Orogenic Belts and Orogenic Sediment Provenance · rock complexes. Five types of such elongated orogenic domains are identiﬁed as the primary building blocks of composite orogens:

[The Journal of Geology, 2007, volume 115, p. 315–334] � 2007 by The University of Chicago. All rights reserved. 0022-1376/2007/11503-0004$15.00

315

Orogenic Belts and Orogenic Sediment Provenance

Eduardo Garzanti,1 Carlo Doglioni,2 Giovanni Vezzoli, and Sergio Ando

Laboratorio di Petrografia del Sedimentario, Dipartimento di Scienze Geologiche e Geotecnologie,Universita di Milano-Bicocca, Piazza della Scienza 4, 20126 Milano, Italy

(e-mail: [email protected])

A B S T R A C T

By selecting a limited number of variables (westward vs. eastward subduction polarity; oceanic vs. continental originof downgoing and overriding plates), we identify eight end-member scenarios of plate convergence and orogeny. Theseare characterized by five different types of composite orogenic prisms uplifted above subduction zones to becomesources of terrigenous sediments (Indo-Burman-type subduction complexes, Apennine-type thin-skinned orogens,Oman-type obduction orogens, Andean-type cordilleras, and Alpine-type collision orogens). Each type of compositeorogen is envisaged here as the tectonic assembly of subparallel geological domains consisting of genetically associatedrock complexes. Five types of such elongated orogenic domains are identified as the primary building blocks ofcomposite orogens: magmatic arcs, obducted or accreted ophiolites, neometamorphic axial belts, accreted paleomarginremnants, and accreted orogenic clastic wedges. Detailed provenance studies on modern convergent-margin settingsfrom the Mediterranean Sea to the Indian Ocean show that erosion of each single orogenic domain produces peculiardetrital modes, heavy-mineral assemblages, and unroofing trends that can be predicted and modeled. Five correspond-ing primary types of sediment provenances (magmatic arc, ophiolite, axial belt, continental block, and clastic wedgeprovenances) are thus identified, which reproduce, redefine, or integrate provenance types and variants originallyrecognized by W. R. Dickinson and C. A. Suczek in 1979. These five primary provenances may be variously recombinedin order to describe the full complexities of mixed detrital signatures produced by erosion of different types ofcomposite orogenic prisms. Our provenance model represents a flexible and valuable conceptual tool to predict theevolutionary trends of detrital modes and heavy-mineral assemblages produced by uplift and progressive erosionalunroofing of various types of orogenic belts and to interpret petrofacies from arc-related, foreland-basin, foredeep, andremnant-ocean clastic wedges.

Introduction

The overall geometry of orogenic belts producedby subduction … is fundamental for provenanceanalysis on a global scale. (Dickinson 1988, p.18)

Orogens are stacks of rock units uplifted above sub-duction zones by tectonic and magmatic processes,which are influenced by subduction geometry aswell as the nature and geological history of con-verging plates. Although distinct types of orogenicprisms may be identified, natural processes are sovaried and complex that any attempt to classifyorogenic belts sounds like a hopeless challenge;

Manuscript received June 15, 2006; accepted December 6,2006.

1 Author for correspondence.2 Dipartimento di Scienze della Terra, Universita La Sapi-

enza, Piazzale Aldo Moro 5, 00185 Roma, Italy; e-mail: [email protected].

each mountain chain has its own peculiarities andstands as a case apart.

As a consequence, modeling the provenance oforogenic sediments and sedimentary rocks is an ar-duous task, and a systematic quantitative treat-ment of orogenic sediment provenance is lacking.Classic provenance models have dealt with this in-tricate problem in only a general way, loosely dis-criminating among three distinct types of orogenicsettings (subduction complex, collision-orogen/su-ture belt, and foreland fold-thrust belt) and threecorresponding types of “recycled orogen prove-nances” (Dickinson and Suczek 1979; Dickinson1985).

This article focuses on topographically elevatedsources of detritus found within thrust belts andarc-trench systems (rather than on the associatedbasins representing sediment sinks; Dickinson

316 E . G A R Z A N T I E T A L .

1988) and proposes a simplified classification oforogenic domains that is intended to represent auseful reference model for a better description andunderstanding of orogenic sediment provenance.

Subduction Polarities and Orogen Types

Current circum-Pacific arcs include east-facingisland arcs and west-facing continental arcs ina consistent pattern that implies net westwarddrift of continental lithosphere with respect tounderlying asthenosphere. (Dickinson 1978, p.1)

Two opposite subduction polarities have long beenrecognized (Nelson and Temple 1972; Uyeda andKanamori 1979). Whereas back-arc spreading ischaracteristic of east-facing arc-trench systems(westward subduction), back-arc thrusting is mostwidespread in west-facing arc-trench systems (east-ward subduction; Dickinson 1978).

Such contrast has often been ascribed to the ageof subducting oceanic lithosphere, with older lith-osphere being denser and consequently prone togenerate a more efficient slab pull (Royden 1993).Systematic analysis of subduction dip and conver-gence rate at the trench, however, fails to show anysignificant correlation between age of downgoinglithosphere and slab inclination (Cruciani et al.2005).

Global tectonics is considered here as funda-mentally controlled by Earth’s rotation, with lith-ospheric plates drifting westward with respect tothe mantle (Bostrom 1971; Dickinson 1978; Scop-pola et al. 2006). Plate motions follow a sinusoidalflow, oriented roughly westward in the AtlanticOcean but turning west-northwestward in the Pa-cific Ocean and finally bending southwestward inAsia (Crespi et al. 2007). Referring to such main-stream of plate motion, we identify two end mem-bers of subduction zones, associated with two fun-damental types of orogens.

Westward subduction zones oppose mantle flowand tend to be steeper. Decollement planes are shal-low and affect only the upper layers of the down-going plate. Low-relief, thin-skinned, singly ver-gent orogens are produced, chiefly consisting ofvolcanic or sedimentary rocks (e.g., Lesser Antilles,Sandwich, Apennines, Carpathians, Banda, Tonga,Marianas, Nankai, Kurili, Aleutians; Doglioni et al.2006).

Instead, eastward subduction zones follow man-tle flow and are less inclined. Decollement planescut across the whole crust, allowing the exhuma-tion of deep-seated rocks. High-relief, thick-

skinned, doubly vergent orogens are produced(Koons 1990; Willett et al. 1993), largely consistingof neometamorphic and plutonic rocks (Andes,Alps, Caucasus, Zagros, Himalayas, Indonesia, Tai-wan, New Zealand Alps; Doglioni et al. 2006).

As a general rule, the subduction hinge movesaway from the upper plate in westward subductionzones and toward the upper plate in eastward sub-duction zones. This dichotomy corresponds withthe distinction between “pull-arc” and “push-arc”orogens (Laubscher 1988). Note that the term “east-ward” is used loosely here to designate subductionzones that follow mantle flow, even if these areactually oriented northeastward in Eurasia (e.g.,Himalayas) because of the undulated mainstreamof plate motion (Crespi et al. 2007). If the net west-erly drift of lithospheric plates is controlled byEarth’s rotation, an astronomical mechanism thatcannot be inverted, this model should be valid alsofor the geologic past.

Geometries of Plate Convergence

If the present is the key to the past, perhapsglobal paleotectonic and paleogeographic recon-structions should be based on the actualistic hy-potheses that … backarc spreading occurs wherearc orogens face east, and that backarc thrustingoccurs where arc orogens face west. (Dickinson1988, p. 20)

Subduction involves a lower downgoing plate andan upper overriding plate, both of which can beeither oceanic or continental. By considering allpossible basic combinations between subductionpolariy (westward vs. eastward) and nature of con-verging plates (oceanic vs. continental), the typesof plate convergence are reduced to eight end mem-bers (fig. 1). The pro side of the orogen is the onefacing the subduction zone.

Westward Intraoceanic Subduction. Westward in-traoceanic subduction is widespread along thewestern Pacific (e.g., Philippines and Marianas) andoccurs in the western Atlantic (Lesser Antilles andSouth Sandwich Islands) and western Mediterra-nean Sea (Aeolian Islands). The age of the oceaniccrust is syn-subduction in the back-arc basin andgenerally much older in the lower plate. The arc-trench system formed above the subduction zoneincludes calc-alkaline igneous rocks (arc massif)and oceanic rocks scraped off the subducting plate(oceanic prism).

Westward Subduction of Oceanic Lithosphere be-neath Continental Lithosphere. This setting is typ-ical of newly formed westward subduction zones,

Figure 1. Eight true-scale schematic diagrams illustrate different styles of orogenic deformation for the eight iden-tified scenarios of plate convergence. Orogens are made of rocks accreted from the lower and/or upper plates; shadesof continental crust highlight upper plate (dark brown) versus lower plate (light brown) contributions. East-facing,singly vergent, and low-relief prisms largely consist of deformed lower-plate rocks (C, D). West-facing, doubly vergent,and high-relief orogens mostly consist of deformed upper-plate rocks in precollisional stages (F), lower-plate rocksbeing massively involved only during final continental collision (G, H). High-pressure neometamorphic rocks, ex-humed in west-facing orogens, are light blue. The profound global asymmetry between east-facing versus west-facingorogens and subduction zones can be fully appreciated only when the mainstream of plate motions is recognized.

