581 Full

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/31271935
Origin of Geochemical Variability by Arc-Continent Collision in the Biru Area,

Southern Sulawesi (Indonesia)
Article in Journal of Petrology · April 2002

DOI: 10.1093/petrology/43.4.581 · Source: OAI
CITATIONS READS
53 435
4 authors, including:
Marlina Elburg Theo van Leeuwen

University of Johannesburg Independant
192 PUBLICATIONS 6,354 CITATIONS 40 PUBLICATIONS 1,103 CITATIONS
SEE PROFILE SEE PROFILE
John Foden
University of Adelaide
181 PUBLICATIONS 8,543 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Mineralogy, geochemistry, and isotope geology of the Pilanesberg Complex and associated dykes View project
Formation, Evolution, Basin Analysis and Metallogeny of the northern East African Orogen (including the Arabian Nubian Shield) View project
All content following this page was uploaded by Marlina Elburg on 10 June 2014.
The user has requested enhancement of the downloaded file.

JOURNAL OF PETROLOGY VOLUME 43 NUMBER 4 PAGES 581–606 2002
Origin of Geochemical Variability by

Arc–Continent Collision in the Biru Area,
Southern Sulawesi (Indonesia)
MARLINA A. ELBURG1∗, THEO VAN LEEUWEN2, JOHN FODEN1

AND MUHARDJO1
1
DEPARTMENT OF GEOLOGY AND GEOPHYSICS, ADELAIDE UNIVERSITY, ADELAIDE, S.A. 5005, AUSTRALIA
2
RIO TINTO EXPLORATION, JAKARTA, INDONESIA
RECEIVED AUGUST 31, 2000; REVISED TYPESCRIPT ACCEPTED SEPTEMBER 28, 2001
Analyses of igneous rocks from the Eocene calc-alkaline and Miocene geochemistry can be modelled by varying contributions
potassic volcanic arc in southwest Sulawesi indicate that magmas from the mantle wedge, fluid from the subducted slab,
became more heterogeneous in their trace element and Pb–Sr–Nd and the presence or absence of a subducted sedimentary
isotopic signature following the collision of the Buton microcontinent component (White & Patchett, 1984; Gill & Williams,
with the arc at >15 Ma. Isotopic ratios become more ‘continental’ 1990; Woodhead & Johnson, 1993; Kepezhinskas et al.,
4 my after the collisional event ( 87Sr/86Sr Ζ0·7085, 143Nd/ 1997; Peate et al., 1997; Alves et al., 1999; Hoogewerff,
144
Nd [0·5125, 206Pb/204Pb Ζ 19·2, 207Pb/204Pb Ζ 15·73, 1999; Pearce et al., 1999; Eiler et al., 2000). Despite
208
Pb/204Pb Ζ39·4). As the overriding plate consists of young agreement on the importance of these components, there
Sundaland crust, whereas the subducted sediment is likely to is still some dispute about the importance of an ocean
have been shed from a compositionally distinct microcontinent of island basalt (OIB) component in the wedge (Stolz et al.,
Australian derivation, we can be certain that the continental isotopic 1990; Edwards et al., 1991) and the implied role of
signature reflects subduction of continental material rather than residual rutile (Ryerson & Watson, 1987; Foley et al.,
crustal contamination. The isotopic compositions of the magmas 2000). If an arc is built upon continental crust, or
can be explained by the melting of a mixed mantle wedge, consisting overlain by a thick layer of sediment, there is a potential
of fluid-fluxed and sediment-modified MORB mantle. In this model, contribution from the upper plate via assimilation–
the maximum amount of sediment added to the mantle source is fractional crystallization (AFC) processes, which can
10%. sometimes be difficult to distinguish from the signature
of subducted sediment (Davidson, 1987; Thirlwall et al.,
1996; Gasparon & Varne, 1998). The only unambiguous
way to distinguish between subducted sediment and
KEY WORDS: arc volcanism; continental collision; Indonesia; radiogenic crustal assimilation is offered by arcs where the isotopic
isotopes; subduction composition of the upper plate is very different from that
of the subducted sediment, or where lateral variation is
clearly controlled by crustal structures.
Additional problems in identifying source components
INTRODUCTION in magmatism are encountered in collisional situations,
Our understanding of the sources contributing to arc where the potential exists for subduction of the con-
magmatism has been greatly increased by detailed trace tinental crust, and addition of material to the region of
element and isotope analysis of subduction-related rocks. magma generation from that source (Hilton et al., 1992;
There is a broad consensus that much of island-arc Van Bergen et al., 1993).
∗Corresponding author. E-mail: marlina.elburg@adelaide.edu.au  Oxford University Press 2002

JOURNAL OF PETROLOGY VOLUME 43 NUMBER 4 APRIL 2002
The extinct arc of western Sulawesi (Brouwer, 1947; the Late Eocene, producing a thick pile of andesitic
Sukamto, 1975; Katili, 1978) offers us the chance of volcanics named the Langi Volcanics (Van Leeuwen,
improving our knowledge of source components in col- 1981). They formed as a response to west-dipping sub-
lisional magmatism by studying the geochemistry of pre- duction (Van Leeuwen, 1981; Yuwono et al., 1985).
to syn-collisional deposits and identifying changes in The upper parts of the Malawa Formation and Langi
magma geochemistry. This study follows on from the Volcanics interdigitate with the Tonasa Limestone For-
work by Elburg & Foden (1999a) in which the studied mation, a sequence of up to 1100 m thickness of platform
collection of samples was divided into pre-, syn- and carbonates, marls and redeposited carbonates of Middle
post-collisional groups, based on K/Ar age and the Eocene to early Middle Miocene age (Van Leeuwen,
perceived timing of the collision between the arc and the 1981; Sukamto, 1982; Wilson, 2000; Wilson & Bosence,
microcontinent of Buton. The present study concerns a 1996). Towards the end of the Early Miocene, magmatic
stratigraphically well-defined suite of samples from the activity of calc-alkaline affinity resumed (Yuwono, 1987;
Biru area (Van Leeuwen, 1981) (Fig. 1), belonging to Yuwono et al. 1988). However, it was soon succeeded by
the Eocene (50 Ma) calc-alkaline arc and the Miocene potassic volcanism in the early Middle Miocene, the
(15–6·3 Ma) potassic volcanic arc. The advantage of onset of which coincided with an Early to Middle Miocene
knowing the stratigraphic order of the deposits is that tectonic event involving a phase of faulting that resulted
we are not solely reliant on K/Ar dating, which cannot in reactivation of earlier faults, and localized tilting and/
always distinguish between sample suites with similar or subaerial exposure of fault blocks (Van Leeuwen,
ages. 1981; Wilson & Bosence, 1996; Wilson, 2000). At Biru,
this tectonic event can be placed at >14–15 Ma on the
basis of palaeontological evidence (Van Leeuwen, 1981)
and is likely to have been caused by the collision between
GEOLOGICAL SETTING western Sulawesi and the northern margin of Australia
As the tectonic history of the area has already been (with subsequent left-lateral translation of the entire col-
discussed by Elburg & Foden (1999a), we will give only lision complex; Daly et al., 1991; Charlton, 2000), or
a brief overview, and identify where our interpretation independently moving terranes that had been detached
has changed since the previous publication. from Australia during Mesozoic rifting (e.g. Pigram &
The oldest exposed rocks are represented by two Panggabean, 1984; Smith & Silver, 1991). The collision,
inliers of tectonically stacked metamorphic, ultrabasic which consisted of a series of highly complex, as yet
and sedimentary lithologies of Jurassic to mid-Cretaceous poorly understood events, was initiated during the Late
age (Sukamto, 1982, 1986; Wakita et al., 1996). They Oligocene–Early Miocene and continues to the present
are overlain with an angular unconformity by deep day (Hall & Wilson, 2000). The collision event probably
marine flysch deposits of the Upper Cretaceous Ba- involved Buton, located in the southern part of the
langbaru Formation (Hasan, 1991). A contemporaneous collision complex (Fig. 1), either as a small separate
distal flysch sequence (Marada Formation; Van Leeuwen, microcontinent (Smith & Silver, 1991; Milsom et al.,
1981) occurs further to the east. These Mesozoic form- 1999) or forming part of a larger fragment (Milsom et al.,
ations are interpreted to have been formed in a fore-arc 2000). Buton docked with Western Sulawesi >15–13 my
basin on a trench slope (Van Leeuwen, 1981; Hasan, ago, as evidenced by a regional unconformity on the
1991), related to a west-dipping subduction zone (Ham- island (Davidson, 1991). This is slightly older than the
ilton, 1979; Parkinson, 1991) that generated magmatism age of 13–11 Ma we previously assigned to the ‘Buton
in south–central Kalimantan (Parkinson et al., 1998). collision’ (Elburg & Foden, 1999a). A sketch map of the
This pattern of subduction ended when a continental tectonic situation just before the collision as envisaged
fragment was thrust underneath Sulawesi at by Hall (1996) is shown in Fig. 2. Potassic magmatism
>115–120 Ma. Remnants of this underthrust plate are continued well after this collisional event, with the young-
preserved as ultra-high-pressure metamorphic complexes est deposits in the Biru area of 6·3 Ma. Further south,
in SW, central and SE Sulawesi (Parkinson et al., 1998). the Lompobatang Volcano was active as recently as
Subduction started to affect SW Sulawesi again in the 1·8 Ma (Polvé et al., 1997; Elburg & Foden, 1999a),
Paleocene. At Bantimala this produced a series of basaltic apparently unrelated to subduction. The tectonic history
to rhyolitic rocks and associated intrusions, named the of the area is further complicated by a second collision,
Alla or Bua Formation (Sukamto, 1986; Yuwono, 1987). between Buton and the submerged Tukang Besi platform,
These grade upwards into an Early to Middle Eocene in the Pliocene around 2–3 Ma (Fortuin et al., 1990; Ali
sequence of marginal marine, coal-bearing siliciclastic et al., 1996). It is unclear whether subduction occurred
rocks of the Malawa Formation (Sukamto, 1982). underneath Buton between 15 and 3 Ma, or whether
In the Biru area, magmatic activity started during the westward movement of the Tukang Besi platform was
Middle Eocene, or possibly earlier, and continued into accommodated by strike-slip faulting. As no record exists
582
ELBURG et al. ARC–CONTINENT COLLISION, BIRU, SULAWESI
Fig. 1. Geological map of the Biru sample area in South Sulawesi (a), with insets showing the location of South Sulawesi (b) and the Biru
area (c).
of subduction-related volcanism in this area around this opportunity to clearly distinguish between upper-crustal
time, the latter alternative seems more likely. contamination and subduction of the microcontinent or
The tectonic history of the area therefore indicates sedimentary material derived from it into the source of
that South Sulawesi is part of Sundaland and consists of magmatism during continental collision.
continental crust that was cratonized in the Mesozoic
(Hamilton, 1979), whereas the colliding microcontinent
of Buton is derived from the much older Australia–New
Guinea continent (Daly et al., 1987). Therefore, the Geology of the Biru area
isotopic signature of the upper plate and the collider are The geology of the Biru area has been described by Van
likely to be very different, and this situation gives us the Leeuwen (1981) and is shown in Fig. 1. The oldest
583
the lower member of the Camba Formation, which crops

