THE GEOLOGICAL SOCIETY

OF AMERICA

Special Paper 400

in New Guinea: The Geotectonics

of Subducting Slab Breakoff

by Mark Cloos
Benyamin Sapiie
Andrew Quarles van Ufford
Richard J. Wei land
Paul Q. Warren
Timothy P. McMahon
Copyright © 2005, The Geological Society of America, Inc. (GSA). All rights reserved. GSA grants
permission to individual scientists to make unlimited photocopies of one or more items from this volume
for noncommercial purposes advancing science or education, including classroom use. For permission to
make photocopies of any item in this volume for other noncommercial, nonprofit purposes, contact the
Geological Society ofAmerica. Written permission is required from GSA for all other forms of capture
or reproduction of any item in the volume including, but not limited to, all types of electronic or digital
scanning or other digital or manual transformation of articles or any portion thereof, such as abstracts,
into computer-readable and/or transmittable form for personal or corporate use, either noncommercial
or commercial, for-profit or otherwise. Send permission requests to GSA Copyright Permissions,
3300 Penrose Place, P.O. Box 9140, Boulder, Colorado 80301-9140, USA.

Copyright is not claimed on any material prepared wholly by government employees within the scope of
their employment.

Published by The Geological Society of America, Inc.

3300 Penrose Place, P.O. Box 9140, Boulder, Colorado 80301 -9140, USA
www.geosociety.org

Printed in U.S.A.

GSA Books Science Editor: Abhijit Basu

Lib1·a•-y of Congress Cataloging-in-Publication Data

Collisional delamination in New Guinea : the geotectonics of subducting slab breakoff I

by Mark Cloos ... [et al.]
p. em. -(Special paper ; 400)
Includes bibliographical references.
ISBN 0-8137-2400-7 (pbk.)
1. Subduction zones--New Guinea. 2. Slabs (Structural geology)--New Guinea. 3.
Geology, Structural--New Guinea. 4 . Orogeny--New Guinea. I. Cloos, Mark. II
Special papers (Geological Society of America) ; 400

QE5 11 .46 .C65 2005

551.1 '36--dc22
2005055351

Cove•·: View looking west from near Puncak Jaya along crest of Central range, west New Guinea.
Folded, Late Miocene New Guinea Limestone Group strata are in the foreground . The Grasberg giant
porphy ry copper-gold deposit is in the background . Photo by Andrew Q uarles van Ufford .

10 9 8 7 6 5 4 3 2 1

11
 Geological Society of America
Special Paper 400
2005

Collisional delamination in New Guinea:

The geotectonics of subducting slab breakoff
Mark Cloos
Benyamin Sapiie
Andrew Quarles van Ufford
Richard J. Weiland
Paul Q. Warren
Timothy P. McMahon
Department of Geological Sciences, University of Texas, Austin, Texas 78712, USA

ABSTRACT

The spine of the island of New Guinea, the Central Range, is a 1300 km long by
100 to 150 km wide mountain belt with numerous peaks over 3 km elevation. This
mountain belt began to form when the Australian passive margin entered a north-
dipping subduction zone in the Middle Miocene. Regional relationships and fieldwork
near the Ertsberg (Gunung Bijih) mining district in the western Central Range are
the basis for making a detailed reconstruction of the events leading up to, and during,
collisional orogenesis.
Most of New Guinea can be divided into five lithotectonic belts. From north to
south, these are an accreted arc terrane, an upturned forearc basement, an underly-
ing metamorphic belt, a highlands fold-and-thrust belt, and a foreland basin. In west-
ern New Guinea, the accreted arc/forearc is the northwestern end of the Melanesian
Arc Terrane that was the product of two phases of subduction volcanism since the
Eocene. This terrane is largely buried under debris shed northwards from the Cen-
tral Range. The crystalline leading edge of the accreted arc/forearc terrane, the Irian
Ophiolite Belt of Jurassic age, is upturned forearc basement forming the north flank
of the Central Range. The ophiolite is underlain by the Ruffaer Metamorphic Belt,
which comprises continental rise and slope sediments and probably some trench axis
deposits that underwent subduction deformation and metamorphism since the Early
Miocene. The metamorphic belt grades into the highlands fold-and-thrust belt, which
contains carbonate shelf strata at least as young as 15 Ma. Kilometer-scale, angular
to rounded folds are the dominant structures. Regional sedimentologic relationships
indicate the highlands area has constituted a 500+ km long landmass since ca. 12
Ma. The southern flank of the western Central Range is a giant 300 km long by 30
km wide basement block that has been thrust southwards since 8 Ma, forming the
Mapenduma anticline. Minor, but widely distributed, magmatism occurred along the
spine of the western highlands from ca. 7.5 to 2.5 Ma. There is abundant evidence for
minor left-lateral strike-slip faulting subparallel to the upturned bedding that was
concurrent with igneous activity at 4–3 Ma.
These relationships, combined with consideration of the mechanical properties
of the crust and lithospheric mantle, are the basis for the construction of a series of
lithospheric-scale cross sections illustrating the process of collisional delamination.
Subduction tectonism and metamorphism began at ca. 30 Ma. Underthrusting of

Cloos, M., Sapiie, B., Quarles van Ufford, A., Weiland, R.J., Warren, P.Q., and McMahon, T.P., 2005, Collisional delamination in New Guinea: The geotectonics
of subducting slab breakoff: Geological Society of America Special Paper 400, 51 p., doi: 10.1130/2005.2400. For permission to copy, contact editing@geosociety.
org. ©2005 Geological Society of America.

1
2 M. Cloos et al.

Australian continental margin sediments was well under way by ca. 15 Ma, when
small isolated islands emerged. Bulldozing of the sediment cover formed an elongate
landmass by ca. 12 Ma, and siliciclastic sediment was shed southwards, overwhelm-
ing carbonate shelf sedimentation. Collisional orogenesis due to the jamming of the
subduction zone and initiation of thick-skinned crust-involved deformation began at
ca. 8 Ma. Magma generation due to asthenospheric upwelling and decompression of
stretched lithospheric mantle occurred from ca. 7.5 to 3 Ma. Contractional deforma-
tion in the western highlands ended at ca. 4 Ma, when this region became a site of
minor northwest-striking, left-lateral strike-slip faulting. Since ca. 2 Ma, offset has
been localized along the Yapen fault zone near the north coast of the island.
Collisional delamination involved the decapitation of the crust, continued sinking
of the subducted lithosphere, and the upwelling of asthenosphere into the rupture as
fast as it separated. This ensuing adiabatic decompression melting manifested itself
as a short-lived magmatic event and up to 2.5 km of vertical uplift, both centered on
the spine of the collision-generated orogenic belt. Collisional orogenesis is still under
way beneath the eastern Central Range, with delamination-generated magmatism in
its waning stages. Starting at ca. 8 Ma, the tear rupturing the subducting Australian
lithosphere propagated from west to east at a rate of ~150 km/m.y.

Keywords: collision, delamination, New Guinea, subduction, orogeny.

 Geological Society of America
Special Paper 400
2005

Chapter 1
Geology of the Central Range of Western New Guinea

INTRODUCTION The Central Range of New Guinea began to form when

the Australian passive margin entered a north-dipping sub-
The spine of the island of New Guinea, the Central Range, duction zone in the Middle Miocene. Convergence resulted
is a major orogenic belt that formed as the result of the con- in the emplacement of a forearc basement complex on top of
vergence between the Australian and Pacific plates (Fig. 1). In deformed passive margin strata and Australian continental
fact, Wegener (1924, p. 67–70) described the island as a site of crust. Distinct stages in the orogeny are dated from regional
Cenozoic mountain building due to continental drift. Dewey sedimentologic, metamorphic, and magmatic relationships.
and Bird (1970) considered New Guinea as the type locality of Australian-Pacific plate motions place tight constraints on the
recent island arc–continent collision. Tectonism has occurred kinematic evolution of the island during the Cenozoic. These
on a grand scale with the New Guinea orogen, comparable in observations, combined with consideration of the mechanical
size to the western U.S. Cordillera (Silver and Smith, 1983). properties of oceanic and continental lithosphere, are the basis

132o E 136o E 140o E 144o E 148o E 152o E

gh

Pacific Ocean Pacific

Trou

Mussau T
0 100 200 mi < ~ 4 Ma
Pacific/Caroline ~25
2
o
Ayu

r at o
0 100 200 300 km / yea 43
Manokwar Plate 11 c
m
Ma t ~2
i Tr ra
~4
oug ea

rou
o h ~ 5 - 3 Ma /y
0 Manus Trench
+ to cm 0o
30

gh
10 cm / year at ~270o? 9 .5
SFZ New Ki
Guin
ea T
renc
li
na
il
Plate
1000 YFZ h u
a
Salawati m T
re
We n
Basin wak
h

c
y asi

Seg Bismarck h
men
Ba aw

Bintinui Basin
Se t Sea
dr

ra IOB Melane MDZ

en

m s i an
C

T WO Arc
LF

ro BTFZ
u B
T

g
h Pk. Jaya, Ter
+ 4884 m
10 IDI r a ne
4o S Irian Ad 4o S
TA F Z F T B 00 m el
BD be
MA rt Bismarck Plate
gh

Banda u
DM Pa
ou

R
Sea Ertsberg pu M
FZ
Mining an
r

Cen u Fi
u T

District tra M KA nn
l R UA T is
F te
Ar

an B h
rr c
EXPLANATION ge e
re
n
Fo i n T
re ita
Ophiolites and ultramafic rocks Br
la New Solomon
Approximate limit of underthrust
Kepulauan Arafura Indonesia n Sea
Ta s m

Tr
TanimbarAustralian continental crust Sea d ob
ria
10

nd o
ug 8 S
Papua New Tro
00

Exposed continental basement POB h

an

Guinea
m

Inactive deformation front

AF

Slow convergence (< 2 cm/yr)

Australian Plate
TB

Rapid convergence (> 2 cm/yr)

Active spreading oceanic ridge
Precambrian Phanerozoic
L
in

Active strike-slip fault zone Wood-

basement basement
e

lark
Inactive strike-slip fault zone
Relative plate motion vector with Coral Rise
respect to the Australian plate
140o E Aust. 144o E Sea 148o E 152o E

Figure 1. Tectonic map of New Guinea, adapted from Hamilton (1979), Cooper and Taylor (1987), Dow et al. (1988), and Sapiie et al. (1999).
AFTB—Aure fold and thrust belt, FTB—fold-and-thrust belt, IOB—Irian Ophiolite Belt, TFB—thrust-and-fold belt, POB—Papuan Ophiolite
Belt, BTFZ—Bewani-Torricelli fault zone, MDZ—Mamberamo deformation zone, YFZ—Yapen fault zone, SFZ—Sorong fault zone, WO—Wey-
land overthrust. Continental basement exposures are concentrated along the southern flank of the Central Range: BD—Baupo Dome, MA—Map-
enduma anticline, DM—Digul monocline, IDI—Idenberg Inlier, MUA—Mueller anticline, KA—Kubor anticline, LFTB—Legguru fold-and-thrust
belt, RMFZ—Ramu-Markham fault zone, TAFZ—Tarera-Aiduna fault zone. The Tasman line separates continental crust that is Paleozoic and
younger to the east from Precambrian to the west.

3
4 M. Cloos et al.

for drawing a series of scaled diagrams illustrating the process (DeMets et al., 1990, 1994). The present-day pole of rotation for
of collisional delamination beneath New Guinea. Pacific-Australian motions is located near New Zealand. Con-
This report is divided into three chapters. Chapter 1 sum- sequently, along the 2000 km length of the Australian-Pacific
marizes the geology of the western Central Range in the area plate boundary in the New Guinea region, the calculated vec-
where we made our regional transect. Chapter 2 concerns the tor for current relative motion varies by <5° in direction and <5
mechanics of collisional delamination—the process we envi- km/m.y. in speed.
sion to account for the pattern—and the timing of deforma- Global plate reconstructions indicate that the relative
tion, metamorphism, sedimentation, and magmatism. These motion between the Australian and Pacific plates changed in the
concepts should be applicable to all similar collisional settings. latest Cenozoic. This recent change in Pacific plate motion has
Chapter 3 presents a series of scaled, lithospheric-scale cross been estimated as occurring at 5 Ma based upon the bend in
sections illustrating the late Cenozoic tectonic evolution of the Hawaiian seamount chain at Oahu (Cox and Engebretson,
western New Guinea. 1985). Analysis of deviations in magnetic anomalies along the
East Pacific Rise indicates the change dates between 5.0 and
GEOLOGY OF WESTERN NEW GUINEA 3.2 Ma (Pollitz, 1986), most specifically at 3.9 Ma (Harbert and
Cox, 1989, 1990). As will be discussed, we believe this change
Geologic Framework is a direct result of the collision that formed the Central Range of
New Guinea. The global plate reconstructions of Scotese et al.
The outline of the island of New Guinea has been described (1988) indicate that the relative Australian-Pacific plate motion
as similar to a bird flying westward with an open mouth (Fig. in the New Guinea region changed in azimuth by ~9° clockwise
1). As a result, the island is commonly geographically divided and increased in speed by ~15 km/m.y. (see arrows on Fig. 1). In
into the Bird’s Head, Neck, Body, and Tail regions. The northern other words, the current local relative motions are more oblique
half of the island is underlain by the Melanesian Arc Terrane that to the strike of the Central Range and slightly faster than they
was built on Mesozoic ocean crust. The southern half is underlain were during the mountain-forming collision.
by Australian continental crust. The collisional welding of the The ocean crust in most of the west-central Pacific Basin is
oceanic terrane onto the Australian continental margin generated Jurassic; however, crust north of New Guinea is early Cenozoic.
the Central Range, the 1300 km by 100 to 150 km mountain- The origin of this area of seafloor is a tectonic conundrum, and it
ous spine with numerous peaks over 3 km elevation. The range occurred before the time period of primary interest in this report.
stretches from the Bird’s Neck (~135° E) to the Bird’s Tail (Papuan More significantly, most of this area of anomalously young sea-
Peninsula, ~146° E) (Dewey and Bird, 1970; Hamilton, 1979). floor was named the Caroline plate, and it has a debatable move-
The northern accreted arc/forearc terrane is largely buried ment history (Weissel and Anderson, 1978). Hegarty et al. (1983)
beneath sediments shed northwards from the Central Range (Fig. identified the Mussau Trough as the eastern boundary between
1). The crystalline leading edge (forearc basement) of this sub- the Pacific and Caroline plates. They concluded that there has
duction system, the Irian Ophiolite Belt, is upturned and exposed been at most a few tens of kilometers of convergence at this
along the lower slopes of the north flank of the Central Range. location, but that it is currently inactive or nearly so (Hegarty
The southern half of the island is composed of rocks depos- and Weissel, 1988) (see Fig. 1). Thus, the Caroline plate (or bet-
ited on top of Australian continental basement. The western ter, microplate) must have moved in a direction and at a rate that
Central Range can be subdivided, from north to south, into three is nearly the same as that of the Pacific plate. Cloos (1992b) con-
parts: the Ruffaer Metamorphic Belt, the highlands fold-and- cluded that the Caroline microplate is simply a broken corner
thrust belt, and the foreland basin. The metamorphic and fold- of the Pacific plate that was a distinct kinematic entity for only
and-thrust belts are composed of deformed Australian passive a short period, between ca. 5 and 3 Ma. As will be discussed,
margin strata. The southern flank of the range contains several this is considered to be a globally minor, but locally significant,
windows into uplifted continental basement (Fig. 1). The fore- adjustment in the collisional tectonic movements that generated
land basin contains a thick wedge of late Cenozoic siliciclastic the Central Range. Movement of the Caroline microplate caused
debris that was shed southwards. These strata blanket carbon- a short-lived episode of strike-slip faulting in the core of the
ate shelf deposits on top of transitional continental crust that was western highlands, which aided the ascent of magmas (Sapiie
thinned during Mesozoic rifting. and Cloos, 2004). However, compared to the many tens of kilo-
meters of contractional deformation prior to and during the col-
Cenozoic Plate-Tectonic Setting lisional orogenesis, a few kilometers of strike-slip movements
are very minor.
The Central Range of New Guinea is almost entirely the
product of late Cenozoic convergence between the Australian Active Tectonism
and Pacific plates. Since ca. 3 Ma, the relative tectonic motion
between these plates for a location in the middle of the Cen- Most of the seismicity occurring beneath the island of New
tral Range is at a rate of 114 km/m.y. along an azimuth of 252° Guinea is clustered near the Huon Peninsula, the Mamberamo
Figure 2. A: The distribution of earthquake epicenters in the New Guinea region for all depths with magnitude M > 4 from 1977 to 1997. Note the very
limited seismicity beneath areas with elevations higher than 1000 m. Source data from the Institute for Research in Seismology (IRIS) catalog, June 1997.
Modified from Sapiie et al. (1999). B: Movement detected during global positioning system (GPS) survey from 1991 to 1997 by Stevens et al. (2002).
The published survey data are only for a six-year period, but the basic kinematics are apparent. The Bird’s Head block with stations at Biak, Manokwari,
and Sorong are moving in directions that parallel those of the Pacific plate. The stations at Timika, Wamena, Merauke, Aru, and Saumlakki are moving
with the Australian plate. Fak Fak is moving nearly parallel to the Biak station but slower. The difference between Biak and Yapen indicates ~5 cm/yr
of left-lateral motion along the Yapen fault zone. The difference between the Kaimana and Timika/Aru stations indicates ~6 cm/yr of left-lateral motion
along the Tarera-Aiduna fault zone. The station at Sentani has a speed and direction of movement that is intermediate to the two major plates, indicating
that some (most?) of the relative motion is accommodated along the eastern segment of the New Guinea Trench (Wewak segment), and the Mamberamo
deformation zone. NGT—New Guinea Trench, BTFZ—Bewani-Torricelli fault zone, MDZ—Mamberamo deformation zone, YFZ—Yapen fault zone,
ST—Seram Trough, TAFZ—Tarera-Aiduna fault zone, AT—Aru Trough.
6 M. Cloos et al.

region, and the Bird’s Neck (Fig. 2A). Notably, there are few A. Depth < 50 km
earthquakes beneath areas with elevation higher than 1000 m. 3°N
Strike-slip Thrust Normal
Abers and McCaffrey (1988) found that the 18 largest events
Pacific Ocean
in their data set were located beneath central and eastern New
0° Mamberamo
Guinea. The large events indicate west-trending, left-lateral, Region
strike-slip faulting near the south end of the Mamberamo region 87161
81229
78115
8207879175
and deep-seated (20 to 50 km deep) west- to northwest-striking 3°S 91186 79114 82088

high angle (up to 65°) reverse faulting along the southern flank of
Pk. Jaya
4884 m

the eastern Central Range (Fig. 3).

1000
m 8723
8615
Bird's 88163
8415
85154
8416
87075
87075
87075 8609
84153
6°S Neck 86043
78203
86043
80177
The Mamberamo region, just west of the international border
90342
88254
87323 85238
88268

and north of the Central Range, is an irregular swampy landscape Aru Arafura
Indonesia
Papua New
Guinea
Trough
with active fields of mud volcanoes (Williams et al., 1984). At 9°S
Sea

depths less than 50 km, earthquake focal mechanisms are domi- Coral
nated by northwest-trending thrust mechanisms (Fig. 3A). Some Sea
12°S
earthquakes occur as deep as ~100 km, and they have a variety
of focal mechanisms (Fig. 3B). We believe this region is a broad Australia 0 300 km
convergent bend, which we term the “Mamberamo deformation 15°S
zone” connecting the Bewani-Torricelli and Yapen left-lateral 130°E 135°E 140°E 145°E 150°E

strike-slip fault zones (Fig. 4) (Sapiie et al., 1999). B. Depth 50 - 100 km

Collisional deformation is ongoing in eastern New Guinea 3°N
Strike-slip Thrust Normal
with rapid uplift forming the famous raised terraces on the Finn-
isterre/Huon Peninsula (Chappell, 1974; Abbott et al., 1994a). 0° Pacific Ocean
The uplifting peninsula is the western end of the forearc block to
the New Britain subduction zone (Davies et al., 1987a; Silver et
al., 1991; Abbott et al., 1994b). The eastern Central Range high- 3°S
Pk. Jaya

lands, located in front of this actively uplifting forearc terrane, 78102

87168
4884 m

1000
87021
are being slowly pushed southwards above a décollement that
m

77191 86288
82073
6°S 2125
78283
7916480264
89149

surfaces near the south flank of the mountain belt.

251
2326 Arafura
Sea
GPS measurements for the period 1991–1997 provide Indonesia Papua New
Guinea

important new information on the active plate kinematics. Sta- 9°S

tions at Sorong and Biak are moving with the Pacific plate, and Coral
stations in Merauke, Timika, and Wamena are moving with the 12°S
Sea

Australian plate (Fig. 2B) (Puntodewo et al., 1994; Stevens et al.,

2002). The stations with movements that are most different from Australia 0 300 km
either the Pacific or Australian plates are the ones at Sentani and 15°S
130°E 135°E 140°E 145°E 150°E
Yapen near the north coast. Relative to the Australian plate, these
stations are moving slowly southwards, indicating that Pacific- Figure 3. Earthquake focal mechanisms in the New Guinea region
from 1987 to 1997. A: Depths <50 km. Note the occurrence of north-
Australian relative movement is localized out to sea along the west-trending thrust-type focal mechanisms under the Mamberamo re-
New Guinea Trench. Convergent movements appear to be fastest gion and the abundance of normal-type mechanisms in the Aru Trough
along the eastern part of the New Guinea Trench, north of the and to the north in the Bird’s Neck region. Thrust-type mechanism are
Mamberamo region, and decreasing to zero to the west. found along the southern flank of the Central Range in Papua New
The GPS data clearly indicate that the Bird’s Head region Guinea. B: Depths 50–100 km. Nearly all of the “deep” events under
New Guinea are shallower than 75 km. Note the concentration under
is moving with the Pacific plate. The southern boundary of this the Mamberamo area and the large variations in orientation and type.
region is the highly active Tarera-Aiduna fault zone (TAFZ, Fig. Source data from Harvard Centroid Moment Tensor (CMT) catalog,
4) with left-lateral strike-slip offsets (McCaffrey and Abers, June 1997. Modified from Sapiie et al. (1999).
1991). Subsidiary movements form a band of normal faulting to
the south along the Aru Trough and more diffuse normal faulting
to the north (Fig. 3) (Sapiie et al., 1999). Extremely oblique con- were concentrated in the western part of the range, and geologic
vergence is occurring along the Seram Trough (Fig. 4) (Cardwell relations for a transect in this area are summarized below, as are
et al., 1980). seismic and GPS data that indicate active tectonic movements.
The overall pattern of active movements indicates that the There are several geologic factors that make the tectonics of the
convergence that formed the mountainous spine of the island eastern Central Range somewhat different from that of the west-
has largely, if not entirely, ended in the western part of the Cen- ern part, and they will be discussed later. The conclusion is that
tral Range, but is ongoing in the eastern part. Our field studies the collisional orogeny forming the spine of the island is time-
 Geology of the Central Range of Western New Guinea 7

Figure 4. Seismotectonic interpretation of New Guinea. Tectonic features include: PTFB—Papuan thrust-and-fold belt, RMFZ—Ramu-Markham fault
zone, BTFZ—Bewani-Torricelli fault zone, MDZ—Mamberamo deformation zone, SFZ—Sorong fault zone, YFZ—Yapen fault zone, RFZ—Ransiki
fault zone, TAFZ—Tarera-Aiduna fault zone, WT—Waipona Trough. Inactive strike-slip fault zones in the core of the Central Range: DFZ—Derewo
fault zone, TFZ—Tahin fault zone, LFZ—Lagaip fault zone. Arrows show relative motion of the Pacific plate and short-lived Caroline microplate with
respect to the Australian plate. Modified from Sapiie et al. (1999) in light of the results of the GPS survey of Stevens et al. (2002).

transgressive. In eastern New Guinea, the orogeny is now at the est because two giant porphyry copper/gold ore districts have so
same stage as the event to the west ~3–4 m.y. ago. A brief discus- far been discovered in intrusive bodies associated with this volca-
sion is now required of a most distinctive aspect of the Central nism (Fig. 5). The largest is the Grasberg and associated orebod-
Range orogeny: collision-related volcanism along the spine of ies, which formed at 3 Ma in the western Central Range (van
the highlands. Nort et al., 1991; MacDonald and Arnold, 1994). Another major
gold deposit formed at Porgera at ca. 6 Ma (Fig. 5). Smaller, but
MAGMATISM IN THE HIGHLANDS no less significant because it indicates ore-forming processes
occurred along the length of the highlands, is the Ok Tedi deposit
The presence of volcanic rocks in the highlands of New formed at ca. 1.5 Ma near the international border in Papua New
Guinea has long been known, but little studied because of the Guinea (Rush and Seegers, 1990).
difficulty of access and deep weathering. This magmatism, gen- Neogene igneous rocks in central New Guinea (136° E to 147°
erally of intermediate composition, is small in volume, but of E) occur along the spine of the Central Ranges (Fig. 5). In Papua
widely scattered occurrence. It is of considerable economic inter- New Guinea and just across the border into Indonesia, Dow (1977)
8 M. Cloos et al.

