16.indd 3 6/11/2012 3:48:37 PM
4
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
Table 16.1: Megalithic and sub-megalithic features and design of the decoration in Pokekea (1), Bewa (2) und Masora (3). Sum of design appearance of
each category is marked in brackets behind the site label
Design / Item Kalambas Lids Stelae Dolmen / circles / Cup marked Stones / Grinding Sum
Site no. (n) Boulder complexes stones Blocks stones (n)
Human faces 1(8) 1(1) 1(1) 3(1) 1(1), 2(2) 18
Cup marks 1(1) 1(1), 3(1) 1(7), 3(8) 1(5) 16
Cup marks and beams 1(1) 1(1) 1(3) 5
Bands 1(1) 1(1) 2
4 bands 1(2) 2
5 bands 1(3) 3
6 bands 1(4) 4
7 bands 1(1) 1(1) 1
Animals 1(2) 1(1) 3(1) 1(2), 2(2) 8
Geometrical forms 3(1) 1
Hearts 3(1) 1
Squares 1(1) 1(1) 2
Central hollow 1(1), 3(1) 3(1) 3
Knob 1(5) 5
Circles 1(1) 1
Floral design 3(1) 1
Site n n n n n n n n
1 Pokekea 21 8 1 4 7 19 60
2 Masora 1 2 3
3 Bewa (Bada) 11 4 1 8 13 1 38
Source: C. Kortemeier / J. Müller.
Crossing Borders 16.indd 4 6/11/2012 3:48:37 PM
5
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
of the centrality and diversity of different sites [Fig.16.3]. Our study focuses on the Pokekea site (S 01°
41’29.52’’, E 120° 12’48.08’’), situated not only on a promontory offering a strategically favourable
location along the bends of the rivers Lumamba and Lengi, but also surrounded by agriculturally fertile
areas [Fig. 16.1]. This site shows the highest density of all the widely distributed megalithic and sub-
megalithic sites in the valley.1
Pokekea offers the opportunity to carry out palynological investigations on peat from the immediate
vicinity and on sediment in-fillings of the kalambas. The two short on-site core profiles and two landscape
profiles, taken with a Russian peat corer (50cm length, 5cm diameter) during a sampling campaign in
August 2006 by Wiebke Kirleis and Herrman Behling close to the archaeological remains, indicate past
human-environment interactions and allow for deeper insight into the land-use of the first megalith-
building Sulawesian societies. Original data from the two on-site cores is presented in this paper, while
the off-site pollen analyses are summarised from a previous paper by two of the authors (Kirleis et al.
2011). As the latter show severe changes in the composition of past vegetation in the Besoa valley, one
of our main research questions was whether a first fire clearance of tropical rainforest about 2000 years
ago can be connected to the construction of megalithic sites.
Fig.16.3: Spatial arrangements at Pokekea: Thiessen polygon structures, S 01°41’29.25’’, E 120°12’48.08’’
(Drawing: C. Kortemeier).
Crossing Borders 16.indd 5 6/11/2012 3:48:38 PM
6
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
Research history
Archaeological Research History
The research history since the late 19th century comprises expeditions, surveys and small excavations in
central Sulawesi (Adriani and Kruyt 1898; Kruyt 1908; Heine-Geldern 1928; Kaudern 1938). Excavations
by Haris Sukendar (1980a, 1980b) on sites in the Bada valley have shown that the kalambas were used
as multiple burial containers. Fragments of human skeletons, skulls, teeth and pottery were discovered.
Later observation of the bones showed that at least ten individuals must have been buried in one jar.
The jars are described as the “metal age megaliths” and as belonging to the metal ages of Southeast
Asia (Bellwood 1979), featuring an ornamentation of anthropomorphic, zoomorphic and geometric
figures (Kaudern 1938; Whitten et al. 2002). In 2009, a visit by two of the authors (Kirleis / Müller)
confirmed the possibility of further investigations in the area. The first results of earlier data collections
are summarized by one of the authors in a thesis at Kiel University (Kortemeier, in prep.).
Archaeobotanical Research History
There is a general lack of archaeobotanical studies in Indonesia, especially in Sulawesi. Recently, several
pollen profiles from central Sulawesi have been analyzed within “Stability of Rainforest Margins,
STORMA” research program funded by the German Research Foundation (Cook et al. 2008a, 2008b;
Kirleis et al. 2011; Haberzettl et al.). These pollen records can be interpreted in the face of human
development in central Sulawesi from the mid-Holocene onwards. Further reconstructions of vegetation
history through pollen analyses on terrestrial material have been carried out in Borneo (Hunt and
Rushworth 2005), Halmahera (Papay Suparan et al. 2001), Java (Stuijts 1993; van der Kaars et al. 2001;
van der Kaars and van den Bergh 2004), Kalimantan (Anshari et al. 2001; Eko Yulianto et al. 2005), New
Guinea (Hope 1983; Haberle et al. 1991), Sumatra (Morley 1982; Flenley 1988; Maloney 1990; Flenley
and Butler 2001) and Sulawesi (Gremmen 1990; Dam et al. 2001; Hope 2001). Along with investigations
of marine cores (van der Kaars 1991, 1998; Wang et al. 1999; van der Kaars et al. 2000, 2010) which give
further information on climate variation and sea-level fluctuation, the results display quaternary vegetation
development in Indonesia in the long run, focusing on major shifts and drifts like the Pleistocene-
Holocene transition. Two main pollen records from Sulawesi deal with long-term vegetation developments
and show hardly any evidence of human activity: palynological research at Lake Tondano, North Sulawesi
[Fig. 16.1], displays the last 33,000 years of vegetation history; at the transition from the Pleistocene
towards the Holocene an increase of precipitation and temperatures furthered the spread of forests and
the rising of lake levels. During the Holocene, human influence was marginal (Dam et al. 2001). Another
pollen record from a swamp near Lake Matano, in South Sulawesi [Fig. 16.1], shows vegetation events
reaching back 75,000 years, demonstrating a forest spread during the last glacial maximum (LGM) about
20,000 years ago. Smaller amounts of micro-charcoal without any further hints of anthropogenic influence
suggest that natural fires occurred throughout the Holocene (Hope 2001).
Archaeological Survey
In the central Sulawesian basins of Bada, Besoa and Napu, 147 megalithic features were observed, of
which 93 cluster in the Besoa valley (Kaudern 1938; Haris Sukendar 1980a; Dwi Yani Yuniawati 2004;
Kortemeier, in prep.).
To conceptualize the development of the megaliths and related sites in combination with completed
and in progress environmental studies, we concentrated our inquiries on the Besoa valley and the site
of Pokekea. The collected data from about 15 sites with 93 megalithic features enabled us to start first
analyses [Fig. 16.1; Table 16.2].
The Pokekea site contains the overall highest number and the widest spectrum of megalithic features
in central Sulawesi. The 60 megaliths feature six categories of architectural types: cup-marked stones,
kalambas, kalamba lids, stelae, ornamented and unornamented blocks and agglomerations of diverse
blocks [Table 16.1]. Most prominent are the 21 kalambas, the highest number in other category.
