Anatomy and genesis of a subduction-related …ea.c.u-tokyo.ac.jp/earth/Members/Isozaki/96Anatomy.pdfAnatomy and evolution of Japanese Islands 291 /II \ N. American Piate Fig. 1 Plate

$Page 1: Anatomy and genesis of a subduction-related …ea.c.u-tokyo.ac.jp/earth/Members/Isozaki/96Anatomy.pdfAnatomy and evolution of Japanese Islands 291 /II \ N. American Piate Fig. 1 Plate$
The Islaizd Arc (1996) 5, 289-320

Thematic Article Anatomy and genesis of a subduction-related orogen:

A new view of geotectonic subdivision and evolution of the Japanese Islands

YUKIO ISOZAKI Department of E a r t h and Planetary Sciences, Tokyo Institute of Technology, 0-okayama,

Megur.0, Tokyo 152, Japari

Abstract The Japanese Islands represent a segment of a 450 million year old subduction-related orogen developed along the western Pacific convergent margin. The geotectonic subdivision of the Japanese Islands is newly revised on the basis of recent progress in the 1980s utilizing microfossil and chronometric mapping methods for ancient accretionary complexes and their high-P/T metamorphic equivalents. This new subdivision is based on accretion tectonics, and it contrasts strikingly with previous schemes based on ‘geosyncline’ tectonics, continent-continent collision-related tectonics, or terrane tectonics. Most of the geotectonic units in Japan are composed of Late Paleozoic to Cenozoic accretionary complexes and their high-PIT metamorphic equivalents, except for two units representing fragments of Precambrian cratons, which were detached from mainland Asia in the Tertiary. These ancient accretionary complexes are identified using the method of oceanic plate stratigraphy. The Japanese Islands are comprised of 12 geotectonic units, all noted in southwest Japan, five of which have along-arc equivalents in the Ryukyus. Northeast Japan has nine of these 12 geotectonic units, and East Hokkaido has three of these units. Recent field observations have shown that most of the primary geotectonic boundaries are demarcated by low-angle faults, and sometimes modified by secondary vertical normal and/or strike-slip faults. On the basis of these new observations, the tectonic evolution of the Japanese Islands is summarized in the following stages: (i) birth at a rifted Yangtze continental margin at ca 750-700 Ma; (ii) tectonic inversion from passive margin to active margin around 500 Ma; (iii) successive oceanic subduction beginning at 450 Ma and continuing to the present time; and (iv) isolation from mainland Asia by back-arc spreading a t ca 20 Ma. In addition, a continent-continent collision occurred between the Yangtze and Sino- Korean cratons a t 250 Ma during stage three. Five characteristic features of the 450 Ma subduction-related orogen are newly recognized here: (i) step-wise (not steady-state) growth of ancient accretionary complexes; (ii) subhorizontal piled nappe structure; (iii) tectonically downward-younging polarity; (iv) intermittent exhumation of high-P/T metamorphosed accretionary complex; and (v) microplate-induced modification. These features suggest that the subduction-related orogenic growth in Japan resulted from highly episodic processes. The episodic exhumation of high-P/T units and the formation of associated granitic batholith (i.e. formation of paired metamorphic belts) occurred approximately every 100 million years, and the timing of such orogenic culmination apparently coincides with episodic ridge subduction beneath Asia.

Key words: accretionary complex, Japan, microfossil mapping, microplate modification, oceanic plate stratigraphy, orogeny, paired metamorphic belts, ridge subduction, subhorizontal nappe, Yangtze.

Accepted for publication April 1996

290 Y. Isoxaki

INTRODUCTION

The modern Japanese Islands geographically com- prise five island arcs: the Kurile, Northeast Japan, Izu-Bonin, Southwest Japan, and Ryukyu arcs, showing complex patterns common in the western Pacific. Four of these form segments of active island arcs between the Eurasian continent and the Pacific Ocean where the seafloor is currently subducting westward beneath Asia; the Izu-Bonin arc is the exception and forms an intra-oceanic arc (Fig. 1). Ongoing subduction processes along these margins add materials to and modify tectonic features of the Asian continental margin. Geological studies in the 1980s revealed that the Japanese Islands evolved under similar tectonic processes to those active today, and that major geologic units exposed on the islands are subduction-related products of the Late Paleozoic to Cenozoic orogenies.

It appears that the orogenic belts in Japan have widened oceanward by about 400 km in 450 million years, by virtue of long-term subduction of the Pacific seafloor along the Yangtze (South China) and Sino-Korean (North China) continental blocks. Ac- cording to the classic categorization of orogenic belts by Dewey and Bird (1970), this 450 million year old orogen of Japan corresponds to a typical example of a Cordilleran-type orogen between a converging pair of continental and oceanic plates, that features a subduction complex, high-P/T schists and coeval granitic batholiths. The Cordil- leran-type orogen is thought to originate from steady-state subduction of an oceanic plate beneath a continental plate, and contrast was emphasized between the Cordilleran-type and the Alpine-Hima- layan style (or continent-continent collision-type) orogen. Although great variation in orogenic styles and possible plate tectonic mechanisms within the so-called Cordilleran orogens has since been described, the distinction between the Cordilleran-type and collision-type orogens still appears justified.

Geologic and tectonic studies in Japan during the last two decades have clarified several new and significant aspects of oceanic subduction-related (or classic Cordilleran-type) orogens. New observations are highlighted in the following five characteristic tectonic features of Japan: (i) intermittent growth of accretionary complexes; (ii) subhorizontal nappe structure; (iii) downward younging polarity; (iv) episodic formation of a tectonic sandwich with a high-PIT unit; and (v) microplate modification. Particularly noteworthy is the episodic (non- steady-state) growth pattern of a subduction- related orogen without a collision of continent or

arc because it clearly contrasts with the previous understandings on steady-state growth of the Cordilleran-type orogen.

This article reviews the latest version of the geotectonic subdivision of the Japanese Islands in view of these developments and discusses its tectonic implications. The implications of this review appear to impact also on subduction-related orogens in general. The new geotectonic subdivision is fundamentally adopted from the summary by Iso- zaki and Itaya (1991) and Isozaki and Maruyama (1991) published in Japanese, and is slightly modified to accommodate recent information. A historical review of studies of orogeny and geotectonic subdivision in Japan is also given in the appendix.

OCEANIC PLATE STRATIGRAPHY (OPS) ANALYSIS FOR ANCIENT ACCRETIONARY COMPLEX (AC)

Nearly 90% of the shallow-level crust of the Japa- nese Islands is occupied by Late Paleozoic to Meso- zoic accretionary complexes and granitic batholiths. This observation suggests that the major orogenic framework of the islands formed a t this time, and that the geotectonic subdivision of the Japanese supracrust depends mainly on the 3D configuration of the accretionary complexes and their high-PIT metamorphosed equivalents.

In this article, the term accretionary complex (AC) is strictly used in the following sense: an AC is a geologic entity which grows in situ in trench and trench inner wall in an active subduction zone as a result of subduction-driven layer-parallel shortening and vertical stacking/thickening of trench-fill materials usually composed of oceanic sediments and underlying volcanic rocks. This definition excludes continental blocks or island arcs, even though they once occurred in an oceanic domain prior to arrival a t the continental margin. The term OPS is an acronym of ‘oceanic plate stratigraphy’ and this represents a sequence of sediments and volcanic rocks accumulated primarily on an oceanic plate prior to subduction-accretion a t trench (Fig. 2). A full spectrum of an ideal OPS is comprised of, in ascending order: (i) MORB basalt; (ii) pelagic/hemipelagic sediments; and (iii) trench-fill turbidites, similar to the rocks recovered from drilling through modern trench floors. In modern examples, almost identical OPS is often preserved also in imbricated thrust packages in the trench inner-wall, that is, the youngest part in an AC. This unique stratigraphy was once called

Anatomy and evolution of Japanese Islands 291

/II \ N. American Piate

Fig. 1 Plate tectonic framework around the present day Japanese Islands Major plate interactions are oceanic subductions along the Nankai trough off southwest Japan and along the Japan trench off northeast Japan The former is an accretionary margin while the latter is an erosional margin with a high convergence rate and rugged-surfaced oceanic plate The Mariana trench represents another non-accretionary margin According to the oblique subduction of the Pacific and Philippine Sea plates southwestward moving fore-arc slivers develop in 3 domains adding secondary across-arc structural features The Miocene back-arc basin Japan Sea, is under destruc- tion along the convergent plate boundary between Eurasian and North American plates while a new back-arc basin is emerging in the Ryukyus associated with rifting of the iconti- nental crust

‘plate stratigraphy’ (Berger & Winterer 1974), however, the adjective ‘oceanic’ is added later in order to exclude sediments derived and accumulated exclu- sively on continental plate. Details of the concept of OPS and the practical example of the OPS analysis can be found in Isozaki et al. (in press b) and Mat- suda and Isozaki (1991).

It is generally easy to recognize the occurrence of modern AC by seismic resea,rch simply because they occur immediately next to active trenches. Ancient examples exposed on land, on the contrary, have usually lost contact with their primary tectonic setting nor wedge-like 3D geometry through later over- printing processes, thus their recognition requires

more specific information, not by external geometry but by internal structure and composition. The most reliable way to identify ancient AC, and to distin- guish them from neighboring ones, is the ‘OPS analysis’ refined mainly in Japan in the 1980s.

Ancient AC exposed on land also possess OPS which usually consists of a sequence in ascending order of greenstones (less than several tens of metres thick), deep-sea pelagic chert (less than 200 m thick), hemipelagic siliceous claystone (less than 100 m thick), and terrigenous clastic rocks such as mudstone and sandstone (more than 200 m thick). As most AC formed by the same teconic process from deep-sea rocks and sediments, they are

rands1 lmudst I C ~ O U S mudstone

bedded chert

- birth at MOR demlae at

I s ~ b d u c l l o n zone t Fig. 2 Simplified ridge subduction system and the concept of Oceanic Plate stratigraphy (OPS) (modified from Mat- suda & lsozaki 1991) Note the age gap between two distinct horizons (solid triangle the horizon between pillowed MORB greenstone and pelagic chert marking the birth of oceanic plate open triangle the

and terrigenous clastics marking the arrivdl at trench) which represents the total travol time of the subducting oceanic plate from mid-oceanic ridge to trench in o t h u words the age of the subducting oceanic plate at trench

auctlon zone

horizon between hemipelagic mudstone Wedge mantle

292 Y. Isoxaki

best distinguished by accurate age determination. Documenting a well-dated OPS for an ancient AC provides significant information on the timing of accretion and the age of the subducted ancient oceanic plate. Microfossil (conodont and radiolaria) dating of component sedimentary rocks is particularly effec- tive in OPS analysis for ancient AC exposed on land because it reveals (i) the accretion age a t trench ( = age of the horizon between hemipelagic sediment and trench-fill turbidite); and (ii) an age of subducted oceanic plate responsible for the accretion ( = duration of pelagic + hemipelagic deposition) (Fig. 2). Thus if documentation of a unique OPS is available with precise dating, it can provide a prime identity for each ancient AC unit that appears as a look-alike to neighboring units.

The OPS analysis for on-land exposed ancient AC has played the main role in the drastic change in geological studies in Japan in the 1980s (Ichikawa et al. 1990), and the research style of detailed field mapping in 1:5000 scale combined with microfossil dating is here called ‘microfossil mapping’ of ancient AC. An interval of one microfossil zone is less than 5 million years in average for the late Paleozoic to Mesozoic, although their resolution for dating is less precisely controlled with few tie points in time scale. However, it is still generally quite useful to distin- guish neighboring units in the field.

In addition, ages of metamorphosed AC of the ‘grey zone’ (see Appendix) that make microfossil dating difficult are determined radiometrically using various methods (K-Ar, Ar-Ar, Rb-Sr, and fission track). In some cases, an AC is accurately dated by two ‘radio ages’ (i.e. radiolarian-based accretion age and radiometrically dated age of subduction-related regional metamorphism). The subduction-related regional low-grade metamorphism usually occurred -10-20 million years later than the former (Takami et al. 1990, 1993; Kawato et al. 1991; Isozaki in press b), and it can add another reference feature for comparison in OPS. This research style combining geochronology dating and detailed field mapping is here called ‘chronometric mapping’ of ancient AC.

Practical examples of these research styles ap- plied to the Permian and Jurassic AC in SW Japan are reviewed in detail in Isozaki (in press a,b) and Kimura (1996).

GEOTECTONIC SUBDIVISION

The newly proposed geotectonic subdivision of the Japanese Islands based on these methods (Figs 3,4) is described here in four sections, based on domains

in southwest Japan, the Ryukyus, northeast Japan, and east Hokkaido that are a t present separated physiographically and/or tectonically by secondary transverse faults. The fundamental scheme of this subdivision is after Isozaki and Maruyama (1991), and is slightly modified according to the latest information. The Izu-Bonin Islands are not described in this article, as they form a young island arc of intra- oceanic nature and have lesser significance to the primary orogenic framework along the Asian continental margin.

In southwest Japan, including Kyushu, Shikoku and western Honshu Islands, the Cenozoic volcano- sedimentary covers are thinner than those in other domains due to the rapid uplift in the Quaternary, and this allows extensive exposure of Late Paleo- zoic to Mesozoic AC and their metamorphic equivalents. The Ryukyus and northeast Japan can be essentially treated as lateral extension of southwest Japan, however, these domains were considerably modified and dislocated by secondary tectonism including movement of fore-arc sliver, back-arc spreading, and arc-arc collision. Thus these two domains will be briefly explained after southwest Japan as its lateral equivalents. The island of Hokkaido is characterized by a unique setting with an arc-arc collision between the Northeast Japan arc and Kurile arc. East Hokkaido that belongs to Kurile arc will be mentioned separately.

In the course of explaining geotectonic subdivision, the term ‘belt’ is used in this article to describe the distribution of geotectonic units in two dimen- sions. When a three-dimensional geologic entity is to be described, non-genetic terms like unit, body, block, complex, and nappe are used. The term ‘terrane’ is avoided here because it may be confused with a prejudicial connotation of allochthoneity (Co- ney et al. 1980; Howell 1985), which has been suggested by Sengor and Dewey (1991) and Hamilton (1991). Refer to Ichikawa et al. (1990) for more detailed descriptions and relevant references of pre- Cretaceous belts; and to Taira et al. (1988,1989) for those of much younger belts. In this long description section, readers interested in tectonics rather than local geology of the islands may read only the description of southwest Japan and skip those of the Ryukyus, northeast Japan, and East Hokkaido because the latter areas represent lateral equivalents of southwest Japan.

