Transition from Subduction to Arc-Continent Collision: Geologic and Neotectonic Evolution of Savu Island, Indonesia

Transition from subduction to arc-continent collision: Geologic and neotectonic evolution of Savu Island, Indonesia

Ron Harris Michael W. Vorkink Department of Geological Sciences, Brigham Young University, Provo, Utah 84602, USA Carolus Prasetyadi Fakultas Teknologi Mineral, Universitas Pembangunan Nasional, Yogyakarta, Indonesia Elizabeth Zobell Nova Roosmawati Department of Geological Sciences, Brigham Young University, Provo, Utah 84602, USA Marjorie Apthorpe Apthorpe Palaeontology Pty Ltd, 69 Bacchante Circle, Ocean Reef, Western Australia 6027, Australia

ABSTRACT The Savu thrust system consists of a series These results inform us that the transi- of active north-directed thrust faults found tion from subduction to collision involves Field analyses of stratigraphy, structure, onshore and offshore the north coast of Savu. (1) strain partitioning away from the sub- and tectonic geomorphology of Savu Island Thrust faults mapped onshore, penetrated by duction zone into the upper plate, (2) forearc defi ne the age and provenance of accreted the Savu #1 well and imaged in vintage seismic closure, (3) structure inherited from the Australian continental margin sequences refl ection profi les, offset the youngest deposits lower plate, (4) initiation of a crustal suture and overlying synorogenic cover, and the of Savu. The uplift history and deformation zone, and (5) uplift and exhumation of the structure, kinematics, and uplift history of pattern associated with the Savu thrust is accretionary wedge. the transition from subduction to collision investigated at a variety of temporal scales. in the eastern Sunda-Banda arc. The results Foraminifera-rich synorogenic deposits indi- INTRODUCTION highlight the dominant infl uence of lower cate low average surface uplift rates until after plate composition and structure in shap- 1.9 Ma ago, when pelagic chalk deposits were Rarely is the critical transition from subduc- ing Savu Island and initiating intraforearc raised from depths of >2500 m to the surface tion to collision preserved in advanced orogens shortening. Provenance and biostratigraphic in fewer than 1 Ma. Island emergence is well or exposed in active ones. This transition is analyses of rocks accreted to the edge of documented by uplifted coral terraces that important because it records the inception of the Sunda-Banda forearc indicate that they encrust the highest ridges to 338 m elevation. mountain building, which includes (1) under- mostly consist of Triassic to Cretaceous syn- U/Th analysis of uplifted coral yields ages of thrusting, shortening, and accretion of passive rift and postrift successions of the Australian 122 ka near sea level, indicating slow uplift continental margin units, (2) emplacement of continental margin. These rocks are similar rates of 0.2 mm/a over the past 400 ka. ophiolites, (3) initiation of synorogenic sedi- in composition and provenance to Gond- Most synorogenic deposits are stripped mentation, (4) development of continental fore- wana sequence units found throughout the from the south coast, exposing parts of the deeps, (5) creation of orogenic topography, and Banda arc to the east, such as the Triassic accretionary wedge. The deeply eroded (6) initiation of subduction polarity reversal Babulu, Jurassic Wai Luli, and Cretaceous nature of this part of the island, combined (Harris, 2004). The active Sunda-Banda arc- Nakfunu Formations of Timor. Previously with its steep fi rst-order stream gradients, continent collision zone of Australasia (Fig. 1) unrecognized units of pillow basalt are found indicates that it underwent rapid rock uplift is used as a modern analog for each of these interlayered with Jurassic beds and incorpo- and exhumation in the past 1–2 Ma. How- inaugural collisional processes (see Animation rated into mélange and mud diapirs. These ever, the south coast region is now subsiding, 1). However, what happens to the Sunda-Banda basalt occurrences have major and trace ele- as evidenced by drowned streams and south- arc-trench system, as it is underthrust by the ment compositions similar to those of Indian tilted, submerging coral terraces. Streams Australian continental margin, remains widely Ocean mid-oceanic ridge basalt and are draining north over the Savu thrust system contested due to few data (i.e., Carter et al., likely associated with Jurassic development show convex-upward patterns with gradi- 1976; Chamalaun and Grady, 1978; Hamilton, of the Scott Plateau volcanic margin. South- ents commonly associated with intermedi- 1979; Bowin et al., 1980; Harris, 1991, 2006; directed thrusting of these units via a duplex ate uplift rates. Flights of coral terraces also Charlton, 1991; Kaneko et al., 2007). thrust system detached the Middle Triassic document growth of the island to the north One of the least understood parts of the section of the underthrust Scott Plateau. above thrust-related folds. Banda orogen is the transition from subduction to collision in the region around Savu Island (Audley-Charles, 1985; Chamalaun et al., Geosphere; June 2009; v. 5; no. 3; p. 152–171; doi: 10.1130/GES00209.1; 12 fi gures; 2 tables; 1 animation.

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

1982; McCaffrey, 1988). This region occupies tion and structure of each geological unit, which latest Pliocene the Scott Plateau of the Aus- the Sunda-Banda arc transition, which extends includes biostratigraphic relations, sandstone tralian continental margin entered the trench, between the eastern Sunda arc island of Flores provenance, U/Pb age of detrital zircon, basalt which increased dramatically its accretionary and western Banda arc island of Wetar (Fig. 1). geochemistry, depth versus age analysis of infl ux and transformed it into a continental Although marine geophysical studies have been foraminifera in synorogenic units, analysis of foredeep and incipient foreland fold-and-thrust conducted in the Savu region (Bowin et al., well data and offshore seismic lines, and tec- belt (Animation 1). 1980; Silver et al., 1983; Breen et al., 1986; tonic geomorphology of uplifted coral terraces The characteristics of the accretionary wedge Karig et al., 1987; Masson et al., 1991; van der and stream profi les. The overall objective is to south of Savu differ from what is found on either Werff, 1995a, 1995b), the fundamental geologi- document the tectonic evolution of Savu, which side. To the west is the 6-km-deep Java Trench, cal and structural makeup is poorly delimited represents the initial, formative stages of arc- tectonically underlain by Jurassic oceanic due to lack of detailed fi eld observations and continent collision and suture development. crust, and to the east is the 2–3-km-deep Timor age determinations. These determinations are Trough, tectonically underlain by continental needed before structural reconstructions of this TRANSITION FROM SUBDUCTION TO crust. Where the Java Trench ends is a diffuse region are possible. Savu also provides a way COLLISION IN THE SAVU REGION zone of deformation that takes on the character- to investigate other key aspects of arc-continent istics of a foreland fold and thrust belt (Audley- collision processes, such as structural style Savu Island provides the earliest glimpse of Charles, 1986a, 2004) with a deformation front (mélange versus fold-thrust), surface uplift pat- mountain building from accretion of Australian of mud ridges (Breen et al., 1986; Masson et terns and rates, timing of deformation (colli- continental margin cover units to the Sunda- al., 1991). Seismic images of the deformation sion propagation rate), and geological hazards Banda arc, and the initiation of collision-related front south of Savu (Breen et al., 1986; Karig (active faults). strain partitioning away from the Java trench. et al., 1987) show the décollement at a strati- To address these questions we conducted a The accretionary system formed initially from graphic level of highly overpressured Jurassic detailed fi eld investigation including geologi- middle Miocene to Pliocene subduction of mudstone that is the major source for mud dia- cal mapping of the island at a scale of 1:25,000. Jurassic and Cretaceous oceanic lithosphere pirs and mélange matrix in Timor (Harris et al., During the mapping we analyzed the composi- at the Sunda-Banda trench (Fig. 2). During the 1998). The deformation front also protrudes to

5 114° 118° 122° 126°000

Sunda Shelf 3000 5000

1000 3000 3000 Banda Sea Basin Wetar Thrust E. Java 00 200 50 Flores Thrust 8°

Bali Lombok 1000 Flores Sumbawa 3000 1000 Savu Basin 0 E. Timor 3000 Sumba 300

0 1000 0 0 Lombok Basin 1 10° Savu W. Timor 1000 1000 1000 3000 200 3000 Forearc Ridge 1000 Rote 5000 inferred position 3000 of trench Java Trench

5000 12° Asia Australian Scott Plateau Continental Shelf Argo Abyssal Plain

200 km Australia Australia 118° 122° 126° Limit of Continental Crust

Figure 1. Map of major tectonic features and domains of the active Sunda-Banda arc-continent collision. Dark gray is forearc basin and light gray is uplifted forearc basement and accretionary wedge. Black with white stippled pattern is exposed forearc basement or Banda terrane (Harris, 1992). Dashed line estimates the position of underthrust Australian continental crust and the Scott Plateau prong. The Sumba Ridge is in the same position as the dashed line west of Sumba and immediately north of Savu. Dark lines with teeth are active thrust faults with teeth on hanging wall. Red triangles are active volcanoes; stars are late Quaternary volcanoes. North-northwest–south-southeast line is cross section in Figure 2. Dotted line is inferred position of the precollisional Sunda-Banda trench, which is now underlain by continental crust and is the Timor Trough.

Geosphere, June 2009 153

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

the south to 100 km beyond its position south the accretionary wedge by underthrusting and of Sumba is well defi ned (Chamalaun et al., 1982; of Sumba and Timor (Fig. 1). Associated with shortening of the Scott Plateau. Underthrusting Pirazzoli et al., 1993), there are no data available this is a change in the accretionary slope angle alone can account for at least 2–3 km of uplift, from the Savu region to test this hypothesis, which to only a few degrees compared to 5°–6° to the which is the difference in bathymetry between is a major focus of our investigation. west and east (Harris, 1991). the Scott Plateau and surrounding oceanic crust. Farther to the east the accretionary wedge in These features are interpreted as a response The structure and orientation of the Scott Pla- Timor rises to 2000–3000 m in elevation. It also to underthrusting of the Scott Plateau and south- teau cause major lateral discontinuities in the exposes collision-related metamorphic rocks ward propagation of the basal décollement into Banda orogen. The Scott Plateau formed dur- exhumed on the north coast of Timor from its weak sedimentary cover units. The change in ing Late Jurassic rifting and has an estimated depths as great as 30 km (Berry and Grady, accretionary slope taper angle is interpreted as crustal thickness of 17–18 km (Symonds et al., 1981). These differences in structural and topo- response to a dramatic reduction in basal trac- 1998). Adjacent magnetic lineations and its rift graphic elevation between Timor and Savu are tion from oceanic crust to mudstone (Harris, basins show that the Scott Plateau was stretched attributed to differences in the thickness of the 1991). The reduction in slope angle is accommo- to the northwest at the same time as it was jux- Australian continental crust that has underthrust dated by a series of top-down-to-the-south half- taposed against oceanic crust along transform these regions, and greater amounts of shortening grabens that form south of Savu (van der Werff, faults on its northeast and southwest sides. The (Harris, 1991). The Australian continental slope 1995a), indicating a phase of wedge collapse. transform-faulted northeast side underthrusts the that has underthrust Timor is as much as 3000 m Sunda-Banda forearc. This collisional geometry above the Scott Plateau (Fig. 1), and is in a Underthrusting of the Scott Plateau means that the northeast-southwest–oriented rift more advanced stage of collision. Several lines basins of the Scott Plateau are colliding with the of evidence (Audley-Charles, 1986b) support The emergence of the Sunda-Banda forearc forearc subparallel to their axes, versus perpen- the fact that the east-northeast–west-southwest on the islands of Sumba and Savu is one of the dicular to their axes, as most reconstructions of continental slope east of the Scott Plateau fi rst most vivid physiographic expressions of the arc-continent collision assume. For example, collided with the Banda arc in the East Timor transition from subduction to collision. The pre- the east-northeast–west-southwest Savu Ridge region and has propagated westward from collisional accretionary ridge (Lombok Ridge) rises where the axis of the Scott Trough, which Timor to Savu at a rate of ~110 km/Ma (Har- to the west of Sumba is at a water depth of is the largest rift basin of the Scott Plateau, ris, 1991). The V-shaped nature of the northwest ~3–4 km. East of Sumba the accretionary ridge underthrusts the Sunda-Banda forearc. Australian continental margin between Sumba rises to form an east-northeast–west-southwest The northwest-trending protrusion of the Scott and Timor predicts that the Savu region, which bathymetric high near sea level marked by the Plateau would have fi rst collided with the east- occupies the point of the V, may represent the islands of Dana, Rai Jua, and Savu. These dif- west Java Trench south of Sumba, causing the col- youngest part of the Banda arc-continent col- ferences in elevation between precollisional and lision to propagate along the northeast edge of the lision, a collisional stage that was happening transitional settings can be attributed to uplift of plateau toward Savu. Although the uplift history south of Sumba and in Timor 2–5 Ma ago.