Crossing Borders 16.indd 6 6/11/2012 3:48:38 PM
7
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
Spatial Arrangements at Pokekea
The spatial arrangement of the 60 megaliths at Pokekea was analyzed using GPS coordinates for each
megalith [compare Table 16.1]. As a first attempt to map spatial patterning, Thiessen polygons were
calculated [Fig.16.3]. As calculated from the point accumulations and the size of polygons, the polygon
structures indicate that the main cluster marks the centre of the Pokekea site, showing a kalamba as the
most central point. In addition to its centrality, the spatial tendency of the site was revealed, using the
evaluation of the standard deviation of all megalithic features (standard deviation ellipse). The entire site
extends to the Northwest, retaining the majority of megalithic features in the North and West of the plain.
However, as cup-marked stones appear only in the northwestern and the southeastern areas but not in the
central jar cluster of Pokekea, different activities might also be responsible for intra-site differences. The
whole Pokekea site measures about 5ha (judging from the current distribution of megaliths), and is more
densely agglomerated than other sites in the Besoa valley [Fig. 16.1].
Thus, Pokekea presents itself as a unique site, not only in the appearance of the megaliths, especially
the kalambas and lids, but also in the dissociation from other sites or single kalamba finds. The site lies
slightly detached from other sites in the landscape, but not only with regards to the spatial arrangement
in the valley. Such a large accumulation and density of features cannot be detected at any other site, nor
can a higher appearance of kalambas representing the outstanding block agglomerations be located at
any other area. In addition, some decorative designs of the Pokekea jars are unique to the entire Besoa
valley and even in central Sulawesi — circles and knobs, which belong to the lid decoration, only occur at
Pokekea. Furthermore, the design combination of cup-marks and beams can also be considered as specific
to Pokekea (Kortemeier, in prep.: 51–2). Internal site structures might indicate a focus on the central
cluster of jars [Fig.16.4]. To summarize, Pokekea represents the highest concentration of megalithic
monuments in the Besoa valley, including those which are the most elaborate. While the average number
of known megalithic structures at each Besoa site is 6.2, Pokekea possesses almost ten times this number.
The 60 Pokekea megaliths are part of an arrangement of 147 megaliths in the three central Sulawesian
basins with megaliths. Both this agglomeration of megaliths, as well as the outstanding combination of
megalithic types and decoration elements, may indicate that Pokekea had a central function within the
Besoa valley. Further research might also indicate if we are dealing with a population agglomeration, an
economic and / or a ritual centre in contrast to smaller dispersed sites.
Decoration Patterns and Ritual Architecture
The kalambas differ in arrangement, ornamentation and size [Fig. 16.2]. Of the 21 kalambas at Pokekea,
52% are in a vertical position and ornamented, thus representing the most common situation. 29% of the
jars are vertical without decoration, while 19% represent other variations, like sub-surface jars. Some
kalambas are situated very close to lids. This indicates that the specific lids might belong to a certain
Table 16.2: Radiocarbon dates, Pokekea on-site pollen profiles. Radiocarbon dates were calibrated with
the Calib radiocarbon calibration program based on the dataset for the southern hemisphere
(Stuiver and Reimer 1993; McCormac et al. 2004)
Lab. code Depth Material 14C age, cal BP AD
[cm] BP (mean;2σ-range) (mean;2σ-range)
Jar no. 74 Erl-10584 23 seed 1251 ± 31 1120 (1052-1184) 830 (766-898)
Erl-10585 43–44 bulk 1197 ± 30 1050 (968-1141) 900 (809-982)
Jar no. 27 Erl-11305 32–33 bulk 949 ± 45 740 (678-804) 1120 (1033-1215)
Erl-11306 45–46 bulk 890 ± 39 830 (735-917) 1210 (1146-1272)
Source: W. Kirleis.
Crossing Borders 16.indd 7 6/11/2012 3:48:38 PM
8
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
kalamba, although the 21 jars exceed the number of lids, of which there are only eight (Kortemeier, in
prep.: 25, 32). The tallest jars’ measurements go up to 2m in height and 1.2m in diameter, while the
smallest specimens are about 0.7m in height and 0.5–0.6m in diameter (Kaudern 1938: 50). Different
styles of decoration exist among the ornamented kalambas. Besides two kalambas with cup-marks and
one with human faces, 10 of the 21 jars carry a band design, although the number of bands differs between
the jars. Only three out of the 13 decorated specimens demonstrate a more complex design. One kalamba
shows more than 25 cup-marks along its thickened rim. Another jar is also engraved with cup-marks,
but accompanied by a “radiating beam” decoration. One of the largest kalambas has eight human faces
along its upper lip (Kortemeier, in prep.: 25–6). Out of the eight kalamba lids at Pokekea, three have
decoration, while five are unadorned. Seven lids are located directly in the central cluster of jars and
the five undecorated lids have a central knob of varying size. Out of the three decorated ones, two show
animal carvings, either arranged in a row across the lid, or radiating around the rim [Fig. 16.2]. The design
of these animal carvings resembles the geometric style of the human face designs found on one kalamba
and several stelae. Another lid features a human face decoration similar to the ones found on the largest
kalamba. The animal figures supposedly resemble monkeys, appearing similar to the Black Sulawesi
Macaque (Macaca nigra) which inhabits the entire island. The notion of the animal figures representing
monkeys is also shared by the natives of Sulawesi (Kaudern 1938: 69).
Besides Pokekea with its 60 megalithic finds, 14 other sites comprising 33 megaliths are known from
the Besoa basin. Of these smaller sites, Masora is one of the most remarkable with respect to decoration
and uniqueness [compare Table 16.1]. The Masora site (01°41’54.5’’ S, 120°14’32.0’’E) is also situated
close to the river Lengi. Two large ornamented blocks resemble a facial design, either human or animal.
The two megaliths are arranged as counterparts and the engraved faces appear to look at each other.
Interpretation Model
Without clear ideas about the chronological development of the above-mentioned sites, it is difficult to
evaluate their meaning. As a working hypothesis, we will describe most of them as former settlement
sites, with both the size and the decoration of the megaliths representing the effort which could be
Fig.16.4: Pokekea site, detail map. NW northwest area; C central cluster of jars; SE southeast area
(Drawing: C. Kortemeier / I. Reese).
Crossing Borders 16.indd 8 6/11/2012 3:48:39 PM
9
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
invested by each household unit / lineage into the display of their ancestry (compare e.g. Bloch 1971;
Bradley 2002). Thus, kalambas, stelae and other features display economic and socio-political abilities.
The agglomeration of such megaliths, in Pokekea for the Besoa valley and close to Bewa for the Bada
valley, contrasts to a dispersed distribution of sites with fewer than five megaliths in the above-mentioned
valleys. Observed stylistic differences between sites and whole basins (Dwi Yani Yuniawati 2001: 21;
Kortemeier, in prep.: 52–3) might be due to individual or kinship differences within the population (for
theoretical background, compare Dietler and Herbich 1998; Müller 2006). As an alternate hypothesis, a
higher number of megalithic features might not only represent a larger population agglomeration, but may
also suggest a longer duration of sites. Therefore, larger sites might also be places and landmarks where
traditions are formed and political decisions made. Nevertheless, without better knowledge about the
duration and chronology of the sites, the reconstruction of economic, political, and social developments
remains unclear. This is also true of the environmental preconditions limiting the demographic size within
these valleys. Besides the vague archaeological hints towards a general typo-chronology of the sites,
palaeo-ecological analyses with absolute dating represent a first step towards a better consideration of
chronology and environment.