SOUTHWEST JAPAN

In southwest Japan the two-dimensional east-west trending zonal arrangement of various geologic


Fig. 3 New geotectonic subdivision of the Japanese islands (modified from lsozaki & Maruyama 1991). The geotectonic subdivision of the Ryukyus and their correlation are shown separately in Fig. 8. Explanatory symbols for geotectonic units used in the figure and text are as follows: Southwest Japan [Rn. Renge belt ~

Sn: Sangun b ; Ak: Akiyoshi b , Mz. Maizuru b ; UT- Ultra-Tanba b (included in Mz in text), M-T . Mino-Tanba b., Ry. Ryoke b ; Sb: Sanbagawa b ; Ch: Chtchibu b (including Northern Chichibu b.; Kurosegawa belt, and Southern Chichibu belt in text); Sh; Shimanto b. (divided into Northern and Southern Shimanto belts in text]; Northeast Japan [Ht-Tk: Hitachi-Takanuki b. ( = Hida b. ) ; Gs. Gosaisho b. ( = Ryoke b . ) , M M : Matsugataira-Motai b. ( = Renge b ); SK. Southern Kitakamt b . ( = Oki b.); NK. Northern Kitakami-Oshima b . ( = Mino-Tanba b.); Kk: Kamuikotan b. ( = Sanbagawa b. ) , Hdk. Hidaka b. ( = Shimanto b. ) , Tokoro b ( = Sanbagawa b . + Shimanto b. ) ; Nm: Nemuro b.]. Solid black area represents ophiolitic zone

units is more obvious if surface cover and granitic intrusions are removed. Southwest Japan comprises 12 distinct geotectonic units (Fig. 5); from oldest to youngest; a 2.0 (;a-250 Ma gneiss complex; a 230 Ma intermediate-pressure type metamorphic complex; a 580-450 Ma ophiolite; a 400- 300 Ma high-P/T schist; a 250 Ma AC; a 230- 200 Ma high-P/T schist; a 180-140 Ma AC; a 120-100 Ma low-P/T metamorphic complex; a 100 Ma high-P/T schist, an 80 Ma AC; and a 40-20 Ma AC. These units are distributed in 15 belts, that is, from the Japa.n Sea side to the Pacific side: Oki belt; Hida b.; 0-eyama b.; Renge b.; Akiyoshi b.; Sangun b.; Maizuru ( + Ultra-Tanba) b.; Mino-Tanba b.; Ryoke b.; Sanbagawa b.; Northern Chichibu b.; Kurosegawa b.; Southern Chichibu b.; Northern Shimanto b.; and Southern Shimanto b. Repeated occurrence of the same unit in more than two belts (i.e. outliers in the form of

klippes and/or tectonic windows as shown in Fig. 6) in southwest Japan causes a mismatch in the number of belts and geotectonic units. For example, the Jurassic AC apparently occur in three belts, that is, the Mino-Tanba b.; N. Chichibu b.; and Southern Chichibu b., separated from each other for up to 50 km, although these are identical in terms of OPS.

The three belts along the Japan Sea side, that is, the Oki, Hida and 0-eyama belts, are intimately linked to Precambrian crusts in nature, and they form the core of the Phanerozoic orogen in southwest Japan (Fig. 3). On the other hand, the other 12 belts surrounding the above three represent zones of subduction-related accretionary growth that account for the 450 million year-long widening and thickening of southwest Japan.

The Oki belt of continental affinity is a t present isolated from mainland Asia, as it was rifted and

294 Y. Isoxaki

A A’ Hokkaido

B’

/ /

C’

h /

/

/

S Ryukyus D D’

\fib Pacific Ocean

C’

/ /

/

NE Japan

SW Japan

1 lOkm 100km

Granites

a Cretaceous and younger ACs

J Jurassic ACs

P Pre-Jurassic ACs

Continent

Shelf sediments

Fig. 4 Across-arc geologic profiles of the Japanese islands along four transects Hokkaido NE Japan SW Japan and southern Ryukyus (modified from lsozaki & Maruyama 1991) Abbreviations for major boundary faults are as follows (I) Thrust (bold) HMT Hidaka Main Thrust HyTZ Hayachine tectonic zone Ng-Hm TL Nagato-Hida marginal tectonic line, I-KTL Ishigaki-Kuga tectonic line BTL Butsuzo tectonic line (11) Strike-slip faults ATL Abashiri tectonic line, TTL Tanakura tectonic line HTL Hatagawa tectonic line H-KTL Hizurne-Kesen nurna tectonic line, M T L Median tectonic line

detached by the Miocene opening of a back-arc basin, the Japan Sea (Jolivert et al. 1994; Otofuji 1996, this issue). Judging from the lithologic/ chronologic similarity to the Precambrian rocks of the Sobaesan massif in South Korea, the Oki belt is regarded as an eastern extension of the Yangtze (South China) craton (Sohma et al. 1990; Isozaki & Maruyama 1991) which represents one of the continental pieces rifted apart from the supercontinent Rodinia a t 750-700 Ma (Powell et al. 1993; Li et al. 1995).

On the other hand, the kyanite-bearing Hida metamorphic rocks are regarded as the northeastern extension of the ultrahigh-pressure to high- pressure metamorphic rocks along the Qinling- Dabie suture (230 Ma continent-continent collision zone) in central China between the Yangtze and Sino-Korean (North China) blocks (Wang et al.

1989; Maruyama et al. 1994; Cong & Wang 1995). The unique occurrence of Middle to Late Paleozoic shelf strata with Boreal fauna in the periphery of the Hida belt (Igo 1990; Kato 1990) suggests a strong link between the Hida belt and Sino-Korean block. The present position of the Hida belt in the central part of Japan is due to later across-arc contraction and juxtaposition (Komatsu 1990), and the primary contact with the Oki belt (the Yangtze block) has been lost. Concerning the geotectonic correlation of the Oki and Hida belts with continental blocks and later tectonic juxtaposition, refer to Isozaki (in press a).

The 450-580 Ma ophiolite of the 0-eyama belt along the southern margin of the Oki belt is the oldest oceanic material in Japan. As its easterly extension in northeast Japan has mid-Paleozoic sedimentary cover of continental shelf facies


Fig. 5 Geotectonic subdivision of southwest Japan (modified from lsozaki & ltaya 1991) Hatched areas represent occurrence of Paleozoic AC, and solid black areas indicate their tectonic outliers on the Pacific side ( = Kurosegawa b) The Nagato-Hida Marginal tectonic line (b) separates the continental Oki belt from the younger Paleozoic-Mesozoic AC belts The lshigaki-Kuga tectonic line (f) divides Paleozoic AC belts and Mesozoic ones in southwest Japan Note its winding surface trajectory indicating a low-angle nature as shown in Fig 6

(Ozawa 1988), the 0-eyama ophiolite probably represents a remnant of the initial oceanic crust that formed through the break-up of Rodinia a t ca 750-700 Ma. Prior to the subduction regime started from 450 Ma a t the latest, a piece of the initial Pacific ocean floor was likely attached to the rifted continental margin of the Yangtze block; that is, the Oki belt (Isoza,ki & Maruyama 1991). The occurrence of the medium-pressure-type amphibolite in the 0-eyama belt probably suggests its involvement in the 250 Ma collision event (Isozaki in press a). The lithologic assemblage, age, and other characteristics of these three units, with continental affinity, are briefly described.

1, Hida belt (Hd); 1.1 Ga to 250 Ma medium- pressure-type metamorphic rocks and 180 Ma granites (Sohma & Kunugiza 1993). The highest metamorphic grade reaches the upper amphibolite facies, only locally to the granulite facies. Proto- liths of paragneiss are composed of sedimentary rocks that most likely accumulated along the passive continental margin. These include peralumi- nous siliciclastic rocks and impure carbonates with a minor amount of mafic igneous rocks. Non- to weakly metamorphosed Middle to Late Paleozoic shelf sequences occur fragmentally in the periphery of the belt; 2, Oki belt (Ok): 2.0 Ga to 250 Ma medium-pressure-type gneiss and granite complex (Suzuki & Adachi 1994). The highest metamorphic grade reaches the upper amphibolite to granulite facies. Protolith is continental sedimentam rocks

with minor amount of mafic igneous rocks; 3, 0-eyama belt (Oe): 450 to 580 Ma dismembered ophiolite composed mostly of serpentinized ultramafic rocks (lherzolitic.harzburgite), metagabbros, and a fragment of medium-pressure-type amphibolite (Arai 1980; Kurokawa 1985).

The belts 4-15 listed below are zones of subduction-related accretionary growth that prac- tically account for the 450 million year-long orogenic widening and thickening of southwest Japan (Figs 5,6) together with underlying granitic batholiths emplaced later. These AC belts including metamorphosed equivalents were successively added to the southeastern margin of the continent, in particular around the Yangtze block. A summary on OPS for these AC units in southwest Japan is shown in Fig. 7. I t is noteworthy that the timing of accretion is generally getting younger oceanward from the 400 Ma meta-AC to the Miocene AC, and that the age of the subducted oceanic plate responsible for accretion has been considerably variable; from -160 million years old for the Jurassic AC to almost zero for the Southern Shimanto AC. For a more detailed description of belts 5-7 and 12, refer to Isozaki (1996a); for a description of belts 8, 11 and 13 refer to Isozaki (in press b). 4, Renge belt (Rn): 400-300 Ma high-P/T schists and associated serpentinite (Nishimura 1990). The highest metamorphic grade reaches the high-pressure amphibolite facies through glaucophane schist facies. The protolith is an AC of unknown age composed of

296 Y. Isoxnki

,- Pre-Jurassic complexes ~

of the Inner Zone

12. Ks presentj

Tectonic Outlier

N a o a t o - Hida

N S

1.

L A INNER ZONE OUTER ZONE -- - -.

B

Fig. 6 Geotectonic profile of southwest Japan (modified from A lsozaki & ltaya 1991, B lsozaki eta / 1992) Profile A shows the subhorizontal nature of the piled nappes of AC with klippes and tectonic windows on the surface Note the downward younging polarity among the AC Block diagram B is an enlargement of central portion of profile A showing correlation of geotectonic units and boundaries across the Median Tectonic Line ( M T L ) Concerning the two-fold nature of the M T L refer to text and Figs 17b c and 18b

greenstones, siliciclastics and chert; 5, Akiyoshi belt (Ak): Late Permian (250 Ma) AC composed of oceanic greenstones mostly of oceanic island basalt (OIB) origin, chert, reef limestone, and terrigenous clastics (Kanmera et al. 1990). This unit suffered from low grade regional metamorphism up to the lower greenschist facies a t 220 Ma; 6, Sangun belt (Sn): 230-210 Ma high-PIT schists (Nishimura 1990). The highest metamorphic grade reaches the high-pressure amphibolite through the glaucophane schist facies. The protolith is an AC, probably containing a part of the 250 Ma Akiyoshi AC. Neighboring schist unit with problematic 200- 180 Ma ages (probably secondarily annealed) are also included here; 7, Maizuru belt (Mz): Middle- Late Permian AC with 280 Ma ophiolite (Hayasaka 1990; Ishiwatari et al. 1990). The unit sometimes discriminated as the ‘Ultra-Tanba belt’ (Caridroit et al. 1981; Ishiga 1990) is included here. The ophiolite suite is dismembered but its primary

thickness is estimated to be -25 km; 8, Mino- Tanba belt (MT): Jurassic AC with a minor amount of latest Triassic and earliest Cretaceous parts (200-140 Ma) (Wakita 1988; Nakae 1993). This unit is composed of oceanic greenstones of OIB origin, deep-sea pelagic chert, reef limestone, and terrigenous clastics. Secondarily mixed AC (olis- tostromes and melanges) occur commonly. This AC unit is tentatively subdivided into three parts; (i) Early Jurassic part (accreted a t 200 Ma; metamorphosed a t 170 Ma); (ii) Middle Jurassic part (accreted a t 170 Ma; metamorphosed a t 140 Ma); and (iii) Late Jurassic part (accreted a t 140-150 Ma; metamorphosed a t 120 Ma); 9, Ryoke belt (Ry): 120-100 Ma low-PIT metamorphic rocks and associated granites (Nakajima in press). The highest grade part includes sillimanite-bearing gneiss. Pro- tolith is mostly composed of Jurassic AC with lesser amount of the pre-Jurassic AC and sediments. Age of granites ranges mostly in 120-


UbdUCted ceanic Plate

4- !45

Fa ra /lo n

Akiyoshi

l 0 w . P m e l . be l l

100 Ma- __

Sangun hrgh P I T me1 belt

0 0 ,-

0 0 N

A

A f

-300 M:

lzanagi

Mino-Tanba-Chichibu (11. 13

I o w - P m e l . belt

k

400 Ma-- ~ _ _

hanba- gawa

met . b e l l h i g h - P I T

~.~ 0 e? - 0

~ ..N a: I ,

----- 4

._ lllllllllll

llllllllll! -

Kula /Pacific /Philip.Sea

& ;I 3 Northern Shimanto ~ Southern Shimanto

l o w - P met be l l t i

metamorphism ~

Legend

* peak O f

trench-1111 turbidites

hemipelagic sediments

pelagic sediments I

I A metagabbro I

Fig. 7 Oceanic plate stratigraphy (OPS) of AC units in southwest Japan and age of subduction-related regional metamorphism (modified from lsozaki & Maruyama 1991) Note the oceanward and tectonically downward younging polarity not only in accretion timing (age of the horizon between hemipelagic mudstone and terrigenous clastics) but also in high-P-T and relevant low-grade metamorphism (shown by asterisk)

70 Ma with eastward younging polarity along the Southwest Japan arc; 10. Sanbagawa belt (Sb): high-PIT metamorphosed Early Cretaceous AC (Banno & Sakai 1989; Takasu & Dallmeyer 1990), well known as the Sanbagawa schists. The highest grade reaches the high-pressure amphibolite facies and radiometric ages concentrate in 100-80 Ma. The high-PIT Sanbagawa belt and the low-P/T Ryoke belt (9) form paired metamorphic belts (Miyashiro 1961); 11, Northern Chichibu belt (Cn): Latest Triassic to Middle Jurassic AC equivalent to the older part of the Jurassic AC in the Mino- Tanba belt (8) (Hada & Kurimoto 1990). This unit is regarded as forming a tectonic outlier of the Jurassic complex of the Mino-Tanba belt; 12, Kurosegawa belt (Kr): Fault-bounded mixture of the pre-Jurassic elements (Yoshikura et al. 1990; Isozaki et al. 1992). Components of above- mentioned belts 3 to 7 occur chaotically as slivers, lenses and/or blocks of various sizes and shapes, enveloped within serpentiriite matrix. As a whole, this unit represents a tectonic outlier of the pre- Jurassic rocks, which occurred on the Asian continent side; 13, Southern Chichibu belt (Cs): Early Jurassic to earliest Cretaceous AC equivalent.to the

younger part of the Jurassic AC in the Mino-Tanba belt (8) and partly to that in the Northern Chichibu belt (11) (Matsuoka 1992); 14, Northern Shimanto belt (Shn): Late Cretaceous scarcely metamorphosed AC composed mostly of terrigenous elastics with lesser amount of oceanic rocks (Taira et al. 1988). Sporadically intervened are thin tectonic slices of melanges that include oceanic greenstones and bedded chert within scaly argillaceous matri- ces; 15, Southern Shimanto belt (Shs): Paleogene and Miocene little metamorphosed AC composed mostly of terrigenous elastic rocks (Taira et al. 1988). A minor amount of tectonic melanges occur in this belt.