Figure 5. Late Cenozoic magmatism in New Guinea. There are two distinct magmatic provinces. Miocene magmatism from 20 to 10 Ma is known
as the Maramuni Arc and intrudes Australian continental basement in eastern New Guinea (Dow, 1977). Similar-age igneous rocks in western
New Guinea are only found in the allochthonous Weyland overthrust and along the western end of the Irian Ophiolite Belt. The younger magmatic
province dates since 7 Ma and is concentrated along the spine of the Central Range. Many Pleistocene volcanic centers, the Fly-Highlands province
(FHP), are present at the eastern end of the Central Range. Major copper and gold ore deposits formed during this phase of magmatism at Grasberg/
Ertsberg, Ok Tedi, and Porgera. Note the lower elevations and lack of magmatism between ~138° E and 140° E in the Central Range. POB—Papuan
Ophiolite Belt (dashed outlines), IOB—Irian Ophiolite Belt (dashed outlines), WO—Weyland overthrust, MV—Minjauh volcanic field. Modified
from McMahon (2001). YFZ—Yapen fault zone, MDZ—Mamberamo deformation zone, BTFZ—Bewani-Torricelli fault zone.

recognized a belt of magmatism into deformed Australian con- onto, the Melanesian Arc Terrane near the Bird’s Head probably
tinental basement. This belt is now divisible into a zone of 20–9 were translated into their present position by plate convergence.
Ma intrusive and volcanic rock, the Maramuni Arc, which occurs An origin 500 km or more to the northeast is possible.
north of a zone of younger than 7 Ma intrusive and volcanic In the western Central Range, nearly all of the intrusive and
rocks, the Fly-Highlands province (McMahon, 2000a, 2000b). volcanic rocks crop out between 136° E and 137° 50′ E. Dated
Nearly 500 km to the west, in and next to the Bird’s Neck, a bodies range from 7.5 to 2.5 Ma (Dow, 1968; Page, 1975; Dow
similar pattern is seen. However, unlike the occurrence in Papua et al., 1986; O’Connor et al., 1994; Parris, 1994; McDowell et
New Guinea, the Maramuni-age igneous rocks, in the Weyland al., 1996; McMahon, 2000a, 2000b). From the eastern end of
overthrust (Dow et al., 1990) and westernmost end of the Irian this province to near the Papua New Guinea border, there is an
Ophiolite Belt (Weiland, 1999), occur within the allochthonous ~300 km long gap where minerals exploration geologists found
arc/forearc terrane. In addition, there are small occurrences of no Tertiary igneous rocks in outcrop or in float (G.V. O’Connor,
young (ca. 5 Ma) volcanic rocks near Nabire north of the Wey- 1996, personal commun.). Notably, the part of the highlands
land overthrust (Fig. 5) (Dow et al., 1986, 1988; Bladon, 1988). lacking Late Tertiary igneous activity is where elevations are
Maramuni magmatism into the continental basement of almost entirely less than 2000 m (Fig. 5).
eastern New Guinea was generated by southwest-dipping sub- In the western Central Range, this very latest Miocene
duction along the Trobriand Trough (Davies et al., 1987b) (Fig. 1). and Pliocene (hereafter Pliocene) volcanism has been best
The equivalent-age magmatic rocks that intruded into, or faulted studied within the Ertsberg mining district (McMahon, 1994a,
 Geology of the Central Range of Western New Guinea 9

1994b, 1994c) and in the Minjauh volcanic field ~30 km to the attempted to reconcile the puzzling tectonic geometry by con-
northwest (Fig. 5) (McMahon, 2001). The magmas intruded cluding that highlands magmatism is not subduction-generated,
and blanketed sedimentary rocks of the Australian continen- at least in the sense of conventional plate-tectonic theory. They
tal margin that were previously deformed during the collision. proposed that this magmatic event was triggered by a steepening
The igneous rocks postdate large-scale folding and some of the of the north-dipping subducted Australian plate to near-vertical.
faulting. Most of the intrusions have high K and LILE (large The mechanical picture is that tighter bending of the subducted
ion lithophile element) contents and low Nb and Ti contents plate caused fracturing and uplift that somehow induced melting
(McMahon, 2001). They are mineralogically and chemically of lithospheric mantle that had been enriched in volatiles during
similar to volcanic arc rocks. Isotopic studies show these mag- a period of Cretaceous subduction beneath the region.
mas incorporated an old, radiogenic component indicative of Hamilton (1979, p. 254) cited the highlands volcanism
Precambrian crust (Housh and McMahon, 2000). Most nota- as the primary evidence that convergence generated the New
bly, a few intrusions in the western Central Range are lam- Guinea Trench by subduction reversal in the Miocene. However,
proitic (McMahon, 2001), unusual compositions that indicate a the subduction reversal model creates several geodynamic enig-
source in lithospheric mantle (Housh and McMahon, 2000). mas. Why was magmatism so short-lived? Why is this magma-
In the eastern Central Range of Papua New Guinea, the tism so concentrated along the spine of the highlands? Subduc-
magmatic rocks in the Fly-Highlands province are of Plio- tion magmas are generated at depths of 100–120 km. It would be
Pleistocene age (Fig. 5). Distinct, glaciated volcanoes are fortuitous for the zone of magma generation to be centered over
still recognizable, and probable Quaternary activity suggests a 1000+ km strike length to precisely parallel the spine of the
several centers are only dormant. These volcanic complexes collision-generated mountain belt. Finally, why would magmas
largely blanket a basement of faulted and folded Australian apparently prefer to intrude the very highest parts of the moun-
passive margin sediments (Mackenzie, 1976; Dow, 1977), but tain belt?
some are tilted as a result of the ongoing tectonic movements A fundamental conclusion of this paper is that this magmatic
(Davies, 1990). The apparently greater abundance of late Neo- belt was not generated by steady-state subduction. Moreover, we
gene volcanism in the eastern highlands (Fig. 5) is not a true conclude that the concentration of magmatism along the axis of
indicator of a greater amount of volcanism. Wide areas are the highlands, and the highest part no less, should actually be
only thinly blanketed by young volcanic flows. A similar, but expected. Because of the tectonic importance placed on the New
slightly older, volcanic blanket in the western highlands would Guinea Trench by Hamilton (1979) and other workers, some dis-
have been removed by erosion. The map pattern in the eastern cussion of this feature is now warranted to clarify its geodynamic
Central Range is probably similar to that in the western Cen- significance.
tral Range at 3 Ma.
Most of the Plio-Pleistocene igneous rocks in Papua New NEW GUINEA TRENCH—A REACTIVATING RELICT
Guinea have arc-type major element compositions (Whalen et
al., 1982; Mackenzie and Johnson, 1984). The 6 Ma intrusions The New Guinea Trench is a bathymetric low that paral-
at the Porgera gold deposit are an exception, for they have trace lels the north coast of the island from ~134° 30′ E to ~144° E
element characteristics that are similar to basalts from intra- (Fig. 1). Most workers apparently agree that there is no obvi-
plate settings (Richards et al., 1990). Another difference is that ous south-dipping Wadati-Benioff seismic zone (e.g., Johnson et
isotopic data from Pleistocene volcanic rocks of the Fly-High- al., 1971; Hamilton, 1979; Cooper and Taylor, 1987; Milsom et
lands province show little, if any, evidence of contamination by al., 1992). For those who advocate recent subduction surfacing
Precambrian crust (Hamilton et al., 1983). This is expected, as at this bathymetric depression, convergence must be either very
the province is east of the Tasman Line (Plumb, 1979a, 1979b), slow, just ended, or just starting. The only deep earthquakes are
the boundary separating Precambrian and Phanerozoic base- scattered occurrences to 100 km depth under the Mamberamo
ment in the Australian continent (Fig. 5). region (Figs. 2A and 3). As discussed, this cluster of earthquakes
Igneous rocks younger than 7 Ma have a distinctive attri- is explainable as due to local convergence along a bend connect-
bute—most are found at or near the highest elevations on the ing the Bewani-Torricelli and Yapen left-lateral strike-slip fault
island (Fig. 5). There is a seeming lack of igneous activity in the zones (Fig. 4) (Sapiie et al., 1999).
middle of the range (138° E to 140° E), where the mountains are Hamilton (1979) showed seismic reflection profiles across the
typically less than 2 km high. We believe this observation has an trench between 141° E and 143° 40′ E (Wewak segment, Fig. 1)
explanation that will be discussed later. that indicate south-dipping underthrusting, and stated that unpub-
Several models have been proposed to explain the origin lished seismic profiles across the trench between 135° E and 139°
of the magmatism in the highlands. Hamilton (1973, 1979) E contained similar features. However, a seismic profile near 135°
argued for south-dipping subduction reversal following arc-con- E indicates that there has been no shortening of the sediment within
tinent collision. Ripper and McCue (1983), Cooper and Taylor the western end of the trench since the middle to late Miocene (Mil-
(1987), and Cullen and Pigott (1989) argued for south-dipping som et al., 1992), and this conclusion is now well supported by
subduction prior to arc-continent collision. Johnson et al. (1978) GPS data (Stevens et al., 2002). In sum, there is evidence for recent
10 M. Cloos et al.

convergence (a few tens of km?) along the eastern end of the New lands to Tembagapura near the Ertsberg mining district. There
Guinea Trench, but not at the western end. are 34 km of nearly continuous roadcut from the base of the
Even more problematical from a tectonics perspective is the mountain to near the crest of the range (Fig. 6). In the mining dis-
fact that the eastern end of the trench lies to the west of most trict, we analyzed more than 20 km of roadcut exposure. North
of the Pleistocene volcanoes of the Fly-Highlands province in of the mining district, where access was by helicopter insertion
Papua New Guinea (Fig. 5). Hamilton (1979) and Cooper and and foot transect, outcrop was largely limited to scattered cliff
Taylor (1987) suggested that south-dipping subduction on the faces, landslide scarps, and river and stream bottoms (Cloos,
New Guinea Trench was somehow linked to that on the Trobri- 1997a, 1997b). The stratigraphy along the access road and mine
and Trough in easternmost New Guinea, north of the Bird’s Tail area was first studied in detail by Martodjojo et al. (1975). For
(Fig. 1). This does not, however, explain the westward truncation the most part, the stratigraphy in our area of field investigation
of the Neogene Bismarck Arc at 145° E, which, in their models, correlates with that in the Bird’s Head and Neck regions a few
overrides the north-facing trench. This geometry is perplexing in hundred kilometers to the west (Pieters et al., 1983), the region
the context of any subduction reversal model. It will be shown of extensive field studies from 1976 to 1982 that were part of the
that magmatism related to north-dipping subduction of the Aus- Irian Jaya Geological Mapping Project, an Indonesia-Australia
tralian plate, in essence the geometry proposed by Johnson et al. collaboration (Dow et al., 1988).
(1978), is the only tectonic scenario proposed to date that is com- From south to north, the lithotectonic belts of the western
patible with all the field relations. Central Range are the foreland basin, the foothills, the Mapen-
In short, there is no kinematic evidence that 100+ km of duma anticline, the highlands fold-and-thrust belt, the Ruffaer
underthrusting occurred at the New Guinea Trench since ca. 10 metamorphic terrane, and the Irian Ophiolite Belt. These will
Ma—an age for subduction initiation required to explain the old- be briefly described below.
est volcanics in the highlands. In addition, a subduction reversal
model does not explain why supposed arc volcanism ended at Foreland Basin—Buried Continental Shelf
ca. 3 Ma in the western highlands, or the recent activity in the
eastern highlands. The foreland basin is a broad area of swampy jungle. A dis-
All of the above leads us to believe that Kroenke (1984) is tributary system of meandering channels near the coast changes
correct in deducing that the New Guinea Trench is a relict of an to shallow braided channels within a few tens of kilometers of the
earlier period of subduction. He concluded this trench was the mountain front. From 50 km (near Timika airport) to as much as
site of south-dipping subduction from the Eocene to early Mio- 100 km inland, the elevation is less than 100 m. The underlying
cene. This caused magmatism forming the now accreted Outer Australian continental margin was rifted in the Triassic (Veevers
Melanesian Arc Terrane that first started to form at ca. 43 Ma, et al., 1991), and from Jurassic to late Cenozoic time strata accu-
when Pacific plate motion changed and subduction began along mulated in a passive margin, slowly subsiding shelf environment
major fracture zones along the western Pacific from New Zea- (Pigram and Panggabean, 1984; Brown et al., 1979; Pigram and
land to Japan (Hilde et al., 1977; Parrot and Dugas, 1980; Casey Symonds, 1991). Uplifted masses of these rift and passive mar-
and Dewey, 1984). The seismic reflection profiles along the east- gin strata occur along the southern flank of the western Central
ern part of the New Guinea Trench (140° E to 145° E) indicate at Range (Fig. 1). Their lithologies record the northward movement
least some recent convergence (Hamilton, 1979). However, just a of Australia toward the equator (Quarles van Ufford, 1996). The
few kilometers of underthrusting can account for all observations passive margin sediments changed from dominantly siliciclastic
known to us. It is concluded that the eastern end of the trench has in the Jurassic to carbonate-bearing in the Late Cretaceous and
been reactivated as the latest tectonic adjustment associated with carbonate-dominated in the Cenozoic as the northern part of the
arc/forearc-continent collision. In other words, subduction rever- continent approached equatorial latitudes. Outcrops and petro-
sal (reactivation) is just beginning at the eastern end of the New leum exploration drill holes near the southern slope indicate that
Guinea Trench and has yet to produce arc magmatism. siliciclastic sediments flooded the shallow carbonate shelf begin-
ning at ca. 12 Ma along the 500+ km strike length that forms the
GEOLOGIC TRANSECT—WESTERN CENTRAL foreland basin (Quarles van Ufford and Cloos, 2005).
RANGE

Our field studies were centered in the core of the western

Central Range, in and near the Ertsberg (Gunung Bijih) min-
ing district, the location of several world-class copper and gold Figure 6. Geology in the Ertsberg (Gunung Bijih) mining district and
porphyry and gold orebodies that formed at 3 Ma (Fig. 1). We along the access road. Diagnostic fossil localities are marked. Zir-
consider it notable that these deposits are at elevations of 3–4 km cons yield late Precambrian fission track ages for three samples in the
greenschist facies slates of the Kariem Formation near the base of the
and within a few kilometers of Puncak Jaya (Carstenz Peak), the mountain. Note the northwest trend of kilometer-scale folds in the min-
highest peak between the Himalayas and the Andes. We mapped ing district. Modified from Quarles von Ufford (1996). HEAT—heavy
and sampled the section along the road from Timika in the low- equipment access trail.
Figure 7. A: Cross section A–A′ located on Figure 6. Note the homoclinal north limb of the Mapenduma anticline. The Ertsberg intrusion in
the mining district is dated at 3 Ma and crosscuts the kilometer-scale folds (McDowell et al., 1996). B: Interpreted movement history for the
formation of the Mapenduma anticline based upon apatite fission track thermochronology (after Weiland and Cloos, 1996). Initial thin-skinned
folding changed to thick-skinned basement-involved deformation. It is likely the Mapenduma anticline formed by reactivating a steeply dipping
Mesozoic normal fault. Formation abbreviations as on Figure 6.
 Geology of the Central Range of Western New Guinea 13

Foothills—Leading Edge of Central Range Orogenic Belt

Period
Epoch
The foothills, with elevations less than 500 m, are located
Lithology Formation Thickness
between the flat-lying foreland and the jagged front of the moun-
tain. West of the mine access road, the foothills include low-relief

Plio
v
folds (Sabins, 1983; Pigram and Panggabean, 1983). Along the v v

Buru Fm.

~6000 m
v
v v v Volc. ?
access road to the mining district, the foothills are underlain by v

Cenozoic
two broad terraces composed of gently south-dipping conglom-
eratic alluvium, the upper Buru Formation (Quarles van Ufford,
1996). Approximately 10 km to the west and the east, upturned

New Guinea
Kais Fm. 1100 m

LS Group
beds of lower Buru (ca. 12 to ca. 4 Ma) and older formations are

1800 m
Sirga Fm. 40 m
folded and exposed by erosion (Parris, 1994).

P EO
Faumai Fm. 300 m
The base of the Buru Formation (and equivalent formations
Waripi Fm. 350 m
to the west and east) are conformable with the shelf limestones of
the New Guinea Limestone Group (Bär et al., 1961; Pigram and Ekmai SS. 650±100 m

Cretaceous
Panggabean, 1983). Tilting of the lower Buru deposits indicates

Kembelangan Group
Piniya Fm.
the older foreland basin deposits were folded or overridden by 1550±300 m

4600±1000 m
thrust blocks. The flat-lying upper Buru Formation (younger than

E L
ca. 4 Ma) indicates the flanks of the mountain were locally buried Woniwogi
1000±200 m
Fm.
in conglomeratic debris (molasse) as movement slowed.

Mapenduma Anticline—Thick-Skinned Deformation L

Kopai Fm.
Jurassic
M 1400±300 m

The most prominent structure along the southern flank of

E

r r
the western Central Range (Fig. 1) is the 280°-trending, ~300
km long basement-cored Mapenduma anticline (Nash et al.,
Triassic

r
r Tipuma Fm. 2000±400 m
1993, Fig. 4 therein). This anticline is the only structure in the r

southwestern Central Range that exposes Precambrian or early r r

Paleozoic sediments and metasediments (Tuaba and Kariem
L

Formations) and metaigneous basement (Awitagoh Formation)

Dev Permian

(Fig. 7). Deep erosion has removed the south limb of the Map- Aiduna Fm. 1750±200 m
enduma anticline at the location of the mine access road. The
E

structure has been confirmed through field mapping to extend

200 m
for ~200 km to the east and 100 km to the west of the road (Par-
950±
L

B
ris, 1994). 2000±400 m
Modio Fm.
Biostratigraphic analyses have detected no stratigraphic 1050±
200 m
Silur

repetitions along the mine access road, where the 15 km wide, A

north-dipping homoclinal limb of the Mapenduma anticline is r r
exposed (Figs. 7A and 8) (Quarles van Ufford, 1996). Dozens
- Early Paleozoic

Tuaba Fm. 950±100 m

of outcrop-scale folds occur, but major folding (>100 m ampli-
Precambrian

tude) is not recognized in the Precambrian to Triassic strata. ~

1600± 2250±
400 m 600 m

Kilometer-scale folding and limited stratigraphic repetition by ~ B

faulting is obvious near the southern crest of the range in the ~ Kariem Fm. 3850±1000 m
Jurassic and younger formations. Bedding dips and formation ~
~ A
thicknesses indicate that at least 8 km of stratigraphic thickness
and as much as 12 km of structural thickness are exposed on the
v ~v ~v ~ v Awitagoh Fm. >200 m

north limb of the Mapenduma anticline (Fig. 7A) (Quarles van

Ufford, 1996; Weiland and Cloos, 1996). Figure 8. Stratigraphic column and approximate formation thicknesses
The southern boundary of the Mapenduma anticline must near the Ertsberg (Gunung Bijih) mining district. The Tipuma Forma-
be a thrust fault (Parris, 1994). The Precambrian or early Paleo- tion was deposited during the Triassic rifting that formed the present
zoic Kariem Formation is structurally higher than the Mesozoic northern edge of the Australian continent. Jurassic and Cretaceous
strata are shallow marine, passive margin deposits. In latest Cretaceous
Kembelangan Group east of the mine access road. The regional time, the northern edge of the Australian continent approached equa-
tilt of beds on the north limb is consistent with a giant 30 km torial latitudes, and carbonate sedimentation dominated, forming the
wide fault-bend fold above an ~30°, north-dipping thrust ramp. New Guinea Limestone Group. From Quarles van Ufford (1996).
14 M. Cloos et al.

Southward movement along the Mapenduma thrust fault can ing along 17 long segments of continuous road outcrop locally
only be estimated. Assuming the basal detachment is currently varies from 5% to 46%. The average shortening accommodated
at a depth of 15 km, a displacement of ~35 km is required to by folding on the northern limb of the Mapenduma anticline is
explain the map pattern. At the base of the ramp, a décolle- estimated as 25% (Quarles van Ufford, 1996). This estimate is
ment would have been at a depth of 20–25 km when movement a minimum, as intraformational thrust faults are also present.
started, that is, within the lower crust. An average rainfall along In the Ertsberg (Gunung Bijih) mining district, forami-
the southern flank of the western Central Range of ~10 m per niferal biostratigraphy of the New Guinea Limestone Group
year causes deep erosion of the southern limb (Fig. 7B). Apatite shows that carbonate shelf sedimentation (Kais Formation,
fission track thermochronology indicates the mid-slope region Fig. 8) was definitely as young as 15 Ma, and could be as young
has been unroofed at rates of 1–2 km/m.y. since the end of the as 12 Ma (Quarles van Ufford, 1996). Folds with near-vertical
Pliocene. Erosional unroofing of the southern flank of the anti- axial planes vary in style from rounded synclines to angular
cline began at ca. 7 Ma (Weiland and Cloos, 1996). anticlines. The average fold axis orientation calculated from
The ramp forming the Mapenduma anticline is believed to a compilation of all the bedding measurements made in the
be a reactivated normal fault that formed during early Mesozoic mining district has a trend of 294° and a plunge of 7°. There
rifting. In Papua New Guinea, abundant well and gravity data are several high-angle reverse faults that are low-angle to the
reveal the presence of Mesozoic normal faults in the subsur- upturned bedding. They could be upturned thrust faults, but
face. They parallel the 310° local trend of the orogen (Fig. 1) they are probably dip-slip faults that formed as the folds tight-
(Australasian Petroleum Company, 1961; St. John, 1970; Ridd, ened to the point that the limbs ruptured. The folds and dip-
1976; Hobson, 1986). In the eastern Central Range, reverse slip faults indicate a shortening direction of ~205°. Horizontal
reactivation of normal faults has been called upon to explain shortening across the 10 km wide mining district (Fig. 6) is ~5
the location of three basement-cored structures that are tens of km by folding and ~1 km by dip-slip faulting. In other words,
kilometers wide (Hill, 1991). The trend of the southern flank of total stratal shortening across this well-mapped 10 km wide
the Central Range appears to be nearly parallel to the trend of area is estimated at ~40% (Quarles van Ufford, 1996).
underlying Mesozoic rift structures. Thus, exposure of Paleo- Overlapping SPOT satellite images provide a 110 km
zoic, Mesozoic, and Cenozoic strata in the Mapenduma fold- wide image along the topographic crest of the western Central
and-thrust structure is most simply explained as due to reacti- Range. On a regional scale, the kilometer-scale folds, such as
vation of a 280°-trending normal fault zone in the crystalline those in the New Guinea Limestone Group in the mining dis-
basement during late Miocene collisional orogenesis and sub- trict, are seen to have a left-stepping en echelon geometry (Fig.
sequent denudation. 9). Similar structures have been recognized east of our field
transect (Nash et al., 1993; Granath and Argakoesoemah, 1989,
Folds in the Southern Central Range—Thin-Skinned p. 82). Regionally, these kilometer-scale folds near the spine
Deformation of the mountain belt have 300° to 310° trends. The left-step-
ping en echelon geometry records a component of left-lateral
Near the crest of the range, the amplitude of folding is at wrenching concurrent with shortening.
the kilometer scale in the Cenozoic carbonate section (Fig. 7) Three steeply dipping, 065°- to 070°-trending strike-slip
(Quarles van Ufford, 1996). Axial planar cleavage occurs in faults in the mining district formed concurrently with folding,
rocks ranging in age from Precambrian (Awitagoh to Tuaba based upon changes in fold shape and truncated strata and fold
Formations) to Triassic (Tipuma Formation). In the small folds axes. These structures are most simply explained as strike-slip
near the base of the mountain and near the core of the Mapen- tear faults that formed during contractional deformation to
duma anticline, cleavage is well developed and dipping north- locally accommodate differences in the pattern of folding and
east. It becomes progressively less common and near-vertical dip-slip faulting along strike (Quarles van Ufford, 1996).
at locations up the mountain slope with scattered occurrences
in the Modio, Aiduna, and Tipuma Formations (Quarles van
Ufford, 1996).
The bedding measurements used to determine the average
fold axis orientation on the northern flank of the Mapenduma
Figure 9. A: Geology of the western Central Range. Note the limited
anticline are distinctly bimodal, indicating chevron-style fold- erosion along the crest of the range as indicated by the widespread
ing. Angular fold hinges are seen at the outcrop scale, and the presence of strata of the New Guinea Limestone Group. En echelon
overall pattern indicates folding at a scale of tens to hundreds folds near the Ertsberg mining district are marked. Metamorphic zones
of meters, an average trend of 130°, and a plunge of 3°. Axial from Warren (1995) and Gray (1995). B: Modern topography. The
planes dip, on average, ~65° to the northeast. Fold trends and southern slope is underlain by upthrusted Australian continental crust.
The northern slope is underlain by upturned Irian Ophiolite. C: Varia-
cleavage orientations indicate an ~210° direction of shortening tion in depth of denudation. Rocks in the Ruffaer Metamorphic Belt
for the small folds along the north limb of the Mapenduma were uplifted from depths of 20–25 km. IOB—Irian Ophiolite Belt.
anticline. The detectable amount of intraformational shorten- Modified from Weiland (1999).
 Geology of the Central Range of Western New Guinea 15

C
16 M. Cloos et al.