318 E . G A R Z A N T I E T A L .

which may develop along the retro side of thick-skinned orogens generated by preexisting eastwardsubduction (e.g., initiation of Barbados subduction;Doglioni et al. 1999). Large crustal slices of the in-active orogen are boudinaged during progressingback-arc extension and dragged eastward while thesubduction hinge migrates away from the upperplate (e.g., Calabria and pre-Pleistocene of northernJapan; Doglioni et al. 1998).

The pro side of the orogen is an accretionaryprism, developed at the expense of the uppermostlayers of the subducting plate and largely repre-sented by oceanic sediments (e.g., Nankai Trough;Moore et al. 1990). Perhaps the best modern ex-ample is northern New Zealand, where the oceanicPacific Plate subducts beneath stretched continen-tal lithosphere (Henrys et al. 2006). The upper-platecrustal extension is propagating southward acrossthe low-topography North Island, unzipping thenascent Taupo back-arc basin (Parson and Wright1996; Beanland and Haines 1998). Subduction po-larity changes farther south, where South Islandis instead characterized by a doubly vergent, high-topography compressional orogen produced by con-tinental collision (Beaumont et al. 1996b).

Westward Subduction of Continental Lithosphere be-neath Oceanic Lithosphere. At the final stage of oce-anic subduction or laterally to an active oceanicsegment (e.g., Banda Arc), thinned continental lith-osphere may be pulled down to 100–250 km (Mullerand Panza 1986; Ranalli et al. 2000), until subduc-tion is eventually throttled by buoyant lithosphereof closer-to-normal thickness (e.g., Southern Apen-nines).

East-facing singly vergent orogens formed abovewestward subduction zones are characterized bylow structural relief (Doglioni et al. 1999). The proside of the orogen is a thin-skinned thrust belt (in-cluding sedimentary sequences originally depositedon a continental paleomargin and frontally accretedturbidites), whereas its retro side is a magmatic arc(better developed if subduction is faster; Tatsumiand Eggins 1995).

Because little sediment is produced by the low-relief orogen and tectonic subsidence is one orderof magnitude greater than for eastward subductions(Doglioni 1994), the sedimentary basin formed bothin front of and above (Ori and Friend 1984) thegrowing accretionary prism typically remainspersistently in deepwater conditions (“foredeep”).Foredeeps formed above westward-subducting con-tinental margins are contrasted here with less sub-sident foreland basins associated with high-reliefAlpine-type collision orogens, which are rapidly ov-

erfilled with shallow-marine to fluvial sediments(Doglioni 1994).

Westward Subduction of Continental Lithosphere be-neath Continental Lithosphere. This case differsfrom the one described above only because exten-sion on the retro side of the orogen is insufficientto tear the crust, and an ensialic back-arc basindevelops (e.g., Pannonian Basin in the rear of theCarpathians; Horvath 1993). This has no system-atic influence on the structure of the orogen and,thus, on terrigenous supply.

Eastward Intraoceanic Subduction. The few mod-ern examples include the Vanuatu (New Hebrides),where a doubly vergent intraoceanic prism has beenproduced by collision of the volcanic arc with oce-anic plateaus, submarine ridges, and seamounts(Taylor et al. 1995; Meffre and Crawford 2001). East-ward subduction also takes place along highlyoblique intraoceanic convergence zones beneaththe Macquarie Ridge (Massell et al. 2000) and theAndaman Sea, a young pull-apart oceanic basin(Curray 2005). The Andaman-Nicobar Ridge is atectonic stack of arc-derived and remnant-oceanturbidites capped by forearc ophiolites (Allen et al.,forthcoming). Ophiolites represent the only sub-aerial exposure of the Macquarie Ridge (Rivizzignoand Karson 2004).

Eastward Subduction of Oceanic Lithosphere beneathContinental Lithosphere. The classic example isthe eastern Pacific, bordered by a continuous high-relief cordillera from Alaska to the Andes (Jaillardet al. 2002). Because viscosity is much higher inthe lower oceanic plate, shortening is mostly con-centrated in the upper continental plate, and theorogen chiefly consists of upper-plate material (ta-ble 1). The lower plate may be subducted entirely,and even tectonic erosion commonly takes placebecause of this marked rigidity contrast (von Hueneand Lallemand 1990; Ranero and von Huene 2000).In contrast, accretion typically occurs where thelower plate is overlain by thick deep-sea turbidites(e.g., Alaska and Sumatra; Dickinson 1995; Inger-soll et al. 2003).

The pro side of the orogen is an arc-trench sys-tem, with a forebelt of variable width cored by base-ment rocks (von Huene 1986). The retro side is athick-skinned thrust belt also involving basementbut frequently propagating toward the foreland inthin-skinned mode (e.g., Rocky Mountains; Ballyet al. 1966).

The orogen generally undergoes compression anduplift, but extension may develop where motion ofthe lower plate has been inverted after subductionof a mid-ocean ridge (e.g., Basin and Range) or where

Journal of Geology O R O G E N I C S E D I M E N T P R O V E N A N C E 319

Table 1. Structural Features of Orogenic Belts Formed in the Eight Identified Scenarios of Plate Convergence

Lithosphere ExamplesSource of accreted

rock unitsProcess of prism

building Orogen type

Westward subduction:O/O Marianas, Sandwich Lower plate Oceanic offscraping Intraoceanic arcC/O N. New Zealand,

CalabriaInherited upper plate �

lower plateOceanic offscraping Boudinaged belt and prism

O/C Banda, S. Apennines Inherited upper plate �lower plate

Continental offscraping Boudinaged belt and prism

C/C Carpathians, N.Apennines

Inherited upper plate �lower plate

Continental offscraping Boudinaged belt and prism

Eastward subduction:O\O Andamans, Vanuatu Upper plate � lower plate Oceanic collision Intraoceanic prismO\C Andes, Cascades Mostly upper plate Continental buckling CordilleraC\O Oman, Papua Upper plate � lower plate Obduction Obduction orogenC\C Himalayas, Alps Upper plate � lower plate Continental collision Collision orogen

Note. Westward relief, thin-skinned, singly vergent, “pull-arc” orogens with rapid subsidence of trench orsubduction p lowforedeep; oceanic (or ensialic) back-arc basin. Eastward relief, thick-skinned, doubly vergent “push-arc” orogenssubduction p highwith slower subsidence of trench or foreland basin; no true back-arc basin. ; . Topographic relief,O p oceanic C p continentaldeformation style, and rock units involved are largely controlled by subduction polarity. The illustrated end-member orogens mayevolve dynamically through geologic time. When a continental margin arrives at the oceanic trench, pull-arc orogens (Laubscher1988) evolve from northern New Zealand type to Banda type (or Carpathian type), ophiolite-capped push-arc orogens from Andamantype to Oman type, and higher-topography push-arc orogens involving continental rocks in the hanging wall of the subduction zonefrom Andean type to Himalayan type.

distinct subplates override the lower plate at dif-ferent velocities (e.g., Aegean Sea; Doglioni 1995).When we consider the undulated mainstream ofplate motions, the Sumatra-Java arc belongs to thiscategory.

Eastward Subduction of Continental Lithosphere be-neath Oceanic Lithosphere. This setting representsthe final stage of eastward intraoceanic subduction,when a continental margin arrives at an intra-oceanic trench. Virtually intact slabs of dense oce-anic upper-plate lithosphere can thus be emplacedonto buoyant continental crust (“obduction”; Cole-man 1971; Karig 1982). Such a process gives rise to“ophiolite-capped thrust belts” (Cawood 1991), thearchetypal example being the Northern Oman oro-gen assembled at Late Cretaceous times (Searle andStevens 1984). The pro side of the obduction orogenis a thick-skinned thrust belt, beneath which con-tinental blueschists and eclogites of the axial beltmay be exposed. Its tectonic lid consists of a com-plete section of forearc mantle and crust, generatedsynchronously with subduction and detached alonga mechanically weak boundary as deep as the as-thenosphere (Spray 1984; Searle and Cox 1999).