Major geotectonic boundaries in southwest Japan (Figs 5,6) are listed with their nature. Also noted in parentheses a re well-documented examples of these boundary faults: (a) boundary between belts 1 and 2 (Hida b./Oki b.): low-angle thrust? (Unazuki suture) activated probably in late Mesozoic; (b) boundary between belts 2 and 3 (Oki b./O-eyama b.), and between belts 2 and 4 (Oki b./Renge b.): low-angle thrust (Nagato-Hida marginal tectonic line); (c) boundary between belts 4 and 5 (Renge b./ Akiyoshi b.): low-angle thrust (Toyogadake thrust;

298 Y. Isoxaki

Kabashima et al. 1994); (d) boundary between belts 5 and 6 (Akiyoshi b./Sangun b.): low-angle thrust (Kitayama thrust), (e) boundary between belts 6 and 7 (Sangun b./Maizuru b.): unexamined; (f) boundary between belts 7 and 8 (Maizuru b./Mino-Tanba b.): low-angle thrust (Ishigaki-Kuga tectonic line); (g) boundary between belts 8 and 9 (Mino-Tanba b./Ryoke b.): high-angle normal fault (Arima- Takatsuki line, Iwakuni fault); essential contact: gradual metamorphic aureole; (h) boundary between belts 9 and 10 (Ryoke b./Sanbagawa b.): primary low-angle thrust (Paleo-Median Tectonic Line activated in the Tertiary) and secondary high-angle strike-slip fault (Neo-MTL active in Quaternary). (i) boundary between belts 10 and 11 (Sanbagawa b./N. Chichibu b.): low-angle thrust (Sasagatani fault; Kawato et al. 1991); 0') boundary between belts 11 and 12 (Northern Chichibu b./Kurosegawa b.): low-angle thrust (Agekura thrust, Nakatsu thrust) = oceanward extension of the boundary fault f (Ishigaki-Kuga tectonic line); (k) boundary between belts 12 and 13 (Kurosegawa b./Southern Chichibu b.): low-angle thrust (Kanbaradani thrust) = oceanward extension of the boundary fault f (Ishigaki-Kuga tectonic line). (1) boundary between belts 13 and 1 4 (S. Chichibu b./N. Shimanto b.): low-angle thrust (Butsuzo Tectonic Line, Tsub- uro thrust; Sasaki & Isozaki 1992). (m) boundary between belts 1 4 and 15 (N. Shimanto b./S. Shi- manto b.): low-angle thrust (Nobeoka thrust, Aki tectonic line).

THERYUKYUS

On the basis of strong similarity of components, the Ryukyu Islands are basically regarded as the southwestern extension of southwest Japan (Fig. 8), however, a north-south trending transverse fault clearly separates the Ryukyus from Southwest Japan. This fault in west Kyushu Island sharply cuts off the zonal arrangement of southwest Japan. Right-lateral off-set of the Late Cre- taceous high-P/T unit in mid- to west Kyushu suggests the boundary fault between southwest Japan and the Ryukyus belongs to the right-lateral strike-slip fault system as well as the Tsushima fault and Yangshan fault in southeast Korea (Yoon & Chough 1995), probably formed in relation to the Miocene opening of the Japan Sea along its western margin.

Geological information on the Ryukyu Islands is limited by the lack of exposures but some geotectonic units are well correlated with those in southwest Japan. Most of the units in this domain are

composed of Late Paleozoic and Mesozoic AC and their metamorphic equivalents. There is no geotectonic unit with continental affinity in this domain except Early Paleozoic ophiolite in west Kyushu. Ac- cording to dredging data, the East China Sea is un- derlain by Precambrian rocks that probably belong to the Yangtze (South China) craton.

The geotectonic units hitherto known from Ryukyus are listed below. For convenience, the numbering and symbols that were used for units in southwest Japan are also used to describe units in the Ryukyus: 3, metagabbro from Nomo point: 450-580 Ma ophiolite ( = Oe) (Igi & Shibata 1979; Nishimura 1990); 6, Tomuru metamorphic rocks: 220 Ma high-P/T schists ( = Sn) (Nishimura 1990); 8, Fusaki Formation: Jurassic AC ( = MT) (Isozaki & Nishimura 1989); 10, Yuan Formation and Takashima schists: Early Cretaceous AC and 60- 90 Ma high-P/T metamorphic equivalents ( = Sb) (Ujiie & Hashimoto 1983); 1 4 , 15, Kunchan Group: Cretaceous and Paleogene AC ( = Sh) (Osozawa 1984).

The following two belt boundaries have been examined on land in Ryukyus; (f) boundary between belts 6 and 8 (Sn/MT): low-angle thrust (Ishigaki-Kuga tectonic line); and (1) boundary between belts 10 and 1 4 (SbIShn): high-angle fault (Butsuzo tectonic line).

NORTHEAST JAPAN

Northeast Japan comprises northeastern Honshu and the western Hokkaido Islands, and is separated from southwest Japan by a left-lateral strike-slip fault called Tanakura tectonic line (T.T.L.), and from eastern Hokkaido by a north- south trending fault in central Hokkaido, respec- tively (Fig. 3). Northeast Japan is also characterized by an apparent zonal arrangement of several geotectonic units on the surface (Fig. 9). This domain has been intensely modified by secondary tectonism, however, in particular by left-lateral strike-slip faults relevant to the Miocene opening of the Japan Sea. This faulting leads to uncertainty in the primary geometry and nature of contact among the component units. However, a comparison with southwest Japan can help in the recon- struction of primary features of this domain. Com- ponent units of northeast Japan are listed through comparison with those in southwest Japan, and short comments will be added for recent advances. As well as the Hida, Oki and 0-eyama belts in southwest Japan, the three belts in northeast Japan, that is, the Hitachi-Takanuki b., Southern

A n a t o m y m i d evolut ion of Jupanese I s l a n d s 299

/ active trench

/ Ishigaki-Kuga Tectonic Line lshigaki island

(6) 2 2 0 ~ . ~ I T SSM.IS

(8) Jurassk AC

Fig. 8 Geotectonic map of the Ryukyus and their link to southwest Japan (modified from lsozaki & Nishimura 1989) Inset map ( B ) of East Asia shows potential along-arc continuity of the Jurassic AC belt for example for -5000 km from west Philippines to Chind/Russia border Another 1000 km northerly extension IS

examined recently in the Kamtchatsuka Koryak region northeast Russia (refer to lsozaki in press b)

Kitakami b., and Miyamori-hayachine b., have strong continental affinities, in particular to the Yangtze craton and the collisional suture between the Yangtze and Sino-Korean cratons. The rest are AC units which later accreted to northeast Japan.

1, Hitachi-Takanuki belt (Ht-Tk): 250 Ma medium-pressure metamorphics (Tagiri 1973; Hiroi & Kishi 1989). The protoliths include Late Paleo- zoic sedimentary rocks accumulated on the continental shelf and volcanic rocks of bimodal characteristics ( = Hd). Granite-related thermal overprint occurred regionally at 110 Ma which is correlated to the Ryoke metamorphism in southwest Japan ( = Ry); 2, Southern Kitakami belt (SK): 440 Ma granite and gneiss, 350 Ma and 250 Ma granites ( = Ok) with Middle to Late Paleozoic sedimentary

covers characterized by marine fauna of Australian (Gondwanan) affinity (Suzuki & Adachi 1993; Kawamura et al. 1990; Kato 1990); 3, Miyamori- Hayachine belt (MH): 450 Ma ophiolite ( = Oe) with Paleozoic sedimentary covers (Ozawa 1988; Okami & Ehiro 1988); 4, Matsugataira-Motai belt (MM): 300-400 Ma high-PIT schists ( = Rn) (Maekawa 1981); 8, Northern Kitakami-Oshima belt (NK- 0s): Jurassic AC ( = MT) + Early Cretaceous AC ( = Sb) (Minoura 1990; Okami & Ehiro 1988); 9, Gosaisho belt (Gs): 110 Ma low-PIT metamorphic rocks and granites ( = Ry). Protoliths include components of the Jurassic AC ( = MT) (Tagiri et al. 1993); 10, Sorachi-Yezo belt (SY): Early Creta- ceous AC + 100 Ma high-PIT (Kamuikotan) schists associated with (Horokanai) ophiolite ( = Sb); 14,

300 Y. Isoxaki

Geotectonic subdivision of Northeast Japan

Pacific Ocean

Mi no -Ta n b a b.)

Paleozoic shelf sediments

High-P/T schists

itachi -Takanuki b.

Fig. 9 Geotectonic map of northeast Japan (modified from lsozaki & Maruyama 1991) Names in parentheses indicate correlative units in southwest Japan Note that the primary subhorizontal geotectonic boundaries are cut by a series of left-lateral strike-slip faults that were activated by the back-arc basin (Japan Sea) opening in the Miocene

Idonnappu belt (Id): Early Cretaceous to early Late Cretaceous AC ( = Sh); 14, 15, Hidaka belt (Hdk): Late Cretaceous to Paleogene AC ( = Sh) partly metamorphosed into low- to medium-pressure type metamorphics associated with 50 Ma migmatite- granite.

Although several units are well correlated to their counterparts in southwest Japan, some of the units

in southwest Japan are apparently missing in northeast Japan, such as Permian AC ( 5 ) , 200 Ma high- P/T schists (6), and 280 Ma ophiolite (7). Their counterparts, however, may possibly be found also in northwest Japan in future as subsurface underlying units bounded by blind thrusts, owing to the subhorizontal structure mentioned below. The geotectonic boundaries between these units in northeast


(iii). The fault (viii) (Hidaka Main thrusts) represents another example of the secondary modification upon the primary structure, and this transpressional fault has been activated probably by ongoing westward collision of the Kurile fore-arc sliver (Kimura 1985, 1996).

Due to severe secondary modifications, primary orogenic structures in northeast Japan have not been fully clarified, however, an analogy in rock type, age and OPS of AC to southwest Japan suggests that a similar subhorizontal piled nappe structure also predominates in northeast Japan which may be represented by a series of subsurface blind thrusts (Fig. 10).

Japan are listed here: (i) boundary between belts 1 and 2 (Hitachi-Takanuki b.lSouthern Kitakami b.): unknown; (ii) boundary between belts 2 and 3 (Southern Kitakami b./Miyamori-Hayachine b.): vertical fault but primarily low-angle thrust (western margin of the Hayachine tectonic zone); (iii) boundary between belts 3 and 8 (Miyamori- Hayachine b./Northern Kitakami b.): vertical fault but primarily low-angle thrust (eastern margin of the Hayachine tectonic zone; Tazawa 1988).

The boundaries (ii) and (iii) appear almost vertical in outcrop but the large-scale sinuous trajectory on the surface suggests a potentially low-angle nature in deeper levels (Fig. 10). The zone between these two faults has been traditionally called the Haya- chine tectonic zone because serpentinized ophiolite occurs in an apparently narrow belt. The eastern margin of this zone ( = b) corresponds to the Ishi- gaki-Kuga tectonic line ( f ) in southwest Japan.

(iv) boundary between belts 1 and 9 (Hitachi- Takanuki b./Gosaisho b.): east-dipping low-angle fault probably activated before the intrusion of the Cretaceous granite; (v) boundary between belts 9 and 4 (Gosaisho b./Matsugataira-Motai b.): left- lateral strike-slip fault (Hatagawa tectonic line); (vi) boundary between belts 4 and 8 (Matsugataira- Motai b./Northern Kitakami-Oshima b.): left- lateral strike-slip fault (Futaba fault); (vii) boundary between belts 8 and 10 (Northern Kitakami- Oshima b./Sorachi-Yezo b ): unknown (concealed beneath Quaternary sediments); (viii) boundary between belts 1 4 and 15 (Idonappu b./Hidaka b.): east-dipping high-angle thrust (Hidaka Main Thrust) on the surface that translates into a low- angle one in deeper level (Ikawa e t al. 1995).

The faults (v) and (vi) are typical examples of sinistral strike-slip faults as well as the Hizume- Kesen’numa tectonic line that activated during the Miocene opening event of the Japan Sea along its eastern margin. These faults clearly cut the primary sinuous boundary faults such as the faults (ii) and

Fig. 10 Profile of northeast Japan Note the left-lateral strike slip faults that cut primary subhorizontal piled nappe structure that includes inferred blind thrusts

EAST HOKKAIDO

East Hokkaido has a rather complicated tectonic history compared to the rest of the Japanese Islands, probably reflecting its peculiar geotectonic condition, sandwiched between two Cenozoic back- arc basins, that is, the Japan Sea and Kurile basin (Fig. 1). The domain boundary between northeast Japan including West Hokkaido and East Hokkaido is inferred in central Hokkaido in the name of the Tokoro fault or Shibetu fault, however, its precise position, geometry and nature are unknown owing to thick Quaternary covers in between. Three geotectonic units are recognized in East Hokkaido, as described below (Fig. 11). There is no Paleozoic and Early Mesozoic unit in East Hokkaido, and this implies a unique geohistory for East Hokkaido: 13, Yubetsu belt (Yb): Cretaceous AC; 10, 14 , Tokoro belt (Tk): Cretaceous AC and high-P/T metamorphic equivalents; 14, 15, Nemuro belt (Nm): Cretaceous-Tertiary shelf sequences with unknown basement probably composed of Mesozoic- Cenozoic crystalline rocks of arc affinity.

These three belts are separated from each other by north-south trending faults. The fault between the Tokoro and Nemuro belts has a strike-slip

I

I I I I Halaaawa

-SW Japan---+l* NE Japan

Hayachine T Z ( 3 O e + 4 R n i 6 Sn?)

..

10 Sb

NE

302 Y. Isoxaki

\ \ Northeast Japan -\-

ldonappu belt -3 \ \

14,(Shimanto b.) - , i \

younging polarity --+ Northern Kitakami-

Oshima b. 8. (Mino-Tanba b.)

__ East Hokkaido ~

North

10 Sorac hi-Y ezo b. .(Sanbagawa b.)

~ Yubetsu b. (Shimanto b.) 14.

\ ' V \

\,Hidaka b?. \\ (shimanto b.) 100 km

14, 15.

Fig. 11 Geotectonic map of Hokkaido (modified from Kimura unpubl data, 1996) Note the polarity reversal of younging direction in AC units across the Hidaka Main Thrust which represents a collision suture between the Kurile arc to the northeast Japan arc

nature, and this fault, called the Abashiri tectonic line, has activated in a right-lateral manner during the opening event of the Kurile back-arc basin. Refer to Kimura (1996) for further details.

The main problem in Hokkaido lies in how to interpret the origin of the parallel-running two coeval blueschists belts, that is, the Kamuikotan b. of northeast Japan and the Tokoro belt of East Hokkaido. The opposite younging polarity in AC between northeast Japan and East Hokkaido (Fig. 11) suggests that a secondary collision/ amalgamation has doubled the primarily linear blueschists belt along the Asian margin. Several other interpretations were also proposed but a final agreement has not yet been reached.

GEOTECTONIC HISTORY OF THE JAPANESE ISLANDS

The time-space relationships among the orogenic units described here suggest that the Japanese Islands have grown oceanward by almost 400 km across the arc since the mid-Paleozoic. On the basis of the new geological data and the revised subdivision of the Japanese Islands, the geotectonic history of the Japanese Islands is summarized below. The history began with rifting of a supercontinent in the late Neoproterozoic, and was fol- lowed by a tectonic inversion shifting from an extensional (Atlantic-type) regime to a convergent (Pacific-type) regime around 500 Ma. Since then an oceanic subduction-related accretion regime has


the core of Japan, which began as a segment of the Yangtze continental margin. In addition, a fragment of the Sino-Korean continental block also participated in the orogenic growth of Japan after the 250 Ma collisional event (Isozaki in press a).

persisted, and is responsible for the oceanward growth of the islands. Figure 12 summarizes this history from the birth of the islands a t ca 700 Ma to their separation from Asia in the Miocene.

BIRTHPLACE OF PROTO-JAPAN These two continental blocks, particularly the Yangtze block, have played the most important roles

The Oki belt in southwest Japan and the Southern Kitakami belt in northeast Japan represents part of

in constraining the configuration of the Japanese Phanerozoic orogen.