Sapiie (1998) and Sapiie and Cloos (2004) provide an anal- block had ended, at least for those rocks in the belt now exposed
ysis of faulting in the mining district. Exceptional exposures at the surface (thermal evolution of subduction zones discussed
along mine roads revealed a comparatively minor, left-lateral in Cloos, 1985).
Riedel shear strike-slip system developed with a trend parallel- Based on field study and laboratory analysis of more than
ing the ~300° local structural grain generated by the upturned 200 outcrop samples from across and along this belt, the bound-
bedding. Contractional deformation in the mining district must ary of the Ruffaer Metamorphic Belt with the highlands fold-
have largely, if not entirely, ended by late Pliocene time, as the and-thrust belt to the south is gradational (Fig. 9) (Warren, 1995;
3 Ma Grasberg intrusion is a nearly circular plug that crosscuts Gray, 1995). Locally, the Derewo fault zone juxtaposes meta-
the axis of a kilometer-scale fold in the Ertsberg mining dis- morphosed and unmetamorphosed Jurassic to Cretaceous rocks
trict (Fig. 6). The left-lateral strike-slip system was concurrent of the Kembelangan Group (Warren, 1995). This fault is recog-
with the igneous activity in the district that occurred from 4.4 nized as a major structure because it forms an easily identifiable
to 2.6 Ma (McDowell et al., 1996). This faulting created path- 200 km long straight valley extending roughly east-west. Nash
ways for intrusion and focused the flow of hydrothermal fluids et al. (1993) suggest the Derewo fault zone is a north-dipping
forming orebodies (Sapiie, 1998). Several contraction-gener- reverse fault separating metamorphosed and unmetamorphosed
ated dip-slip faults were reactivated as left-lateral, strike-slip rocks, based primarily upon interpretation of satellite imagery.
faults. This distinctive structural episode caused a cumulative Field-based observation in the Ruffaer Metamorphic Belt led
left-lateral offset across the mining district of a kilometer or Warren (1995) to conclude that the Derewo fault zone is a near-
two at most (Sapiie, 1998). vertical strike-slip fault zone that only locally marks the bound-
ary between metamorphosed and unmetamorphosed rocks. The
Ruffaer Metamorphic Belt—Metamorphosed Passive magnitude of offset is unknown, but its length suggests move-
Margin Strata ment of at least 10 km and, more likely, several tens of kilome-
ters. A left-lateral sense of slip is inferred for the Derewo fault
The distribution of metamorphic rocks in the western Cen- zone by analogy with the Sorong-Yapen fault zones to the north
tral Range, the Ruffaer Metamorphic Belt, was first mapped by and late-stage strike-slip faulting to the south in the mining dis-
Dow et al. (1986) and Nash et al. (1993) using satellite imagery trict (Sapiie and Cloos, 2004).
and aerial photographs confirmed with limited field transects.
The first systematic petrologic studies of this slate and phyllite Irian Ophiolite Belt—Accreted Oceanic Lithosphere
terrane were by Warren (1995) and Gray (1995) (Fig. 9). They
showed that protoliths for most metamorphic rocks were simi- The Irian Ophiolite Belt in the western Central Range was
lar to the Jurassic and Cretaceous shales and siltstones exposed mapped using imagery by Dow et al. (1986). The first petrologic
in the fold-and-thrust belt. Several Cretaceous ammonites were study for rocks from the Irian Ophiolite was by Weiland (1999),
found in the slate belt (Warren, 1995). The presence of parago- who made three field transects in the belt. He found serpentinized
nite, phengitic muscovite, chloritoid, spessartine garnet, and the ultramafics at the higher elevations along the southern part of the
near absence of biotite indicate high-pressure greenschist facies belt. Rodingitic mafic dikes occur in some outcrops. The north-
metamorphism (Fig. 9). Maximum temperature and pressure ern flank, in the swampy lowlands, is composed of mafic intru-
estimates for the slates and phyllites are in the range of 250–350 sives and volcanic rocks. Outcrops are very scarce in the deeply
°C and 5–8 kbar (corresponding to maximum depths of ~25 km). weathered basalts.
The southern part of the belt is slate, whereas the northern part The structural setting indicates that the Irian Ophiolite Belt,
is locally phyllitic. This observation, combined with increasing like the Papuan Ophiolite Belt in the Bird’s Tail region (Davies,
illite crystallinity, widespread paragonite, and scattered occur- 1971), is a slab of ocean crust and upper mantle that has been
rence of spessartine garnet and chloritoid in the northern part of uplifted and tilted northwards ~30°. The southern edge has been
the belt, indicates the Irian Ophiolite Belt to the north was the unroofed to mantle depths, that is, from at least 7 km beneath
heat source that caused metamorphism. the ocean floor (Fig. 9). Because Jurassic seafloor was probably
Weiland (1999) provided the first geochronological study of at water depths of ~5 km, the ultramafic rocks now occurring at
the belt. Whole-rock K-Ar ages for 15 samples with strong pre- elevations of 1 km must have risen at least 13 km.
ferred orientations and petrographic evidence of thorough recrys- The ophiolite belt contains scattered occurrences of horn-
tallization are between 28 and 20 Ma. This age range is consid- blende amphibolite (Warren, 1995) that yield Jurassic Ar isotopic
ered the time of peak metamorphic temperatures. A single sample ages (Weiland, 1999). These black rocks form very distinctive
was coarse-grained enough for complete mineral separation, and cobbles in the river deposits. The hornblende-plagioclase ± gar-
a K-Ar white mica age of 21 Ma was obtained. We conclude that net assemblage indicates that these cobbles are mafic rocks that
metamorphic recrystallization in the exposed rocks started at ca. became foliated and lineated at temperatures of 500–800 °C and at
28 Ma, with northward subduction beginning slightly earlier. By pressures less than 5 kbar—high-temperature/low-pressure con-
ca. 20 Ma, the forearc region had sufficiently cooled that meta- ditions (Warren, 1995). Weiland (1999) concluded this dynamic
morphic recrystallization driven by residual heat in the forearc amphibolite facies metamorphism occurred at a location far out in
 Geology of the Central Range of Western New Guinea 17

the Pacific Basin where crustal slices or mafic dikes were caught deformation first caused one or more islands to emerge near the
in an oceanic transform fault zone near the formative spreading junction of the plates (the forearc high). The Makats Formation
ridge. Similar gneissic amphibolites have been dredged along the is reported to contain clasts of “metamorphic rocks, mica schist,
Mid-Atlantic Ridge (Honnorez et al., 1984). [and] slates …” (Visser and Hermes, 1962, p. 100–106) indi-
Dioritic plutons with arc-type chemistry are present in the cating deep (>10± km) denudation of the source landmass. The
western and eastern parts of the Irian Ophiolite Belt (Weiland, first appearance of metamorphic detritus is of great interest, for
1999). In the northwesternmost corner of the belt, there are the this would place important constraints on the rate of unroofing.
previously mentioned occurrences of plutons and volcanics of Unfortunately, the age of the first strata containing metamorphic
Maramuni age (20–10 Ma), for which the structural setting is debris is uncertain. Nonetheless, the middle Miocene age of the
unclear (Fig. 9). There is no known evidence that this magmatic basal Makats deposits indicates a substantial supply of silici-
complex was intruded into the ultramafic rock exposed to the clastic sediment to the south that was transported north into the
south. We suspect that this magmatic complex is an alloch- forearc basin. These biostratigraphic data indicate that landmass
thonous sheet thrust onto the back of the Irian Ophiolite Belt erosion began at ca. 15 Ma. This precedes, by several million
(this interpretation is shown on Fig. 9). A similar imbrication is years, the beginning of widespread synorogenic sedimentation
documented in the much better studied and exposed Weyland to the south and on the Australian continental basement at ca. 12
overthrust to the west, where arc-type magmatic rocks were Ma (Quarles van Ufford and Cloos, 2005).
thrust on top of the highlands fold belt (Dow and Sukamto,
1984a; Dow et al., 1988; Dow et al., 1990). We also believe SHORTENING ESTIMATES AND PLATE-TECTONIC
that these occurrences, along with the Cenderawasih Bay RATES
embayment, are recent and probably still developing structural
complications at the western end of the Central Range. The The magnitude of shortening recorded by observed folding
most recent movements are manifestations of the switch of the and detectable thrust imbrication in the western highlands is
Bird’s Head block from the Australian plate to the Pacific plate estimated to be ~80 km over the 120 km width of the western
(see earlier discussion). Central Range (Quarles van Ufford, 1996). In the eastern high-
In the eastern portion of the Irian Ophiolite Belt, a dioritic lands of Papua New Guinea, regional shortening is dominated
plutonic complex is intruded into the ultramafic terrane. K-Ar by thrust imbrication with subsidiary folding (Hobson, 1986).
hornblende and biotite ages of 35–28 Ma have been obtained The amount is estimated as ~100 km (Hill, 1991). These are
for part of this complex near Dabera (Fig. 9) (Weiland, 1999). obviously minimum estimates, as undetected structures must
This plutonic complex must have formed during the episode of be present. Even if the shortening was 200 km, this move-
south-dipping subduction that formed the New Guinea Trench ment is but a small fraction of the total convergence between
(Weiland, 1999). The presence of the Dabera plutonic complex the Pacific and Australian plates along the trend of the Central
at the leading edge of the forearc block indicates that north-dip- Range since 25 Ma.
ping subduction beneath the Outer Melanesian Arc started at ca. The Central Range developed in an obliquely convergent
30 Ma. Rupture initiation was by subduction reversal along the setting. The plate margin orientation at the time of subduction
axis of the line of arc volcanism (Quarles van Ufford and Cloos, initiation is uncertain, but it probably was northwest-trending.
2005). This is mechanically reasonable, if not expected, because The strike of the western Central Range is ~N80W. The trend
arc plutonism would thermally weaken this zone of lithosphere. of folds in the western highlands is N55W. Taking the average
Another effect is that when still-hot arc terrane becomes the trend of these strike orientations, and a relative plate motion vec-
leading edge, or forearc block, of a new subduction zone, it has tor of 9 cm/yr along an azimuth of S65W (from Scotese et al.,
a high heat content. This, in turn, causes higher than typical peak 1988), the normal component of convergence is calculated as
temperatures for a given depth and increases the total volume of ~50 km/m.y. and the strike-parallel component as 70 km/m.y. for
rock affected by metamorphic recrystallization during the initia- the period of at least 30–4 Ma. At this rate, the measured 80–100
tion of north-dipping subduction. Along with modest subduction km of shortening (or even 200 km) across the highlands fold-
speeds of ~5 cm/yr, the presence of cooling plutons and thus an and-thrust belts could have easily occurred in less than 4 m.y. As
abnormal heat content in the hanging wall of the new subduc- with all other subduction zones, most of the convergence is not
tion zone would explain why slates and phyllites of the Ruf- directly recorded in the geology of the accretionary complex.
faer Metamorphic Belt attained high-pressure greenschist facies Prior to the subduction of the edge of the Australian conti-
metamorphic rocks. nent, convergence was accommodated by movements in a shear
An important field relation concerning sedimentation of zone (the subduction channel as defined by Shreve and Cloos,
the Irian Ophiolite Belt forearc basement occurs in the Mam- 1986) beneath the accreted rocks forming the metamorphic belt.
beramo region near the international border (Fig. 1). The Makats Once deformation of continental rise strata began, a progressively
Formation contains abundant siliciclastic debris. The base of thickening layer of passive margin sediments was bulldozed. The
this formation is early Middle Miocene (16–14 Ma) (Visser and top part was accreted, and the basal section, several hundred
Hermes, 1962, p. 100–111). This unit records when subduction meters in thickness, was subducted to mantle depths. Accretion
18 M. Cloos et al.

by offscraping widens an accretionary complex, whereas under- core of the western highlands: Once collision began and crystal-
plating adds to the base and thickens it (see Cloos and Shreve, line basement became uprooted, the transcurrent component of
1988a, 1988b, for more details). Using this terminology, the motion began to manifest itself as northwest-trending, left-lat-
rocks of the Ruffaer Metamorphic Belt were underplated beneath eral, strike-slip faulting (Sapiie and Cloos, 2004). As will be seen,
the ophiolite belt. The estimate of 200 km of shortening in the the delamination of the lithospheric mantle and the change in the
fold-and-thrust belt only applies to deformed shelf deposits and, force balance on the edge of the Pacific plate as collisional oro-
thus, to convergence after continental crust began to underthrust. genesis progressed worked in concert to cause a distinct change
Therefore, many hundreds of kilometers of intraoceanic subduc- in structural response.
tion must have previously occurred to account for the previous In this chapter, the pattern and timing of sedimentation,
20+ m.y. of convergence. metamorphism, magmatism, and deformation was summarized
Overall, folds in the western Central Range trend ~40° for the western Central Range of New Guinea. Any model for the
oblique to the direction of plate convergence. Oblique conver- tectonic evolution of western New Guinea must account for these
gence readily accounts for the generation of left-stepping en observations. In Chapter 2, mechanical concepts are discussed
echelon fold trains in the core of the highlands. The extent to and summarized to provide a basis for predicting the response
which the folds rotated to their present orientation is probably of the underlying lithospheric mantle during the subduction of a
10°–20°, but this is uncertain. One phenomenon is clear in the continental margin.
 Geological Society of America
Special Paper 400
2005

Chapter 2
Mechanics of Subducting Slab Breakoff by Collisional Delamination
LITHOSPHERIC DELAMINATION

The basic concept that lithospheric mantle could separate

and sink away from continental crust was first proposed by Bird
(1978). The “delamination” process was invoked to explain the
timing of metamorphism and the generation of synorogenic
magmas in the core of the Himalayas (Fig. 10A). Bird (1979)
proposed that delamination was also a simple explanation for
the vertical uplift of the Colorado Plateau and associated sparse
magmatism.
According to Bird (1978, 1979), lithospheric mantle can
peel away from the overlying continental crust and sink vertically
because it is cooler and more dense than the underlying astheno-
sphere. The rate of delamination tearing depends strongly upon
the viscosity of the asthenosphere that must flow into the propa-
gating crack (Bird and Baumgardner, 1981). A similar sinking
trajectory is shown in the delamination models of Collins (1994)
and Collins and Vernon (1994). A variant of this process was
proposed by Houseman et al. (1981), Platt and England (1994),
and Schott and Schmeling (1998). These workers concluded that
thickened roots of lithospheric mantle form under contractional
orogenic belts. It is envisioned that where overthickened masses
of cold lithospheric mantle are generated, blobs detach as “litho-
spheric drips” that sink vertically into the asthenosphere.
Sacks and Secor (1990) proposed that following the jam-
ming of a subduction zone by a continent or thick arc complex,
the subducted plate would continue to sink along its previous
inclined trajectory (Fig. 10B). They argue that the overall force
balance driving the subducted end of the plate downwards is
changed little by the collisional jamming. Consequently, a subter-
ranean rifting of the subducted lithosphere must nucleate where
plate bending is tightest. By analogy to continental rifting, both
“pure-shear” ductile necking and “simple-shear” detachment Figure 10. Models for lithospheric delamination in collisional moun-
faulting were considered to be possible end members of this rift- tain melts. A: Himalayan orogen (Bird, 1978, Fig. 9 therein). B: Ap-
ing behavior. palachian orogen (Sacks and Secor, 1990, Fig. 1 therein). C: European
Plate rupture and sinking along an inclined trajectory has Alps (Davies and von Blanckenburg, 1995, Fig. 1 therein). Note the
differing ways in which plate rupture is envisioned to occur.
been termed “slab breakoff” by Davies and von Blanckenburg
(1995) and von Blanckenburg and Davies (1995). In their model,
they make a force balance / lithosphere strength analysis that Central Range of New Guinea. Cloos et al. (1994) referred to this
leads them to envision that rupture occurs by a crack propagat- process as “collisional delamination.” We believe delamination
ing upwards from the base and nearly perpendicular to the plate is the preferable descriptor of the slab breakoff process, for it
(Fig. 10C). They also conclude that rupture does not localize in implies that separation is strongly influenced by the mechanical
the region of tightest bending, rather it occurs at greater depths anisotropies in the lithosphere. It will be argued that the mechani-
where subduction speeds remain faster. cal layering of the sedimentary cover, the crystalline crust, and
We concluded that subterranean plate rifting must also have the mantle each play a predictable role in controlling the struc-
occurred in the subducting Australian lithosphere beneath the tural evolution before and during collisional orogenesis at a sub-
19
20 M. Cloos et al.

duction zone—a sequence of events well recorded in the geology dent upon confining pressure, and fluid pressure/effective stress
of western New Guinea. Two distinctive geologic phenomena are principles should also apply (Fig. 11). As permeability, fluid con-
well recorded in the rock record of New Guinea that constrain the tent, and fluid pressure gradients determine the rates and scales
timing of collisional delamination. They also should be found in of equilibration, the mechanical role of fluids must be large in
and near more ancient, deeply eroded, collisional orogenic belts: porous, water-rich sediments but comparatively minor in anhy-
late-stage igneous activity along the axis of the orogenic belt and drous mantle rocks.
profound changes in regional sedimentation patterns. At greater depth, temperatures are higher, and differential
stresses that cause elastic strains are slowly relieved by distor-
RUPTURE OF THE LITHOSPHERE tion (flow) of crystals (Fig. 12A). Dislocation motion controls the
creep rate under a wide range of temperature conditions (Carter,
Our field observations and isotopic ages summarized earlier 1976). Under common tectonically imposed strain rates and
are the basis for reconstructing the timing of geologic events at greenschist facies metamorphic temperatures of ~300 °C, quartz
and near Earth’s surface during the collision forming the Cen- flows; at temperatures of ~600 °C and higher, olivine flows
tral Range of western New Guinea. The mechanical response of (Tsenn and Carter, 1987). “Flow laws” are experimentally cali-
materials at depth must be inferred. On this subject, our kinematic brated relationships between measured differential stress (flow
picture of lithospheric behavior differs significantly from that of strength) and imposed strain rate, with the viscosity as the derived
previous workers who advocate delamination or slab breakoff as coefficient of proportionality. Flow laws have been determined
a tectonic process. For this reason, our mechanical reasoning is for quartzites and some other rocks to approximate continental
described in detail below. crust and dunites to approximate the mantle (Kirby, 1983; Carter

Fracture and Flow

There are four end-member rheological behaviors in the

solid earth: elastic, brittle, plastic, and viscous. On the short time
scales of seismic wave propagation, the entire earth responds
elastically, or largely so. Earthquakes and brittle behavior involve
planar faults. Plastic and viscous behaviors are both forms of dis-
tributed flow.
Where earth materials are sufficiently cool (<300–400 °C),
differential stress (σ1 – σ3) and resultant elastic strains can slowly
build up over time scales of years to centuries. When the differ-
ential stresses exceed the fracture strength or, much more com-
monly, the frictional resistance to renewed slip on planes of weak-
ness (fault reactivation), rapid movement occurs, and earthquake
waves are generated. Ruptures commonly propagate upward into
zones of weak materials that are only able to sustain small differ-
ential stress between events (e.g., Cloos, 1992a). Displacement
along faults involves friction, and thus, resistance increases with
increasing normal stress on the slip planes. Higher fluid pressure
can lessen and even entirely negate the effects of higher confin-
ing pressure because it proportionately decreases the “effective”
normal stresses (Hubbert and Rubey, 1959).
Conventional fracture strength analysis is rooted in Mohr-
Coulomb theory. This analysis appears to apply fairly well for
short-term phenomena such as the rapid loading of rocks by
earthquake-generating movements, but in the case of orogenic Figure 11. Diagram illustrating the concept that the “peak strength”
of the lithosphere is at depths of 10–25 km. Peak strength occurs
processes occurring over million-year time scales, it only pro- where the pressure-dependent downward-increasing fracture strength
vides an upper limit on the differential stress that can be sus- limit intersects the upward-increasing, temperature-dependent flow
tained. There are significant, but poorly calibrated, time-depen- strength. At the depths where the lithosphere has peak strength, large
dent effects that cause fractures to form at “subcritical” differen- elastic strains can develop. This is where the largest earthquakes in a
tial stresses (Atkinson, 1987). The field of engineering fracture given area are nucleated, other factors being equal. Peak strength oc-
curs at shallower depths where temperature gradients are higher, local
mechanics is well established but essentially undeveloped for strain rates are slower, or fluid pressures are lower. In most tectonic
processes occurring on geologic time scales. Nonetheless, long- regimes, the dominant factor controlling the depth of peak strength is
term fracture strengths should be linearly, or nearly so, depen- probably the temperature gradient.
 Mechanics of Subducting Slab Breakoff by Collisional Delamination 21

extrapolations depends upon whether the same grain-scale defor-

mation mechanisms that operate in the experiments also operate
in the natural conditions of interest. The trace water content (at
ppm amounts) has also been shown to have large effects (Kohlst-
edt et al., 1995). Although rigorous quantitative extrapolation is
limited by the multitude of variables (Rutter and Brodie, 1991),
it is clear that as temperatures increase at greater depths, the flow
strength of silicate materials decreases exponentially (Fig. 11). At
near-surface temperatures, the viscosities of silicate rocks are so
large that any flow is negligible, and the differential stresses and
elastic strains can build up to the point where fracture-causing
earthquakes occur.

Plastic and Viscous Flow

Many workers use the terms plastic and viscous as syn-

onyms for the ductile flow of materials. However, in materials
science, these terms correspond to distinct end-member mechan-
ical behaviors. Ideal plastic materials have a differential stress
(yield strength) that must be exceeded before any flow begins.
Ideal viscous materials, on the other hand, flow wherever differ-
ential stresses are nonzero. Plastic and viscous are the mechani-
cal approximations that best characterize the long-term behavior
of lithospheric mantle and asthenosphere, respectively (Oxburgh
and Turcotte, 1976).
The plastic flow regime in the lithosphere occurs where tem-
peratures are sufficiently small that differential stresses less than
some finite value, the yield strength (σy), can be supported indef-
initely. Yield strength exponentially decreases with increasing
temperature. Whenever differential stress attains the magnitude
of σy, distortion begins (Fig. 12A). In an ideal plastic material,
flow occurs instantly to reduce the ambient differential stress to
Figure 12. Factors affecting plastic and viscous flow. Plastic materials
have a finite differential stress they can support—the yield strength. values less than σy, at which point distortion immediately ceases.
A range of force imbalances and pressure gradients can be supported. In reality, plastic materials will flow at rates that depend upon the
Viscous materials have a negligible yield strength, and flow will occur viscosity of the material. Yield strength and associated viscosity
in response to even small gravitationally induced pressure gradients. increase with decreasing temperature. In homogeneous, isotro-
Lithospheric mantle behaves as a plastic material and can generally pic materials, plastic or plastico-viscous yielding is concentrated
transmit the push and pull forces induced by plate movement. Asthe-
nospheric mantle behaves viscously. A: The effect of temperature on in directions where the shear stress is at maximum values. This
yield strength. B: Most flowing geologic materials have power-law occurs in bands that are oriented at 45° from the local σ1 and σ3
(nonlinear) relationships between differential stress and strain rate. A stress directions.
consequence of this behavior is that the viscosity is different at slow The viscous regime is where temperatures are sufficiently
and fast strain rates. The concept of “equivalent viscosity” refers to high that any differential stress causes flow at a rate that depends
the viscosity of a material at a specific range of strain rate. For silicate
minerals, the power-law relationship is commonly a factor of about on the viscosity (Fig. 12B). Flow strength and viscosity decrease
three (n = 3). This means the equivalent viscosity at slow strain rates exponentially with increasing temperature. In an ideal (Newto-
is much higher than at fast strain rates. As a result, there is a large nian) viscous material, flow occurs at a constant rate regardless
gradient in strain when shearing motions are imposed on one of the of differential stress magnitude. Real rocks are significantly non-
boundaries of the flowing material. Newtonian, and the rate of flow depends upon the magnitude of
the imposed differential stress or strain rate. In environments of
constant load, such as inactive seamounts, differential stress is
and Tsenn, 1987). The laboratory experiments show an exponen- imposed, and this determines the rate of flow in the underlying
tial dependence on temperature and strain rate. Appropriate tem- mantle. In subduction zones, a displacement rate is imposed that
perature conditions can be simulated, but the experiments must determines the magnitudes of the differential stresses generated
be done at laboratory strain rates that are more than five orders in the zone of flow. The “equivalent viscosity” is the resistance to
of magnitude faster than natural conditions. The validity of the flow at specified conditions of temperature and differential stress
22 M. Cloos et al.