Eastward Subduction of Continental Lithosphere be-neath Continental Lithosphere. This setting repre-sents the final stage of eastward oceanic subduc-tion, when two continental margins collide.Doubly vergent Alpine-type orogenic prisms withhigh structural and topographic relief are thus pro-duced (Doglioni et al. 1999). The axial part of theorogen includes slivers of strongly thinned outer

continental margins and adjacent oceanic litho-sphere that underwent high-pressure metamor-phism in the early subduction stage of collision(Beaumont et al. 1996a). Because thrust planes cutacross the lithosphere, deeply subducted eclogiticrocks can be exhumed in a few million years (Ru-batto and Hermann 2001; Baldwin et al. 2004).Thick-skinned thrust belts formed along both ex-ternal sides of the orogen include inner parts ofcollided paleomargins underlain by continentalcrust of closer-to-normal thickness as well as fron-tally accreted foreland-basin clastics. In externalbelts, close to the contact with the axial belt, oro-genic metamorphism may reach amphibolite facies(Frey and Ferreiro Mahlmann 1999).

Collision orogens are by no means all alike, andmajor differences are displayed by two of the best-known examples belonging to the same orogenicsystem, the Alps and the Himalayas. The Hima-layas have a much better developed arc-trench sys-tem, which can be traced for ∼3000 km from Pak-istan to Myanmar (Gansser 1980). The axial belt ofthe Alps includes a stack of both continental andoceanic nappes showing widespread high-pressuremetamorphism. In contrast, in the Himalayas, theaxial belt is confined to a narrow wedge largelyconsisting of amphibolite-facies metasediments ex-truded between a main thrust zone at the base anda major detachment system at the top (Hodges2000); oceanic units are lacking, and eclogites havebeen recognized only locally so far (Lombardo andRolfo 2002). The retro side of the Alps consists of


Figure 2. Detrital modes, heavy-mineral assemblages, and compositional trends for the five identified types ofprimary provenances. Shown are sample means of modern sand provinces in areas with arid climate and/or high relief(sources cited in text); arc-derived volcano-plutonic sands include 32 Quaternary circum-Pacific suites compiled byMarsaglia and Ingersoll (1992). Samples with transitional subprovenance are gray in all diagrams. ,Q p quartz F p

, grains (including Lc carbonate lithics; symbols have thick or very thick outlines iffeldspars L p lithic Lc/QFL% ≥or ≥50, respectively). and felsic metaigneous lithic grains; , metavolcanic,20 Lm p metasedimentary Lv p volcanic

and mafic metaigneous lithic grains; lithic grains (serpentinite, serpentine schist);Lu p ultramafic Ls plithic grains. ; ; transparentsedimentary A p amphiboles P � O � S p pyroxenes � olivine � spinel &tHM p other

heavy minerals. The 90% confidence regions about the mean were calculated after Weltje (2002). Provenance fieldsafter Dickinson (1985; arc, block, orogen); field of recycling afterMA p magmatic CB p continental RO p recycledGarzanti et al. (2006).

a well-defined, 50–100-km-wide, thick-skinnedthrust belt, whereas the deformed retro side of theHimalayas is much broader and includes the vastTibetan plateau as well as a series of highly elevatedmountain chains (e.g., Hindukush-Karakorum, Pa-mir, and Tian Shan). Both external belts of the Alpsinclude deep-sea turbidites ranging in age from Cre-taceous to Miocene, whereas foreland-basin clas-tics accreted along the front of the Himalayaschiefly include Neogene fluvial sediments (Bur-bank et al. 1996; Najman 2006).

Classification of Orogenic Sediment Provenances

Given the potential diversity of recycled oro-genic sediment, it is a severe challenge to devisea scheme for its identification and classificationthat has empyrical validity for interpretation ofthe sedimentary record. (Dickinson 1985, p. 347)

Because orogenic belts are composite structures in-cluding diverse rock complexes assembled in variousways by geodynamic processes, orogenic detritusembraces a varied range of signatures. Unravelingprovenance of clastic wedges accumulated in fore-land basins, foredeeps, or remnant-ocean basins is,therefore, an arduous task that requires a sophisti-cated conceptual reference scheme.

The Dickinson Model. Orogenic detritus deriveseither from volcano-plutonic rock suites generatedalong active intraoceanic and continental arcs(magmatic-arc provenance) or from mainly sedi-mentary or metamorphic rocks tectonically up-lifted within subduction complexes (subductioncomplex provenance), foreland fold-thrust belts(foreland uplift provenance), and collision orogens(collision orogen provenance; Dickinson and Suc-zek 1979). Subduction complex provenance is typ-ified by chert-rich detritus from offscraped oceanicslivers and abyssal-plain sediments, whereas fore-land uplift provenance is characterized by varied

sedimentary lithics, moderately high quartz, andminor feldspars and volcanic lithic grains. Sedi-ments from collision orogens also have typicallyintermediate quartz contents, high quartz/feldsparratio, and abundant sedimentary and metasedi-mentary lithic fragments (Dickinson and Suczek1979; Dickinson 1985).

Dickinson and Suczek (1979) and Dickinson(1985, 1988) recognized the intrinsically compositenature of detritus shed from orogenic belts into as-sociated sedimentary basins but did not attempt toestablish clear conceptual and operational distinc-tions among the three identified types of recycledorogenic provenances. For instance, subductioncomplex and collision orogen provenances mayboth include detritus from ophiolitic melanges, andcollision orogen and foreland uplift provenancesmay include detritus from magmatic arc remnantsin relationship with relief distribution and drainagepatterns (Dickinson and Suczek 1979, p. 2176–2178; Dickinson 1985, p. 350).

Observations from modern settings, where allfactors affecting sediment composition can be ver-ified, reveal discrepancies with the Dickinsonmodel. Ophiolitic detritus is only marginally dealtwith. Large, subaerially exposed subduction com-plexes chiefly consist of offscraped turbidites andthus typically shed abundant shale/slate grainsrather than chert (Garzanti et al. 2002a; E. Gar-zanti, data from the Indo-Burman Ranges). Detritusfrom remnants of continental paleomargins incor-porated within thrust belts shows close affinitieswith anorogenic detritus of continental block prov-enance (Garzanti et al. 2003, 2006). Collision oro-gens produce a wide range of lithic/quartzolithic toquartzofeldspathic signatures, including all possi-ble mixtures of both first-cycle and multicycle de-tritus from neometamorphic, paleometamorphic,plutonic, volcanic, and sedimentary rocks. Furtherproblems are caused by recycling of clastic wedges

322 E . G A R Z A N T I E T A L .

accreted at the orogenic front (Garzanti et al. 2002a,2004b, 2005).

As pointed out by Ingersoll (1990, p. 733), most ofthese complexities cannot be succesfully handledbecause “third-order fields (magmatic arc, recycledorogen, and continental block) are useful for conti-nental-scale analyses,” whereas “second-order fields(e.g., undissected, transitional, and dissected arcs ofDickinson 1985) are useful at the scale of mountainranges and basins.” In order to improve on the res-olution of provenance models, this article focuseson the structure of single orogenic belts rather thanon whole continents and ocean basins.

A Refined Two-Step Model for Orogenic SedimentProvenance. The scheme presented here is basedon the simple observation that the complicated tec-tonic structure of composite orogens results fromthe juxtaposition and superposition of a limitednumber of geological domains, each including ge-netically associated rock complexes elongated sub-parallel to the orogen’s strike. Because major riversreach well into the core of the composite orogen,sediments supplied to the associated basins are in-variably a mixture of detritus from different typesof such linear domains (Dickinson and Suczek1979). If mixed detrital signatures are too varied tobe modeled directly, then their complexity can behandled by operating in two steps. We focus firston detrital modes produced by each distinct typeof orogenic domain (primary orogenic provenances)and only subsequently recombine the appropriatetypes of primary provenances to model detritalsuites recorded in sedimentary basins (compositeorogenic provenances).

Five types of orogenic domains are identified hereas the primary building blocks of composite oro-genic prisms: (1) magmatic arcs (autochthonous orallochthonous sections of arc crust), (2) accreted orobducted ophiolites (largely intact allochthonoussections of oceanic lithosphere), (3) neometa-morphic axial belts (polydeformed slivers of distalcontinental margin crust and adjacent oceanic lith-osphere that have undergone high-pressure to high-temperature metamorphism during subduction andsubsequent exhumation), (4) paleomargin remnants(only weakly metamorphosed allochthonous sec-tions of continental basement and/or overlyingplatform to pelagic strata), and (5) orogenic clasticwedges (accreted foreland-basin, foredeep, or rem-nant-ocean-basin terrigenous sequences).

As shown by provenance studies in modern oro-genic settings from the Mediterranean Sea to theIndian Ocean, detritus produced by the erosion ofeach single orogenic domain is characterized byunique detrital modes, heavy-mineral assemblages

and unroofing trends, which can be predicted andmodeled. The five types of orogenic domains thuscorrespond to five types of primary sediment prov-enance (fig. 2).