out in the northern part of the Biru unit. This unit
contains a middle to late Middle Miocene nannoplankton
fauna (Hasibuan, 1996). Higher in the sequence con-
glomerates are found containing material derived from
the late Middle Miocene Sopo Volcanics (see below) and
older formations.
The potassic volcanic rocks can be divided into six
units (Table 1), each having distinct lithologic, petro-
logical (Table 2) and geochemical (Table 4, see below)
characteristics. These are the Sopo Volcanics (SV; al-
ternating basalt to basaltic andesite lavas and vol-
caniclastics, and associated dykes), Marara Ignimbrite
(MI; pyroclastic flow deposits of dacitic composition),
Bila Volcanics (BV; polymict volcaniclastics with in-
tercalations of leucite-bearing tephrite lava flows in the
middle part of the sequence), Kahu Volcanics (KV;
volcaniclastics and interbedded lavas of basaltic to andes-
itic composition), Ulubila Volcanics (UV; andesitic to
dacitic volcaniclastics and lavas), and Lemo Volcanics
(LV; breccias, subordinate lavas and dykes of andesitic
composition). K/Ar dating results indicate that the six
units fall into two main age groups, namely,
>11·2–10·3 Ma (SV, MI and BV) and >7·6–6·2 Ma
(KV, UV and LV). The intervening period was char-
acterized by differential block-faulting and erosion. We
Fig. 2. Tectonic sketch map of the area just before the collision of the will refer to the older group as ‘early post-collisional’ and
west Sulawesi arc with the Buton microcontinent, after Hall (1996). It the younger group as ‘late post-collisional’.
should be noted that the Buton–Tukang Besi Platform (BTB) is here The Marara Ignimbrite and the lower and middle
indicated as a single entity, although recent work suggests that it consists
of two separate microcontinents. South Sulawesi area indicated by box.
parts of the Bila Volcanics are the equivalent of, re-
PNG, Bird’s Head area of Papua New Guinea. spectively, the E, F and G members of the Pammesurang
Volcanics of Van Leeuwen (1981). The Kahu Volcanics
and Ulubila Volcanics correspond to his Walanae Vol-
rocks exposed in the area are a flysch-like sequence canics, excluding the occurrences found in the Walanae
of greywackes, arkosic sandstones and shales of Late Depression, which we have grouped with the Walanae
Cretaceous age (Marada Sandstone Formation). These Formation.
are overlain (presumably above an angular conformity) Intrusive rocks are extensively exposed in the catch-
by the Langi Volcanics, a series of propylitized andesitic ment areas of the Biru and Bulubuluk rivers. These were
lavas and volcaniclastics that were formed in Paleocene(?) called Biru Granodiorite by Van Leeuwen (1981), but
to Eocene times and pre-date the collisional event (see renamed the Biru Intrusive Complex (BIC) in this paper
below). The Tonasa Limestone Formation, which ranges as our investigations have shown that the intrusive body
in age between Late Eocene and earliest Middle Miocene, consists of at least two suites: a Middle Eocene horn-
conformably overlies the volcanic series. The top part of blende–biotite monzonite–tonalite suite, which forms the
this formation contains significant amounts of volcanic bulk of the intrusive complex and is genetically related
material, including lava intercalations, which are referred to the Langi Volcanics, and an Upper Miocene clino-
to in this paper as the Pake Volcanics. These are the pyroxene–biotite syenite–monzonite suite, named here
only syn-collisional samples analysed from this area. The the Biru Syenite. A K/Ar cooling age of 8·4 Ma was
three formations were folded and block faulted during obtained, but the actual emplacement age could be older,
the early Middle Miocene and subjected to erosion before which could explain the isotopic similarities between the
being covered by a thick sequence of sedimentary and Biru Syenite and the Sopo Volcanics. Nishimura (cited
potassic volcanic rocks during the Middle to Late Mio- by Van Leeuwen, 1981) obtained a zircon fission-track
cene. Block faulting continued during this period, re- age of 19 ± 3·4 Ma for a granodiorite sample, suggesting
sulting in local tilting of the strata. The oldest part of that the intrusive complex may contain a third suite of
the cover sequence is a sedimentary unit belonging to Early–Middle Miocene age.
584
Table 1: Stratigraphic and volcanological characteristics of the igneous units in the Biru area
Unit Abbreviation K/Ar age ( Ma) Thickness Field appearance Classification Lower stratigraphic Upper stratigraphic
ELBURG et al.
(m) contact contact
Langi Volcanics LSv <400 lava flows, volcaniclastics, basalt–(trachy)-andesite Marada Sandstone Tonasa Limestone
dykes (angular unconformity) (conformable)
Langi Intrusives LSi 50–52 tonalite–granodiorite
Pake Volcanics PV (15; based on basalt
palaeontology)
Sopo Volcanics SV 10·5–11·2 250 lava flows, volcaniclastics, basaltic andesite, Tonasa Limestone —
585
dykes (NNW–ENE strike) shoshonite (angular unconformity)
Marara Ignimbrite MI 10·3 Ζ110 ignimbrite (columnar jointing) dacite Marada Sandstone BV (conformable)
(angular unconformity)
Bila Volcanics BV 10·8 175 volcaniclastics, tuffaceous marl, leucite tephrite, LV, BIC (angular KV (unconformity)
lava pyroxene basalt unconformity),
MI (conformable)
Biru Syenite BIC 8·4 intrusives syenite, monzonite
Kahu Volcanics KV 7·0 (Ar loss)–7·6 Ζ100 volcaniclastics, lava flows basalt–andesite BV (disconformity) UV, LV
Ulubila Volcanics UV 6·2 Ζ50 volcaniclastics, lava flows andesite–dacite KV (conformable) LV
Lemo Volcanics LV 6·3–7·0 >350 volcanic breccias, tuffs, lavas, andesite UV, KV, BIC —
dykes
ARC–CONTINENT COLLISION, BIRU, SULAWESI
Table 2: Petrographic characteristics of the units studied
Unit Texture Olivine Clino- Ortho- Plagioclase K-feldspar Hornblende Biotite Others Groundmass Secondary
pyroxene pyroxene minerals
Langi porphyritic 0–15% ± 60–77% + (in trachyte) 0–1% ± mt, ap ± qz fsp, glass ± qz chl, ep, sm,ill, ser,
Volcanics cc, actinolite
Langi holocrystalline 35–58% 6–50% 1–10% 3–12% mt, ap, zirc as above
Intrusives
Pake porphyritic 10% 77% mt (3%), ap (<1%) glass, 10% 70%: ab, ep, sm
Volcanics
Sopo porphyritic augite/ Ζ20% labradorite– common in Ζ2% ap (1–2%), plag, mt, px, sm, zeolite, chl, cc
Volcanics diopside, andesine, associated mt (2–4%) ap (K-fsp, bt), (mostly in
JOURNAL OF PETROLOGY
12–30% 65–70% dykes glass (0–45%) volcaniclastics)

Marara porphyritic augite, rare common, + + qz, mt glass, cpx, 15–25%, sm-illite,
Ignimbrite omphacite Ζ15mm opaques, bt, cc, zeolite, silica
minerals
586
Bila aphanitic, + + + rare mt, lc (in tephrite) glass (px, bt) cc, chalc (in
Volcanics porphyritic vesicles)
VOLUME 43
Biru holocrystalline augite, 1–2% andesine–albite orthoclase 3–6% sphene, ap, mt, — weak: ser, ill–sm, cc
Syenite 6–10% (20–40%) (45–55%) qz (<5%)
Kahu porphyritic ± augite Ζ20% ± andesine– ± ± mt, ap, glass (Ζ60%), trace–90%: chl, sm,
Volcanics (phenocryst) labradorite (phenocryst) (phenocryst, sphene (acc.) often devitrified zeolite, haematite,
NUMBER 4
(30–85%) groundmass) cc, tr

Ulubila micro- <1% augite minor andesite: andesine dacite: andesite: resorbed, dacite: quartz glass, microlites minor: sm, cc,
Volcanics crystalline, hypersthene (70–80%); dacite: sanidine, 3–5% 3–5% microlites san, zeolite
porphyritic andesine–albite Ζ40mm plag
Lemo porphyritic 0–5% augite– labradorite– perhaps ?xenocryst Ζ2% glass, plag, px minor: sm, cc, chl,
Volcanics diopside, andesine, deuteric zeolite
APRIL 2002
6–15% 50–70%
ap, apatite; ab, albite; bt, biotite; cc, calcite; chalc, chalcedony; ep, epidote; lc, leucite; mt, magnetite; plag, plagioclase; px, pyroxene; qz, quartz; san, sanidine; tr,
tridymite; sm, smectite; ill, illite; ser, sericite; chl, chlorite; fsp, feldspar.
Table 3: K–Ar data for igneous rocks from the Biru area
Sample Rock type Material %K 40

Ar (×10−11 mol/g) 40
Ar∗/40Artotal Age (Ma)
Lemo Volcanics
T ol basalt WR 2·68 2·8819 0·709 6·3 ± 0·3
B 45 bt–cpx andesite WR 3·81 4·4575 0·430 6·7 ± 0·1
B 85 cpx andesite WR 3·94 4·8190 0·500 7·0 ± 0·1
Ulubila Volcanics
B 340 cpx–bt tuff Bt 6·85 0·7394 0·263 6·2 ± 0·2
Kahu Volcanics
B 78 bt–hbl andesite WR 3·33 4·0247 0·480 7·0 ± 0·1
B 74 cpx–bt andesite WR 7·57 9·9503 0·610 7·6 ± 0·2
B 47 bt–ol basalt WR 2·14 2·7709 0·500 7·5 ± 0·1
Biru Intrusive Complex
B 21 cpx–bt syenite WR 2·91 4·2969 0·570 8·5 ± 0·2
B 37 bt granite Bt 7·22 10·3965 0·530 8·3 ± 0·2
Bila Volcanics
B 342 lc tephrite Lc 13·67 2·5566 0·658 10·8 ± 0·1
Marara Ignimbrite
B 369 ignimbrite San 7·98 1·4286 0·877 10·3 ± 0·1
Sopo Volcanics
B 24 cpx basalt WR 3·3 6·0237 0·670 10·5 ± 0·2
B 27 bt–cpx andesite WR 2·16 4·0470 0·620 10·8 ± 0·2
O1 hbl andesite (dyke) Hbl 1·03 1·9627 0·212 11·2 ± 0·5
Langi Intrusives
B 360 bt monzonite Bt 6·69 6·1504 0·938 52·3 ± 0·5
B 370 bt monzonite Bt 6·28 5·4945 0·900 49·8 ± 0·4
WR, whole rock; Bt, biotite; Lc, leucite; San, sanidine; Hbl, hornblende.
The structure of the Biru area (Van Leeuwen, 1981) metaborate fusion rather than acid digestion to place the
is dominated by a series of steeply dipping to vertical sample into solution. This fusion technique ensures that
north- to NW-trending faults, including the West all major elements can be determined. Accuracy for the
Walanae Fault, which show both vertical and horizontal major element oxides is better than 0·01%. Analyses of
displacements. A second, less prominent fault system international standards show that accuracy for the trace
consists of a series of steeply dipping to vertical normal elements is better than 20% for Tb and Nb at levels
faults with northeasterly to easterly trends. Both fault lower than 8 ppm, better than 15% for Cr, Ce, Pr, Hf,
systems were probably already active in early Middle Ta and Th, better than 10% for V, Cu, Cs, Ba, Er, Tm
Miocene times and continued to be active intermittently and Lu, and better than 5% for Ni, Rb, Sr, Y, Zr, Nd,
up to the Pliocene. Sm, Eu, Gd, Dy, Ho, Yb, Pb, U and Nb at levels higher
The field and petrographic characteristics of the units than 100 ppm. Trace element analyses of standards are
studied have been tabulated in Tables 1 and 2. K–Ar included in Table 4.
age data are given in Table 3. Sr, Nd and Pb isotope ratios were analysed at the
Department of Geology and Geophysics by thermal
ionization mass spectrometry (TIMS) on a Finnigan
ANALYTICAL TECHNIQUES MAT 262 system in static mode. All ground samples
A selection of samples from the Eocene to Upper Miocene were leached in 3N HCl for 30 min at >100°C. The
units were analysed by Actlabs (Canada) for major oxides supernatant was pipetted off, the sample washed in
and a wide range of trace elements by fusion–inductively deionized water and the water pipetted off. The residue
coupled plasma and fusion–inductively coupled plasma- was then analysed for its isotopic composition. In selected
mass spectrometry methods, respectively, using a lithium cases, the leachate was analysed too (Table 5).
587
Table 4: Selected whole-rock analyses
Ulubila Ulubila Ulubila Bila Bila Kahu Kahu Kahu