or strain rate. In ideal homogeneous materials, viscous flow is asthenosphere occurs wherever pressure gradients are not at
evenly distributed throughout the deforming body. In a section of equilibrium values. Where movements thin or thicken the over-
real rock, flow is concentrated in weak zones and along boundar- lying lithospheric plate or where erosion and sedimentation cause
ies that move. variations in loading, flow occurs in the asthenosphere to accom-
modate changes in the load of the overlying lithosphere and its
The Lithosphere: Brittle/Plastic Behavior cover of sediment.
The base of the lithosphere is a rheological boundary that is
Oceanic lithosphere is best approximated as having an upper primarily temperature-controlled. The asthenosphere is the source
elastic/brittle and underlying elastic/plastic structure (Fig. 11). of ocean ridge basaltic magmas that erupt at temperatures of
Most of the time, ambient differential stresses are less than the 1100–1200 °C (Philpotts, 1990), indicating a minimum tempera-
fracture strength near the surface or the flow strength at depth. ture for the source region. Strong supporting observations for dif-
Internal density differences due to composition and temperature fering mechanical behavior with depth come from mantle nodules
are rarely sufficient to cause movement, and heat flow through in basalts from rift zones or ocean islands in mid-plate settings
the lithosphere is primarily by conduction. In continental areas, (Finnerty and Boyd, 1987; Nixon and Davies, 1987). Nodules with
thermal gradients are in the range of ~30 °C/km in the crust textures indicating annealing and long-term static conditions and a
where the heat-producing elements are concentrated, and as low lithospheric source typically come from depths less than ~100 km
as 5 °C/km in the underlying lithospheric mantle. In oceanic lith- (P < 30 kbar) and typically record temperatures less than 1100 °C.
osphere, thermal gradients are systematically related to plate age. Nodules with sheared fabrics indicating recent flow and an astheno-
Near ocean ridges, thermal gradients are as high as 100 °C/km. spheric source are more common from greater depths and typically
Where oceanic lithosphere is ca. 80 Ma, thermal gradients are record temperatures ~1200–1400 °C.
near 10 °C/km. The transition from lithospheric mantle with some strength
An important exception is the plutonic/volcanic arc envi- to viscously flowing asthenosphere should occur over a fairly
ronment where intruding magmas advect heat upwards and ther- narrow zone. The gradient in temperature downwards causes
mally weaken the wall rock. In long-lived subduction systems, a flow strength to decrease downwards. Laboratory experiments
major zone of weakness typically exists along the line of the arc show that silicate minerals have differential stress-strain rate
in the overriding plate. relationships that are significantly nonlinear (Fig. 12B). For
Where tectonic movements generate differential stresses small differential stresses, the flow rate is slow and the effec-
that exceed the local strength of the lithosphere, deformation can tive viscosity high. For larger differential stresses, the flow rate
create variations in topography and crustal thickness. Once this is faster and the effective viscosity lower. As a result, differen-
occurs, gravity alone can generate sufficient differential stresses tial shearing movements from normal plate motions are con-
that drive local movements and continue long after plate tecto- centrated into the hotter, basal part of the lithosphere. The base
nism has ceased (extensional collapse) (Dewey et al., 1993). The of the lithosphere should be thought of as a zone across which
rate of these gravity-driven movements slows progressively as there is a large strain rate gradient anytime the lithosphere and
erosion and sedimentation flatten topography and deeper-seated asthenosphere are moving with respect to one another. Flow-law
“isostatic” adjustments occur. relationships indicate that the gradation from negligible flow
In mantle materials, dislocation glide is the dominant defor- (high viscosity) to significant flow (low viscosity) under most
mation mechanism in the regime of plastic flow. This is a condi- plausible conditions of plate movement should occur across an
tion where the flow of rock is limited by the ability of disloca- interval of ~100 °C. This would correspond to a thickness inter-
tions to migrate across olivine and pyroxene crystals. Because val of ~10 km, assuming a conductive thermal gradient of 10
of the increase in temperature with depth, the yield strength of °C/km in the lower lithosphere.
mantle materials decreases progressively downwards. The frac- There is a well-established phenomenon that bears on the
ture strength and flow strength intersect, forming a band with thickness of the shear zone at the bottom of the lithosphere.
“peak strength” typically at depths of 10–25 km in old oceanic Geophysicists have long correlated the top of the asthenosphere
lithosphere (Fig. 11). The yield strength in this depth range is with the top of the “low-velocity zone.” Beneath most oceanic
sufficient to transmit horizontally the typical differential stresses and some continental areas, the low-velocity zone has a fairly
arising from the push and pull forces that drive plate motion. abrupt top at ~100 km depth and gradational bottom at ~250 km
depth (Anderson, 1989, p. 51). The low-velocity zone, named
The Asthenosphere: Viscous Response for the slower travel times and attenuation of seismic waves, is
thought to be caused by the presence of a small amount of melt,
The asthenosphere is where temperatures are sufficiently indicating temperature conditions just below the solidus. Less
high that rock flow by distortion of olivine and pyroxene readily than 1% melt can account for these geophysical observations
occurs. Dislocation climb is the dominant grain-scale deforma- (O’Connell and Budiansky, 1977). Although the magnitude of
tion mechanism, a condition attained where dislocation move- the mechanical effect of partial melting is not well calibrated, it
ment is substantially aided by atomic diffusion. Flow in the is clear that the bulk viscosity is lower if any melt is present and
 Mechanics of Subducting Slab Breakoff by Collisional Delamination 23

progressively decreases as the amount of partial melt increases Where fluid pressures are elevated because of prograde meta-
(Kohlstedt and Zimmerman, 1996). If, as it appears, the base of morphic dehydration or distributed partial melting, the effect
the lithosphere is where the change from near-solidus to slightly of higher confining pressures is negated. At depth, high enough
suprasolidus conditions occurs, the base of the lithosphere can temperatures are attained so that plastic flow (distributed
be symbolized as an isotherm that approximates a zone perhaps strain) occurs rather than fracture. As temperatures increase,
only a few kilometers thick. flow strength exponentially decreases. The largest differen-
Since the advent of plate tectonics, most geophysicists tial stress, or “peak strength,” in the earth must occur at the
believe that the viscosity in the asthenosphere is sufficiently crossover between downward-increasing fracture strength and
low that steady convective flow occurs in this part of the mantle. upward-increasing plastic flow strength (Fig. 11) (Brace and
The near chemical uniformity of ocean ridge basalts indicates Kohlstedt, 1980; Kohlstedt et al., 1995). For a given area, the
that global-scale homogenization of the asthenosphere occurs largest differential stresses, consequent elastic strains, and
on the 200–400 m.y. time scale. This directly indicates that thus the largest earthquakes should develop near the depth
the asthenosphere is well mixed, perhaps even at the kilome- of the strength crossover. Strain rate effects on flow strength
ter scale. This strongly suggests that buoyancy forces that arise are qualitatively similar, but opposite, to those of temperature.
from density differences due to temperature and compositional High fluid pressure conditions, either aqueous or magmatic,
gradients readily drive internal circulation. Consequently, heat enable fracturing at great depths (Fig. 11).
flow in the asthenosphere is primarily by convection, and ther- Strong supporting evidence for the concept of a zone of
mal gradients are nearly adiabatic (~0.5 °C/km). peak strength in the lithosphere comes from the observation
There is a chemical matter of special importance to issues that most large (M > 7) earthquakes originate at depths of 10–
of magmagenesis. Most partial melt in the asthenosphere 25 km (Sibson, 1989; Scholz, 1990). At shallow depths, rocks
should be buoyant and tend to infiltrate upwards, or pool and are too weak to support the buildup of large elastic strains.
then intrude as dikes, into the overlying lithospheric mantle. In areas of high thermal gradient, the deepest earthquakes are
Small amounts of magma have small heat contents. Thus, when shallower and vice versa. The extreme case of lower thermal
rising into lithospheric mantle, whether by infiltration or as gradient occurs in subduction zones, where most large earth-
small dikes, they would rapidly cool and solidify. This process quakes are thrust-type and originate at depths of 25–60 km.
can produce dramatic chemical modification of the lower litho- The extreme case of higher thermal gradients occurs near
spheric mantle over time. Lithospheric mantle becomes “fertil- ocean ridges, where the largest earthquakes are normal-type
ized” (metasomatized) with K, S, Cl, Ti, Fe, P, LREEs (light and originate at depths less than 10 km.
rare earth elements), and other elements that are “incompatible” The effects of temperature gradient, strain rate, and fluid
with the olivine, pyroxene, and garnet in the asthenosphere and pressure on peak strength in oceanic and continental litho-
hence are concentrated in small-degree partial melts that intrude sphere are schematically summarized on Figure 13. In areas
upward (Roden and Murthy, 1985; McKenzie, 1989; Menzies, with normal thermal gradients, peak strength lies within the
1990). These elements are “stored” in the lithospheric mantle upper mantle for old oceanic lithosphere, but within the crust
within phlogopite, amphibole, rutile, ilmenite, apatite, various for all but the thinnest continental crust. The crust under the
sulfides, and exotic minerals. Consequently, magmas that are continents is always ~30 km thick at the locations near the
derived from, or pass through, metasomatized lithosphere will shoreline and typically ~40 km thick in the interior (Mooney et
generally have distinctive trace element and isotopic composi- al., 1998). The geophysical properties of crust under the shelves
tions. Thus, a component of melt from the lithospheric mantle is intermediate to that of typical continental and oceanic ter-
is geochemically distinctive and detectable in magmas from ranes, and this region is commonly referred to as underlain
continental rift zones (Hawkesworth et al., 1990; Gibson et al., by “transitional crust” (Bond and Kominz, 1988). Transitional
1992). As will be discussed, significant quantities of magma crust forms during continental rifting when normal faulting,
should be derived from a lithospheric mantle source in areas of erosion of uplifted blocks, and ductile stretching thin the crust,
collisional delamination. and mafic intrusions increase its density.
The key conclusion is that the strength profile of “nor-
PEAK STRENGTH IN THE CRUST AND MANTLE mal” 30+ km thick continental lithosphere is distinctly differ-
ent from that of oceanic lithosphere (Fig. 13). The mechani-
The strength profile of lithosphere requires further eval- cal implication is that normal continental crust is only weakly
uation because there are dramatic mechanical differences coupled to the underlying lithospheric mantle. In the interior of
between areas capped by continental versus oceanic crust. continents, the brittle middle part of the crust is underlain by a
Where temperatures are low, the strength of rocks that com- plastic zone that overlays the viscously behaving lower crust.
pose the crystalline basement is high because friction resists It is not necessary for the lower crust to be uniform in
internal distortion. The differential stress required for slip on composition to behave viscously. Only parts of it must contain
preexisting fractures or other weak anisotropies, or the forma- significant quantities of quartz (>10%?). It is highly likely that
tion of entirely new fractures, increases linearly with depth. in some areas, the lower crust is entirely made of granitic rock,
24 M. Cloos et al.

Figure 13. Peak strength profiles for the lithosphere. Peak strength in normal oceanic lithosphere is within the upper mantle. In areas where
continental crust has thickness greater than ~25 km, there are two “strong” horizons. The upper horizon is in the crust. The lower one is at the
Moho in the uppermost part of the mantle. The brittle upper crust sits atop a viscous lower crust. In areas where the crust is “transitional,” that is,
between ~25 and 10 km in thickness, the peak strength is in the upper mantle. The lower part of continental crust that is ~15–25 km thick behaves
plastically, and detachment from the lithospheric mantle roots can occur during collisional orogeny. Continental crust thinner than ~15 km thick
is effectively welded to the mantle, and the deep subduction of the edges of continents should be commonplace over geologic time.

and flow should be distributed. In others, it probably only con- FORCE TRANSMISSION
tains layers or zones of quartz-rich material, and these zones
will control the nature of flow in the lower crust. Where the It is evident that the largest push and pull forces of plate
lower continental crust is quartz-poor, the behavior may be tectonics are laterally transmitted at depths of 10–25 km. This is
plastic rather than viscous. within the uppermost mantle of oceanic lithosphere. The overly-
Only where continental crust is thinner than ~15–20 km ing crust and lower lithospheric mantle respond to movements in
is it effectively “welded” to the underlying mantle (Fig. 14). the strong, uppermost mantle. Distortions can occur slowly and
Taking this perspective, continental crust is progressively less aseismically, or rapidly concurrent with movements generating
strongly coupled to the underlying mantle toward the interior, earthquakes. The high strength of oceanic lithosphere compared
a phenomenon that should have profound ramifications for tec- to the typical differential stresses related to plate motion is evi-
tonic behavior during the subduction of continental margins dent from the observation that far away from plate boundaries,
and subsequent collisional orogenesis. oceanic plates are nearly aseismic, and the overlying blanket of
 Mechanics of Subducting Slab Breakoff by Collisional Delamination 25

Figure 14. Schematic diagram illustrating the differences in mechanical behavior from oceanic to normal continental crust. Transitional crust
underlying continental margins is subductable. Overlying sediments (rise, slope, shelf) are only weakly attached and will become bulldozed
against the hanging-wall buttress as the underlying crystalline basement underthrusts and then subducts. Collisional tectonism will detach con-
tinent fragments underlain by lower crust that behaves plastically. Imbrication of detaching crust imparts forces on the edge of the continent
that can be transmitted in crust viscously coupled to the mantle for many tens to several hundred kilometers toward the interior. Modified from
Cloos (1993).

oceanic sediments is flat-lying. A notable and active exception, sediment thickness, small angular unconformities in the near-
almost certainly related to the collision forming the Himalayas, horizontal stratigraphy, and/or fault offsets or arching or doming
is buckling and scattered seismicity of the seafloor in the central or basin subsidence record mid-plate vertical movements that are
Indian Ocean (Weissel et al., 1980; Zuber, 1987). In this unusual rarely more than a few kilometers. The overall 500+ m.y. stability
case, the giant Indian-Australian plate is contorting ~2000 km of the interior of continents is evident from the flat-lying Paleo-
seaward of the collision-generated Himalayas. zoic strata that cover large areas. Thus, the upper part of conti-
The situation for the continents is clearly very different. nental lithosphere must have a yield strength that is greater than
There are broad areas of scattered seismicity in western North the differential stresses generally imparted by the pull force(s)
America, Asia, Europe, and Africa. In the interior of continents, transmitted updip from the sinking end(s) of the descending plate
episodes of vertical movements have occurred far from plate and the push forces applied along some edges.
boundaries. Some movements cause uplift, forming arches or
domes. Others cause subsidence, forming basins. Compared to PLATE BENDING DURING SUBDUCTION
structures observed in mountain belts, however, the interiors of
large areas of continents have generally only been affected by Considerable plastic distortion must occur in the lower litho-
relatively small vertical displacements. Variations in regional spheric mantle when it bends to subduct. It is evident that the
26 M. Cloos et al.

bending and unbending of oceanic lithosphere occurs steadily for However, continental detachment should typically begin
many tens or even 100+ m.y. The lower part of oceanic litho- well before 30+ km thick crust starts to bend downwards to sub-
sphere must first contract as the hinge forms and then extend as duct. Crustal detachment should initiate where positively buoy-
it unbends to straighten and continue descent into the mantle. ant lithosphere begins to turn downwards to move to mantle
Where the bending is tight because the subducting plate dip is depths. This is where the crust is 15–20 km thick and transitional
steep, the ocean crust and uppermost mantle fracture. Normal in nature. The lower part has a finite yield strength and thus
faults with scarps up to a few hundred meters high are common responds plastically (see Fig. 13). By the time continental crust
in the seafloor between the base of the trench slope and the top that is weakly attached to the underlying mantle is forced to turn
of the outer rise. Their formation is concurrent with normal-type, downwards to subduct, crustal buoyancy generates substantial
bending-induced earthquakes detected near the outer rise (Isacks resistive forces that are transmitted updip and laterally toward
et al., 1968). Although the near-surface rupture of descending the interior of the continent. These are the collisional forces
oceanic lithosphere is commonplace, it has no obvious direct that cause uplift, crustal thickening, and other tectonic move-
long-term effect on the overall subduction process. This indicates ments many tens to hundreds of kilometers landward of a trench.
that faulted oceanic crust remains welded to, and subducts with, All of this reasoning leads to the expectation that the effects of
the underlying upper mantle. mechanical layering during subduction zone jamming—the pro-
As evident from the Himalayas and the European Alps, the cess of collisional delamination—combine to create a predict-
situation is very different where continent-capped lithosphere able sequence of events, well recorded in the surface geology of
enters a subduction zone. Where 30+ km thick continental crust New Guinea.
is involved, there are two distinct horizons of peak strength: an
upper zone of peak strength in the middle of the crystalline crust BUOYANCY OF THE LITHOSPHERE
and a lower zone in the uppermost mantle (Fig. 13). These hori-
zons of strength are separated by the lower crust that has a flow Nearly all subduction (except where the oceanic lithosphere
strength that should typically decrease downwards as tempera- is younger than ca. 10 Ma) is driven by the inherent negative
ture increases, but compositional factors may dominate. Where buoyancy of oceanic lithosphere compared to the underlying
continental crust is thinner than ~15–20 km, that is, beneath the asthenosphere. But before proceeding, the density distribution of
continental shelf and slope, the crust in contact with lithospheric lithosphere and the location of the negative buoyancy requires
mantle responds plastically with a yield strength that generally some discussion. The lithosphere is strong, for it is the cooler
should increase as crust thickness decreases. This mechanical outer boundary layer of the mantle. As the temperature of the
picture (Fig. 14) has considerable tectonic implications for crustal mantle at the base of the lithosphere is almost the same as that
response as a continental margin enters a subduction zone. at the top of the asthenosphere, the density is nearly the same.
Crust underlying the continental slopes and outermost shelf Because the variation in compressibility as a function of temper-
areas is effectively welded to the lithospheric mantle, and the ature is comparatively small (Anderson, 1989, p. 86), the density
peak strength is in the upper mantle. The faster the speed of sub- at depth is almost entirely a result of temperature difference. The
duction, the deeper such crust can be subducted before heating equivalent 1 bar density profile for the mantle is shown in Figure
weakens the attachment to the underlying mantle. Where con- 15. Most of the negative buoyancy of an oceanic plate is gener-
tinental crust is of 30+ km thickness, a zone, 10 km or so thick, ated in the cooler, upper 30 km of lithospheric mantle.
in the lower crust is typically warm enough to flow viscously. The sinking plate is most dense where it is also the strongest.
Crystalline middle crust with significant fracture and plastic flow Consequently, peak strength corresponds to where the magnitude
strength, and any overlying sediments with negligible strength, of the updip pull force from plate sinking is also a maximum. We
are effectively floating on the lower crust. Where subduction to believe this is the factor controlling the response of the descend-
mantle depths is attempted, normal-thickness continental crust ing plate during collisional delamination.
must detach from its mantle underpinning, because the buoyancy In Chapter 3, the geological observations of Chapter 1 will
forces resisting the subduction of continental lithosphere are as be integrated with the mechanical principles discussed in Chapter
large as those driving oceanic lithosphere downwards (Cloos, 2 to generate a series of scaled, lithospheric-scale cross sections
1993). This factor alone can explain the long-term preservation of the collisional delamination event that created the Central
of old continental crust. Range of New Guinea.
Figure 15. A: Temperature profile of the lithosphere. “Potential temperature” is the equivalent temperature at surface pressure “corrected” for
the adiabatic temperature increase with depth that results from compression. The lithosphere is a conductive thermal boundary layer between the
surface and the asthenosphere that convects freely and maintains a near-adiabatic temperature gradient. B: Bulk density profile of the lithosphere.
Absolute densities are uncertain because exact rock composition is uncertain, but for oceanic lithosphere and asthenosphere, the relative density
contrasts are well constrained. The high density of the uppermost mantle (up to 3.39 g/cm3) is due to lower temperature and resultant thermal
contraction. The lower part of lithospheric mantle is nearly at the temperature of the asthenosphere and thus has nearly the same density. Den-
sity increase due to compression is not included, for it is a comparatively minor second-order phenomenon that affects hot and cold mantle to a
similar magnitude. From this figure it is apparent that the negative buoyancy of the lithosphere resides in the upper 20 km or so of the mantle.
Compared to the density contrasts that drive subduction, the extreme low density of continental crust will resist subduction proportional to the
amount of mass involved. Nonetheless, the deep subduction of the thinned edges of continents (transitional crust) occurs shortly before col-
lisional orogenesis begins. During normal subduction, the downward shearing forces imparted by the sinking of the oceanic lithosphere drags a
thin layer of sediment to the depths of arc magmagenesis. Parameters are consistent with those in Cloos (1993).
 Geological Society of America
Special Paper 400
2005

Chapter 3
Palinspastic Reconstruction of Collisional Delamination
in New Guinea
COLLISIONAL DELAMINATION—SCALED CROSS adjustment to the load. The base of the trench slope marks the
SECTIONS seaward extent of deformation of the hanging-wall block. Where
not buried by sediment, an outer rise, commonly 100–200 m high,
Lithospheric-scale cross sections of the Central Range is located ~100 km farther seaward. The outer rise marks the sea-
orogeny were drawn to reflect the events recorded in the geol- ward extent of deformation of the descending plate. Collisional
ogy of New Guinea (discussed in Chapter 1 and summarized in orogenesis only begins when positively buoyant lithosphere is
Table 1), the consideration of the mechanical layering of con- forced to bend downwards and begin true subduction. In the case
tinental lithosphere, and the physical constraints on collisional of New Guinea, the component of head-on convergence was ~5
orogenesis in Cloos (1993). In this case, the collision resulted cm/yr, and about five million years elapsed between when con-
from north-directed subduction of ocean-crust-capped litho- tinental-crust-capped lithosphere began to underthrust and when
sphere, which changed to transitional Australian-crust-capped crystalline basement began to be uprooted because of collisional
lithosphere. The essential geodynamic requirement is that steady orogenesis.
subduction occurs as long as the descending plate has a bulk den- Scaled cross sections are presented to take into account all
sity greater than the underlying asthenosphere (negatively buoy- known timing constraints for the development of the orogeny
ant). When the incoming plate has a bulk density less than the that formed the western Central Range of New Guinea. As the
underlying asthenosphere (positively buoyant), the subduction ages of most events are known within time spans of 2 m.y., our
zone eventually jams. Normal oceanic lithosphere, oceanic litho- cross sections are drawn at 2 m.y. increments.
sphere capped by short-lived or young island arcs (<~20 m.y.), The sections are drawn using relative Australian-Pacific
and continental margins underlain by thin crust are inherently plate motions that indicate convergence at a speed of ~50 km/
subductable because the bulk density of the crystalline crust and m.y. prior to 5 Ma (see Scotese et al., 1988). The cross sections
lithospheric mantle is greater than that of the underlying asthe- are drawn from the Central Range foreland to beyond the New
nospheric mantle. Guinea Trench, and the events generate the geologic relations
An important concept is that sedimentary rocks are not depicted along section B–B′ on Figure 9. Reference points for
welded to the crystalline basement. Sediments simply ride on the sequential comparison of position are the relict New Guinea
top of the plate and thus do not add to lithospheric buoyancy. Of Trench for the overriding plate and the Timika airport for the
course, loading by sediment deposition or unloading by erosion subducting Australian plate.
do cause vertical isostatic adjustments. Several concepts need refinement to add focus on funda-
At current shorelines, continental crust typically has a thick- mental tectonic processes. The use of the terms “accretionary
ness of ~30 km. With this crustal thickness, the lithosphere has prism” for steady-state subduction zones and “collisional moun-
a positive buoyancy that is nearly the same magnitude as the tain belt” for cases of attempted continental subduction are well
negative buoyancy that drives steady subduction of 50+ Ma established in the literature, but the distinction of where steady
oceanic lithosphere. All but the youngest oceanic lithosphere is subduction ends and collision begins is not.
inherently subductable. On passive margins, neutral lithospheric Sediments deposited upon oceanic basement (abyssal plain
buoyancy occurs somewhere beneath the continental shelf. Shelf and rise deposits) are deformed at the base of the trench slope.
width varies from a few tens of kilometers to many hundreds of During steady convergence, a layer of sediment, up to 500 m or
kilometers, as off southeastern South America (Falklands area) so thick, and in some cases the entire incoming sediment pile,
and northern Australia (the New Guinea region). subducts to the depths of arc magmagenesis. Once continental
Underthrusting occurs before “true” subduction that involves shelf and slope deposits enter the realm of subduction-driven
the downward movement of crustal materials to mantle depths. deformation, the bulk of the pile is simply offscraped into a
Sediment offscraping occurs at the front of the trench slope. steadily growing mass. Offscraping and underplating occur with
Material that moves under the hanging-wall block is simply no change in the speed of convergence (see discussion in Cloos
underthrusting until it reaches the point where true subduction and Shreve, 1988a, 1988b). Once thick slope and shelf deposits
begins. In the zone of underthrusting, sediments are bulldozed, are involved, the offscraping deformation is far from a steady-
and the deflection of the descending plate is simply an isostatic state condition because an ever thickening wedge of sediment
29
 TABLE 1. GEOLOGIC EVENTS ALONG THE NORTHERN AUSTRALIAN CONTINENTAL MARGIN
Western New Guinea (136°E to 141°E) Eastern New Guinea (141°E to 147°E)
Indonesia Papua New Guinea
Present-day
• Pacific plate north of area • Bismarck microplate northeast of area; Solomon
• Strike-slip along Yapen fault zone microplate east of area
• No Pleistocene volcanism • Convergent tectonism uplifts Huon Peninsula
• Little seismicity in highlands • Widespread Pleistocene volcanism
• Scattered seismicity in highlands
• Thrust faulting along southern margin of highlands

4 Ma
• Shortening of Irian fold-and-thrust belt ending • Shortening/unroofing of crystalline basement under
• Volcanism ending along spine of western highlands Papuan thrust-and-fold belt begins
• Conglomeratic sedimentation begins • Widespread volcanism along spine of eastern highlands

ca. 5 to 3 Ma
•• Caroline microplate north of island

7 Ma
•• Pacific plate north of island
• Shortening/unroofing of crystalline basement under Irian fold- • Renewed uplift of Papuan Peninsula
and-thrust belt begins
• Volcanism begins along spine of western highlands

10 Ma
• No igneous activity into Australian crust • Ending of Maramuni arc magmatism into Australian
basement

12 Ma
•• Siliciclastic sediments flood carbonate shelf deposits of New Guinea Limestone Group

15 Ma
•• Beginning of siliciclastic deposition on oceanic forearc basement (Makats Formation)
• Peak Maramuni arc magmatism

ca. 25 to 20 Ma
• Subduction forming rocks in Ruffaer Metamorphic Belt from • Carbonate sedimentation blankets Sepik complex
metamorphism of Australian rise and slope sediments • Siliciclastic deposition continues in Aure Trough
• Beginning of Maramuni arc magmatism into Australian
basement by subduction at Trobriand Trough

ca. 33 to 31 Ma
•• Global sea level falls 100 m and forms unconformity on most of Australian shelf

ca. 38 to 30 Ma
• Collisional orogeny in eastern New Guinea forms Papuan Peninsula,
Sepik deformational and metamorphic belt, uplifts Papuan Ophiolite
• Aure Trough is relict trench

ca. 43 Ma
•• Change in Pacific plate motion forms west-dipping Izu-Bonin-Mariana and
Tonga-Kermadec subduction zones and Outer Melanesian Arc Terranes

62 to 56 Ma
• Coral Sea rifting forms Gulf of Papua

ca. 70 Ma
•• Australian plate arrives at northerly latitudes, and widespread carbonate sediment accumulation begins on shelf

ca. 100 to 80 Ma
•• Southwest-dipping subduction along northeastern edge of Australian plate;
amount of magmatism unknown, but probably minor

ca. 200 to 100 Ma

•• Slow passive margin subsidence, widespread mature quartz sandstone and shale (Kembelangan Group)

ca. 240 to 200 Ma

•• Rifting of northern margin of Australia begins and truncates Tasman orogenic belt and
forms broad 1000+ km wide shelf underlain by thin continental crust

ca. 440 to 350 Ma

• Precambrian basement blanketed by cratonal shelf strata • Tasman orogeny forms mid-Paleozoic mountain
Paleozoic belt that is deeply eroded
• = eastern or western half of island only
•• = islandwide, or global phenomenon
 Palinspastic Reconstruction of Collisional Delamination in New Guinea 31

is bulldozed. To differentiate this phase of tectonism, the term decapitation and other mechanical adjustments due to delamina-
“precollision complex” is used in this report. The progressively tion of the subducting lithosphere. From this perspective, it is
greater influx of passive margin strata (larger sediment supply evident that a transient phase of thin-skinned deformation dur-
as defined by Shreve and Cloos, 1986) would quickly cause a ing steady subduction predates thick-skinned deformation during
precollision complex to become emergent as the pile grows in collisional orogenesis.
thickness and width.
“Orogeny” is used in the classical sense as a term to describe 25 Ma: Intraoceanic Subduction
the growth of subaerial mountains, or in other words, a tec-
tonized landmass that will be subjected to erosion. In the case Jurassic oceanic crust that formed after Triassic rifting of
of the present-day island of New Guinea, the precollision com- the northern edge of the Australian plate was subducted beneath
plex was entirely submarine until small islands emerged at ca. a south-facing intraoceanic subduction zone (Fig. 16A). This
15 Ma. Emergent uplifts of passive margin strata with a lateral phase of convergence in the western Pacific started when north-
extent comparable to the present island started to shed large vol- ward subduction began beneath the Outer Melanesian arc,
umes of siliciclastic materials at ca. 12 Ma. This is considered the product of at least two phases of arc magmatism since the
the beginning of the Central Range orogeny proper. All of this Eocene (Coleman and Packham, 1976; Kroenke, 1984; Mil-
deformation is “thin-skinned” during the steady underthrusting som, 1985; Packham, 1996). This plate reorganization is almost
that predates the collisional jamming, at which time it becomes certainly due to the jamming of the Outer Melanesian subduc-
“thick-skinned.” The Australian crystalline basement underlying tion zone by the Ontong Java Plateau, by ca. 25 Ma (Kroenke,
the western Central Range did not become involved in the defor- 1984; Wells, 1989). The New Guinea Trench (NGT in Fig. 16A)
mation until ca. 8 Ma. This event marks the beginning of crustal and the magmatic complex that dates between 35 and 30 Ma