I _ - -

rifling ,-

I 1.0-07Ga I

Asthenospheric mantle l i I 1

Break-up of supercontinent Rodinia I and birth of Pacific (750-700 Ma)

(b)

Y ~ J W ~ ' Passive marain sediments Remnant of orimilive

Onset of initial oceanic subduction

(tectonic inversion) (-500 M a )

Calc-alkaline 0-eyama fore-arc ophtolite volcanic arc --, ,,

Omanward accretionay growth in the Paleozoic (400-250 Ma)

lCollision of YangtzeISino-Korea

Hida gneiss (collision comDlex) Ok Oe Rn Ak 6n.c: Mz V.T Sb SIB

system along continental

Fig. 12 Series of simplified cartoons (a-g) showing 700 million years of geotectonic evolution of the Japanese Islands (a) The birth of Japan along a Proterozoic rifted continental margin of Yangtze by the break-up of the supercontinent Rodinia and the birth of the Pacific Ocean at 750-700 Ma. (b) After widening of the proto-Pacific Ocean basin, tectonic inversion occurred to initiate an intraoceanic subduction zone off the rifted Yangtze margin around 500 Ma, leaving a small fragment of primitiveoceaniccrust (0-eyamaophiolite). (c) Thearc-trench system wasestablished by ca450 M a a n d gaveacalc-alkal ineoverprint on thefore-arcophiolite. (d) The subduction zone matured to accommodate the oldest AC which corresponds to the protoliths of the 400 Ma high-P/T Renge schists. The high-P/T Renge belt together with coeval granite belt (fragmented) form the oldest set of paired metamorphic belts. (e) Successive oceanic subduction widened the accretionary edifice during the Late Paleozoic (the Renge and Akiyoshi belts). On the opposite side of the Yangtze block, another continental block (Sino-Korea) collided and thrust over the Yangtze at around 250 Ma. (f) Oceanward growth on the Pacific side continued during the Mesozolc and Cenozolc to widen and thicken the accretionary edifice in the form of subhorizontal piled nappes. The accretionary growth was punctuated by the episodic arrival of mid-oceanic ridge roughly every 100 milion years, which generated 3 to 4 sets of sandwich structure of high-P/T nappe between low-pressure AC units above and below (e.9. Sangun and Sanbagawa belts; see text for details). (9) By the back-arc opening of the Japan Sea at 20 Ma, Japan detached from mainland Asia; a continental arc becamean island arc The Hida belt represents a remnant of the 250 M a collision suture between the Yangtze and Sino-Korean blocks that partly features an ultrahigh-pressure metamorphic belt.

304 Y. Isoxuki

The relation of the Yangtze and Sino-Korean continental margins to protoJapan goes back to nearly 750-700 Ma when it is believed that the supercontinent Rodinia started rifting apart (Hoff- man 1991; Dalziel 1992; Powell et al. 1993). Judg- ing from the patterns of radial dyke swarms and rifted basins (Park et al. 1995; Bond et al. 1984), the breakup of Rodinia was probably triggered by a superplume rising to the surface from the mantle/ core boundary. Consequently, the proto-Pacific ocean was born in 750-700 Ma by the rifting of Rodinia, and several continental fragments, including the Sino-Korean and Yangtze blocks were dispersed in various directions (Fig. 12a). Due to similarities in Neoproterozoic to Cambrian stratigraphy, Laurentia (North America) and Australia are probably the conjugate continental block(s) to the rifted Yangtze margins (Li et al. 1995). How- ever, the paleoposition of the Sino-Korean block with respect to Rodinia remains highly enigmatic. The birthplace of Japan was probably located some- where in the northern periphery of Rodinia in mid- latitudes a t about 700 Ma. Most of the dispersed continental fragments once again assembled to form another (semi-)supercontinent, Gondwana- land, about 500 Ma (Hoffman 1991; Dalziel 1992), however, some large blocks such as Laurentia, Sibe- ria and Baltica were isolated from the superconti- nental mass. Likewise, the absence of evidence for late Neoproterozoic to Cambrian collision in China suggests that both the Sino-Korean and Yangtze blocks, including proto-Japan, were also isolated from Gondwanaland, although Early to mid-Paleo- zoic faunal provincialism suggests that the Yangtze block (including proto-Japan) and Australia were close neighbors (Burrett et al. 1990; Kato 1990).

Due to limited exposure and later tectonic modification in Japan, there is little evidence to conclude that continental rifting, such as extensional fault system, rift-related bimodal volcanism, and rift- related sedimentary sequences were active a t this time. Stratigraphical and paleontological studies suggest that the small distribution of Early to Middle Paleozoic (Ordovician to Devonian) terrigenous clastic/carbonate sequences in the periphery of the Hida belt, Hitachi-Takanuki belt, Southern KitakamiIMatsugataira-Motai belt, and the Kurosegawa belt all represent remnants of continental shelf facies accumulated along the rifted Proterozoic continental margins. The Korean peninsula and/or mainland China may have preserved such features of Rodinian rifting, however, further structural and stratigraphical analyses are needed.

INVERSION FROM PASSIVE TO ACTIVE MARGIN

With the exception of Precambrian gneissic clasts in younger sediments, the oldest unit of oceanic affinity in Japan is the 580 Ma ophiolite in the O-eyama belt, southwest Japan. Its occurrence in the periphery of the Oki belt (ancient Yangtze margin) suggests that this unit is a remnant of proto-Pacific oceanic crust. The O-eyama ophiolite and its equivalent in the Miyamori-Hayachine belt in northeast Japan have a bimodal distribution of radiometric ages; one a t 580 Ma and another a t 480-450 Ma (Ozawa 1988; Nishimura & Shibata 1989). Isozaki and Maruyama (1991) explained the age distribution of the oldest ophiolite in Japan in the following way.

The tectonic history of the O-eyama ophiolite is two-fold: (i) following rifting with the emplacement of nascent oceanic crust, a MORB-like oceanic crust formed at the proto-Pacific mid-oceanic ridge around 580 Ma (Fig. 12a); and (ii) these rocks were then intruded around 450 Ma by calc-alkaline volcanism of arc affinity when a new intra-oceanic subduction was initiated (Fig. 12b,c).

O-eyama is regarded as the only example of a fore-arc ophiolite in Japan, while other ophiolitic rocks are regarded as accreted fragments of ancient seamounts, rises and plateaus (Isozaki et al. 1990b; Kimura & Maruyama in press). Between 580 Ma and 480 Ma, tectonics in proto-Japan changed dramatically from riftlridge-related extension to subduction-related compression, and passive margins changed rapidly to active margins. This tectonic inversion probably corresponds to global plate reorganization, in particular to the opening of the proto-Atlantic (Iapetus) ocean on the opposite side of the globe. Accretion of the O-eyama ophiolite can be explained by either of the following two mechanisms: (i) initiation of a landward dipping subduction zone within the primary oceanic crust (see Fig. 12b,c); or (ii) collision of an island arc system from the ocean side and a reversal of subduction polarity.

ACCRETIONARY GROWTH OF THE JAPANESE ISLANDS

After the tectonic inversion around 500 Ma, proto- Japan began a state of accretionary growth that persists today. Within 50 million years after initiation of intra-oceanic subduction, the arc-trench system featured AC, high-PIT schists (metamorphosed AC), and granitic batholiths (Fig. 12d,e) that is, in the Renge belt plus Kurosegawa belt (=tectonic outlier of the former) in southwest

Amtomy a n d evolution of Japanese I s lands 305

Sino-Korean and Yangtze blocks collided along the Qinling-Dabie suture, generating an ultrahigh- pressure metamorphic belt, best preserved in the Dabie mountains and Shandon province in China (Wang et al. 1989; Maruyama e t al. 1994; Cong & Wang 1995). The region of ultrahigh-pressure metamorphism appears to be restricted to the central part of the 3000 km long suture, suggest- ing that a promontory collision may have locally generated ultrahigh-pressure conditions (Isozaki in press a). East of the Shandon peninsula, the coeval 250 Ma regional metamorphism is detected solely in the Hida and Hitachi-Takanuki belts in Japan; they are characterized by medium-pressure-type metamorphism. Although these belts lack ultrahigh-pressure metamorphism, from a geographical viewpoint (Fig. 13), they may also correspond to an eastern extension of a collision-related metamorphic belt (Sohma et al. 1990), including the Imjin- gang or Ogchon zones in Korea (Ernst e t al. 1988; Lan e t al. 1995). The Hida belt and its equivalent occur a t the eastern terminal of the suture. Data from this belt indicate a decrease in grade and magnitude of a collision-related metamorphic belt. After 250 Ma, accretionary growth in Japan radi- ated from a core composed of the amalgamated Sino-Korea plus Yangtze blocks. Thus while minor in extent, the Japanese Islands do contain elements of a non-accretionary, continent-continent collision- type orogen.

Japan, and in Southern Kitakami and Matsuga- taira-Motai belts in northeast Japan. In particular, the oldest AC in Japan is the protolith of the 450- 400 Ma high-P/T schists. The high-P/T schists and coeval granitic rocks are elements of a paired metamorphic belt (with a high-P/T belt on the ocean side and a low-P/T belt on the continent side; Miyashiro 1961).

Following the oldest 450 Ma unit, Late Paleo- zoic, Mesozoic and Cenozoic AC were formed through subsequent subduction. At least several major oceanic plates have subducted beneath the Yangtze margin, leaving more than 10 distinct AC belts. Numerous oceanic fragments derived from subducted oceanic plates, including deep-sea sediments and seamount-derived basaltslreef limestone, were accreted to Japan. Details of the accretion processes during the Permian and Juras- sic periods are reported in Isozaki (in press a,b).

Accretionary growth apparently has been not continuous. Including the youngest AC now under construction along the Nankai trough off southwest Japan, total accretionary growth is nealy 400 km in across-arc width (Fig. 12f), not taking into account the material loss by subduction-erosion. Thus the overall AC-dominated orogen in Japan has grown oceanward for almost 400 km during the 450 million years (-100 km per 100 million years).

As all of the AC units in Japan were formed in situ by subduction along the Yangtze (South China) continental margin, they are autochthonous to Asia, with the exception of small oceanic fragments peeled off from subducting oceanic crust and accreted landward into AC (Isozaki et al. 1990b). I t thus appeared that the Japanese islands do not represent a collage of ‘suspect or exotic terranes’ that had existed prior to subduction and accretion processes (Coney et al. 1980).

I t was a mere 20 million years ago when the Japanese Islands obtained their present configuration through back-arc spreading (Fig. 12g). How- ever, the accretionary growth of the Japanese Islands will likely continue until other continental blocks, such as Australia or North America, collide against Asia to form a future supercontinent (Maruyama 1994).

REMNANT OF CONTINENT-CONTINENT COLLISION

While most of the geotectonic units in Japan are the results of oceanic subduction, the Hida belt and the Hitachi-Takanuki belt are remnants of continent-continent collision. At about 250 Ma, the

DISCUSSION

The newly revised geotectonic subdivision and reconstructed tectonic history of the Japanese Is- lands clarifies the properties of the Cordilleran-type orogenic growth. Five tectonic features are described following these observations in Japan. They are: (i) step-wise growth of AC; (ii) subhorizontal nappe structure; (iii) downward younging polarity; (iv) a tectonic sandwich of high-P/T metamorphic units; and (v) secondary tectonic modifications. These may represent the principal characteristics of subduction-related orogenic belts in general.

STEP-WISE GROWTH OF AC UNITS

The 450 million-year-old subduction-related history of the Japanese Islands is characterized by the intermittent formation of AC, clearly detected by OPS analysis (Fig. 7). Several intervals lack AC (Carboniferous-Early Permian, Early-Middle Tri- assic, and late Early Cretaceous), and this empha-

306 Y. Isoxaki

Jurassic AC

sizes the overall zonal (piled nappe) arrangement of AC in Japan (Figs 5,6,8-11). Secular changes in relative plate motion between the Asian continent (or its precursory fragment) and subducted oceanic plates may be responsible for such episodicity because AC cannot form from a highly oblique subduction or in an along-arc strike-slip regime (Maruyama & Sen0 1986).

On the other hand, extensive modern AC can be constructed where a large amount of sediments are supplied and the subduction rate is moderately low. No accretion or tectonic erosion (subduction- erosion) occurs when a trench is starved of sediments or the subduction rate is too high (von Huene & Lallemand 1990; von Huene & Scholle

Fig. 13 Tectonic framework of 250 M a East Asia showing a continent-continent collision orogen between the two Precambrian cratons that is the Sino-Korean and Yangtze blocks (modified from lsozaki & Maruyama 1991) The collision suture fea- turing the 250 M a ultrahigh-pressure rneta- morphic belt of the Qinling-Dabie zone in China extends eastward to the Hida belt and Hitachi-Takanuki belt in Japan character ized by the 250 M a medium-pressure metamorphism The pathway of the suture in the Korean peninsula is still controversial as two alternative interpretations are proposed, one along the Ogchon zone and the other along the Imjingan zone

1991). Thus the apparent episodicity in the formation of AC in Japan may also indicate the episodic preservation of AC rather than formation. None- theless such episodic growth of AC is a primary characteristic of oceanic subduction-related orogens that persist for 100 million years or longer.

The formation of high-PIT metamorphic and coeval granitic belts also appears episodic as these occur in highly restricted time intervals. As they occur roughly every 100 million years, their formation may also be episodic and/or periodic.

In contrast, the classic concept of the Cordil- leran-type orogeny proposed by Dewey and Bird (1970) assumed a steady-state orogenic process, including the simultaneous formation of subduction


Subhorizontal piled nappe structures generally characterize continent-continent collision-type orogens, like the European Alps or the foreland fold- and-thrust belt in the Appalachian and Canadian Rocky mountains. Development of similar structures in an arc-trench setting dominated by subduction tectonics is noteworthy and should be tested in other AC-dominated orogens older than 100 million years.

complexes ( = AC), granitic batholiths, regional metamorphic belts, and relevant geologic structures. Later publications, h.owever, emphasized the non-steady-state nature of the arc-subduction zone setting (Dewey 1980). Although various interpretations are given, causes for the episodicity in subduction-related orogeny has not been well clarified. A possible cause for the episodic exhumation of high-PIT metamorphosed AC and coeval granitic activity associated with low-P/T regional metamorphism will be discussed later.

SUBHORIZONTAL PILED NAPPE STRUCTURE OF AC

Most of the geotectonic boundaries between adjacent AC units in southwest Japan are low-angle faults, and therefore the AC units are considered to be piled nappes (Fig. 6). Recent microfossil/ chronometric mapping documented the occurrences of regional klippes and windows of these nappes even in thickly vegetated areas. These observations indicate that the subhorizontal piled nappe structure governs the fundamental tectonic framework of the 450 million-year-old orogen in southwest Japan (this study: see also Hara et al. 1977; Charvet et al. 1985; Faure 1985). Primary orogenic structures in the Ryukyus and northeast Japan have not yet been fully mapped but a subhorizontal piled nappe structure controlled by subsurface blind thrusts has been predicted (Tazawa 1988; Isozaki & Nishimura 1989; Fig. 10).

To a first approximation, such a subhorizontal piled nappe structure is consistent with subduction- related deformations, particularly to the subhorizontal shortening caused by underplating through which new materials are added to the sole of the previously formed accretionary wedge through step-wise activation of a subhorizontal decolle- ment. The size of an individual ancient AC nappe is nearly 200 km in width across the arc, similar to the size of the widest modern AC wedge (von Huene & Scholle 1991).

The documentation of a predominant subhorizontal structure in Japan is contrary to the traditional view that vertical tectonics dominated horizontal tectonics. There are, in fact, some along-arc vertical faults of strike-slip nature (Neo-M.T.L., T.T.L.), but most of them became active in the Cenozoic and were driven by microplate activities. The total amount of displacement along the vertical fault, however, is too small to account for the present zonal arrangement of belts in Japan that extend along the arc for more than 1000 km.