Our scheme expands on the original Dickinsonmodel, with the following main modifications: (1)magmatic arc provenance is unchanged, (2) ophio-lite provenance is recognized as a new provenancetype, (3) axial belt provenance is defined as the mostdistinctive type of collision orogen provenance, (4)marked affinities between foreland uplift prove-nance (a type of orogenic sediment provenance inDickinson and Suczek 1979) and continental blockprovenance (the anorogenic sediment provenancein Dickinson and Suczek 1979) are emphasized, (5)clastic wedge provenance is newly defined, in orderto single out, and put emphasis on, the thorny prob-lem of sediment recycling. This latter provenancetype consists entirely of recycled orogen-deriveddetritus, whereas the former four types chiefly con-sist of first-cycle detritus (with the limited excep-tions of grains recycled from forearc basin strataexposed in arc domains or from sandstone-bearingpassive-margin strata intercalated within paleo-margin successions).

Such a moderate increase in complexity is heldas necessary and sufficient to appropriately handlethe full variety of cases observed along modern sub-duction zones. The proposed scheme is flexibleenough to reproduce the complete range of mixeddetrital signatures issued during erosional denu-dation of composite orogens and to predict the evo-lution of detrital modes and heavy-mineral assem-blages as recorded in space and time by clasticwedges deposited in arc-related, foreland, foredeep,and remnant-ocean basins (table 2).

Primary Provenances and Unroofing Trends

Data for modern marine and terrestrial sandsfrom known tectonic settings provide standardsto evaluate the effect of tectonic setting on sand-stone composition. (Dickinson and Suczek1979, p. 2164)

The information contained in this section is basedon provenance studies of modern sediments pro-duced and deposited in areas characterized mostlyby arid climate and/or high relief. The describedcompositional signatures can thus be considered asunaffected by significant chemical weathering anddiagenetic dissolution and to faithfully reflectplate-tectonic setting and lithology of source ter-ranes. Such primary detrital modes can be used asa reference for assessing provenance of ancient clas-

Table 2. The Complex Problem of Modeling Orogenic Provenance is Handled by Operating in Two Steps

Primary orogenicprovenances

Composite orogens

Indo-Burman-typesubduction complexes

Apennine-typethin-skinned orogens

Oman-typeobduction orogens

Andean-typecordilleras

Alpine-typecollision orogens

Magmatic arc Minor (largely recycled) Dominant locally (retro side) Minor locally (retro side) Dominant (pro side) Significant locally (early stage rundissected arc; late stage rdissected arc)

Ophiolite Significant locally (retro side) Dominant locally (subduc-tion complex remnants)

Dominant (early stage) Significant locally (pro side) Significant locally (early stage)

Axial belt Insignificant Dominant locally (boudi-naged remnants)

Significant locally (latestage)

Significant locally (pro side) Dominant (late stage)

Continental block Significant locally (incorpo-rated terranes)

Major Dominant locally (latestage)

Dominant (retro side) Major (late stage)

Clastic wedge Dominant (remnant-oceanturbidites)

Major (foredeep turbidites) Minor Major (retro side) Major (foreland-basin clastics)

Note. The diagnostic detrital modes produced by each distinct orogenic domain are first identified (primary orogenic provenances) and next appropriately recombined to describedetrital suites recorded by arc-related, foreland, foredeep, and remnant-ocean basin fills (composite orogenic provenances). Locally significant to locally dominant provenancesmay characterize first- to second-order-scale samples, whereas detritus of mixed provenance is invariably recorded at third-order scale (e.g., big rivers; Ingersoll 1990).


Figure 3. Detrital modes of modern sands from the Apenninic thin-skinned orogen. Shown are river and beachsamples derived from one single geological domain or subdomain (Garzanti et al. 2002a; first-order sampling scaleof Ingersoll 1990). Also shown are fields for the 10 largest rivers on each side of the orogen (E. Garzanti, unpublisheddata; second-order sampling scale of Ingersoll 1990). All of these carry recycled detritus of clastic wedge provenance,associated with lithic sedimentary detritus of undissected continental block provenance (Adriatic rivers draining thepro side of the orogen) or with feldspatholithic to arkosic detritus of magmatic arc provenance (Thyrrhenian riversdraining the retro side of the orogen). Detritus from remnants of the Alpine subduction complex and axial belt arelocally dominant (Northern Apennines and Calabria, respectively). Petrographic parameters, symbols, provenancefields, and confidence regions about the mean as in figure 2. Scale m; all photos with crossed polars.bar p 250

tic suites deposited in comparable geodynamic set-tings or for inferring compositional modificationscaused by intense weathering in hot humid cli-mates or diagenesis.

High-resolution bulk petrography data were col-lected by the Gazzi-Dickinson method (Ingersoll etal. 1984; Garzanti and Vezzoli 2003) on river andbeach sands derived from single homogeneous oro-genic domains (first- to second-order samplingscales of Ingersoll 1990). Further quantitative in-formation on heavy-mineral assemblages is pro-vided in Garzanti and Ando (2007b).

Magmatic Arc Provenances. Volcanic detritusfrom basaltic, andesitic, and rhyodacitic lavas andignimbrites representing the arc cover consists ofvolcanic lithic grains, plagioclase, and pyroxenes(fig. 3G, Undissected Magmatic Arc Provenance;Marsaglia and Ingersoll 1992). Plutonic detritusfrom diorite-granodiorite batholiths representingthe plutonic roots of the arc massif chiefly includesquartz, plagioclase, K-feldspar, and mainly blue-green hornblende (figs. 3F, 4D, Dissected MagmaticArc Provenance). The ideal compositional trend re-corded by terrigenous sequences accumulated inforearc and other arc-related basins during unroof-ing of the arc massif is, therefore, characterized bythe progressive increase of quartz, K-feldspar, andblue-green hornblende at the expense of volcaniclithic grains and pyroxenes (Dickinson 1985; Gar-zanti and Ando 2007b).

Ophiolite Provenance. Tectonically accreted orobducted sections of oceanic lithosphere, which es-caped subduction and orogenic metamorphism,shed mafic to ultramafic detritus with peculiar pet-rographic and mineralogical features (Nichols et al.1991). Distinct signatures characterize sedimentssupplied during unroofing of progressively deeperstratigraphic levels of the multilayered oceaniclithosphere. Pillow lavas and sheeted dikes of theupper crust shed lathwork volcanic to altered dia-base lithic grains and clinopyroxene or low-grademinerals grown during oceanic metamorphism(e.g., actinolite and epidote). Orthopyroxene-phyric

boninite grains may be common in detritus fromsuprasubduction-zone ophiolites (fig. 5D; Bloomeret al. 1995; Garzanti et al. 2000). Plutonic rocks ofthe lower crust supply cumulate, gabbro, and pla-giogranite rock fragments, calcic plagioclase, diop-sidic clinopyroxene, and green/brown hornblende;hypersthene grains may reflect the hydrous, arc-related character of late-stage magmatic sourcerocks. Serpentinized mantle harzburgites supplylizardite-serpentinite grains with pseudomorphiccellular texture, enstatitic orthopyroxene, olivine,and rare chrome spinel (figs. 3E, 5E; Garzanti et al.2002b).

Axial Belt Provenance. Detritus supplied by theaxial pile of neometamorphic nappes representingthe central backbone of collision orogens is influ-enced by several factors, including metamorphicgrade of source rocks and relative abundance of con-tinental versus oceanic protoliths. Metasedimen-tary cover nappes shed lithic to quartzolithic de-tritus, including metapelite, metapsammite, andmetacarbonate grains of various ranks (figs. 3A, 4B,5A); only amphibolite-facies metasediments supplyabundant heavy minerals (e.g., almandine garnet,staurolite, kyanite, sillimanite, and diopsidic cli-nopyroxene). Continental basement nappes shedhornblende-rich quartzofeldspathic detritus (figs.3B, 4A). Largely retrogressed blueschist to eclogite-facies metaophiolites supply albite, metabasite,and foliated antigorite-serpentinite (serpentineschist) grains (fig. 4C), along with abundant heavyminerals (e.g., epidote, zoisite, clinozoisite, actin-olitic to barroisitic amphiboles, glaucophane, om-phacitic clinopyroxene, and lawsonite).

Increasing metamorphic grade and/or deeper tec-tonostratigraphic level of source rocks may be re-flected by (a) increasing rank of metamorphic rockfragments (as indicated by progressive developmentof schistosity and growth of micas and other indexminerals; metamorphic index of Garzanti and Vez-zoli 2003), (b) increasing feldspars, (c) increasingheavy-mineral concentration (heavy-mineral con-centration index of Garzanti and Ando 2007a), (d)


Figure 4. Detrital modes of modern sands derived from the Alpine collision orogen. Shown are river samples derivedfrom one single geological domain or subdomain (Garzanti et al. 2004b, 2006). Major rivers (Rhone, Rhein, Inn,Salzach, Mur, Drau, Adige, and Po) carry mixed detritus mostly supplied by axial belt neometamorphic rocks and bycontinental block paleometamorphic and sedimentary rocks in various proportions. Petrographic parameters, symbols,provenance fields, and confidence regions about the mean as in figure 2. Profile modified after Polino et al. (2002).Scale m; all photos with crossed polars.bar p 250

increasing hornblende, changing progressively incolor from blue/green to green/brown (hornblendecolor index of Garzanti et al. 2004b), and (e) suc-cessive appearance of chloritoid, staurolite, kya-nite, fibrolitic, and prismatic sillimanite (metased-imentary minerals index of Garzanti and Ando2007b).

