Comparison of sedimentary processes on adjacent
passive and active continental margins offshore of SW
Taiwan based on echo character studies
J.-K. Chiu and C.-S. Liu
Institute of Oceanography, National Taiwan University,Taipei,Taiwan, ROC
ABSTR AC T
Echo character recorded on Chirp sub -bottom sonar data from o¡shore of SW Taiwan was analysed to examine and compare the sedimentary processes of adjacent passive and active continental margin settings. Sea£oor echoes in the study area are classi¢ed into four types: (1) distinct echoes, (2) indistinct echoes, (3) hyperbolic echoes and (4) irregular echoes. Based on the mapped distribution of the echo types, the sedimentary processes o¡shore of SW Taiwan are di¡erent in the two tectonic settings. On the passive South China Sea (SCS) margin, slope failure is the main process on the upper continental slope, whereas turbidite deposits accumulate in the lower continental slope. In contrast, the submarineTaiwan orogenic wedge is characterized by ¢ll-and- spill processes in intraslope basins on the upper slope, and mass-transport deposits are observed in the canyons and on the lower Kaoping slope. This di¡erence is largely caused by the huge in£ux of terrigenous sediments into the submarineTaiwan orogenic wedge compared with the passive SCS continental margin. In the latter, loading and movement of theTaiwan orogenic wedge has had a signi¢cant e¡ect on the sea£oor morphology and has triggered retrogressive failures. Gas hydrate dissociation may have enhanced the slope failure processes at some locations.
INTRODUCTION
Investigations of sedimentary processes have been con-ducted extensively on both passive (e.g. Chough et al., 2002; Casas et al., 2003; Mienert & Weaver, 2003) and active continental margins (e.g. MacDonald, 1993; Whitmore et al., 1999; Orpin, 2004; Yu & Huang, 2006). However, few studies have focused on direct comparison of the sedimen-tary processes on adjacent passive and active continental margins, as the present study does. For example, McAdoo et al. (2000) analysed submarine landslides in both western (active) and eastern (passive) US continental slopes, but the four areas they studied are far apart; thus, each area may have its own controlling factors on sedimentary pro -cesses besides the di¡erent tectonic settings.
The area o¡shore of SW Taiwan is the place where the passive South China Sea (SCS) continental margin meets the Taiwan orogenic wedge (Fig. 1). Active arc- continent collision has created the high mountains of Taiwan (Teng, 1990; Huang et al., 1999). Frequent earthquakes and high precipitation during typhoon seasons have caused rapid mass-wasting and high sediment output to the seas around Taiwan (Hovius et al., 2000; Dadson et al., 2003). For exam-ple, Damuth (1979, 1980a) described ¢elds of large
migrat-ing sediment waves on the seaward wall of the Manila Trench, which were deposited by turbidity currents trans-porting large volumes of sediments from Taiwan.Thus, the area o¡shore of SW Taiwan is an ideal place to study and compare sedimentary processes in continental margins containing both passive and active tectonic environments.
During the past 50 years, seabed imaging and character-ization using very high-resolution seismic (echo character) data have provided critical information that helps to under-stand the complex processes of transporting sediments from shore, across the continental margin, to the deep ocean basin.Very high-resolution sub-bottom pro¢ler data (mainly 3.5 kHz and chirp sonar) have been routinely used to reveal the echo character (also termed seismic facies) of sea£oor sediments to better understand the sedimentary processes in deep- sea environments (e.g. Damuth, 1975, 1980a; Damuth & Hayes, 1977; Damuth et al., 1983; Gaullier & Bellaiche, 1998; Lee et al., 2004; see Damuth, 1980b for a review), and more recently continental shelves (e.g. Chough et al., 2002; Shinn et al., 2007). These and other studies have demonstrated that distinct types of echoes can be identi¢ed on the sub-bottom pro¢les, and classi¢ed and mapped. These di¡erent types of echo character re£ect the combined e¡ects of sea£oor topography, subsurface geometry and sediment types and processes (Damuth, 1980b; Chough et al., 2002; Lee et al., 2005).
Damuth (1979, 1980a) studied the Quaternary sedi-mentation processes of the SCS on a regional scale using
Correspondence: Prof. Char-Shine Liu, Institute of Oceanogra-phy, National Taiwan University, Taipei, Taiwan, ROC. E-mail: [email protected]
echo - character mapping and piston cores. These studies included the slope and basin £oor in most of the present study area, but data were very widely spaced. In particular, the studies showed that large volumes of sediment derived from Taiwan were transported by turbidity currents down the Manila Trench and were deposited as high migrating sediment waves on the western wall of the trench. Yu & Lee (1992) used 3.5 kHz sub -bottom pro¢le data to study the echo characteristics of the shelf and upper slope tran-sition o¡ SW Taiwan.They suggested that shallow-marine sedimentation predominates on the Kaoping shelf and in the southern Taiwan Strait where currents and tides have modi¢ed the relict Late Pleistocene sands and Holocene muds. The continental slope in their study area is domi-nated by downslope processes that change laterally. How-ever, their study was limited to the shelf and upper slope area, where the water depth is generallyo1000 m.
Since 2002, densely spaced multichannel seismic re£ec-tion data, together with chirp sonar pro¢le data, have
been collected in a broad area o¡shore of SW Taiwan (Fig. 2) for gas hydrate investigation (Liu et al., 2006). These data sets cover the area where the active Taiwan orogenic wedge encroaches on the rifted SCS continental margin, thus providing good opportunities to investigate the structural variations and sedimentary processes in both passive and active tectonic environments. A study of tectonic/structural features of the region is reported in another paper (Lin et al., accepted). Here, sedimentary processes of the region are examined based on the chirp sonar and seismic re£ection data. We ¢rst give an overview of the sedimentary processes o¡shore SW Taiwan, mainly based on seismic images and sea£oor echo characteristics. A map of echo characteristics within the study area is constructed that reveals di¡erent sedi-mentary processes in di¡erent tectonic and sedisedi-mentary environments. Finally, we discuss the possible mechan-isms of the di¡erent sedimentary processes and their geological implications. 119°30' 120°00' 120°30' 121°00' 21°00' 21°30' 22°00' 22°30' 23°00' 118° 120° 122° 124° 18° 20° 22° 24° 26° 28° China Taiwan SCS Continent al Shelf Deform ation Front SCS Continental S lope K aoping Shelf Kao ping C anyo n Penghu Canyon 1000 2000 3000 1000 2000 3000 Kaoping Slop e Taiwan Liuchieuyu Island P enghu C anyon Formos a Can yon Fangliao Canyon Longitude (E) Lati tude (N)
Fig. 1. Geological setting of the study area o¡shore SW Taiwan. The black dashed line represents the deformation front separating the passive South China Sea (SCS) continental margin from the active submarineTaiwan orogenic wedge. Bathymetry contours are at 250 m intervals.The box inset shows the map area.