Volcanics Volcanics Volcanics Tephrite Tephrite Volcanics Volcanics Volcanics
B 350 B 469 B 472 B 328 B 342 B 74 B 475 B 485
SiO2 55·33 62·64 55·08 42·28 42·52 50·33 59·57 49·99

Al2O3 16·1 17·93 19·33 16·51 16·63 16·88 15·44 16·07
Fe2O3∗ 6·12 3·56 5·97 10·75 10·75 9·59 6·12 10·17
MnO 0·15 0·05 0·22 0·24 0·25 0·14 0·1 0·17
MgO 2·87 0·56 1·83 4·88 4·73 4·39 3·39 4·5
CaO 7·57 1·8 4·97 12·94 13·09 8·73 5·59 10·29
Na2O 4·11 5·14 3·55 1·24 1·61 3·18 2·87 2·91
K2O 4·77 5·65 4·39 5·74 5·81 2·17 2·99 2·52
TiO2 0·74 0·48 1·11 1·51 1·47 1·02 0·77 1·11
P2O5 0·47 0·15 0·34 1·32 1·31 0·52 0·18 0·31
LOI 2·1 0·83 2·5 2·47 2·14 1·88 1·94 1·65
Total 100·33 98·78 99·3 99·88 100·31 98·97 98·95 99·68
mg-no. 48·15 23·75 37·77 47·34 46·56 47·55 52·31 46·70
Sc 13 5 9 9 8 23 18 27
V 118 51 117 608 616 232 136 240
Cr 54 32 53 38 35 37 87
Co 6 14 4 39 39 29 26
Ni n.d. n.d. n.d. 14 20 17 32
Cu 41 25 11 662 686 178 80
Zn 56 51 25 119 128 68 54
Ga 24 22 21 21 22 20 18
Rb 320 173 221 255 248 56 90
Sr 469 923 725 2041 2149 895 693
Y 26 26 72 49 46 19 37
Zr 633 361 418 99 81 106 155
Nb 89 38 50 8 8 3·9 6
Cs 6·1 3·5 2·8 9·9 10 3·3 8·8
Ba 581 1470 1330 2276 2299 952 1030
La 102 74 77 53 54 27 54
Ce 179 139 157 110·8 111·4 52·1 79
Pr 15·9 12·6 14·2 13·17 12·89 5·36 9·96
Nd 54 47 56 63·8 63·2 27·3 41
Sm 9 8·6 12 15 14·8 5·5 8·9
Eu 1·8 2·14 2·7 4·6 4·6 1·6 1·9
Gd 5·8 5·9 11 12·3 11·9 4·8 8·1
Tb 1 1 1·9 1·8 1·7 0·7 1·3
Dy 5 4·9 10 8·9 8·8 3·8 6·8
Ho 0·9 0·9 2 1·6 1·6 0·7 1·3
Er 2·9 2·7 6·1 4·8 4·7 1·9 3·7
Tm 0·44 0·39 0·9 0·6 0·57 0·26 0·52
Yb 2·8 2·4 5·6 3·8 3·6 1·8 3·2
Lu 0·43 0·37 0·89 0·55 0·53 0·26 0·5
Hf 14 8 9·7 3·1 2·7 3·1 4·3
Ta 4·51 2·46 2·73 1·11 1·11 0·3 0·4
Pb 64 42 48 84 103 12 36
Th 50 17 21 29·4 28·9 8·4 11
U 12 2·7 5·6 10 9·9 2·4 2·1
588
Lemo Lemo Lemo Marara Marara Pake Sopo Sopo

Volcanics Volcanics Volcanics Ignimbrite Ignimbrite Volcanics Volcanics Volcanics
B 48 B 86 B 338 B 42 B 369 B 53 B 27 B 54
SiO2 57·96 57·57 57·3 58·18 58·56 50·49 49·84 48·51

Al2O3 15·78 16·18 14·81 19·5 19·49 18·53 17·47 18·08
Fe2O3 6·55 7·21 8·06 3·21 3·25 8·89 9·63 10·11
MnO 0·16 0·12 0·12 0·2 0·22 0·37 0·18 0·21
MgO 3·27 3·17 4·2 0·62 0·63 2·8 3·56 3·82
CaO 5·82 6 6·61 2·01 2·86 6·28 9·5 9·76
Na2O 3·23 3·05 2·88 3·86 4·34 2·79 2·77 3·06
K2O 4·08 3·79 3·56 7·09 7·09 4·07 3·24 2·07
TiO2 0·78 0·9 0·89 0·59 0·57 0·87 0·79 0·88
P2O5 0·44 0·42 0·52 0·11 0·1 0·05 0·37 0·45
LOI 1·2 1·1 0·89 3·77 3·56 4·1 2·29 2·44
Total 99·44 99·61 99·81 99·13 100·68 99·24 99·8 99·54
mg-no. 49·72 46·54 50·79 27·67 27·74 38·41 42·27 42·80
Sc 15 16 28 1 1 19 22 17
V 144 154 181 77 79 257 285 290
Cr 85 69 109 32 27 40 44 37
Co 22 19 25 2 2 21 25 27
Ni 18 10 40 n.d. 20 46 6 n.d.
Cu 64 43 126 n.d. 13 95 113 182
Zn 53 36 66 103 100 270 89 59
Ga 19 19 19 20 18 19 19 21
Rb 136 135 125 244 206 88 47 32
Sr 771 805 761 383 379 3054 1524 2116
Y 25 26 24 35 35 18 20 25
Zr 171 185 131 282 275 88 63 66
Nb 7·7 7·7 6 13 12 3 1·1 1·5
Cs 3·1 4·4 3·9 11·8 10·6 n.d. 0·5 0·5
Ba 977 1000 1033 72 110 1165 920 926
La 38·9 41·4 31·4 60 61·6 11·0 18·7 17·3
Ce 73·5 78·6 58·7 112·9 125·8 21·4 37·3 35·4
Pr 7·04 7·76 6·26 11·75 12·08 2·63 4·04 4·11
Nd 34·7 37·3 28·7 48·4 47·4 13·9 21·9 23·4
Sm 6·4 7 6·3 9·1 9·4 3·6 4·9 5·1
Eu 1·52 1·63 1·97 1·9 1·98 1·9 1·56 1·64
Gd 5·7 6 5·8 9·8 9·9 2·7 4·5 5·1
Tb 0·8 0·9 0·8 1·2 1·1 0·5 0·6 0·7
Dy 4·3 4·8 4·2 5·9 5·9 3 3·8 4·3
Ho 0·8 0·9 0·8 1·1 1·1 0·6 0·8 0·8
Er 2·5 2·6 2·5 3·5 3·5 1·7 2·3 2·4
Tm 0·35 0·38 0·32 0·5 0·51 0·25 0·3 0·35
Yb 2·3 2·5 2·1 2·9 3·2 1·7 2·1 2·2
Lu 0·37 0·39 0·33 0·49 0·51 0·25 0·3 0·34
Hf 5 5·3 3·7 6·1 6·6 2·3 1·9 1·9
Ta 0·59 0·6 0·49 0·95 0·96 0·17 0·1 0·12
Pb 7 n.d. 33 120 114 60 8 29
Th 15·3 15·6 12·1 29·2 29·6 4·3 5·0 3·5
U 4·1 4·0 3·5 8·2 10·3 1·5 1·7 1·3
589
Table 4: continued
Biru Biru Langi Langi Langi B Langi B Standard Standard

Syenite Syenite Intrusives Intrusives
B 21 B 22 B 360 B 370 B 18 B 19 W2 MRG1
SiO2 55·6 56·11 64·92 66·7 54·74 57·87

Al2O3 17·75 18·3 16·03 15·95 16·31 17·5
Fe2O3 6·47 6·23 4·27 4·26 7·99 5·95
MnO 0·13 0·12 0·07 0·07 0·19 0·1
MgO 1·93 1·82 1·78 1·84 4·45 2·56
CaO 6·21 6·03 3·92 6·95 8·94 5·73
Na2O 3·78 4·01 4·62 4·46 2·46 2·86
K2O 4·81 4·9 2·04 1·89 0·16 1·96
TiO2 0·62 0·65 0·52 0·5 0·87 0·59
P2O5 0·34 0·32 0·25 0·15 0·19 0·13
LOI 0·95 0·61 0·62 0·77 1·21 2·69
Total 98·78 99·28 99·03 100·53 97·53 97·95
mg-no. 37·14 36·65 45·22 46·10 52·45 46·01
Sc 9 9 9 8 23 13
V 172 166 63 64 198 139 264 523
Cr 54 50 109 67 145 65 107 433
Co 16 15 10 10 26 17 42 87
Ni n.d. n.d. 45 26 9 9 68 194
Cu 62 67 17 240 48 13 112 136
Zn 49 30 17 0 34 60 83 188
Ga 18 18 17 16 19 21 18 18
Rb 106 92 34 33 3 38 19 7
Sr 1302 1312 421 416 598 689 196 271
Y 20 24 15 13 22 14 22 13
Zr 172 133 138 110 112 76 99 104
Nb 4·1 4·8 5 4 3·2 1·1 6 20
Cs 2 1·7 1·1 1·1 0·2 3·2 0·9 0·6
Ba 1029 1004 301 281 88 198 174 49
La 25·2 26·3 13·8 13·8 15·6 7·0 11 9·9
Ce 48·3 54·4 30·1 29·7 35·05 15·86 24 27
Pr 4·76 5·59 3·01 2·85 3·761 1·815 2·92 3·73
Nd 24·2 29·3 12·4 12 18·8 9·6 14 19
Sm 4·9 5·8 2·6 2·4 4·34 2·6 3·3 4·5
Eu 1·45 1·58 0·9 0·83 1·246 0·872 1·11 1·45
Gd 4·4 5·2 2·6 2·6 4·02 2·53 3·6 4·1
Tb 0·6 0·8 0·4 0·3 0·63 0·41 0·7 0·6
Dy 3·6 4·5 2·4 2·1 3·88 2·46 3·8 3·0
Ho 0·7 0·8 0·5 0·4 0·74 0·49 0·8 0·5
Er 2 2·5 1·4 1·3 2·39 1·46 2·3 1·2
Tm 0·29 0·34 0·22 0·2 0·33 0·21 0·35 0·11
Yb 2 2·3 1·4 1·3 2·36 1·44 2·1 0·7
Lu 0·32 0·36 0·22 0·2 0·361 0·23 0·31 0·09
Hf 4·2 3·6 3·6 2·9 3·36 2·35 2·6 3·8
Ta 0·27 0·37 0·43 0·39 0·234 0·1 0·4 0·9
Pb 38 42 8 7 n.d. 7 9 n.d.
Th 12·9 6·6 5·1 5·2 3·5 1·9 n.d. n.d.
U 4·5 1·9 1·9 1·1 0·8 0·6 n.d. 0·5
n.d., not detected. Fe2O3∗, all iron expressed as Fe2O3.
590
Table 5: Isotope data for Biru samples
206 207 208 206 207 208 206 207 208