Figure 16. A: Lithospheric-scale cross section at 25 Ma. The Outer Melanesian Arc Terrane contains an old arc complex formed during south-
dipping subduction at the New Guinea Trench from ca. 30 Ma, when reversal occurred and a new arc began to form by north-dipping subduction.
Sediment derived from the Australian continent is deforming, recrystallizing, and accreting to the base of the forearc block, the Irian Ophiolite
Belt, forming the phyllitic rocks now exposed in the Ruffaer Metamorphic Belt. The Arafura Sea and southern New Guinea are underlain by
a broad, 1000 km wide zone of transitional continental crust that was stretched during Triassic rifting. Crust thickness and density profiles are
illustrated along with the position of neutral lithospheric buoyancy (see Cloos, 1993). AUS—Australian plate, PAC—Pacific plate. (Figure 16
continues on following pages.)
32 M. Cloos et al.

at Dabera in the Irian Ophiolite Belt (Fig. 9) are relicts of the phyllitic part of the metamorphic belt is best characterized as
older southwest-dipping subduction system. The initial forearc penetratively distributed, as would be expected if the protoliths
block (now the uplifted Irian Ophiolite) and associated trench were poorly lithified, Oligocene trench axis deposits.
were probably similar to the modern Mariana Trench (Bloomer,
1983) except for the fact that the hanging-wall block contained 15 Ma: Emergence of Isolated Islands
significant residual heat when subduction began because the arc
was just extinguished. Between 25 and 15 Ma, 500 km of oceanic lithosphere was
Subduction reversal occurred at ca. 30 Ma. Soon after, conti- subducted beneath the Outer Melanesian Arc Terrane. As the
nentally derived detritus that entered the trench were subducted. Australian continent approached, a progressively greater thick-
The deeply accreted materials are now exposed as the Ruffaer ness of abyssal plain and rise strata entered the subduction zone.
Metamorphic Belt. Peak temperatures of ~350 °C were attained By 15 Ma, small islands had emerged because the accretionary
between ca. 28 and 20 Ma (Weiland, 1999). The petrology of complex in front of the forearc block was 20 km or so thick (Fig.
the slates and phyllites indicate high-pressure greenschist facies 16B). It is most likely that emergence of small islands resulted
conditions with recrystallization at depths of 15–25 km (Warren, from the bulldozing of outer slope deposits of Jurassic to early
1995). The more phyllitic rocks are nearer the Irian Ophiolite Miocene age.
Belt (position A in Fig. 16A), and the cooling arc complex in the The evidence for subaerial exposure of the top of the precol-
hanging-wall block was a heat source causing metamorphism of lision complex comes from the deposition of the siliciclastic-rich
the underplated materials. Makats Formation (Fig. 16B) in the eastern North Coast Basin.
Whether the metamorphosed sediments are Jurassic to The basal Makats deposits are early Middle Miocene, 16–14 Ma.
Oligocene distal rise strata or distal Oligocene turbidites fun- By ca. 15 Ma, small islands along the spine of a rising forearc
neled far northwards along the trench is unclear. No distinctive high at the crest of a large accretionary complex were shedding
stratigraphic horizons have been found, hindering recognition debris northwards onto the forearc block. This material was
of any larger pattern of faulting and folding. Deformation in the uplifted and is now exposed in the Mamberamo region in the

15 Ma : Emergence of Isolated Islands on Forearc High

South North

accretionary
complex of
Mesozoic and Cenozoic Makats Formation
rise/slope sediments in forearc basin
Positive Negative
buoyancy buoyancy
Timika
shelf slope NCB NGT

Fore
arc
+ -

50 km/m.y.
forearc block
cooled; Ruffaer
Metamorphics
under
100 km Irian Ophiolite thermally
thinned
NCB = North Coast Basin lithosphere
NGT = New Guinea Trench under arc

Figure 16 (continued). B: Lithospheric-scale cross section at 15 Ma. Approximately 500 km of convergence has occurred since 25 Ma. The new
volcanic arc is well established, and the underlying lithosphere is thinner from the upward advection of heat in rising magma. Subduction of
cold oceanic lithosphere has caused substantial cooling of the base of the forearc block, the Irian Ophiolite, and metamorphic rocks are no longer
forming at the present depths of erosion into the Ruffaer Metamorphic Belt. Continental crust is shown about to underthrust, and rise strata from
Jurassic to Miocene age have been bulldozed, forming a thick precollision complex that is locally emergent and subject to erosion. Siliciclastic
detritus is shed northwards from isolated bathymetric highs to accumulate as the Makats Formation on top of the oceanic forearc block. The
trench axis depression is a barrier to the southward transport of sediment onto the shelf.
 Palinspastic Reconstruction of Collisional Delamination in New Guinea 33

north-central part of New Guinea near the international border. complex reached sufficient size to become a widespread source
The paleogeography was probably similar to that near Barbados of siliciclastic detritus (Fig. 16C). Continental slope and outer
Island in the Lesser Antilles subduction system. shelf deposits were bulldozed into a 500+ km long landmass,
Debris must also have been shed southwards, but this early hundreds of meters high, that shed detritus to the north and
stage of uplift and erosion is not recorded on the Australian con- south. The leading edge of the Australian continent was forced
tinental margin. A trench would act as a barrier to southward to bend downwards to subduct into the mantle.
sediment transport, and all trapped sediment would have been This is the date for initiation of the orogeny forming the
immediately underthrust. western Central Range because this is when widespread subaer-
The subduction of cold oceanic lithosphere refrigerated the ial erosion caused a distinct regional change in sedimentation.
base of the forearc block. By the time continental crust began to A trench no longer exists, possibly because it was filled with
underthrust, thermally driven metamorphic recrystallization was debris, but more likely because a thick pile of continental shelf
no longer occurring at depths less than 25 km (see Cloos, 1985, deposit had entered the zone of deformation. The carbonate shelf
for discussion of thermal evolution of subduction zones). Erosion (New Guinea Limestone Group) became widely flooded with
of the emergent parts of the forearc high is the first stage in the siliciclastic detritus (Buru Formation).
unroofing that eventually leads to the exposure of the Ruffaer From a geotectonic perspective, we think it is more appro-
Metamorphic Belt rocks from beneath the Irian Ophiolite Belt. priate to describe the deforming mass as a “precollision com-
At this time, erosion had probably only removed a kilometer or plex” because crystalline basement and lithospheric mantle
so of the rocks, and the metamorphic belt exposed today was still are not involved. The cover of sediment is becoming bulldozed
deeply buried beneath the forearc block. into a pile while the negatively buoyant Australian lithosphere
is steadily subducting. The mechanical behavior of the deform-
12 Ma: Beginning of the Central Range Orogeny ing sediment pile is changing, however. Before this time, most
of the strata involved in the deformation were recently depos-
Between 15 and 12 Ma, 150 km of plate convergence ited, lenticular, poorly lithified trench axis deposits, followed
occurred via the subduction of the oceanic lithosphere and tran- by rise and outer slope deposits that were also poorly lithified
sitional crust of the Australian plate. By 12 Ma, the accretionary because they were never deeply buried. Now, the deformation

Figure 16 (continued). C: Lithospheric-scale cross section at 12 Ma. The precollision complex has become an elongate landmass extending east-
west for 500+ km. Mud-rich continental slope and outer shelf deposits are eroding, and debris is shed northwards into the North Coast Basin on
top of oceanic basement and southwards to flood the carbonate shelf deposits on top of the Australian continental basement. The deformation is
thin-skinned as continental crust continues to underthrust and then bends downwards to subduct.
34 M. Cloos et al.

involves the thick pile of outer shelf strata, the bottom part At this time, underthrusting Australian continental base-
of which is composed of Mesozoic age strata that had time ment was bending downwards to subduct. The leading edge of
to compact ad lithify. Shelf deposits typically accumulate as the Australian continent was thin enough and thermal gradi-
well-bedded formations that extend laterally for many tens to ents were low enough that the crust was sufficiently welded to
hundreds of kilometers. At low temperatures, well-lithified, the underlying mantle that it readily moved to mantle depths.
well-layered formations act as beams that buckle and break as Most of the overlying pile of slope and shelf sediments are off-
the forces of convergence are transmitted through them. scraped while some of the lower part is underplating and thick-
At this stage, the growth of a precollision complex is far ening the orogenic belt.
from steady-state. The volume of sediments caught up in the
deformation is steadily increasing as the landward-thickening 10 Ma: Thin-Skinned Deformation
wedge of continental margin strata moves into the subduction
zone. The subduction bulldozing of a layered, highly aniso- Between 12 and 10 Ma, 100 km of plate convergence has
tropic stratigraphy generates kilometer-scale folds and fault occurred via subduction of transitional crust. At 10 Ma, Austra-
offsets. lian continental lithosphere continues to underthrust the precol-
The overall result is that the subduction bulldozing of a lision complex and then bend downwards to subduct (Fig. 16D).
layered, highly anisotropic stratigraphy generates folds and As the zone of neutral lithospheric buoyancy has only begun to
faults, that is, thin-skinned deformation. The factor(s) that underthrust, convergence continues unimpeded. Underthrusting
control whether folding or faulting is the dominant response crystalline basement is thicker and progressively less strongly
are not obvious, but in New Guinea a correlation is apparent. coupled to its mantle roots.
In the western Central Range, Precambrian basement under- The growth of the precollision complex still only involves
lies the strata, and folding is dominant. In the eastern Central thin-skinned deformation of the sedimentary cover. Folding is
Range of Papua New Guinea, Paleozoic basement underlies the dominant response in the top of the pile. Well-lithified outer
the strata, and imbricate thrusting is more common. shelf deposits of Cenozoic carbonate strata form the spectacular

Figure 16 (continued). D: Lithospheric-scale cross section at 10 Ma. Lithosphere with positive buoyancy is underthrusting the precollision
complex, but subducting continental lithosphere still has negative buoyancy. Thin-skinned deformation continues with the formation of kilome-
ter-scale folds as thick layers of carbonate shelf strata are bulldozed. The thickening of the sediment pile has formed an elongate landmass with
elevations up to ~2 km. From this stage onwards, bulldozing acts to widen the mountain belt more than causing an increase in surface eleva-
 Palinspastic Reconstruction of Collisional Delamination in New Guinea 35

kilometer-scale folds now found in the highlands. Convergence forearc basin. As the shallow waters near the rising landmass are
is oblique, and regionally the large folds develop a left-step- filled in, the shorelines migrate to the north and south. The rocks
ping en echelon pattern. Deeper in the pile are well-cemented of the Ruffaer Metamorphic Belt are moving slowly toward the
Jurassic-Cretaceous quartzose sandstones in the Kembelangan surface as the Irian Ophiolite Belt is progressively jacked up by
Group. These units were attached more firmly to the underly- underplating and subjected to ever increasing rates of erosion.
ing crust and underthrust to deeper levels before detaching. The
thick, well-cemented sandstone units appear to have resisted 8 Ma: Collisional Delamination Begins
folding and instead imbricated as southwest-vergent thrust
sheets. As this occurs, reverse or thrust faults propagate into the Between 10 and 8 Ma, 100 km of plate convergence has
overlying folds. occurred via the subduction of transitional crust. At 8 Ma, col-
The bulldozing of the incoming continental margin strata lisional orogeny actually begins with the uprooting of Australian
rapidly widens and thickens the pile. If the pile became 25 km crystalline basement. Thick-skinned deformation occurs because
thick, the resulting landmass would become ~2 km high. In the continental lithosphere of positive buoyancy was underthrust to
tropical climate, the limestones of the New Guinea Limestone the point that for continued movement, it must bend downwards
Group are largely removed by dissolution. The remaining silici- to subduct into the mantle (Fig. 16E). Almost concurrently,
clastic components and the debris from the Kembelangan and two mechanical adjustments occur: decapitation of the buoy-
older strata are shed to the south, flooding the northern shelf with ant continental basement blocks and thinning of the underlying
sand, silt, and shale (lower Buru Formation). Debris that was shed lithospheric mantle. The evidence for crustal involvement is the
to the north (adding to the Makats Formation) accumulated in the initiation of movement that formed the giant Mapenduma anti-

Figure 16 (continued). E: Lithospheric-scale cross section at 8 Ma. Collisional delamination begins as lithosphere with positive buoyancy
reaches the position where it must bend downward for subduction to continue. The first effect is the thick-skinned imbrication of continental
crust with decoupling in the lower crust. Shallower movements are localized by reactivation of faults formed during Mesozoic rifting. The pull
from the negative buoyancy of the lithosphere is transmitted updip through the cold, strong, upper part of the lithospheric mantle that continues
to subduct. Rupture of the lithospheric mantle nucleates in the region of high bending strains and propagates upwards (dashed zone in figure).
36 M. Cloos et al.

Figure 16 (continued). F: Lithospheric-scale cross section at 6 Ma. Continental blocks are detaching and moving more slowly than the mantle
underpinnings. Thick-skinned imbrication of basement occurs, and shortening by folding has largely ceased because the plate movements are
accommodated at deeper levels. The imbrication of continental blocks is much like the cars in a train wreck. Tectonic force from impingement of
the detached crust with the base of the hanging wall is transmitted southwards into previously undeformed Australian continent. The Mapenduma
anticline forms where the most southerly block is pushed up a reactivated Mesozoic normal fault. Another effect is that shortening in the arc
area inhibits the rise of magma and shuts off volcanism. As the lithospheric mantle continues to subduct, the ability of the mantle to respond by
plastic stretching is exceeded, and plate rupture occurs. Asthenosphere upwells as fast as lithospheric separation occurs, and magma is generated
by adiabatic decompression melting. Some magma is generated by the decompression melting of lithospheric mantle. Mafic magma tends to
pool near the base of the crust, forming magma chambers, heating wall rock, and assimilating continental crust. Some of this crustally contami-
nated magma rises to erupt along the spine of the highlands. Convergence continues to broaden the mountain belt. The removal of lithospheric
mantle causes an abrupt steepening of mountain slopes and an isostatic rise of as much as 2.5 km. This vertical uplift has a profound effect on the
pattern of denudation as orographically induced precipitation is concentrated on the flanks of the mountain. The steepening of mountain slope
and enhanced precipitation initiates widespread conglomeratic (molassic) sedimentation along the mountain flanks. AUS—Australian plate,
PAC—Pacific plate.

cline (Fig. 16F). The evidence for lithospheric mantle extension the southernmost imbricated crustal block begins to rise ~20 km
is the volcanism that starts at 7 Ma along the spine of the western south of the deformation front separating folded from unfolded
highlands. shelf strata.
The impingement of buoyant, weakly coupled crust with the The new toe for the mountain front was probably localized
base of the overriding plate initiates thick-skinned imbrication of where a Mesozoic normal fault was reactivated as a thrust ramp.
crystalline continental basement. In effect, a “train wreck” has The scale of crustal involvement is enormous. The Mapenduma
begun with the detachment of blocks of crust above shear zone(s) anticline is a 300 km long, 30 km wide fault-bend fold cored
in the ductile lower crust. Movement tends to be localized where by crystalline basement (Fig. 9) (see Weiland and Cloos, 1996).
ancient faults are reactivated in the brittle middle crust. The tip of Similar basement-cored uplifts that are ~50 km long formed to
 Palinspastic Reconstruction of Collisional Delamination in New Guinea 37

the west and east (Fig. 1). As the crystalline basement blocks Housh and McMahon, 2000). The volumetrically minor, but
move, the overlying precollision complex of deformed passive widespread, collision-generated magmatism that intruded into
margin strata begins to raft southwards (Fig. 7). Folding defor- deformed Australian passive margin strata is well dated in the
mation almost certainly slows and may even cease as plate con- western Central Range as a short-lived event between 7 Ma and
vergence is accommodated within the ductile lower crust. 2.5 Ma (O’Connor et al., 1994; McDowell et al., 1996).
Below the level of crustal decapitation, another important Magma intrusion and volcanism is concentrated along the
phenomenon begins to manifest itself. The pull force from the spine of the highlands because the highest elevations develop
negative buoyancy of the subducting lithosphere continues to above the zone of maximum asthenospheric upwelling and thus
transmit updip through the cold, strong, upper 30 km or so of the maximum magma generation. Mafic magma from the mantle
lithospheric mantle. Because relatively large differential stresses will pool and form chambers near the base of the crust. Lower
already exist in the area of bending, the hinge plays a critical crustal materials are assimilated by the engulfment of pieces
role in localizing plate separation. What starts as ductile neck- and convective mixing of melted wall rock. The generation of
ing (Fig. 16E) rapidly leads to rupture (Fig. 16F). The subduc- magma chambers and wall-rock heating further weakens the
tion pull force remains steady, but the local differential stresses crust and aids decoupling of the crust from the mantle.
rapidly increase as the neck thins. Whether ruptures propagate In the case illustrated for the western Central Range (Figs.
upwards or downwards is unclear, but the net result is that the 16E, 16F, and 16G), only a relatively limited stretching of the
cool, dense, and strong tongue of upper lithospheric mantle pulls underlying mantle occurs, and asthenosphere is shown to rise to
out from beneath the decapitated and imbricated blocks of crust the base of the crust. In other locations, mantle stretching may
and above the wedge of weak, nearly neutrally buoyant lower be substantial, and upwelling asthenosphere need not directly
lithospheric mantle. The asthenosphere is already at a condition contact the base of the crust. Whether the asthenosphere upwells
of incipient melting, and upwelling causes more melt to form. directly to the base of the crust or not must play a major role
As the underlying wedge of lithospheric mantle stretches and in controlling the amount of lower crustal melting that occurs.
decompresses, partial melting is initiated. It will be argued that along-strike variations in the degree of
stretching of lower lithospheric mantle can explain the lack of
6 Ma: Thick-Skinned Deformation and Magmatism magmatism and the lesser topography of the middle segment of
the Central Range.
Between 8 and 6 Ma, 100 km of additional plate conver-
gence has occurred. At 6 Ma, collisional delamination of the 4 Ma: Collisional Delamination is Complete
lithosphere is well under way. Thick-skinned deformation cre-
ates the south-vergent Mapenduma anticline. The effects of the Between 6 and 4 Ma, ~100 km of plate convergence should
collision are also transmitted northward, causing contraction of have been distributed across the island; however, the collision
the thermally weakened lithosphere beneath the arc, and volca- has also begun to affect areas far to the north. At ca. 4 Ma, delam-
nism along the north coast of the island shuts off. It is probable ination of the Australian lithosphere is complete (Fig. 16G).
that some convergence was accommodated by reactivation of Consequently, asthenospheric upwelling has ended, and magma
movement along the old subduction shear zone that reaches the generation is in the final stages. Magma chambers in the lower
surface at the New Guinea Trench. Beneath the orogenic belt, crust are solidifying faster than new melt is added. The front of
lithospheric mantle is replaced as asthenosphere rapidly upwells isostatic uplift has propagated northwards, and the edge of the
into the gap in the ruptured plate. As this occurs, the mountain- Irian Ophiolite Belt is tilted upwards as it rises vertically ~2 km.
ous belt with peaks that were 2–3 km or so tall and generated by At this time, weak asthenosphere underlies the Central
10 m.y. of contractional deformation undergoes a vertical iso- Range, and heating of the crust both conductively and from
static uplift in less than 2 m.y. that is as much as 2.5 km higher advection by rising magmas is at a maximum rate. Conse-
(timing and amounts discussed below) above the zone of maxi- quently, the rock strength beneath the mountain belt is low, and
mum upwelling. In other words, the core of the orogenic belt seismicity should be limited. This phenomenon would explain
rapidly rose vertically about as much as the surface rose during the perplexing scarcity of seismicity now detected in the Timor
the earlier prolonged phase of folding and faulting. region as noted by Milsom et al. (1983).
The upwelling asthenosphere undergoes adiabatic decom- The collision is having tectonic effects far to the north of the
pression melting in amounts proportional to the local amount of island. By the middle Cenozoic, plate motions between the Aus-
rise (volumes discussed below). The wedge of rising lower litho- tralian, Philippine, and Pacific plates were such that a prong of
spheric mantle extending north from beneath the Australian con- the Pacific plate extended between New Guinea and the Mariana
tinent is the piece that was beneath the pulled-out, strong tongue subduction zone. The collision caused a sufficient change in the
of uppermost mantle. This piece also undergoes decompression force balance on the prong that it broke off and became a kine-
melting, but to a far smaller degree. In western New Guinea, the matically distinct entity. The broken piece, the Caroline micro-
lithospheric mantle is Precambrian, which contributed a small, plate (Fig. 1), moved more westerly and slightly more slowly
but isotopically distinctive component to the magmas (see than the parent Pacific plate.
38 M. Cloos et al.

For the period of ca. 5–3 Ma (discussed more later), this It is inferred that initial strike-slip movements from Aus-
caused some Pacific-Caroline convergence at the Mussau tralian-Caroline plate interactions were concentrated in the
Trough and Caroline-Australian transcurrent motion along highlands because this is where the lithosphere was weakest.
New Guinea. The transcurrent motion was accommodated At this time, the crust directly overlies upwelled asthenosphere,
by strike-slip offsets distributed across the highlands (e.g., and magma has advected much heat to very shallow depths.
detected in the mining district [Sapiie and Cloos, 2004] and Near the surface, the nature of left-lateral strike-slip motion
along the Derewo fault zone in the metamorphic belt [Warren, in the highlands is strongly controlled by the local mechanical
1995]) and in the area of the recently extinct arc, forming the anisotropy of the upturned bedding and major high-angle fault
Yapen fault zone. zones. Magmas rise from lower crustal chambers commonly

Figure 16 (continued). G: Lithospheric-scale cross section at 4 Ma. Collisional delamination is complete, and the Australian continent abuts the
Caroline microplate. The underthrusting and imbrication of continental crust and the vertical isostatic rise driven by delamination cause addi-
tional uplift and upturning of the leading edge of the Pacific plate. The mantle underpinnings are exposed, forming the Irian Ophiolite Belt (IOB).
The mud-rich rocks overlying the underplated Ruffaer Metamorphic Belt (RMB) are highly erodible and rapid erosional unroofing is under way
with debris shed northwards into the North Coast Basin (NCB), which largely buries the accreted arc terrane. Jamming of the subduction zone
changes the force balance, and the prong of the Pacific plate between the Philippine and Australian (AUS) plates breaks off, temporarily forming
the Caroline microplate (CAR). The kinematics are such that left-lateral strike-slip motion occurs in the highlands underlain by thinned litho-
sphere, heated crust, and magma chambers, and in the recently extinct arc now located at the present-day north coast of New Guinea. Strike-slip
faulting in the highlands intersects lower crustal magma chambers, and passive intrusions occur along pull-apart pathways. Strong orographically
induced precipitation concentrates erosion along the southern and northern flanks of the highlands.
 Palinspastic Reconstruction of Collisional Delamination in New Guinea 39

by passive intrusion into pull-apart pathways along strike-slip 2 Ma: Postcollision

faults. Concurrent faulting and intrusion had a profound effect
on hydrothermal fluid flow. Cu-Au mineralization in the Erts- Between 4 and 2 Ma, ~200 km of left-lateral strike-slip
berg mining district occurs as both porphyry- and skarn-type motion was distributed across the island (see Fig. 1). By 2 Ma,
deposits at ca. 3 Ma (McDowell et al., 1996). the sinking plate has moved past the position of the Yapen fault
The two kilometers or so of rapid isostatic uplift steep- zone (Fig. 16H), and the magmatic event set off by delamination
ens slopes and greatly enhances orographically induced pre- has ended. The upwelled asthenosphere under the highlands is
cipitation, which concentrates erosion along the flanks of the cooling and converting to lithospheric mantle.
highlands (Weiland and Cloos, 1996). The south limb of the Strike-slip movements across the highlands probably only
Mapenduma anticline and the upturned ophiolitic rocks overly- totaled a few tens of kilometers before the faults deactivated
ing the metamorphic belt on the north slope are most affected. because plate motion became localized along the Yapen fault
The steepening of slopes lead to conglomeratic sedimentation zone. This fault zone is located along the axis of the just extin-
(molassic) along the flanks of the mountain belt (the boulder- guished, but still hot arc. This localization is due, in part, to the
rich upper Buru Formation). cooling and strengthening of the lithosphere beneath the high-

Figure 16. H: Lithospheric-scale cross section at 2 Ma. Plate motion is now focused along the Yapen fault zone in the center of the recently
extinct arc. This probably occurred because this zone of weakness had a trend that could accommodate the imposed movements as the corner
of the Caroline microplate ruptured, forming the Bismarck plate, and the corner of the Australian plate ruptured, forming the Solomon micro-
plate. The collisional delamination-generated magmatic event ends in the highlands as the lower crustal magma chamber solidifies. Upwelled
asthenosphere cools and transforms into lithospheric mantle. This drives a slow regional subsidence of the highlands that will continue for tens
of millions of years or until other plate-tectonic movements are initiated. Deep erosion is still concentrated on the flanks of the mountain belt.
RMB—Ruffaer Metamorphic Belt, AUS—Australian plate, PAC—Pacific plate.
40 M. Cloos et al.