GEOLOGICAL BACKGROUND
The area o¡shore of SW Taiwan is where two distinctive tectonic provinces, the passive SCS continental margin on the west and the Luzon subduction complex on the east, meet (Liu et al., 1997, 2004; Huang et al., 1999) (Figs 1 and 2). The structural features are quite di¡erent in these two tectonic provinces: ENE^WSW-trending normal faults are observed in the passive continental margin, whereas foldandthrust structures dominate in the oro -genic wedge (Liu et al., 1997). The boundary of these two provinces is the deformation front, which extends northward from the northern end of the Manila Trench to the Tainan area onshore Taiwan (Liu et al., 2004). In addition, complex systems of mud diapirs are observed in the Kaoping shelf and the upper Kaoping slope (Sun & Liu, 1993).
Morphologically, the passive SCS continental margin consists of a shelf, slope and rise (Liu et al., 1998). The SCS continental shelf is the southwestward extension of the Taiwan Strait shelf. It is shallow (o240 m deep), relatively £at and has a width of about 200 km. The SCS continental slope is incised by a series of submarine
canyons and gullies. East of the deformation front, the submarineTaiwan orogenic wedge consists of the Kaoping shelf and Kaoping slope. The Kaoping shelf is a narrow and shallow shelf that is the seaward extension of the coast-al plain of SW Taiwan (Yu & Wen, 1992).The Kaoping slope has a very irregular surface that can be divided into an upper slope domain and a lower slope domain (Reed et al., 1992; Liu et al., 1998). The former has four main NE^SW-trending submarine canyons, whereas the latter is charac-terized by closely spaced N^S- to NW^SE-trending ridges and troughs formed by structural deformation (Yu & Song, 2000).
The gradients of the sea£oor in the study area vary from 21 to 601 (Fig. 3). Steep slopes characterize the side walls of most submarine canyons in both the orogenic wedge and the passive continental margin provinces. In the passive margin province, the upper continental slope is much steeper than the lower continental slope east of about 1191240E, but this is not the case further west (Fig. 3). In contrast, in the submarine Taiwan orogenic wedge province, the sea£oor of the upper slope domain appears to be broad and gentle, except in submarine canyons,
119°00' 21°30' 1000 1000 2000 2000 3000 3000 100 0 100 0 2000 2000 300 0 300 0 Fig. 6 Fig. 5 Fig. 8 Fig. 7 Fig. 9 Longitude (E) Latitude (N) 119°30' 120°00' 120°30' 121°00' 23°00' 22°30' 22°00' 21°00'
Fig. 2. Location map of the chirp sonar pro¢les used in this study, shown as thin lines. Thick lines show the locations of the pro¢les shown in the ¢gures as labelled. The bathymetry map is the same as in Fig. 1.
whereas in the lower slope domain, steep slopes have been observed along the £anks of many anticlinal ridges.
DATA PREPARATION
High-resolution bathymetry, chirp sub-bottom pro¢le and seismic re£ection data were used in this study to examine sea£oor morphology and sedimentary processes. Regional bathymetry (Fig. 1) is based on the 500 m grid TAIDBMv6 DTM data set (Liu et al., 1998), with a recent improvement from newly collected single-beam bathy-metric data from the SCS continental margin west of 1191E (Liu, 2007). Swath bathymetry data collected by the R/VAtalante during the1996 ACTsurvey (Lallemand et al., 1997) provided high-resolution (100 m grid) bathymetric coverage in the central portion of the study area where the SCS continental margin meets the Taiwan orogenic wedge (Liu et al., 2004). Bathymetry maps in this paper were generated using GMT (Wessel & Smith, 1993).
Approximately 9948 km of chirp sub-bottom pro¢le data have been analysed (Fig. 2). These data were collected by the R/V Ocean Researcher 1 (OR1) during seven cruises from 2002 to 2006 in an area from 211300to 221300N and 1181400to 1201350E. Most of the survey lines are oriented in a NW^SE direction with a line spacing of about 1.86 km. The sub -bottom pro¢le data were collected by a hull-mounted Bathy-2000P chirp sonar system. An 8 kHz bandwidth FM chirp waveform was used during the survey. This system can provide high-resolution sea£oor re£ection data with an 8 cm vertical resolution and bottom penetration of up to 200 m under ideal conditions.
Selected multi channel seismic (MCS) re£ection pro -¢les were used in this study to reveal the sedimentary structures of the region. The data were collected using a high-resolution MCS re£ection system. A solid streamer of 48 or 24 channels with a 12.5 m channel interval was used to receive seismic signals, and the seismic source was an airgun array consisting of three Bolt airguns with a total volume of 475^775 in3, depending on the sizes of the air guns used on di¡erent cruises. The guns were ¢red every 10 s, which gives a nominal shot interval of 25 m. All the MCS data were processed at the Institute of Oceanogra-phy, National Taiwan University, using the ProMAX and SIOSEIS seismic data processing systems. The typical seismic data processing sequence comprised trace editing, spiking noise removal, water column mute, geometry set-up, amplitude compensation, predictive deconvolution, band-pass ¢ltering, velocity estimation, normal move-out correction, CMP stacking and F-K time migration.
ECHO CHARACTER CLASSIFICATION
AND MAPPING
Using the chirp sonar pro¢le data (Fig. 2), we have identi-¢ed and classiidenti-¢ed eight discrete echo types on the basis of re£ection characters (e.g. clarity, continuity, amplitude and geometry of bottom and sub bottom echoes). Our classi¢ -cation is modi¢ed from Damuth (1975, 1980a, b) and Yu & Lee (1992) to accommodate echo types observed in our study area. Four categories of echo types are recognized in this study: (1) Distinct echoes (Types I-1 and I-2), (2) Indistinct echoes (Types II-1 and II-2), (3) Hyperbolic
Fig. 3. Topographic gradient (slope) map of the study area. High gradients occur in the upper slope of the eastern portion of the passive SCS margin, and along submarine canyon walls and ridge £anks. SCS, South China Sea.
echoes (Type III) and (4) Irregular echoes (Types IV-1, IV-2 and IV-3) (Fig. 4).We describe each echo type in detail be-low, and Table 1 summarizes the description, occurrence and geological interpretation of each echo type.
Distinct echoes
Type I-1Echoes show a sharp sea£oor re£ection, with a few to no sub -bottom echoes (Fig. 4a). The sea£oor is generally £at or gently undulating, with a slightly irregular relief. Type
I-1 echoes occur mostly on the continental shelf (Figs 5^7 and 10) where coarse- grained, sandy deposits are common (Chen, 1983; Lee et al., 2004).
Type I-2
Echoes are characterized by distinct bottom echoes with continuous, parallel, generally conformable sub -bottom re£ections (Fig. 4b).The sea£oor is generally £at to undulat-ing or rollundulat-ing. Penetration ranges from 40 to 100 m. At a few locations, the continuous re£ections are interrupted by
Fig. 4. Chirp sonar characters of di¡erent echo types. (a) Distinct bottom echo with either no or a few sub -bottom re£ectors (type I-1) and indistinct bottom echo and bushy, prolonged sub-bottom echoes (type II-1). (b) Distinct bottom echo with continuous, parallel internal re£ectors (type I-2). (c) Wavy, prolonged sub-bottom echoes (type II-1). (d) Large single or irregular overlapping hyperbolae with widely varied vertex elevations (type III-2). (e) Several slope failure planes (SFP) within a short distance (type IV-1). (f) Some blanking zones interrupt the continuous re£ectors (type IV-2). (g) Sea £oor transparent blanketing layer. (type IV-3).