DOWNWARD YOUNGING POLARITY IN AC NAPPES

The along-arc zonal arrangement of these ancient AC is most clearly demonstrated in southwest Japan (Fig. 14a). Remarkably, these AC show an oceanward younging polarity in map view without windows and klippes. In addition, ages of regional metamorphic belts and granitic batholiths also suggest an oceanward younging polarity. This apparent polarity is a function of the piled nappe structure of the AC (Figs 6,12f). Downward younging polarity is also recognized within an individual AC nappe; this was clearly demonstrated in the Juras- sic AC, which comprises six or more subnappes (Isozaki in press b). Such oceanward and downward younging polarity is consistent with the growth patterns of modern AC.

Although detailed information is scarce, the Ryukyus and northeast Japan plus East Hokkaido also appear to display younging polarity. The present zonal arrangement in northeast Japan, however, does not fit with the oceanward younging polarity in southwest Japan (Fig. 14b). This is due to secondary strike-slip faulting related to the opening of the Japan Sea in the Miocene which modified the primary orogenic configuration.

TECTONIC SANDWICH OF HIGH-P/T AC UNIT

There are three geotectonic units composed of high-P/T schist in the Japanese Islands formed a t around 450-300 Ma, 200 Ma, and 100 Ma. The 450-300 schists can be further subdivided into two distinct units. All of these high-P/T units also occur as subhorizontal nappes (Fig. 6). The most striking feature of the high-P/T nappes is their sandwich-like structure, where a high-PIT nappe is tectonically interleaved between two unmetamorphosed AC nappes (Maruyama 1990; Isozaki & Maruyama 1991). The nappes composed of high- P/T metamorphosed units can be traced laterally for more than 500 km along the arc, even though they are usually thinner than 2 km. For example, the 100-70 Ma Sanbagawa schists (Sb) occur as a

308 Y. Isoxaki

SW-Japan / Ryukyus NE Japan

Sino - Korean crafon

Precambrian craton collision zone

ophiolite / high-P/T metamorphosed AC non- to weakly metamorphosed AC

Shn

\ \

low-PrT metamorphosed part

Fig. 14 Cartoon showing stepwise oceanward growth of southwest Japan with a high-P/T sandwich structure after removing windows and klippes (see text) in

a map view and a contrasting feature in northeast Japan which was secondarily modified by strike-slip shredding lo obscure the primary structure

nappe sandwiched between the overlying Jurassic AC (MT) and the underlying Cretaceous AC (Sh) (Kawato et al. 1992; Sasaki & Isozaki 1992).

There are high pressure gaps between the high- P/T nappe in the middle of the sandwich and the adjacent AC nappes. These gaps, which are greater than several kilobars in pressure, correspond to 10-20 km of crustal thickness. In order to preserve these pressure gaps without metamorphic anneal- ing, the tectonic insertion of the high-P/T nappe must have been rapid and caused by tectonic exhumation of a high-PIT nappe into a low- pressure domain. In fact, all of the high-P/T nappes are bounded by a set of subhorizontal faults along their top and bottom surfaces, however, the sense of dislocation between these faults is opposite. To compensate for the pressure gap between the hanging wall and foot wall, the upper fault is normal while the bottom fault is reverse. Synchro- nous activation of such paired faults is believed to result in the insertion of the thin high-P/T nappe into the low-pressure domain.

These observations strongly suggest that the exhumation of the high-PIT nappes was tectonic

and episodic rather than caused by steady-state processes wherein buoyant uplift of a metamorphic domain together with over- and underlying unmetamorphosed units was proposed. This episodic tectonic exhumation mechanism called ‘wedge ex- trusion’ was first suggested by Maruyama (1990; see also Maruyama in press). A very similar model was also proposed for Himalayan medium-pressure gneisses by Burchfiel et al. (1992). The downward younging polarity among AC is not disturbed by the intermittent intercalation of high-P/T nappes. This suggests that subduction-driven burial of the protolith AC, high-PIT metamorphism, and tectonic exhumation of metamorphosed AC all occurred in the same structural horizon, likely along the Wadati-Benioff plane. The large-scale sandwich (containing a thin well-done steak) with downward younging polarity cannot be formed through any other known exhumation process (Suppe 1974; Cloos 1982; Platt 1986).

Based on radiometric ages determined by various methods, the timing of metamorphic peak temper- ature and the subsequent cooling history of high- P/T units in Japan (Itaya & Takasugi 1989; Nishi-

Anatomy and eTolution qf Japanese Islands 309

gests an eastward along-arc younging polarity for these intrusions (Nakajima e t al. 1990; Kinoshita 1995). This observation is consistent with the oblique subduction of the Kula/Pacific ridge beneath Asia as mentioned above (i.e. northeastward passage of the TTR triple junction of the Eurasia/ Kula/Pacific plates, because the subducted ridge- related slab window may have been a heat source).

The coincident timing of high-P/T nappe exhumation and granite belt formation suggests a causal relationship to ridge subduction. Further research is needed to elucidate the mechanism of the subhorizontal ejection of high-PIT nappes under a buoyant subduction regime induced by the movement of young oceanic crust. The preservation of high-PIT units in the fore-arc is possible even under high heat flow from the subducted ridge if the exhumation of the high-PIT nappe preceded the arrival of the ridge-crest a t the trench. The metamorphic/cooling ages of the high-P/T unit (100-70 Ma) are slightly older than the subduction timing of the brand new oceanic plate (ca 75 Ma), and this may support the interpretation provided above.

Prior to the 100-70 Ma event, exhumation of the high-PIT nappe occurred probably three times in the 600 million year history of the Japanese Islands (Fig. 15). Such episodic occurrences of the high- PIT sandwich structure suggests that these were also consequences of episodic ridge-subduction in the Paleozoic and early Mesozoic (Fig. 16). The distribution of pre-Cretaceous granites in Japan is highly limited; most of them are fragmented. Nonetheless the ages of the granites that are preserved suggest episodic formation of a granitic belt in the intervals of 450-400 Ma, 350 Ma, and 250 Ma.

If those three pre-Cretaceous geologic episodes are regarded as the result of episodic ridge subduction, the tectonic history of the Japanese Islands

mura e t al. 1989; Shibata & Nishimura 1989; Takasu & Dallmeyer 1990), the exhumation of high-P/T nappes occurred episodically. I t is noteworthy that these episodic events apparently coin- cide with the formation of granitic batholith in low-P/T metamorphic belts on the continent side (Isozaki & Maruyama 1991; Fig. 15). The best example of such paired metamorphic belts, described by Miyashiro (1961), is the 100-70 Ma high-P/T Sanbagawa belt on the trench side and the low-P/T Ryoke belt on the arc side of southwest Japan.

The formation of the Cretaceous paired belts was probably related to subduction of the Kula/Pacific ridge along the Asian margin. An observation on contemporary OPS in mid-Cretaceous AC in the Shimanto belt in southwest Japan (Taira e t al. 1988) suggested a progressive decrease in the age of the subducted slab in the Late Cretaceous (Fig. 16). At about 75 Ma, an AC was formed by subduction of a brand new oceanic plate (a mid- oceanic ridge per se ) .

In addition, based on hot spot tracks, the reconstructed paleo-plate motion in the Pacific domain over the last 150 million ,years indicates that the Kula-Pacific mid-oceanic ridge subducted obliquely beneath the eastern Asian continental margin around 100-70 Ma (Engebretson e t al. 1985). Thus a TTR (trench-trench-ridge) triple junction may have migrated from southwest to northeast off-shore of Japan.

The phenomenon of ridge-subduction and its tectonic effects on the continental margin have long been discussed (Pitman & Hays 1968; Wilson 1973; Uyeda & Miyashiro 1974); however, its tectono-thermal effects and orogenic consequences are still controversial (Farrar & Dixon 1993; Thor- kelson 1994). A recent compilation on the age of Mesozoic granite emplacement in East Asia sug-

Stepwi$e oceanward propaRation of BS belt and granite belt Fig. 15 Simplified diagram of geotectonic profile of the Japanese Islands (modified from lsozaki & Maruyama 1991) Note the oceanward and tectonically downward N younging growth of AC nappes incluijing high-P/T blueschist (BS) nappe Formation of the sandwich structure of BS nappe occurs intermittently roughly every 100 million years, accompanying coeval granite batholith belt with low-P/T metamorphic belt which is usually positioned -100-200 km continentward from coeval BS belt The iriter- mittent oceanward propagation of the BS belt and granite belt appears to have occurred when mid-oceanic ridges episodically col-

b

High-PIT metamorphic AC (BS) nappe

Low-pressure A C iiappe

Granite intrusron

/

tided and subducted beneath proto-Japan

310 Y. Isoxaki

Age ( M a )

Formation of high-PIT sandwich

(tectonic exhumation)

Age of subducted

400 300 200 100 0 , I

450-40018 330-280Ma 230-210~a 100-70Ya

Renge I1 Sang"" Sanbsgawa 'engel

* * * * -~ I

I / - - 0 1

,' (\ I , io

15

s labdeduced f rom OPS of AC

interacted oceanic plate

Ridge subduction

can be simplified to 450 million years of oceanward accretionary growth punctuated four times by ridge subduction. For each event, granites and associated low-P/T metamorphic belts formed on the continent side of Japan, probably separated by 200 km or more from the coeval high-PIT schist belt (Fig. 15). The nearly 100 million year period- icity in sandwich formation may correspond to the consumption of a major oceanic plate dissected by mid-oceanic ridges on both sides. The Japanese Islands experienced subduction of a t least five major oceanic plates (from young to old, Pacific, Kula-Izanagi, Farallon, an unnamed older oceanic plate, and the plate which originated during initial rifting; Fig. 16).

The anatomy of the subduction-related orogen in Japan and the tectonic interpretation of these structures strongly contradict earlier views of 'Cordilleran-type' orogeny. The most significant difference lies in the recognition of episodicity in orogenic activity related to ridge subduction. Pre- vious models for exhumation of high-P/T units assumed a steady-state and long-term process related to subduction of the normal ocean floor. For such an oceanic subduction-related orogenic process involving episodic culmination by ridge subduction, Isozaki and Maruyama (1 99 1) introduced a new term, 'Miyashiro-type orogeny'. The name is given after Akiho Miyashiro's outstanding contributions to subduction zone tectonics, particularly the first perception of paired metamorphic belts (Miyashiro 1961) prior to the birth of plate tectonics in the mid-l960s, and a keen perspective on ridge subduction and relevant geologic phenomena in the 1970s (Uyeda & Miyashiro 1974).

,' '\\ 7

- LOO ,' ?

~ 150

-- 200Ma- w e 7 Unnamed Farallon $ 1 Izanagi ,,*" Kula $ 1 P a c i f i c p

?nn 77n

SECONDARY MODIFICATION: MICROPLATE TECTONICS

The primary orogenic structures of the Japanese Islands are controlled by the spatial arrangement

Fig. 16 Timetable of ridge subduction and orogenic culmination in Japan indicated by formation of the high-P/T sandwich structure, age of subducted oceanic slab deduced from OPS analysis and paleoplate interaction based on hot spot track analysis (modified from Isozaki & Maruyama 1991) The names for oceanic plates that interacted with Japan are adopted from the reconstructed plate interaction in the Pacific domain by Engebret- son e t a / (1985) Note the temporary coinci- dence among the zero age of subducted oceanic plate at 75 M a exhumation of the high-P/T Sanbagawa metamorphic belts, and passage of KulaiPacific ridge off Japan

of main components (i.e. ancient AC, regional metamorphic rocks and granitic batholiths). Subduction-related orogens in Japan are characterized by subhorizontal piled nappes of AC with downward younging polarity, and by the episodic occurrence of a sandwich structure of high-PIT nappes and unmetamorphosed AC. The primary orogenic features, however, were modified or de- stroyed in some cases by secondary tectonic processes. In particular, microplate-related tectonics has the profound ability to secondarily reorganize primary structures (Miyashiro 1982). In Japan, there are three important secondary tectonic processes (Fig. 17; fore-arc sliver movement, back-arc basin opening, and arc-arc collision) that are all orogenic manifestations of microplate activities (Isozaki 1989; Isozaki & Maruyama 1991).

A fore-arc sliver is a decoupled microplate of a frontal arc driven by a strike-slip component of the oblique subduction of an oceanic plate. In the Japanese Islands, there are three active fore-arc slivers (the South Ryukyu sliver, Nankai sliver in southwest Japan, and Kurile sliver in East Hok- kaido; Fig. 1). Along-arc movement of fore-arc slivers can create three distinct features that de- stroy the primary orogenic edifice (i.e. along-arc strike-slip fault on are-side margin, and across-arc compressional structure in front, plus an extensional one in rear of the sliver). The best example of an along-arc strike-slip fault is the Quaternary Neo-M.T.L. in the Kii peninsula and on Shikoku Island that demonstrates a remarkably linear surface trajectory for more than 500 km (Figs 3,5). This right-lateral strike-slip fault was driven by the westward movement of the Nankai fore-arc sliver (Fig. 17b), cutting the older low-angle feature of the Tertiary paleo-M.T.L. (Isozaki 1989; Yamakita et al. 1995; Fig. 6). This westward sliver translation is also associated with across-arc compression along the Bungo Strait between Shikoku


(a) Arc-arc collisian

(b) Fore-arc sliver

Philippine Sea Plate

(c) Opening of back-arc basin I’ A\

Fig. 17 Three representative modes of secondary modification of primary orogenic structure related to microplate tectonics (modified from lsozaki & Maruyama 1991). (a) Formation of syntaxis in the older accretionary orogenic system and accretion of exotic arc crust occur at an arc-arc collision front around the Izu peninsula where the Izu arc penetrates almost perpendicularly to the Southwest Japan arc (b ) Formation of along-arc strike-slip fault (Neo-M T.L ) and across-arc extensional and compressional structure by along-arc movement of the Nanakai fore-arc sliver (c) Formation of along-arc extensional structure (Miocene rifted basins with bimodal volcanism), compressional structures (thrusting along paleo-M.T L ) and the strike-slip fault (T T.L.) occurred when the Japan segment detached from mainland Asia through the rifting-opening of the Japan Sea. Note these Neogene to Quaternary structures have modified considerably the pre-existing major structures of the ca 450 Ma accretionary orogen

and Kyushu Islands, and across-arc extension in Ise Bay. Across-arc compression related to a fore- arc sliver is best observed in the elevated Hidaka mountains in central HoIkkaido. Kimura (1985) explained the exposure of lower crustal rocks of the Hidaka belt (Komatsu et al. 1989) as a tectonic manifestation of westward frontal collision of the Kurile fore-arc sliver to northeast Japan. Seismic reflection research (Ikawa et al. 1995) recently documented the crust-cutting detachment surface that dips eastward from the Hidaka main thrust (Fig. 11).