Banda Orogen exposed in Timor. A stratig- duplex system of large isoclinal, recumbent raphy of lime and sand is scaled to thick- folds, and out of syncline thrusts. Initially the nesses observed for cover sequences of the upper thrust sheets and faults dip between Scott Plateau, while plasticine clay is used to 20°–30°, but with accretion of new thrust sheets represent rigid forearc basement. to the front of the stack, older sheets are tilted The rheological properties of these mate- progressively more arcward until they become rials are well documented (i.e., Huiqi et al., near vertical and even slightly overturned. 1992) and provide reasonable approxima- Maximum uplift occurs in the imbricate tions of contrasts in mechanical strength fan above the leading edge of the backstop. found between Australian cover rocks and This part of the accretionary wedge would forearc basement. The model is scaled be the fi rst to emerge above sea level and at a ratio of 1 cm = 1 km. A 1 cm layer of stripped by erosion, as observed in Savu. Animation 1. Sand box experiment movie plasticine clay extends on top of a 3.7-cm The backstop is also uplifted and folded into illustrating various phases of the arc- thick section of alternating white lime and a concave down geometry. A highly attenu- continent collision observed throughout red lime/sand layers (Gondwana Sequence ated mélange-like zone develops immedi- the Banda Arc. You will need Windows lower cover units of Australian continental ately beneath the plasticine forearc nappe Media Player or a multimedia player such margin). In front of the plasticine is a 1.2-cm due to very high shear strain. as Quicktime Media Player to view this thick section of alternating units (Australian Insertion of the fl at fl exible backstop fi le. If you are viewing the PDF of this paper Passive Margin Sequence of the upper cover into the layered section produces two dis- or reading it offl ine, please visit http://dx.doi of the Australian continental margin). Shear tinct structural domains: frontal accretion .org/10.1130/GES00209.S1 or the full-text strengths are 161 and 176 (Pa) for the lime above and duplexing beneath the backstop. article at http://geosphere.gsapubs.org to view and sand/lime layers, respectively. The same patterns of accretion and back- the animation. The experiment and others like it produce stop deformation are found throughout the The model is designed to test the infl uence an upper structural level of trench-ward verg- Banda arc-continent collision (Harris, 1991). of inserting a fl at-fl exible backstop (forearc ing imbricate thrust sheets of only the upper Only the early stages of the model are like the basement) into the accretionary wedge as sections of the cover in front and above the Savu phase of collision. Later stages are more observed in Savu and older parts of the backstop. The lower section of cover forms a emblematic of the Timor region.

154 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

North of Savu the accretionary wedge abuts -14 against the Sumba Ridge via the north-directed SSE Savu thrust (Reed et al., 1986). The Sumba Ridge extends southeast from Sumba to Savu and northeast from Savu to Timor. The eastern n = 5039 Mud Diapirs -10

Def. Front Def. arm of the V is uplifted and exposed throughout Timor as a series of high-level crystalline nappes

Latitude with Asian afﬁ nity sedimentary and igneous cover (e.g., Molengraaff, 1912; Brouwer, 1925). The tectonic signiﬁ cance of these rocks, and -6 their association with the Sunda-Banda forearc

200 400 600 Depth (km) Depth and Sumba, was recognized by Audley-Charles (1985) and Carter et al. (1976). We interpret the

64.3 +/- 0.2 mm/a Sumba Ridge as the forearc crustal outline of the nda terrane, green—mélange, yellow— nda terrane, green—mélange, Accretionary Wedge northern V-shaped edge of the underthrust Aus- tralian continental margin. Where the continental margin is thin, such as the Scott Plateau (Fig. 2),

sitioning system velocities are relative to the Sunda relative sitioning system velocities are the amount of uplift is much less. This interpreta- tion differs from previous models that represent Mélange

29.1 +/- 0.9 mm/a the low-density rocks underlying Sumba as a Savu Scott Plateau continental fragment trapped in the upper plate (Hamilton, 1979; Chamalaun et al., 1982). Other models have considered the possibility that Sumba represents an anomalously thick part of Lithospheric Mantle

Savu Thrust Savu forearc basement (i.e., Rutherford et al., 2001). However, none of these interpretations relate the uplifted forearc rocks exposed in Sumba to Sumba Ridge underthrusting of the continental Scott Plateau. In summary, the tectonic features associated with underthrusting of the Scott Plateau and the transition from subduction to collision in the Savu region include: (1) initiation of Forearc Forearc northward-verging thrust systems (Flores and

Savu Basin Savu Savu thrusts), (2) migration of shallow seismicity (<50 km) away from the trench into the Jurassic Oceanic Crust Jurassic forearc and backarc, (3) uplift of the forearc basin (Sumba) and ridge (Savu), (4) abrupt wid- ening of the accretionary wedge by >100 km, (5) accretionary wedge collapse (surface slope

Lithospheric Mantle angle reduction from 5° at Lombok Ridge to 2° south of Savu), (6) formation of slope basins, and (7) extensive mud diapirism. Features 1–3 we relate to increased coupling due to positive buoyancy of the underthrust continental plate.

22.2 +/- 0.3 mm/yr Features 4–7 we interpret as a consequence of adjustments made by the accretionary wedge as a result of collision initiation. Some of these W. Flores W. include basal décollement weakening as Aus- tralian continental slope units enter the trench Sunda/Banda Arc (Harris et al., 1998).

Global Positioning System Measurements and the Distribution of Strain

High-resolution global positioning system

Banda Sea Floor (GPS) measurements reveal the distribution of strain throughout the transition from subduction Flores Thrust to collision (Genrich et al., 1996; Nugroho et al., 80 90 70 40 50 60 20 30

0 100 Backarc Km 2009). These measurements show that relative to NNW -0.2 +/- 1.2 mm/yr hor. = vert. hor. Shelf (backarc) in the direction of shortening (Nugroho et al., 2009). White—continental crust, black—oceanic striped—Ba et al., 2009). of shortening (Nugroho in the direction Shelf (backarc) Figure 2. Crustal cross section through the transition from subduction to collision (see Fig. 1 for line of section). Global po subduction to collision (see Fig. 1 for the transition from section through 2. Crustal cross Figure earthquake hypocenters outlining subducting slab. wedge collapse. Blue dots are extensional basins from the Sunda Shelf the motion of the Sunda-Banda

Geosphere, June 2009 155

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

arc is increasingly coupled with the Australian GEOLOGY OF SAVU to its excellent exposure and unique characteris- plate eastward along orogenic strike (Fig. 1). tics, the Pedaro sandstone is an important strati- GPS stations within the Asian plate show little From fi eld mapping (Fig. 3) and petrographic, graphic marker unit on Savu. to no motion with respect to one another and biostratigraphic, radiometric, and geochemical Zobell (2007) investigated the provenance of the rest of Southeast Asia, except for arc islands sample analysis, we identify four lithotectonic sandstone units within the Babulu Formation immediately west of Sumba, which move from units in Savu (Figs. 3 and 4). They include pre- of Savu and Timor by petrographic studies and 14 to 17 ± 1 mm/a in the same direction as the collisional continental margin cover sequences U/Pb age analysis of detrital zircon grains. Bab- Australian plate relative to stable Asia. North of of the Scott Plateau that accreted to the edge ulu Formation sandstone is texturally immature, Sumba and Savu the motion of the Sunda-Banda of the Asian plate, which are overlain by syn- with large framework grains that are subangular arc increases to 23–25 ± 1–2 mm/a, and the inac- orogenic deposits. Scott Plateau continental to subrounded and plot as quartz wackes to lithic tive Sunda-Banda arc islands north of Timor margin cover sequences include Triassic and wackes on the classifi cation diagram of Williams move as much as 44 mm/a in the same direction Jurassic siliciclastic rocks that are similar to the et al. (1982). Sandstone provenance classifi ca- as the Australian plate (Nugroho et al., 2009). Babulu and Wai Luli Formations of the prerift tions (Dickinson et al., 1983) indicate a recycled Outer arc islands of the accretionary wedge Gondwana sequence (Harris et al., 2000), and orogen provenance. All samples have signifi cant show a similar trend. Sumba, Savu, Rote, and the postrift Australian passive margin sequence amounts of fresh twinned feldspar, mica, and Timor all move parallel to the Australian plate at (Haig et al., 2007), which includes rocks simi- lithic fragments, indicating that the sandstone increasing rates to the east that are slightly higher lar to the Nakfunu Formation of Timor (Figs. 4 beds of the Babulu Formation were deposited than motions of adjacent parts of the Sunda- and 5). Mélange with blocks from most of these close to their source region (Zobell, 2007). Banda arc. Where subduction is dominant south units and others not found in outcrop (Fig. 5G) U/Pb age determinations of detrital zir- of Java and Bali, most convergence is taken up surround a thrust-faulted package of Gondwana con grains yield major peaks at 301 Ma and at the trench. Increased coupling between the sequence units (Fig. 3). All of these lithologies 1882 Ma ago (Zobell, 2007). The youngest upper and lower plates in the oblique collision are repeated by east-northeast–west-southwest– grain analyzed is 234.6 ± 4 Ma old and the old- activates deformation throughout the arc-trench trending thrust sheets directed north-northwest est grains are Archean, with a maximum age of system where the buoyant Scott Plateau is under- and south-southeast (Fig. 3). Unconformably 2725.3 ± 37.6 Ma old. The most likely source thrust near Sumba and Savu, and the Australian overlying synorogenic units (Figs. 5A and 5H) region that satisfi es all of the criteria observed is continental slope is underthrust beneath Timor. include foraminifera-rich deep-water chalk and Argoland, which rifted from the Australian con- Timor and the adjacent volcanic islands to the uplifted coral terraces, which provide useful tinent during Jurassic breakup of Gondwana. north move as much as 47 ± 1 mm/a relative to data for reconstructing the islands uplift history. Deposition of the Babulu Formation most likely the Asian plate in the same direction as Australia, occurred in proximal nearshore to shelf break which is two-thirds of the rate of the Australian Babulu Formation (Sawyer et al., 1993) through turbidity currents plate (Nugroho et al., 2009). from prograding deltas (Bird and Cook, 1991). Baselines measured in the direction of The Late Triassic Babulu Formation in West structural transport (north-northwest–south- Timor was fi rst described by Gianni (1971), Wai Luli Formation southeast) from stable Australia across the Savu and consists mostly of interbedded sandstone portion of the accretionary wedge, western and shale. Approximately 250 m of Babulu-like The Wai Luli Formation in East Timor was Flores, and the backarc, document where most successions are repeatedly exposed in thrust fi rst described by Audley-Charles (1968), from of the collisional strain is distributed during the sheets near the southern coast in central Savu exposures in the Wai Luli River valley of East initiation of collision. Although the majority of (Figs. 5A, 5B). Most of the section consists of Timor. Similar units are also found in Savu, and motion is still taken up near the deformation thinly bedded limestone with chert nodules, consist of a basal pink calcilutite overlain by front, at least 7 ± 1 mm/a of convergence is mea- shale, siltstone, and thick sandstone beds. A a series of pillow basalt and thick successions sured between the north coast of Savu and west- marker bed in the middle part of the section of massive gray mudstone with minor silt and ern Flores, and another 22 mm/a between Flores consists of a 10-cm-thick packstone containing sandstone beds (Figs. 5C, 5D). and the Sunda Shelf (Nugroho et al., 2009). Halobia sp., which is a well-documented index The upper boundary of the pink calcilutite We infer that most of this convergence is local- fossil for the Carnian–Norian (Gianni, 1971; layer is marked by a discontinuous string of pil- ized along the Savu and Flores thrust systems Bird and Cook, 1991). low basalt outcrops as much as tens of meters in (Fig. 2), with the Savu thrust only accounting The limestone section of the Babulu Forma- diameter. The consistent stratigraphic position for approximately one-third of the motion on the tion in Savu is sandwiched between two mas- of this unit and the lateral continuity of calcilu- Flores thrust. Deformation associated with the sive to bedded sandstone units. The basal unit tite sections around it indicate only minor inter- Flores thrust system is well documented (Silver is a bedded, olive to tan, micaceous, moderately nal deformation. We interpret the pillow basalt et al., 1983). Estimates from seismic profi les to poorly sorted sandstone with iron concre- unit as a product of Late Jurassic rift-related indicate from 30 to 60 km of horizontal motion tion layers (Fig. 5A). The upper massive unit is magmatism along the northwest Australian of western Flores relative to the backarc. In east- red, coarse-grained sandstone named here the margin (see following discussion of basalt). The ern Flores little if any deformation occurs along Pedaro sandstone after Pedaro village in central Scott Plateau is a well-known igneous province this thrust system. If we divide the total amount Savu (Fig. 5B). (Crawford and von Rad, 1994). Similar basalt of shortening estimated along the Flores thrust The Pedaro Member also has concretions units are also found in the Banli #1 well on by the present rate of shortening measured ranging from 5 cm to 1 m in diameter that are the south coast of West Timor, which drilled across it (22 mm/a), it formed ca. 1.4–2.7 Ma often stained red to black. At several locations through a section similar to that found in Savu ago. This is most likely a minimum estimate, bedding-plane surfaces of the Pedaro sandstone (Sani et al., 1995). because there is evidence that rates increase as have large, 10–20-cm-wide, 5-cm-high mullion Stratigraphically overlying the basalt unit the thrust propagates. structures produced by bedding-plane slip. Due is ~200 m of varicolored mudstone. The best

156 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

exposures of this facies are found in south-central originally described in southwest Germany by Basalt Savu (Fig. 5D). Popcorn textures on weathered Klement (1960) to range into the Late Juras- surfaces indicate the abundance of expandable sic. The species brackets the age of the sample Basalt units of Savu are found in two dis- clay. Thin siltstone interbeds outline recumbent, from Oxfordian to Kimmeridgian. This age is tinctive associations. In the northern part of isoclinal folds throughout this facies that compli- younger than samples from the Wai Luli For- the island, basalt is exposed in a sheet >2 km2 cate estimates of its original thickness. mation analyzed in West Timor, which yield a in areal extent and is surrounded by mélange. One sample from the mudstone facies yields consistent Hettangian–Callovian age (Sawyer Within the mélange basalt occurs as oblate- 25 different species of spores and pollen, most et al., 1993). In this regard, it is important to shaped blocks 1–2 m in diameter. In the south- of which are long ranging in time. The only note that rifting migrated from northeast to ern part of Savu, basalt layers are found as short time range species is the dinoﬂ agellate southwest along the northwest Australian con- laterally continuous 10–30-m-thick interbeds Rigaudella aemula. This series of dinoﬂ agel- tinental margin, consistent with younger ages of the Wai Luli Formation (Fig. 5C), and as lates belongs to a genus similar to Polystepha- for rift-related features of the Scott Plateau blocks as much as 1 m in diameter near active nophorus or Systematophoraore, which was (Symonds et al., 1998). mud volcanoes.