Environmental Reconstruction from On-Site and Off-Site Pollen Analyses
The On-Site Pollen Analyses at Pokekea
Two short on-site core profiles, taken from the interior of two different kalambas at the Pokekea site, were
evaluated to supplement the archaeological investigations [Fig. 16.2]. Samples were taken and prepared
in the laboratory of the University of Kiel and analyzed following standard laboratory methods (Fægri
et al. 1989). Pollen percentage calculations are based on the terrestrial pollen sum. Depending on pollen
preservation, 150 to 200 terrestrial pollen grains define the total for the calculations conducted using Tilia
2.0 software. Visualization is carried out with Tilia Graph View. Both profiles present sediment in-filling
rather than in-situ natural soil development. From radiocarbon dating on bulk sediment and one seed, a
chronological classification for the cores, but also a terminus ante quem for the kalambas, are determined
[Table 16.2].
Pollen diagrams MP-2a [Fig.16.5] and MP-4a [Fig.16.6] feature similarities in pollen-contributing
taxa and in lithology layers. The color of the lowest parts, which is darker than the upper parts in both
profiles, is notable. Since material from the surrounding areas was filled into the kalambas, the soil
characteristics were also transferred into the jars. The two major soil types of central Sulawesi are
vertisol and ferralsol, soils characteristic for eluviations and alteration processes. The leaching of humic
substances, silicon dioxide and nutrients, leads to secondary quartz sand in the upper soil sections,
enforced by intense rain events. These secondary quartz sand fragments appear in both sediment cores
in the form of mica. Where the kalambas were not covered by lids, further eluviation processes occurred
in the interior of the jars after intense rain events, and also had an influence on the sediment core colors
(Kortemeier, in prep.).
Pollen diagram MP-2a [Fig.16.5] is divided into five pollen assemblage zones (PAZ). Two
radiocarbon dates show an age of around AD 900 for the bottom of the core and AD 830 for the section
with the high amounts of micro-charcoal [Table 16.2]. The lowermost PAZ 74-1 is dominated by Poaceae
including cereals (the latter hidden in the pollen curve labeled Poaceae >37µm). Together, both curves
make up an average of 70% of the terrestrial pollen. Grewia, Evodia, Boehmeria, Castanopsis-type and
Macaranga are the main contributors of tree pollen; furthermore Dodonea, Rutaceae and Acer also occur.
Typha represents the wetland plants, Davallia, the ferns. The assemblages of PAZ 74-2 are similar, but
a slight decrease is observable within the main tree pollen taxa. In PAZ 74-3 Blumeodendron, Ilex and
Rosaceae supplement the tree pollen. Most obvious however, is the jump in the number of micro-charcoals
and non-pollen palynomorphs (NPP) in this PAZ. While the values for NPP and micro-charcoal decline
in the course of PAZ 74-4, the highest diversity in tree pollen and herbs is shown. The uppermost PAZ
74-5 shows the highest tree pollen values of the whole record due to an increase of Macaranga pollen
input; Poaceae, including cereals, reach only about 40% of the terrestrial pollen.
Crossing Borders 16.indd 9 6/11/2012 3:48:39 PM
10
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
0
10
20
30
40
50
60
70
80
90
100
110
Depth [cm]
0
1000
2000
3000
4000
5000
6000
Sample ages [cal B.P.]
9480
1800
590
Dates [cal B.P.]
20 40 60 80 100
Trees and palms
Indeterminate
Varia
Grasslands
Sum Podocarpaceae
20 40 60
Castanopsis-type
Sum Euphorbiaceae
10
Myrica-type
10
Melastomateceae
10
Trema
10
Ilex
20 40 60 80
Sum Poaceae
20
Poaceae >37 µm
312
136 305
310
315
301
335
322
312
304
317
302
297
299
301 341
350 356
311
308
211
Sum terrestrial pollen
20 40 60
Lycopodium cernuum
20 40
111
110
176
95
Charred fragments >10 µm/ml
PAZ
BT-3b
BT-3a
BT-2b
BT-2c
BT-2a
BT-1
Wiebke Kirleis
x 105
Stratigraphy
white silty clay (Munsell chart 5, Y R, 8/1) black silty clay, organic, calcareous (Munsell chart 5, Y R, 2.5/1)
Fig.16.5: Percentage pollen diagram MP-2a, kalamba 74 at Pokekea [see Figure 16.2] (Diagram: C. Kortemeier).
Crossing Borders 16.indd 10 6/11/2012 3:48:40 PM
11
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
Fig.16.6: Percentage pollen diagram MP-4a, kalamba 27 at Pokekea [see Figure16.2] (Diagram: C. Kortemeier).
8
24
40
56
72
88
104
120
136
152
168
184
200
216
232
248
264
280
296
312
328
Depth [cm]
0
200
400
600
800
1000
1200
1400
1600
1800
Sample ages [cal B.P.]
1864
1779
1514
Dates [cal B.P.]
20 40 60 80 100
Trees and shrubs
Indeterminable
Varia
Grasslands
20
Podocarpaceae sum
20
Castanopsis-type
20
Euphorbiaceae sum
8
Myrica-type
8
Melastomataceae
8
Trema
8
Ilex
20 40 60 80
Sum Poaceae
20 40
Poaceae > 37 µm
200
Sum terrestrial pollen
20
Lycopodium cernuum
8
Monochoria
Oogonia of Characeae
Charred fragments of grasses
20 40 60 80
101153
pollen concentration
PAZ
POK-1
POK-2
POK-3
POK-4
POK-5
Wiebke Kirleis
(poor)
.
.
.
.
.
.
.
.
x 103
+++
+++
+++
+++
+++
+++
+++
[grains/ml]
+++
+++
+++
+++
+++
+++
+++
+++
+++
+++
+++
+++
+++
+
+
+
Stratigraphy
Limus detrituosus with mica, Munsell chart 10 YR, 2/1 Turfa herbacea, decomposed, Munsell chart 10 YR, 2/1 Turfa herbacea, poorly decomposed, Munsell chart YR 10, 2/1
Crossing Borders 16.indd 11 6/11/2012 3:48:40 PM
12
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
Pollen diagram MP-4a [Fig.16.6] is divided into four PAZ. Two radiocarbon dates show an age
around AD 1210 for the bottom of the core and AD 1120 for the section with higher amounts of micro-
charcoal [Table 16.2]. In the following section we consider both as an indication for the youngest age of
this kalamba, although the dates are inverse, because the range of dates of the standard deviations overlap.
The lowermost PAZ 27-1 is characterized by Poaceae and cereal values summing up to around 60% of
the terrestrial pollen. Boehmeria, Macaranga and Euphorbia dominate the tree pollen spectrum, which is
completed by Acer and Castanopsis-type. Micro-charcoal values are low (max. 25%). Towards PAZ 27-2,
tree pollen declines; Acer is the most important remaining species with values up to 8%, and all the other
tree species almost disappear. Poaceae and cereals are dominant, spores and NPP increase slightly and
micro-charcoal reached the highest values of the entire record. PAZ 27-3 shows the highest diversity of
tree pollen for this record. Macaranga, Euphorbia, Acer, Castanopsis-type, Rosaceae, Evodia, Mallotus,
Acalypha, Blumeodendron, Grewia and Ilex occur, but Boehmeria is no longer present. Nevertheless,
Poaceae plus cereals dominate the assemblages while the values of micro-charcoal decline. The uppermost
PAZ 27-4 is characterized by tree pollen values reaching 70%; the main contributors are Macaranga,
Evodia, Mallotus and Grewia. Poaceae plus cereals show with about 30%, the lowest values of the whole
record. Micro-charcoal is absent.