lands, but more likely it is due to major plate-tectonic adjustment Melting in the asthenospheric mantle occurs because
in the neighborhood. The Caroline microplate became reattached there is an upwelling of as much as 100 km, resulting in adia-
to the Pacific plate as the Bismarck Sea spreading center started batic decompression as great as 30 kbar. This melt component
to open at 3.5 Ma (Fig. 1) (Taylor, 1979). This spreading center should have major, trace, and isotopic compositions similar to
directly links westward to the Bewani-Torricelli strike-slip zone mafic magmas at ocean ridges (McKenzie and Bickle, 1988).
and from there to the Yapen-Sorong fault zones via a broad con- Ocean ridge magmas are remarkably similar globally, and the
vergent kink in the Mamberamo area (Fig. 1). expected composition of asthenospheric melts beneath New
Guinea is considered known. Asthenospheric upwelling occurs
Present Day as fast as the lithosphere tear develops and the subducted
plate continues to sink away. The magma generation process
The current tectonic movements affecting New Guinea is physically the same as that beneath continental rift zones
are summarized in the beginning of this report, and only a few (e.g., Red Sea, Rio Grande). However, beneath New Guinea,
points need to be added. Roughly concurrent with the beginning the subterranean rifting event lasted only a few million years.
of spreading in the Bismarck Sea and the reattachment of the The reason plate separation occurs at very near the previous
Caroline microplate to the Pacific plate, the direction of Aus- speed of subduction is that the forces (negative buoyancy) driv-
tralian-Pacific plate motions changed by ~9° clockwise. Conse- ing the sinking of the subducted plate are little changed. In
quently, convergence in the region is more oblique than before fact, as rupturing is preceded by the decapitation of the crust,
ca. 4 Ma (Fig. 1). The change in the relative motion is due more the negative buoyancy of the subducting lithosphere actually
to a change in the Pacific plate (Cox and Engebretson, 1985; Pol- increases near the ruptured end. Because plate sinking is fast,
litz, 1986; Harbert and Cox, 1989, 1990) than the Australian plate melting during collisional delamination is more comparable
(Wessel and Kroenke, 2000; see also Scotese et al., 1988). to that at an ocean spreading center where spreading occurs
The lithosphere under the highlands should thicken at a at speeds of cm/yr rather than typical continental rifts where
rate proportional to the square root of time (Parsons and Sclater, separation commonly occurs slowly and episodically at speeds
1977). The thermal constant for thickening in this setting is of mm/yr for many millions of years.
approximated by analogy with the conductive thickening of As the bulk of primary melting occurs in the upwelled
oceanic lithosphere with age. Ten to 20 km of upwelled asthe- asthenospheric mantle, the volume of magma generation can be
nosphere should have converted to lithospheric mantle since 4 estimated by analogy with models for magma generation at rift
Ma. Cooling and lithospheric thickening causes an ever slowing zones (Pedersen and Ro, 1992). The cornerstone observation is
regional subsidence of the highlands that should now total ~400 that seafloor spreading generates a column of mafic crust that
m. Because a mountain belt is a site of erosion, this subsidence is is 7 ± 1 km thick regardless of spreading rate (McKenzie and
not readily detectable. Bickle, 1988; Stein and Stein, 1992). In the case of continental
Erosion is fastest along the flanks of the mountain belt rifting, the amount of melt generation depends upon the speed
because of orographically concentrated precipitation (Fig. 16G). of separation because the “walls” surrounding the upwelling
Stratigraphic relations indicate that at least 8 km and perhaps a asthenosphere are comparatively cool, and conductive heat
structural thickness of 12 km has been removed from the south loss is significant.
limb of Mapenduma anticline since 8 Ma. In the area of the Irian Using the rift zone melt model of Pedersen and Ro (1992),
Ophiolite Belt, as much as 25 km of material has been denuded. the height of the magma column is estimated for different
On the north flank, erosional unroofing began at ca. 15 Ma, but thicknesses of remaining lithosphere (Fig. 17). In western New
was probably minor until ca. 12 Ma. Average denudation rates Guinea, only crustal material remains, and the corresponding
on both flanks of the mountain belt have been in the range of 1–2 stretching factor, β, is ~5 (taking an original lithospheric thick-
km/m.y. (1–2 mm/yr) (see Fig. 9). ness of ~100 km and a present crustal thickness of ~20 km). A
melt column thickness of ~5 km appears probable. This corre-
COLLISIONAL DELAMINATION AND INDUCED sponds to a cross-sectional area of magma of ~500 km 2 per km
MAGMATISM strike length of delaminated area. Where lithospheric mantle
stretches, there is less upwelling, and the amount of melt gen-
Magmatism induced by collisional delamination has the erated is less. Where the remaining lithosphere is about half
distinctive attribute of occurring during and after the very lat- the original thickness (β ≈ 2), the melt column thickness is ~2.5
est contractional movements; that is, latest synorogenic and/or km. With lesser amounts of upwelling, the speed of subduction
postorogenic. In the physical model outlined above, there are and hence delamination is a proportionally more significant
one definite and two probable sources of melt: asthenosphere, factor. No melt should be generated in the upwelled astheno-
lithospheric mantle, and crust (Housh and McMahon, 2000). sphere if 50% lithospheric thinning occurs at speeds slower
The magmatic variations found in New Guinea are summa- than ~2 cm/yr.
rized below to highlight the obvious implications for the analy- Decompression of the lower lithospheric mantle also gen-
sis of similar igneous activity in other orogenic belts. erates primary mafic magma. Because this part of the mantle
 Palinspastic Reconstruction of Collisional Delamination in New Guinea 41

walls of large chambers. As batches of magma escape from the

mantle, they commonly intrude into, and mix with, variably
differentiated magmas in lower crustal chambers. Measure-
ment of Pb, Nd, and Sr isotopic ratios by Housh and McMahon
(2000) led them to conclude there was a component of at least
10%, to as much as 75%, crustally derived material incorpo-
rated into the suite of volcanic rocks they analyzed from across
the western highlands.
The conclusion is that a great diversity of magma composi-
tion should be expected along the strike of an orogen created by
collisional delamination. In the most detailed study of volcanic
rocks in western New Guinea, at the Minjauh volcanic field,
McMahon (2001) reported four distinct suites in an area less
than 25 km across: lamprophyres, syenites, shoshonites, and
calc-alkaline rocks. Their bulk, trace, and isotopic chemistries
record derivation from both lithospheric and asthenospheric
mantle followed by varying degrees of crustal incorporation.
Overall, the most distinctive chemical attribute of the primary
magmas was a high potassium content, which was diluted to
varying degrees by crustal assimilation.
The pattern and amount of upwelling depends upon the
degree of lithospheric stretching before tearing and the speed
Figure 17. Magma volumes during collisional delamination. Seafloor of delamination. The bulk of the primary magma should be
spreading involves complete plate separation and generates a layer of from the asthenosphere, and along strike it should be isotopi-
oceanic crust that is typically 7 km thick. Subterranean plate rupture cally similar. The rapid rise of this magma provides the source
during collisional delamination is more analogous to continental rift-
ing, where cold lithosphere separates into two pieces. Consequently, of heat for most of the crustal assimilation. The trace element
the volume of magma generated depends upon the rate of lithospheric and isotopic composition of lithospheric melts is a factor that
separation. Complete delamination beneath New Guinea occurs in 3 can be highly variable from place to place. The same is true
or 4 m.y. The lithospheric stretching factor, β, varies depending upon of crustal melt components. A great isotopic diversity has
how much lithosphere thinning occurred. With crustal imbrication been documented for the collisional magmatism in the west-
and complete removal of lithospheric mantle as shown in Figure 16G,
the β factor is ~5 in 3 m.y., and a magma column height of ~5–6 km ern (Housh and McMahon, 2000) and eastern (Hamilton et al.,
is expected. Where the lower lithospheric mantle has stretched and 1983) highlands, a strike length of more than 700 km. Addi-
the postcollision lithosphere thickness is about one-third the original tionally, in New Guinea, there is a clear isotopic manifestation
thickness, the β factor is ~2 and the melt column thickness is signifi- of the fact that the lithospheric basement is Precambrian in the
cantly less. Melt column thickness is calculated using the method of west and Phanerozoic in the east.
Pedersen and Ro (1992) and assuming an initial asthenosphere poten-
tial temperature (Fig. 15) of 1300 °C. Locations and approximate ag- A 300 km long stretch along the middle of the Central
gregate column thickness of dispersed melt are shown on Figures 16F Range (138° E to 141° E) is notable for an apparent lack of
and 16G. lf is final length; li is initial length igneous activity and elevations generally less than 2000 m
(Fig. 5). Although there may seem to be no reason to expect
a correlation between magmatism and elevation, the process
of collisional delamination actually predicts one. Complete
is fertilized over time by subduction events and other forms of removal of lithospheric mantle (and maximum magma genera-
mantle metasomatism, these melts are chemically (trace ele- tion) causes a vertical isostatic uplift of ~2 km beyond that due
ment and isotopically) distinctive and greatly different from to contraction-generated thickening of the crust and sediment
one part of the planet to another (Menzies et al., 1987; Hawkes- cover. Where lithospheric mantle remains, there will be pro-
worth et al., 1990; Foley, 1992). The amount of this type of portionally less isostatic uplift and less magma generation.
magma is small but readily detectable from isotopic analysis. Greater ductile necking during plate separation leaves
Measurement of Pb, Nd, and Sr isotopic ratios led Housh and more residual lithospheric mantle under the collisional moun-
McMahon (2000) to estimate there was a 2%–3% component tain belt (and less vertical isostatic uplift). If residual mantle
of lithospheric melt in a suite of volcanic rocks from across the was ~50 km thick (~50% thinning), a melt column ~2 km thick
western highlands of New Guinea. is expected (Fig. 17). But before this melt could manifest itself
Primary mafic magmas become modified as crustal com- at the surface, it must penetrate the residual lithospheric man-
ponents are assimilated by the incorporation of xenolithic tle. Most likely, melt rising beneath the amagmatic middle sec-
pieces or by the melting and convective entrainment of the tor of the Central Range solidified at depth.
42 M. Cloos et al.

It appears likely that the crust can be a direct source of

granitic magma where asthenosphere upwells to the Moho.
Kay and Kay (1993) argued that this occurred in the Cenozoic,
forming the Altiplano in the Andes of western South America.
Collins and Vernon (1994) argue that this also occurred in the
Paleozoic beneath the Lachlan fold belt of eastern Australia.
The kinematics of delamination envisioned by these workers is
different from that described above, but the geologic effects are
the same. Whether or not there is typically some stretched resid-
ual lithospheric mantle left in areas of collisional delamination
is unknown, but it seems very likely that stretching and residual
lithospheric mantle are greater where the speed of subduction is
slower. However, other factors must be involved also to produce
the variations in magmatism observed along the strike of the
Central Range of New Guinea.

COLLISIONAL DELAMINATION AND

SEDIMENTATION

The effect on regional sedimentation patterns from the col-

lisional delamination event in New Guinea has two character-
istics that should be recognizable in the rock record of ancient
orogenic belts: changes in provenance and style of deposition.
The reason for the first change is obvious. The generation
of topography during contractional tectonism creates uplifted
Figure 18. Average surface elevation and changes in average surface el-
areas that become new sediment sources. An increase in land evation corresponding to collisional delamination and removal of 50%
elevation of only 1 km has a large effect on airflow and localiz- or 100% of the lithospheric mantle. With an original total lithospheric
ing precipitation. Orographically induced precipitation should mantle thickness of 100 km and a imbricated crust thickness of 30 km,
be concentrated on the flanks of the uplift, and prevailing wind the initial mean surface elevation of ~700 m will increase ~1500 m
directions can greatly skew the erosion pattern to one side or with removal of 50% of the lithospheric mantle and ~2500 m with the
removal of 100% lithospheric mantle. Model parameters: lithospheric
another of a mountain belt (Beaumont et al., 1989; Hoffman mantle density = 3.30 g/cm3, crust density = 2.75 g/cm3, asthenosphere
and Grotzinger, 1993). The relationship of prevailing wind density = 3.23 g/cm3, consistent with those in Cloos (1993).
direction and the strike of the mountain belt will play a major
role controlling the overall pattern of denudation. The erosion
of a rising landmass will show up in the rock record as a change
in sedimentation pattern and types of clastic debris. In western (lower Buru Formation) changed to boulder-rich deposition
New Guinea, this occurred at ca. 12 Ma when shelf limestone (upper Buru Formation). A distinct change in depositional styles
deposition gave way to siliciclastic sedimentation with a source in orogenic belts has long been recognized as the phenomenon
to the north. of molassic sedimentation late in the history of mountain belts
The reason for the second sedimentological change is evi- (Pettijohn, 1975, p. 580).
dent from the physical changes during collisional delamination.
As the cooler, more dense lithospheric mantle is replaced by TIME-TRANSGRESSIVE OROGENESIS:
asthenosphere, a vertical isostatically driven uplift as great as PROPAGATING TEAR
2.5 km can occur very rapidly (Fig. 18). Where rock units are
horizontal and uplifted topography is gentle, as in the Colorado The Central Range–forming collision is slightly younger
Plateau, the effects of entrenchment of river channels induced in eastern New Guinea (Papua New Guinea) than in western
by rapid uplift that is spectacularly evident. Where topography New Guinea. With the exception of volumetrically minor, but
is irregular, as in a collision-generated fold-and-thrust belt, the economically significant, intrusive activity at Porgera, dated
effects of enhanced incision is far less obvious because there is magmatism in the eastern highland (Fly-Highlands province) is
a scarcity, if not a lack, of paleohorizontal markers. Nonethe- Pleistocene, and some of the volcanic centers may not be extinct.
less, the steepening of slopes would enhance erosion by land- In some areas, faults offset young volcanic flows (Davies, 1990,
sliding and produce an overall coarsening of sediment near the p. 262). The formation of basement-cored structures in Papua
mountain flanks. In western New Guinea, this occurred at ca. New Guinea appears to have begun between 4 and 2 Ma based
4 Ma when relatively fine-grained siliciclastic sedimentation on apatite fission track thermochronology (Hill and Gleadow,
 Palinspastic Reconstruction of Collisional Delamination in New Guinea 43

1989). Seismic activity indicates that convergent motions are West of New Guinea, in the Timor region, the Australian
active along the southern flank of the eastern Central Range continental margin is in the early stages of jamming the east-
(Fig. 3A). ern segment of the Sumatra-Java-Banda subduction zone (Mil-
The corresponding timing of each stage in Papua New som et al., 1983; McCaffrey et al., 1985; Milsom and Audley-
Guinea, the eastern end of the Central Range, is sequentially ~3 Charles, 1986; Silver et al., 1986). Timor consists of uplifted
m.y. younger. The younger timing is explainable as a result of a and imbricated Australian continental material. The arc north
west to east propagation of the tear in the subducting plate (Fig. of Timor is inactive, and there is detectable thrust deforma-
19) at a rate of ~150 km/m.y. (discussed below). tion in the backarc area. We consider the Timor region to be

Figure 19. Schematic diagram illustrating the tear propagating from west to east beneath the Central Range of New Guinea at ca. 5 Ma. Rupture
began at ca. 8 Ma beneath the Central Range, and the tear reached the eastern end at ca. 3 Ma, propagating at a rate of ~150 km/m.y. Beneath
the western Central Range, removal of lithospheric mantle appears to have been complete enough to generate much topography higher than 2
km and enough magma to cause significant volcanism in the western highlands. In the middle of the Central Range, from ~138° E to 140° E, it
appears that ductile stretching of the lithospheric mantle was substantial and thinning was less, perhaps to only one-half the original thickness
(not illustrated in this figure). In this area, the average relief is less than 2 km, less magma was generated, and none was able to reach the surface.
Beneath the eastern Central Range, the effects are similar to those in the western end, but collisional movements are still ongoing. A reason for
the anomalous behavior of the middle segment of the Central Range is unknown, but it may be related to the change from Precambrian litho-
sphere to the west to Phanerozoic lithosphere in the east, along what is known as the Tasman Line (Fig. 1).
44 M. Cloos et al.

comparable to the condition near the western Central Range differences in structural style along orogenic strike. The mag-
at ca. 7 Ma. nitude of lithospheric mantle stretching and the thickness of
residual lithospheric mantle must play a role in influencing the
ALONG-STRIKE VARIATIONS IN STRUCTURAL STYLE nature of large-scale deformation during the later stages of col-
lisional orogenesis. As discussed, the volume of magma gener-
The dominant style of deformation varies along the 1300 ation should vary in proportion to the amount of residual litho-
km long strike of the Central Range. In the area of our field spheric mantle. Likewise, whether the asthenosphere upwells
transect, a giant basement anticline, the Mapenduma anticline, to the Moho and rapidly heats and weakens the lower crust, or
formed, and the highlands are dominated by en echelon fold- an intervening layer of lithospheric mantle remains and limits
ing with subsidiary thrust faulting. Approximately 300 km east the rate and magnitude of lower crustal heating, must have a
of our transect, Granath and Argakoesoemah (1989, p. 82) and profound effect on late-stage tectonic movements.
Granath et al. (1991) concluded the structure is dominated by Determination of whether preexisting Precambrian, Paleo-
kilometer-scale thrust sheets with a left-stepping en echelon zoic, Mesozoic, or Cenozoic structural grains, stratigraphic
geometry. The geology of the eastern Central Range in Papua variations, or variations from heating and magmatism due to
New Guinea is the best understood segment of all because of delamination are the dominant controlling factor in the large-
drilling during hydrocarbon exploration (Carman and Carman, scale deformation patterns of collisional orogenesis in New
1990, 1993; Buchanan, 1996; Buchanan et al., 2000). Here, both Guinea is beyond the resolution of this investigation. But one
thin and thick-skinned thrust faulting occur with subsidiary fold- point is clear: Collisional delamination provides an explana-
ing (Hobson, 1986; Hill, 1989, 1991). tion for the change from thin-skinned to thick-skinned defor-
Primary factors that must have controlled the dominant mation in many orogenic belts.
structural style along the strike of the Central Range were the
differing basement on the two sides of the Tasman Line (Fig.
1) and the nature of basement weaknesses generated (or reacti-
vated) during Mesozoic rifting.
The western and eastern parts of the Central Range are
underlain by fundamentally different basement rock. Near the
international border at ~141° E, the Tasman Line separates base-
ment intensely deformed in the late Paleozoic and intruded in
the Mesozoic (Davies, 1990) from the Precambrian metasedi-
mentary and metaigneous basement that underlies most of the
Australian craton (Plumb, 1979a, 1979b; Hamilton, 1979, Fig.
120 therein). This boundary marks the western edge of the early
Paleozoic Tasman orogenic belt along the eastern Precambrian
edge of the Australian craton (Fig. 1). The basement of eastern
New Guinea must have a significant north-trending structural
grain from this major event.
Granath and Argakoesoemah (1989) believe differences in
the predeformational basement geometry controlled whether the
mountain belt is thin- or thick-skinned. Hill (1991), Buchanan
and Warburton (1996), Hill et al. (1996), and McConachie et
al. (2000) believe the thrust sheets that involve basement are
inverted Mesozoic normal faults.
Another factor is that the history of rifting and passive
margin development differed between western and eastern New Figure 20. Amounts of lithospheric thickening since delamination. The
Guinea (Granath and Hermeston, 1993). Early Mesozoic exten- rate of thickening depends primarily upon the thickness of the litho-
sphere (crust + mantle) at the end of delamination. Where the residual
sion with a roughly east trend occurred along the length of the
lithosphere is thinner, the initial rate of thickening is faster. The dif-
present Central Range. In contrast, late Mesozoic–early Ceno- ferences in the rate of thickening progressively decrease with time.
zoic rifting and opening of the Coral Sea propagated along a Where the initial postdelamination crustal thickness was ~30 km (the
northwest trend near the southeastern part of the range (Weissel thickest zone of imbricated crust and thickened sediment cover), ap-
and Watts, 1979). Both events must have produced normal fault proximately the upper 20 km of the upwelled asthenosphere has cooled
and been converted to lithospheric mantle. The rate of thickening, or
zones that were pronounced mechanical anisotropies likely to
“healing,” is approximated by analogy with that for thickening oceanic
become reactivated during collisional deformation. lithosphere formed at spreading ridges. In reality, the rate of thickening
But there is another factor, only recognizable in the con- is somewhat faster because the forearc block was previously cooled to
text of collisional delamination, that could cause significant below normal temperatures by subduction.
 Palinspastic Reconstruction of Collisional Delamination in New Guinea 45

LITHOSPHERIC HEALING AFTER DELAMINATION based upon the time when pull-apart spreading became active in
the Salawati Basin (Fig. 1) (Froidevaux, 1978; Charlton, 1996).
The delamination of the lithospheric mantle beneath the core These plate motion changes have caused the ongoing col-
of the highlands led to the rapid upwelling of the asthenosphere. lision in eastern New Guinea to differ from that in the west.
The upwelled material is cooling and converting to lithosphere at Spreading in the Bismarck Sea has created the Bismarck plate,
a rate that depends upon the thickness of the overlying sediment and the Finnisterre/Huon–New Britain Arc forearc is no longer
cover, crust, and residual lithospheric mantle. The rate of thick- part of the Pacific plate. The consequence of this is that the geom-
ening of the lithospheric mantle under mountain belts is approxi- etry of ongoing collision in eastern New Guinea is much less
mated by analogy with the thickening of oceanic lithosphere over oblique than it was before 3 Ma.
time. The initial rate of lithospheric mantle thickening is esti- New Guinea may be a bit unusual in that the later part of
mated to be similar to that occurring beneath oceanic lithosphere the collisional delamination process was concurrent with a major
of similar thickness (Fig. 20). Assuming an initial crustal thick- change in overall tectonic movements to ones dominated by strike
ness of 20 km beneath the spine of the western Central Range and slip (explained below). In many, if not most, collision zones, it
100% removal of the underlying lithospheric mantle, a 10–20 appears convergent motion continues after collisional delamina-
km thickness of asthenosphere has already been converted into tion begins. When this occurs, lithospheric shortening will occur
lithospheric mantle beneath the western Central Range. within the core of the collision-generated mountain belt and
Cooling and conversion of asthenospheric mantle to litho- within the region of the recently extinct arc. Under the arc, there
sphere will cause a steady subsidence of the mountain belt. is an ~50 km wide zone of thermally thinned lithosphere. Under
Again, the analogy with ocean spreading ridges, is appropriate. the mountain belt, there is typically an ~100 km wide zone lack-
After complete removal of lithospheric mantle, regional subsid- ing lithospheric mantle. Convergence at subduction speeds of a
ence will occur at an average rate of ~100 m/m.y. for the first few cm/yr would take only a few million years to contract the
10 m.y. after the event, steadily slowing thereafter. Of course, thermally thinned lithosphere under the arc region and fill the
mountain peaks can still be rising as the regional base level is gap under the arc and orogenic belt. As this occurs, lithospheric
lowering. shortening must occur within both areas. If the motion of the two
plates continues as before, a new subduction zone must form,
PLATE-TECTONIC CHANGES RELATED TO THE and it should typically nucleate within the thermally weakened
COLLISION FORMING NEW GUINEA: NEARBY AND area of oceanic arcs. Once a new subduction zone is established,
TRANSOCEANIC steady convergence, but with a reversal of polarity, can continue
for many tens of millions of years. In the case of New Guinea,
The jamming of a subduction zone will change the force bal- subduction reversal appears to be just starting after an ~5 m.y.
ance on the edges of the colliding plates. It would seem that the period of adjustment dominated by strike-slip, transform move-
magnitude of the change should be proportional to the length of ments. The New Guinea Trench is reactivating from the east to
the collision zone and the geometry of the relative motions. Two the west (Figs. 1 and 4).
end-member tectonic responses appear likely: continued plate The change in force balance was not only sufficient to break
convergence as before (at least for a short time) prior to subduc- pieces off the nearby corners of the Pacific and Australian plates,
tion reversal behind the oceanic arc, or a fundamental change but enough to cause an ~9° clockwise rotation of the giant Pacific
in the relative motion between the colliding plates. The latter plate. The effects are detectable across the Pacific Basin. The San
occurred, at least to some degree, as the aftermath of the oblique Andreas system of California (Pollitz, 1988; Harbert, 1991; Page
collision forming New Guinea. and Brocher, 1993), the Queen Charlotte system in southeastern
As discussed, the Caroline microplate formed as the prong Alaska (Fitzgerald et al., 1993; Hyndman and Hamilton, 1993),
of the Pacific plate wedged between the Australian and Philippine and the Alpine system of New Zealand (Adams, 1981) have all
plates, broke off, and began to move as a distinct kinematic entity become slightly convergent transform margins since the Pliocene.
at ca. 5 Ma. This movement produced strike-slip faulting in the Significant adjustments along the East Pacific Rise, besides small
collision-generated New Guinea highlands and in the still-warm changes in spreading rate and optimal transform orientation, are
core of the accreted arc terrane. At ca. 3.5 Ma, two other plate the rupture of plate corners along the transform forming the Eas-
ruptures occurred nearby, which must also have been a manifes- ter and Juan Fernandez microplates (Searle et al., 1993). Because
tation of the collision forming the island (Fig. 4). A corner of subduction zones ring most of the Pacific, the small change in
the Pacific plate broke off, forming the Bismarck microplate Pacific plate motion makes only a small change in the speed and
(Taylor, 1979), and a propagating tear in the northeast corner of direction of convergence that by itself is probably essentially
the Australian plate formed the Woodlark spreading center east undetectable in the rock record. However, in some places, such
of the island (Weissel et al., 1982; Honza et al., 1987). As these as the Alaska-Aleutian Trench, the uplift of coastal mountains
plate ruptures occurred and the spreading center opened, strike- has greatly increased the sediment supply to the trench, and this
slip activity became concentrated along the Yapen-Sorong and in turn can cause a profound change in accretionary patterns (see
Bewani-Torricelli fault zones, which became active at ca. 4 Ma the role of sediment supply in Shreve and Cloos, 1986).
46 References Cited

CONCLUSIONS Vice Presidents Dave Potter and Steve Van Nort encour-
aged the overall program and provided exceptional advice.
The formation of the island of New Guinea has long been Mine District Geologists Tom Collinson, Dave Mayes, Chuck
cited as the product of a Cenozoic arc-continent collision. Brannon, Al Edwards, and Larry Segretson provided guidance
Geologic studies in the Central Range of western New Guinea and assistance for the studies in the Ertsberg (Gunung Bijih)
(near Puncak Jaya, 4884 m) have revealed field and timing mining district. Head Exploration Geologists Peter Doyle,
relationships, which combined with mechanical considerations Gary Artmont, Wahyu Sunyoto, Gary O’Connor, and Sugeng
leads to a refined model for the tectono-magmatic effects of Widodo provided guidance and assistance for our regional
collisional orogenesis. studies that rode on the coattails of the minerals exploration
North-dipping subduction of the oceanic end of the Aus- program.
tralian plate began at ca. 30 Ma at an intraoceanic subduc- Geologists and related staff who provided discussion
tion zone, following a subduction reversal event beneath the and special assistance during one or more phases of our work
Outer Melanesian Arc. Sediment accretion and metamorphism include Kris Hefton, Jay Pennington, George MacDonald,
began soon after as continental sediment was transported far Uttu Mekel, Keith Parris, Bambang Trisetyo, Lasito Soebari,
out to sea along the trench. Peak metamorphic temperatures Peter Sedgwick, Ian Tasiran, Art Ona, Mark Gilliam, Boedi-
were attained between 28 and 20 Ma. The top of an accretion- yono, Imants Kavalieris, Mike Burson, Larry Johnson, Soeba-
ary prism formed small islands by ca. 15 Ma. By ca. 12 Ma, a gio Sopareman, Louie Bell, Sanusi Sukarya, Bambang Irawan,
widespread landmass was shedding siliciclastic detritus to the Eddy Suwardy, Erwin Siregar, Judihandri, Gandhi Hermawan,
south, flooding the carbonate shelf. Jamming of the subduc- Nurhamid, Nurhadi, Eddy Bastomi, Herry Susanto, Djoko
tion zone by the underthrusting of the Australian continental- Waskito, Djuharlan, Budi Cahyono, Rudy Poedjono, Roga
crust-capped plate began at ca. 8 Ma, with the detachment and Pendawalima, Syaiful Bahri, Doug MacKenzie, Frachrudin
southward displacement of the large basement block forming Ali, Clark Arnold, Allen Schappert, Hamish Campbell, Joe
the Mapenduma anticline. The subducted oceanic end of the McPherson, Ben Coutts, Dean Wadsworth, Fanfan, Slamet
Australian plate continued to sink and broke off. The subter- Riyanto, Steve Hughes, Ahmed, Ari Suldjana, and Rahmat
ranean rifting, or delamination, of the Australian plate caused Wirawan.
a magma generation event beneath the western Central Range An extra special thanks to our Papuan field assistants,
from 7 to 4 Ma. Concurrently, the upwelling of asthenosphere Ayub “Obi” Zagani, Amelius Beanal, Julianus Magal, Etinus
caused a rapid, isostatically driven vertical uplift as great as 2.5 Tabuni, Dominikus Mom, Tiranus Beanal, Benny Dolame,
km of the collision-generated fold belt. This caused a profound Peniel Kogoya, Petrus Mote, Emi Tabuni, Lanus Owaling, and
change in regional sedimentation, forming molassic boulder many others who showed us the natural wonders of their home-
beds near the mountain flanks. land.
Collisional delamination is a common and fundamental We also thank all of our colleagues at our collaborating
plate-tectonic process that occurs when subduction zones are university, the Institut Teknologi Bandung (ITB). We specifi-
jammed by continental margins or very large oceanic arc/pla- cally thank Professors Iman Soengkowo, R.P. Koesoemadinata,
teau complexes. Subterranean plate rifting is as fundamental Rubini Soeria-Atmadja, Yahdi Zaim, and Emmy Suparka.
to the end of the long-term history of steady subduction as the We thank Fred McDowell, Jim Connelly, Todd Housh,
rifting process is to the beginning of seafloor spreading and W.R. Muehlberger, and W.R. Dickinson for reviews of this
long-term, steady subsidence of a passive margin. paper. Ertsberg Project Contribution No. 25.