121° 48' 121° 50' 1 2 3 Geologic Map of Savu

Line of Section 4 5 (Figure 12) 42 10° 30'

35 50 58 38 30 N 1 38 30

25 36 40 50 38 70 55 17 20 10° 35' 70 52 42 50 20 50 53 50 60 40 35 42 50 24 35 48 62 52 40 46 60 42

50 2 40 59 5 km 4 5 Structural Domains 3 Viqueque Group Wai Luli Fm. Babulu Fm.

Coral Terraces Grey mudstone Pedaro Sandstone Melange Basalt Shale, Limestone and Sandstone Batu Putih Fm. Pink Calcilutite Olive Micaceous Sandstone

Figure 3. Geologic map of south-central Savu. Unit descriptions are given in Figure 4 and text. Light green is lower Babulu colluvium. Lower hemisphere stereographs of bedding-plane attitudes are given for each of ﬁ ve structural domains. The general transport direction from fold-normal lines and asymmetry is 345° ± 15°.

Geosphere, June 2009 157

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

Whole-rock geochemical analyses of these overlap between Savu basalt and basalt of the Nakfunu Formation of Timor (Fig. 5E). Although basalt occurrences were conducted using Sunda-Banda forearc (Harris, 1992). no fossils were found to confi rm its age, the X-ray fl uorescence (Siemens SRS 303) in Basalt of Jurassic age forms a large welt that lithostratigraphic nature of the chert-rich beds order to test for differences in composition covers an area of 25,000 km2 of Scott Plateau is not documented from any part of the Austra- and tectonic affi nity. All samples are tholei- (Ramsay and Exon, 1994). The source of the lian continental margin sequences (Sawyer et al., itic ferrobasalt with very similar abundances basalt is interpreted as decompressional melt- 1993). However, some chert-rich rocks are found of major and trace elements (Table 1), and ing of the mantle during fi nal breakup and the in cover sediments of the Asian affi nity Banda plot in the mid-oceanic ridge basalt (MORB) onset of seafl oor spreading (Crawford and von terrane (Haile et al., 1979). We favor a Nakfunu fi eld near the boundary with ocean island Rad, 1994). The only samples from the Scott Formation correlation due to the clear associa- basalt (OIB) (Shervais, 1982) or near the Plateau available for study thus far are from tion of the chert-rich beds with other Australian border of the MORB and within-plate basalt dredge halls and drill core. However, Savu pro- affi nity units. The Nakfunu Formation in Timor (WPB) fi elds (Figs. 6A, 6B). There is very lit- vides several large exposures of the stratigraphy is commonly associated with accreted rise and tle compositional difference in Zr/Nb versus and morphology of these deposits, which most slope deposits of the Australian continental mar- Zr/Y between Savu basalt and Indian Ocean likely represent submarine fl ows that poured out gin best exposed in the Kolbano region of West MORB (Fig. 6C), which plots between basalt over a developing rift zone along the edge of the Timor, and are known as the Australian passive from the Argo abyssal plain and Scott Plateau Scott Plateau during Callovian–Oxfordian time. margin sequence (Haig et al., 2007). However, (Crawford and von Rad, 1994). Basalt found in most parts of Timor where the Nakfunu For- in Permian and Triassic Gondwana sequence Nakfunu Formation mation is exposed, it is overlain by thick depos- units in Timor (Berry and Jenner, 1982) over- its of Cretaceous–Miocene carbonate and shale lap other rift-related basaltic fi elds, but not the Small outcrops of thinly bedded red shale and (Sawyer et al., 1993). None of these other parts much more iron-rich Savu basalt. There is no ribbon chert are similar to the Early Cretaceous of the Australian passive margin sequence is

Savu Stratigraphy Depositional Age Period Depositional Timor Stratigraphy environment (Ma) environment

Viqueque Group 1.8 Quat. Open Coral terraces Member- Grey to blue coral covered with Marine 5.3 Open redish orange soil min. thickness 100 m. Marine Viqueque Formation

Batu Putih Formation - white to pink foram-rich chalk and 23.8 Neogene marl, underlain by thinly bedded tuff and marl (150m). Tectonic and diapiric Blocks of mixed compositions and sizes incased processes Melange in a scaly-clay matrix. Paleogene Wai Luli Formation Missing Early 65 Distal turbidite Cretaceous to Ofu formation Wai Luli Member-Dark grey to tan expandable clay in a deep marine Middle Miocene mudstone interlayered with thin 10-50 cm interbedded 143 siltstone and sandstone minumum thickness 300m. Distal calciturbidite Marginal Marine, Cret. Menu Formation Pillow basalt member - meters to tens of meter sized Deltaic Estuarine in a deep marine block of basalt with pillow texture and 155 Cont. Slope interpillow sediment draps. Grey and red mudstone Nakfunu Formation upper 5m (minimum) RO Pink Calcilutite - massive 3-5m thinly Nonmarine, Deltaic to Shelf to bedded, pink calcilutite probably bedded calcilutite. 0.6 0.9 1.2 1.5 Fluvial/ Oe Baat Formation tectonically thickned in most areas. Lower 5m massive Pink Calcilutite Floodplain Jurassic Babulu Formation 205 Open to starved Pedaro sandstone - Immature, poorly sorted Wai Luli Formation Nonmarine, Marine massive red sandstone abundant iron alteration, sandstone Fluvial/Deltaic concretions of all sizes (20m). Deltaic to shelf/ Bedded shale-poorly exposed(15m). Babulu Formation slope Thinly bedded siltstone and shale (15 m.) Aitutu Formation Open to starved Graded beds of angular and poorly sorted sandstone (22m). Nonmarine, Thinly bedded sandstone and shale (15m). Fluvial/Deltaic Marine

Limestone - fossil rich micrite layers (30 m). Turbidity currents Fluvial/Deltaic Triassic in shallow to Niof Formation deep marine Finely bedded bio-clastic siltstone and shale (20 m). Savu Data Thinly bedded purple, tan, and green non calcerous Shallow Shelf Timor Data siltstone and shale (38 m). Maubisse Formation

Olive micaceous sandstone with round, boulder sized 250 sandstone concretions (40 m).

Figure 4. Permian to present stratigraphy of Savu compared with that of Timor. Most of the Australian continental units found on Savu are correlative with the Babulu, Wai Luli, and Viqueque Formations of Timor. Palynomorph data from the Wai Luli Formation on Savu indicate an age of Kimmeridgian–Oxfordian. Depositional environments are defi ned by comparing the abundance of kerogen type and wood content. Vitrinite refl ectance (Ro) data from Timor (Harris et al., 2000) with Savu measurements show little to no thermal effects from the collision. Increasing Ro with depth most likely refl ects precollisional depth of burial on the Scott Plateau (see text).

158 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision AB Babulu Fm. Repeated Pedaro SS tilted coral terrace

C Wai Luli Fm. D pink calcilutite pillow basalt

H Stage 5e Terrace? Erosional Terrace

Geosphere, June 2009 159

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

recognized in Savu. We offer two possibilities from a depth of >7 km below the sea surface in to explain this stratigraphic anomaly. (1) Expo- the northern Savu Basin to 1.4 km below the sea TABLE 1. SAVU BASALT GEOCHEMISTRY sures of the Australian passive margin sequence surface east of the north coast of Savu (Reed et Wt% Sv-30 Sv-56 Sv-83 Sv-4 are rare throughout the Banda orogen due to the al., 1986). Likewise, the acoustic basement of SiO2 48.0 46.5 47.7 49.4 fact that only the hinterland parts of the orogen the Lombok forearc to the west, where oceanic TiO2 2.6 2.7 2.0 1.7 are exposed, and most of the high structural level lithosphere is subducting, is ~5 km below the sea Al2O3 12.8 15.5 13.1 11.2 rocks are eroded away. This may be why only surface (van Weering et al., 1989). We interpret Fe2O3 16.9 16.6 14.9 15.2 MnO 0.3 0.3 0.2 0.6 the lowest part of the Australian passive margin these differences in structural level of forearc MgO 6.0 6.0 7.3 8.3 is exposed in Savu. (2) Another possibility is that basement in part due to buoyancy differences CaO 6.1 2.7 8.7 4.0 much of the Australian passive margin sequence between underthrust oceanic and thinned conti- Na2O 3.2 4.9 3.2 2.4 was removed from the Scott Plateau by marine nental lithosphere (Fig. 2). The Scott Plateau is K2O 1.8 0.7 0.2 0.0 currents, which are very strong through this nar- 2.5 km shallower than the adjacent ocean basin, P2O5 0.2 0.2 0.2 0.2 Total 97.8 96.0 97.4 93.1 row ocean gateway (Reed et al., 1986). Seismic and where both enter into the Java trench the lines across the Scott Plateau show little to no deformation front is lifted from 6 km to 3.5 km ppm Cl 0 265 0 0 cover rocks deposited above the Gondwana depth, or by 2.5 km. However, to lift the forearc Sc 47 62 46 43 sequence (Symonds et al., 1998). Dredge hauls basement to 1.4 km, as it is found on the north V 545 556 443 329 also ﬁ nd only Miocene chalk directly overlying coast of Savu, an additional 1.1 km of uplift Cr 74 178 163 111 Triassic and Jurassic deposits (Crawford and von is needed. The structure of Savu indicates that Ni 76 103 88 77 Cu 260 305 246 157 Rad, 1994). crustal thickening by thrusting and folding may Zn 127 142 107 92 make up the difference. In other words, under- Ga 18 21 18 17 Sumba Ridge thrusting and shortening of the Scott Plateau Rb 35 12 2 0 has lifted the southern edge of the Sunda-Banda Sr 308 219 268 168 Y 43 36 37 30 One of the most important contributions of forearc basement from a precollisional depth of Zr 167 172 122 133 seismic reﬂ ection studies in the Savu region is ~4.5 km to 1.4 km below the sea surface. Nb 9 8 6 9 the imaging of the Sumba Ridge (Silver et al., The Sumba Ridge is well exposed on Timor Ba 189 32 77 3 1983), which represents the southern part of the Island and forms deeply eroded peaks up to Ce 27 18 20 16 Sunda-Banda forearc. The Sumba Ridge rises 2500 m above sea level. Beneath these forearc

Ti/100

A B Sv-30 Figure 5. Photographs of units and struc- 60 Sv-56 tural features of Savu. (A) Looking west WPB Sv-83 along the south coast of Savu at resistant IAT Sv-4 outcrops of Babulu Formation sandstone and shale. Note southward-tilted single Argo Abyssal Plain MORB (n = 10) coral terrace unconformably overlying 40 CAB Zr Y*3 exhumed Triassic units. (B) Looking west at thrust repetition of the Pedaro Member Zr/Nb of the Babulu Formation (resistant succes- Indian Ocean Permian-Triassic MORB (n = 19) sions). (C) String of pillow basalt layers Basalt Timor (n = 11) 20 interbedded with pink calcilutite of the Wai Luli Formation. (D) Interlayered pink and white mudstone and iron-rich sandstone. Scott Plateau n = 8 n = 7 Unit is isoclinally folded, with boudinage Banda Forearc of resistant sandstone beds. (E) Looking 0 west at dark red-bedded chert most likely 1 2 4 6 8 Zr/Y of the Cretaceous Kolbano Formation. (F) Looking west at a steeply dipping fore- Figure 6. Geochemistry of Savu basalt. Sv-83 and Sv-4 are samples from the largest occur- limb of a fold in bedded sandstone layer rence of basalt, which is in northern Savu and surrounded by mélange. Sv-30 and Sv-56 of the Babulu Formation. Fold vergence are from basalt interlayered with the Wai Luli Formation (Fig. 5C). (A) High ﬁ eld strength is toward the south, as in E. (G) Mélange trace element compositions for Savu basalt overlap Indian Ocean mid-oceanic ridge basalt with blocks of Eocene limestone (above (MORB) (Ludden and Dionne, 1992) and plot between Argo abyssal plain and Scott Pla- notebook), basalt, sandstone, and green- teau compositions (Crawford and von Rad, 1994), which all overlap with Permian– Triassic schist. (H) Panorama of uplifted coral ter- basalt units found in Timor (Berry and Jenner, 1982). There is no similarity with forearc races near Aimau. U/Th age is from lower basalt from the Savu Basin (Harris, 1992, 2006). (B) Savu basalt plots near the boundary of terrace. Bedded units are forereef deposits MORB and within-plate basalts (WPB) on the tectonic discriminate diagram of Pearce and capped by coral in growth position. Cann (1973). IAT—island arc tholeiite; CAB—calc-alkaline basalt.