Three interpretation models were developed based on three initial assumptions that were tested
through palynological investigations of both on-site pollen profiles. The first assumption is that all bone
remains from the former use of the jars (as burial containers) must have been expelled before new material
was filled in. In both pollen cores, no anthropogenic remained, charred bones or floral grave goods were
found instead. The second assumption is that the kalamba content is anthropogenically filled-in material
from the surrounding areas. The third assumption concerns the cover of the jars; whether a lid covered
the jars or not is uncertain, as wooden or braided plant lids might have existed as well. Thus we assume
that the kalambas were initially closed with a lid. We made three attempts at interpretation based on these
initial statements.
Results and Interpretation: Anthropogenic or Natural Kalamba Fillings?
Model one assumes contemporary anthropogenic filling layers, two filling sections for MP-2a and three
filling sections for MP-4a. The first filling for MP-2a (pollen assemblage zone (PAZ) 74-1 to PAZ 74-
2) comprises pollen of primary lowland forest (Grewia, Boehmeria, Mallotus, Asteraceae liguliflorae),
pollen of montane rainforest (Evodia, Castanopsis-type, Macaranga, Dodonea) as well as pollen of open
landscape areas (Poaceae, Typha). This range of pollen either suggests pollen input from the region by
wind or sediment collection from montane and lowland forest regions as well as the surroundings of the
site. The occurrence of Typha either indicates an influence from the nearby river vegetation or shows the
growth of Typha directly in the kalamba, if the jar was not covered by a lid. Some taxa such as Grewia,
Boehmeria and Dodonea, which possess useful traits, could have been intentionally deposited. The
second filling of MP-2a (PAZ 74-3 to PAZ 74-5) indicates open landscape, forest clearance by fire and
the creation of secondary forest. Noticeable is the appearance of new taxa (Ilex), which were not present
in the first filling section. The uppermost section resembles modern pollen rain and shows a local pollen
signal. In particular, Macaranga, as an indicator for shifting cultivation (Eichhorn 2006), reaches high
values in PAZ 74-5.
The MP-4a pollen record is separated into three anthropogenic filling sections. PAZ 27-1 indicates
species from primary lowland forest as well as those from montane rainforest and open landscape and
thus is similar to the first infilling of the kalamba MP-2a. Boehmeria, Macaranga and Euphorbia show
the highest values among the trees and shrubs. As already mentioned for MP-2a, the material of the first
filling might have been collected from different areas or could be the result of regional pollen input
through wind and insects. Noticeable in PAZ 27-2 is the decline or disappearance of the formerly present
tree taxa, while the values of Poaceae remain high. The material of the second filling could have been
taken from the nearby surrounding area, which explains the high amount of Poaceae and low values
of tree pollen. The values of charred fragments again suggest fire events in the jar’s surroundings. The
diversity of taxa increases in filling section three (PAZ 27-3 to PAZ 27-4); this could be the result of
Crossing Borders 16.indd 12 6/11/2012 3:48:41 PM
13
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
collecting the filling material from different areas again, or might, similar to MP-2a, indicate modern
pollen rain from the immediate vicinity of the jar.
A second interpretative model asserts that the different fillings of the kalambas were not necessarily
connected. This model suggests different fillings related to ritual activities or cultural sacrifices. The
kalambas would have then been emptied completely before they were filled again. The appearance of
single pollen grains of particular species could be a result of earlier filling remains at the bottom or on
the inner walls of the jars. The idea of emptying and refilling the kalambas during rituals, for example,
burial rituals or for a sacrifice, could explain the partly incoherent taxa values. If the kalambas were
used as sacrifice as well as burial containers, special cultural objects and special, meaningful plants were
possibly placed inside the jars.
The third interpretative model assumes that the kalambas were never covered by a lid and that no
anthropogenic filling took place. According to that model, the filling is comprised of pollen input in a
naturally developed soil. This does not exclude the possibility that the jars could have been emptied at
some point. The high amounts of charred fragments are then interpreted as remains from the burning
of surrounding grasslands and paddy fields. The fumes could have been effectual enough to transport
such an amount of charred fragments that makes the arrangement in layers found in the jars noticeable
(Kortemeier, in prep.: 81–3).
Off-Site Pollen Analyses in Besoa Valley
The two off-site cores at 1km and 4km distance from the archaeological site Pokekea provide the first
clue about the environmental development and human impact in the Besoa valley. Results from previous
investigations by two of the authors (Kirleis et al. 2011) are summarized here.
Prehistoric Forest Clearing and Grassland Maintenance in Besoa Valley
The pollen record from the northeastern part of Besoa valley shows the regional vegetation dynamics of
the last 10,000 years [Fig.16.7]. About 2,000 years ago, an abrupt change of the vegetation composition
occurred. Montane rainforests, dominated by Castanopsis (or Lithocarpus, that is included into the
Castanopsis-type), were replaced by grassland vegetation. As is shown by the high input of charred
fragments, the vegetation shift was initiated and maintained through fire clearing. Meanwhile, slightly
higher values of large Poaceae pollen grains and the secondary forest species Trema indicate human
activity in the area. The new vegetation distribution in the valley, mainly open grassland with some
patches of forest vegetation, remains stable. Slightly higher values of indeterminable pollen grains in the
summary diagram (68cm depth) coincide with a drawback in the Poaceae curve. It reflects an effect of bad
pollen preservation, causing a lot of crushed, hardly identifiable Poaceae pollen grains. Thus, the general
tendency of vegetation dynamics shows that, until modern times, periodic burning has been hindering
the recovery of montane rainforests.
Human Impact versus Climate Change
A second off-site pollen record originates from the floodplain of the meandering rivers Lumamba and
Lengi in the vicinity of the megalithic Pokekea site. It completes the picture of the regional pollen record
from the northeastern part of Besoa valley because it covers only the last 2000 years and thus has a
higher time resolution [Fig.16.8]. In general, we have traced a human landscape with open grassland
vegetation and cereal cultivation throughout the last 2,000 years. Small patches of montane rainforest
with Castanopsis or Lithocarpus as main tree species are randomly distributed in the open vegetation
area. However, the most striking feature of this pollen record is the section POK-4, where the sediment
composition is changing from lacustrine to peaty at 100cm depth and the pollen concentrations decrease
to nil due to hydrological variations. The poor pollen concentrations are the result of an elevated oxygen
supply to the peat, due to water-level fluctuations, that furthers decomposition of organic material.
Thus, the peat growth indicates an aridification of the floodplain area and may have been caused by
Crossing Borders 16.indd 13 6/11/2012 3:48:41 PM
14
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
Fig.16.7: Excerpt from the percentage pollen diagram Bariri Tower, BT, 1331m a.s.l., 01°39.247’S, 120°11.364’ E (Diagram: W. Kirleis).
0
5
10
15
20
25
30
35
40
45
Depth in cm
900 (809-982)
830 (766-898)
Radiocarbon dates in A. D.