Back-arc spreading is another tectonic process that has occurred frequently in the western Pacific since the latest Mesozoic, and is also a powerful modifier of primary orogenic structures in Japan. For example, when the Japan Sea opened in the Miocene, it split pre-existing orogenic structures into several blocks (Figs 9,10,14). Although a cer-

tain amount of rotation was involved, the opening of the Japan Sea basin has been attributed to dislocation of a pair of north-south running strike- slip faults along the basin margin (Fig. 17c). The chaotic alignment of belts in northeast Japan is primarily due to the series of left-lateral strike- slips along the eastern margin of the Japan Sea. The T.T.L. presently dividing southwest Japan and northeast Japan, and the parallel Futaba and Hata- gawa faults are typical examples of an eastern margin strike-slip fault that obliquely dissects an older along-arc zonal arrangement. In contrast, a segment of the right-lateral strike-slip fault marking the western margin in east Korea was recently noted (Yoon & Chough 1995). In addition, a domain of compression in the fore-arc may have been associated with back-arc spreading (Fig. 17c). It is difficult to explain the side-by-side juxtaposition of the low-P/T Ryoke metamorphic belt and high-PIT

312 Y. Isoxaki

Sanbagawa metamorphic belt in modern southwest Japan (Fig. 5 ) . These belts should have been separated from each other by a t least 100-200 km when they formed in the Late Cretaceous arc- trench system (Fig. 18a). The origin of the low- angle Paleo-M.T.L. between these two belts (Fig. 6) may be related to development of a horizontal detachment surface in the arc crust and to trenchward dislocation of the upper crust along

this surface, probably by back-arc spreading (Isozaki & Maruyama 1991; Fig. 18b). For further details on the opening of the Japan Sea, refer to Jolivert et al. (1994), Otofuji (1996) and Yamash- ita et al. (1996). The present-day opening of the Okinawa trough (Kimura et al. 1988) is still in a nascent stage of back-arc spreading (Fig. 1). This extensional regime appears to propagate northward into mid-Kyushu (note the Beppu-Shimabara

6 c usually 100-200 k m (100-701 a? ca 5 0 k m ___b.

Akryorhr-Sangun A B C pre-Juraralc

S posslble external llmlt of

volc. tront trench

-

-

h i g h - P / T Sanbagawa met. Mlno- lanba

Juraralc complex

100 k m

subducted Kula-Paclflc

rldge

Orlglnal conflguratlon of the Cretaceous

palred metamorphlc belta in SW Japan

p i~ - n'B' = Intra-arc shortenlng

1 = dlslocatlon 01 delslchmenl Iault I 130-40 Ma] fore-arc contraction 1

A ' , B' C'

100 Ma trench 100 Ma voic front

trench Paleo-MTL volcanlc [rant

Shlmanto Palaogana complex

back -arc

extension

mlsalnp volume of the Ryokc fore-arc rcplon

= accreted volume 0 1 the Shlmanto complex

Secondary juxtapositlon of hlgh-P/T and low-P/T motamorphlc belts

Fig. 18 Schematic cross-section showing the primary configuration of the Cretaceous arc-trench system in southwest Japan (a) and secondary modification

involving intra-arc shortening associated with back-arc (Japan Sea) spreading (b) Note the juxtaposition of the Cretaceous paired metamorphic belts induced by the trenchward translation of upper arc-crust along an intracrustal horizontal detachment surface


tica, and eastern Australia, as the Paleo-Pacific ocean floor was subducted along these continental margins after the 750-700 Ma breakup of Rodinia. Reconnaissance studies (Sedlock & Isozaki 1990; Isozaki et al. 1992; Isozaki & Blake 1994) including a compilation of previous works for the Klamath-Franciscan belt in California (Isozaki & Maruyama 1992; Maruyama et al. 1992) suggests fundamental similarities between the geotectonic evolution of California and that of Japan. I t appears promising to extrapolate the above-described new aspects and analyzing schemes for ancient AC to other orogenic belts of different time-space co-ordinates in the Earth’s history, including the Precambrian.

graben with active Aso and Unzen volcanoes), accordingly, Kyushu Island, now composed of several geotectonic units, may eventually split into two.

Arc-arc collisions are the third process that can modify the primary geotectonic structures in Ja- pan. The ongoing collision of the Izu-Bonin arc against the southwest Japan arc is a typical example of this process (Fig. 17a). The northward buoyant subduction of an intra-oceanic arc can indent into a pre-existing orogenic structure, leaving a clear V-shaped mark called ‘orogenic syntaxis’, in the west of Tokyo (Fig. 4). In addition, the accretion of arc crust basement has been achieved in a step-wise manner, adding a significant volume in central Japan (Taira et al. 1989). Several tectonic blocks of intra-oceanic arc origin around Mt. Fuji (Amano 1986) thus represent bona f ide ‘allochthonous or exotic terranes’ in Japan.

CONCLUSION

The Japanese Islands represent a segment of a Phanerozoic subduction-related orogen developed along the western margin of the Pacific Ocean. Significantly, this orogen has grown oceanward nearly 400 km during the last 450 million years. The anatomy of the Japanese orogen, best preserved in southwest Japan, is characterized by a subhorizontal piled nappe structure, which involves multiple AC nappes including those of high-P/T metamorphosed AC. Orogenic growth is unusual in that it includes: (i) step-wise growth of AC units; (ii) tectonically downward younging polarity; and (iii) intermittent sandwich structure of high-PIT nappes. Episodic subduction of mid-oceanic ridges ( =migration of TTR triple junctions) appears to explain the episodic exhumation of high-P/T nappes and the formation of granite/low-P/T metamorphic belts every 100 million years. Micro- plate tectonic processes such as the movement of fore-arc slivers, back-arc spreading and arc-arc collisions, secondarily modified the primary structure of Japan.

In order to establish a general model for subduction-related orogenic processes, the new geotectonic model for the origin of Japan needs to be tested on other orogens formed in similar tectonic environments. For example, the occurrence of a subduction-related orogen characterized by a similar anatomy would be expected in Circum- Pacific Phanerozoic orogens in western North America, western South America, Western Antarc-

ACKNOWLEDGEMENTS

The author would like to thank S. Maruyama, Y. Nishimura, T. Itaya, T. Matsuda and many stu- dents from Yamaguchi University for a decade- long discussion in various tectonic aspects of the Japanese Islands. Sincere thanks are also due to A. J. Kaufman, S. Maruyama, G. Kimura, and L. Ivany who constructively reviewed the manu- script. P. F. Hoffman gave valuable comments on the neo Proterozoic distribution of continental blocks.

REFERENCES ARAI S. 1980. Dunnite-harzburgite-chromite complexes

as refractory residue in the Sangun-Yamaguchi zone, western Japan. Journal of Petrology 21, 141-65.

AMANO K. 1986. Southern Fossa Magna as multiple collision belt. Chikyu (Earth Monthly) 8 , 581-85 (in Japanese).

BANNO S. & SAKAI C. 1989. Geology and metamorphic evolution of the Sambagawa belt, Japan. In Daly J. S., Cliff R. A. & Yardley B. W. D. eds. Evolution of Metamorphic Belts, Geological Society of London Special Publication 43, 519-32.

BERGER w. H. & WINTERER E. L. 1974. Plate stratigraphy and the fluctuating carbonate line. In Hsu K. J. & Jenkyns H. eds. Pelagic Sediments on Land and Under the Sea. International Association of Sedimentology, Special Publication 1, 11-48. Black- well Science, Oxford.

BOND G. C., NICKESON P. A. & KOMINZ M. A. 1984. Breakup of a supercontinent between 625 Ma and 555 Ma: New evidence and implications for continental histories. Earth and Planetary Science Letters 70, 325-45.

BURCHFIEL B. c., CHEN Z., HODGE K. V. et al. 1992. The southern Tibetan detachment system, Himalayan oro-

314 Y. Isoxaki

gen: Extension contemporaneous with and parallel to shortening in a collisional mountain belt. Geological Society of America Special Paper 269, 1-41.

BURRETT C., LONG J . & STRAIT B. 1990. Early-Middle Paleozoic biostratigraphy of Asian terranes derived from Gondwana. I n McKerrow W. S. & Scotese C. R. eds. Paleozoic paleogeography and biogeography. Geological Society of London Memoir 12, 63-174.

CARIDROIT M., CHARVET J. & ICHIKAWA K. 1985. The ultra-Tamba zone, a new unit in the Inner Zone of Southwest Japan: Its importance in the nappe structure after the example of the Maizuru area. Chikyu- Kagaku 39, 210-19.

CHARVET J., FAURE M. & GUIDI A. 1985. Some tectonic and tectogenetic aspects of SW Japan. An alpine-type oregen in a n island-arc position. hi Nasu N., Koba- yashi K., Uyeda S., Kushiro I. & Kagami H. eds. Formatioil of Actiz9e O c e m Margins, pp. 791-817. Terra Science Publications, Tokyo.

CLOOS M. 1982. Flow melange, Numerical modelling and geologic constraints on their origin in the Fran- ciscan subduction complex, California. Geological So- ciety of America Bulletin 93, 330-45.

CONEY P., JONES D. L. & MONGER J. W. H. 1980. Cordilleran suspect terranes. Nature 288, 329-33.

CONG B. & WANG Q. 1995. Ultra-high-pressure metamorphic rocks in China. Episodes 18, 91-94.

DALZIEL I. W. D. 1992. On the organization of American plates in the Neoproterozoic and breakout of Lauren- tia. G S A Today 2, 237, 240-41.

DEWEY J. F. 1980. Episodicity, sequence, and style at convergent plate boundaries. Geological Associattow of Canada, Special Paper 20, 553-73.

DEWEY J. F. & BIRD J. M. 1970. Mountain belts and the new global tectonics. Journal of Geophysical Re- search 75, 2625-47.

ENGEBRETSON D., Cox A. & GORDON R. G. 1985. Relative plate motions between oceanic and continental plates in the Pacific basin. Geological Society of America Special Paper 206, 1-59.

ERNST W. G., CAO R. L. & JIANG J. Y. 1988. Reconnais- sance study of Precambrian metamorphic rocks, northeastern Sino-Korean shield, Peoples’s Republic of China. Geological Society of America Bulletin 100, 692-701.

FARRAR E. & DIXON J. M. 1993. Ridge subduction: Kinematics and implications for the nature of mantle upwelling. Canadian Journal of E a r t h Sciences 30, 893-907.

FAURE M. 1985. The pre-Cretaceous structure of the Outer Belt of Southwest Japan. Tectonics 5 , 1089- 114.

GEOLOGICAL SURVEY OF JAPAN 1992. G e o l o g ~ a l At las of Japan 2nd edn. Asakura Shoten, Tokyo.

HADA S. & KURIMOTO C. 1990. Northern Chichibu Tw- rane. I n Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. 1990. Pre-Cretaceous Terranes of Japan, pp. 165-183. Publication of IGCP #224, Osaka.

HAMILTON W. B. 1991. On terrane analysis. In Dewey J . F., Gass I. G., Curry G. B., Harris N. B. W. & Sengor A. M. C. eds. Allochthonous Terranes, pp. 55-65. Cambridge University Press, Cambridge.

HARA I., HIDE K., TAKEDA K., TSUKUDA E., TOKUDA M. & SHIOTA T. 1977. Tectonic movement in the Sam- bagawa belt. In Hide K. ed. The Sambagawa Belt, pp. 309-90. Hiroshima University Press, Hiroshima (in Japanese with English abstract).

HAYASAKA Y. 1987. Study on the late Paleozoic-early Mesozoic tectonic development of western half of the Ineer Zone of Southwest Japan. Report of Earth Sciences, Hiroshima University 27, 119-204.

HAYASAKA Y. 1990. Maizuru Terrane. I n Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre- Cretaceous Terranm of Japan, pp. 81-95. Publica- tion of IGCP #224, Osaka.

HIROI Y. & KISHI. 1989. P-T evolution of Abukuma metamorphic rocks in north-east Japan: Metamorphic evidence for ocean crust obduction. I?? Daly J . S., Cliff R. A. & Yardley B. W. D. eds. Eiioli*.tion of Metamor- p h ic Belts, Geological Society of London, Special Publication 43, 481-86.

HOFFMAN P. F. 1991. Did the breakout of Laurentia turn Gondwanaland inside-out? Science 252, 1409-12.

HOWELL D. G. (ed.) 1985. Tectonostratigraphic Ter- mnPs of the Circum-Pacific Region. Circum-Pacific Council Energy and Mineral Resources, Houston.

ICHIKAWA K. 1984. Tectonic development of the basement structure of Eastern Asia. I. Iu Huzita K. ed. Tectonic Belts of A s i a , pp. 223-38, Kaibundo, Tokyo (in Japanese).

ICHIKAWA K., MIZUTANI S., HARA I., HADA S. & YAO A. (eds) 1990. Pre-Cretaceous Terraxes qf Japan. Pub- lication of IGCP #224, Osaka.

ICHIKAWA K., SHIMAZU M. & FUJITA Y. (eds) 1970. Geologic Developmerit of the Japanese Islands. Tsukiji-shokan Publications, Tokyo, (in Japanese).

IGI s., HATTORI H. & SHIBATA K. 1979. Nomo metagabbro complex and their 450 m.y. ages as a clue t o the basement geology - a proposal of ‘Saihi Structural zone’ including the pre-Silurian basements in west- ernmost part of the Japanese Islands. In Kano, H. ed., The basement of the Jupunew Islalzds, 261-80 (in Japanese with English abstract).

IGO H. 1990. Paleozoic strata in the Hida ‘Gaien’ belt. In Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre-Cretaceous Terranes of Japan, pp. 41-8. Publication of IGCP #224, Osaka.

IKAWA T., ARITA K., I T 0 T. et al. 1995. Vibroseis profiling and tectonics of the Hidaka collision zone. Abstracts 102nd Annual Meeting of Geological Socwty cf Japan, pp. 241 (in Japanese).

ISHIGA H. 1990. Ultra-Tamba terrane. Irl Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre- Cretaceous Tewanes of Japan, pp. 97-107, Publica- tion of IGCP #224, Osaka.

ISHIWATARI A., IKEDA Y. & KOIDE Y. 1990. The Yakuno

Anatomy and wolution of Japanese Islands 315

during cooling and deformation. Contributions to Mineralogy and Petrology 100, 281-90.

JOLIVERT L., TAMAKI K. & FOURNIER M. 1994. Japan Sea, opening history and mechanism: A synthesis. Journal of Geophysical Research 99B, 22 237-59.

Wrangellia: A displaced terrane in northwestern North America. Canadian Jownal of Earth Sciences 14, 2565-77.

KABASHIMA T., ISOZAKI Y. & NISHIMURA Y. 1993. Find of boundary thrust between 300 Ma high-P-T type schists and weakly metamorphosed Permian accretionary complex in the Nagato tectonic zone, South- west Japan. Journal of Geological Society of Japan 99, 877-80 (in Japanese).

KANMERA K. 1976. Correlation between modern and past geosynclinal sediments I, 11. Kuguku 46, 284- 91, 371-78 (in Japanese).

KANMERA K. 1980. Paleozoic to Paleogene tectonic evolution. In Kanmera K., Hashimoto M. & Matsuda T. eds. Geology of Japan, pp. 325-350. Iwanami Publications, Tokyo (in Japanese).

KANMERA K., SANO H. & ISOZAKI Y. 1990. Akiyoshi Terrane. In Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre-Cretaceous Terranes of Japan, pp. 49-62. Publication of IGCP #224, Osaka.

KATO M. 1990. Paleozoic corals. In Ichikawa K., Mizu- tani S., Hara I., Hada S. & Yao A. eds. Pre- Cretaceous Terranes of Japan, pp. 307-12. Publica- tion of IGCP #224, Osaka.

KAWAMURA M., KATO M. & KITAKAMI PALEOZOIC RE- SEARCH GROUP 1990. Southern Kitakami terrane. In Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre-CretacPous Terranes of Japan, pp. 249-66. Publication of IGCP #224, Osaka.

KAWATO K., ISOZAKI Y. & ITAYA T. 1991. Geotectonic boundary between the Sanbagawa and Chichibu belts in central Shikoku, Southwest Japan. Journal of Geological Society of Japan 97, 959-75 (in Japanese with English abstract).