ACKNOWLEDGMENTS REFERENCES CITED

This paper is based primarily upon theses and dissertations Abbott, L.D., Silver, E.A., Thompson, P.R., Filewicz, M.V., Schneider, C., and
Abdoerrias, 1994a, Stratigraphic constraints on the development and tim-
by students at the University of Texas, Austin, that were ing of arc-continent collision in northern Papua New Guinea: Journal of
made possible by the assistance of the P.T. Freeport Indonesia Sedimentary Research, v. 64, p. 169–183.
(PTFI) exploration and mine geology departments. The 200+ Abbott, L.D., Silver, E.A., and Galewsky, J., 1994b, Structural evolution of a
modern arc-continent collision in Papua New Guinea: Tectonics, v. 13,
geoscientists who staffed these departments during the 1990s p. 1007–1034, doi: 10.1029/94TC01623.
provided the foundation for selecting sites for our fieldwork Abers, G., and McCaffrey, R., 1988, Active deformation in the New Guinea
that was followed by laboratory analysis of collected samples. fold-and-thrust belt: Seismological evidence for strike-slip faulting and
basement involved thrusting: Journal of Geophysical Research, v. 93,
We especially thank James R. Moffett, president and CEO of p. 13,332–13,354.
Freeport McMoRan and Robert E. Boyer, dean and professor Adams, C.J., 1981, Uplift rates and thermal structure in the Alpine fault zone
of geological sciences at the University of Texas at Austin, and Alpine schists, southern Alps, New Zealand, in Price, N.J., and
McClay, K.R., eds., Thrust and nappe tectonics: Geological Society [Lon-
whose vision made the Ertsberg Project a unique industry- don] Special Publication 9, p. 211–222.
academia initiative to better understand the geology of western Anderson, D.L., 1989, Theory of the Earth: Boston, Blackwell Scientific
New Guinea. Publications, 366 p.
 References Cited 47

Atkinson, B.K., 1987, Introduction to fracture mechanics and its geophysical ical Society of America Bulletin, v. 85, p. 553–570, doi: 10.1130/0016-
applications, in Atkinson, B.K., ed., Fracture mechanics of rock: Lon- 7606(1974)85<553:GOCTHP>2.0.CO;2.
don, Academic Press, p. 1–26. Charlton, T.R., 1996, Correlation of the Salawati and Tomori Basins, eastern
Australasian Petroleum Company, 1961, Geological results of petroleum Indonesia: A constraint on left-lateral displacements of the Sorong fault
exploration in western Papua, 1937–1961: Geological Society of Austra- zone: Geological Society [London] Special Publication 106, p. 465–
lia Journal, v. 8, p. 1–133. 481.
Bär, C.B., Cortel, H.J., and Escher, A.E., 1961, Geological results of the Star Cloos, M., 1985, Thermal evolution of convergent plate margins: Thermal
Mountains (“Sterrengebergte”) Expedition: Nova Guinea, Geology, modeling and re-evaluation of isotopic Ar-ages for blueschists in the
v. 4, p. 40–99. Franciscan complex of California: Tectonics, v. 4, p. 421–433.
Beaumont, C., Fullsack, P., and Hamilton, J., 1989, Erosional control of active Cloos, M., 1992a, Thrust-type subduction-zone earthquakes and seamount
compressional orogens, in McClay, K.R., ed., Thrust tectonics: New asperities: A physical model for seismic rupture: Geology, v. 20, p. 601–
York, Chapman and Hall, p. 1–18. 604, doi: 10.1130/0091-7613(1992)020<0601:TTSZEA>2.3.CO;2.
Bird, P., 1978, Initiation of intracontinental subduction in the Himalaya: Jour- Cloos, M., 1992b, Origin of the Caroline block and plate: Tectonic response
nal of Geophysical Research, v. 83, p. 4975–4987. to changes in Pacific plate motion at 43 and 4 Ma: Geological Society of
Bird, P., 1979, Continental delamination and the Colorado Plateau: Journal of America Abstracts with Programs, v. 24, no. 7, p. A185.
Geophysical Research, v. 84, p. 7561–7571. Cloos, M., 1993, Lithospheric buoyancy and collisional orogenesis: Subduction
Bird, P., and Baumgardner, J., 1981, Steady propagation of delamination of oceanic plateaus, continental margins, island arcs, spreading ridges,
events: Journal of Geophysical Research, v. 86, p. 4891–4903. and seamounts: Geological Society of America Bulletin, v. 105, p. 715–
Bladon, G.M., 1988, Catalogue, appraisal, and significance of K-Ar isotopic 737, doi: 10.1130/0016-7606(1993)105<0715:LBACOS>2.3.CO;2.
ages determined for igneous and metamorphic rocks in Irian Jaya: Irian Cloos, M., 1997a, Geology and the Grasberg: A model for joint industry and
Jaya Geological Mapping Project, Geological Research and Develop- academic research: Geotimes, v. 42, no. 9, p. 19–22.
ment Centre, Indonesia, in cooperation with the Bureau of Mineral Cloos, M., 1997b, Anatomy of a mine: The discovery and development of
Resources, Australia, on behalf of the Department of Mines and Energy, Grasberg: Geotimes, v. 42, no. 1, p. 16–20.
Indonesia, and the Australian Development Assistance Bureau, 75 p. Cloos, M., and Shreve, R.L., 1988a, Subduction-channel model of prism accre-
Bloomer, S.H., 1983, Distribution and origin of igneous rocks from the land- tion, melange formation, sediment subduction, and subduction erosion
ward slopes of the Mariana Trench: Implications for its structure and at convergent plate margins: 1. Background and description: Pure and
evolution: Journal of Geophysical Research, v. 88, p. 7411–7428. Applied Geophysics, v. 128, p. 455–500, doi: 10.1007/BF00874548.
Bond, G.C., and Kominz, M.A., 1988, Evolution of thought on passive conti- Cloos, M., and Shreve, R.L., 1988b, Subduction-channel model of prism accre-
nental margins from the origin of geosynclinal theory (ca. 1860) to the tion, melange formation, sediment subduction, and subduction erosion
present: Geological Society of America Bulletin, v. 100, p. 1909–1933, at convergent plate margins: 2. Implications and discussion: Pure and
doi: 10.1130/0016-7606(1988)100<1909:EOTOPC>2.3.CO;2. Applied Geophysics, v. 128, p. 501–545, doi: 10.1007/BF00874549.
Brace, W.F., and Kohlstedt, D.L., 1980, Limits on lithospheric stress imposed Cloos, M., McMahon, T.P., Quarles van Ufford, A., Sapiie, B., Warren, P.Q.,
by laboratory experiments: Journal of Geophysical Research, v. 85, and Weiland, R.J., 1994, Collisional delamination in New Guinea: Geo-
p. 6248–6252. logical Society of America Abstracts with Programs, v. 26, no. 7, p. A-
Brown, C.M., Pigram, C.J., and Skwarko, S.K., 1979, Mesozoic stratigraphy 502.
and geological history of Papua New Guinea: Palaeogeography, Pal- Coleman, P.J., and Packham, G.H., 1976, The Melanesian borderlands and
aeoclimatology, Palaeoecology, v. 29, p. 301–322, doi: 10.1016/0031- India-Pacific plates’ boundary: Earth Science Reviews, v. 12, p. 197–
0182(79)90087-7. 233, doi: 10.1016/0012-8252(76)90005-2.
Buchanan, P.G., editor, 1996, Petroleum exploration, development and pro- Collins, W.J., 1994, Upper- and middle-crustal response to delamina-
duction in Papua New Guinea: Proceedings of the Third Papua New tion: An example from the Lachlan fold belt, eastern Australia:
Guinea Petroleum Convention, Port Moresby, 9–11 September 1996, Geology, v. 22, p. 143–146, doi: 10.1130/0091-7613(1994)022<0143:
822 p. UAMCRT>2.3.CO;2.
Buchanan, P.G., and Warburton, J., 1996, The influence of preexisting basin Collins, W.J., and Vernon, R.H., 1994, A rift-drift-delamination model of con-
architecture in the development of the Papuan fold and thrust belt: tinental evolution: Paleozoic tectonic development of eastern Australia:
Implications for petroleum prospectivity, in Buchanan, P.G., ed., Petro- Tectonophysics, v. 235, p. 249–275, doi: 10.1016/0040-1951(94)90197-
leum exploration, development and production in Papua New Guinea: X.
Proceedings of the 3rd Papua New Guinea Petroleum Convention, Port Cooper, P., and Taylor, B., 1987, Seismotectonics of New Guinea: A model for
Moresby, 9–11 September 1996, p. 89–109. arc reversal following arc-continent collision: Tectonics, v. 6, p. 53–68.
Buchanan, P.G., Grainge, A.M., and Thornton, C.N., editors, 2000, Papua Cox, A., and Engebretson, D., 1985, Change in motion of Pacific plate at 5
New Guinea’s petroleum industry in the 21st century: Proceedings of Myr BP: Nature, v. 313, p. 472–474, doi: 10.1038/313472a0.
the Fourth Papua New Guinea Petroleum Convention, Port Moresby, Cullen, A.B., and Pigott, J.D., 1989, Post-Jurassic tectonic evolution of Papua
29–31 May 2000, 593 p. New Guinea: Tectonophysics, v. 162, p. 291–302.
Cardwell, R.K., Isacks, B.L., and Karig, D.E., 1980, The spatial distribution Davies, H.L., 1971, Peridotite-gabbro-basalt complex in eastern Papua: An
of earthquakes, focal mechanisms solutions, and subducted lithosphere overthrust plate of oceanic mantle and crust: Australian Bureau of Min-
in the Philippine and northeastern Indonesian islands: American Geo- eral Resources, Geology and Geophysics Bulletin No. 128, 48 p.
physical Union Monograph 23, p. 1–35. Davies, H.L., 1990, Structure and evolution of the border region of New
Carman, G.J., and Carman, Z., editors, 1990, Petroleum exploration in Papua Guinea, in Carman, G.J., and Carman, Z., eds., Petroleum exploration in
New Guinea: Proceedings of the First Papua New Guinea Petroleum Papua New Guinea: Proceedings of the First Papua New Guinea Petro-
Convention, Port Moresby, 12–14 February 1990, 597 p. leum Convention, Port Moresby, February 12–14, 1990, p. 245–269.
Carman, G.J., and Carman, Z., editors, 1993, Petroleum exploration in Papua Davies, H.L., Lock, J., Tiffi n, D.L., Honza, E., Okuda, Y., Murakami, F.,
New Guinea: Proceedings of the Second Papua New Guinea Petroleum and Kisimoto, K., 1987a, Convergent tectonics in the Huon Peninsula
Convention, Port Moresby, 31 May–2 June 1993, 687 p. region, Papua New Guinea: Geo-Marine Letters, v. 7, p. 143–152.
Carter, N.L., 1976, Steady-state flow of rocks: Reviews of Geophysics and Davies, H.L., Honza, E., Tiffi n, D.L., Lock, J., Okuda, Y., Keene, J.B.,
Space Physics, v. 14, p. 301–360. Murakami, F., and Kisimoto, K., 1987b, Regional setting and structure
Carter, N.L., and Tsenn, M.C., 1987, Flow properties of continental lithosphere: of the western Solomon Sea: Geo-Marine Letters, v. 7, p. 153–160.
Tectonophysics, v. 136, p. 27–63, doi: 10.1016/0040-1951(87)90333-7. Davies, J.H., and von Blanckenburg, F., 1995, Slab breakoff: A model of litho-
Casey, J.F., and Dewey, J.F., 1984, Initiation of subduction zones along sphere detachment and its test in the magmatism and deformation of
transform and accreting plate boundaries, triple-junction evolution, collisional orogens: Earth and Planetary Science Letters, v. 129, p. 85–
and forearc spreading centres—Implications for ophiolitic geology 102, doi: 10.1016/0012-821X(94)00237-S.
and obduction: Geological Society [London] Special Publication 13, DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S., 1990, Current plate
p. 269–290. motions: Geophysical Journal International, v. 101, p. 425–478.
Chappell, J., 1974, Geology of coral terraces, Huon Peninsula, New Guinea: A DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S., 1994, Effect of recent
study of quaternary tectonic movements and sea-level changes: Geolog- revisions to the geomagnetic reversal time scale on estimates of current
48 References Cited

plate motions: Geophysical Research Letters, v. 21, p. 2191–2194, doi: Harbert, W., 1991, Late Neogene relative motions of the Pacific and North
10.1029/94GL02118. American plates: Tectonics, v. 10, p. 1–15.
Dewey, J.F., and Bird, J.M., 1970, Mountain belts and the new global tectonics: Harbert, W., and Cox, A., 1989, Late Neogene motion of the Pacific plate:
Journal of Geophysical Research, v. 75, p. 2625–2647. Journal of Geophysical Research, v. 94, p. 3052–3064.
Dewey, J.F., Ryan, P.D., and Anderson, T.B., 1993, Orogenic uplift and collapse, Harbert, W., and Cox, A., 1990, Correction to “Late Neogene motion of the
crustal thickness, fabrics and metamorphic changes: The role of eclogites: Pacific plate”: Journal of Geophysical Research, v. 95, p. 5171.
Geological Society [London] Special Publication 76, p. 325–343. Hawkesworth, C.J., Kempton, P.D., Rogers, N.W., Ellam, R.M., and van Cal-
Dow, D.B., 1968, A geological reconnaissance in the Nassau Range, west New steren, P.W., 1990, Continental mantle lithosphere, and shallow level
Guinea: Geologie en Mijnbouw, v. 47, p. 37–46. enrichment processes in the Earth’s mantle: Earth and Planetary Sci-
Dow, D.B., 1977, A geological synthesis of Papua New Guinea: Australian Bureau ence Letters, v. 96, p. 256–268, doi: 10.1016/0012-821X(90)90006-J.
of Mineral Resources, Geology and Geophysics Bulletin 201, 41 p. Hefton, K., and Pennington, J., 1994, Geologic map of COW-A (unpublished):
Dow, D.B., and Sukamto, R., 1984a, Western Irian Jaya: The end-product of P.T. Freeport Indonesia, scale 1:10,000, 1 sheet.
oblique plate convergence in the Late Tertiary: Tectonophysics, v. 106, Hegarty, K.A., and Weissel, J.K., 1988, Complexities in the development of
p. 109–139, doi: 10.1016/0040-1951(84)90224-5. the Caroline plate region, western equatorial Pacific, in Nairn, et al.,
Dow, D.B., and Sukamto, R., 1984b, Late Tertiary to Quaternary tectonics of eds., The ocean basins and margins, volume 7B: The Pacific Ocean:
Irian Jaya: Episodes, v. 7, p. 3–9. New York, Plenum Press, p. 277–301.
Dow, D.B., Robinson, G.P., Hartono, U., and Ratman, N., 1986, Geologic map Hegarty, K.A., Weissel, J.K., and Hayes, D.E., 1983, Convergence at the
of Irian Jaya, Indonesia: Bandung, Indonesia, Geological Research and Caroline-Pacific plate boundary: Collision and subduction: American
Development Centre, Indonesian Ministry of Mines and Energy, scale Geophysical Union Monograph 27, v. 2, p. 326–348.
1:1,000,000, 2 sheets. Hilde, T.W.C., Uyeda, S., and Kroenke, L., 1977, Evolution of the west-
Dow, D.B., Robinson, G.P., Hartono, U., and Ratman, N., 1988, Geology of Irian ern Pacific and its margin: Tectonophysics, v. 38, p. 145–165, doi:
Jaya: Irian Jaya Geological Mapping Project, Geological Research and 10.1016/0040-1951(77)90205-0.
Development Centre, Indonesia, in cooperation with the Bureau of Mineral Hill, K.C., 1989, The Muller anticline, Papua New Guinea: Basement-cored,
Resources, Australia, on behalf of the Department of Mines and Energy, inverted extensional fault structures with opposite vergence: Tectono-
Indonesia, and the Australian Development Assistance Bureau, 298 p. physics, v. 158, p. 227–245, doi: 10.1016/0040-1951(89)90326-0.
Dow, D.B., Harahap, B.H., and Sufni, H.A., 1990, Geology of the Enarotali Hill, K.C., 1991, Structure of the Papuan fold belt, Papua New Guinea: Amer-
sheet area, Irian Jaya: Geological Research and Development Centre, ican Association of Petroleum Geologists Bulletin, v. 75, p. 857–872.
Department of Mines and Energy, Bandung, Indonesia, scale 1:250,000, Hill, K.C., and Gleadow, A.J.W., 1989, Uplift and thermal history of the Pap-
1 sheet, 57 p. uan fold belt, Papua New Guinea: Apatite fission-track analysis: Austra-
Finnerty, A.A., and Boyd, F.R., 1987, Thermobarometry for garnet peridotites: lian Journal of Earth Sciences, v. 36, p. 515–539.
Basis for the determination of thermal and compositional structure of Hill, K.C., Simpson, R.J., Kendrick, R.D., Crowhurst, P.V., O’Sullivan, P.B.,
the upper mantle, in Nixon, P.H., ed., Mantle xenoliths: New York, John and Saefudin, I., 1996, Hydrocarbons in New Guinea, controlled by
Wiley and Sons, p. 381–402. basement fabric, Mesozoic extension and Tertiary convergent margin
Fitzgerald, P.G., Stump, E., and Redfield, T.F., 1993, Late Cenozoic uplift of tectonics, in Buchanan, P.G., ed., Petroleum exploration, development
Denali and its relation to relative plate motion and fault morphology: Sci- and production in Papua New Guinea: Proceedings of the 3rd Papua
ence, v. 259, p. 497–499. New Guinea Petroleum Convention, Port Moresby, 9–11 September
Foley, S., 1992, Vein-plus-wall-rock melting mechanisms in the lithosphere 1996, p. 63–76.
and the origin of potassic alkaline magmas: Lithos, v. 28, p. 435–453, doi: Hobson, D.M., 1986, A thin-skinned model for the Papuan thrust belt and
10.1016/0024-4937(92)90018-T. some implications for hydrocarbon exploration: Australian Petroleum
Froidevaux, C.M., 1978, Tertiary tectonic history of the Salawati Area, Irian Exploration Association Journal, v. 26, p. 214–224.
Jaya, Indonesia: American Association of Petroleum Geologists Bulletin, Hoffman, P.F., and Grotzinger, J.P., 1993, Orographic precipitation, ero-
v. 62, p. 1127–1150. sional unloading, and tectonic style: Geology, v. 21, p. 195–198, doi:
Gibson, S.A., Thompson, R.N., Leat, P.T., Dickin, A.P., Morrison, M.A., Hen- 10.1130/0091-7613(1993)021<0195:OPEUAT>2.3.CO;2.
dry, G.L., and Mitchell, J.G., 1992, Asthenosphere-derived magmatism in Honnorez, J., Mevel, C., and Montigny, R., 1984, Occurrence and signifi-
the Rio Grande rift, western USA: Implications for continental break-up: cance of gneissic amphibolites in the Vema fracture zone, equatorial
Geological Society [London] Special Publication 68, p. 61–89. Mid-Atlantic Ridge: Geological Society [London] Special Publication
Granath, J.W., and Argakoesoemah, R.M.I., 1989, Variations in structural style 13, p. 121–130.
along the eastern Central Range thrust belt: Proceedings, Indonesian Honza, E., Davies, H.L., Keene, J.B., and Tiffi n, D.L., 1987, Plate boundar-
Petroleum Association, v. 18, p. 79–89. ies and evolution of the Solomon Sea region: Geo-Marine Letters, v. 7,
Granath, J.W., and Hermeston, S.A., 1993, Relationship of the Toro Sandstone p. 161–168.
Formation and the Alene Sands of Papua to the Woniwogi Formation of Houseman, G.A., McKenzie, D.P., and Molnar, P., 1981, Convective instabil-
Irian Jaya, in Carman, G.J., and Carman, Z., eds., Petroleum exploration ity of a thickened boundary layer and its relevance for the thermal evo-
and development in Papua New Guinea: Proceedings of the Second Papua lution of continental convergent belts: Journal of Geophysical Research,
New Guinea Petroleum Convention, Port Moresby, May 31–June 2, 1993, v. 86, p. 6115–6132.
p. 201–206. Housh, T., and McMahon, T.P., 2000, Ancient isotopic characteristics of Neo-
Granath, J.W., Soofi, K.A., and Mercer, J.B., 1991, Applications of SAR in gene potassic magmatism in western New Guinea (Irian Jaya, Indone-
structural modeling of the Central Ranges thrust belt, Irian Jaya, Indo- sia): Lithos, v. 50, p. 217–239, doi: 10.1016/S0024-4937(99)00043-2.
nesia: Eighth thematic conference on geologic remote sensing, April 29– Hubbert, M.K., and Rubey, W.W., 1959, Role of fluid pressure in mechanics of
May 2, 1991, Denver, Colorado: Proceedings of the Thematic Conference overthrusting faulting: Geological Society of America Bulletin, v. 70,
on Remote Sensing for Exploration Geology, v. 8, p. 105–116. p. 115–166.
Gray, A.E., 1995, Petrology of the Ruffaer Metamorphic Belt, Rotanbrug map Hyndman, R.D., and Hamilton, T.S., 1993, Queen Charlotte area Ceno-
sheet (scale 1:125,000), central Irian Jaya, Indonesia [B.S. honors thesis]: zoic tectonics and volcanism and their association with relative plate
Austin, University of Texas, 93 p. motions along the northeastern Pacific margin: Journal of Geophysical
Hamilton, P.J., Johnson, R.W., Mackenzie, D.E., and O’Nions, R.K., 1983, Research, v. 98, p. 14,257–14, 277.
Pleistocene volcanic rocks from the Fly-Highlands province of western Isacks, B.L., Oliver, J., and Sykes, L.R., 1968, Seismology and the new global
New Guinea: A note on new Sr and Nd isotopic data and their petroge- tectonics: Journal of Geophysical Research, v. 73, p. 5855–5899.
netic implications: Journal of Volcanology and Geothermal Research, Johnson, R.W., Mackenzie, D.E., and Smith, I.E.M., 1971, Seismicity and late
v. 18, p. 449–459, doi: 10.1016/0377-0273(83)90020-3. Cenozoic volcanism in parts of Papua New Guinea: Tectonophysics,
Hamilton, W., 1973, Tectonics of the Indonesia region: Geological Society of v. 12, p. 15–22, doi: 10.1016/0040-1951(71)90064-3.
Malaysia Bulletin, v. 6, p. 3–10. Johnson, R.W., Mackenzie, D.E., and Smith, I.E.M., 1978, Delayed partial
Hamilton, W., 1979, Tectonics of the Indonesian region: U.S. Geological Sur- melting of subduction-modified mantle in Papua New Guinea: Tectono-
vey Professional Paper 1078, 345 p. physics, v. 46, p. 197–216, doi: 10.1016/0040-1951(78)90114-2.
 References Cited 49