20 40 60 80 100
Trees and Shrubs
Varia and Indet
Herbs
10
Grewia
1020
Evodia
1020
Boehmeria
10
Castanopsis- Type
20 40
Macaranga
10
Euphorbia
10
Mallotus
10
Dodonea
10
Rutaceae, Type 1
10
Rutaceae, Type 2
10
Acer
10
Acalypha
10
Blumeodendron
10
Phyllocladus
10
Ilex
10
Rosaceae
20 40 60
Poaceae p.p.
20 40 60
Poaceae>37ȝm (incl. cereals)
10
Asteraceae (liguliflorae)
10
Polygonaceae, Type 2
10
Polygonaceae
1020
Typha
10
Varia
10
Type 27
10
Indeterminate
5
Thelypteris
20 40
Davallia
20
Spores p.p.
10
NPP Type 1
10
NPP Type 3
10
NPP Type 5
1020
NPP Type 7
20
NPP p.p.
10
Fungi
50 100 150
Charcoal>10ȝm, absolutcountings
190
190
192
179
183
148
147
146
142
177
170
159
167
168
Terrestrial pollen sum
PAZ
74 - 5
74 - 4
74 - 3
74 - 2
74 - 1
Varia SporesHerbsTrees & Shrubs
Pokekea Profile MP- 2A
Kalamba no. 74
Analyses: C. Kortemeier 2010
Lithology
Crossing Borders 16.indd 14 6/11/2012 3:48:41 PM
15
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
5
10
15
20
25
30
35
40
45
50
Depth in cm
1210 (1146-1272)
1120 (1033-1215)
Radiocarbon dates in A.D.
20 40 60 80 100
Trees and Shrubs
Varia and Indet
Herbs
10203040
Boehmeria
102030
Macaranga
1020
Euphorbia
10
Acer
10
Castanopsis - Type
10
Rosaceae
102030
Evodia
10
Rutaceae, Type 2
10
Mallotus
10
Acalypha
10
Blumeodendron
1020
Grewia
10
Ilex
10
Dodonea
10
Rutaceae, Type 1
20 40
Poaceae p.p.
20 40
Poaceae > 37ȝm (incl. cereals)
10
Asteraceae (liguliflorae)
10
Polygonaceae
10
Type 16
10
Type 26
10
Type 27
10
Type 28
10
Type 29
10
Type 31
1020
Indeterminate
20
Spores p.p.
10
Thelypteris
10
Davallia
10
NPP Type 1
10
NPP Type 2
10
NPP Type 3
10
NPP Type 4
1020
NPP Type 5
10
NPP Type 6
10
NPP Type 7
1020
NPP p. p.
10
Fungi
50 100 150
Charcoal >10 ȝm, absolut countings
217,0
207,0
140,0
208,0
61,0
196,0
200,0
186,0
157,0
163,0
173,0
169,0
173,0
Terrestrial pollen sum
PAZ
27 - 4
27 - 3
27 - 2
27 - 1
SporesVariaHerbsTrees & Shrubs
Pokekea Profile MP- 4A
Kalamba no. 27
Analyses: C. Kortemeier 2010
Lithology
Fig.16.8: Excerpt from the percentage pollen diagram Pokekea, POK, 1,213 m a.s.l., S 01°41.530’, E 120°12.758’ (Diagram: W. Kirleis).
Crossing Borders 16.indd 15 6/11/2012 3:48:41 PM
16
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
either human impact or drought. This drying-up of the floodplain coincides with human activity at the
adjacent archaeological site of Pokekea. Radiocarbon dating applied to a seed from the interior of one
large stone jar [Table 16.2] provides a terminus ante quem of around AD 830 for the site, whereas three
radiocarbon dates on bulk material are accordant, despite being slightly younger. Even the erection of
the monuments may have been contemporaneous with the hydrological changes in the floodplain. The
people at the Pokekea site might have changed the courses of the rivers Lengi and Lumamba to get better
access to the site. However, drought cannot be excluded as a possible trigger of hydrological changes in
the floodplain. The drying could also be the result of a local or a regional to inter-regional phenomenon.
However, variations in past micro-climates are difficult to trace. We have to consider a strong inter-linkage
of climate events and human activity that intensify one another. The time resolution of the regional pollen
record at Bariri Tower is not sufficient to elucidate small-scale changes.
On a macro-regional scale, records of sea surface temperatures (SST) in the western Pacific warm-
pool (WPWP) are available. They show the dynamics of the El Niño Southern Oscillation (ENSO)
and the East Asian monsoon (Oppo et al. 2009; Linsley et al. 2010) that both affect the environment
of central Sulawesi as well as part of the Indonesian archipelago. Measurements of Mg / Ca ratios
and d18O on planktonic foraminifera from the marine core MD9821-60 in the Makassar Strait provide
data that show a 400-year-long period with warm temperatures and high salinities starting around AD
1000. This is equivalent to the northern hemisphere medieval warm period (Newton et al. 2006). This
evidence is supported by measurements of d15N ratios from the marine core BJ8-03-102GGC from Kau
Bay, Halmahera, where a basin stagnation was observed during the period AD 1000–250 that indicates
less ENSO-like conditions (Langton et al. 2008). Evidence for a short-term drought period from ca. AD
1275-325 is shown by geochemical analyses on a laminated sediment core from East Java (Crausbay et
al. 2006). However, this drought period may not necessarily have affected Sulawesi because of micro-
regional differences. Furthermore, compared to the Pokekea record, the dating is slightly too late and
older material was not available from the Java sediment core. More convincing is the correlation between
stratigraphical change towards peat with good pollen preservation due to wetter conditions at 56cm core
depth and the general cooling trend of the Little Ice Age (LIA) beginning around AD 1400. Again it is
the foraminifera record of the Makassar Strait marine core MD9821-60 (Newton et al. 2006) that shows
evidence for increased precipitation south of the equator during the LIA (AD 1400–850). Additional
evidence for wetter climate during the LIA is available from El Junco Lake on the Galapagos Island
San Cristobal (Sachs et al. 2009). This is in contrast to observations for the (sub-)tropical regions north
of the equator, where increased aridity is indicated during the LIA (Wang et al. 1999; Watanabe et al.
2001; Haug et al. 2001; Sachs et al. 2009). The contrasting latitudinal precipitation / aridity pattern is
interpreted as being indicative of a pronounced and rapid southwards displacement of the inter-tropical
convergence zone (ITCZ) that reached its southernmost position during the LIA (Newton et al. 2006;
Sachs et al. 2009). Direct evidence for drivers of the hydrological change in the catchment of the rivers
Lengi and Lumamba in the Besoa Valley around AD 1000 is still missing. Future investigations have to
show whether human-induced change of the river course or drought have to be taken into consideration
as the main trigger of the hydrological change.
Bringing Together of On-Site and Off-Site Data at Pokekea
The environment of the Pokekea site offers the rare opportunity to combine palaeo-ecological on-site and
off-site analyses. The three pollen profiles POK, MP-2A and MP-4A [Figs. 16.6–16.8] are available to
evaluate human activity in the vicinity of the site.