Continental Block Provenance. Allochthonousplatform to pelagic strata, representing the tecton-ically dismembered remnants of sedimentary suc-cessions originally deposited on a continental pa-leomargin, supply diverse sedimentary to low-rankmetasedimentary grains (e.g., limestone, dolostone,chert, shale, slate, and metacarbonate; figs. 3H, 4F;fig. 5B, 5C, Undissected Continental Block Prov-enance), locally associated with quartz, feldspars,or volcanic/metavolcanic rock fragments recycledfrom interbedded siliciclastic or volcaniclasticunits. Tectonically imbricated basement units shedquartzolithic to quartzofeldspathic sands with mi-cas, garnet, staurolite, kyanite, sillimanite, horn-blende, epidote, or pyroxenes (fig. 4E, DissectedContinental Block Provenance).

Detritus supplied by paleomargin remnants in-corporated within thick-skinned thrust belts variesmarkedly during unroofing of deep-seated base-ment rocks. Heavy-mineral concentration progres-sively increases, and detrital modes ideally changefrom lithic sedimentaclastic or locally sedimenta-clastic-volcaniclastic (cover sequences) to quart-zolithic, quartzofeldspathic, and feldspathic meta-morphiclastic signatures (greenschist-facies togranulite-facies basement units; Garzanti et al.2006). Instead, thin-skinned orogens associatedwith westward subduction zones mostly incorpo-rate cover strata, which invariably supply lithicsedimentary detritus (with various amounts of re-cycled quartz).

Clastic Wedge Provenance. Sand recycled fromfluvial to turbiditic foreland-basin, foredeep, orremnant-ocean-basin clastic sequences tend to re-produce the composition of orogen-derived (andthus typically quartzolithic; Dickinson 1985) par-ent sandstones, generally with significant additionof labile mudrock grains (fig. 3C, 3D; fig. 4G; Cav-

azza et al. 1993; Fontana et al. 2003). Because un-stable minerals are extensively dissolved duringdiagenesis of parent sandstones, recycled heavy-mineral assemblages include only stable to ultra-stable species and have very low concentrations(Garzanti et al. 2002a; Garzanti and Ando 2007a).

Composite Orogenic Provenances

Complex orogenic belts may include all threekinds of provenance in subparallel linear beltswhich may jointly contribute mixed detritus tovaried successor basins. Arc-derived detritusmay also be incorporated into such mixed suites.(Dickinson and Suczek 1979, p. 2176)

In order to illustrate how primary provenancesmay combine in different scenarios of plate con-vergence and give rise to composite orogenic prov-enances, we follow an exemplary rather than ex-haustive approach. We describe the signatures ofsediments supplied by a large subduction complex(Indo-Burman Ranges and Andaman Islands), by athin-skinned orogen produced by westward sub-duction (Apennines), and by three types of thick-skinned composite orogens produced by eastwardsubduction of continental-beneath-oceanic (Omanobduction orogen), oceanic-beneath-continental(Andean cordillera), and continental-beneath-con-tinental lithosphere (Alpine and Himalayan col-lision orogens).

Detritus from the Indo-Burman-AndamanSubductionComplex. Subduction complexes large enough tobe exposed subaerially and become significantsources of terrigenous detritus are typically formedby tectonic accretion above trenches choked withthick sections of remnant-ocean turbidites (Inger-soll et al. 2003). This is the case of the outer ridgeextending from the Indo-Burman Ranges to off-shore Sumatra, which largely consists of accretedabyssal-plain sediments ultimately derived fromthe rising Himalayas and locally overthrust by vol-caniclastic and ophiolitic forearc sequences (Curray2005; Allen et al., forthcoming). Modern sands fromthe Indo-Burman Ranges and Andaman Islands con-

328 E . G A R Z A N T I E T A L .

sist of quartz and feldspars recycled from turbiditicsandstones, with various amounts of shale/slategrains shed from turbiditic mudrocks (clastic wedgeprovenance). Ultramafic and mafic detritus is sup-plied locally from accreted forearc ophiolites (ophi-olite provenance). Additional volcanic detritus andchert grains are recycled from arc-derived and deep-water sediments of the forearc basin (fig. 6).

Detritus from the Apenninic Thin-skinned Orogen.Modern sands from the Apennines are derived fromdiverse source rocks, including pelagic cherty lime-stones and carbonate platforms of the Apulian pa-leomargin (undissected continental block prove-nance), foredeep turbidites accreted along the proside of the orogen (clastic wedge provenance), andvolcanic or locally plutonic arc rocks exposed alongits retro side (magmatic arc provenance). Becausewestward Apenninic subduction started in the latePaleogene along the retro side of eastward Alpinesubduction (Doglioni et al. 1998), the compositeApenninic orogen is capped by the proto-Alpinesubduction complex, including remnant-ocean tur-bidites and accreted ophiolites, and incorporatesboudinaged greenschist-facies to amphibolite-facies remnants of the Alpine axial belt (fig. 3). Be-cause of modest tectonic uplift, erosional unroofingis limited and the spatial distribution of detritalsignatures is largely controlled by geological in-heritance (Garzanti et al. 2002a).

Detritus from the Oman Obduction Orogen. Mod-ern sands from the Oman obduction orogen aremainly derived from mafic and ultramafic rocksoccupying the highest structural position of the tec-tonic pile. Sedimentary rock fragments, subordi-nate quartz, and metamorphic detritus are suppliedby frontally accreted to deeply subducted conti-nental margin rocks exposed in tectonic windows(fig. 5; Garzanti et al. 2002b). Where and when ero-sion bites into deeper structural levels, detritus ofophiolite provenance is thus mixed with, and fi-nally ideally replaced by, detritus of continentalblock and axial belt provenances.

Detritus from the Andean Cordillera. Modernsands from the Andes show a marked asymmetry.Volcano-plutonic detritus is dominant along thepro side of the cordillera, whereas quartzolithic toquartzose detritus with abundant metamorphiclithic grains characterize its retro side (fig. 7; Potter1994). Sediments in the Peru-Chile Trench rangefrom undissected-arc provenance adjacent to areasof active volcanism, to transitional-arc and dis-sected-arc provenances where the batholithic rootsof the inactive arc massif have been uplifted andwidely exposed along the Cordillera Occidental(Yerino and Maynard 1984; Thornburg and Kulm

1987). Sediments shed by metamorphic to granitoidbasement rocks and Paleozoic to Mesozoic strataof the Cordillera Oriental and Subandean thrustbelt display continental block provenance andare invariably mixed with subordinate volcano-plutonic detritus from the arc massif (DeCelles andHertel 1989). Recycling of weathered orogenic de-tritus (clastic wedge provenance) leads to a markedincrease in quartz across subequatorial lowlands(Johnsson et al. 1988).

Detritus from the Alpine and Himalayan Collision Or-ogens. Modern sands from the Alps and the Him-alayas chiefly include quartz, feldspars, metamor-phic rock fragments, micas, and amphibole-garnet-epidote heavy-mineral assemblages, reflect-ing major supply from partially retrogressed high-pressure (e.g., Penninic Domain) to high-tempera-ture (e.g., Greater Himalaya) neometamorphicrocks of the high-topography and rapidly exhumedaxial belt. Similar signatures, however, may char-acterize first-cycle detritus from paleometa-morphic basements representing remnants of thecontinental margins caught in collision, as well aspolycyclic detritus recycled from orogen-derivedclastic wedges accreted along the mountain front(fig. 4; Garzanti et al. 2004b, 2006).

In foreland-basin to remnant-ocean-basin succes-sions, neometamorphic detritus of axial belt prov-enance can be differentiated from paleometa-morphic detritus of continental block provenanceonly by using appropriate detrital geochronologytechniques (Najman 2006). As a further complexity,allochthonous remnants of continental paleomar-gins not only are commonly overthrust by the axialbelt and confined to external parts of the orogen(Helvetic Domain, Lesser Himalaya) but may alsolie structurally above it, as a “tectonic lid” (Aus-troalpine Domain, Tethys Himalaya). Externalbelts may directly face foreland basins on both sidesof the orogen, where the axial belt is narrow andtopographically subdued (e.g., Maritime and East-ern Alps).