Ta b le 1. De sc ript io n an d in te rpre ta ti o n of ec h o ty p es T y p e L in e dra wing De sc ript io n O cc u rre nc e In te rpre ta ti o n I-1 S h ar p se a£ o or re£ ecto r w it ho ut or wi th a fe w sub -b otto m ech o es P as si v e an d act iv e co n ti n en tal sh elf C o ar se -g rain ed se di me n ts (C h en ,1 983 ; L ee et al ., 20 0 4 ) I-2 D ist in ct b o tt o m ec ho wi th co n ti nuo us , p arall el in te rn al re£ ecto rs ; the p en et ra tio n de pt h ra n g es fro m 4 0 to 10 0 m ; se a£ o o r is g en eral ly £a t In tr a -sl o p e b as in of th e K ao p ing S lop e and th e lo w er p ar t of th e p assi v e co n ti n enta l slo p e P elag ic or h emip el ag ic de p o si ti on (L ee et al ., 19 9 9 ); su bma rine fan II -1 In di st in ct b otto m ech o an d bus h y, p rol on ge d sub -b o tto m ec ho es U p per C h in a co n ti n en ta l sl o pe C re ep sed im ent s II -2 W av y, p rolo ng ed su b -b otto m ec h o es ; the p en et ra ti o n d epth va ri es fr o m 10 to 5 0 m T ro u g h so r d ep re ss io n so f th e Ka o p in g S lo p e Tu rb id it es III S ingle or ir re g ula r ov er lap p ing h y p erb ola e w it h w id el y v ar ie d v er te x el ev at io ns an d n o sub -b o tto m re £ ecto rs S u m m it s an d o utc ro p s of th e K ao p in g S lo p e an d sm all h ill s or ru g g ed are a o f the upp er C h in a con ti n en tal sl op e Ba se me n t h ig h or ou tcr op s (D amut h, 19 8 0a ), E sc ar p m en ts IV -1 S lo p e fa ilu re p lan es wi th in a sh o rt d is ta n ce ; do w n -slo p e slu mp s b y th e g ra v it y sl idin g St ee p slo p e an d g ull ie s o f th e S o uth C h in a S ea (SCS ) co nti n enta l m ar gi n and ca n y o n sy st em s o f the K ao p ing S lop e S lo p e fai lu re (L ee et al ., 20 02 , 2 0 05) IV -2 A co u st ic bl an ki ng zo n e b et w ee n co n ti nuo us re£ ecto rs ; th e wi d th of th e b la n k in g sig na ls v ar ie d fro m 2 0 0 m to se v eral ki lo me tr es . M u d d ir p ir zo n e o f the upp er Ka o p in g S lo p e G ass y sed im ent s (C h iu et al ., 20 06 ) IV -3 U p p er m o st tra ns pa re n t bl an ki ng la y er b el ow th e se a £o o r L o w er S C S co nti n en ta l slo p e an d arou nd th e lo w er se ct io n of th e P engh u C an y o n D eb ris £o w d ep o sits (B ry an t & R o em er ,1 98 3 )
faults or submarine mud volcanoes. Type I-2 echoes are mostly observed in the slope basins of the Kaoping slope and the lower part of the SCS continental slope (Figs 6^8 and 10). This echo type indicates persistent sedimentation.
Indistinct echoes
Type II-1Echoes are characterized by indistinct bottom echoes with discontinuous, prolonged sub-bottom echoes (Fig. 4a).
Fig. 5. Chirp sonar pro¢le OR1-719-1 (b) and corresponding seismic re£ection pro¢le (c). The inset (a) shows an enlarged chirp sonar image of the selected section. This pro¢le runs across the slope of the SCS passive margin (see Fig. 2 for pro¢le location). The interpreted echo types and their respective extents along this pro¢le are given above the chirp sonar pro¢le. BSR, Bottom Simulating Re£ectors; SCS, South China Sea.
Fig. 6. Chirp sonar pro¢le OR1-754 -26. The insets show enlarged chirp sonar images of selected sections. The large arrow indicates where the upper pro¢le connects to the lower pro¢le. See Fig. 2 for pro¢le location. The interpreted echo types and their respective extents along this pro¢le are given above the chirp sonar pro¢le.
Fig. 7. Chirp sonar pro¢le OR1-647-20. This pro¢le runs across the upper Kaoping slope (see Fig. 2 for pro¢le location). The inset shows the enlarged chirp sonar images of the sections indicated by the box where ¢ll-and- spill features can be clearly seen. The interpreted echo types and their respective extents along this pro¢le are given above the chirp sonar pro¢le.
They are mainly recorded from the SCS continental slope with a relatively smooth sea£oor (Figs 5, 6 and 10).
Type II-2
Echoes are characterized by indistinct to prolonged bottom echoes and continuous to discontinuous, prolonged sub-bottom echoes (Fig. 4c). Penetration varies from 10 to 50 m. These echoes are generally observed in the troughs or the depressions of the slopes on both passive and active mar-gins (Figs 9 and10) and are interpreted as turbidites (Lee etal., 2002).
Hyperbolic echoes
Type IIIEchoes are characterized by irregular single to overlapping hyperbolae with widely variable vertex elevations (Fig. 4d). This echo type shows no sub-bottom re£ectors and is observed on records from small hills and other areas of
rugged relief (Figs 7, 9 and10).Type III echoes are generally recorded from rugged basement highs or rock outcrops (Damuth, 1980a, b; Lee et al., 2002, 2005).
Irregular echoes
Type IV-1Echoes show distinct bottom echoes with truncated or vended sub -bottom re£ections beneath steeply dipping sea£oor and conformable sub-bottom re£ections beneath the adjacent £at sea£oor (Fig. 4e). They occur on steep slopes or along canyon walls (Figs 5, 8 and 9). In some cases, the unconsolidated sediments near the sea£oor were sheared and thinned along a slide plane.
Type IV-2
Echoes are characterized by the occurrence of acoustic blanking zones, which often interrupt the adjacent continu-ous re£ectors (Fig. 4f). Type IV-2 echoes are interpreted as representing gassy sediments (Fig. 7), and the acoustic
Fig. 8. Chirp sonar pro¢le OR1-647-2 (a) and the corresponding seismic re£ection pro¢le (d).The NW^SE-trending pro¢le lies near the mouth of the Kaoping River on the Kaoping slope (see Fig. 2 for pro¢le location). Parallel, continuous re£ectors that were bent by the growth of a structural ridge can be seen in inset (b). Inset (c) shows slope failure deposits of the Kaoping canyon (KPC) wall.The interpreted echo types and their respective extents along this pro¢le are given above the chirp sonar pro¢le. BR, Buried Ridge; PHC, Penghu canyon.
blanking is caused by an abrupt change of the re£ection coe⁄cient at the gas- ¢lled sediment boundary. The width of the blanking zone varies from 200 m to several kilometres.