For better comparison of the three profiles, we assume that, in general, the vegetation development
shown in POK has been stable throughout the last 2000 years. Obvious changes in the record occur
only due to hydrological variation around 1000 years ago that have resulted in poor pollen preservation
(PAZ-4), but in the following POK-5, pollen composition resembles the previous PAZ (POK-1 to POK-
3). Thus, in order to directly compare the records, which have different time depths, it was necessary to
compress the data from the off-site pollen diagram. Presence-absence data of identifiable taxa were used
for comparison, since this remained stable over the entire time-depth of the pollen diagram. Therefore, it
Crossing Borders 16.indd 16 6/11/2012 3:48:41 PM
17
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
32
9
3
10
0
5
10
15
20
25
30
35
pollen records: POK (off-site), Kalamba 27 and Kalamba 74
n taxa out of 54
off-site only all sites Kalamba 74 and off-site Kalambas only
Fig. 16.9: Presence and distribution of pollen from 54 taxa in the on-site and off-site pollen profiles
at Pokekea (Diagram: W. Kirleis).
can be treated as analogous to the information in the on-site profiles from the kalambas. The comparison
of 54 taxa in all three records is shown in Figures 16.9 and 16.10. It is noticeable that 32 out of the 54
taxa only occur in the off-site profile and just nine are shown in all three records. However, the matching
nine taxa include the main contributors to the pollen assemblages from the off-site pollen profile (like
cereals or other Poaceae (named as Poaceae p.p., with p.p. = pro parte), and the Castanopsis-type that
grow in the Besoa basin) and thus show the inter-connectivity of the sediments in the stone jars and the
surroundings. Useful plants are included in those taxa occurring only in the kalambas; for example, the
fibre plant Boehmeria, and Dodonea, of which the timber is used and red dye is produced from the fruit.
Other plants can be interpreted as floral grave goods, such as the perfume plant Evodia with its leaves
rich in essential citrus oil.
To conclude, the pollen assemblages from the sediment infillings of the kalambas show the
“background noise” of the past local to regional scale vegetation development and in addition, hint to a
deposition of particular plants in the jars.
Conclusions
Following the results of the radiocarbon dating of the two short on-site core profiles from the interior of
the Pokekea kalambas (AD 766–898 earliest and AD 1146–272 latest [Table 16.2]), a terminus ante quem
for the kalamba erections is given. Therefore the jars could date to the time period of Island Southeast
Asia when metallurgy had already commenced, and they belong to the metal ages of Southeast Asia
(Bellwood 1985: 313; Glover 1998 / 1999). It cannot yet be determined if the erection of the megaliths
can be linked with the first forest clearance about 2000 years ago. However, in general, the Sulawesi
megaliths are connected with the realm of death in the sense of building a monument for deceased
ancestors (Dwi Yani Yuniawati 2001: 3, 20).
Regarding the spatial arrangement, the central character of some sites and the meaning of the size
and extent of decoration of the megaliths, our initial investigations show that in each central Sulawesian
basin, an agglomeration of megaliths (e.g. Pokekea) exists in contrast to the dispersed pattern of single or
small groups of megaliths. We identify them as central domestic sites with ritual activities and probably
Crossing Borders 16.indd 17 6/11/2012 3:48:42 PM
18
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
an intra-site spatial organization, which could
be the focus of further research.
Concerning the decoration patterns, band
designs are characteristic of the Pokekea jars
alone. The megalithic stones and the lid designs
represent the local fauna, in most cases monkeys
(Kortemeier, in prep.: 89). Thus, investment in
ornamentation was used to display the local
environment on monuments. The investment in
energy, in terms of their large size (highly vis-
ible even from a distance) and elaborate decora-
tion of kalambas probably displayed the power
of Pokekea inhabitants within the basin. The
material from the inside of the jars is possibly
an anthropogenic filling, collected for cultural
reasons; natural in-filling of the sediments is
less probable. Intentional deposition of plants
into the jars is a reasonable conclusion from the
palynological evidence. Whether the kalambas
were covered by lids cannot yet be determined;
wooden discs or lids of braided plants might
have existed and would have decomposed.
Parallels exist between the pollen assem-
blages in the kalamba interior and the Pokekea
off-site pollen profile, such as high values
of Poaceae and cereals in all three profiles.
Euphorbiaceae, Castanopsis-type pollen and
pollen of Ilex appear in all profiles, but with
different values and in different time periods.
One congruent factor among all profiles is
the similar percentage (average 50–60%) of
grasslands and herbs as compared to trees
(Kirleis et al. 2011; Kortemeier, in prep.:
70–80). Furthermore, the off-site profiles show
that human activity in the region, for about
2000 years, may be related to the megalithic
sites. Changes in the peat composition of one
profile further hint at climate deterioration
towards the LIA, with the latter perhaps linked
to the southwards shift of the ITCZ that leads to
increased precipitation south of the equator.
This study has produced a preliminary
chronological framework for the kalambas, as
well as positing various hypotheses regarding
their use and socio-economic relevance. Ad-
ditionally, the palaeo-environmental data has
shown clear evidence of human influence in
the region. As such, these initial investigations
on the megalithic landscape of central Sulawesi
show a high potential for detailed studies aimed
towards a deeper understanding of human-
environment interactions in the region. Further
Acer
Evodia
Blumeodendron
Grew ia
Mallotus
Rosaceae
Dodonea
Boehmeria
Thelypteris
Polygonaceae
Typha
Davallia
Phyllocladus
Castanopsis-Typ
Macaranga
Euphorbia
Poaceae p. p.
Cereals indeterminata
Asteraceae, liguliflorae
Ilex
Acalypha
Rutaceae p. p.
Dacrycarpus
Melastomataceae
Trema
Alchornea
Bischofia
Erythrina
Elaeocarpaceae
Moraceae/Urticaceae p. p.
Syzygium
Lycianthes
cf . Aphanante
cf . Araliaceae
cf . Combretaceae
cf . Sapotaceae
Sterculiaceae
cf . Myricaceae
Palms
Asteraceae tubulif lorae
Brassicac eae p. p.
cf . Apiaceae
cf . Ranunculaceae
Chenopodiaceae-Amaranthaceae
Lamiaceae p. p.
Cyperaceae
cf . Equisetum
Eichhornia-type
Cyatheaceae sum
cf . Nephrolepis
Lycopodiaceae
Lycopodium cernuum
Anthoceros
Hymenophyllum
Kalamba 27
Kalamba 74
POK, off -site
Fig. 16. 10: List of 54 taxa occuring in the on-site and off-
site pollen profiles at Pokekea based on presence-absence
data (Diagram: W. Kirleis).
Crossing Borders 16.indd 18 6/11/2012 3:48:42 PM
19
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
research combining the two lines of evidence will also provide improved insights into the specific role
the kalambas played for early Sulawesian societies.
Acknowledgements
Special thanks go to Mr. Iksam, State Museum Central Sulawesi, Palu, Indonesia, for his introduction
to the megalithic sites in central Sulawesi. Research was partly carried out as part of subproject C7 of
the Collaborative Research Centre “Stability of Rainforest Margins” (STORMA). In this context, we
gratefully acknowledge the logistic support from STORMA’s Indonesian partner universities in Bogor
and Palu, Institut Pertanian Bogor (IPB) and Universitas Tadulako (UNTAD), the Ministry of Education
in Jakarta (DIKTI) and the Indonesian Institute of Sciences (LIPI). Upgrading our microfossil reference
collection on Indonesian species and its integration into the online Australasian Pollen and Spores Atlas
(APSA) was supported by an EU Synthesis travel grant, NL-TAF-3869, and two grants by the ARC
Environmental Futures Network, Australia. We are grateful to Nicole Taylor, Kiel, for her linguistic input
and to Ines Reese, Kiel, for improving the quality of the figures and tables.