The composition of foreland-basin sedimentsmay change from volcaniclastic, ophioliticlastic, orsedimentaclastic/low-rank metasedimentaclasticin early syn-collisional stages, when detritus fromvolcanic arcs and subduction complexes may bedominant (“Taiwan stage”; Dorsey 1988; Garzantiet al. 1996; Najman and Garzanti 2000), to high-rank neometamorphiclastic at later collisionalstages, when the axial metamorphic core of the oro-gen starts to be rapidly exhumed (White et al. 2002).Focused erosion of rapidly uplifted gneiss domesmay then produce huge volumes of hornblende-richquartzofeldspathic detritus that typically exceed


Figure 5. Detrital modes of modern sands from peri-Arabian obduction orogens. Shown are river and beach samplesderived from one single geological domain or subdomain (Garzanti et al. 2000, 2002a). Crust-derived feldspatholithicdetritus to mantle-derived lithic detritus of ophiolite provenance is shed by the oceanic upper plate, whereas sedi-mentary to neometamorphic metasedimentary and metavolcanic detritus of continental block to axial belt prove-nances is shed by the continental lower plate. Petrographic parameters, symbols, and confidence regions about themean as in figure 2. Scale m; all photos with crossed polars.bar p 250

the storage capacity of associated foreland basinsand can even reach totally unrelated sedimentarybasins thousands of kilometers away (Ingersoll etal. 2003; Garzanti et al. 2004a, 2004b). Volcanic orophiolitic detritus becomes volumetrically insig-

nificant, but contributions from dissected-arc mas-sifs may remain locally prominent (Garzanti et al.2005). Because of the progressive lateral growth ofexternal belts, which shield the foreland basin fromaxial belt detritus, compositional trends may revert

330 E . G A R Z A N T I E T A L .

Figure 6. Detrital modes of modern sands from the Indo-Burman-Andaman subduction complex. Shown are singlesamples from major rivers and beaches (E. Garzanti, unpublished data). Recycled detritus of clastic wedge provenanceincludes various amounts of shale/slate lithic grains eroded from turbiditic mudrocks. Sands from the Indo-BurmanRanges include minor volcanic-arc detritus and, invariably, negligible chert ( ,Lv/QFL% p 3 � 2 Lch/QFL% p

). Ophiolitic detritus is shed locally from forearc slivers found at the top of the orogenic stack in the Andaman0.5 � 0.5Islands, where beach sands may locally contain chert (symbols have thick outline if andLch/QFL% p 5–10

and very thick outline for the only chert-rich sample where and ).Lch/L% p 10–15 Lch/QFL% p 38 Lch/L% p 68Petrographic parameters, symbols, provenance fields, and confidence regions about the mean as in figure 2.

in time to sedimentaclastic/low-rank metasedi-mentaclastic (White et al. 2002). Detrital signaturesof foreland-basin sediments are controlled by theentry points of high-rank neometamorphiclasticdetritus carried by major rivers draining the axialbelt and thus vary irregularly along strike and arestrongly dependent on drainage changes (Muttoniet al. 2003; Najman et al. 2003).

Conclusions

Provenance interpretations for sedimentary as-semblages can be addressed most effectively inthe context of global paleogeographic patternsinferred from paleotectonic reconstructions andcan be used to test such reconstructions. (Dick-inson 1988, p. 22)

Orogens formed at convergent plate margins, rep-resenting topographically elevated sources of de-tritus, are composite geological structures of greatcomplexity, including diverse rock units assem-bled in various ways by geodynamic processes.Orogenic sediments thus embrace a large range ofmixed signatures, including variable proportions

of both first-cycle and multiciycle detritus fromneometamorphic, paleometamorphic, sedimen-tary, and igneous rocks. Unraveling the prove-nance of orogen-derived terrigenous successions isconsequently an arduous task that requires a de-tailed but, at the same time, simple and flexiblereference model.

In order to establish a classification of orogenicbelts and orogenic sediment provenances, we reducethe numerous possible plate interactions observedalong subduction zones by selecting a limited num-ber of variables (westward vs. eastward subductionpolarity; oceanic vs. continental downgoing andoverriding plates). Eight possible scenarios of plateconvergence are thus recognized, each characterizedby the tectonic assembly of a distinct type of com-posite orogen (fig. 1). As a further and most impor-tant simplification, we represent the structure oforogenic belts as a hierarchy of genetically associ-ated rock complexes generated by tectonic and mag-matic processes along subduction zones. The diver-sity of composite orogens is thus seen as resultingfrom juxtaposition and superposition of a limitednumber of subparallel geological domains.

Five types of such elongated orogenic domains


Figure 7. Detrital modes of modern sands from the Andean cordillera. Feldspatholithic suites of magmatic arcprovenance characterize the west-facing pro side of the orogen (modes of deep-sea samples after Yerino and Maynard1984 and Thornburg and Kulm 1987). Instead, quartzolithic sands of continental block to clastic wedge provenance,showing increasing degree of chemical weathering toward lower equatorial latitudes (Johnsson et al. 1988), characterizethe orogen’s east-facing retro side (modes of river samples after DeCelles and Hertel 1989). Petrographic parameters,symbols, provenance fields, and confidence regions about the mean as in figure 2.

are identified as the primary building blocks ofcomposite orogenic prisms: (1) magmatic arcs (au-tochthonous or allochthonous arc crust), (2) ob-ducted or accreted ophiolites (allochthonous oce-anic lithosphere), (3) neometamorphic axial belts(subducted continental margin crust or adjacentoceanic lithosphere), (4) paleomargin remnants (al-lochthonous continental crust), and (5) orogenicclastic wedges (allochthonous foreland-basin, fore-deep, or remnant-ocean-basin fills).

Detritus produced by erosion of each of theseprimary orogenic domains is characterized by spe-cific detrital modes, heavy-mineral assemblages,and unroofing trends, which can be predicted andmodeled. The five primary orogenic domains thuscorrespond to five (four chiefly first-cycle and onepolycyclic) primary types of sediment prove-nances; this refines and expands on the classicmodel proposed by the Dickinson school in thelate 1970s (fig. 2). As a final step, the complexityof detrital signatures produced by each type ofcomposite orogen as a whole may be representedas resulting from a limited number of combina-tions of the five primary provenances in variousproportions (figs. 3–7).

This relatively simple scheme is flexible enoughto describe the complete range of mixed detritalsignatures observed along modern subductionzones from the Mediterranean Sea to the Indian andPacific oceans. It is thus proposed here as a con-ceptual tool to model the composition of sedimentsproduced by erosional denudation of compositeorogenic systems and to predict the evolution ofdetrital modes and heavy-mineral assemblages asrecorded in space and time by clastic successionsdeposited in arc-related, foreland, foredeep, andremnant-ocean sedimentary basins.

A C K N O W L E D G M E N T S

This work benefited through the years from nu-merous fruitful discussions with, and precious ad-vice from, W. Dickinson, R. Ingersoll, Y. Najman,A. Basu, S. Critelli, P. DeCelles, A. Di Giulio, D.Fontana, B. Lombardo, C. Malinverno, M. Mange,M. Minoli, R. Valloni, H. von Eynatten, G. Weltje,and G. Zuffa. Careful, stimulating reviews by R.Ingersoll and P. Cawood are gratefully ac-knowledged.

332 E . G A R Z A N T I E T A L .

R E F E R E N C E S C I T E D

Allen, R.; Najman, Y.; Carter, A.; Bandopadhyay, P. C.;Chapman, H.; Bickle, M.; Garzanti, E.; Vezzoli, G.;Ando, S.; and Foster, G. Forthcoming. New con-straints on the sedimentation and uplift history of theAndaman-Nicobar accretionary prism, South Anda-man Island. Geol. Soc. Lond. Spec. Publ.

Baldwin, S. L.; Monteleone, B. D.; Webb, L. E.; Fitzgerald,P. G.; Grove, M.; and Hill, E. J. 2004. Pliocene eclogiteexhumation at plate tectonic rates in eastern PapuaNew Guinea. Nature 431:263–267.

Bally, A. W.; Gordy, P. L.; and Stewart, G. A. 1966. Struc-ture, seismic data, and orogenic evolution of southernCanadian Rocky Mountains. Bull. Can. Petrol. Geol.14:337–381.

Beanland, S., and Haines, J. 1998. The kinematics of ac-tive deformation in the North Island, New Zealand,determined from geological strain rates. N. Z. J. Geol.Geophys. 41:311–323.

Beaumont, C.; Ellis, S.; Hamilton, J.; and Fullsack, P.1996a. Mechanical model for subduction-collisiontectonics of alpine-type compressional orogens. Ge-ology 24:675–678.

Beaumont, C.; Kamp, P. J. J.; Hamilton, J.; and Fullsack,P. 1996b. The continental collision zone, South Island,New Zealand: comparison of geodynamical modelsand observations. J. Geophys. Res. 101:3333–3359.

Bloomer, S. H.; Taylor, B.; MacLeod, C. J.; Stern, R. J.;Fryer, P.; Hawkins, J. W.; and Johnson, L. 1995. Earlyarc volcanism and the ophiolite problem: a perspectivefrom drilling in the western Pacific. In Taylor, B., andNatland, J. H., eds. Active margins and marginal ba-sins of the western Pacific. Washington, DC, AGUMonogr. Ser. 88, p. 1–30.