KIMURA G. 1985. Oblique subduction and collision. Fore- arc tectonics of the Kuril Arc. Geology 14, 404-7.

KIMURA G. 1996. Collision orogeny at arc-arc junctions in the Japanese Islands. The Island Arc 5, 262-75.

KIMURA G. & MARUYAMA S. Plume-related ophiolites in Japan and episodic activity of the southern acific superplume. The Island Arc (in press).

KIMURA M., UYEDA S., KATO Y. et al. 1988. Active thermal mounds in the Okinawa Trough back-arc basin, Japan. Tectonophysics 145, 319-24.

KINOSHITA 0. 1995. Migration of igneous activities related to ridge subduction in Southwest Japan and East Asian continental margin from the Mesozoic to the Paleogene. Tectonophysics 245, 25-35.

KOBAYASHI T. 1941. The Sakawa orogenic cycle and its bearing on the origin of the Japanese Islands. Jour- nal of Faculty of Science, Imperial University of Tokyo, Section 2 5, 219-578.

JONES D. L., SILBERLING N. J. & HILLHOUSE J. 1977.

ophiolite, Japan. Fragments of Permian island arc and marginal basin rrust with a hot spot. In Malpas J. G., Moore E. M., Ponayiotou A. & Xenophontos C. eds. Ophiolites. Oceanic Crust Analogue, pp. 497- 506. Geological Survey Department, Nicosia.

ISOZAKI Y. 1989. Convergent processes of microplates: Application of major plate convergence to microplate tectonics. Report of Institute of Kuroshio Sphere, Kochi University 4, 59-75.

ISOZAKI Y. a. Contrasting two types of orogen in Permo-Triassic Japan: Accretionary versus collisional. The Island Arc (in press).

ISOZAKI Y. b. Jurassic accretion tectonics of Japan. The Island Arc (in press).

ISOZAKI Y. & BLAKE M. c‘. 1994. Biostratigraphic constraints on formation and timing of accretion in a subduction complex. An example from the Franciscan complex in northern California. Jowrnal of Geology 102, 283-96.

ISOZAKI Y., HASHIGUCHI T. & ITAYA T. 1992. The Kurosegawa klippe: An examination. Journal of Geo- logical Society of Japan 98, 917-41 (in Japanese with English abstract).

ISOZAKI Y. & ITAYA ‘I’. 1991. Pre-Jurassic klippe in northern Chichibu Belt in west-central Shikoku, Southwest Japan. Kurosegawa Terrane as a tectonic outlier of the pre-Jurassic rocks of the Inner Zone. Journal of Geological Society of Japan 97, 431-50 (in Japanese with English abstract).

ISOZAKI Y., ITAYA T. & KAWATO K. 1990. Metamorphic ages of Jurassic accretionary complex in the Northern Chichibu belt, Southwest Japan. Journal of Geologi- cal Society of Japan 96, 557-60 (in Japanese).

ISOZAKI Y. & MARUYAMA s. 1991. Studies on orogeny based on plate tectonics in Japan and new geotectonic subdivision of the Japanese Isands. Journal of Geo- graphy 100, 697-761 (in Japanese with English abstract).

ISOZAKI Y. & MARUYAMA S. 1992. New geotectonic subdivision of the Franciscan-Klamath region redefined by formation age of accretionary complexes. American Association of Petroleum Geology Bulle- tin 76, 423.

ISOZAKI Y., MARUYAMA S. & FURUOKA F. 1990. Ac- creted oceanic materials in Japan. Tectonophysics 181, 179-206.

ISOZAKI Y. & MATSUDA T. 1980. Age of the Tanba Group along the Hozugawa ‘anticline’, western hills of Kyoto, Southwest Japan. Journal of Geosciences, Osaka City University 23, 115-34.

ISOZAKI Y. & EISHIMUKA Y. 1989. Fusaki Formation, Jurassic subduction-accretiion complex on Ishigaki Island, southern Ryukyus and its geologic implication to Late Mesozoic convergent margin of East Asia. Memoir of Geological Socic7ty of Japan 33, 259-75 (in Japanese with English abstract).

ITAYA T. & TAKASUGI H. 1989. Muscovite K-Ar ages of the Sanbagawa schists, Japan and argon depletion

316 Y. Isoxaki

KOIKE T. 1979. Biostratigraphy of Triassic conodonts. I n Koike T. ed. Biostratigruphy of Permian and Triassic Conodonts and Holothurian Sclerites in Japan, pp. 21-77. Committee of Prof. M. Kanuma Retirement Ceremony, Tokyo (in Japanese).

KOMATSU M. 1990. Hida ‘Gaien’ belt and Joetsu belt. In Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre-Cretaceous Terranes of Japan, pp. 25-40, Publication of IGCP #224, Osaka.

S. 1989. Evolution of the Hidaka metamorphic belt, northern Japan. I n Daly J. S., Cliff R. A. & Yardley B. W. D. eds. Evolution of Metamorphic Belts, Geological Society of London Special Publication

KUROKAWA K. 1985. Petrology of the Oeyama ophiolite complex in the Inner Zone of Southwest Japan. Scientific Report of Ni igata liniversity, ser. E 6, 37-113.

LAN C-Y., LEE T., ZHOU X-H. & KWON S-T. 1995. Nd isotopic study of Precambrian basement of South Korea. Evidence for Early Archean crust? Geology 23, 249-52.

LI Z. X., ZHANG L. & POWELL C. McA. 1995. South China in Rodinia: part of the missing link between Australia-East Antarctica and LAurentia? Geology 23, 407-10.

MAEKAWA H. 1981. Geology of the Motai Group in the southwestern part of the Kitakami mountains. Jour- nal of Geological Society of Japan 87, 543-54 (in Japanese with English abstract).

MARUYAMA S. 1990. Denudation process of high- pressure metamorphic belt. Abstracts of 9 7th Aniiual Meeting of Geological Society of Japan, 484.

MARUYAMA S. 1994. Plume tectonics. Jourraal of Geo- logical Society of Japan 100, 24-49.

MARUYAMA S. Pacific-type orogeny revisited: Miyashiro- type orogeny proposed. The Island A r c (in press).

MARUYAMA S., ISOZAKI Y., TAKEUCHI R., TOMINAGA N., TAKESHITA H. & ITAYA T. 1992. Application of K-Ar mapping method on the Franciscan Complex in Northern California and redefined subdivision and boundaries. American Association of Petroleum Geo- logy Bulletin 76, 425.

MARUYAMA s., LIOU, J. G. & ZHANG R. 1994. Tectonic evolution of the ultrahigh-pressure (UHP) and high- pressure (HP) metamorphic belts from central China. The Island A r c 3, 112-21.

MARUYAMA S. & SEN0 T. 1986. Orogeny and relative plate motions: Example of the Japanese islands. Tectonop hysics 127, 3 05-29.

MATSUOKA A. 1992. Jurassic-Early Cretaceous tectonic evolution of Southern Chichibu terrane, southwest Japan. Paleogeography, Paleoclimatology, Paleoecol-

MATSUDA T. E. & ISOZAKI Y. 1991. Well-documented travel history of Mesozoic pelagic chert: From mid-

KOMATSU M., OSANAI Y., TOYOSHIMA T. & MIYASHITA

43, 487-93.

ogy 96, 71-88.

oceanic ridge to subduction zone. Tectonics 10, 475- 99.

MATSUDA T. E. & UYEDA S. 1971. On the Pacific-type orogeny and its model: Extension of the paired metamorphic belts concept and possible origin of marginal seas. Teetonophysics 11, 5-27.

MINATO M., GORAI M. & FUNAHASHI M. (eds) 1965. The Geologic Development of the Japanme Islands. Tsu- kiji Shokan, Tokyo.

MINOURA K. 1990. The pre-Cretaceous geology and tectonics of Northern Kitakami region. I n Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre-Cretaceous Terrunes of Japayi, pp. 267-79. Pub- lication of IGCP #224, Osaka.

MIYASHIRO A. 1961. Evolution of metamorphic belts. Joatrnal of Petrology 2, 277-311.

MIYASHIRO A. 1982. Theory of orogeny based on plate tectonics. I n Miyashiro A., Aki K. & Sengor A. M. C. eds. Orogeny, pp. 49-119. John Wiley, Chichester.

MIZUTANI S. 1987. Mesozoic terranes in the Japanese islands and neighbouring East Asia. In Leitch E. C. & Scheibner E. eds. Terrane accretion and orogexac belts, American Geophysical Union, Geodynamic Series 19, 263-73.

NAKAE S. 1993. Jurassic accretionary complex of the Tamba Terrane, Southwest Japan, and its formative process. Journal of Geosciences, Osaka Ci ty Univer- sity 36, 15-70.

NAKAJIMA T. Regional metmorphic belts of the Japanese Islands. The Island A r c 5 (in press).

NAKAJIMA T., SHIRAHASE T. & SHlBATA K. 1990. Along-arc variation of Rb-Sr and K-Ar ages of Cre- taceous granitic rocks in Southwest Japan. Contribu- tions to Mineralogy and Petrology 104, 381-9.

NAKASEKO K. (ed) Proceedings of the first Japanese radiolarian symposium. News of Osaka Micropaleoii- tologists Special Volume 5, pp. 485 (in Japanese with English abstract).

NISHIMURA Y. 1990. ‘Sangun metamorphic rocks’: Ter- rane problem. I n Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre-Cretaceous Terranes of Japan, pp. 63-79. Publication of IGCP #224, Osaka.

1989. Depositional age and metamorphic history of 220 Ma high P/T type metamorphic rocks. An example of the Nishiki-cho area, Yamaguchi Prefecture, Southwest Japan. Memoir of Geologscal Society of Japan 33, 143-66 (in Japanese with English abstract).

NISHIMURA Y., & SHIBATA K. 1989. Modes of occurrence and K-Ar ages of metagabbroic rocks in the “Sangun metamorphic belt”, Southwest Japan. Mem- oir of Geological Society of Japan 33, 343-57 (in Japanese with English Abstract).

OKAMI K. & EHIRO M. 1988. Review and recent progress of studies on the Pre-Miyakoan sedimentary rocks of the Northern Kitakami massif, Northeast

NISHIMURA Y., ITAYA T., ISOZAKI Y. & KAMEYA A.

Anatonzy and evolutioyi of Jupanese Islands 317

SUPPE J . 1974. Interrelationships of high-pressure metamorphism, deformation and sedimentation in Franciscan tectonics, U.S.A. Abstracts 24th Znterna- tional Geological Congress, 55 2-5 9.

SUZUKI H., ISOZAKI Y. & ITAYA T. 1990. Tectonic super- position of the Kurosegawa Terrane upon the Sanba- gawa Metamorphic Belt in eastern Shikoku, Southwest Japan. K-Ar ages of weakly metamorphosed rocks in northern Kamikatsu Town, Tokushima Prefecture. Journal of Geological Society of Japan 96, 143-53 (in Japanese with English Abstract).

SUZUKI K. & ADACHI M. 1993. 240 Ma CHIME ages of monazite and zircon from the Hirasawa granitic mass in the Southern Kitakami terrane. Journal of E a r t h & Planetary Sciences, Nagoya University 40, 1-16.

SUZUKI K. & ADACHI M. 1994. Middle Precambrian detrital monazite and zircon from the Hida gneiss on Oki-Dogo island, Japan: Their origin and implications for the correlation of basement gneiss of Southwest Japan and Korea. Tectonophysics 235, 277-92.

TAGIRI M. 1973. Metamorphism of Paleozoic rocks in the Hitachi district, southern Abukuma plateau, Japan. Science Reports of Tohoku University, Series 111 12, 1-67.

TAGIRI M., SAT0 H., MATSUMURA E. & NEMOTO H. 1993. Late Mesozoic low-P/high-T metamorphism preceding emplacement of Cretaceous granitic rocks in the Gosaisho-Takanuki district, Abukuma metamorphic belt. Thp Island A r c 3, 152-69.

TAIRA A., KATTO J., TASHIRO M., OKAMURA M. & KODAMA K. 1988. The Shimanto belt in Shikoku, Japan: Evolution of Cretaceous to Miocene accretionary prism. Modern Geology 12, 5-46.

TAIRA A., SAITO Y. & HASHIMOTO M. 1983. The role of oblique subduction and strike-slip tectonics in evolution of Japan. I n Hilde T. W. C. & Uyeda S. eds. Geodynamics of the Western Pacific-Indonesian regions. American Geophysical Union Geodynamic Series 11, 303-16.

TAIRA A., TOKUYAMA H. & SOH W. 1989. Accretion tectonics and evolution of Japan. I n Ben-Avraham Z. ed. The Evolution of the PaciJc Ocean Margins, pp. 100-23. Oxford University Press, New York.

TAKAMI M., ISOZAKI Y., ITAYA T. & NISHIMURA Y. 1990. Geochronology of weakly metamorphosed Ju- rassic accretionary complex (the Kuga Group) in eastern Yamaguchi Prefecture, Southwest Japan. Journal of Geological Society of Japan 96, 669-81 (in Japanese with English abstract).

TAKAMI M., ISOZAKI Y., NISHIMURA Y. & ITAYA T. 1993. Effect of detrital white mica and contact metamorphism to K-Ar dating of weakly metamorphosed accretionary complex: An example of Jurassic accretionary complex in eastern Yamaguchi prefecture, Southwest Japan. Journal of Geological Society of Japan 99, 545-63. (in Japanese with English abstract).

TAKASU A. & DALLMEYER R. D. 1990. 40Ar/3gAr min-

Japan. Chikyu-Kagaku 43, 187-201 (in Japanese with English abstract).

OSOZAWA S. 1984. Geology of the Amami Oshima, central Ryukyu islands, with special reference to effect of gravity transportation on geologic structure. Science Report of Tohoku IJniversity, 2nd series 54, 165-89.

OTOFUJI Y. 1996. Large tectonic movement of the Japan Arc in late Cenozoic times inferred from palleomag- netism: Review and synthesis. The Island A r c 5, 229-49.

OZAWA K. 1988. Ultramafic 1,ectonite of the Miyamori ophiolitic complex in the Kitakami mountains, North- east Japan. Hydrous upper mantle in a n island arc. Contribution of Minera1og:y and Petrology 99, 159- 75.

PARK J. K., BUCHAN K. L. & HARLAN s. s. 1995. A proposed giant radiating dyke swarm fragmented by the separation of Laurentia and Australia based on paleomagmatism of ca 780 Ma mafic intrusions in western North America. Earth and Planetanj Sci- erice Letters 132, 129-39.

PITMAN w. c. 111 & HAYES D. E. 1968. Sea-floor spreading in the Gulf of Alaska. Journal of Geophys- ical Research B73, 6571-80.

PLATT J. P. 1986. Dynamics of orogenic wedges and uplift of high-pressure metamorphic rocks. Geological Society of America. Bulletzn 97, 1037-53.

POWELL c. MCA., LI Z. X., MCELHINNY M. W., MEERT J. G. & PARK J. K. 1993. Paleomagnetic constraints on timing of the Neoproterozoic breakup of Rodinia and the Cambrian formation of Gondwana. Geology

SAITO Y. & HASHIMOTO M. 1982. South Kitakami Region: An allochthonous terrane in Japan. Journal of Geophysical Research 87, 3691-96.

SASAKI H. & ISOZAKI Y. 1992. Low-angle thrust between the Sanbagawa and Shimanto belts in central Kii peninsula, Southwest Japan. Journal of Geologi- cal Soiety of Japan 98, 57-60 (in Japanese).