160 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

nappes is a duplex thrust system of mostly the Sumba Ridge, since there are no Cretaceous gies, including some that do not crop out on the Permian and Triassic Gondwana sequence units clastic units on the Scott Plateau. island (Fig. 5G). The best exposure we found is (Harris et al., 1998). We infer a similar struc- Eocene carbonate is identifi ed in Rai Jua (the in a stream cut bank 5 km west of Seba, where tural style for the uplift of the Sumba Ridge in island immediately west of Savu), and as blocks blocks of indurated sandstone (Triassic?), the Savu region. We project the southern edge in mélange encountered by the Savu #1 well. crinoid-rich limestone (Permian?), and sparse of the Sumba Ridge beneath the highest topog- The Rai Jua occurrences have Camerina and blocks of metamorphic and igneous rocks raphy on Savu because mélange exposed in this Discocyclina, which are also found in Eocene (Banda terrane?) fl oat in a matrix of expandable region includes blocks of Sumba Ridge litholo- limestone on Sumba (Grady et al., 1983). In scaly clay. Blocks are mostly subangular and as gies (Fig. 5G). Sumba, Eocene limestone is interlayered with large as 4 m in diameter. Matrix material has andesitic and dacitic volcanic fl ows and tuffs. characteristics similar to those of the mudstone Sunda-Banda Forearc Units The age and composition of these rocks are facies of the Wai Luli Formation. associated with the Great Indonesian Arc that In the Savu #1 well, mélange is encountered The Sumba Ridge is correlated with out- precedes the Miocene to present Sunda-Banda twice below a repeated section of late Miocene– crops in Sumba (van der Werff et al., 1995b) arc-trench system (Rutherford et al., 2001). In Pliocene deep-water chalk: it is found at a depth and Timor (Harris, 2006) that expose forearc Animation 1, the Banda terrane is represented of 290–725 m, and again at 900–1224 m. In cover and basement of the Asian affi nity Sunda- by black plasticine. these intervals the well drilled through sev- Banda arc. These units include deformed eral different types of blocks, which include Cretaceous —Miocene arc-derived sedimentary Mélange (1) coarsely crystalline limestone with large, and volcanic rocks, which overlie greenschist late-early Miocene benthonic foraminifera to amphibolite facies metamorphic rocks with Mélange is found in a discontinuous belt (340–346 m); (2) carbonate with a rich assem- Eocene cooling ages (Harris, 2006; Standley on the northern edge of exposures of accreted blage of upper Eocene foraminifera (440– and Harris, 2009). The Savu #1 well (Fig. 7) Australian continental margin units (Figs. 3 483 m); (3) Oligocene fauna associated with encountered blocks with similar compositions and 4) and in the Savu #1 well on the north an interval of mixed pebbles that include slate, to units of the Sumba Ridge. It stopped drill- coast. Due to the mud-dominated matrix of the chlorite schist, and quartz veins (~944 m); and ing at a depth of 1227 m in Cretaceous clastic mélange, it is poorly exposed as rolling slopes (4) Cretaceous clastic units at the bottom of the units, which we infer may have been the top of with lag deposits of blocks of mixed litholo- well (1227 m). All of these rock types are found

122° 52' 122° 00' Northern Savu 5 km

Figure 9

Thrust 7 Thrust 6 Thrust 5

Savu # 1 well 10° 25'

Thrust 5 71 Thrust 6 Thrust 4

101 Aimau survey Thrust 5 N 68 Aimau coral 118 terrace106 survey st 4 Thru 187 135 65 78

Active thrust (deforms surface) Terraces Fault 3 Blind thrust deforms units < 0.3 Ma 218 Coral Deposits 10° 30' 42 Thrust ramp anticline Batu Putih Fm. 110 Inter-thrust syncline Melange 20 Thrust 2 152 Ship track lines Inferred Contact

338 Elevation (m) Foram Samples 338 Thrust 1

Figure 7. Structural map of northern Savu. Rectangle marks location of uplifted coral terrace survey (Fig. 10) near Aimau and star the location of the Savu #1 well. Gray lines are ship tracks of seismic reﬂ ection lines including Figure 9 on far left.

Geosphere, June 2009 161

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

in the Banda terrane and have no Australian younger ages of the upper Batu Putih Formation this sedimentary section is the offshore equiva- continental margin equivalent (Harris, 2006). refl ect a much later uplift of these islands than lent of the Batu Putih Formation. Sediment In southeastern Savu, mélange is locally indicated from the synorogenic sedimentary thickness estimates of as much as 1000 m from found associated with active hot water mud vol- record of Timor. seismic profi les are calculated using P-wave canoes surrounded with lag deposits of basalt, The absence of turbiditic synorogenic units velocities of 1900–2300 m/s, which are reported crinoidal limestone, micrite, and sandstone, exposed on the islands west of Timor indicates for the uppermost sequence in the Savu Basin which have Australian continental margin asso- the youth of islands where the initial stages of (van der Werff, 1995a). ciations. The highly altered basalt blocks have accretionary wedge emergence are observed. Applying age data for the Batu Putih Forma- Ti, Sr, Zr, and Y contents similar to those of The part of the accretionary wedge exposed in tion that we report here also provides a way to basalt interlayered with the Wai Luli Formation; Savu and Rote did not have a source for turbid- estimate sedimentation rates, which is neces- this also supports a Wai Luli Formation source itic sand. However, now that these islands are sary for defi ning the age and duration of defor- for much of the mélange. Mud volcanoes are eroding it is likely that such deposits are accu- mational events affecting the Batu Putih For- also documented in seismic profi les north and mulating in slope basins offshore. mation. Strong bottom currents throughout the west of Savu, near the toe of the accretionary The Batu Putih Formation in Savu consists region have stripped sediments down to con- wedge (Breen et al., 1986; Karig et al., 1987) mostly of foraminifera-rich, homogeneous solidated bedrock in places (Reed et al., 1986). and throughout Timor (Barber et al., 1987; Har- white chalk. The chalk is interlayered locally However, successions as much as 1000 m thick ris et al., 1998). with thinly bedded tuffaceous siltstone and accumulated in 5.6 Ma, which is a sedimen- The structural position of mélange near the marl. Pleistocene coral and fl uvial gravels tation rate of 178 m/Ma. The 240-m-thick top of the Australian continental margin thrust unconformably overlie the Batu Putih Forma- section drilled and dated in the Savu #1 well stack and the degree of block rounding and tion (Fig. 5H). The thickest sections of the Batu yields sedimentation rates of 171 m/Ma. Two mixing are similar to the Bobonaro mélange Putih Formation exposed on Savu are >100 m unconformities identifi ed in the well indi- of the Timor region. The Bobonaro mélange is thick, and are found mostly in the eastern part cate that this is a minimum rate. These rates interpreted as forming along the suture between of the island (Figs. 3 and 7). The Savu #1 well are slightly higher than modern sedimenta- Asian affi nity forearc thrust sheets above and a encountered 240 m of chalk and calcilutite that tion rates measured in the Timor Trough of thrust duplex of Australian continental margin overlies mélange. Ages from foraminifera in 140 m/Ma (Ganssen et al., 1989). units below (Harris et al., 1998). A similar situ- units drilled by the well indicate a disconformity The maximum age of the Batu Putih Forma- ation may exist in northern Savu. at 63–70 m depth with N22/N23 (Pleistocene- tion implies that the part of the accretionary Holocene) calcilutite intermixed with coral wedge these sediments overlie formed by at Viqueque Group debris directly overlying N19 (4.2–5.2 Ma ago) least 5.6 Ma ago. The contact between accreted age calcilutite (Fig. 5H). The base of the Batu units and the overlying Batu Putih Formation The Viqueque Group of Timor (Audley- Putih in the Savu #1 well is lower N18/N19 is exposed in many places on the islands of Charles, 1986a) represents a synorogenic sedi- (5.6–5.0 Ma ago). Sumba, Savu, Rote, and Timor. In the forearc mentary sequence that overlies rocks of the Detailed analysis of foraminifera of the Batu basin region of Sumba the basal age of chalk Sunda-Banda arc and Banda orogen. The basal Putih Formation in outcrop (Roosmawati and deposits is around the same as the age of vol- Batu Putih Formation (Audley-Charles, 1968) is Harris, 2009) indicates that the oldest units have canic arc itself (ca. 10 Ma old; Abbot and Cha- a pelagic chalk with some tuffaceous and calci- a well-preserved assemblage of Globorotalia malaun, 1981). The Batu Putih Formation in lutite horizons that were deposited in the forearc tumida tumida and Globorotalia merotumida, Savu and Timor has the same basal age of N18 basin, and over the southward-expanding accre- characteristic of Zone N18/N19 (5.6–5.0 Ma (5.6 Ma old), which we interpret as the time that tionary wedge. Collision propagation from ago). Although the sections studied on Savu this part of the accretionary wedge formed. Timor to the southwest (Harris, 1991) predicts have similar ages to those sampled by the well, Quaternary coral unconformably overlies that ages of the upper Batu Putih Formation the faunal assemblages are slightly different. most of the units of Savu (Fig. 5A). Coral ter- will vary throughout the region. In the forearc Outcrop ages near the top of the section are race deposits are grayish-blue to white, and are basin deposits exposed on Sumba (Fortuin et mostly N21 (3.35–1.92 Ma ago), which is deter- commonly covered with a thin red soil. Flights al., 1997), pelagic chalk (Waikabubak Forma- mined by the fi rst appearance of Globorotalia of coral terraces are found in northern Savu as tion) yields foraminifera ages as old as middle tosaensis and the last appearance of Sphaeroi- high as 338 m in elevation. The most complete Miocene (N15) (all foraminifera zones and dinellopsis seminulina seminulina and Globoro- section of terraces is well exposed on the north ages are according to Berggren et al., 1995; and talia multicamerata. The highest samples in the coast of Savu near Aimau (Figs. 5H and 8), geologic time scale [GTS2004] of Gradstein et section have Globorotalia truncatulinoidses, which we surveyed and analyzed (see discus- al., 2005) and as young as latest Pliocene (N21, Globorotalia crassaformis, and Neogloboqud- sion of uplift history for results). At the base of 3.35–1.92 Ma ago). In the accretionary wedge rina dutertei that indicate an age of upper N21 some terraces is a conglomerate with calcilutite, region of Timor (Audley-Charles, 1986a) and and lower N22 (3.0 to younger than 1.92 Ma sandstone, limestone, and micrite clasts. Savu (Roosmawati and Harris, 2009), the base ago). N22/N23 ages were found at a depth of of the Batu Putih Formation is latest Miocene 55–63 m underlying and intermingled with STRUCTURE AND NEOTECTONICS (N18, 5.6–5.2 Ma ago); however, the age of the Quaternary coral in the Savu #1 well. top of the unit youngs to the west (Roosmawati Vintage seismic lines shot in the mid 1970s Analyses of structural features in Savu pro- and Harris, 2009). In Timor the top is overlain along the north coast of Savu (donated by Inter- vide information for reconstructing its struc- by distal turbidites of N20, 4.20–3.35 Ma ago. national Oil) show continuous high-amplitude tural evolution and geometry. Combining these In Rote and Savu there are no overlying turbi- refl ections of cover sedimentary units over data with our studies of the recent history and dites, but pelagic chalk continues into the Pleis- acoustic basement (Fig. 8). Observations from patterns of uplift and subsidence of the island tocene (Roosmawati and Harris, 2009). These the Savu #1 well and our mapping indicate that provide additional information. Savu and Rai

162 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

Jua form the top of an east-northeast–west- al., 2002). As the continental margin encroaches the orientation of fold hinge lines and thrust southwest ridge that parallels the structural upon the Sunda-Banda arc, some of these rift faults (Fig. 3). Based on these observations we grain of the northwest Australian continental basins are inverted to form anticlines with fold claim that the pattern of deformation in Savu margin. Extensional structures of the passive axes oriented parallel to east-northeast–west- is strongly inﬂ uenced by structural inheritance margin initiated in the late Paleozoic and con- southwest rift basins (Doré and Stewart, 2002). rather than the convergence direction, or the tinued throughout the Mesozoic until seaﬂ oor In many ways, the Savu–Rai Jua ridge can irregular orientation of the thrust front. In more spreading moved the rift access away from the also be considered an inverted rift basin of the advanced stages of collision south of Timor, the margin ca. 155 Ma ago (Symonds et al., 1998). Scott Plateau; bedding attitude measurements thrust front also rotates to parallelism with the Several rift basins formed that are mostly ori- throughout Savu show a dominantly east- structural grain of the northwest Australian con- ented east-northeast–west-southwest (Keep et northeast–west-southwest strike, which is also tinental margin (Fig. 1).

SSW NNE

0 2 km A 0 Seconds (TWTT) 1

0 2 km 0 B T- 4 T- 5 T- 6 T- 7 Seconds (TWTT) 1

Figure 8. Seismic refl ection profi le shot off northwest coast of Savu in 1975 by International Oil. (A) Unmigrated record section. (B) Interpreted section showing deformed Batu Putih Formation (blue) and inferred thrust faults (T). Sedimentation rate is determined based on velocities of 2100 m/s. Folds show onlapping (growth) from ca. 2 Ma ago to present. However, the top 150 m of the profi le is obscured by bubble pulse bottom refl ectors (yellow). TWTT—two- way traveltime.