Note
1. In contrast to the widespread use of the term “megaliths” for Sulawesian standing stones, kalambas and boulder ag-
glomerations, some of these would be called only “sub-megalithic” in European terms. These sub-megalithic features
were constructed of boulders less than 1m in diameter or length, while boulders usually are labeled “megalithic” if
their dimensions extend beyond the aforementioned standard. For simplicity, we use the term “megalith” for mega-
lithic and sub-megalithic structures as is common in most of the literature about this Sulawesian phenomenon.
References
Adriani, N. and Kruyt, A.C. Von Poso naar Parigi, Sigi en Lindoe. Tijdschrift voor Zendingswetenschap 42 (1898):
369–535.
Anshari, G., Kershaw, A.P. and Kaars, S. van der. A late Pleistocene and Holocene pollen and charcoal record from peat
swamp forest, Lake Sentarum Wildlife Reserve, West Kalimantan, Indonesia. Palaeogeography, Palaeoclimatology,
Palaeoecology 171 (2001): 213–28.
Bellwood, P. Man’s conquest of the pacific. The prehistory of Southeast Asia and Oceania. New York: Oxford University
Press, 1979.
______. Prehistory of the Indo-Malaysian Archipelago. Sydney: ANU E-Press, 1985.
Bintarti, D.D. More on urn burials in Indonesia. Indo-Pacific Prehistory Association Bulletin 19. Melaka Paper 3 (2000):
73–5.
Bloch, M. Tombs, ancestral villages and kinship organization in Madagascar. London / New York: Seminar Press, 1971.
Bradley, R. The Past in Prehistoric Societies. London: Routledge. 2002.
Cook, E., Haberzettl, T., Kaars, S. van der, Kirleis, W., Gutknecht, B., Ir. Iskandar and Behling, H. High resolution multi-
proxi records of palaeoenvironmental change through the last millennium from Lakes Lindu and Kalimpaa, Central
Sulawesi, Indonesia. In Tropical Rainforests and Agroforests under Global Change, eds. M. Grosse, W. Lorenz,
Surya Tarigan, Adam Malik. Proceedings International Symposium, 5–9 Oct. 2008, Kuta, Bali, Indonesia. Reihe der
Universitätsdrucke, Göttingen: Universitätsverlag Göttingen, 2008a, p. 104.
Cook, E., Kirleis, W. and Behling, H. 122: A high resolution multi-proxi record of palaeoenvironmental change through
the last millennium at Lake Lindu, Central Sulawesi, Indonesia. Terra Nostra 2008 / 2, IPC-XII / IOPC-VIII Bonn,
Germany 2008, Berlin: Geo-Union der Alfred-Wegener-Stiftung (2008b): 53.
Crausbay, S., Russell, J. and Schnurrenberger, D. A ca. 800-year lithologic record of drought from sub-annually laminated
lake sediment, East Java. Journal of Paleolimnology 35 (2006): 641–59.
Dam, R.A.C., Fluin, J., Papay Suparan and Kaars, S. van der. Palaeo-environmental developments in the Lake Tondano
area (N. Sulawesi, Indonesia) since 33.000 yr BP. Palaeogeography, Palaeoclimatology, Palaeoecology 171 (2001):
147–83.
Dietler, M. and Herbich, I. Habitus, Technique, Style: An Integrated Approach to the Social Understanding of Material
Culture and Boundaries. In The Archaeology of Social Boundaries, ed. M.T. Stark. Washington / London: Smithsonian
Institute Press, 1998, pp. 232–63.
Crossing Borders 16.indd 19 6/11/2012 3:48:42 PM
20
WIEBKE KIRLEIS, JOHANNES MÜLLER, CORINNA KORTEMEIER, HERMANN BEHLING AND SANTOSO SOEGHONDO
Dwi Yani Yuniawati. Stone Burials, one of the Megalithic Remains in Sulawesi. Aspects of Indonesian Archaeology 26
(2001): 1–33.
Dwi Yani Yuniawati, Bagyo Prasetyo, Retno Handini and Romania, L.G. Penelitan situs-situs megalitik di kawasan Lembah
Besoa, Kecamatan Lore Tengah Kabupaten Poso, Provinsi Sulawesi Tengah [Research on megalithic sites in the
region of Lembah Besoa, District Lore Tengah, Regency Poso, Province central Sulawesi]. Jakarta: PUSLIT Proyek
Penelitian dan Pengembangan Arkeologi 2004 [unpublished].
Eichhorn, K. Plant diversity after rain-forest fires in Borneo. Journal of plant taxonomy and plant geography. Blumea
supplement 18. Leiden: National Herbarium Nederland, Universiteit Leiden branch, 2006.
Eko Yulianto, Rahardjo, A.T., Dardji Noeradi, Siregar, D.A. and Hirakawa, K.A. Holocene pollen record of vegetation
and coastal environmental changes in the coastal swamp forest at Batulicin, South Kalimantan, Indonesia. Journal
of Asian Earth Science 25 (2005): 1–8.
Fægri, K., Iversen, J., Kaland, P.E. and Krzywinski, K. Textbook of pollen analysis, 4th ed., Chichester / New York: John
Wiley & Sons, 1989.
Flenley, J.R. Palynological evidence for land use changes in South-East Asia. Journal of Biogeography 15 (1988):
185–97.
Flenley, J.R. and Butler, K. Evidence for continued disturbance of upland rain forest in Sumatra for the last 7000 Years of
an 11,000 year record. Palaeogeography, Palaeoclimatology, Palaeocology 171 (2001): 289–305.
Glover, I.C. The archaeological past of island Southeast Asia. In Message in stones. Statues and sculptures from tribal
Indonesia in the collections of the Barbier-Mueller Museum, ed. J.P. Barbier. Geneva: Musée Barbier-Mueller /
Milano: Skira, 1998 / 99, pp. 17–34.
Gremmen, W. H. E. Palynological investigations in the Danau Tempe depression, southwest Sulawesi (Celebes), Indonesia.
Modern Quaternary Research in Southeast Asia 11 (1990): 123–34.
Guide Lore Lindu. Guide Lore Lindu. Directorate General of Forest Protection and Nature Conservation. Lore Lindu
National Park Central Sulawesi. A visitor’s guide 2001. Sulawesi, 2001 [unpublished].
Haberle, S., Hope, G.S. and Defretes, Y. Environmental changes in the Baliem valley, montane Irian Jaya, Republic of
Indonesia. Journal of Biogeography 18 (1991): 25–40.
Haberzettl, T., St-Ongea, G., Behling, H. and Kirleis, W. Improving late Holocene radiocarbon-based chronologies by
matching paleomagnetic secular variations to geomagnetic field models – an example from Lake Kalimpaa (Sulawesi,
Indonesia). In Magnetostratigraphy: not only a dating tool, ed. Geological Society of London.
Haris Sukendar. Laporan Penelitian Kepurbakalaan di Sulawesi Tengah [Research Report on the prehistory of Central
Sulawesi]. Berita Penelitian Arkeologi 25. Jakarta: Proyek Penelitian dan Penggalian Purbakala, Departemen P &
K, 1980a.