Group Sample Pb/204Pb 2 SE Pb/204Pb 2 SE Pb/204Pb 2 SE Pb/204Pb Pb/204Pb Pb/204Pb Pb/204Pb Pb/204Pb Pb/204Pb
fc fc fc initial initial initial
Ulubila Volcanics B 350 18·796 0·002 15·614 0·003 38·766 0·009 18·841 15·670 38·952 18·827 15·670 38·933
ELBURG et al.
Ulubila Volcanics B 469 18·696 0·001 15·625 0·001 38·774 0·003 18·741 15·681 38·960 18·736 15·681 38·950
Ulubila Volcanics B472 18·610 0·001 15·570 0·001 38·541 0·004 18·655 15·626 38·726 18·647 15·626 38·716
Bila Tephrite B 328 18·915 0·001 15·629 0·001 38·967 0·004 18·961 15·685 39·154 18·946 15·685 39·140
Bila Tephrite B 342 18·964 0·002 15·603 0·002 38·933 0·004 19·010 15·659 39·120 18·998 15·659 39·108
Kahu Volcanics B 74 19·125 0·002 15·629 0·001 39·039 0·009 19·171 15·686 39·227 19·155 15·685 39·209
Kahu Volcanics B 475 19·178 0·002 15·624 0·002 39·014 0·006 19·224 15·680 39·201 19·219 15·680 39·193
Kahu Volcanics B 485 18·817 0·001 15·616 0·001 38·799 0·003 18·862 15·673 38·985 18·846 15·673 38·967
591
Lemo Volcanics B 48 18·916 0·001 15·606 0·001 38·843 0·002 18·962 15·662 39·029 18·919 15·661 38·978
Lemo Volcanics B 86 18·941 0·001 15·639 0·002 38·970 0·005 18·986 15·696 39·157 18·928 15·695 39·084
Lemo Volcanics B 338 19·185 0·001 15·671 0·001 39·241 0·002 19·231 15·727 39·429 19·223 15·727 39·421
Marara Ignimbrite B 42 18·630 0·001 15·576 0·001 38·558 0·002 18·675 15·632 38·743 18·667 15·632 38·733
Marara Ignimbrite B 369 18·611 0·002 15·550 0·002 38·472 0·006 18·656 15·606 38·656 18·645 15·606 38·646
Pake Volcanics B 53 18·464 0·002 15·534 0·002 38·328 0·005 18·508 15·589 38·512 18·504 15·589 38·508
Sopo Volcanics B 27 18·396 0·001 15·526 0·001 38·239 0·003 18·440 15·582 38·422 18·414 15·582 38·397
Sopo Volcanics B 54 18·424 0·001 15·533 0·002 38·278 0·005 18·468 15·589 38·462 18·462 15·589 38·457
Biru Syenite B 21 18·533 0·001 15·553 0·001 38·417 0·003 18·577 15·609 38·601 18·566 15·609 38·591
Biru Syenite B 22 18·505 0·001 15·539 0·001 38·360 0·002 18·549 15·595 38·544 18·545 15·595 38·539
Langi Intrusives B360 18·380 0·001 15·536 0·001 38·282 0·002 18·424 15·591 38·466 18·289 15·591 38·350
Langi Intrusives B370 18·352 0·001 15·513 0·001 38·229 0·003 18·396 15·569 38·413 18·307 15·569 38·278
ARC–CONTINENT COLLISION, BIRU, SULAWESI
Langi Volcanics B18 18·383 0·001 15·536 0·001 38·311 0·003 18·427 15·592 38·495 18·332 15·591 38·368
Langi Volcanics B19 18·286 0·001 15·520 0·001 38·186 0·002 18·329 15·576 38·369 18·285 15·576 38·320
Table 5: continued
87 87 87 143 143
Group Sample Sr/86Sr 2 SE Sr/86Sr Sr/86Sr 2 SE Nd/144Nd 2 SE Nd/144Nd 2 SE
initial l/r l/r
Ulubila Volcanics B 350 0·707950 0·000010 0·707898 0·512478 0·000007

Ulubila Volcanics B 469 0·705642 0·000010 0·705592 0·512664 0·000006
Ulubila Volcanics B472 0·705156 0·000012 0·705106 0·705715 l 0·000009 0·512728 0·000007 0·512700 l 0·000008
Bila Tephrite B 328 0·706213 0·000011 0·706157 0·706233 l 0·000012 0·512635 0·000009 0·512629 l 0·000007
Bila Tephrite B 342 0·706226 0·000011 0·706175 0·512629 0·000007
Kahu Volcanics B 74 0·706131 0·000011 0·706113 0·706127 l 0·000012 0·512678 0·000009 0·512703 l 0·000054
Kahu Volcanics B 475 0·708467 0·000012 0·708429 0·512489 0·000008
JOURNAL OF PETROLOGY
Kahu Volcanics B 485 0·705479 0·000010 0·705455 0·512672 0·000006 0·512667 r 0·000009
Lemo Volcanics B 48 0·707173 0·000011 0·707127 0·707091 l 0·000011 0·512573 0·000011 0·512559 l 0·000006
Lemo Volcanics B 86 0·707077 0·000011 0·707034 0·512599 0·000006
592
Lemo Volcanics B 338 0·708159 0·000011 0·708116 0·512510 0·000007
Marara Ignimbrite B 42 0·705671 0·000012 0·705383 0·512742 0·000019
VOLUME 43
Marara Ignimbrite B 369 0·705612 0·000013 0·705367 0·706484 l 0·000016 0·512708 0·000006 0·512703 l 0·000005
Pake Volcanics B 53 0·705016 0·000014 0·704999 0·704781 l 0·000014 0·512919 0·000008 0·512911 l 0·000009
Pake Volcanics B 53 r 0·705036 0·000014 0·512922 0·000007
Sopo Volcanics B 27 0·703846 0·000012 0·703832 0·703876 l 0·000013 0·512909 0·000008 0·512935 l 0·000006
Sopo Volcanics B 54 0·704263 0·000010 0·704256 0·512944 0·000011
NUMBER 4
Biru Syenite B 21 0·704230 0·000011 0·704202 0·704368 l 0·000010 0·512882 0·000009 0·512864 l 0·000007
Biru Syenite B 22 0·704265 0·000013 0·704206 0·512874 0·000006
Langi Intrusives B360 0·703592 0·000015 0·703428 0·512900 0·000006
Langi Intrusives B370 0·703628 0·000015 0·703466 0·512879 0·000006
Langi Volcanics B18 0·703795 0·000012 0·703785 0·512847 0·000006 0·512842 l 0·000005
APRIL 2002
Langi Volcanics B19 0·703813 0·000011 0·703701 0·703845 r 0·000011 0·512867 0·000007
fc, fractionation corrected; l/r, leachate or repeat.

The long-term average for the in-house Nd standard Volcanics (Elburg & Foden, 1999a). Positive anomalies
( J&M specpure Nd2O3) is 0·511603 ± 9 (1 of total are observed for K, Pb and Sr (except for the Marara
population, n = 105). The La Jolla standard gave Ignimbrite and Ulubila Volcanics).
0·511828 ± 11 (n = 9) and BCR-1 gave 0·512593 ± Chondrite-normalized rare earth element (REE) pat-
16 (n = 12). Typical blanks are of the order of 100–200 terns (Fig. 5) display moderate to high enrichment of
pg for Nd. The average for the NBS987 Sr standard is light REE (LREE) over heavy REE (HREE). The Langi
0·710258 ± 18 (n = 56). Typical Sr blanks are better Volcanics show the least LREE enrichment, and the
than 2 ng. Although this is a high blank, it is negligible degree of enrichment increases from BV basalt and SV
compared with a typical sample size of 10–100 g of Sr. samples, through KV and LV samples to BV leucite
All Pb isotopic analyses were performed at approximately tephrite, MI and UV samples. None of the suites show
the same temperature of 1150°C, and a mass frac- significant HREE fractionation. The MI, LV, UV and
tionation factor of 0·08% per a.m.u. was used, based on some of the KV samples display negative Eu anomalies,
replicate analyses of the NBS981 Pb standard. Typical whereas the other KV samples and samples from the SV
Pb blanks are of the order of 300 pg. and BV units show smooth patterns. The one sample of
K–Ar dating of samples B360, B370, O1 and T was the Pake Volcanics shows a positive Eu anomaly. As this
performed at AMDEL analytical laboratories, on a modi- sample is altered, with epidote and albite as secondary
fied MS-10 mass spectrometer, following techniques de- minerals (Table 2), we think that this positive anomaly
scribed by Webb et al. (1986). Duplicate K analyses agree is more likely to be related to alteration than to reflect
within 0·5% and duplicate Ar analyses within 1%. The a primary feature of the sample.
errors given are for the analytical uncertainty at 1 SD. The Langi Volcanics and Intrusives display the geo-
The other samples were dated at the Institute of Geo- chemical characteristics most typical of arc volcanics,
logical and Nuclear Sciences Ltd, Rafter Laboratories, with low Nb and Ti contents, and a calc-alkaline trend
on a modified MS-10 mass spectrometer, following in- in an SiO2 vs FeO∗/MgO diagram. In many variation
house methods developed by C. Adams. Errors of age diagrams, the Langi samples overlap with those from the
measurement are 2 SD and combine individual errors oceanic Sangihe arc of northern Sulawesi.
of potassium (±0·5%) 38Ar spike calibration (±0·3%), Our dataset includes a single, altered, basaltic rock from
Ar isotope measurement (<0·1%) and instrument mass the Pake Volcanics and it is unclear how representative of
discrimination variation (for 40Ar/36Ar <0·3%). the suite is this sample. Immobile elements such as Ti,
Nb, Zr and Y overlap the fields of those of the slightly
younger Sopo Volcanics.
The Sopo Volcanics are geochemically fairly similar
MAJOR AND TRACE ELEMENT to the ‘pre-collisional’ samples analysed by Elburg &
GEOCHEMISTRY Foden (1999a), with which they are now thought to be
Whole-rock analyses of the Biru samples cover the petro- contemporaneous. Sr contents are very high (often >2000
graphic spectrum from basaltic to dacitic rocks (SiO2 ppm), and Ba contents are also elevated. With respect
contents between 45 and 67 wt %; Fig. 3, Table 4). The to Nb, Ti, Zr and Y, the Sopo Volcanics do not differ
Miocene samples are characterized by high total alkalis much from the Langi Volcanics and Intrusives.
(Na2O + K2O >5 wt %), high K2O/Na2O ratios (>0·6 The Marara Ignimbrite stands out in many Harker
at 50 wt % SiO2, >1·0 at 55 wt % SiO2), high but variation diagrams, by having rather variable contents
variable Al2O3 (14·3–20·8 wt %), low TiO2 (<1·5 wt %), of Na2O, K2O, Ba, Rb and Sr, and low CaO contents.
and low mg-numbers (<54). Most of the Biru samples These elements are mobile during alteration, and there
plot in the shoshonitic field of the K2O–SiO2 classification appears to be a broad correlation between loss on ignition
diagram of orogenic lavas of Peccerillo & Taylor (1976; and values for these elements. We therefore think that
Fig. 3). Only the Eocene Langi samples fall within the these variations are more related to alteration than to
field for low- to medium-K rocks. the primary igneous composition of these samples.
The Miocene suites have the following trace element The Bila Tephrite is unusual with respect to its very
characteristics in common. Compatible elements occur low SiO2 content, combined with relatively low contents
in relatively low concentrations (Cr <220 ppm, Ni <70 of MgO and compatible trace elements such as Ni and
ppm, Co <40 ppm). Normal mid-ocean ridge basalt (N- Cr. Its TiO2, V, P2O5, Cu and Ba content, on the other
MORB)-normalized trace element diagrams (Fig. 4) show hand, are much higher than for the other suites, and
well-defined negative anomalies for Ta, Nb and Ti, and Th, Sr and Zn are also high for the silica content of the
patterns that are both steep and enriched in large ion samples.
lithophile elements (LILE). The Nb–Ta anomaly is least The Biru Syenite and Kahu Volcanics do not show
pronounced in samples from the Ulubila Volcanics, and any particularly unusual geochemical features. Like all
in this respect they resemble the Pliocene Lompobatang post-Eocene suites from this area, they are fairly rich in
593
Fig. 3. Variation of selected major elements (MgO, K2O, TiO2) and trace elements (Sr, Zr, Ni) vs SiO2. Legend ordering follows stratigraphic
succession. The field for North Sulawesi samples (oceanic arc) is from Elburg & Foden (1998). Fields for other South Sulawesi pre-, syn- and
post-collisional (Lompobatang) samples from Elburg & Foden (1999a). Sediment samples from Vroon et al. (1996) and Elburg & Foden (1999b).
K2O, Rb, Ba and Th. Levels of Nb and Y are only The youngest suite in the Biru area, the Lemo Vol-
slightly higher than in the Eocene Langi Volcanics and canics, does not show any of the unusual Nb, Zr or Th
Intrusives. enrichments seen in the slightly older Ulubila Volcanics.
The Ulubila Volcanics contain high but variable con- With respect to most elements, the Lemo Volcanics fall
tents of Nb and Zr, elements that are typically low in within the same trend as the Biru Syenite or Kahu
abundance in arc volcanics. Some of the enrichment of Volcanics.
Nb and Zr may result from the relatively fractionated
character of the samples, but they are also high in these
elements when compared with samples of equivalent
SiO2 content from other units. Rb and especially Th are ISOTOPE GEOCHEMISTRY
also high, whereas Ba and Sr occur at levels comparable In terms of Sr and Nd isotopic signature (Table 5, Fig.
with those of the other Miocene suites. Although the 6), the Biru samples fall roughly in the same range as
overall pattern of enrichments is similar to that seen in the samples from South Sulawesi analysed by Elburg &
the Pliocene Lompobatang volcano further south, the Foden (1999a). The one sample of the syn-collisional
Ulubila Volcanics show more extreme enrichments in Pake Volcanics has a relatively high 87Sr/86Sr ratio for
Nb and Zr. The two suites also differ in their Rb and its Nd isotopic ratio when compared with other samples
K2O contents (higher in the Ulubila) and Sr and P2O5 from South Sulawesi. Although this sample was leached
contents (higher in the Lompobatang) before analysis, alteration was severe, and it is possible
594
Fig. 4. N-MORB normalized trace element patterns for individual samples at similar intermediate SiO2 content, except Bila Tephrite (42%
SiO2). Normalization factors from Sun & McDonough (1989).
that the elevated Sr isotopic ratio reflects alteration rather In terms of Pb isotopes (Fig. 7), the Biru samples
than a primary igneous value. The Sopo Volcanics are extend the trend defined by the ‘syn-collisional’ samples
the most primitive in terms of Nd isotopic composition, from further south, with differences most pronounced in
207
although the Langi samples extend to somewhat lower Pb/204Pb ratios. The Eocene Langi Volcanics and
87
Sr/86Sr ratios. The early post-collisional Sopo Volcanics Intrusives stand apart from the other samples because of
plot very close to the field for the ‘pre-collisional’ samples their elevated 207Pb/204Pb ratios relative to their 206Pb/
204
of Elburg & Foden (1999a), of which the age is poorly Pb ratios. In this respect they show more similarity to
known. The Biru Syenite is also fairly primitive in its Sr the igneous rocks from North Sulawesi (Elburg & Foden,
and Nd isotope ratios, but the other post-collisional 1998). It is interesting to note that the Biru samples
samples from Biru fall within the same range as the extend to more radiogenic Pb isotope signatures than
7–12 Ma [termed ‘syn-collisional’ by Elburg & Foden sedimentary rocks from South and Central Sulawesi,
(1999a)] samples from elsewhere in South Sulawesi. The especially with respect to 206Pb/204Pb and 208Pb/204Pb.
samples with the most ‘continental’ Nd and Sr isotopic This means that exposed sedimentary rocks in Sulawesi
signature overlap with sediments from South Sulawesi are not viable end members to explain the ‘continental’
(Vroon et al., 1996; Elburg & Foden, 1999a). isotopic signatures of the Biru samples. Another argument
595
Fig. 5. Chondrite-normalized REE patterns for the same samples as

in Fig. 4. Normalization factors from Taylor & McLennan (1985).
against crustal contamination (in an AFC-type process)

is the lack of correlation between any index of frac-
tionation and isotopic signature. Within several suites,
the samples with the most radiogenic Sr isotopic signature
are those with the highest content of MgO (Tables 4
and 5), which is the opposite from what would be expected Fig. 7. Variation of 207Pb/204Pb and 208Pb/204Pb vs 206Pb/204Pb (initial)
for assimilation combined with fractional crystallization. for the Biru magmatic rocks. Legend and data sources as in Fig. 3.
broad temporal changes are also reflected in some aspects

ISOTOPIC VARIATION THROUGH of geochemical change. If Nd isotopic ratios are plotted
TIME against the age of the samples it is obvious that the late
The Biru samples define a broad envelope of isotopic post-collisional samples from Biru have more variable,
change through time, with 143Nd/144Nd ratios decreasing and, on average, lower 143Nd/144Nd ratios than the pre-,
to their lowest values at 6 Ma. The decidedly post- syn- and early post-collisional samples. However, al-
collisional samples of Lompobatang have higher values though there is a broad trend of isotopic change with
again. Sr isotopes mirror the trend of Nd, rising to their time, it is not one of simple unidirectional change towards
highest values at 6 Ma and then falling (Fig. 8). These more ‘continental’ isotopic ratios. For example, the Biru
143
Fig. 6. Variation of Nd/144Nd vs 87Sr/86Sr (initial) for the samples analysed and for North and South Sulawesi. Legend and data sources as
in Fig. 3.
596
Sulawesi samples never approach the very ‘continental’

values seen in the post-collisional rocks from the Biru
area.
GEOCHEMICAL VARIATION
THROUGH TIME
Comparison of the geochemical characteristics of the
various units is hampered by their difference in frac-
tionation stage, and we have therefore attempted to
determine crucial elemental concentrations and ratios at
55% SiO2 (Fig. 9). This is in some cases only a rough
estimate, as some units do not extend to these silica
contents (i.e. Bila Tephrite), and it is not always possible
to unambiguously interpret the trend of each element or
elemental ratio with SiO2 content. However, if we attempt
this exercise, we see that there is, again, no unambiguous
progression in time, although all Miocene volcanics have
higher K2O contents than the Eocene Langi Volcanics
and Intrusives at 55% SiO2. On average, Nb/Y and Ce/
Yb ratios are higher too. La/Nb ratios can be both
Fig. 8. Variation of 87Sr/86Sr and 143Nd/144Nd (initial) vs age. Legend
higher (Bila Tephrite, Kahu Volcanics; even when taking
and data sources as in Fig. 3. Data for ‘syn-collisional’ rocks from the error on the Nb analyses into account) and lower
Elburg & Foden (1999a) not included because of uncertainty in their (Ulubila Volcanics) than in the Langi samples.
age. Changes in isotopic and chemical compositions are
often observed in across-arc transects, such as described
for the Sunda Arc by Hoogewerff et al. (1997). It could
syenite, which falls between the Bila Tephrite and the therefore be argued that the position of the Biru area
Kahu Volcanics with respect to its K/Ar age, has higher changed with respect to the subducting slab through
143
Nd/144Nd ratios than either of these deposits. This time, and that this caused the geochemical and isotopic
discrepancy could also be explained by the emplacement changes observed. We do not think that this is a satis-
age of the syenite being older than its K/Ar cooling age, factory explanation, as the across-arc changes described
although it might be excessive to postulate a gap of 3 my by previous workers (Leeman et al., 1990; Woodhead &
between intrusion and cooling through the biotite closing Johnson, 1993; Hoogewerff et al., 1997) do not induce
temperature. If data from the Lompobatang volcano are more pronounced isotopic and chemical variability, but
incorporated into this diagram we can see that 143Nd/ rather a unidirectional change. This scenario would also
144
Nd values return to appreciably higher values >13 my necessitate very sudden changes in slab geometry in a
after the collisional event. The trends seen in Sr and very short time, as units that have virtually in-
Pb isotopes are very similar, with ‘continental’ values distinguishable K/Ar ages (Ulubila Volcanics and Lemo
becoming more important just after the collisional event, Volcanics) have vastly different isotopic and geochemical
whereas the much later Lompobatang samples show a signatures. We therefore contend that the observed geo-
return to values close to those for syn- and pre-collisional chemical changes reflect a time progression, and thereby
suites. a response to arc–continent collision.
The isotopic trend through time in the Biru area
contrasts markedly with the trend seen in North Sulawesi
(Fig. 8), indicating that the collision with the Buton
microcontinent and other continental fragments did not DISCUSSION
have an effect on the northerly area. The isotopic com- Nature of the mixing end members
position of samples from North Sulawesi remains virtually The discrepancy in isotopic compositions between sedi-
unchanged from 14 Ma until recent times, when the ments from South Sulawesi and the Biru samples is also
isotopic signature becomes more variable, probably be- clearly visible in a diagram of 143Nd/144Nd vs 206Pb/204Pb
cause of cessation of active subduction of the Molucca (Fig. 10). It is often envisaged that some mixing processes
Sea plate (Elburg & Foden, 1998). This is marked by a take place in the mantle wedge between a mantle com-
decrease in the Nd isotopic signature, but the northern ponent similar to that of a MORB source, and a fluid,
597
Fig. 10. Variation of 143Nd/144Nd vs 206Pb/204Pb (initial) ratio, showing

the distinction between the sediments from Sulawesi (stars) and the
Biru samples. Legend and data sources as in Fig. 3.
of, for instance, the Lemo Volcanics by this type of

process, we need two crustal end members with distinct
Pb isotopic characteristics. Another option is that the
array seen in the Lemo Volcanics represents a single
mixing line. In that case, a MORB source is unlikely to
be the isotopically primitive end member for this mixing
array, as its elemental Pb/Nd ratios are too low to give
rise to this kind of array when sedimentary material (with
very high Pb/Nd ratios compared with a MORB source)
is mixed with it. In that case we would need an end
member with isotopic characteristics similar to a MORB
source, but with Pb/Nd ratios closer to those of sediments
or the melts or fluids derived therefrom. Arc volcanics
themselves have suitable Pb/Nd ratios to be one end
member for this kind of process. The mixing process
could therefore be envisaged as taking place between an
arc-type magma or fluid-fluxed mantle component (which
are themselves a two-component mixture of MORB
source and fluid), and a sedimentary end member. How-
ever, this sedimentary end member cannot be represented
by the exposed sedimentary rocks in South Sulawesi, as
they do not have the appropriate Pb isotopic signature.
Fig. 9. Variation of K2O, Nb/Y, Ce/Yb and La/Nb vs age of the To explain the mixing array seen in the Pb isotope
various units, recalculated to the value at 55% SiO2, to avoid influence diagrams we need a component that has higher Pb isotope
of crystal fractionation. Legend and data sources as in Fig. 3. It should ratios than exposed sedimentary rocks from Sulawesi.
be noted that the Lemo Volcanics (×) stratigraphically overlie the high
Nb/Y Ulubila Volcanics (Φ) and are therefore younger.
Although we could postulate that this component is
present in the Sulawesi crust at a deeper level, this
solution invokes something like a deus ex machina. Material
melt-like or solid sedimentary end member. If this mixing that has the appropriate Pb isotope composition is found
process were the cause of the ‘continental’ isotopic sig- among sediments of the Banda Arc (Vroon et al., 1995),
nature in some of the Biru samples, the Nd–Pb isotope which have an Australian cratonic origin, instead of the
mixing arrays would be strongly curved (Fig. 11), with Sundaland origin of the exposed Sulawesi sedimentary
first an increase in Pb isotope values at high and near- rocks. Although we have no data on the sediments
constant Nd isotope ratios, followed by a decrease in that were present in the trench of the West Sulawesi
143
Nd/144Nd at high and nearly constant Pb isotope ratios subduction zone, there are reasons to think that these
(e.g. Vroon et al., 1993). This results from the very high contained a significant component of Australian origin.
Pb/Nd ratios of fluids or sediments compared with a The main reason is that these sediments are likely to
MORB source (Miller et al., 1994). To explain the array have been shed from the microcontinents of Australian
598
Fig. 11. Variation of 143Nd/144Nd vs 206Pb/204Pb to illustrate the types of mixing curves that could explain the isotopic composition of the Biru
samples. Field for Banda sediment from Vroon et al. (1995) and North Australian sediment from P. Z. Vroon & M. A. Elburg (unpublished
data, 1994).
origin that collided with the volcanic arc at a later of the volcanic arc, and Australian-type sediment be-
stage. Tectonic reconstructions, based on stratigraphic coming dominant as the Buton microcontinent ap-
similarities between Buton and the islands on the Aus- proaches the arc.
tralian continental shelf, such as Timor, generally classify
Buton as being part of the Australian plate (e.g. Hall,
1996). It is therefore probable that the sediments that Involvement of an OIB component?
were subducted shortly before, and during collision with The high Nb contents, high Nb/Zr and relatively low
the Buton microcontinent had an Australian origin, and La/Nb ratios of the Ulubila Volcanics compared with
could have had the appropriate Pb isotopic signature to the other samples could be explained by the involvement
explain the Pb–Pb mixing array. of an OIB-type component in magma petrogenesis. This
The Eocene Langi samples have relatively high 207Pb/ end member has been invoked in several studies of
204
Pb ratios for their 206Pb/204Pb ratios compared with Indonesian volcanics (Wheller et al., 1987; Vroon et al.,
the other samples from South Sulawesi, or from other 1993), most notably to explain the geochemical signature
Indonesian areas such as the Banda Arc (Vroon et al., of the high-K samples of the volcano Muriah in Java
1993). They overlap in Pb isotopic composition with (Edwards et al., 1991). The high-K samples of Muriah
12–14 Ma igneous samples from the oceanic Sangihe arc are characterized by relatively high 208Pb/204Pb ratios at
in North Sulawesi (Elburg & Foden, 1998). The shape a given 207Pb/204Pb ratio, an observation that can also
of the array for the other Biru volcanics and for the be made for Indian Ocean OIB (Fig. 13). This therefore
Banda Arc samples is characterized by a great variation makes a very strong case for the involvement of an OIB-
in 206Pb/204Pb ratios for a limited variation in 207Pb/204Pb, type component in magma genesis. The situation in the
Biru sector is different, as all samples, irrespective of their
thereby creating a relatively flat trend (Fig. 12). This
Nb/Zr ratio, fall on a straight mixing line between
kind of trend reflects involvement of a sedimentary end
MORB-type mantle and a sedimentary component,
member of Australian origin. The opposite trend is
which could be represented by Australian continental
observed in igneous suites from Halmahera and the material, such as has been dredged in the Banda Arc
Philippines, where there is significantly more variation (Vroon et al., 1995) or collected in North Australia
in 207Pb/204Pb ratios than in 206Pb/204Pb, thereby creating (Hoogewerff, 1999). If the elevated Nb contents of the
a steep to almost vertical field. This is generally taken to Ulubila Volcanics were the result of involvement of an
reflect involvement of a sedimentary end member with OIB-type component, we would expect to observe a
high 207Pb/204Pb ratios for their 206Pb/204Pb ratios, being shift towards the field for OIB and the Muriah high-K
sediment from the Pacific or South China Sea (Fig. 12) volcanics. As this is not the case, we do not think that
rather than from the Australian continent. We propose an OIB-type mantle has been involved in the petrogenesis
that the distinction seen between the Eocene and Mio- of any of the igneous rocks from the Biru area.
cene–Pliocene Biru samples has a similar origin, with The Lompobatang volcanics, which also display rel-
Pacific-type sediment being subducted early in the history atively high Nb/Zr ratios (Fig. 14), do not show any shift
599
Fig. 12. Variation of 207Pb/204Pb vs 206Pb/204Pb for selected sample suites. Data sources in this figure and subsequent figures for arcs and
sediments: Morris et al. (1983); Mukasa et al. (1987); Ben Othman et al. (1989); Edwards et al. (1991); McDermott et al. (1993); Vroon et al. (1993,
1995). Fields for I-MORB from database of Elburg & Foden (1999b). NHRL, Northern Hemisphere Reference Line.
Fig. 13. Variation of 208Pb/204Pb vs 207Pb/204Pb (initial) for the Biru samples and I-OIB, I-MORB [data sources in Elburg & Foden (1998)] and
Muriah (Edwards et al., 1991). Symbols and shading as in Fig. 3. Fields for I-OIB from database of Elburg & Foden (1999b).
in Pb isotopic compositions towards the field for OIB 1985), or those from the Banda trough (Fig. 14: 0·07–0·1,
either. For these volcanics Elburg & Foden (1999a) argued Vroon et al., 1995) or the North Australian continent
that they represented melting of the subcontinental litho- (0·02–0·05, P. Z. Vroon & M. A. Elburg, unpublished
spheric mantle. An alternative option, that high Nb/Zr data, 1994). The isotopic data could be interpreted to
contents could represent addition of a partial melt of mean that the Ulubila Volcanics contain a variable
sediment to the mantle wedge, is not considered likely amount of a sediment in their source. The fact that Nb/
as the Lompobatang volcanics show isotopically less Zr ratios are higher than those for average sediment
influence of a sedimentary component than the syn- would imply that the sediment was added to the mantle
collisional samples from the same area, which have lower wedge as a partial melt, rather than as a wholesale
Nb/Zr ratios (Fig. 14). The Ulubila Volcanics and the addition. High Nb/Zr ratios in partial melts can be
Lompobatang volcanics display some important differ- achieved only if rutile is not a residual phase, as this
ences: trace element and isotopic ratios are distinctly mineral has a significantly higher solid–melt distribution
more variable within the Ulubila Volcanics, and their coefficient for Nb than for Zr ( Jenner et al., 1994). Partial
isotopic signature trends towards more ‘continental’ val- melting experiments on pelagic clay show that rutile is
ues. Nb/Zr ratios in the Ulubila are higher (0·09–0·14) not a residual phase (Nichols et al., 1996). It is possible
than in the Lompobatang volcanics, but also higher than that rutile would be stable in more silicic environments
in global sediments (average of 0·09, Taylor & McLennan, (Ryerson & Watson, 1987), thereby lowering the Nb/Zr
600
Fig. 14. Variation of 87Sr/86Sr (initial) vs Nb/Zr ratio for the Biru samples, Lompobatang and syn-collisional samples from South Sulawesi, and
Sulawesi sediment. The range in Nb/Zr ratios is given for sediments from the Banda Arc (Vroon et al., 1995), and from North Australia (P. Z.
Vroon & M. A. Elburg, unpublished data, 1994).
ratio of the partial melt, but there is also evidence that Table 6a: Mixing end members for
zircon would be stable in these circumstances ( Johnson
& Plank, 1999), and this would result in an increase in mixing of fluid-modified mantle wedge
the Nb/Zr ratio of the melt. It is therefore possible that and sediment (Fig. 15a and b)
addition of partial melt of sediments to the source of
the volcanics is responsible for the high Nb/Zr and Fluid-modified mantle Sediment
‘continental’ isotopic values of the Ulubila Volcanics.
Sr (ppm) 100 150
Nd (ppm) 2 20
Quantification of the mixing process Pb (ppm) 1 16
If our ideas about the mixing process and end members 87
Sr/86Sr 0·7035 0·73
(simple mixing between fluid-modified mantle wedge and 143
Nd/144Nd 0·51295 0·5120
Australian sediment, based on within-suite trends for Pb 206
Pb/204Pb 18·45 19·8
and Nd isotopes; see above) are correct, we should be
able to quantify the amount of sediment that has been Fluid-modified mantle is based on the isotopic and trace
mixed into the mantle wedge, and constrain its isotopic element compositions of the Sopo Volcanics. Sediment based
composition. It is indeed possible to construct simple on values for North Australian sediments (P. Z. Vroon & M. A.
Elburg, unpublished data, 1994). (See text for discussion.)
mixing curves that follow the isotopic trends seen in the
Biru volcanics (Fig. 15a and b; end members given in
Table 6a). For the fluid-modified mantle wedge, we took already modified by addition of a fluid before sediment
the isotopic composition of the Sopo Volcanics, of which was added, thereby significantly increasing the Pb and
the trace element data show they were mainly influenced Sr concentrations in the mantle.
by a fluid component. The trace element composition At present, we do not have any constraints on the
was taken as being 10 times less than that of the Sopo composition of the sediment that was subducted. Tectonic
Volcanics, roughly equivalent to assuming that the vol- reconstructions suggest that the Buton microcontinent
canics were formed by 10% of melting. A satisfactory came from the New Guinea area of the Australian plate
mixing curve was achieved only if the sediment was taken (Hall, 1996), and is younger, and therefore isotopically
to be of North Australian derivation (P. Z. Vroon & M. less evolved, than the sedimentary end member with
A. Elburg, unpublished data, 1994). The mixing models which the modelling was achieved. We are therefore not
constrain the maximum amount of sediment added to convinced that the proposed mixing model is correct.
the source as being 10%. This is fairly high, but this To circumvent the necessity of postulating this iso-
follows from the assumption that the mantle wedge was topically very evolved sedimentary end member, we made
601
Fig. 15. Variation of 143Nd/144Nd vs 206Pb/204Pb (a) and 87Sr/86Sr (b) showing the results of the mixing model between fluid-modified MORB-
source and sediment to explain the isotopic variation of the Biru samples. Tick marks indicate the amount of added sediment in weight percent.
End members given in Table 6a.
a mixing model between fluid-modified and sediment- samples show greater isotopic and geochemical vari-
modified (perhaps best envisaged as resulting from fault- ability. The timing of the onset of the geochemical
controlled interleaving) mantle before melting (Fig. 16), variability is hard to determine as we do not have
with the sedimentary end member being similar to that complete time coverage. If the collisional event is indeed
of sediments from the Banda Trough. The end members taken to have occurred around 13–15 Ma, coincident
of this mixing process are given in Table 6b. Although with faulting in the Biru area, the apparent geochemical
this is a process that can explain the isotopic variation impact postdates this event by several million years, as
in the Biru samples, the physical aspects of this process the Marara Ignimbrite, at 10·5 Ma, is the first unit to
are poorly constrained. We suggest that melting of sedi- show more continental isotopic ratios. A time lag between
ment, and thereby transfer into the sub-arc mantle, may the subduction of a geochemical tracer into the arc and
be facilitated by collision-induced stalling of the subducted the expression thereof in arc volcanics has also been
slab, thereby raising the temperature of slab and sub- observed by Turner et al. (1997), who found that 2–4 my
ducted sediments at a given depth. The isotopic variation elapsed before the geochemical signature was seen in
we observe in the Biru samples may result from melting erupted volcanics.
of a multiply contaminated mantle, or by mixing of The dataset presented here conforms to the idea by
melts from parts of the mantle that have had distinct Turner et al. (1997) that samples that show the greatest
contamination histories. influence of a fluid component in their trace element
ratios (e.g. high Ba/Th ratios) have isotopic compositions
that deviate least from those expected for the mantle
CONCLUSIONS wedge. This is taken to be evidence that the fluid has
Analyses of pre- to post-collisional samples from the been derived by dewatering of the subducted slab, rather
Biru area show that the compositions of post-collisional than any entrained sedimentary component. Samples
602
Table 6b: Mixing end members for mixing of MORB mantle, sediment and fluid-modified mantle
(Fig. 16)
MORB mantle Sediment Mix A1 + B Fluid-modified Fluid-modified