SEDLOCK R. L. & ISOZAKI Y. 1990. Lithology and biostratigraphy of Franciscan-like chert and associated rocks in west-central Baja California, Mexico. Geological Society of America Bulletin 102, 852-64.

SENGOR A. M. C. & DEWEY J. F. 1991. Terranology. Vice or virtue. In Dewey J. F., Gass I. G., Curry G. B., Harris N. B. W. & Sengor A. M. C. eds. Allochonous Terranes, pp. 1-21. Cambridge University Press.

SOHMA T. & KUNUGIZA K. 1993. The formation of the Hida nappe and the tectonics of Mesozoic sediments: tectonic evolution of the Hida region, central Japan. Memoir of Geological Socit7ty of Japan 42, 1-20 (in Japanese with English abstract).

SOHMA T., KUNUGIZA K. & TERABATYASHI M. 1990. The Hida metamorphic belt. Excursion Guidebook, Geological Society of Jupan, 27-58 (in Japanese).

SHIBATA K. & NISHIMURA Y. 1989. Isotopic ages of the Sangun crystalline schists, Southwest Japan. Memoir of Geological Society of Jupan 33, 317-341.

21, 889-92.

318 Y. Isoxaki

era1 age constraints for the tectonothermal evolution of the Sanbagawa metamorphic belt, central Shikoku, Japan: A Cretaceous accretionary prism. Tectono- physics 185, 11-39.

TANAKA K. 1980, Kanoashi Group, an olistostrome, in the Nichihara area, Shimane prefecture. Journal of Geological Society of Japan 86, 613-28 (in Japanese with English abstract).

TAZAWA J. 1988. Paleozoic-Mesozoic stratigraphy and tectonics of the Kitakami mountains, northeast Ja- pan. Chikyu-Kagaku 42, 165-78 (in Japanese with English abstract).

THORKELSON D. J. 1994. Ridge subduction: Kinematics and implications for the nature of mantle upwelling: Discussion. Canadian Journal of Earth Sciences 31,

UJIIE H. & HASHIMOTO Y. 1983. Geology and radiolarian fossils in the inner-zone of ‘Motobu belt’ of Okinawa islands. Chikyu (Earth Monthly) 5 , 706-12 (in Japanese).

UYEDA S. & MIYASHIRO A. 1974. Plate tectonics and the Japanese islands. Geological Society of America Bulletin 85, 1159-70.

VON HUENE R. & LALLEMAND s. 1990. Tectonic erosion along the Japan and Peru convergent margins. Geo- logical Society of America Bulletin 102, 704-20.

VON HUENE R. & SCHOLLE D. W. 1991. Observation at convergent margins concerning sediment subduction, subduction erosion, and the growth of continental crust. Review of Geophysics 29, 279-316.

WAKABAYASHI J. 1992. Nappes, tectonics of oblique plate convergence, and metamorphic evolution related to 140 million years of continous subduction. Fran- ciscan complex, California. Journal of Geology 100, 19-40.

WAKITA K. 1988. Origin of chaotically mixed rock bodies in the Early Jurassic to Early Cretaceous sedimentary

1486-89.

APPENDIX: HISTORICAL REVIEW ON STUDIES OF OROGENY AND GEOTECTONIC SUBDIVISION OF THE JAPANESE ISLANDS

In order to establish a geotectonic subdivision for a particular region, conventional field mapping on a regional scale is essential, and this usually requires painstaking work by numerous geologists over an extended period of time, particularly in thickly vegetated regions like the Japanese Islands that are a t a humid mid-latitude. In the 100-year history of geological research in Japan, many efforts have been made to synthesize all available geological information into an overall tectonic framework. Although later criticized and discarded, several outstanding geotectonic syntheses were compiled, and these all came together with the revised summary of the geotectonic subdivision of the Japanese Islands. The history of geological studies of the

complex of the Mino terrane, central Japan. Bulletin of Geologacal Survey of Jupan 39, 675-757.

WANG X., LIOU J. G. & MAO H. K. 1989. Coesite-bearing eclogites from the Dabie mountains in central China. Geology, 17, 1085-8.

WILSON J. T. 1973. Mantle plume and plate motions. Tectonop hysics 19, 149-6 4.

YAMAKITA s., ITOH T., TANAKA H. & WATANABA H. 1995. Early Oligocene top-to-west motion along the Sashu fault, a low-angle oblique thrust of the Paleo- Median Tectonic Line, east Kyushu, Japan. Journal of Geological Society of Japan 101, 978-88 (in Japanese with English abstract).

YAMASHITA N. 1957. The Mesozoic era I. Geology booklet 10, Chidanken, Tokyo (in Japanese).

YAMASHITA s., TATSUMI Y. & NOHDA s. 1996. Tempo- ral variation in primary magma compositions in the NE Japan arc: its bearing on evolution of mantle wedge. The Island Arc 5 276-88.

YAMATO-OMINE RESEARCH GROUP, 1981. Paleozoic and Mesozoic rocks in central Kii mountains. Excursion Guidebook, 35th Annual Meeting of Chidanken, pp. 88 (in Japanese).

YAO A., MATSUDA T. & JSOZAKI Y. 1980. Triassic and Jurassic radiolarian assemblages from the Inuyama area central Japan. Journal of Geosciences Osaka City University 23, 135-55.

YOON S. H. & CHOUGH S. K. 1995. Regional strike slip in the eastern continental margin of Korea and its tectonic implications for the evolution of Ulleung Basin, East Sea (Sea of Japan). Gdogical Society qf America Bulletin 107, 83-97.

YOSHIKURA S., HADA S. & ISOZAKI Y. 1990. Kurosegawa Terrane. In Ichikawa K., Mizutani S., Hara I., Hada S. & Yao A. eds. Pre-Cretaceous Terranes of Japan, pp. 185-201, Publication of IGCP #224, Osaka.

Japanese Islands can be roughly divided into four stages, based on the orogenic concept that governed the understanding of contemporary geologists; these are: (i) the pre-geosyncline concept stage (1865-1940); (ii) the stage of importing the geosyncline concept (1941-1955); ( 5 ) the stage of popularization of geosyncline-based orogeny (1956-1975); and (iv) the stage of plate tectonics- based orogeny (1976-present).

The first stage is represented by activities of E. Naumann, B. Lyman and other foreign geologists who imported modern geology as well as other sciences to Japan immediately after the Meiji revo- lution in 1868, after two century-long diplomatic isolation. During this stage, foreign and domestic geologists recognized fundamental structures of the Islands, particularly major tectonic boundaries such as the Median Tectonic Line (M.T.L.) and Itoigawa-Shizuoka Tectonic Line (I.-S. T.L.); how-


the importance of distinguishing orogenic phase in H. Stille’s sense. His summary was the first application of the then world-popular geosyncline concept to a terra incognita named Japan. His achievement, however, marks a prime milestone in geological studies in Japan and his subdivision provided a foundation for later works. One point to note is his emphasis on nappe structures through- out southwest Japan, as this idea was revived in the 1990s in the plate tectonic framework.

The third stage was governed by geologists of the post-World War I1 generation who strongly criticized Kobayashi’s summary. They emphasized the higher resolution and accuracy of their data set through detailed mapping and the advantage of their new tectonic models (Yamashita 1957: Minato et al. 1965: Ichikawa et al. 1970), however, their perspective was more or less the same as that of Kobayashi, that is, popularization of the classic geosynclinal concept with minor modification in the context of regional geology. Although the world according to fixism was still there, the quality of the regional geological description of the Japanese Islands improved much during this stage, with the use of megafossil dating, particularly fusulinid biostratigraphy. Regional distribution of geotectonic units and locations of mutual boundaries were clarified, and the geotectonic subdivision made in this stage appears very similar to the present one in map view except for ‘vertical ’ major boundary faults (Fig. 19b). Accordingly, sporadically exposed older mid-Paleozoic rocks, including high-grade metamorphic rocks and serpentinites, were all explained as geosynclinal basement rocks squeezed out from deeper levels along putative ‘crust- penetrating vertical faults’. In the early 1970s, several avant-garde geologists from Japan were the first to propose plate tectonic interpretations for the evolution of the Japanese Islands, with special emphasis on subduction-related tectonics (Matsuda & Uyeda 1971: Uyeda & Miyashiro 1974). The majority of scientists, however, still held conservative understandings linked to the geosyncline concept, and the geotectonic subdivision was not revised from the plate tectonic viewpoint.

The most significant reform in the geotectonic subdivision occurred in the fourth stage. Nearly a decade after the fundamental construction of plate tectonics in the late 1960s, younger Japanese geologists started to prefer mobilism rather than fixism. Down a long and winding road with many arguments, including the conversion from a geosyncline world to a plate tectonics world, both in individuals and in society (Kanmera 1976), re-

ever, the overall geotectonic subdivision of the Islands was still a rough sketch. In hindsight, this era up to the 1930s in Japan may be called a time of fundamental ‘find-and-describe’ in preparation for the following stage of importing the concept of the geosyncline.

During the second stage, T. Kobayashi was the first Japanese geologist to synthesize a grand view of the geotectonic evolutim of the Islands on the basis of the stratigraphic and megafossil age data (Kobayashi 1941). He recognized most of the important geotectonic units currently known, but his subdivision (Fig. 19a) was strongly biased by the geosyncline concept. In particular, he emphasized

Akiyoshi Orogen Sakawa Orogen ~~~~~~

M T L

I c t o S c a l e l Opa Nappa

:b ) l ch ikawa era / 1970

N Shimanto Orogen S ~- Honrhu Orogcn ___ Cp Mz MT Ry Sb Ch Hdk Mr

k m M T L [i 30

. , 5 0 km, ,

Fig. 19 Classic examples of geotectonic map and profile of the Japanese Islands (modified from (a): Kobayashi 1941, (b): lchikawa et a/. 1970). Compare these classic geotectonic subdivisions and profiles backgrounded by geosynclinal viewpoint with the current version shown in Figs 3,4 Before 1975, major geotectonic units in Japaii were all explained as Precambrian sialic basement and overlying Paleozoic to Mesozoic geosynclinal sediments. Hd. Hida b , Ok: Oki b . , Cg: Chugoku b., Mz, Maizuru b., MT: Mino-Tanba b.; Ry: Ryoke b.; Sb: Sanbagawa b.; Ch: Chichibu b.; Hdk. Hidakagawa b. ( = Sh: Shimanto b.) ; Mr. Muro b . ( = NK. Nakamura b.); Km: Kitakami marginal b.; KI: Kitakami b ~ Sm. Soma b.; Ab: Abukuma b., As: Ashio b.; Jo: Joetsu b.

320 Y. Isoxaki

gional geologic information was constantly accumulated on a nationwide basis. The great reform came in two waves, the first one in the early 1980s and the second in the early 1990s.

The first wave was brought by two factors combined: (i) the remarkable enhancement in microfossil (conodont and radiolaria) biostratigraphy, in particular the high resolution dating of the pre- Cretaceous or ‘so-called Paleozoic eugeosynclinal sedimentary rocks’ in Japan (Koike 1979; Isozaki & Matsuda 1980; Tanaka 1980; Yao e t al. 1980, Yamato-Omine Research Group 198 1; Nakaseko et al. 1982); and (ii) the acceptance of a subduction- accretion concept (Kanmera 1980; Taira et al. 1983) which was constructed mainly through deep- sea drilling and regional seismic profiling on modern AC. As a result, most of the ‘so-called Paleo- zoic eugeosynclinal rocks’ in Japan were revealed to be Jurassic AC by the mid-1980s. Age of accretion for each AC unit was precisely dated, and this enabled mutual discrimination of neighboring and similar-looking Paleozoic to Cenozoic AC in Japan. This reform initiated a considerable redraft- ing of the geotectonic history of the Islands in terms of accretion tectonics, as well as their geotectonic subdivision. Utility of microfossil dating for ancient AC and the relevant results in geotectonic subdivision of Japan up to the late 1980s are summarized in Ichikawa et al. (1990).

On the other hand, concerning tectonic interpretation, the ‘allochthonous or suspect terrane’ concept (Jones et al. 1977; Coney et al. 1980; Howell 1985) invaded Japan in the early 1980s and allowed many geologists in Japan to believe the occurrence of exotic continental and/or oceanic blocks and to use the term ‘terrane’ to describe various orogenic units (Saito & Hashimoto 1982; Mizutani 1987; Ichikawa et al. 1990). On the geotectonic subdivision, the ‘terrane’-based understanding of the attitude of boundary faults between the orogenic units is of note because most of the ‘terrane boundaries’ were regarded as vertical faults of a strike-slip nature related to ‘terrane dispersion’ (Taira et al. 1983). The nappe tectonics, on the contrary, was revived also in the early 198Os, and its significance with subhorizontal boundary faults was emphasized by French and domestic geologists (Hara et al. 1977; Yamato-Omine Research Group 1981; Faure 1985; Charvet et al. 1985; Hayasaka 1987). French geologists, in particular, interpreted the nappe- related subhorizontal structure as a result of an ancient continent-microcontinent collision, however, evidence for the putative collided micro-

continent per se was not persuasive. Independent from these works, relative plate motions were partly reconstructed for the late Mesozoic and Cenozoic Pacific region by Engebretson et al. (1985), and the correlation between the plate interactions in East Asia and orogenic events re- corded in Japan was first discussed by Maruyama and Sen0 (1986). Owing to the mixed effects of these various interpretations and existence of problematic ‘gray zones’ mentioned below, understanding of the 3D structure and tectonic evolution of the Japanese orogen was in a state of confusion in the 1980s.

The second wave in the early 1990s that provided the final tool for redrawing the subdivision of the Islands came in the form of success in chronometric dating of weakly metamorphosed AC. AC well-recrystallized by regional metamorphism, such as the Sanbagawa blueschists, had already been dated by radiometric methods in the 1970s, however, less recrystallized metamorphic AC had not been dated a t all owing to difficulty in mineral separation techniques. Similarly, while microfossil analysis is powerful in dating non- to weakly metamorphosed AC, those metamorphosed to the greenschists facies were mostly left untouched owing to difficulty in microfossil extraction. Such a dilemma had left a considerable number of undated weakly metamorphosed AC units, the ‘gray zones’, in Japan even after the 198Os, and this delayed the completion of the geotectonic subdivision of the Japanese Islands. In the late 1980s, an advanced technique of fine-grained mineral separation was developed (Nishimura et al. 1989) that solved the ‘gray zone’ problem (Isozaki et al. 1990a, 1992; Isozaki & Itaya 1991; Suzuki et al. 1990; Takami et al. 1990; Kawato et al. 1991). Not only did this permit the dating of the greenschist facies AC rocks, but it allowed subdivision of high-PIT regional metamorphosed units on the same basis as non- to weakly metamorphosed ones (Isozaki & Maruyama 1991). By 1993, most of the important geotectonic boundaries were clearly defined by age difference and were re-examined in the field. This clarified that these boundary faults, that is, the nappes, are essentially subhorizontal, except for several vertical ones of secondary origin. In other words, there is no deep crust-cutting vertical fault zones nor a collage of exotic terranes in Japan. The predominance in subhorizontal structures of the Japanese Islands are currently accepted by many, and are utilized, for example, in the latest version of a geologic map of the islands issued by the Geological Survey of Japan (1992).

Documents

Anatomy and genesis of a subduction-related …ea.c.u-tokyo.ac.jp/earth/Members/Isozaki/96Anatomy.pdfAnatomy and evolution of Japanese Islands 291 /II \ N. American Piate Fig. 1 Plate