Type IV-3
These consist of sharp bottom echoes with a thin transpar-ent or a non-re£ective layer on top of a zone of indistinct to prolonged sub-bottom echoes (Fig. 4g).This echo type has been interpreted to represent debris £ow deposits (Bryant & Roemer, 1983).
Echo character map
A map showing the distribution of the di¡erent echo types was compiled to help de¢ne the sedimentary processes in the di¡erent tectonic and depositional environments in the study area (Fig. 10). For example, type I-1 echoes are distributed on the shelf areas of both the passive continen-tal margin and the orogenic wedge provinces (Figs 5, 6 and 10), re£ecting the sandy sediments that dominate shelf deposits. Type III echoes are limited to ridge summits and their steep £anks (Figs 7, 9 and 10). They re£ect hard sea£oor outcrops on ridge crests. Type IV-1 echoes are observed along the paths of submarine canyons and
Fig. 9. Chirp sonar pro¢le OR1-716 - 4 (c) and the corresponding seismic re£ection pro¢le (d). The NE^SW-pro¢le goes from the Kaoping canyon to the lower Kaoping slope (see Fig. 2 for pro¢le location). Echo characters of the slope basin deposits can be seen in the enlarged chirp sonar images in inset (a, b).The interpreted echo types and their respective extents along this pro¢le are given above the chirp sonar pro¢le.
channels (Figs 5, 8 and 9), indicating frequent slope failures in steep canyon walls. Type IV-2 echoes appear in mud diapir zones where gas- charged sediments and mud volcanoes are observed.
Besides the echo types associated with speci¢c sedi-mentary processes mentioned above, three main types of echoes are observed in the passive SCS continental slope: types I-2, II-1 and III. Type III echoes are predominant in the continental slope area bordering the orogenic wedge (Figs 5 and 10), but where the continental slope is further away from the Taiwan orogenic wedge, type III echoes occur only in a mid- slope zone (Figs 6 and 10).This distribution pattern correlates well with the sedimentary pro -cesses described in the slope failure section below, and many escarpments were observed on the chirp sonar pro¢les where type III echoes occur. Types II-1 and I-2 echoes are recorded in the western half of the SCS conti-nental margin in the study area (Fig. 10). Type II-1 echoes, which represent mass-transport deposits, most likely creep, occur in the upper slope domain (Figs 5, 6 and 10), whereas type I-2 echoes in the lower slope domain re£ect pelagic or hemipelagic deposition (Figs 5 and 10). The distribution pattern of echo type I-2 implies that the lower continental slope is a stable depositional environment.
In the Taiwan orogenic wedge province, the distribut-ion of echo types is more complex (Fig. 10). Apart from the type I-1 echoes along the narrow Kaoping shelf and the echo characters associated with submarine canyons (type IV-1), mud volcanoes (type IV-2) and hard ridge tops (type III), the dominant echoes in the upper Kaoping slope domain are type I-2, whereas the type III echoes are widely distributed in the lower Kaoping slope (Figs 7 and 8^10). This distribution pattern contrasts with the passive SCS continental margin where type III echoes are observed on the upper and mid- continental slope
(Figs 5, 6 and 10), whereas the type I-2 echoes are mostly observed on the lower continental slope (Figs 7, 8, and 10). Type II-2 echoes are also found in some of the slope basins on the lower Kaoping slope (Figs 9 and 10) where the sediments are mostly turbidites. Echos of type IV-3 are observed on either side of the deformation front where the Penghu submarine canyon emerges from the fold and thrust zone of the lower Kaoping slope (see Figs 1 and 10).
SEDIMENTARY PROCESSES
Slope failures in the passive margin
Submarine slope failures are observed throughout the passive SCS continental margin. Canyons and gullies formed by mass-wasting processes dissect the upper SCS slope, and generally trend down slope perpendicular to the isobaths. Intercanyon areas form a series of parallel slope ridges (Fig. 3). Slope failures that transport sediments from the shelf and upper slope to deeper environments are key sediment transport processes in the passive SCS continental margin.
Figure 5 presents a dip - oriented chirp sonar and seismic pro¢le across the passive SCS continental margin. A strong Bottom Simulating Re£ector (BSR) is clearly visible on the seismic pro¢le, about 300 ms below the sea£oor in the lower part of the continental slope on the seismic pro¢le (Fig. 5c). Slope failures are evident as steep cli¡s at the shelf break and at several places along the pro¢le down to 1700 m water depth. The heights of the cli¡s that could be the head scarps or sidewalls of slope failures vary from 10 to over 100 m. Rugged sea£oor, where numerous scars and slump deposits lie, presents as small irregular hy-perbolic echoes (type III echoes) on the chirp sonar pro¢le.
West of 1191240E, most slope failures occur on the mid-dle to lower part of the SCS continental slope.The sea£oor is smooth and the gradient is gentle on the upper conti-nental slope (Fig. 6), whereas numerous slope failures are observed below 750 m water depth. A thin blanketing layer with a transparent zone is observed on the lower part of the upper slope. A small slope failure is observed at about 400 m water depth on the up - slope side of this thin layer; it may represent slump/debris £ow deposits related to the failure. Compared with the pro¢le farther east (Fig. 5), fewer slope failures are identi¢ed on this portion of the slope (Fig. 6), and most of the slope failures occur on the mid to lower slope.
Di¡erences in the occurrence of slope failures and sea£oor gradients from NE (nearTaiwan) to SW (away from Taiwan) along the passive SCS continental margin may re-£ect the in£uence of the impingingTaiwan orogenic wedge on the passive SCS continental margin. We suggest that loading by the orogenic wedge has steepened the conti-nental slope to the east of about 1191240E, promoting slope failure and canyon- cutting processes along the shelf edge and in the upper slope area.West of about 1191240E, where the SCS continental margin is no longer a¡ected by the
Fig. 10. Distribution of echo types o¡shore of SW Taiwan. See text for interpretation.Thick lines give the locations of the chirp sonar pro¢les shown in Figs 5^9.
Taiwan orogenic wedge, slope failures are mostly located in the lower slope area. In fact, in the SCS continental slope away from theTaiwan orogenic wedge, an apparent change of the slope gradient is observed around the 2000 m iso bath.This boundary separates the steeper upper slope do -main from the gentler lower slope do-main. Mass-wasting processes are mostly erosive in the upper continental slope, as a few or only thin deposits are observed there. In contrast, both chaotic and discontinuous seismic facies interpreted as slump/debris- £ow deposits and turbidite deposits with continuous, parallel internal re£ectors are observed in the lower continental slope domain where BSRs (e.g. Fig. 5) are frequent (Liu et al., 2006).