Geosphere, June 2009 163

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

Cross Section axis of the Sumba Ridge (Reed et al., 1986). north and form two different dip domains that In our cross section we have inferred a slightly represent a gently dipping top limb (20°–30°) The composite north-northwest-south- deeper top adjacent to the Savu thrust system. and more steeply dipping (40°–50°) back limb southeast cross section (Fig. 9) of central Savu Movement of the Savu thrust system is young of folds associated with south-directed thrust- combines stratigraphic and structural relations enough to deform even the youngest coral ter- ing. Some steeply dipping (Fig. 5F) and even mapped out on the island, encountered in the races (younger than 120 ka). Where erosion has overturned (Fig. 5G) forelimbs are found, but Savu #1 well, and imaged in seismic refl ection stripped most of the coral canopy we fi nd north- mostly as parasitic folds. No hanging-wall or profi les offshore. Starting in the offshore north verging fold asymmetry in the northernmost footwall cutoffs, which are needed to estimate of Savu, seismic data document a system of exposure of Pedaro sandstone. We interpret this amounts of slip, are identifi ed except on para- north-directed thrust faults of Pleistocene age thrust sheet as the highest structural level of the sitic folds. However, the lack of any older parts (Figs. 7 and 8). The Savu #1 well reveals that Savu thrust system. In the backlimb syncline of the Gondwana sequence between the thrust these faults place mélange over the Batu Putih of this thrust are remnants of a coral terrace at sheets indicates that they detach near the Middle Formation. Drilling near the south coast of 293 m elevation. Triassic and form a duplex beneath younger Timor documented >2 km of mélange beneath The central and southern parts of Savu pro- units and mélange (Animation 1). These data the Batu Putih Formation (Audley-Charles, vide excellent laterally continuous exposure of are also consistent with the relative thinness of 1986a). Block compositions from the well and Australian continental margin units accreted to thrust sheets and informs us that there are likely in mélange outcrops in Savu and Rai Jua reveal the edge of the Sunda forearc (Figs. 5A–5D). multiple detachments and zones of duplexing a signifi cant component of Sumba Ridge forearc Our structural studies of these units identifi ed within the Gondwana sequence, as found in the material, which indicates that the southern edge three thrust-repeated sections of mostly Upper Timor region (Harris, 1991). of the Sunda-Banda forearc most likely extends Triassic and Jurassic Australian continental The cross section indicates that two major beneath these outcrops. It is possible the Savu margin units that are ~2 km thick. We infer geological domains divide Savu: the north is #1 well encountered the top of the forearc at its that most lower Cretaceous and younger units dominated by mélange, and synorogenic depos- total depth of 1227 m. Seismic images off the are part of the roof thrust above the duplex sys- its overlying the edge of the Sunda-Banda north coast of Savu indicate that forearc base- tem of Middle Triassic to Jurassic rocks. Most forearc, and the south is dominated by accreted ment is within 1400 m of the sea surface at the rocks incorporated into the duplex system dip fragments of the Australian continental margin.

NNW Savu Island SSE Cretaceous Nakfunu Fm.

Savu Thrust Highest System Coral Fm. Terrace luli Batu Wai uli Fm. 0.5 Putih Fm. Coral Wai l Tilted Coral 0.5 Terraces Wai luli Fm. Terrace Sea level Sea level Pedaro SS 0.5 . 0.5 Pedaro SS u Fm. 1.0 1.0 Pedaro SS

1.5 Lower Babulu Fm 1.5 Lower Babul

Wai luli Fm. Km 2.0 Mélange 2.0 Km 2.5 2.5 Forearc Basement 3.0 3.0

3.5 3.5

4.0 4.0

4.5 4.5

5.0 5.0

5.5 5.5 Lower Gondwana Seq. Duplex Middle Triassic Décollement 6.0 6.0 horizontal = vertical

Figure 9. Composite north-northwest–south-southeast section across Savu (see Fig. 3 for line of section). The structure of northern section is according to seismic proﬁ les off the north coast and from drilling data. Folds and thrust faults are determined from ﬁ eld mapping (Fig. 3). Dip measurements are given by black dots with line pointing in direction of dip. Black bodies in Wai Luli Formation are basalt. Detachment depth is estimated from the projected depth of the Sumba Ridge and stratigraphic thickness of units incorporated into thrust sheets. Cretaceous units are interpreted as a roof thrust for the Upper Triassic–Jurassic duplex. Assuming no subduction erosion, the southern edge of the forearc was the precollisional Sunda-Banda trench, which is now abandoned due to accretion of Australian continental margin in front and beneath it. SS—sandstone; Seq.—sequence.

164 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

The area between the two is a nascent collisional with inferred ages of 780–850 ka, and other Batu consistent with a shortening rate of ~3–4 mm/a, suture zone. Along orogenic strike in Timor, the Putih Formation units with younger foramin- which is six times the rate of uplift. suture zone is exposed at deeper structural levels ifera. Fault 3 is mapped on the basis of abruptly The eastern part of the Savu Basin disappears (Harris et al., 1998). Several fragments of the truncated and tilted coral terraces, but its sense where it is buried by north-directed thrust faults Sunda-Banda forearc (Banda terrane) structur- of shear is unclear. Thrust 4 projects onshore of the Timor orogenic wedge. Off of the west ally overlie a duplex system of mostly Permian from seismic refl ection profi les on either side coast of West Timor a sequence of north-directed and Triassic Gondwana sequence units accreted of Savu (Fig. 8) and may also account for the thrusts is similar to thrusts west of Savu, some of from the edge of the Australian continental mar- abrupt scarp that cuts across terraces at ~100 m which break to the seafl oor surface (Karig et al., gin. Thick occurrences of mélange mark the elevation near the Aimau traverse (Fig. 7). 1987; van der Werff, 1995b). Farther to the east mostly low-angle suture zone between arc and These structural relations also indicate Pleisto- the thrust system causes the nearly complete continental affi nity units (Harris et al., 1998). cene activity along the Savu thrust. closure of the forearc basin, and brings accreted This same suture zone in also recognized in The International Oil seismic grid shot off- Australian continental margin deposits to within Savu, where the rear of the accretionary wedge shore (Fig. 7) shows thrust faults breaking to the 20 km of the Banda arc island of Atauro (Har- impinges upon and overrides the Sumba Ridge surface and creating seafl oor topography, which ris, 1991; Snyder et al., 1996). All along the and Savu Basin along the Savu thrust. indicates Holocene displacement along the Savu north coast of East Timor are extensive fl ights thrust. Movement along thrust 5, as predicted by of Pleistocene coral terraces with estimated Savu Thrust System GPS results (Nugroho et al., 2009), produces as uplift rates of ~0.5 mm/a (Chappell and Veeh, much as 250 m of vertical offset on the seafl oor 1978). Many of the terraces are warped by short The Savu thrust is part of a system of at least (Fig. 8). This structural relief increases eastward (10–15 km) wavelength folds indicating shallow four south-dipping thrust faults imaged by seis- to 550 m west of Savu #1, then decreases far- crustal deformation processes (Cox et al., 2006), mic studies between Savu and Sumba (Silver et ther east. The topography of the north coast of similar to those observed along the Savu thrust. al., 1983; Reed et al., 1986; Karig et al., 1987; Savu mimics this pattern, suggesting a direct link van der Werff, 1995b). In the context of these between topography and recent movement along UPLIFT HISTORY other thrust faults, the Savu thrust is the north- the Savu thrust. The Savu #1 well may have pen- ernmost of what may be interpreted as a for- etrated thrust 5, which could intersect the thrust The emergence of Savu from a precollisional ward-propagating thrust sequence from south fault found in the well if it dips ~30° south. forearc ridge at depths of 3000–4000 m below to north. In front of the thrust faults are diapirs Combined movement along blind thrusts 6 sea level to an island as much as 340 m above rising from acoustic forearc basement that are as and 7 deform well-imaged sections of the off- sea level is recorded by thermal history indica- much as 2 km across (van Weering et al., 1989). shore equivalent of the Batu Putih Formation all tors in uplifted Triassic units, depth versus age Seismic images of the faults show hundreds of of the way to the seafl oor. The transition from analysis of benthic and planktonic foraminifera meters of undeformed sedimentary successions predeformational to syndeformational units in synorogenic deposits (Roosmawati and Har- overlying fault tips consistent with no slip at (constant thickness versus onlapping) is diffi cult ris, 2009), uplifted coral terraces, and patterns least during the late Quaternary (Silver et al., to resolve with both folds. However, the extent of longitudinal stream profi les. These different 1983). Although there is shallow seismic activ- of obvious onlapping units indicates that these records yield uplift rate and deformation pattern ity in the region, none has been directly attrib- folds may have been active for at least 1–2 Ma. information from different spatial and temporal uted to the Savu thrust system (McCaffrey et al., However, most structures show little to no indi- scales. For example, analyses of thermal his- 1985). The age of initiation of the Savu thrust cation of growth, indicating they are very young tory indicators and foraminifera record uplift of system is unknown, and whether the fault is still and activated <1 Ma ago. the forearc ridge over a time span of 1–5 Ma, active is debated (McCaffrey et al., 1985; Reed Using the line length of various refl ections whereas coral terraces record uplift from et al., 1986). The easternmost strand of the Savu that can be traced over folds, we calculate a approximately sea level to 300 m during only thrust is projected through northern Savu and minimum rate of shortening over the past 1 Ma the past 50–800 ka. Records over such a broad offshore to the east, where it dies out before it of 3.3 mm/a. This rate is a minimum estimate range of time in an active orogen afford the rare reaches the islands of Rote and Timor (Crostella because only the deformation associated with opportunity to connect patterns of deformation and Powell, 1976; Reed et al., 1986). folding is represented, not the total slip along found in bedrock units with their Pliocene– To better understand the Savu thrust sys- faults. GPS baselines measured across the Pleistocene synorogenic sedimentary cover. tem we investigated the onshore and offshore widest part of the Savu Sea in the direction of regions of the north coast where it is projected thrust transport show there is 6–8 ± 0.9 mm/a of Thermal History Indicators into Savu. The Savu #1 well and structural fi eld convergence between Savu and western Flores mapping of northern Savu demonstrates long- (Nugroho et al., 2009). The Savu thrust system Vitrinite refl ectance measurements of car- term north-directed thrusting. The tight seismic is the only structure identifi ed along the baseline. bonaceous shale and mudstone indicate sub- grid (Fig. 7) shows that most structural features On either side of the Savu Sea there is evi- mature maximum paleotemperatures for Trias- are generally oriented east-northeast–west- dence of greater amounts of closure of the sic and Jurassic units accreted from the Scott southwest, subparallel to the structural grain Sunda-Banda forearc. For example, the Savu Plateau (Fig. 4). This contrasts with slightly of the Scott Plateau and Australian continental Sea ends to the west due to deformation in the higher vitrinite refl ectance values found in age- margin to the east. Sumba region, where active thrust faults are equivalent units in Timor (Harris et al., 2000). Geomorphic patterns provide evidence of inferred based on seismic refl ection profi les The differences can be explained by the little to Pleistocene deformation along the northern part (van der Werff et al., 1995a) and extensive no postrift passive margin sediment overlying of Savu. Mélange is thrust northward over Batu fl ights of uplifted coral terraces (Pirazzoli et al., Gondwana sequence units of the Scott Plateau Putih Formation units along thrust 1. Thrust 2 1993). Age analysis of the coral indicates uplift (Symonds et al., 1998) versus the northwest places Batu Putih Formation over coral deposits rates of at least 0.5 mm/a during the past 1 Ma, Australian continental margin accreted to

Geosphere, June 2009 165

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

Timor. The Cretaceous–Pliocene sections on the the rear of the Sunda-Banda forearc accretion- relative to the land surface that we use here to continental slope that collided with Timor are at ary wedge was abruptly uplifted from a depth track rates and patterns of deformation during least 1.0–1.5 km thick. of ~2500–4000 m to near the surface before the Pleistocene. We surveyed uplifted coral Low vitrinite values also rule out heating N22/N23. Surface uplift rates of 2–4 mm/a are terraces on both the north and south coasts of associated with tectonic burial during accretion. required to get the accretionary wedge to the Savu (Fig. 11B), and analyzed U/Th ages of A large scatter of apatite fi ssion track ages and surface over such a short time interval (Fig. 11). pure aragonite samples from these sections. The uniformly short track lengths found in Trias- This abrupt increase in uplift rate followed samples were fi rst analyzed by X-ray diffraction sic units accreted to Timor (Harris et al., 2000) by slowing indicates to us that the accretion- for degree of alteration from aragonite to calcite. demonstrate that most of these units were in the ary wedge encountered an abrupt bathymetric Five samples with ≥99% aragonite were found. partial annealing zone (>110 °C) before they feature such as the transform faulted northeast However, four of these are from the centers of were accreted, and were uplifted to the surface edge of the Scott Plateau, which has a relief of Tridacna shells, which yielded unreliable ages before new tracks formed (Harris et al., 2000). 2–3 km. Emergence of the island above sea level due to highly variable U values. is well documented by fl ights of uplifted Pleis- The only acceptable age is from coral on Depth Versus Age Analysis of Foraminifera tocene coral terraces that overlie the Batu Putih the north coast collected from an erosional ter- Formation. race 4 ± 1 m above sea level (high tide mark – Analysis of thousands of foraminifera from tidal range/2). It has an age of 121,593 ± 736 a Batu Putih Formation chalk samples (see sam- Uplifted Coral Terraces (Table 2). This age corresponds to the beginning ple locations; Fig. 7) provide a way to defi ne of the last major interglacial sea-level highstand long-term depth versus age histories of synoro- Coral terraces mostly surround the island (MIS 5e), which at that time was ~0–5 m above genic sedimentary units deposited over the Savu of Savu, and cap its highest ridges to 338 m where it is now (Lambeck and Chappell, 2001). accretionary wedge (Roosmawati and Har- in elevation (Figs. 7 and 11A). These terraces Although the coral most likely grew 3–4 m ris, 2009). All samples have both benthic and provide a nearly continuous record of sea level below sea level, at face value, this result indicates planktonic foraminifera, which provide a rough estimate of depth and age, respectively, of the Batu Putih Formation throughout Savu (Fig. 9). However, most samples consist of <2% benthic Pliocene Pleistocene foraminifera, which is characteristic of abyssal Zanclean Mioc. Piacenz. Gelas. depths (Van Marle et al., 1987). N18 N 19 N 20 N 21 N 22 N 23 Nearly all samples from Savu of N18 to N21 5 4 3 2 1 0 Age (Ma) age (5.5–1.9 Ma ago) yield deep-marine benthic foraminifera (Fig. 10). Most of these show evi- +500 dence of dissolution and low diversity with only 0.3 - 0.8 mm/a the most dissolution-resistant species remaining Sea Level 0 (Roosmawati and Harris, 2009). Dissolution in the lysocline is found at depths of ~4000 m in 500 Indonesia (Van Marle et al., 1987). Some late N21 and early N22 samples near 1000 the top of the chalk show greater species diversity with little to no evidence of dissolution. 2–3 mm/a 1500 We interpret this shallowing-upward trend as a consequence of slow uplift of the accretionary Depth (m) wedge between 5.5 and 1.0 Ma ago. Although 2000 the resolution is poor, these deposits were not uplifted more than 1500 m (0.5 mm/a) during 2500 this time interval, on the basis of the occurrences of Globorotalia truncatulinoides, of N22 (1.92 0.5 mm/a 3000 to 0.7 Ma), along with characteristic deep-water species, such as Pullenia bulloides, Oridorsalis Lysocline 3500 umbonatus, and Palnulina wuellerstorfi . These deep-water benthic foraminifera indicate that 4000 deep conditions similar to the crest of the precollisional Lombok accretionary wedge to the 4500 west (3–4 km) persisted during N22. An unconformity overlies the deep-water sec- Figure 10. Surface uplift rates from depth versus age analysis of fora- tion of the Batu Putih Formation where much minifers. Each sample is represented by its upper depth limit (fi lled of zones N21 and N22 are missing. Above the rectangle) based on benthonic foraminifers, and age range (horizon- unconformity are late N22/N23 (younger than tal line) based on planktonic foraminifers. The combination of these 1 Ma) neritic species (0–200 m depth) intermin- data provides minimum surface uplift rates. Parts of the curve above gled with coral (Roosmawati and Harris, 2009). sea level are fi xed by U-Series age analysis of uplifted coral terraces. These relations indicate that after 1.92 Ma ago, Stars are data from the Savu #1 well.