(A1) (B) (C) mantle (A2) mantle (D)
Nd (ppm) 0·73 20 1·7 0·73 2

Pb (ppm) 0·03 10 0·52 0·6 1
143
Nd/144Nd 0·513 0·512 0·51241 0·513 0·51295
206
Pb/204Pb 18·2 19·5 19·43 18·2 18·45
Fig. 16. Variation of 143Nd/144Nd vs 206Pb/204Pb to illustrate the mixing model between fluid-modified mantle and sediment-modified mantle.
End members given in Table 6b. The sediment-modified mantle (C) was produced by mixing (curve 1) a MORB source (A1) with sediment (B).
Curve 2 was produced by mixing fluid-modified MORB source (with higher Pb concentrations, but the same isotopic compositions as unmodified
MORB source: A2) with C. Curve 3 reflects mixing between a fluid-modified MORB source, similar in isotopic composition to the Sopo
Volcanics (D) and C. The assumption that fluid-modified mantle could still have pristine Pb isotopic values (end member A2) is based on the
study by Turner et al. (1997), who found that the arc volcanics from Tonga–Kermadec that have the most obvious fluid signature have low Pb
isotopic ratios. Shaded areas are the same as in Fig. 15.
with lower Ba/Th ratios show more continental isotopic is by melting of a mixture of fluid-modified and sediment-
values, indicative of a sedimentary component in magma modified mantle.
genesis. The high Pb isotopic values of some Biru units The question of whether the continental isotopic sig-
preclude that this sedimentary component is of Sundaland nature that is observed reflects subduction of sediment
origin. Although it is possible that Australian crust is or of the leading edge of the Buton microcontinent itself
present in the upper plate owing to underthrusting in cannot be resolved with this dataset. In the case of the
the Cretaceous (Parkinson et al., 1998), our preferred Sunda–Banda Arc, several workers have argued that the
interpretation is that this sediment was subducted and continental signature in the volcanics reflects subduction
added to the mantle wedge. Simple mixing models of of the continental crust itself (Hilton et al., 1992; Van
MORB source and sediment can explain the observed Bergen et al., 1993) on the basis of helium isotopes and
isotopic variations, but necessitate the involvement of the uplift of the arc. Helium isotope data are not available
sedimentary end members with different isotopic com- from South Sulawesi, but it is obvious that uplift in
positions to explain the Pb isotopic variation within in a the area has not been pronounced. This contrasts with
single magmatic unit. Mixing models between fluid- Central Sulawesi, where there is clear evidence for high
modified MORB source and sediment can explain the rates of uplift and isotopic signatures characteristic of the
observed isotopic variations with a single sedimentary Australian subcontinental lithospheric mantle (Bergman
end member, but this sediment must have very evolved et al., 1996; Elburg & Foden, 1999b). In the absence of
isotopic ratios, similar to sediment from North Australia. this evidence in South Sulawesi, we would suggest that
Perhaps the best way to explain the observed variations the continental signature here is related to subduction of
603
sediment rather than the continental crust itself. The Brouwer, H. A. (1947). Geological Explorations in the Island of Celebes.
reason for this sedimentary signature to become more Amsterdam: North Holland.
pronounced in the post-collisional volcanics may be re- Charlton, T. R. (2000). Tertiary evolution of the Eastern Indonesian
Collision Complex. Journal of Asian Earth Sciences 18, 603–631.
lated to halting of the subduction process, thereby al- Daly, M. C., Hooper, B. G. D. & Smith, D. G. (1987). Tertiary plate
lowing the slab and subducted sediment more time to tectonics and basin evolution in Indonesia. Jakarta: Indonesian
heat up at a given depth in the mantle. This could Petroleum Association. Proceedings of the Petroleum Association, 16th
cause melting of subducted sediment and allow for more Annual Convention, pp. 399–427.
effective transport of sediment from the slab to the Daly, M. C., Cooper, M. A., Wilson, J., Smith, P. G. & Hooper, B. G.
mantle wedge. Evidence for partial melting of sediments D. (1991). Cenozoic plate tectonic and basin evolution in Indonesia.
is sometimes found by an increase in Nb/Zr ratios of Marine and Petroleum Geology 8, 2–21.
Davidson, J. P. (1987). Crustal contamination versus subduction zone
the erupted volcanics (Vroon et al., 1993), but in the case
enrichment: examples from the Lesser Antilles and implications for
of South Sulawesi there is no clear correlation between mantle source compositions of island arc volcanic rocks. Geochimica
Nb/Zr ratios and the isotopic signature of the volcanics. et Cosmochimica Acta 51, 2185–2198.
It is therefore unclear whether the high Nb/Zr ratios of Davidson, J. W. (1991). The geology and prospectivity of Buton Island,
the Ulubila Volcanics reflect partial melting of sediment S.E. Sulawesi, Indonesia. Jakarta: Indonesian Petroleum Association.
or the addition of a component from the subcontinental Proceedings of the Indonesian Petroleum Association, 20th Annual Convention,
lithospheric mantle, as has also been argued for the pp. 209–233.
Lompobatang Volcano. Edwards, C., Menzies, M. & Thirlwall, M. (1991). Evidence from
It is interesting to note that continental collision does Muriah, Indonesia, for the interplay of supra-subduction zone and
intraplate processes in the genesis of potassic alkaline magmas.
not impose unidirectional geochemical change in the
Journal of Petrology 32, 555–592.
post-collisional volcanics, but rather introduces greater Eiler, J. M., Crawford, A., Elliott, T., Farley, K. A., Valley, J. W. &
geochemical variability. This would argue for the semi- Stolper, E. M. (2000). Oxygen isotope geochemistry of oceanic-arc
contemporaneous tapping of discrete mantle reservoirs, lavas. Journal of Petrology 41, 229–256.
rather than a wholesale change in the geochemistry of Elburg, M. A. & Foden, J. (1998). Temporal changes in arc magma
the magma source. A similar collisional event may also geochemistry, Northern Sulawesi, Indonesia. Earth and Planetary Science
explain the extreme geochemical variability seen in Cent- Letters 163, 381–398.
ral Sulawesi (Bergman et al., 1996; Elburg & Foden, Elburg, M. A. & Foden, J. (1999a). Geochemical response to varying
tectonic settings: an example from southern Sulawesi (Indonesia).
1999b).
Geochimica et Cosmochimica Acta 63, 1155–1172.
Elburg, M. & Foden, J. (1999b). Sources for magmatism in Central
Sulawesi: geochemical and Sr–Nd–Pb constraints. Chemical Geology
156, 67–93.
ACKNOWLEDGEMENTS
Foley, S. F., Barth, M. G. & Jenner, G. A. (2000). Rutile/melt partition
David Bruce is acknowledged for his assistance in the coefficients for trace elements and an assessment of the influence of
isotope laboratory. This work was carried out while rutile on the trace element characteristics of subduction zone
M.A.E. was a recipient of an ARC Australian Post- magmas. Geochimica et Cosmochimica Acta 64, 933–938.
doctoral Fellowship. Helpful reviews were provided by Fortuin, A. R., De Smet, M. E. M., Hadiwasatra, S., van Marle, L.
Simon Turner and Manfred van Bergen. J., Troelstra, S. R. & Tjokrosapoetro, S. (1990). Late Cenozoic
sedimentary and tectonic history of south Buton, Indonesia. Journal
of Southeast Asian Earth Sciences 4, 107–124.
Gasparon, M. & Varne, R. (1998). Crustal assimilation versus subducted
REFERENCES sediment input in west Sunda arc volcanics: an evaluation. Mineralogy
Ali, J. R., Milsom, J., Finch, E. M. & Mubroto, B. (1996). SE Sundaland and Petrology 64, 89–117.
accretion: palaeomagnetic evidence of large Plio-Pleistocene thin- Gill, J. B. & Williams, R. W. (1990). Th isotope and U-series studies
skin rotations in Buton. In: Hall, R. & Blundell, D. J. (eds) Tectonic of subduction-related volcanic rocks. Geochimica et Cosmochimica Acta
Evolution of Southeast Asia. Geological Society, London, Special Publications 54, 1427–1442.
106, 431–443. Hall, R. (1996). Reconstructing Cenozoic SE Asia. In: Hall, R. &
Alves, S., Schiano, P. & Allègre, C. J. (1999). Rhenium–osmium isotopic Blundell, D. J. (eds) Tectonic Evolution of Southeast Asia. Geological Society,
investigation of Java subduction zone lavas. Earth and Planetary Science London, Special Publications 106, 153–184.
Letters 168, 65–77. Hall, R. & Wilson, M. E. J. (2000). Neogene sutures in eastern Indonesia.
Ben Othman, D., White, W. M. & Patchett, J. (1989). The geochemistry Journal of Asian Earth Sciences 18, 781–808.
of marine sediments, island arc magma genesis and crust–mantle Hamilton, W. (1979). Tectonics of the Indonesian Region. US Geological Survey
recycling. Earth and Planetary Science Letters 94, 1–21. Professional Paper 1078, 345 pp.
Bergman, S. C., Coffield, D. Q ., Talbot, J. P. & Garrard, R. A. (1996). Hasan, K. (1991). The Upper Cretaceous flysch succession of the
Tertiary tectonic and magmatic evolution of western Sulawesi and Balangbaru Formation, Southwest Sulawesi. Jakarta: Indonesian
the Makassar Strait, Indonesia: evidence for a Miocene continent– Petroleum Association. Proceedings of the Indonesian Petroleum Association,
continent collision. In: Hall, R. & Blundell, D. J. (eds) Tectonic Evolution 20th Annual Convention, pp. 183–208.
of Southeast Asia. Geological Society, London, Special Publications 106, Hilton, D. R., Hoogewerff, J. A., van Bergen, M. J. & Hammerschmidt,
391–429. K. (1992). Mapping magma sources in the east Sunda–Banda arcs,
604
Indonesia: constraints from helium isotopes. Geochimica et Cosmochimica Parkinson, C. D., Miyazaki, K., Wakita, K., Barber, A. J. & Carswell,
Acta 56, 851–859. D. A. (1998). An overview and tectonic synthesis of the pre-Tertiary
Hoogewerff, J. A. (1999). Magma genesis and slab–wedge interaction very-high-pressure metamorphic and associated rocks of Java, Su-
across an island arc–continent collision zone, East Sunda Arc, lawesi and Kalimantan, Indonesia. The Island Arc 7, 184–200.
Indonesia. Ph.D. thesis, Utrecht University. Pearce, J. A., Kempton, P. D., Nowell, G. M. & Noble, S. R.
Hoogewerff, J. A., Van Bergen, M. J., Vroon, P. Z., Hertogen, J., (1999). Hf–Nd element and isotope perspective on the nature and
Wordel, R., Sneyers, A., Nasution, A., Varekamp, J. C., Moens, H. provenance of mantle and subduction components in western Pacific
L. E. & Mouchel, D. (1997). U-series, Sr–Nd–Pb isotope and trace- arc–basin systems. Journal of Petrology 40, 1579–1611.
element systematics across an active island arc–continent collision Peate, D. W., Pearce, J. A., Hawkesworth, C. J., Colley, H., Edwards,
zone: implications for element transfer at the slab–wedge interface. C. M. H. & Hirose, K. (1997). Geochemical variations in Vanuatu
Geochimica et Cosmochimica Acta 61, 1057–1072. arc lavas: the role of subducted material and a variable mantle
Jenner, G. A., Foley, S. F., Jackson, S. E., Green, T. H., Fryer, B. J. wedge composition. Journal of Petrology 38, 1331–1358.
& Longerich, H. P. (1994). Determination of partition coefficients Peccerillo, A. & Taylor, S. R. (1976). Geochemistry of Eocene calc-
for trace elements in high pressure–temperature experimental run alkaline volcanic rocks from the Kastamonu area, northern Turkey.
products by laser ablation microprobe–inductively coupled plasma- Contributions to Mineralogy and Petrology 58, 63–81.
mass spectrometry (LAM–ICP-MS). Geochimica et Cosmochimica Acta Pigram, C. J. & Panggabean, H. (1984). Rifting of the northern margin
58, 5099–5103. of the Australian continent and the origin of some microcontinents
Johnson, M. C. & Plank, T. (1999). Dehydration and melting ex- in eastern Indonesia. Tectonophysics 107, 331–353.
periments constrain the fate of subducted sediments. Geochemistry, Polvé, M., Maury, R. C., Bellon, H., Rangin, C., Priadi, B., Yuwono,
Geophysics, Geosystems 1, 1999GC000014. S., Joron, J. L. & Soeria Atmadja, R. (1997). Magmatic evolution
Katili, J. A. (1978). Past and present geotectonic setting of Sulawesi, of Sulawesi: constraints on the Cenozoic geodynamic history of the
Indonesia. Tectonophysics 45, 289–322. Sundaland active margin. Tectonophysics 272, 69–92.
Kepezhinskas, P., McDermott, F., Defant, M. J., Hochstadter, A., Ryerson, F. J. & Watson, E. B. (1987). Rutile saturation in magmas:
implications for Ti–Nb–Ta depletion in island-arc basalts. Earth and
Drummond, M. S., Hawkesworth, C. J., Koloskov, A., Maury, R.
Planetary Science Letters 86, 225–239.
C. & Bellon, H. (1997). Trace element and Sr–Nd–Pb isotopic
Smith, R. B. & Silver, E. A. (1991). Geology of a Miocene collision
constraints on a three-component model of Kamchatka Arc petro-
complex, Buton, eastern Indonesia. Geological Society of America Bulletin
genesis. Geochimica et Cosmochimica Acta 61, 577–600.
103, 660–678.
Leeman, W. P., Smith, D. R., Hildreth, W., Palacz, Z. & Rogers, N.
Stolz, A. J., Varne, R., Davies, G. R., Wheller, G. E. & Foden, J. D.
(1990). Compositional diversity of Late Cenozoic basalts in a transect
(1990). Magma source components in an arc–continent collision
across the southern Washington Cascades: implication for subduction
zone: the Flores–Lembata sector, Sunda arc, Indonesia. Contributions
zone magmatism. Journal of Geophysical Research 95, 19561–19582.
to Mineralogy and Petrology 105, 585–601.
McDermott, F., Defant, M. J., Hawkesworth, C. J., Maury, R. C. &
Sukamto, R. (1975). Geologic Map of Indonesia, Ujung Pandang Sheet, scale
Joron, J. L. (1993). Isotope and trace element evidence for three
1:1 000 000, sheet 8. Bandung: Indonesian Geological Survey.
component mixing in the genesis of the North Luzon arc lavas
Sukamto, R. (1982). Geologi lembar Pangkajene dan Watampone bagian barat,
(Philippines). Contributions to Mineralogy and Petrology 113, 9–23.
Sulawesi. Bandung: Geological Research and Development Centre.
Miller, D. M., Goldstein, S. L. & Langmuir, C. H. (1994). Cerium/ Sukamto, R. (1986). Tectonik Sulawesi Selatan dengan acuan khusus
lead and lead isotope ratios in arc magmas and the enrichment of ciri-ciri himpunan batuan daerah Bantimala. Dissertation, ITB,
lead in the continents. Nature 368, 514–520. Bandung.
Milsom, J., Ali, J. & Sudarwono (1999). Structure and collision history Sun, S.-s. & McDonough, W. F. (1989). Chemical and isotopic sys-
of the Buton continental fragment, Eastern Indonesia. AAPG Bulletin, tematics of oceanic basalts: implications for mantle composition and
83, 1666–1689. processes. In: Saunders, A. D. & Norry, M. J. (eds) Magmatism in the
Milsom, J., Thurow, J. & Roques, D. (2000). Sulawesi dispersal and Ocean Basins, Geological Society, London, Special Publications 42, 313–345.
evolution of the Northern Banda Arc. Jakarta: Indonesian Petroleum Taylor, S. R. & McLennan, S. M. (1985). The Continental Crust: its
Association. Indonesian Petroleum Association, Proceedings of the 27th Annual Composition and Evolution. Oxford: Blackwell Scientific.
Convention, pp. 495–505. Thirlwall, M. F., Graham, A. M., Arculus, R. J., Harmon, R. S. &
Morris, J. D., Jezek, P. A., Hart, S. R. & Gill, J. B. (1983). The Macpherson, C. G. (1996). Resolution of the effects of crustal
Halmahera island arc, Molucca Sea collision zone, Indonesia: a assimilation, sediment subduction, and fluid transport in island arc
geochemical survey. In: Hayes, D. E. (ed.) The Tectonic and Geologic magmas: Pb–Sr–Nd–O isotope geochemistry of Grenada, Lesser
Evolution of Southeast Asian Seas and Islands. Washington, DC: American Antilles. Geochimica et Cosmochimica Acta 60, 4785–4810.
Geophysical Union, pp. 373–387. Turner, S., Hawkesworth, C., Rogers, N., Bartlett, J., Worthington,
Mukasa, S. B., McCabe, R. & Gill, J. B. (1987). Pb-isotopic compositions T., Hergt, J., Pearce, J. & Smith, I. (1997). 238U–230Th disequilibria,
of volcanic rocks in the West and East Philippine island arcs: presence magma petrogenesis, and flux rates beneath the depleted Tonga–
of the Dupal isotopic anomaly. Earth and Planetary Science Letters 84, Kermadec island arc. Geochimica et Cosmochimica Acta 61, 4855–4884.
153–164. Van Bergen, M. J., Vroon, P. Z. & Hoogewerff, J. A. (1993). Geo-
Nichols, G. T., Wyllie, P. J. & Stern, C. R. (1996). Experimental chemical and tectonic relationships in the east Indonesian arc–
melting of pelagic sediment, constraints relevant to subduction. In: continent collision region: implications for the subduction of the
Bebout, G. E., Scholl, D. W., Kirby, S. H. & Platt, J. P. (eds) Australian passive margin. Tectonophysics 223, 97–116.
Subduction: Top to Bottom. American Geophysical Union Special Publication Van Leeuwen, T. M. (1981). The geology of southwest Sulawesi with
96, 293–298. special reference to the Biru area. In: Barber, A. & Wiryosujono, S.
Parkinson, C. D. (1991). The petrology, structure and geologic history (eds) The Geology and Tectonics of Eastern Indonesia. Bandung: Geological
of the metamorphic rocks of Central Sulawesi. Ph.D. thesis, Uni- Research and Development Centre, Special Publication 2, pp.
versity of London. 277–304.
605
Vroon, P. Z., Van Bergen, M. J., White, W. M. & Varekamp, J. C. White, W. M. & Patchett, J. (1984). Hf–Nd–Sr isotopes and in-
(1993). Sr–Nd–Pb isotope systematics of the Banda Arc, Indonesia: compatible element abundances in island arcs: implications for
combined subduction and assimilation of continental material. Journal magma origins and crust–mantle evolution. Earth and Planetary Science
of Geophysical Research 98, 22349–22366. Letters 67, 167–185.
Vroon, P. Z., Van Bergen, M. J., Klaver, G. J. & White, W. M. Wilson, M. E. J. (2000). Tectonic and volcanic influences on the
(1995). Strontium, neodymium, and lead isotopic and trace-element development and diachronous termination of a Tertiary tropical
signatures of the East Indonesian sediments: provenance and im- carbonate platform. Journal of Sedimentary Research 70, 310–324.
plication for Banda Arc magma genesis. Geochimica et Cosmochimica Wilson, M. E. J. & Bosence, D. W. J. (1996). The Tertiary evolution
Acta 59, 2573–2598. of South Sulawesi: a record in redeposited carbonates of the Tonasa
Vroon, P. Z., Van Bergen, M. J. & Forde, E. J. (1996). Pb and Nd Limestone Formation. In: Hall, R. & Blundell, D. J. (eds) Tectonic
isotope constraints on the provenance of tectonically dispersed Evolution of Southeast Asia. Geological Society, London, Special Publications
continental fragments in east Indonesia. In: Hall, R. & Blundell, D. 106, 365–389.
J. (eds) Tectonic Evolution of Southeast Asia. Geological Society, London, Special Woodhead, J. D. & Johnson, R. W. (1993). Isotopic and trace-element
Publications 106, 445–453. profiles across the New Britain island arc, Papua New Guinea.
Wakita, K., Sopaheluwakan, J., Miyazaki, K., Zulkarnain, I. & Munasri Contributions to Mineralogy and Petrology 113, 479–491.
(1996). Tectonic evolution of the Bantimala Complex, South Su- Yuwono, Y. S. (1987). Contribution à l’étude du volcanism potassique
lawesi, Indonesia. In: Hall, R. & Blundell, D. J. (eds) Tectonic Evolution de l’Indonésie, exemples du Sud-Ouest de Sulawesi et du volcan
of Southeast Asia. Geological Society, London, Special Publications 106, Muria ( Java). Thèse de troisième cycle, Université de Bretagne
353–364. Occidentale, Brest.
Webb, A. W., Thomson, B. P., Blissett, A. H., Daly, S. J., Flint, R. B. Yuwono, Y. S., Bellon, H., Soeria-Admadja, P. & Maury, R. C. (1985).
& Parker, A. J. (1986). Geochronology of the Gawler Craton, South Neogene and Pleistocene volcanism in South Sulawesi. Proceedings
Australia. Australian Journal of Earth Sciences 33, 119–143. Ikatan Ahli Geologi Indonesia 14, 169–179.
Wheller, G. E., Varne, R., Foden, J. D. & Abbott, M. J. (1987). Yuwono, Y. S., Maury, R., Soeria-Atmadja, R. & Bellon, H. (1988).
Geochemistry of Quaternary volcanism in the Sunda–Banda arc, Tertiary and Quaternary geodynamic evolution of South Sulawesi:
Indonesia, and three-component genesis of island-arc basaltic constraints from the study of volcanic units. Geologi Indonesia Jakarta
magmas. Journal of Volcanology and Geothermal Research 32, 137–160. 13, 32–48.
606
View publication stats