Bostrom, R. C. 1971. Westward displacement of the lith-osphere. Nature 234:536–538.

Burbank, D. W.; Beck, R. A.; and Mulder, T. 1996. TheHimalayan foreland basin. In Yin, A., and Harrison,T. M., eds. The tectonic evolution of Asia. New York,Cambridge University Press, p. 149–188.

Cavazza, W.; Zuffa, G. G.; Camporesi, C.; and Ferretti,C. 1993. Sedimentary recycling in a temperate climatedrainage basin (Senio River, north-central Italy): com-position of source rock, soil profiles, and fluvial de-posits. In Johnsson, M. J., and Basu, A., eds. Processescontrolling the composition of clastic sediments.Geol. Soc. Am. Spec. Pap. 284:247–261.

Cawood, P. A. 1991. Processes of ophiolite emplacementin Oman and Newfoundland. In Peters, T.; Nicolas,A.; and Coleman, R. G., eds. Ophiolite genesis andevolution of oceanic lithosphere. Dordrecht, Kluwer,p. 501–516.

Coleman, R. G. 1971. Plate tectonic emplacement of up-per mantle peridotites along continental edges. J. Geo-phys. Res. 76:1212–1222.

Crespi, M.; Cuffaro, M.; Doglioni, C.; Giannone, F.; and

Riguzzi, F. 2007. Space geodesy validation of the globallithospheric flow. Geophys. J. Int. 168:491–506

Cruciani, C.; Carminati, E.; and Doglioni, C. 2005. Slabdip vs. lithosphere age: no direct function. EarthPlanet. Sci. Lett. 238:298–310.

Curray, J. R. 2005. Tectonics and history of the AndamanSea region. J. Asian Earth Sci. 25:187–232.

DeCelles, P. G., and Hertel, F. 1989. Petrology of fluvialsands from the Amazonian foreland basin, Peru andBolivia. Geol. Soc. Am. Bull. 101:1552–1562.

Dickinson, W. R. 1978. Plate tectonic evolution of NorthPacific rim. J. Phys. Earth 26(suppl.):1–19.

———. 1985. Interpreting provenance relations from de-trital modes of sandstones. In Zuffa, G. G., ed. Prov-enance of arenites. NATO Advanced Studies InstituteSeries 148. Dordrecht, Reidel, p. 333–361.

———. 1988. Provenance and sediment dispersal in re-lation to paleotectonics and paleogeography of sedi-mentary basins. In Kleinspehn, K. L., and Paola, C.,eds. New perspectives in basin analysis. Berlin,Springer, p. 3–25.

———. 1995. Forearc basins. In Busby, C. J., and Ingersoll,R. V., eds. Tectonics of sedimentary basins. Cam-bridge, Blackwell, p. 221–261.

Dickinson, W. R., and Suczek, C. A. 1979. Plate tectonicsand sandstone composition. Am. Assoc. Petrol. Geol.Bull. 63:2164–2172.

Doglioni, C. 1994. Foredeep versus subduction zones. Ge-ology 22:271–274.

———. 1995. Geological remarks on the relationshipsbetween extension and convergent geodynamic set-tings. Tectonophysics 252:253–267.

Doglioni, C.; Carminati, E.; and Cuffaro, M. 2006. Simplekinematics of subduction zones. Int. Geol. Rev. 48:479–493.

Doglioni, C.; Harabaglia, P.; Merlini, S.; Mongelli, F.; Pec-cerillo, A.; and Piromallo, C. 1999. Orogens and slabsvs. their direction of subduction. Earth Sci. Rev. 45:167–208.

Doglioni, C.; Mongelli, F.; and Pialli, G. 1998. Boudinageof the Alpine belt in the Apenninic back arc. Soc.Geol. Ital. Mem. 52:457–468.

Dorsey, R. J. 1988. Provenance evolution and unroofinghistory of a modern arc-continent collision: evidencefrom petrography of Plio-Pleistocene sandstones, east-ern Taiwan. J. Sediment. Petrol. 58:208–218.

Fontana, D.; Parea, G. C.; Bertacchini, M.; and Bessi, P.2003. Sand production by chemical and mechanicalweathering of well lithified siliciclastic turbidites ofthe Northern Apennines (Italy). In Valloni, R., andBasu, A., eds. Quantitative provenance studies in Italy.Mem. Descr. Carta Geol. Ital. 61:51–60.

Frey, M., and Ferreiro Mahlmann, R. 1999. Alpine meta-morphism of the Central Alps. Schweiz. Mineral. Pe-trogr. Mitt. 79:135–154.

Gansser, A. 1980. The Peri-Indian suture zone. In Au-


boin, J.; Debelmas, J.; and Latreille, M., eds. Geologiedes chaines alpines issues de la Tethys. Mem.B.R.G.M. 115:140–148.

Garzanti, E., and Ando, S. 2007a. Heavy-mineral con-centration in modern sands: implications for prove-nance interpretation. In Mange, M., and Wright, D.eds. Heavy minerals in use. Developments in Sedi-mentology Series. Vol. 58. Amsterdam, Elsevier, p.567–598.

———. 2007b. Plate tectonics and heavy-mineral suitesof modern sands. In Mange, M., and Wright, D. eds.Heavy minerals in use. Developments in Sedimen-tology Series. Vol. 58. Amsterdam, Elsevier, p. 809–833.

Garzanti, E.; Ando, S.; and Scutella, M. 2000. Actualisticophiolite provenance: the Cyprus Case. J. Geol. 108:199–218.

Garzanti, E.; Ando, S.; and Vezzoli, G. 2006. The conti-nental crust as a source of sand (southern Alps cross-section, northern Italy). J. Geol. 114:533–554.

Garzanti, E.; Ando, S.; Vezzoli, G.; and Dell’Era, D. 2003.From rifted margins to foreland basins: investigatingprovenance and sediment dispersal across desert Ara-bia (Oman, UAE). J. Sediment. Res. 73:572–588.

Garzanti, E.; Canclini, S.; Moretti Foggia, F.; and Petrella,N. 2002a. Unraveling magmatic and orogenic prove-nances in modern sands: the back-arc side of the Apen-nine thrust-belt (Italy). J. Sediment. Res. 72:2–17.

Garzanti, E.; Critelli, S.; and Ingersoll, R.V. 1996. Paleo-geographic and paleotectonic evolution of the Hima-layan Range as reflected by detrital modes of Tertiarysandstones and modern sands (Indus transect, Indiaand Pakistan). Geol. Soc. Am. Bull. 108:631–642.

Garzanti, E., and Vezzoli, G. 2003. A classification ofmetamorphic grains in sands based on their compo-sition and grade. J. Sediment. Res. 73:830–837.

Garzanti, E.; Vezzoli, G.; and Ando, S. 2002b. Modernsand from obducted ophiolite belts (Oman, U.A.E.). J.Geol. 110:371–391.

Garzanti, E.; Vezzoli, G.; Ando, S.; France-Lanord, C.;Singh, S. K.; and Foster, G. 2004a. Sediment compo-sition and focused erosion in collision orogens: theBrahmaputra case. Earth Planet. Sci. Lett. 220:157–174.

Garzanti, E.; Vezzoli, G.; Ando, S.; Paparella, P.; and Clift,P. D. 2005. Petrology of Indus River sands: a key tointerpret erosion history of the Western HimalayanSyntaxis. Earth Planet. Sci. Lett. 229:287–302.

Garzanti, E.; Vezzoli, G.; Lombardo, B.; Ando, S.; Mauri,E.; Monguzzi, S.; and Russo, M. 2004b. Collision-oro-gen provenance (western and central Alps): detritalsignatures and unroofing trends. J. Geol. 112:145–164.

Henrys, S.; Reyners, M.; Pecher, I.; Bannister, S.; Nishi-mura, Y.; and Maslen, G. 2006. Kinking of the sub-ducting slab by escalator normal faulting beneath theNorth Island of New Zealand. Geology 34:777–780.

Hodges, K. V. 2000. Tectonics of the Himalaya and south-ern Tibet from two perspectives. Geol. Soc. Am. Bull.112:324–350.

Horvath, F. 1993. Towards a mechanical model for the

formation of the Pannonian basin. Tectonophysics226:333–357.

Ingersoll, R. V. 1990. Actualistic sandstone petrofacies:discriminating modern and ancient source rocks. Ge-ology 18:733–736.

Ingersoll, R. V.; Bullard, T. F.; Ford, R. L.; Grimm, J. P.;Pickle, J. D.; and Sares, S. W. 1984. The effect of grainsize on detrital modes: a test of the Gazzi-Dickinsonpoint-counting method. J. Sediment. Petrol. 54:103–116.