166 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

that there is little uplift on the north coast rela- Holocene terraces associated with the pres- tive to sea level since at least 121,593 a ago. ent highstand are as much as 1150 m in width However, stratigraphic analysis of the erosional off of Point Bali (Fig. 7), but <100 m wide on terrace demonstrates that it is carved into bed- the north coast. A similar pattern is found in the

ded forereef sediments shed from a higher reef more than 20 uplifted terraces above the present † complex (Figs. 5H and 11C). Above the forereef coral reef. Matching coral ages with elevations Initial

sediment is a coral rampart in front of a lagoon yields a best-ﬁ t minimum uplift rate solution of U 234 δ (corrected) 4–6 m deep. The architecture of the uplifted ter- 0.2–0.3 mm/a (Fig. 11B). At this low rate, some Th atomic

race is consistent with a single reef complex at lower highstands (dashed) are in the shadow of 232 least 20 m thick that was shedding coral from the later higher ones, and may only be represented Th /

barrier ridge down a forereef slope (Fig. 11C). by small notches, such as those found in stages 230 Sea level at the time the reef was formed (stage 9 and 11. The proﬁ le across the northern part of 5e, ca. 122 ka ago) was at least as high as the Savu (Fig. 10A) shows some very wide terraces at Th age (yr)

top of the coral rampart, which varies from 18 100 m, 130–150 m, 180–220 m, and 300–330 m. (corrected) 230 to 22 m in elevation (Fig. 11). Wave action, from The minimum uplift rate of 0.2 mm/a provides a subsequent sea-level highstands (stage 5c and poor match between known sea level high-stands 5a, or stage 3), is responsible for eroding into and terrace elevations above 68–72 m. Higher the thick stage 5e coral reef as it was uplifted. uplift rates of at least 0.4–0.5 mm/a are needed The wave-cut terrace is also uplifted and warped, to ﬁ t broad terraces with major isotope stages Th ages assume the initial Th ages assume the initial U value of 3.8. The errors are arbitrarily indicating recent deformation. (Fig. 11A). The match of sea level high-stands Th age (yr) 230 238 230 (uncorrected) Th /

A Profile of major coral terraces of N. Savu 232

North South U . Corrected 238 300-330 m (23-25) –1 y –10 300 Th / (activity) (activity)

190-220 m (19, 21) 230 120-150 (15,17) 200 . 100 m (13) Elev. (m) Elev. 234 xT 100 λ 68 m (11) x e x

0 U* = 1.55125 x 10 234

0 5 10 15 20 25 30 Th DATING RESULTS δ 238 λ 230 Aimau Profile Distance from N. Coast (km) measured , (measured) U –1 234 y δ

100 –6 ) = –6 TABLE 2. Th initial Elev. (m) 232 B Aimau Coral Terrace Profile 68- 72 U 234

47-52 50 δ Th / 34

0.2 mm/yr 230 = 2.8263 x 10 18-22 (atomic x 10 234 λ 0 3-5 ,

0 –1 100 m y –6 Th age (T), i.e.,

5c Th 230 (ppt)

5e 5a Detail of Wave Eroded Terrace 232 -50 C lagoon 11 9 7 5 Sea Level at ~120 ka meters b.s.l. wave 20 eroded – 1] x 1000. – 18 coral = 9.1577 x 10 δ O stages and forereef 10

-100 230 sediment λ sea level curve U activity . Those are the values for a material at secular equilibrium, the bulk earth with base . 238 σ –6 0 m (ppb) U) U) Present Sea Level 400 300 200 100 0 238 Time (ka) 121,593 ± 736 yrs old forereef coral U / on wave eroded terrace 234 was calculated based on was

Figure 11. Proﬁ les of uplifted coral terraces on the north coast of Savu near Aimau (see initial : The error is 2 U = [( Fig. 7 for location). (A) Proﬁ le from coast to highest terrace with elevations and inferred U 234 234 number Sample δ δ † oxygen isotope stages. Thrust faults correlate with those in Figure 7. (B) Detail of lowest Note 7/24-2C#1 32.4 ± 0.1 276 ± 6 1672 ± 36 149 ± 13 0.8635 ± 0.0044 144619 ± 4060 144416 ± 4052 224 ± 20

7/24-2A-I 7/24-2A-II 7/30-1A 229.4 ± 0.3 230.1 ± 0.2 * 652 ± 7 632 ± 7 2033 ± 3 73652 ± 521 5941 ± 64 6132 ± 66 7.0 ± 0.3 150.5 ± 2.3 148.2 ± 1.7 1.0226 ± 0.0043 1.0210 ± 0.0029 147.1 ± 1.7 215180 ± 3052 0.0154 ± 0.0007 215716 ± 2135 215116 ± 3050 215653 ± 2134 1476 ± 64 276.5 ± 4.9 272.7 ± 3.5 554 ± 468 147.3 ± 1.8 7/23/01BR#2B 1692 ± 2 12281 ± 24 1706 ± 6 99.8 ± 1.4 0.7497 ± 0.0023 121778 ± 732 121593 ± 736 140.7 ± 2.0 assumed to be 50%.

ratio of 4.4 ± 2.2 x 10 part of proﬁ le A correlating terraces with sea-level highstands according to U/Th age of 7/23-3D 167.6 ± 0.2 237 ± 5 10958 ± 246 104.8 ± 2.2 0.9367 ± 0.0045 193309 ± 2759 193274 ± 2758 180.9 ± 4.1 lowest surface (3–5 m) (b.s.l.—below sea level). Assuming constant uplift yields a best-ﬁ t correlation of 0.2 mm/a. (C) Detail of lowest terraces in B. Terrace at 3–5 m is a wave- eroded platform cut into the stage 5e terrace, exposing forereef sediments. Eroded parts of reconstructed forereef are shown (dashed).

Geosphere, June 2009 167

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

with terraces may also be complicated by thrust the strike of the island. Longitudinal profi les from on elevation versus log distance from the divide. faults inferred on the basis of juxtaposition of seven different rivers on Savu (Fig. 12) were pro- South-draining streams have lower order, high- different uplift patterns (Fig. 7). duced from 1:25,000 topographic maps and plot- gradient profi les. North-draining stream profi les All of the terraces we surveyed are laterally ted according to the method of Merritts and Vin- show only one concave profi le and the rest are warped, with the highest elevations near and cent (1989). Stream gradient points were taken distinctly convex. Stream gradients and degree of south of the Aimau region (Fig. 7). Terraces every 500 m from the coast to drainage divides. convexity generally increase toward the Aimau decrease slightly in elevation to the west, where Stream profi les differ on either side of the upwarp, where coral terraces also achieve their they wrap around the northwest coast and change east-northeast–west-southwest—trending maximum elevations. In most drainages, differ- geometry abruptly above an elevation of ~30– divide of Savu (Fig. 12). Rivers south of the ences in lithology show no detectable effect on 40 m. The change corresponds with where thrust divide show linear to slightly concave profi les stream gradient. Climate variation on the island 3, which is found in offshore seismic profi les, intersects with the coastline (Fig. 7). We infer that thrust 3 tracks beneath the terraces and joins with another recently active thrust segment found Aimau off Point Bali at the eastern tip of the island. We upwarp also infer that the well-developed wide terraces Loko Diagama on Point Bali are on the hanging wall of thrust 3. However, to the south, the broad terraces of Point Bali are truncated against older coral terraces that tilt eastward from 78 m elevation to sea level over N a distance of only 3 km (Fig. 7). We infer a fault at this location to explain these discontinuities, Drainage divide but the sense of shear is not clear. On the south coast of Savu, an extensive coral carapace is tilted seaward as much as 5° (Fig. 5A) from elevations of ~25 m to below sea 5 km level over a distance of 250–300 m. Tridacna A B shells found on this terrace did not produce reli- 250 250 able ages. However, the fact that the terrace is at least 25 m above sea level indicates that the 200 200 south coast has also undergone some Pleisto- cene uplift. The lack of multiple fl ights of ter- 150 150 races, and the southward tilt into the sea, indi- 100 100 cates that uplift was slow or short lived, and did not keep pace with the north coast. 50 The western coastal regions of Savu have 50 coral terraces that dip as much as 8° westward 0 0 beneath the narrow Rai Jua Strait between Savu 1,000 10,000 0 5,000 10,000 and Rai Jua Islands. On the east coast of Rai Jua coral terraces are observed that dip ~5° east- C D 250 250 ward, back toward the strait. The corresponding dips on each side of the strait help us rule out 200 200 slumping as a cause of the deformation. Instead, we interpret the deformation pattern as a result 150 150 of differences in amounts of slip along strike of 100 various segments of the Savu thrust. In this case, mElevation, m Elevation, 100 medium high the strait may represent a fault segment boundary where the tips of two thrust faults end or 50 50 low slightly overlap, leaving a slip defi cit. In summary, we interpret the deformation pat- 0 0 1,000 10,000 0 5,000 10,000 tern of coral terraces, with continuous uplift on Distance from Divide, m Distance from Divide, m the north coast, uplift followed by southward tilt on the south coast, and warping that corresponds Figure 12. Longitudinal stream profi les of Savu. Various dashed lines correspond to loca- to topography, as a result of activity along the tions of streams on map with black lines for streams north of the divide (A and B) and gray Savu thrust system over the past 800 ka. lines representing streams south of the divide (C and D). Data acquired every 500 m within drainages. Horizontal scales are semilogarithmic on left and arithmetic on right. Colored Stream Longitudinal Profi les lines represent profi le estimates of relative uplift rate (Merritts and Vincent, 1989). Plots show that uplift rates were high in the recent past along the south coast, but now only have As Savu emerged above sea level, drainages low-order streams to base level. North-draining streams show intermediate uplift rates developed in response to differential uplift across with some convex-shaped profi les.

168 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

is also negligible due to little to no rain shadow. south of Savu (Fig. 1), which means that the Scott tle lag time between initial uplift ca. 3–4 Ma ago We consider tectonics as the major variable con- Plateau arrived at the trench ca. 1.3–1.7 Ma ago. as the distal edge of the continental slope col- trolling stream profi le gradient and shape. The collision of the Scott Plateau with the lided with the Sunda-Banda trench, and a later Comparisons of Savu stream profi les with Sunda-Banda arc has also modifi ed other fea- phase of rapid uplift signaling the arrival of the other studies (Merritts and Vincent, 1989) indi- tures of the arc-trench system. For example, Australian shelf at the trench ca. 0.2 Ma ago (De cate mostly medium rates of uplift north of the the distance between the forearc ridge crest and Smet et al., 1990). Between these two pulses of divide (Fig. 12), and those streams have sig- the arc is consistently ~230–240 km throughout uplift the accretionary wedge subsided. These nifi cantly higher uplift rates than south-directed the eastern Sunda arc, which is only 50–60 km along-strike variations in timing, and the epi- streams that show only evidence of past uplift from the trench (van der Werff et al., 1995b). sodic advance of the orogen that they cause, (steep low-order stream gradients). The asym- At ~118°E (just west of Sumba) these distances are very diffi cult to resolve in reconstructing metry of drainage basin length north and south begin to change. A distinct westward bend of the older, more mature orogens. However, these of the divide also demonstrates higher rates 5000 m depth contour of the lower plate is found. events may explain some of the disparate ages of uplift to the north. Overall, the analysis of North of this shallower section of the lower of similar features and structural discontinuities stream profi les corroborates other evidence plate, the Java trench and Savu accretionary in older orogens. for the northward advance of the Scott Plateau ridge are indented ~40 km to the north (Fig. 1). beneath Savu during the Quaternary. Sumba rises in what was the forearc basin, and CONCLUSION directly to the north of western Sumba the arc COLLISION INITIATION AND is inactive. In the backarc north of Sumba, the The Savu region provides a rare glimpse of EVOLUTION IN THE SAVU REGION Flores thrust is also defl ected northward >30 km neotectonic associations of the transition from (Silver et al., 1986). We interpret these features subduction to collision. Complexities result To estimate the timing of events that shaped as modifi cations to the eastern Sunda arc by col- from lower plate structural and stratigraphic Savu, we reconstruct the position of the Scott lision of the westernmost Scott Plateau protru- discontinuities that cause waves of crustal short- Plateau through time and compare it with the sion that currently is underthrusting the Sunda ening and uplift followed by subsidence. The uplift history. We use the northeast edge of the forearc beneath Sumba (Fig. 1). geology of the Savu region is separated into two Scott Plateau as a reference point, which cor- The estimated time of collision initiation at very different domains divided by a develop- responds with the axis of the Sumba Ridge the Sunda-Banda deformation front and the ing suture zone between units of Australian and ~10 km north of the Savu (Fig. 1). Palinspastic observed collisional modifi cations to the arc- Asian affi nity. Australian affi nity units correlate reconstruction of the position of the northeast trench system are consistent with other results with the Gondwana sequence that is also found edge of the Scott Plateau is possible because we present here, such as (1) foraminifera analy- in Rote and Timor. Asian affi nity units correlate we know the convergence rate between it, the ses that document rapid uplift from precolli- with those exposed in Sumba and other parts volcanic arc, and Savu. Long-term rates of con- sional depths of a accretionary ridge to the sur- of the Sunda-Banda arc, and high-level thrust vergence between Australia and Eurasia are face later than 1.90 Ma ago, (2) uplifted coral sheets in Timor. estimated as 71 km/Ma at 014° (Kreemer et al., terraces as old as 800 ka, (3) faults that offset The shortening history of the Savu region 2003). These rate estimates based on geological these Quaternary deposits, and (4) an active involves thrust duplex development of mostly features are nearly identical to GPS measure- fold-and-thrust belt tracking the northward pro- Triassic–Jurassic synrift Gondwana sequence ments between Australia (Darwin) and Eurasia gression of the Sumba Ridge. sedimentary and volcanic rocks accreted from of 69 ± 0.3 mm/a at 016° (Nugroho et al., 2009). The shape of the Australian continental the subducting Scott Plateau by top-to-the- Both geological data and GPS measurements margin changes south of Savu to a more north- southeast thrusting. The direction of shortening also document that the forearc and arc move northeast–south-southwest direction, which is controlled by the inherited east-northeast– north-northeast relative to Eurasia. Savu moves is nearly parallel to its convergence direction. west-southwest structural grain of the Austra- at a rate of 31 ± 1 mm/a and the western part This implies that the shelf-slope break that is lian continental margin. The southern edge of of Flores at a rate of 23 mm/a (Nugroho et al., currently colliding with West Timor will not the Sunda-Banda forearc forms the backstop 2009). We infer that the difference in motion arrive at the deformation front south of Savu for of the accretionary wedge via the Savu thrust between Savu and western Flores (8 ± 1 mm/a) another 6.0 Ma. South of Sumba the Scott Pla- system. Increased coupling between the under- is the deformation rate on the Savu thrust sys- teau has already underthrust the Sunda-Banda thrust Scott Plateau and the overriding Sunda- tem. The net motion of the Scott Plateau rela- forearc and now oceanic lithosphere is again Banda forearc actively drives the rear of the tive to Savu once it arrives at the Java Trench subducting beneath the deformation front. The accretionary wedge northward over the Sunda- is 40 mm/a (71 mm/a – 31 mm/a). At this rate, next part of the Australian continental margin to Banda forearc basin at rates of ~8 mm/a. This the Scott Plateau would have arrived in south- collide with Sumba will likely be the Exmouth northward motion is accommodated by the ern Savu ca. 0.82 Ma ago and the north coast of Plateau, which has a large northwest-south- active Savu thrust system, which offsets Qua- Savu at 0.25 Ma ago. east–oriented crustal prong very similar to the ternary deposits on the north coast and is well In order to determine when the Scott Plateau Scott Plateau (Longley et al., 2002). During this imaged by seismic refl ection profi les offshore. fi rst collided with the Sunda-Banda trench south event ~9–10 Ma from now, the arc- continent The uplift history of Savu documents the of Savu, we reconstruct the position of the trench collision may extend as far as Bali, but the arrival and passing of the northeast edge of the based on the consistent 290–305 km arc-trench rest of the Sunda arc will most likely remain a Scott Plateau beneath the Savu region between distance of the eastern Sunda arc. At this dis- subduction zone unless plate boundaries in the 0.8 and 0.2 Ma ago. Foraminifera from syno- tance from the active volcanic arc north of Savu, region reorganize. rogenic deposits reveal pelagic depths of 2.5– the trench would have been at latitude 11.2° to In contrast, the confi guration of the Australian 4.0 km between 5.6 Ma and more recently than 11.4°S. The Sumba Ridge is 97–118 km from continental margin south of West Timor (lower 1.9 Ma ago. By roughly 1 Ma ago, neritic spe- the inferred position of the Sunda-Banda trench angle of collisional obliquity) results in very lit- cies were present that document uplift of the

Geosphere, June 2009 169

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Harris et al.

M.E., and Shackleton, N.J., 1995, Late Neo- Dickinson, W.R., Beard, L.S., Brakenridge, G.R., Erjavec, synorogenic section to within 200 m of the sur- gene chronology; new perspectives in high- J.L., Ferguson, R.C., Inman, K.F., Knepp, R.A., face in <1 Ma. By ca. 0.8 Ma ago, coral terraces resolution stratigraphy: Geological Society of Lindber, F.A., and Ryberg, P.T., 1983, Provenance formed in the central part of the island above the America Bulletin, v. 107, p. 1272–1287, doi: of North American Phanerozoic sandstones in 10.1130/0016-7606(1995)107<1272:LNCNPI> relation to tectonic setting: Geological Soci- deepest structural levels of the exposed accre- 2.3.CO;2. ety of America Bulletin, v. 94, p. 222–235, doi: tionary wedge. The island continued to grow Berry, R.F., and Grady, A.E., 1981, Deformation and meta- 10.1130/0016-7606(1983)94<222:PONAPS> morphism of the Aileu Formation, north coast, 2.0.CO;2. northward at a rate of ~2 m/ka, as recorded by East Timor and its tectonic signifi cance: Jour- Dore, A.G., and Stewart, I.C., 2002, Similarities and differ- fl ights of uplifted coral. The youngest coral ter- nal of Structural Geology, v. 3, p. 143–167, doi: ences in the tectonics of two passive margins; the races are erosional features into the massive 10.1016/0191-8141(81)90011-0. Northeast Atlantic margin and the Australian North Berry, R.F., and Jenner, G.A., 1982, Basalt geochemistry West Shelf, in Keep, M., and Moss, S.J., eds., Sedi- MIS 5e (122 ka) coral complex. as a test of tectonic models of Timor: Geological mentary basins of Western Australia: Petroleum Savu Island documents that initiation of intra- Society of London Journal, v. 139, p. 593–604, doi: Exploration Society of Australia Symposium Pro- forearc convergence, and partitioning of strain 10.1144/gsjgs.139.5.0593. ceedings, v. 3, p. 89–117. Bird, P.R., and Cook, S.E., 1991, Permo-Triassic successions Fortuin, A.R., van der Werff, W., and Wensink, H., 1997, away from the deformation front, and is one of of the Kekneno area, West Timor: Implications for Neogene basin history and paleomagnetism of a the earliest transitional phases from subduction paleogeography and basin evolution: Journal of rifted and inverted forearc region, on- and offshore Southeast Asian Earth Sciences, v. 6, p. 359–371, Sumba, eastern Indonesia: Journal of Asian Earth to collision. Although this phase of deforma- doi: 10.1016/0743-9547(91)90081-8. Sciences, v. 15, p. 61–88. tion commonly occurs in subduction zones due Bowin, C., Purdy, G.M., Johnston, C., Shor, G., Lawyer, L., Ganssen, G., Troelstra, S.R., Faber, B., van der Kaars, to collision of seamounts and other lower plate Hartono, H.M.S., and Jezek, P., 1980, Arc-continent W.A., and Situmorant, M., 1989, Late Quaternary collision in the Banda Sea region: American Asso- palaeoceanography of the Banda Sea, eastern Indo- obstructions, it signals the beginning of the end ciation of Petroleum Geologists Bulletin, v. 64, nesia piston cores (Snellius-II Expedition, Cruise of subduction in the eastern Sunda-Banda arc. p. 868–918. G5): Netherlands Journal of Sea Research, v. 24, Similar phases of deformation occurred earlier Breen, N.A., Silver, E.A., and Hussong, D.M., 1986, p. 491–494. Structural styles of an accretionary wedge south Genrich, J.F., Bock, Y., McCaffrey, R., Calais, E., Stevens, both west and east of Savu Island due to the of the island of Sumba, Indonesia, revealed by C.W., and Subarya, C., 1996, Accretion of the V-shaped geometry of the northeast-southwest SeaMarcII side scan sonar: Geological Society southern Banda arc to the Australian plate mar- of America Bulletin, v. 97, p. 1250–1261, doi: gin determined by Global Positioning System Australian continental margin and the northwest - 10.1130/0016-7606(1986)97<1250:SSOAAW> measurements: Tectonics, v. 15, p. 288–295, doi: southeast protrusion of the Scott Plateau. 2.0.CO;2. 10.1029/95TC03850. Brouwer, H.A., 1925, The geology of the Netherlands East Gianni, L., 1971, The geology of the Belu district of Indo- ACKNOWLEDGMENTS Indies: Lectures delivered as exchange-professor nesian Timor [M.Ph. thesis]: London, University of at the University of Michigan in 1921–1922: New London, 122 p. York, Macmillan, 160 p. Gradstein, F.M., Ogg, J.G., and Smith, A.G., 2005, A geo- This research was funded by grants from the Carter, D.J., Audley-Charles, M.G., and Barber, A.J., 1976, logic time scale 2004: Cambridge, Cambridge Uni- National Science Foundation (EAR-0337221) and the Stratigraphical analysis of island arc-continental versity Press, 610 p. Kennedy Center for International Studies at Brigham margin collision in eastern Indonesia: Geological Grady, A.E., Abbott, M.J., Chamalaun, F.H., and von der Young University. Universitas Pembangunan Nasi- Society of London Journal, v. 132, p. 179–198, doi: Borch, C.C., 1983, The Banda Arc, Sumba to onal Yogyakarta sponsors our research in Indonesia. 10.1144/gsjgs.132.2.0179. Timor: A review: Geological Society of Australia We appreciate the fi eld and surveying assistance of Chamalaun, F.H., and Grady, A.E., 1978, The tectonic evo- Journal, v. 9, p. 23–25. lution of Timor: A new model and its implications Haig, D.W., McCartain, E., Barber, L., and Backhouse, J., my son Nathan Harris, and the invaluable logistical for petroleum exploration: Australian Petroleum 2007, Triassic–Lower Jurassic foraminiferal indi- support of Geneviève Duggan. Helpful reviews were Exploration Association Journal, v. 18, p. 102–108. ces for Bahaman-type carbonate-bank limestones, provided by Andrew Meigs, Michael Audley-Charles, Chamalaun, F.H., Grady, A.E., von der Borch, C.C., and Cablac Mountain, East Timor: Journal of Forami- and Ron Berry. Hartono, H.M.S., 1982, Banda arc tectonics: The niferal Research, v. 37, p. 248–264, doi: 10.2113/ signifi cance of the Sumba Island (Indonesia), in gsjfr.37.3.248. REFERENCES CITED Watkins, J.S., and Drake, C.L., eds., Studies in con- Haile, H.S., Barber, A.J., and Carter, D.J., 1979, Mesozoic tinental marine geology: American Association of cherts on crystalline schists in Sulawesi and Timor: Petroleum Geologists Memoir 34, p. 361–375. Geological Society of London Journal, v. 136, Abbot, M.J., and Chamalaun, F.H., 1981, Geochronology Chappell, J., and Veeh, H., 1978, Late Quaternary tec- p. 65–70, doi: 10.1144/gsjgs.136.1.0065. of some Banda arc volcanics, in Barber, A.J., and tonic movements and sea-level changes at Hamilton, W., 1979, Tectonics of the Indonesian region: U.S. Wiryosuyono, S., eds., The geology and tectonics of Timor and Atauro Island: Geological Society Geological Survey Professional Paper 1078, 345 p. eastern Indonesia: Bandung, Indonesia, Geological of America Bulletin, v. 89, p. 356–368, doi: Harris, R.A., 1991, Temporal distribution of strain in the Research and Development Center Special Publica- 10.1130/0016-7606(1978)89<356:LQTMAS> active Banda orogen: A reconciliation of rival tion 2, p. 253–268. 2.0.CO;2. hypotheses: Journal of Southeast Asian Earth Sci- Audley-Charles, M.G., 1968, The geology of Portuguese Charlton, T.R., 1991, Postcollision extension in arc-continent ences, v. 6, p. 373–386, doi: 10.1016/0743-9547 Timor: Geological Society of London Memoir 4, 75 collision zones, eastern Indonesia: Geology, v. 19, (91)90082-9. p., doi: 10.1144/GSL.MEM.1968.004.01.01. p. 28–31, doi: 10.1130/0091-7613(1991)019<0028: Harris, R.A., 1992, Peri-collisional extension and the forma- Audley-Charles, M.G., 1985, The Sumba enigma: Is PEIACC>2.3.CO;2. tion of Oman-type ophiolites in the Banda Arc and Sumba a diapiric forearc-nappe in process of for- Cox, N.L., Harris, R.A., and Merritts, D., 2006, Quaternary Brooks Range, in Parson, L.M. et al., eds., Ophio- mation?: Tectonophysics, v. 119, p. 435–449, doi: uplift of coral terraces from active folding and lites and their modern oceanic analogues: Geo- 10.1016/0040-1951(85)90049-6. thrusting along the northern coast of Timor-Leste: logical Society of London Special Publication 60, Audley-Charles, M.G., 1986a, Timor-Tanimbar Trough: Eos (Transactions, American Geophysical Union), p. 301–325. The foreland basin to the evolving Banda orogen, v. 87, abs. T51D-1564. Harris, R.A., 2004, Geodynamic patterns of ophiolites and in Allen, P.A., and Homewood, P., eds., Foreland Crawford, A.J., and von Rad, U., 1994, The petrology, geo- marginal basins of the Indonesian and New guinea basins: International Association of Sedimentolo- chemistry and implications of basalts dredged from regions, in Dilek, Y., and Robinson, P.T., eds., gists Special Publication 8, p. 91–102. the Rowley Terrace–Scott Plateau and Exmouth Ophiolites in Earth history: Geological Society of Audley-Charles, M.G., 1986b, Rates of Neogene and Qua- Plateau margins, northwestern Australia: AGSO London Special Publication 218, p. 481–505. ternary tectonic movements in the Southern Banda Journal of Australian Geology & Geophysics, v. 15, Harris, R.A., 2006, Rise and fall of the eastern Great Indo- arc based on micropaleontology: Geological Soci- p. 43–55. nesian Arc recorded by the assembly, dispersion ety of London Journal, v. 143, p. 161–175, doi: Crostella, A.A., and Powell, D.E., 1976, Geology and and accretion of the Banda Terrane, Timor: Gond- 10.1144/gsjgs.143.1.0161. hydrocarbon prospects of the Timor area: Indo- wana Research, v. 10, p. 207–231, doi: 10.1016/j.gr Audley-Charles, M.G., 2004, Ocean trench blocked and nesian Petroleum Association Proceedings, v. 2, .2006.05.010. obliterated by Banda forearc collision with Aus- p. 149–171. Harris, R.A., Sawyer, R.K., and Audley-Charles, M.G., tralian proximal continental slope: Tectonophysics, De Smet, M.E.M., Fortuin, A.R., Troelstra, S.R., van Marle, 1998, Collisional mélange development: Geologic v. 389, p. 65–79, doi: 10.1016/j.tecto.2004.07.048. L.J., Karmini, M., Tjokrosapoetro, S., and Hadi- associations of active mélange-forming processes Barber, A.J., Audley-Charles, M.G., and Carter, D.J., 1987, wasastra, S., 1990, Detection of collision-related with exhumed mélange facies in the western Banda Thrust tectonics in Timor: Geological Society of vertical movements in the Outer Banda Arc (Timor, orogen, Indonesia: Tectonics, v. 17, p. 458–479, Australia Journal, v. 24, p. 51–62. Indonesia), using micropaleontological data: doi: 10.1029/97TC03083. Berggren, W.A., Hilgen, F.J., Langereis, C.G., Kent, Journal of Southeast Asian Earth Sciences, v. 4, Harris, R.A., Kaiser, J., Hurford, A.J., and Carter, A., 2000, D.V., Obradovich, J.D., Raffi , I., Raymo, p. 337–356. Thermal history of Australian passive margin

170 Geosphere, June 2009

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021 Transition from subduction to arc-continent collision

sequences accreted to Timor during late Neogene cal Research, v. 90, p. 4511–4528, doi: 10.1029/ collision: Journal of Geophysical Research, v. 88, arc-continent collision, Indonesia: Journal of Asian JB090iB06p04511. p. 7429–7448, doi: 10.1029/JB088iB09p07429. Earth Sciences, v. 18, p. 47–69, doi: 10.1016/ Merritts, D., and Vincent, K.R., 1989, Geomorphic Silver, E.A., Breen, N.A., Prasetyo, H., and Hussong, S1367-9120(99)00036-X. response of coastal streams to low, intermediate D.M., 1986, Multibeam study of the Flores back- Huiqi, L., McClay, K.R., and Powell, D., 1992, Physical and high rates of uplift, Mendocino triple junction arc thrust belt, Indonesia: Journal of Geophysi- models of thrust wedges, in McClay, K.R., ed., region, northern California: Geological Society cal Research, v. 91, p. 3489–3500, doi: 10.1029/ Thrust Tectonics: New York, Chapman and Hall, of America Bulletin, v. 101, p. 1373–1388, doi: JB091iB03p03489. p. 71–81. 10.1130/0016-7606(1989)101<1373:GROCST> Snyder, D.B., Prasetyo, H., Blundell, D.J., Pigram, C.J., Bar- Kaneko, Y., Maruyama, S., Kadarusman, A., Ota, T., Ishi- 2.3.CO;2. ber, A.J., Richardson, A., and Tjokosaproetro, S., kawa, M., Tsujimori, T., Ishikawa, A., and Oka- Molengraaff, G.A.F., 1912, On recent crustal movements 1996, A dual doubly-vergent orogen in the Banda moto, K., 2007, On-going orogeny in the outer in the Island of Timor and their bearing on the Arc continent-arc collision zone as observed on arc of Timor-Tanimbar region, eastern Indonesia, geological history of the East Indian Archipelago: deep seismic refl ection profi les: Tectonics, v. 15, in Santosh, M., and Maruyama, S., eds., Island Koninkijke Nederlandse Akademie van Weten- p. 34–53, doi: 10.1029/95TC02352. arcs past and present: Gondwana Research, v. 11, schappen, v. 15, p. 224–235. Standley, C., and Harris, R., 2009, Banda forearc basement p. 218–233. Nugroho, H., Harris, R., Amin, W.L., and Maruf, B., 2009, accreted to the NW Australian continental margin: Karig, D.E., Barber, A.J., Charlton, T.R., Klemperer, Active plate boundary reorganization in the Banda A geochemical, age and structural analysis of the S., and Hussong, D.M., 1987, Nature and dis- arc-continent collision: insights from new GPS Lolotoi metamorphic complex of East Timor: Tec- tribution of deformation across the Banda Arc- measurements: Tectonophysics, doi: 10.1016/j.tecto tonophysics, doi: 10.1016/j.tecto.2009.01.034. Australian collision zone at Timor: Geological .2009.01.026. Symonds, P.A., Planke, S., Frey, O., and Skogseid, J., 1998, Society of America Bulletin, v. 98, p. 18–32, doi: Pearce, J.A., and Cann, J., 1973, Tectonic setting of basic Volcanic evolution of the western Australian conti- 10.1130/0016-7606(1987)98<18:NADODA> volcanic rocks determined using trace element anal- nental margin and its implications for basin devel- 2.0.CO;2. yses: Earth and Planetary Science Letters, v. 19, opment: Petroleum Exploration Society of Australia Keep, M., Clough, M., Langhi, L., and Moss, S.J., 2002, p. 290–300, doi: 10.1016/0012-821X(73)90129-5. Proceedings, v. 2, p. 33–54. Neogene tectonic and structural evolution of the Pirazzoli, P.A., Radtke, U., Hantoro, W.S., Jouannic, C., van der Werff, W., 1995a, Cenozoic evolution of the Savu Timor Sea region, NW Australia, in Keep, M., and Hoang, C.T., Causse, C., and Borel-Best, M., 1993, Basin, Indonesia: Forearc basin response to arc/ Moss, S.J., eds., Sedimentary basins of Western A one million-year-long sequence of marine ter- continent collision: Marine and Petroleum Geol- Australia: Petroleum Exploration Society of Austra- races on Sumba Island, Indonesia: Marine Geol- ogy, v. 12, p. 247–262, doi: 10.1016/0264-8172 lia Symposium Proceedings, v. 3, p. 341–353. ogy, v. 109, p. 221–236, doi: 10.1016/0025-3227 (95)98378-I. Klement, K., 1960, Dinofl aggellaten und Hystrichosphaeri- (93)90062-Z. van der Werff, W., 1995b, Structure and morphotectonics of deen aus dem Malm Suedwestdeutschlands unter Ramsay, D.C., and Exon, N.F., 1994, Structure and tectonic his- the accretionary prism in the eastern Sunda/western besonderer Beruecksichtigung stratigraphisch wich- tory of the northern Exmouth Plateau and Rowley Ter- Banda Arc: Journal of Southeast Asian Earth Sci- tiger Arten: Palaeontologishe Zeitschrift, v. 34, p. 11. race; outer North West Shelf: AGSO Journal of Austra- ences, v. 11, p. 309–322, doi: 10.1016/0743-9547 Kreemer, C., Holt, W.E., and Haines, A.J., 2003, An inte- lian Geology & Geophysics, v. 15, p. 55–70. (94)00038-G. grated global model of present-day plate motions Reed, D.L., Silver, E.A., Prasetyo, H., and Meyer, A.W., Van Marle, L.J., van Hinte, J.E., and Nederbragt, A.J., 1987, and plate boundary deformation: Geophysi- 1986, Deformation and sedimentation along Plankton percentage of the foraminiferal fauna in cal Journal International, v. 154, p. 8–34, doi: a developing terrane suture: Sunda forearc, seafl oor samples from the Australian-Irian Jaya 10.1046/j.1365-246X.2003.01917.x. Indonesia: Geology, v. 14, p. 1000–1003, doi: continental margin, eastern Indonesia: Marine Lambeck, K., and Chappell, J., 2001, Sea level change 10.1130/0091-7613(1986)14<1000:DASAAD> Geology, v. 77, p. 151–156, doi: 10.1016/0025-3227 through the last glacial cycle: Science, v. 292, 2.0.CO;2. (87)90089-2. p. 679–686, doi: 10.1126/science.1059549. Roosmawati, N., and Harris, R., 2009, Long-term surface van Weering, T.C.E., Kusnida, D., Tjokrosapoetro, S., Lubis, Longley, I.M., Buessenschuett, C., and Clydsdale, L., 2002, uplift history of the active Banda arc-continent col- S., and Kridohato, P., 1989, Slumping, sliding and The North West Shelf of Australia—A Woodside lision: Depth and age analysis of foraminifera from the occurrence of acoustic voids in recent and sub- perspective, in Keep, M. and Moss, S. eds., Sedi- Rote and Savu Islands, Indonesia: Tectonophysics, recent sediments of the Savu forearc basin, Indo- mentary basins of Western Australia: Petroleum doi: 10.1016/j.tecto.2009.04.009. nesia: Netherlands Journal of Sea Research, v. 24, Exploration Society of Australia Symposium Pro- Rutherford, E., Burke, K., and Lytwyn, J., 2001, Tectonic p. 415–430, doi: 10.1016/0077-7579(89)90119-1. ceedings, v. 3, p. 27–88. history of Sunda Island, Indonesia, since the Late Williams, H.F., Turner, F.J., and Gilbert, C.M., 1982, Petrog- Ludden, J.N., and Dionne, B., 1992, The geochemistry of Cretaceous and its rapid escape into the forearc in raphy, an introduction to the study of rocks in thin oceanic crust at the onset of rifting in the Indian the Miocene: Journal of Asian Earth Sciences, v. 19, section (second edition): San Francisco, W.H. Free- Ocean, in Gradstein F.M., and Ludden, J.N., Pro- p. 453–479, doi: 10.1016/S1367-9120(00)00032-8. man and Co., 626 p. ceedings of the Ocean Drilling Program, Scientifi c Sani, K., Jacobson, M.L., and Sigit, R., 1995, The thin- Zobell, E., 2007, Origin and tectonic evolution of Gondwana results, Volume 123: College Station, Texas, Ocean skinned thrust structures of Timor: Jakarta, Indo- sequence units accreted to the Banda Arc: A struc- Drilling Program, p. 791–799. nesian Petroleum Association 24th Annual Conven- tural transect through central East Timor [M.S. the- Masson, D.G., Milsom, J., Barber, A.J., Sikumbang, N., tion, Proceedings, p. 277–293. sis]: Provo, Utah, Brigham Young University, 75 p. and Dwiyanto, B., 1991, Recent tectonics around Sawyer, R.K., Sani, K., and Brown, S., 1993, The stratigra- the island of Timor, eastern Indonesia: Marine phy and sedimentology of West Timor, Indonesia: and Petroleum Geology, v. 8, p. 35–49, doi: Indonesian Petroleum Association Proceedings, 10.1016/0264-8172(91)90043-Z. p. 534–574. McCaffrey, R., 1988, Active tectonics of the eastern Sunda Shervais, J.W., 1982, Ti-V plots and the petrogenesis of and Banda Arcs: Journal of Geophysical Research, modern and ophiolitic lavas: Earth and Plan- v. 93, p. 15,163–15,182. etary Science Letters, v. 59, p. 101–118, doi: McCaffrey, R., Molnar, P., and Roecker, S.W., 1985, 10.1016/0012-821X(82)90120-0. Microearthquake seismicity and fault plane solu- Silver, E.A., Reed, R., McCaffrey, R., and Joyodiwiryo, MANUSCRIPT RECEIVED 14 SEPTEMBER 2008 tions related to arc-continent collision in the east- Y., 1983, Back arc thrusting in the eastern Sunda REVISED MANUSCRIPT RECEIVED 31 JANUARY 2009 ern Sunda arc, Indonesia: Journal of Geophysi- arc, Indonesia: A consequence of arc-continent MANUSCRIPT ACCEPTED 10 FEBRUARY 2009

Geosphere, June 2009 171

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/5/3/152/3337883/i1553-040X-5-3-152.pdf by guest on 25 September 2021