581 Full

Uploaded by

Copyright:

Available Formats

581 Full

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

581 Full

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Origin of Geochemical Variability by Arc-Continent Collision in the Biru Area,

Article in Journal of Petrology · April 2002

Marlina Elburg Theo van Leeuwen

SEE PROFILE SEE PROFILE

The user has requested enhancement of the downloaded file.

Origin of Geochemical Variability by

MARLINA A. ELBURG1∗, THEO VAN LEEUWEN2, JOHN FODEN1

∗Corresponding author. E-mail: marlina.elburg@adelaide.edu.au  Oxford University Press 2002

the lower member of the Camba Formation, which crops

(m) contact contact

12–30% 65–70% dykes glass (0–45%) volcaniclastics)

(30–85%) groundmass) cc, tr

Sample Rock type Material %K 40

Table 4: Selected whole-rock analyses

Ulubila Ulubila Ulubila Bila Bila Kahu Kahu Kahu

SiO2 55·33 62·64 55·08 42·28 42·52 50·33 59·57 49·99

Lemo Lemo Lemo Marara Marara Pake Sopo Sopo

SiO2 57·96 57·57 57·3 58·18 58·56 50·49 49·84 48·51

Biru Biru Langi Langi Langi B Langi B Standard Standard

SiO2 55·6 56·11 64·92 66·7 54·74 57·87

n.d., not detected. Fe2O3∗, all iron expressed as Fe2O3.

206 207 208 206 207 208 206 207 208

Ulubila Volcanics B 350 0·707950 0·000010 0·707898 0·512478 0·000007

fc, fractionation corrected; l/r, leachate or repeat.

Fig. 5. Chondrite-normalized REE patterns for the same samples as

against crustal contamination (in an AFC-type process)

broad temporal changes are also reflected in some aspects

Sulawesi samples never approach the very ‘continental’

Fig. 10. Variation of 143Nd/144Nd vs 206Pb/204Pb (initial) ratio, showing

of, for instance, the Lemo Volcanics by this type of

MORB mantle Sediment Mix A1 + B Fluid-modified Fluid-modified

Nd (ppm) 0·73 20 1·7 0·73 2

View publication stats

You might also like