Fill-and-spill processes in the orogenic wedge
Active tectonics are a major control on the sedimentary processes in the submarineTaiwan orogenic wedge. Three large submarine canyons run across the Kaoping shelf and slope (Fig. 1). From east to west, they are the Fangliao can-yon (Yu & Lu, 1995), the Kaoping cancan-yon (Liu et al., 1993) and the Penghu canyon system (Chuang & Yu, 2002). These submarine canyons have conveyed large amounts of sediments from theTaiwan mountain belt to the deep SCS basin. However, structural lows between imbricated folds and thrusts and diapiric ridges have formed intraslope basins and sediment depocentres on the orogenic wedge. Similar to the intraslope basins of the northern Gulf of Mexico,‘¢ll-and- spill’depositional processes (e.g. Prather et al., 1998; Toniolo et al., 2006a, b) have dominated sedi-ment dispersal during slope developsedi-ment (Chiang et al., 2004; Yu & Huang, 2006).Basin ¢lls are characterized by a distinct upward change in seismic facies beginning with a basal, convergent-baselapping facies, succeeded by chaotic facies and topped by parallel and draping facies (Figs 8 and 9). Sediments from the prograding shelf and upper slope are transported and deposited mainly in the con¢ned basal accom-modation space of intraslope basins, forming convergent-baselapping seismic facies in the early stage of basin ¢ll. In the later stages of basin development, sediments have been transported to a lower slope distal to the sediment source, progressively ¢lling and spilling intraslope basins in a generally southwestward direction (Yu & Huang, 2006). These elongated basins parallel the structural trend of the orogenic wedge, and have a high length-to-width ratio.
The ‘¢ll-and- spill’ process can be observed on the chirp sonar pro¢les. Figure 7 shows a chirp sonar pro¢le that runs from the Kaoping shelf to the upper Kaoping slope. Two ridges (labelled A and B) can be identi¢ed in the southwestern portion of the pro¢le. Ridge A has acted as a sill that controls the transport direction of the sedi-ments. In£ux of sediment from the Taiwan mountain belt into this region via the Kaoping River is extremely high (49 MT/year; Dadson et al., 2003). These sediments have ¢lled the intraslope basin on the NE side of Ridge A, and sediments are now spilling over this ridge and into the
in-traslope basin between Ridges A and B (Fig. 7). Slope fail-ures have occurred on the £anks of Ridge B and side walls of the submarine canyon between ridges A and B. Some uppermost transparent layers (type IV-3 echoes) appear to be debris £ow deposits, suggesting that the displaced slope sediments travelled relatively short distances.
Along- slope variations of sea£oor sedimentary and morphological features are revealed in Fig. 8, where the chirp and seismic pro¢les show ridges, canyons (Penghu canyon and Kaoping canyon) and an intraslope basin. The seismic image shows that the submarine canyons are bounded by anticlinal ridges, and there is a buried ridge underneath the intraslope basin ¢ll. The thickness of the intraslope basin sediment is more than 2 s (two -way travel time) on either side of the buried ridge (Fig. 8). On the chirp sonar image, the type I-2 echoes with continuous, parallel internal re£ections of the sea£oor sediment in the intraslope basin are more than 50 m deep (Fig. 8). Some of the re£ections are deformed by the £anks of the folded ridges, revealing active tectonism in this region.
There are at least four NW^SE-trending intraslope basins between the Kaoping canyon and the Penghu canyon on the lower Kaoping slope. Sediment transport depends on the submarine canyon and channel systems connecting these basins. Figure 9 presents a chirp sonar pro¢le with the corresponding seismic image of the lower slope domain. Two slope basins near the Kaoping canyon have already been ¢lled with sediment. The third intra-slope basin shown at the SW end of the pro¢le (Fig. 9) has an active channel at the foot of a ridge, and the sediment in this basin is much thicker than in the two shallower intraslope basins.
Comparison of the morphology of the canyons at either end of the pro¢le in Fig. 9 shows that the Kaoping canyon has a steep V- shaped cross-pro¢le, whereas the small unnamed submarine canyon in the SW is much wider and shallower.There are slump deposits in the Kaoping canyon observed on the chirp sonar pro¢le, but the Kaoping canyon is mainly erosive at the present time (Fig. 9). On the other hand, there are thick sediments beneath the £oor of the small submarine canyon, which implies that part of the sediments transported by turbidity currents along submar-ine canyons are deposited in the distal slope basins.
DISCUSSION
The effect of sediment supply
Sediment supply has a major impact on deposition. Sediment sources are quite di¡erent in the two tectonic provinces o¡shore of SW Taiwan. On the passive SCS continental margin, sediments are mainly terrestrial and biological in origin (Yen & Lundberg, 2006), and the amount of sediment transported into this region is much less than the supply to the submerged Taiwan orogenic wedge. Although the sediments of the passive SCS
conti-nental margin mainly come from China, Liu et al. (2008) have shown that a signi¢cant portion of the detrital sedi-ment on the SCS continental margin actually came from Taiwan, as similar clay mineral assemblages were observed from the Taiwan orogenic wedge to the northeastern SCS continental margin.This suggests that deep currents have carried re- suspended ¢ne sediments from Taiwan into the northern SCS continental margin, similar to what Ludmann et al. (2005) suggested for the SCS slope of the Dongsha Island region located to the west of the study area. However, sediment waves indicative of these deep bottom currents (Zhong et al., 2007) were not observed in our study area, and downslope processes appear to domi-nate the SCS continental slope.
In contrast to the SCS passive margin, rapid tectonic uplift and high erosion rates have led to large volumes of sediments from the Taiwan mountain belt being trans-ported into the submarine orogenic wedge region o¡ southern Taiwan since Pliocene times (Covey, 1984; Chiang et al., 2004). Submarine canyons have provided convenient pathways for part of the Taiwan- derived sediments to the deep ocean, whereas the rest of the terrigenous sediments have ¢lled the Kaoping slope basins. This has smoothed the sea£oor of the upper Kaoping slope, except where submarine canyons are present. The equilibrium pro¢le of the continental slope is apparent in the upper slope do -main of the orogenic wedge. On the other hand, in the SCS continental margin province, due to limited sediment supply, the head scarps of slope failures persist for a long period of time and the slope gradient will increase locally due to slope failure. The gradient map (Fig. 3) shows that the passive SCS margin has a steeper upper slope, whereas the orogenic wedge has a steeper lower slope.
The echo character map (Fig. 10) re£ects the gradient changes. The upper Kaoping slope is dominated by con-tinuous and parallel re£ections (echo type I-2), especially in the submarine fan that extends from the mouth of the Kaoping River. The sea£oor echo characters of the lower Kaoping slope reveal abundant slope failures on the £anks of anticlinal ridges with irregular echoes (type III echoes). The passive SCS continental slope also presents a distinct zonal distribution of erosion and deposition as revealed on the echo character map, but the pattern is opposite to that in the Kaoping slope region: slope failure is the dominant process in the upper SCS continental slope whereas turbi-dite deposition has shaped on the lower SCS continental slope. Damuth (1980a) placed the boundary separating his echo types 3A and 1B at about the 2000 m isobath. It coin-cides with the boundary we have found here.
Evolution of retrogressive failures
Along the passive continental slope, the occurrence of echo type III diminishes southwestward (Fig. 10). This suggests that the occurrence of slope failures may follow a trend. Sultan et al. (2004) proposed a model of retrogres-sive failure of the continental slope when sea level drops,
but the tectonics of the SCS continental margin in the study area might be an alternative driver of retrogressive failures. The northeastern corner of the passive SCS continental margin is a¡ected by loading of the Taiwan orogenic wedge (Lin & Watts, 2002). Extensional crustal deformation has occurred along the continental margin due to lithospheric bending. In this location, the SCS con-tinental slope has steepened here due to the impinging orogenic wedge, and this may have facilitated initial slope failure. Once the initial failure has occurred, the tempera-ture and pressure of the upper part of the sediments will change, leading to a second slope failure. As a conse-quence, retrogressive slope failures occur progressively, moving their head wall towards the upper part of the slope.
Gas hydrate dissociation effect
Widely distributed BSR have been observed beneath the sea£oor in the o¡shore area of SW Taiwan (Chi et al., 1998; Liu et al., 2006); very high methane concentrations in the bottom sea water and in the pore £uid of the sea£oor sedi-ment have also been measured at many coring sites in the study area (Chuang et al., 2006), suggesting that gas hydrates are present beneath the sea£oor. When the climate warms (e.g. in Late Quaternary), gas hydrates in the stability zone could dissociate and produce free gas in the sediment pore space, reducing the strength of the sediments on the upper slope and increasing the probabil-ity of slope failures. McAdoo et al. (2000) showed that most of the submarine landslides on the passive US continental margin occurred between the 1000 and 2000 m isobaths. However, the headwalls of slumps and slides observed in our study area are much shallower. The locations of the slide headwalls are mostly at water depths around 500 m in the SCS continental margin. This could be explained by reduction of shear strength of sediments due to gas hydrate dissociation at shallower depths within the SCS passive continental margin. However, a direct link between slope failures and gas hydrate dissociation (i.e. seeing the gas) is yet to be observed.
In contrast, in the orogenic wedge province, slope fail-ures that appear most likely to be triggered by gas hydrate dissociation are limited to canyons’ walls or steep £anks of some ridges. Although extremely high methane concen-trations have been detected in the bottom sea water and in cored sediments o¡shore SW Taiwan (Chuang et al., 2006; Yang et al., 2006), and gassy sediments are widely distribu-ted in the study area (Chiu et al., 2006), the slope basins separated by NW^SEtrending ridges in the upper Kaop -ing slope have been ¢lled with sediments to form a gently dipping, smooth sea£oor that is not prone to slope failure. Therefore, the e¡ects of gas hydrate dissociation on sedi-mentary processes in passive and in active margins are somewhat di¡erent.
CONCLUSIONS
A detailed analysis of chirp sonar data o¡shore SW Taiwan reveals distinctive echo types that correlate well with local sedimentary features. Eight echo types have been identi-¢ed based on their clarity, continuity, depth of penetration, bottom and sub-bottom echo geometry, etc. These echo types are grouped into four categories in this study: Type I echoes have distinct sea£oor echoes that may or may not show strati¢ed sub -bottom re£ectors depending on the grain size of the sea£oor sediments; Type II echoes are in-distinct echoes that are interpreted to be formed by creep sediments or turbidite deposits; Type III echoes have hy-perbolic shapes that re£ect hard, irregular or very rough sea£oor topography; and Type IV echoes are irregular echoes that represent special sedimentary features such as slope failures, gassy sediments and debris £ow deposits. An echo character map that we compiled in this study reveals di¡erent sedimentary processes in the tectonic and depositional environments of the study area. Echo type I-1 occurs on the passive SCS continental shelf and the Kaoping shelf, suggesting that the shelf deposits are mostly coarse-grained sands and silts. Echo type I-2 is observed in the lower SCS continental slope and in the upper Kaoping slope, indicating deposition in those areas. Type II-1 echoes, which represent creep sediments, are observed in the western part of the upper slope domain of the SCS continental margin in our study area, whereas type II2 echoes, which represent turbidite deposits, ap -pear in the slope basins of the lower Kaoping slope. Type III echoes that suggest either rock outcrops, or very rough sea£oor topography, are observed on top of several sub -marine ridges, and on the upper and mid- slope area of the SCS continental margin and the lower Kaoping slope of theTaiwan orogenic wedge.Type IV-1 echoes that repre-sent slope failures appear along submarine canyons and their tributaries. Type IV-2 echoes that may be associated with gases in the sediments are distributed where submar-ine mud volcanoes have been observed. Type IV-3 echoes appear along either side of the lower section of the Penghu canyon, indicating the location of massive debris £ow de-posits there. There is a major di¡erence between the sedi-mentary environments of the passive SCS continental margin and the submarine orogenic wedge o¡shore SW Taiwan. The echo distribution map clearly indicates that in the passive SCS continental margin province, the sedimentary environment in the upper to mid- slope is erosional in nature, dominated by downslope processes, whereas the lower continental slope of the SCS continen-tal margin is depositional in nature. On the other hand, in the active submarine Taiwan orogenic wedge, the upper slope domain appears to be a depositional environment, except in submarine canyons, whereas the lower slope do -main is erosional in nature, except in slope basins.
The main reason for this huge contrast is the supply of terrigenous sediment. Huge amounts of sediment have been carried into the submarine Taiwan orogenic wedge
o¡shore of SW Taiwan, and ¢ll-and- spill processes have smoothed the sea£oor topography of the upper Kaoping slope. In the lower Kaoping slope, spilled sediments have ¢lled only part of the slope basins, and the steep slopes of the ridge £anks and submarine canyon walls generate fre-quent slope failures. The passive SCS continental margin receives considerably less sediment from either China or Taiwan. There, the steep slope of the upper continental slope forms an erosional environment whereas mass-wast-ing materials are deposited in the lower continental slope and continental rise, where the sea£oor slope is much gen-tler. The downslope processes are especially active in the eastern half of the SCS continental margin adjacent to the Taiwan orogenic wedge. We suggest that this is due to the e¡ect of orogenic wedge loading, causing extensional structural deformation along this part of the SCS conti-nental margin. The bending and faulting of the continen-tal slope due to orogenic wedge loading may trigger retrogressive failures on the upper continental slope. Away from the orogenic wedge, fewer slope failures have occurred in the SCS continental upper slope domain. Disso -ciation of gas hydrates that are widely distributed in both the SCS passive continental margin and the active oro -genic wedge in the study area could aid the development of slope failures, but their e¡ects must be local, if they ex-ist, as no direct link between gas hydrate dissociation and massive landslides was observed.
ACKNOWLEDGEMENTS
We would like to thank the Central Geological Survey, Ministry of Economic A¡airs, ROC. for organizing and supporting the gas hydrate investigation programme that made this study possible.The chirp sonar and seismic data used in this study were collected by the R/V Ocean Re-searcher 1 of the Institute of Oceanography, National Tai-wan University, with excellent technical support from S. D. Chiu. Philippe Schnurle processed the seismic data, and K. R. Lai and S. Y. Liu helped in preparing the chirp sonar data. Swath bathymetry data were collected by the R/ VAtalante during the ACT cruise led by S. Lallemand and C. S. Liu. Critical comments and suggestions made by J. E. Damuth, N. Hovius and J. Malavieille signi¢cantly im-proved this paper. This study is supported by the Central Geological Survey, Ministry of Economic A¡airs, ROC under grants 5226902000 - 05-94 - 01.
REFERENCE
Brya nt,W.R. & Roemer, L.B. (1983) Structure of the continen-tal shelf and slope of the northern gulf of Mexico and its geohazards and engineering constrains. In: CRC of Handbook of Geophysical Exploration at Sea (Ed. by R.A. Geyer), pp. 123^ 218. CRC press, Boca Raton, FL.
Casas, D., Ercilla, G., Baraza, J., Alonso, B. & Maldonada, A. (2003) Recent mass-movement processes on the Ebro continental slope (NW Mediterranean). Mar. Petrol. Geol., 20, 445^457.
C h e n, M.P. (1983) Physical properties and depositional environ-ments of continental slope sedienviron-ments, southern Taiwan Strait. Acta Oceanogr.Taiwanica, 14, 42^63.
C h i,W.C., Reed, D.L., Liu, C.S. & Lundberg, N. (1998) Distri-bution of the bottom- simulating re£ector in the o¡shore Tai-wan collision zone. Terr. Atmos. Ocean. Sci., 9, 779^794. C h i a n g, C.S., Yu, H.S. & Chou, Y.W. (2004) Characteristics of
wedge-top depozone of the southern Taiwan foreland basin. Basin Res., 16, 65^78.
C h i u, J.K.,Tseng,W.H. & Liu, C.S. (2006) Distribution of gassy sediments and mud volcanoes o¡shore southwest Taiwan. Terr. Atmos. Ocean. Sci., 17, 703^722.
C h oug h, S.K., Kim, J.W., Lee, S.H., Shinn,Y.J., Jin, J.H., Suh, M.C .& Lee, J.S. (2002) High-resolution acoustic characteris-tics of epicontinental sea deposits, central- eastern Yellow Sea. Mar. Geol., 188, 317^331.
C h ua n g, C.Y. & Yu, H.S. (2002) Morphology and canyon form-ing processes of upper reach of the Penghu Submarine Canyon o¡ southwestern Taiwan. Terr. Atmos. Ocean. Sci., 13, 91^108. C h ua n g, P.C., Yang, T.F., Lin, S., Lee, H.F., Lan, T.F.F.,
Ho n g, W.L., Liu, C.S., Chen, J.C. & Wang, Y. (2006) Extremly high methane concentration in bottom water and cored sediments from o¡shore southwestern Taiwan. Terr. Atmos. Ocean. Sci., 17, 903^920.
C ovey, M. (1984) Lithofacies analysis and basin reconstruction, Plio -Pleistocene western Taiwan foredeep. Petrol. Geol. Taiwan, 20, 53^83.
Da d s o n, S.J., Hovius, N., Chen, H., Dade,W.B., Hiseh, M.L., Wi llett, S.D., Hu, J.C., Horng, M.J., Chen, M.C., Stark, C .P., Lague, D. & Lin, J.C. (2003) Links between erosion, runo¡ variability and seismicity in the Taiwan orogen. Nature, 426, 648^651.
Da m uth, J.E. (1975) Echo character of western equatorial Atlan-tic £oor and its relationship to dispersal and distribution of terrigenous sediments. Mar. Geol., 18, 17^ 45.
Da m uth, J.E. (1979) Migrating sediment waves created by tur-bidity currents in the northern South China Basin. Geology, 7, 520^523.
Da m uth, J.E. (1980a) Quaternary sedimentation processes in the South China Basin as revealed by echo - character mapping and piston- core studies. In:TheTectonic and Geologic Evolution of Southeast Asian Seas and Islands (Ed. by D.E. Hayes), Amer. Geo-phys. Union GeoGeo-phys. Monogr., 23, 105^126.
Da m uth, J.E. (1980b) Use of highfrequency (3.5^12 kHz) echo -grams in the study of near-bottom sedimentation processes in the deep sea: a review. Mar. Geol., 38, 51^75.
Da m uth, J.E. & Hayes, D.E. (1977) Echo character of east Brazi-lian continental-margin and its relationship to sedimentary processes. Mar. Geol., 24, 73^95.
Damuth, J.E., Jacobi, R.D. & Hayes, D.E. (1983) Sedimentation processes in the northwest Paci¢c Basin revealed by echo-charac-ter mapping studies. Geol. Soc. Am. Bull., 94, 381^395.
Gau lli e r,V. & Bellaiche, G. (1998) Near-bottom sedimenta-tion processes revealed by echo - character mapping studies, northwestern Mediterranean Basin. AAPG Bull., 82, 1140^1155. Hovi us, N., Stark, C.P., Chu, H.T. & Lin, J.C. (2000) Supply and removal of sediment in a landslide- dominated mountain belt: Central Range, Taiwan. J. Geol., 108, 73^89.
Hua n g, C.Y.,Wu,W.Y., Chang, C.P., Tsao, S., Yuan, P.B., Lin, C .W.& Xia, K.Y. (1999) Tectonic evolution of accretionary prism in the arc- continent collision terrane of Taiwan.Tectono-physics, 281, 31^51.
Lalle man d, S.E., Liu, C.S. & The ACT Scientific Crew (1997) Swath bathymetry mapping reveals details of the active arc-continent collision o¡shoreTaiwan. EOS,Trans. AGU, 78, 173^175. Le e, S.H., Bahk, J.J., Chough, S.K., Back, G.G. & Yoo, H.S. (2004) Late quaternary sedimentation in the Ulleung Iner-plain Gap, East Sea (Korea). Mar. Geol., 206, 225^248. Le e, S.H., Chough, S.K., Back, G.G. & Kim, Y.B. (2002) Chirp
(2^7 kHz) echo characters of the South Korea Plateau, East Sea: style of mass movement and sediment and sediment gravity £ow. Mar. Geol., 184, 227^247.
Le e, S.H., Chough, S.K., Back, G.G., Kim, Y.B. & Sung, B.S. (1999) Gradual downslope change in high-resolution acoustic characters and geometry of large- scale submarine debris lobes in Ulleung Basin, East Sea (Sea of Japan), Korea. Geo-Mar. Lett., 19, 254^261.
Le e, T.G., Hein, J.R., Lee, K., Moon, J.W. & Ko, Y.T. (2005) Sub - sea£oor acoustic characterization of seamounts near the Ogasawara Fracture Zone in the western Paci¢c using chirp (3^7 kHz) sub -bottom pro¢les. Deep-Sea Res. I, 52, 1932^1956. Li n, A.T., Liu, C.S., Lin, C.C., Schnurle, P., Chen, G.Y.,
Li ao, W.Z., Teng, L.S., Chuang, H.J. & Wu, M.S. Tectonic features associated with the overriding of an accretionary wedge on top of a rifted continental margin: an example from Taiwan. Mar. Geol., (accepted).
Li n, A.T. & Watts, A.B. (2002) Origin of the West Taiwan basin by orogenic loading and £exure of a rifted continental margin. Jour. Geophys. Res., 107, B9, 2185, doi: 10.1029/2001JB000669. Li u, C.S. (2007) Geological exploration of gas hydrate resources
o¡shore southwestern Taiwan ^ Geophysical investigation (4/ 4). Central Geological Survey Report 96 -26 Summary, 176pp. (in Chinese with English abstract)
Li u, C.S., Deffontaines, B., Lu, C.Y. & Lallemand, S. (2004) Deformation patterns of an orogenic wedge in the transition zone from subduction to collision o¡shore southwestern Tai-wan. Mar. Geophys. Res., 25, 123^137.
Li u, C.S., Huang, I.L. & Teng, L.S. (1997) Structure features o¡ southwestern Taiwan. Mar. Geol., 137, 305^319.
Li u, C.S., Liu, S.Y., Lallemand, S.E., Lundberg, N. & Reed, D.(1998) Digital elevation model o¡shore Taiwan and its tec-tonic implications. Terr. Atmos. Ocean. Sci., 9, 705^738.
Li u, C.S., Lundberg, N., Reed, D. & Huang, Y.L. (1993) Mor-phological and seismic characteristics of the Kaoping Submar-ine Canyon. Mar. Geol., 111, 93^108.
Liu, C.S., SchnYºrle, P., Wang, Y., Chung, S.H., Chen, S.C. & Hsiuan, T.H. (2006) Distribution and characters of gas hydrate o¡shore of southwesternTaiwan.Terr. Atmos. Ocean. Sci., 17, 615^644. Li u, Z., Tuo, S., Colin, C., Liu, J.T., Huang, C.-Y., Selvaraj, K., Chen, C.-T.A., Zhao, Y., Siringan, F.P., Boulay, S. & C h e n, Z. (2008) Evaluation detrital sediment contribution of Taiwan to the northern South China Sea. Mar. Geol., doi: 10.1016/j.margeo.2008.08.003.
Lu d m a n n, T., Wong, H.K. & Berglar, K. (2005) Upward £ow of North Paci¢c Deep Water in the northern South China Sea as deduced from the occurrence of drift sediments. Geophys. Res. Lett., 3, L05614, doi: 10.1029/2004GL021967.
MacDonald, D.I.M. (1993) Controls on sedimentation at conver-gent plate margins. Spec. Publs Int. Ass. Sediment., 20, 225^257. McAd o o, B.G., Pratson, L.F. & Orange, D.L. (2000)
Submar-ine landslide geomorphology, US contSubmar-inental slope. Mar. Geol., 169, 103^136.
Mienert, J. & Weaver, P, Eds. (2003) European Margin Sediment Dy-namics. Springer-Verlag, Berlin, 309pp.
O rpi n, A.R. (2004) Holocene sediment deposition on the pov-erty- slope margin by the muddy Waipaoa River, East Coast New Zealand. Mar. Geol., 209, 69^90.
Prath e r, B.E., Booth, J.R., Steffens, G.S. & Craig, P.G. (1998) Classi¢cation, lithologic calibration, and stratigraphic succession of seismic facies of intraslope basins, deep -water Gulf of Mexico. AAPG Bull., 82, 701^728.
R e e d, D.L., Lundberg, N., Liu, C.S. & Kuo, B.Y. (1992) Struc-tural relations along the margins of the o¡shoreTaiwan accre-tionary wedge ; implication for accretion and crustal kinematics. Acta Geol.Taiwan., 30, 105^122.
Sh i n n,Y.J., Chough, S.K., Kim, J.W. & Wpp, J. (2007) Develop -ment of depositional systems in the southeastern Yellow Sea during the postglacial transgression. Mar. Geol., 239, 59^82. Su lta n, N., Cochonat, P., Foucher, J.P. & Mienert, J. (2004)
E¡ect of gas hydrate melting on sea£oor slope instability. Mar. Geol., 213, 379^ 401.
Su n, S.I. & Liu, C.S. (1993) Mud diapirs and submarine channel deposits in o¡shore Kaohsiung-Hengchun, southwest Taiwan. Pet. Geol.Taiwan, 28, 1^14.
Te n g, L.S. (1990) Geotectonic evolution of late Cenozoic arc-continent collision in Taiwan. Tectonophysics, 183, 57^76. To n io lo, H., Lamb, M. & Parker, G. (2006a) Depositional
tur-bidity currents in diapiric minibasins on the continental slope: formulation and theory. J. Sediment. Res., 76, 783^797. To n io lo, H., Parker, G.,Voller,V. & Beauboeuf, R. (2006b)
Depositional turbidity currents in diapiric minibasins on the continental slope: experiments ^ numerical simulation and upscaling. J. Sediment. Res., 76, 798^818.
We s s e l, P. & Smith, W.H.F. (1993) The GMT- system version 2.1.4. Technical reference and Cookbook, SOEST/NOAA. Wh itm o re, G.P., Crook, K.A.W. & Johnson, D.P. (1999)
Sedi-mentation in a complex convergent margin: the Papua New
Guinea collision zone of the western Solomon Sea. Mar. Geol., 157, 19^ 45.
Ya n g, T.F., Chuang, P.C., Lin, S., Chen, J.C., Wang, Y. & C h u n g, S.H. (2006) Methane venting in gas hydrate potential area o¡shore of SW Taiwan: evidence of gas analysis of water column samples. Terr. Atmos. Ocean. Sci., 17, 933^950.
Ye n, J.Y. & Lundberg, N. (2006) Sediment compositions in o¡shore southern Taiwan and their relations to the source rocks in modern arc-continent collision zone. Mar. Geol., 225, 247^263.
Yu, H.S. & Huang, Z.Y. (2006) Intraslope basin facies and sedi-mentary processes in the Kaoping slope, o¡shore southwes-tern Taiwan. Terr. Atmos. Ocean. Sci., 17, 659^677.
Yu, H.S. & Lee, J.T. (1992) Echo character of the shelf-upper slope transition o¡ southwestern Taiwan margin and it’s relationship to Quaternary sedimentation. Acta Oceangr.Taiwan., 29, 45^58. Yu, H.S. & Lu, J.C. (1995) Development of the shale diaper-
con-trolled fangliao canyon on the continental- slope o¡ southwes-tern Taiwan. J. Southeast Asian Earth Sci., 11, 265^276.
Yu, H.S. & Song, G.S. (2000) Submarine physiographic features in Taiwan region and their geological signi¢cance. J. Geol. Soc. China, 43, 267^286.
Yu, H.S. & Wen, Y.H. (1992) Physiographic characteristics of the continental margin o¡ southwestern Taiwan. J. Geol. Soc. China, 35, 337^351.
Zh o n g, G., Li, Q., Hao, H. & Wang, L. (2007) Current status of deep -water sediment wave studies and the South China Sea perspectives. Adv. Earth Sci., 22, 907^913 (in Chinese with Eng-lish abstract).
Manuscript received 17 June 2008; Manuscript accepted 9 September 2008