Kay, R.W., and Kay, S.M., 1993, Delamination and delamination magmatism: Menzies, M.A., 1990, Petrology and geochemistry of the continental mantle:
Tectonophysics, v. 219, p. 177–189, doi: 10.1016/0040-1951(93)90295-U. An historical perspective, in Menzies, M.A., ed., Continental mantle:
Kirby, S.H., 1983, Rheology of the lithosphere: Reviews of Geophysics and Oxford, Clarendon Press, p. 31–54.
Space Physics, v. 21, p. 1458–1487. Menzies, M., Rogers, N., Tindle, A., and Hawkesworth, C.J., 1987, Metaso-
Kohlstedt, D.L., and Zimmerman, M.E., 1996, Rheology of partially molten matic and enrichment processes in lithospheric peridotites, an effect of
mantle rocks: Annual Reviews of Earth and Planetary Science Letters, asthenosphere-lithosphere interaction, in Menzies, M.A., and Hawkes-
v. 24, p. 41–62, doi: 10.1146/annurev.earth.24.1.41. worth, C.J., eds., Mantle metasomatism: London, Academic Press,
Kohlstedt, D.L., Evans, B., and Mackwell, S.J., 1995, Strength of the litho- p. 313–361.
sphere: Constraints imposed by laboratory experiments: Journal of Geo- Milsom, J., 1985, New Guinea and the western Melanesian arcs, in Nairn, et
physical Research, v. 100, p. 17,587–17,602, doi: 10.1029/95JB01460. al., eds., The ocean basins and margins, Volume 7A: The Pacific Ocean:
Kroenke, L.W., 1984, Cenozoic tectonic development of the southwest New York, Plenum Press, p. 551–605.
Pacific: United Nations Economic and Social Commission for Asia and Milsom, J., and Audley-Charles, M.G., 1986, Post-collision isostatic readjust-
the Pacific, Committee for Coordination of Joint Prospecting for Min- ment in the southern Banda arc: Geological Society [London] Special
eral Resources in South Pacific Offshore Areas (U.N. ESCAP, CCOP/ Publication 19, p. 353–364.
SOPAC), Technical Bulletin No. 6, 126 p. Milsom, J., Audley-Charles, M.G., and Barber, A.J., and Carter, D.J., 1983,
MacDonald, G.D., and Arnold, L.C., 1994, Geological and geochemical Geological-geophysical paradoxes of the eastern Indonesia collision zone:
zoning of the Grasberg igneous complex, Irian Jaya, Indonesia: Jour- American Geophysical Union Geodynamics Series, v. 11, p. 401–411.
nal of Geochemical Exploration, v. 50, p. 143–178, doi: 10.1016/0375- Milsom, J., Masson, D., Nichols, G., Sikumbang, N., Dwiyanto, B., Parson,
6742(94)90023-X. L., and Kallagher, H., 1992, The Manokwari Trough and the western
Mackenzie, D.E., 1976, Nature and origin of late Cenozoic volcanoes in west- end of the New Guinea Trench: Tectonics, v. 11, p. 145–153.
ern Papua New Guinea, in Johnson, R.W., ed., Volcanism in Australasia: Mooney, W.D., Laske, G., and Masters, T.G., 1998, Crust 5.1: A global crustal
Amsterdam, Elsevier, p. 221–238. model at 5° × 5°: Journal of Geophysical Research, v. 103, p. 727–747,
Mackenzie, D.E., and Johnson, R.W., 1984, Pleistocene volcanoes of the west- doi: 10.1029/97JB02122.
ern Papua New Guinea highlands: Morphology, geology, petrography, Nash, C.R., Artmont, G., Gillan, M.L., Lennie, D., O’Connor, G., and Parris,
and modal chemical analyses: Australian Bureau of Mineral Resources, K.R., 1993, Structure of the Irian Jaya mobile belt, Irian Jaya, Indone-
Geology and Geophysics Report 246, 271 p. sia: Tectonics, v. 12, p. 519–535.
Martodjojo, S., Sudradjat, D., Subandrio, E., and Lukman, A., 1975, The geol- Nixon, P.H., and Davies, G.R., 1987, Mantle xenolith perspectives, in Nixon,
ogy and stratigraphy along the road cut Tembagapura, Irian Jaya (unpub- P.H., ed., Mantle xenoliths: New York, John Wiley and Sons, p. 741–756.
lished): Institute of Technology, Bandung, Indonesia Report, 51 p. O’Connell, R.J., and Budiansky, B., 1977, Viscoelastic properties of fluid-
McCaffrey, R., and Abers, G.A., 1991, Orogeny in arc-continent collision: saturated cracked rocks: Journal of Geophysical Research, v. 82,
The Banda arc and western New Guinea: Geology, v. 19, p. 563–566, p. 5719–5735.
doi: 10.1130/0091-7613(1991)019<0563:OIACCT>2.3.CO;2. O’Connor, G.V., Soebari, L., and Widodo, S., 1994, Upper Miocene-Pliocene
McCaffrey, R., Molnar, P., Roecker, S.W., and Joyodiwiryo, Y.S., 1985, Micro- magmatism of the Central Range Mobile Belt, Irian Jaya, Indonesia:
earthquake seismicity and fault plane solutions related to arc-continent Fourth Asia/Pacific Mining Conference, p. 1–27.
collision in the eastern Sunda arc, Indonesia: Journal of Geophysical Oxburgh, E.R., and Turcotte, D.L., 1976, The physico-chemical behavior
Research, v. 90, p. 4511–4528. of the descending lithosphere: Tectonophysics, v. 32, p. 107–128, doi:
McConachie, B., Lanzilli, E., Kendrick, D., and Burge, C., 2000, Extensions 10.1016/0040-1951(76)90088-3.
of the Papuan basin foreland geology into eastern Irian Jaya (west Papua) Packham, G., 1996, Cenozoic SE Asia: Reconstructing its aggregation and
and the New Guinea fold belt in Papua New Guinea, in Buchanan, P.G., reorganization: Geological Society [London] Special Publication 106,
et al., eds., Papua New Guinea’s petroleum industry in the 21st century: p. 123–152.
Proceedings of the 4th Papua New Guinea Petroleum Convention, Port Page, B.M., and Brocher, T.M., 1993, Thrusting of the central California
Moresby, 29–31 May 2000, p. 219–237. margin over the edge of the Pacific plate during the transform regime:
McDowell, F.W., McMahon, T.P., Warren, P.Q., and Cloos, M., 1996, Pliocene Geology, v. 21, p. 635–638, doi: 10.1130/0091-7613(1993)021<0635:
Cu-Au-bearing igneous intrusions of the Gunung Bijih (Ertsberg) dis- TOTCCM>2.3.CO;2.
trict, Irian Jaya, Indonesia: K-Ar geochronology: Journal of Geology, Page, R.W., 1975, Geochronology of Late Tertiary and Quaternary mineral-
v. 104, p. 327–340. ized intrusive porphyries in the Star Mountains of Papua New Guinea
McKenzie, D.P., 1989, Some remarks on the movement of small melt fractions and Irian Jaya: Economic Geology and the Bulletin of the Society of
in the mantle: Earth and Planetary Science Letters, v. 95, p. 53–72, doi: Economic Geologists, v. 70, p. 928–936.
10.1016/0012-821X(89)90167-2. Parris, K., 1994, Preliminary geological data record Timika (3211) 1:250,000
McKenzie, D.P., and Bickle, M.J.J., 1988, The volume and composition of sheet area, Irian Jaya (unpublished): P.T. Freeport Indonesia, 38 p.
melt generated by extension of the lithosphere: Journal of Petrology, Parrot, J.F., and Dugas, F., 1980, The disrupted ophiolitic belt of the southwest
v. 29, p. 625–679. Pacific: Evidence of an Eocene subduction zone: Tectonophysics, v. 66,
McMahon, T.P., 1994a, Pliocene intrusions in the Ertsberg (Gunung Bijih) p. 349–372, doi: 10.1016/0040-1951(80)90249-8.
mining district, Irian Jaya, Indonesia: Petrography, geochemistry, tec- Parsons, B., and Sclater, J.G., 1977, An analysis of the variation of ocean floor
tonic setting [Ph.D. thesis]: Austin, University of Texas, 299 p. bathymetry and heat flow with age: Journal of Geophysical Research,
McMahon, T.P., 1994b, Pliocene intrusions of the Gunung Bijih (Ertsberg) v. 82, p. 803–827.
mining district, Irian Jaya, Indonesia: Petrography and mineral chemis- Pedersen, T., and Ro, H.E., 1992, Finite duration extension and decompres-
try: International Geology Review, v. 36, p. 820–849. sion melting: Earth and Planetary Science Letters, v. 113, p. 15–22, doi:
McMahon, T.P., 1994c, Pliocene intrusions of the Gunung Bijih (Ertsberg) 10.1016/0012-821X(92)90208-D.
mining district, Irian Jaya, Indonesia: Major and trace element chemis- Pettijohn, F.J., 1975, Sedimentary rocks (3rd edition): New York, Harper and
try: International Geology Review, v. 36, p. 925–946. Row, 628 p.
McMahon, T.P., 2000a, Magmatism in an arc-continent collision zone: An Philpotts, A.R., 1990, Principles of igneous and metamorphic petrology:
example from Irian Jaya (western New Guinea), Indonesia: Buletin Englewood Cliffs, New Jersey, Prentice Hall, 498 p.
Geologi, Jurusan Teknik Geologi—Institute Teknologi Bandung, v. 32, Pieters, P.E., Pigram, C.J., Trail, D.S., Dow, D.B., Ratman, N., and Suka-
no. 1, p. 1–22. mto, R., 1983, The stratigraphy of western Irian Jaya: Bulletin of the
McMahon, T.P., 2000b, Origin of syn- to post-collisional magmatism in New Geologic Research and Development Centre (Bandung, Indonesia), v. 8,
Guinea: Buletin Geologi, Jurusan Teknik Geologi—Institute Teknologi p. 14–48.
Bandung, v. 32, no. 2, p. 89–104. Pigram, C.J., and Panggabean, H., 1983, Geological data record, Waghete
McMahon, T.P., 2001, Origin of a collision-related ultrapotassic to calc-alka- (Yapekopra) 1:250,000 sheet area, Irian Jaya: Irian Jaya Geological
line magmatic suite: The latest Miocene Minjauh volcanic field, Irian Mapping Project, Geological Research and Development Centre, Indo-
Jaya, Indonesia: Buletin Geologi, Jurusan Teknik Geologi—Institute nesia, in cooperation with the Bureau of Mineral Resources, Australia,
Teknologi Bandung, v. 33, no. 2, p. 47–77. on behalf of the Department of Mines and Energy, Indonesia, and the
50 References Cited

Australian Development Assistance Bureau, 126 p. Scholz, C.H., 1990, The mechanics of earthquakes and faulting: Cambridge,
Pigram, C.J., and Panggabean, H., 1984, Rifting of northern margin of the UK, Cambridge University Press, 439 p.
Australian continent and the origin of some microcontinents in east- Schott, B., and Schmeling, H., 1998, Delamination and detachment of a litho-
ern Indonesia: Tectonophysics, v. 107, p. 331–353, doi: 10.1016/0040- spheric root: Tectonophysics, v. 296, p. 225–247, doi: 10.1016/S0040-
1951(84)90257-9. 1951(98)00154-1.
Pigram, C.J., and Symonds, P.A., 1991, A review of the timing of the major Scotese, C.R., Gahagan, L.M., and Larson, R.L., 1988, Plate tectonic recon-
tectonic events in the New Guinea orogen: Journal of Southeast Asian structions of the Cretaceous and Cenozoic ocean basins: Tectonophys-
Earth Sciences, v. 6, p. 307–318, doi: 10.1016/0743-9547(91)90076-A. ics, v. 155, p. 27–48, doi: 10.1016/0040-1951(88)90259-4.
Platt, J.P., and England, P.C., 1994, Convective removal of lithosphere beneath Searle, R.C., Bird, R.T., Rusby, R.I., and Naar, D.F., 1993, The development of
mountain belts: Thermal and mechanical consequences: American two oceanic microplates: Easter and Juan Fernandez microplates, East
Journal of Science, v. 293, p. 307–336. Pacific rise: Geological Society [London] Journal, v. 150, p. 965–976.
Plumb, K.A., 1979a, The tectonic evolution of Australia: Earth Science Shreve, R.L., and Cloos, M., 1986, Dynamics of sediment subduction, melange
Reviews, v. 14, p. 205–249, doi: 10.1016/0012-8252(79)90001-1. formation, and prism accretion: Journal of Geophysical Research, v. 91,
Plumb, K.A., 1979b, Structure and tectonic style of the Precambrian shields p. 10,229–10,245.
and platforms of northern Australia: Tectonophysics, v. 58, p. 291–325, Sibson, R.H., 1989, Earthquake faulting as a structural process: Journal of
doi: 10.1016/0040-1951(79)90314-7. Structural Geology, v. 11, p. 1–14, doi: 10.1016/0191-8141(89)90032-1.
Pollitz, F.F., 1986, Pliocene change in Pacific-plate motion: Nature, v. 320, Silver, E.A., and Smith, R.B., 1983, Comparison of terrane accretion in
p. 738–741, doi: 10.1038/320738a0. modern Southeast Asia and the Mesozoic North American Cordil-
Pollitz, F.F., 1988, Episodic North America and Pacific plate motions: Tecton- lera: Geology, v. 11, p. 198–202, doi: 10.1130/0091-7613(1983)11<198:
ics, v. 7, p. 711–726. COTAIM>2.0.CO;2.
Puntodewo, S.S.O., McCaffrey, R., Calais, E., Bock, Y., Rais, J., Subarya, C., Silver, E.A., Breen, N.A., Prasetyo, H., and Hussong, D.M., 1986, Multibeam
Poewariardi, R., Stevens, C., Genrich, J., Fauzi, Zwick, P., and Wdow- study of the Flores backarc thrust belt, Indonesia: Journal of Geophysi-
inski, S., 1994, GPS measurements of crustal deformation within the cal Research, v. 91, p. 3489–3500.
Pacific-Australia plate boundary zone in Irian Jaya, Indonesia: Tectono- Silver, E.A., Abbott, L.D., Kirchoff-Stein, K.S., Reed, D.L., Berstein-Taylor,
physics, v. 237, p. 141–153. B., and Hilyard, D., 1991, Collision propagation in Papua New Guinea
Quarles van Ufford, A.I., 1996, Stratigraphy, structural geology, and tecton- and the Solomon Sea: Tectonics, v. 10, p. 863–874.
ics of a young forearc-continent collision, western Central Range, Irian Stein, C.A., and Stein, S., 1992, A model for the global variation in oceanic
Jaya (western New Guinea), Indonesia [Ph.D. thesis]: Austin, University depth and heat flow with lithospheric age: Nature, v. 359, p. 123–129,
of Texas, 8 sheets, 420 p. doi: 10.1038/359123a0.
Quarles van Ufford, A., and Cloos, M., 2005, Cenozoic tectonics of New Stevens, C.W., McCaffrey, R., Bock, Y., Genrich, J.F., Pubellier, M., and
Guinea: American Association of Petroleum Geologists Bulletin, v. 89, Subarya, C., 2002, Evidence for block rotations and basal shear in the
p. 119–140. world’s fastest slipping continental shear zone in NW New Guinea:
Richards, J.P., Chappell, B.W., and McCulloch, M.T., 1990, Intraplate-type American Geophysical Union Geodynamics Series, v. 30, p. 87–99.
magmatism in a continent-island-arc collision zone: Porgera intru- St. John, V.P., 1970, The gravity field and structure of Papua and New Guinea:
sive complex, Papua New Guinea: Geology, v. 18, p. 958–961, doi: Journal of the Australian Petroleum Exploration Association, v. 10,
10.1130/0091-7613(1990)018<0958:ITMIAC>2.3.CO;2. p. 41–55.
Ridd, M.F., 1976, Papuan basin—On-shore, in Leslie, R.B., et al., eds., Eco- Taylor, B., 1979, Bismarck Sea: Evolution of a back-arc basin: Geology, v. 7,
nomic geology of Australia and Papua New Guinea: 3. Petroleum: Aus- p. 171–174, doi: 10.1130/0091-7613(1979)7<171:BSEOAB>2.0.CO;2.
tralian Institute of Mining and Metallurgy Monograph 7, p. 478–494. Tsenn, M.C., and Carter, N.L., 1987, Upper limits of power law creep of rocks:
Ripper, I.D., and McCue, K.F., 1983, The seismic zone of the Papuan fold Tectonophysics, v. 136, p. 1–26, doi: 10.1016/0040-1951(87)90332-5.
belt: Bureau of Mineral Resources Journal of Australian Geology and van Nort, S.D., Atwood, G.W., Collinson, T.B., Flint, D.C., and Potter, D.R.,
Geophysics, v. 8, p. 147–156. 1991, Geology and mineralization of the Grasberg copper-gold deposit:
Roden, M.F., and Murthy, V.R., 1985, Mantle metasomatism: Annual Reviews Mining Engineering, v. 43, p. 300–303.
of Earth and Planetary Sciences, v. 13, p. 269–296, doi: 10.1146/annurev. Veevers, J.J., Powell, C.McA., and Roots, S.R., 1991, Review of seafloor
ea.13.050185.001413. spreading around Australia: I. Synthesis of the patterns of spreading:
Rush, P.M., and Seegers, H.J., 1990, Ok Tedi copper-gold deposits, in Hughes, Australian Journal of Earth Sciences, v. 38, p. 373–389.
F.E., ed., Geology of the mineral deposits of Australia and Papua New Visser, W.A., and Hermes, J.J., 1962, Geological results of the exploration for
Guinea: Melbourne, Australia, Australasian Institute of Mining and oil in the Netherlands New Guinea: Koninklijk Nederlands Geologisch
Metallurgy Monograph 14, v. 2, p. 1747–1754. Mijnbouwkundig Genootschap Verhandelingen, Geologische Serie,
Rutter, E.H., and Brodie, K.H., 1991, Lithosphere rheology—A note of cau- v. 20, 256 p.
tion: Journal of Structural Geology, v. 13, p. 363–367, doi: 10.1016/0191- von Blanckenburg, F., and Davies, J.H., 1995, Slab breakoff: A model for
8141(91)90136-7. syncollisional magmatism and tectonics in the Alps: Tectonics, v. 14,
Sabins, F.F., 1983, Geologic interpretation of space shuttle radar images of p. 120–131, doi: 10.1029/94TC02051.
Indonesia: American Association of Petroleum Geologists Bulletin, Warren, P.Q., 1995, Petrology, structure, and tectonics of the Ruffaer Meta-
v. 67, p. 2076–2099. morphic Belt, west-central Irian Jaya, Indonesia [M.A. thesis]: Austin,
Sacks, P.E., and Secor, D.T., 1990, Delamination in collisional orogens: University of Texas, 339 p.
Geology, v. 18, p. 999–1002, doi: 10.1130/0091-7613(1990)018<0999: Wegener, A., 1924, The origin of continents and oceans: New York, Dutton,
DICO>2.3.CO;2. 212 p. [Translated from German third edition by Skerl, J.G.A.].
Sapiie, B., 1998, Strike-slip faulting, breccia formation and porphyry Cu-Au Weiland, R.J., 1999, Emplacement of the Irian Ophiolite and unroofi ng of
mineralization in the Gunung Bijih (Ertsberg) mining district, Irian the Ruffaer Metamorphic Belt of Irian Jaya, Indonesia [Ph.D. thesis]:
Jaya, Indonesia [Ph.D. thesis]: Austin, University of Texas, 304 p., 4 Austin, University of Texas, 526 p.
plates. Weiland, R.J., and Cloos, M., 1996, Pliocene-Pleistocene asymmetric unroof-
Sapiie, B., and Cloos, M., 2004, Strike-slip faulting in the core of the Central ing of the Irian fold belt, Irian Jaya, Indonesia: Apatite fission-track ther-
Range of west New Guinea: Ertsberg mining district, Indonesia: Geo- mochronology: Geological Society of America Bulletin, v. 108, p. 1438–
logical Society of America Bulletin, v. 116, p. 277–293, doi: 10.1130/ 1449, doi: 10.1130/0016-7606(1996)108<1438:PPAUOT>2.3.CO;2.
B25319.1. Weissel, J.K., and Anderson, R.N., 1978, Is there a Caroline plate?: Earth
Sapiie, B., Natawidjaya, D.H., and Cloos, M., 1999, Strike-slip tectonics of and Planetary Science Letters, v. 41, p. 143–158, doi: 10.1016/0012-
New Guinea: Transform motion between the Caroline and Australian 821X(78)90004-3.
plates, in Busono, I., and Alam, H., eds., Developments in Indonesian Weissel, J.K., and Watts, A.B., 1979, Tectonic evolution of the Coral Sea
tectonics and structural geology: Proceedings of Indonesian Associa- basin: Journal of Geophysical Research, v. 84, p. 4572–4582.
tion of Geologists, Volume I, 28th Annual Convention, Jakarta, Indone- Weissel, J.K., Anderson, R.N., and Geller, C.A., 1980, Deformation of the Indo-
sia, 30 November–1 December 1999, p. 1–15. Australian plate: Nature, v. 287, p. 284–291, doi: 10.1038/287284a0.
 References Cited 51

Weissel, J.K., Taylor, B., and Karner, G.D., 1982, The opening of the Wood- Economic Geology and the Bulletin of the Society of Economic Geolo-
lark basin, subduction of the Woodlark spreading system, and the evo- gists, v. 77, p. 592–616.
lution of northern Melanesia since mid-Pliocene time: Tectonophysics, Williams, P.R., Pigram, C.J., Dow, D.B., and Amirrudin, 1984, Melange pro-
v. 87, p. 253–277, doi: 10.1016/0040-1951(82)90229-3. duction and the importance of shale diapirism in accretionary terranes:
Wells, R.E., 1989, Origin of the oceanic basalt basement of the Solomon Nature, v. 309, p. 145–146.
Islands arc and its relationship to the Ontong Java Plateau—Insights Zuber, M.T., 1987, Compression of oceanic lithosphere: An analysis of intra-
from Cenozoic plate motion models: Tectonophysics, v. 165, p. 219–235, plate deformation in the central Indian basin: Journal of Geophysical
doi: 10.1016/0040-1951(89)90048-6. Research, v. 92, p. 4817–4825.
Wessel, P., and Kroenke, L.W., 2000, Ontong Java Plateau and late Neogene
changes in Pacific plate motion: Journal of Geophysical Research,
v. 105, p. 28,255–28,277, doi: 10.1029/2000JB900290.
Whalen, J.B., Britten, R.M., and McDougall, I., 1982, Geochronology and
geochemistry of the Frieda River Prospect area, Papua New Guinea: MANUSCRIPT ACCEPTED BY THE SOCIETY 18 JULY 2005

Printed in the USA

 Contents
A b.!>lracl

Chapter I. Geol~~ of Ihe Cen1rol Runge of Western New G uinea

lnlroducaion
Geology ot \\e,aem 'loc'' Gumen
Geologu: Fro.~rnc,~orl
Ccno1oi\.' Pl.uc-Tc..:tonrc Scurn~
\ 1:11\ c TeliOill\111
Magm:u"m rn lhc lhghlund'
'lei\ Gurm:.1 Trench A Rc.K:II\illing Rc:licl
Gcolot!ic Tr.tn,l"CI \\c,lcm Ccnlral Rilnt:c
Foreland Bu"n Buried Con1inental Shell
FuoahiiJ,- l..cat.lrng Edge of Ccmntl R;mge Orogenic Bell
M~lll'!ndum<~ Anuchnc Th1cl..· Sl..mncd Dcformauon
Fold~ rn the Southern Cen1rol Runge - Thin-Skinned Deformmion
Rufi,Lcr Metumurrhrc Bch Me~amurpll<>~d P:L,~lve Margin S1n11u
lrtclll OJ)hr<llitc fldt A.:crct~.-'\1 0~-cunic l .i ll)u-,phc~
Shoncnlnp E~ummc~ and Pl~tc Tectonic Rtth:~

Chnpttr 2. Mt-chunrc..., or Subdu cling Slah Breakoff by Collisional Dellimhtalion

Lilho,phenc Dclaminatron
Rup1ure of the l.ithu'phere
Fnn.:lllrc ;md Fhn'
Pl.l,IIC and Vio,cou~ nm\
The Lilho,phere Bnule I,Ja.,tic Beha' ror
The \'thcnu,phere Vi..cuu' Re,pon'e
Pc<lk S1rcngth in the Cru'l .md \Iande
Fon:c Tr.tn,mi,\ion
Plate BenJrng dunng Suhduc1ion
Buo}anq of 1hc Lllhthphcr~

Cbupter 3. Poliru.pru.tlc Rl'Con, truction of Collbional Delaminudon in 'ew Guinea

Colli'il)lllll O.:lamrnalmn - Sc;Jied Cn"' Secuon'

25 Ma: lmrucx:canu: Subduction
15 Ma~ Emer~.tcncc of l,olmcd !\lund<,
12 Mw Bcg,annrng uf the Ccnu,tl Range Orugt:n}
I() M01: Thin-Sklnnc!d Dcfclnmuion
8 Mtl: Colli'lonal Dclnmin.uion Bellin'
6 ~I a: Thick-SI.innet.l Ddormatiun unu Mul:!m:tli~m
-4 ~b: C(llii,ional O.!l.rmin.Ltion '' Cumplele
2 Mu. Po,tcolll,ion
Pre,cnl Day
Cnlli~ional Dcltunrnalion und lnduL-cd Mapmmi'm
Colh,ionul 0\:l:.rmrn:uron und Sedimem.nion
1imc-Tmn,prc''>l\C OroJ!CIIC'" Propaj!.ltmp Tear
Along·Stnl.e \iirialion' in Structuml Sl) le
Ltlho..pht:nc llc:ahng a her Otlamin;uwn
Plutc T~.-'1:tonic Ch.mgc' Rclat\.'tllo the CoJIJ,Jon f-orming New Gulllca: :-.!carl>) Wld Tr..tn'>()Ccunic

ISBN 0·8 137-2400-7

8 THE GEOLOGICAL SOCIETY
- OF AMERICA

3300 Penrose Place • P.O. Box 9140 • Boulder, CO 80301·9140, USA