______. Discription on the megalithic tradition of Indonesia. Berkala Arkeologi 8 (1), Yogyakarta: Balai Arkeologi
Yogyakarta (1980b): 1–30.
Haug, G.H., Hughen, K.A., Sigman, D.M., Peterson, L.C. and Roehl, U. Southwards migration of the intertropical
convergence zone through the Holocene. Science 293, 1 (2001): 304–8.
Heine-Geldern, R. Die Megalithen Südostasiens und ihre Bedeutung für die Klärung der Megalithenfrage in Europa und
Polynesien. Special printing from Anthropos 23 (1928): 276–315.
Hope, G.S. The vegetational changes of the last 20,000 years at Telefomin, Papua New Guinea. Singapore Journal of
Tropical Geography 4 (1983): 25–33.
______. Environmental change in the Late Pleistocene and later Holocene at Wanda site, Soroako, South Sulawesi,
Indonesia. Palaeogeography, Palaeoclimatology, Palaeoecology 171 (2001): 129–45.
Hunt, C.O. and Rushworth, G. Holocene human impact in wet tropical lowland forest: Evidence from Niah, Sarawak,
Malaysian Borneo. Quaternary Research 64 (2005): 460–8.
Kaars, S. van der. Palynology of eastern Indonesian marine piston-cores: A late quaternary vegetational and climatic record
for Australasia. Palaeogeography, Palaeoclimatology, Palaeoecology 85 (1991): 239–302.
______. Marine and terrestrial pollen records of the last glacial cycle from the Indonesian region: Bandung basin and Banda
Sea. Palaeoclim Data Model 3 (1998): 209–19.
Kaars, S. van der, Wang, X., Kershaw, A.P., Guichard, F. and Setiabudi, D.A. A late Quaternary palaeoecological record
from the Banda Sea, Indonesia: Patterns of vegetation, climate and biomass burning in Indonesia and northern
Australia. Palaeogeography, Palaeoclimatology, Palaeoecology 155 (2000): 135–53.
Kaars, C. van der, Penny, D., Tibby, J., Fluin, J., Dam, R.C.A. and Suparan, P. Late Quaternary palaeoecology, palynology
and palaeolimnology of a tropical lowland swamp: Rawa Danau, West-Java, Indonesia. Palaeogeography,
Palaeoclimatology, Palaeoecology 171 (2001): 185–212.
Kaars, S. van der and Bergh, van den G.D. Anthropogenic changes in the landscapes of west Java (Indonesia) during
historic times, inferred from a sediment and pollen record from Teluk Banten. Journal of Quaternary Science 19, 3
(2004): 229–39.
Crossing Borders 16.indd 20 6/11/2012 3:48:42 PM
21
THE MEGALITHIC LANDSCAPE OF CENTRAL SULAWESI, INDONESIA
Kaars, S. van der, Tapper, N. and Cook, E.J. Observed relationships between El Niño-Southern Oscillation, rainfall
variability and vegetation and fire history on Halmahera, Maluku, Indonesia. Global Change Biology 16, 1 (2010):
705–14.
Kaudern, W. Megalithic Finds in Central Celebes. Ethnographical Studies in Celebes. Results of the author’s expedition
to Celebes 1917–20. Göteborg: Elanders Boktryckeri Aktiebolag, 1938.
Kirleis, W., Pillar, V.D. and Behling, H. Human-environment interactions in mountain rainforests: Palaeobotanical evidence
from Central Sulawesi, Indonesia. Vegetation History and Archaeobotany 20 (2011): 165–79. DOI: 10.1007 / s00334-
010-0272-0.
Kortemeier, C. Tracing jars. The megalithic landscape of Central Sulawesi. An archaeological overview and palynological
investigations, in prep.
Kruyt, A.C. Nadere gegevens betreffende de oudheden aangetroffen in het landschap Besoa. Tijdschrift Koniglijk
Nederlands Aardrijksk. Genootschap 26, Leiden, 1908.
Langton, S.J., Linsley, B.K., Robinson, R.S., Rosenthal, Y., Oppo, D.W., Eglinton, T. I., Howe, S.S., Djajadihardja, Y.S. and
Syamsudin,F.3500yrrecordof centennial-scaleclimate variabilityfrom theWesternPacificWarmPool.Geology
36 (2008): 795–8.
Linsley,B.K.,Rosenthal,Y.and Oppo,D.W.Holoceneevolution oftheIndonesian throughflow andthewestern Pacific
warm pool. Nature Geoscience 3 (2010): 578–83.
Maloney, B.K. Grass pollen and the origins of rice agriculture in North Sumatra. Modern Quaternary Research in Southeast
Asia 11 (1990): 135–61.
Morley, R.J. A palaeoecological interpretation of a 10,000 year pollen record from Danau Padang, Central Sumatra,
Indonesia. Journal of Biogeography 9 (1982): 151–90.
Müller, J. Soziale Grenzen und die Frage räumlicher Identitätsgruppen in der Prähistorie. In Soziale Gruppen – kulturelle
Grenzen. Die Interpretation sozialer Grenzen in der Prähistorischen Archäologie, eds. S. Burmeister and N. Müller-
Scheessel. Münster: Waxmann, 2006, pp. 101–17.
Newton,A.,Thunell,R.andStott,L.Climate andhydrographicvariabilityintheIndo-PacificWarmPoolduringthelast
millennium. Geophysical Research Letters 33 (2006): L19710. doi:10.1029/2006GL027234.
Oppo, D.W., Rosenthal, Y. and Linsley, B.K. 2,000-year-long temperature and hydrology reconstructions from the Indo-
Pacificwarmpool.Nature 460, 1 (2009): 113–6.
Sachs, J.P., Sachse, D., Smittenberg, R.H., Zhang, Z., Battisti, D.S. and Golubic, S. Southward movement of the Pacific
intertropical convergence zone AD 1400–1850. Nature Geoscience 28 Jun. 2009, DOI: 10.1038 / NGEO554.
Stuijts, I.L. Late Pleistocene and Holocene vegetation of West Java, Indonesia. Rotterdam: A. Balkema, 1993.
Suparan, P., Dam, R.A.C., Kaars, S. van der and Wong, T.E. Late Quaternary tropical lowland environments on Halmahera,
Indonesia. Palaeogeography, Palaeoclimatology, Palaeoecology 171 (2001): 229–58.
Wang, X., Kaars, S. van der, Kershaw, A.P., Bird, M. and Jansen, F. A record of vegetation and climate through the last three
glacial cycles from Lombok Ridge core G6-4, eastern Indian Ocean, Indonesia. Palaeogeography, Palaeoclimatology,
Palaeoecology 147 (1999): 241–56.
Watanabe, T., Winter, A. and Oba, T. Seasonal changes in sea surface temperature and salinity during the Little Ice Age in
the Caribbean Sea deduced from Mg / Ca and 18O / 16O ratios. Marine Geology 173 (2001): 21–36.
Whitten, T., Muslimin Mustafa and Henderson, G.S. The Ecology of Sulawesi. The Ecology of Indonesia Series 4. Jakarta:
Periplus Editions, 2002.
Crossing Borders 16.indd 21 6/11/2012 3:48:42 PM