Ingersoll, R. V.; Dickinson, W. R.; and Graham, S. A.2003. Remnant-ocean submarine fans: largest sedi-mentary systems on Earth. In Chan, M. A., and Ar-cher, A. W., eds. Extreme depositional environments:mega end members in geologic time. Geol. Soc. Am.Spec. Pap. 370:191–208.

Jaillard, E.; Herail, G.; Monfret, T.; and Worner, G. 2002.Andean geodynamics: main issues and contributionsfrom the 4th ISAG, Gottingen. Tectonophysics 345:1–15.

Johnsson, M. J.; Stallard, R. F.; and Meade, R. H. 1988.First-cycle quartz arenites in the Orinoco River basin,Venezuela and Colombia. J. Geol. 96:263–277.

Karig, D. E. 1982. Initiation of subduction zones: impli-cations for arc evolution and ophiolite development.In Leggett, J. K., ed. Trench forerarc geology. Geol. Soc.Lond. Spec. Publ. 10:563–576.

Koons, P. O. 1990. Two-sided orogen: collision and ero-sion from the sandbox to the Southern Alps, New Zea-land. Geology 18:679–682.

Laubscher, H. P. 1988. The arcs of the Western Alps andthe Northern Apennines: an updated view. Tectono-physics 146:67–78.

Lombardo, B., and Rolfo, F. 2002. Two contrasting eclo-gite types in the Himalayas: implications for the Him-alayan orogeny. J. Geodyn. 30:37–60.

Marsaglia, K. M., and Ingersoll, R. V. 1992. Composi-tional trends in arc-related, deep-marine sand andsandstone: a reassessment of magmatic-arc prove-nance. Geol. Soc. Am. Bull. 104:1637–1649.

Massell, C.; Coffin, M. F.; Mann, P.; Mosher, S.; Frohlich,C.; Duncan, C. S.; Karner, G.; Ramsay, D.; and Lebrun,J.-F. 2000. Neotectonics of the Macquarie Ridge Com-plex, Australia-Pacific plate boundary. J. Geophys.Res. 105:13,457–13,480.

Meffre, S., and Crawford, A. J. 2001. Collision tectonicsin the New Hebrides (Vanuatu). Island Arc 10:33–50.

Moore, G. F.; Shipley, T. H.; Stoffa, P. L.; Karig, D. E.;Taira, A.; Kuramoto, S.; Tokuyama, H.; and Suyehiro,K. 1990. Structure of the Nankai Trough accretionaryzone from multichannel seismic-reflection data. J.Geophys. Res. 95:8753–8765.

Muller, S., and Panza, G. F. 1986. Evidence of a deep-reaching lithospheric root under the Alpine Arc. InWezel, F. C., ed. The origin of arcs. (Developments ingeotectonics, Vol. 21). Amsterdam, Elsevier, p. 93–113.

Muttoni, G.; Carcano, C.; Garzanti, E.; Ghielmi, M.; Pic-cin, A.; Pini, R.; Rogledi, S.; and Sciunnach, D. 2003.

334 E . G A R Z A N T I E T A L .

Onset of major Pleistocene glaciations in the Alps.Geology 31:989–992.

Najman, Y. 2006. The detrital record of orogenesis: a re-view of approaches and techniques used in the Him-alayan sedimentary basins. Earth Sci. Rev. 74:1–72.

Najman, Y., and Garzanti, E. 2000. Reconstructing earlyHimalayan tectonic evolution and paleogeographyfrom Tertiary foreland basin sedimentary rocks,northern India. Geol. Soc. Am. Bull. 112:435–449.

Najman, Y.; Garzanti, E.; Pringle, M.; Bickle, M.; Stix, J.;and Khan, I. 2003. Early-Middle Miocene paleodrain-age and tectonics in the Pakistan Himalaya. Geol. Soc.Am. Bull. 115:1265–1277.

Nelson, T. H., and Temple, P. G. 1972. Mainstream man-tle convection: a geologic analysis of plate motion.Am. Assoc. Petrol. Geol. Bull. 56:226–246.

Nichols, G.; Kusnama, S.; and Hall, R. 1991. Sandstonesof arc and ophiolite provenance in backarc basin, Hal-mahera, eastern Indonesia. In Morton, A. C.; Todd, S.P.; and Haughton, P. D. W., eds. Developments in sed-imentary provenance studies. Geol. Soc. Lond. Spec.Publ. 57:291–303.

Ori, G. G., and Friend, P. F., 1984. Sedimentary basinsformed and carried piggyback on active thrust sheets.Geology 12:475–478.

Parson, L. M., and Wright, I. C. 1996. The Lau-Havre-Taupo back-arc basin: a southward-propagating, mul-tistage evolution from rifting to spreading. Tectono-physics 263:1–22.

Polino, R.; Dela Pierre, F.; Borghi, A.; Carraro, F.; Fioraso,G.; and Giardino, M. 2002. Note illustrative dellacarta geologica d’Italia alla scala 1 : 50,000: foglio 132–152–153, “Bardonecchia.” Torino, Litografia Geda Ni-chelino, 128 p.

Potter, P. E. 1994. Modern sands of South America: com-position, provenance and global significance. Geol.Rundsch. 83:212–232.

Ranalli, G.; Pellegrini, R.; and D’Offizi, S. 2000. Timedependence of negative buoyancy and the subductionof continental lithosphere. J. Geodyn. 30:539–555.

Ranero, C., and von Huene, R. 2000. Subduction erosionalong the Middle America convergent margin. Nature404:748–752.

Rivizzigno, P. A., and Karson, J. A. 2004. Structural ex-pression of oblique seafloor spreading in the Mac-quarie Island ophiolites, Southern Ocean. Geology 32:125–128.

Royden, L. H. 1993. The tectonic expression slab pull atcontinental convergent boundaries. Tectonics 12:303–325.

Rubatto, D., and Hermann, J. 2001. Exhumation as fastas subduction? Geology 29:3–6.

Scoppola, B.; Boccaletti, D.; Bevis, M.; Carminati, E.; andDoglioni, C. 2006. The westward drift of the litho-sphere: a rotational drag? Geol. Soc. Am. Bull. 118:199–209.

Searle, M. P., and Cox, J. 1999. Tectonic setting, origin,and obduction of the Oman ophiolite. Geol. Soc. Am.Bull. 111:104–122.

Searle, M. P., and Stevens, R. K. 1984. Obduction pro-cesses in ancient, modern and future ophiolites. InGass, I. G.; Lippard, S. J.; and Shelton, A. W., eds.Ophiolites and oceanic lithosphere. Geol. Soc. Lond.Spec. Publ. 13:303–319.

Spray, J. G. 1984. Possible causes and consequences ofupper mantle decoupling and ophiolite displacement.In Gass, I. G.; Lippard, S. J.; and Shelton, A. W., eds.Ophiolites and oceanic lithosphere. Geol. Soc. Lond.Spec. Publ. 13:255–268.

Tatsumi, Y., and Eggins, S. 1995. Subduction zone mag-matism. Cambridge, Blackwell, 211 p.

Taylor, F. W.; Bevis, M. G.; Schutz, B. E.; Kuang, D.; Recy,J.; Calmant, S.; Charley, D.; et al. 1995. Geodetic mea-surements of convergence at the New Hebrides islandarc indicate arc fragmentation caused by an impingingaseismic ridge. Geology 23:1011–1014.

Thornburg, T. M., and Kulm, L. D. 1987. Sedimentationin the Chile Trench: petrofacies and provenance. J.Sediment. Petrol. 57:55–74.

Uyeda, S., and Kanamori, H. 1979. Back-arc opening andthe mode of subduction. J. Geophys. Res. 84:1049–1061.

von Huene, R. 1986. Seismic images of modern conver-gent margin tectonic structure. Am. Assoc. Petrol.Geol. Stud. 26:1–60.

von Huene, R., and Lallemand, S. 1990. Tectonic erosionalong the Japan and Peru convergent margins. Geol.Soc. Am. Bull. 102:704–720.

Weltje, G. J. 2002. Quantitative analysis of detritalmodes: statistically rigorous confidence regions in ter-nary diagrams and their use in sedimentary petrology.Earth Sci. Rev. 57:211–253.

White, N.; Pringle, M.; Garzanti, E.; Bickle, M.; Najman,Y.; Chapman, H.; and Friend, P. 2002. Constraints onthe exhumation and erosion of the High HimalayanSlab, NW India, from foreland basin deposits. EarthPlanet. Sci. Lett. 195:29–44.

Willett, S.; Beaumont, C.; and Fullsack, P. 1993. Me-chanical model for the tectonics of doubly vergentcompressional orogens. Geology 21:371–374.

Yerino, L. N., and Maynard, J. B. 1984. Petrography ofmodern marine sands from the Peru-Chile Trench andadjacent areas. Sedimentology 31:83–89.

Documents

Orogenic Belts and Orogenic Sediment Provenance · rock complexes. Five types of such elongated orogenic domains are identiﬁed as the primary building blocks of composite orogens: