Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
I SUMMARY
Plate Tectonics ant] the Continents A Review 1 INTRO D UCTION B. CLARK BURCHFIEL Massachusetts Institute of Technology The plate-tectonics model has had marked success in ex- plaining the first-order features of the earth; it has been particularly successful in explaining the nature and age of the crust of the ocean basins. The model accounts for the distribution of earthquakes and volcanoes along plate margins and the nature of deformations occurring along the margins of the continents. The model, however, has had less success in accounting for geological activity within plates. especially the continental portions, and our confidence in the usefulness of the model diminishes as the age of continental crust increases. The model has been applied in an attempt to explain the tectonic features of the continents and to test the limi- tations ofthe model in time and space. The accompanying papers in this volume illustrate this application and the limitations and methods used to study the continental crust. The characteristics of plate boundaries, along which the major deformation seems to occur, and the na- ture of tectonic events that are not associated with plate boundaries are summarized in this paper. Chapter 2 treats the problems of understanding the pre-Mesozoic folder 15 r than ~200 million years (m.y.)] evolution of the conti- nents, which represents the majority of continental crust. PLATE BOUNDARIES The study of modern plate boundaries allows us the op- portunity to examine the geological and geophysical ac- tivity at the boundary and contemporaneous action ity adja- cent to the boundary. Studies should be directed toward establishing relationships between plate-bounda~y activ- ity and the wide range of geological and geophysical events, so that plate boundaries can be treated as dynamic systems. We know from modern and recent plate-bounda~y activ- ity that conditions at the boundary may change rapidly. on the order of a few million years, and in order to relate geological events to plate-boundary activity contempo- raneity of events must be established. Changes in plate- boundary activity may lead to partial destruction, defor- mation, or overprinting of the geological features of an older system by a younger system. Interrelations between the various parts of each system become lost or blurred.
16 Such changes are the reason why the details of an ancient plate-boundary system become hard to decipher. It is only through the study of modern plate-boundary systems that we can begin to comprehend fully the geological, geo- physical, and geochemical processes and their interrela- tions within these dynamic systems. The end product of such studies would be to understand the processes and geo- logical results of these systems from the earth's surface into the underlying mantle. We are a long way from this level of understanding, but continued investigations into modem plate-boundary systems must continue as they fonn Me fundamental basis for plate-tectonic analyses. Even though there is great diversity in modern plate- boundary systems, they do not represent the complete range of potential plate-boundary systems. Therefore, geological investigation of ancient plate-boundary sys- tems may demand interpretations different from those of modern systems. Plate-bounda~y systems can be divided into divergent, transfonn, and convergent types. Relative motion across plate boundaries is oRen not orthogonal. particularly at convergent boundaries; thus more complex combinations of these types may also exist. DIVERGENT BOUNDARY SYSTEMS Divergent motion within continental lithosphere causes extension of the lithosphere, producing elongate fault- controlled depressions called riRs. If divergence contin- ues' oceanic crust may develop and the continental l~tho- sphere becomes part of two diverging plates. In many cases ricing ceases before oceanic crust is developed, and the nbc and its related features remain within continents. Where divergent motion continues and oceanic crust is fanned, the no strictures become buried and fonn the basement for passive continental margins. Burke (see chapter 4) has outlined many of the features associated with continental rifts and has emphasized that to evaluate these divergent systems properly we must establish con- temporaneity of events within the system, the evolution of riR systems, and the temporal stn~ctum1 relations of riRs to adjacent oceanic crust. A modern example of a rift system that has led to the development of continental separation, connation of oce- anic crust, and a divergent plate boundary is the East African Red Sea, Gulf of Aden rift system. Studies in this region have produced good, yet still incomplete, un- derstanding of the evolution of a divergent plate- boundary system. The East African rifts are in the riR stage of development, whereas the Gulf of Aden has en- tered the drifting stage. The Red Sea appears to be in a transitional stage. From studies ofthe East African~lfofAden area (see Chapter 4) ant! from geophysical studies of continental margins (see Chapter 5), an evolution of a divergent plate- boundary system can be sketched. During the early riding stage, the system consists of a broad linear belt of active normal faults that produce elevated and depressed parts B. CLARK BURCHFIEL ofthe crust end characteristic patterns of sedimentary rock types that commonly include great thickness of salt. This early stage is also accompanied by volcanism, both al- kaline and tholeiitic. and significant geothel~al activity. The lithosphere is thinned over a broad linear zone, and the asthenosphere rises beneath the riced zone. Dunng continued ricing, this system enters the driR stage, in which oceanic crust fonns between the separated conti- nental lithosphere. Volcanism, structures, and associated sedimentary basins may extend for hundreds of kilometers away from the site that eventually develops into a diver- gent plate boundary. Modification of continental litho- sphere must take place beneath the entire affected region. The divergent system thus affects a broad region of conti- nental lithosphere, whereas divergent activity is more restricted within oceanic lithosphere. The relationship between upper crustal brittle faulting and deeper litho- sphenc structural and chemical behavior is at present poorly known and is certainly an important area for fixture studies. As the noted continental margins drift away from the spreading ndge they become less active and finally form a passive continental margin, where the transition from continental to oceanic lithosphere takes place within a single plate. Early subsidence along the passive margin leads to the formation of continental shelf sedimentary rocks, which overlap early-stage nk deposits. Further subsidence can be explained by thermal decay away Tom a spreading edge, and the formation of a constructional passive margin consisting of shelf slope ant rise morphol- ogy underlain by characteristic sedimentary-rocic assem- blages. The continued subsidence of some continental margins indicates that they are not truly "passive," but important dynamic processes are still active at the interface between continental and oceanic lithosphere. Sediment loading has been proposed as a mechanism for continued subsi- dence, but it is insufficient to provide the proper magni- tude of subsidence. Bott (1971) has argued that continued subsidence along some passive margins may be the result of lateral flow in the lithosphere across the ocean- continent interface. Such ductile flow could produce its own characteristic structures and modification of a pas- sive margin that are superposed on those of the earlier rift stage. By the time a passive margin has completed its subsidence due to thermal decay, it is outside the influx ence of the divergent boundary system. Its further devel- opment is in an intraplate setting, and the processes that affect it properly should be ~liscussed under the topic of non-plate-boundary systems. But since passive-margin development is clearly part of the evolution of divergent systems it is considered here. Investigations of modem passive margins at all stages of development are still in their infancy, and a great deal is yet to be discovered about their dynamics. Ancient passive margins are well known from the late Precambrian to the Recent. The marginal pants of most deformed belts, now present as part of the internal struc-
Plate Tectonics and the Continents: .A Rechew ture of continents, contain rock assemblages and struc- tures that indicate that they developed as part of a divergent dynamic system. Similar passive-margin sedi- mentary sequences are known as eroded remnants from some middle Precambrian deformed belts, which sug- gests that deep erosion could remove all sedimentary evidence for former passive margins. Some structures, ig- neous activity. and lithospheric modification by a di~er- gent dynamic system could, however still be preserved. Some middle Precambrian deformed belts show no sedi- mentary record of a former passive margin, and the ques- tion of whether one was ever present may have to rely on studies of the geological and geophysical characteristics of the crust beneath passive-margin sedimentary se- quences. Archean passive margins are unknown, which, along with much evidence, has suggested to many scien- tists that during most of Archean time continental litho- sphere was thin and ductile (see Chapter 15~. If plate- tectonic processes were active in the Archean, they may have operated at rates and scales on lithosphere and mantle with characteristics unlike those of later Pre- cambrian and Phanerozoic time. Some ancient passive margins are eroded where deeper parts ofthe lithosphere are exposed. In the southern Alps. an entire section of the continental crust and several ki- lometers of its upper mantle are exposed. Geological evi- dence indicates that this area was the site of a divergent plate boundary that evolved into a passive margin during Mesozoic time. Study of ancient passive margins permits the direct examination of geological features produced at greater depths than those of modern passive margins. During the early riding stage, several rids may develop that do not progress to the fo~`ation of ocean crust. Dur- mg the driPring phase they become inactive but retain the geological features characteristic of rifting, such as exten- si~e nonnal faulting, igneous activity, coarse elastic and e`aporitic deposition, followed by local overlap by pas- si~e-margin sediments. Studies of these failed rifts are important because they yield evidence for the evolution of adjacent passive margins and oceanic terranes which for ancient e.Yamples may be completely destroyed. Failed rids, commonly referred to as aulacogens. are s<~metimes reaction ated by other dynamic systems fanning either local mountain belts. .~bsiding basins (see Chapter A. fir the locus `~f modern intraplate seismicitv and ~ult- ing (see Chapter `). \tanv Ailed rifts contain important gas and `,il reserves as well as being related locally to lither types ot fire deposits. T RA TV S FO R M B O UN D A R Y S Y S T E ~ S Transtorrn h`~ndaries, ~ here taco lithosphere plates slide past each <'ther horizontally, that affect c<,ntinental lithosphere often Grin wide and complex belts `,t de~r- mation. The San Andreas Fault ot Calitomia has linen regarded as marking the transiorrn plate b<~ndarv be- tu een the Pacitic and N<,rth .\merican plates. It is I,<~` e~ en <'my fine of n meres north est trending tumults 17 of similar right-slip displacement. Some of these faults have large displacements and may have served as the plate boundary prior to the development of the present San Andreas Fault. It is more likely that the relative mo- tion between the two plates is a much broader zone, stretching from the marine California borderlands inland to at least western Nevada, a zone perhaps 500 km wide. Thus, this transform boundary, lice most others within the continental crust. is a broad zone of plate interaction and may be called more properly a transform boundary system. Although tTansforrn boundary systems contain a wide variety of associated geological features, the San Andreas system can serve as an example. Associated with the numerous strike-slip faults are areas affected by exten- sion, such as pull-apart basins, or compression, such as the folds and thrusts within the Transverse Ranges (Crowell, 1974~. ~rge-scale right-lateral structural bending and faulting in southwestern Nevada and adjacent California are probably part of the San Andreas transform system (Figure 1.11. Blake et al. (1978) have reviewed the ;NTeo- gene history of coastal California and have shown that the development of the oil-rich Neogene basins is related to evolution of the San Andreas system. In addition. recon- structions of Mesozoic and early Cenozoic paleogeog- naphy are complicated by the scrambling of paleogeo- graphic belts and associated rotations of these temnes by strike-slip faulting of late Cenozoic age (see Chapter 3) related to this system. Tra~el-time data from teleseismic studies of the upper mantle in southern California by Hadley and Kanamori ( 1977) suggest that the San Andreas Fault does not offset ~ ridgelike body of high-velocity mantle extending from 40 t<, perhaps lOO km. The San Andreas Fault apparently is confined to the upper 40 km of the continental litho- sphere. The ridge of high-velocity upper mantle extends 150 km east of the San Anclreas before ending. Hadley and Kanamori suggest that the upper lithosphere beneath soothed California is decoupled along a surface of hori- zontal shear. perhaps marked by low velocity, and that the plate bo~ndarv at depth is further east than at the surface (Figure 1.~). These interpretations are significant because they raise the Cation of correlation between tipper lithosphere and deeper plate and mantle dynamics. The dec`~pling at fine or more layers in the lithosphere has been suggested for other areas and other types of plate- b`>~ndary systems (see Chapters 8 and 9; .\rmbruster et aI., l9 ~ 8). This problem of lithospheric decoupling and its etEects on correlating shallow and deep plate motions and dynamics is clearly an important question that demands farther attention. The San Andreas Fault and other faults of this so stem has e been the toc Ilk of studies concerned ~ ith earthquake prediction and associated geological hazards. A ~ ide Ronnie ot steadies hare been directed toward understand- ing the dynamics of strain buildup. Cult initiation. and movement. Beca~lse the San Andrea system has been active from ;Il)OIlt 30 mu ago to the Recent. the opportu-
18 B. CLARK BURCHFIEL it, - ~ ~ - ~^ . ~ . . . _ _; +; _~ ... . hi_ ~ ¢~__~7~~ ..~: `~ .~ ,.~ ~ ~ _ Hi_ _~ I_ __ FIGURE 1.1 Looking northwest across southem Nevada ant across the Las Vegas Valley shear zone. The shear zone is marked by right-lateral bending of all pre-\liocene rocks and marlcs the displacement near the eastern side of Me San Andreas transform system. The bending and faulting along the shear zone account for a minimum of 70-km differential motion of late Miocene Age. nity exists to understand the devils of this transfonn system through time and its imprint on continental geology. Many large faults of proven strike-slip displacement are known in the geological record, such as the Great Glen Fault in Scotland. Other large linear faults or zones of faults. such as the Kings Keweah Suture in California (Saleeby, 1977) and broad Proterozoic shear zones of the midcontinent region of the United States (Warner, 1978) have been interpreted as ancient transforms. The fact is that the importance of transfonn boundary systems in the evolution of continents is essentially unknown. Where ancient aseismic displacement may have occurred over a broad zone, it is possible that we might not recognize it. A deformed belt such as the Precambrian Limpopo Belt of southern Africa is an intracratonic erogenic belt resulting from horizontal shear between two undeformed crustal blocks. The deformation described by Coward et al. (1976) occurs across about 100 km of terrane and took place at amphibolite and locally granulite metamorphic grade. Structural evidence indicates that the deformation was mainly by horizontal crystal shear. Can this defonna- tional style be the deeper expression of transform motion? An answer cannot be given wit}, our present state of knowledge of transfonn systems within continental litho- sphere. CONVERGENT BOUNDARY SYSTEMS Convergent plate boundaries, where one lithosphenc plate passes beneath another, are Me sites of subduction or recycling back into the mantle of the lithosphere created at spreading ridges. The stable configuration at a convergent boundary is where oceanic lithosphere is sub- ducted because it is more dense than the underlying as- thenosphere. Where oceanic lithosphere is subducted beneath oceanic lithosphere a volcanic island arc chain is present. Island arcs or continents may Boron past of a sub- ducting plate and may eventually be drawn into a subduc- tion zone. Because they are formed from material less
Plate Tectonics and the C`'ntinent.s:.A Reeled ~ a_ SO rv AVOID~ \ \ ,~,~,,,,1J FIGURE 1.2 Interpretive diagram of the proposed divergence of the crust and mantle boundaries along the San Andreas Fault in southern California (from Hadley and Kanamori, 1977. re- printed Mom the Bulletin of the Geological Society of America, with pez.~'ission). dense than oceanic crust, they are more difficult to sub- duct. The result is arc-arc, arc~ontinent' or continent- continent collisions. Study of modern subduction boundaries demonstrates that geological, geophysical, and geochemical events and features may extend for several thousand kilometers Tom the boundary. From studies of events and features at mod- ern convergent boundary systems, it is clear that ancient analogs exist within continents, and it appears that most material that forms continental lithosphere has been gen- erated or reworked in convergent.boundary systems. It is within such boundary systems that rocks formed within an oceanic setting are added to continents. NONCOLLISIONAL CONVERGENT SYSTEMS In modern noncollisional convergent systems geological features related to the system occur mainly within the o~e~iding plate and only to a minor degree within the lower plate. In fact, those features in the lower plate that are part of the systemextension faults caused by flexing of the lithosphere seaward of the subduction zoneare mostly lost during subduction; this is not the case during collisional events where the subducted plate is not oce- anic lithosphere. It is the association of geological fea- tures of the overriding plate that Is commonly presented in the geological record and leaves a record of plate con- vergence. Hamilton (Chapter 3) has reviewed briefly some of the elements of convergent systems. The outermost elements of the system are the trench and outer nonvolcanic ridge fanned from deformed packages of sedimentary rocks containing occasional slivers of oceanic lithosphere, an outer or forearc basin of undeformed sedimentary rocks, and volcanic arc consisting of a characteristic sprite of vol- canic rocks. Behind the volcanic arc, three dynamic set- 19 tines .~re pos.sit~le: (1) ``here the <overriding plate is in extension, leading to the development of a marginal sea floored by young oceanic crust or more distributed exten- sion leading to e..ctended continental crust such as the Basin and Range province (~e Chapters ~ and 9); <_) where the overriding plate is in compression, leading to the development of a wide belt of compressiona1 detorma- tion behind the arc such as in the modem Andes (see Chapter 6~; and (3) where the overriding plate is neither in extension nor compression and little or no bacic-arc structural activity takes place. The dynamics of these modern systems are being pieced together from studies in all branches of earth science. Geophysical research (see Chapter 10) has begun to define the geometry, thickness, and behavior of the subducted slab; the state of the lithosphere and astheno- sphere above the subducted slab; the nature of the thermal regime both above and within the subducted slab; and the general stress field within the overriding plate. Geochemical research has focused on problems related to the origin of the magmas and what data the volcanic arc rocks can contribute to our understanding about the nature of the source material (see Chapter 13), what modifications take place during assent of the mag- mas. and their final emplacement (see Chapter 152~. Geol- ogists have begun to establish the paleogeographic en- vironments of the sedimentary and volcanic sequences, the defonnational and metamorphic history of the dif- ferent parts of the system, and the range in evolutionary patterns for convergent systems. This is by no means a complete list of important research conducted in modern convergent systems but only a sampling to exemplify the multidisciplinary approach that must be undertaken to understand the dynamics of such complex systems. Within continents, ancient noncollisional systems can be identified as far back as the early Proterozoic or the Archean, although the earliest examples are significantly different from modern systems, and some workers have suggested that they have formed from dynamic systems that should be considered unrelated to plate tectonics. Mesozoic examples for each type of system are common and well established, Paleozoic examples are also com- mon but many are less well established, and Proterozoic examples are present but with a few exceptions are poorly understood. The problem becomes even more acute in the Archean where some elements of plate-boundary systems can be recognized. lout the assemblage of geologi- c`~1 features present in modern systems is lacking. `\ hether these Precambrian belts were formed by plate-boundary systems as we know them today, by modifications of modern systems, or by systems unique to the Precambrian is still an important question (see Chapters 52 and 1~) and one that certainly needs considerable attention because milch of the continental lithosphere consists of Precam- bnan belts of activity. ~7oncollisional cons ergent systems can affect conti- nental growth and evolution sometimes for more than 1000 km from the plate boundary. The Mesozoic and
JO Cenozoic development of the southern part of the western United States can serve as an example. Subduction of oceanic lithosphere beneath the continental margin ofthe western United States began about 240 m.y. ago and has continued. at least locally, to the present. Dunug much of Mesozoic time it appears that the convergent boundary system generated compression within the overriding plate and produced structural and magmatic effects more than 1000 km eastward from the surface trace of the plate boundary (see Chapters 3 and 6~. The sequence of ter- ranes from west to east across the western United States consists of a western terrane composed of accreted oce- ~nic sedimentary and volcanic rocks, `~ central terr`~ne composed open Andean-type magmatic arc, an stem terrace composed of belts of folds and east-directed thrusts where older continental basement was reactivated, find locally ~ foreland belt of faulted Precambrian cry.stal- line rocks fonned by older plate-boundary systems. The relationship between all these terranes is poorly under- stood, but correlation with modern systems such as the Andes will allow ~ betterunderstanding ofthe dynamics of such systems. Study of the Cordil reran belt ~ ill permit us to study relations between these terranes at crusta1 1~` els not exposed in the modem Andes. These two types <,f studies need to proceed together During the later part of the early Cenozoic time, con- ~ergence continued and magmatic effects took place across the western United States, but only minor deforrna- tion occurred, with the exception of the Rio Grande Rift that began about 29 m.y. ago (see Chapter 14~. From about 20 m.y. ago to the present, extensional tectonics have dis- ~upted the continental crust, folding the Basin and Range province. Extension began in a back-arc sewing and evolved complexly so that today part of the extensional terrane is being overprinted by the San Andreas transfonn system (see Chapters 8 and 9~. This late Cenozoic development ofthe Basin and Range extension has been recognized by several generations of earth scientists, and many explanations for the extension involve dynamics that would more properly be discussed under the category of divergent or transform plate- boundary systems. Recent studies have indicated that the extension is an integral part of a convergent system. Re- cent studies in the Basin ant! Range province exemplify how new and exciting discoveries await us in unexplored territory and how the melding of data from all earth- science disciplines can yield completely new insights into the development and evolution of the area. Field studies about 10 years ago led to the discovery that many subhorizontal faults, long considered Mesozoic thrust faults, were Cenozoic in age. A few years later it was determined that some areas affected by high-grade meta- morphism and deformation, again considered Mesozoic in age, were Cenozoic in age and locally could be related to the subhonzontal low-angle faults. Tying together ail the available geophysical. geological. and geochemical data known for the Basin and Range province, Davis (see Chapter 8) and Eaton (see Chapter 9) have developed B. CLARK BURCHFIEL models suggesting that the cr tstal rocks above to km e.,c- tend by Iystic normal faulting and are detached along a variably thick transition zone from the underlying crust that is extended by ductile flow and dike intrusion Ero- sional levels within the Basin and Range province have exposed perhaps all three crystal levels. The ideas ex- pressed in these models are new and unproved but demonstrate that the potential still exists for major new discoveries in regions that have been studied for several generations. Pre-.Mesozoic examples of noncollisional plate-bound- ary systems are poorly known largely because the ;ll- timate end of continental-margin evolution is collision by either island arcs or other continental fragments. These events commonly overprint the terranes generated prior to collision. Because a Paleozoic ocean basin was present in the site of the modern Pacific, evidence for Paleozoic noncollisional convergent systems is present around the Pacific margin, such as in Australia, South China (see Chapter 16), Canada, and parts of South America and Ant- arctica. Noncollisional systems are probably present within some Precambrian deformed belts, but because coIlisional events have overprinted these systems, they have become difficult to recognize. Only through care fill unraveling of the overpnuting events and recognition of the characteristic features of the noncollisional systems can these earlier systems be reconstructed. COLLISIONAL CONVERGENT BOI;NDARY SYSTEMS Island-arc terranes are fanned by noncollisional conver- gent boundary systems, but, because oceanic cmst ult - mately is subducted. the end product of all island-err development is collisional with a continent. The dy- namics of island-arc systems can be studied Tom modern examples, but the features of these systems become dis- rupted during their inevitable collision. Dunng arc- continent collision, new crust is added to the continent, but in many cases collided arcs are reworked by later superposed plate-boundary systems before they become ~ part of typical continental lithosphere. Continent- continent collisions simply rearrange continental ma- terial, but the processes leading to collision may add some new continental material. Several arc - Continent collision systems are at present active and in different stages of evolution. In most arc~ontinent collisions, the features of noncollision systems are present in the island arc, but where conti- nental lithosphere is present in the subducting plate and enters the subduction zone, considerable defonnation may take place within the subducting plate, extending the effects ofthe convergent boundary system. The following modem convergent systems show an increase in the extent of disruption of the subducted! continental litho- sphere: (1) northwest Australia, (2) Taiwan, and (3) north- ern New Guinea. In each case, the continental passi~e- margin sedimentary rocks, and in the latter two cases en en
Plate Tectonics and the Continents: A Re';zeu~ the underlying continental crust, have been cut by thrust faults locally as much as 20~ km from the site of the collision boundary. Because island arcs are backed up by oceanic lithosphere, following collision with a continent, the new ocean~ontinent interface is potentially subject to the development of a new plate-boundary system. If the geometry of the new convergent system subducts oceanic lithosphere beneath the collided arc, rocks of the collided system are overprinted by a noncollisional continental- margin arc system and become deformed and metamor- phosed to form the polyphase terranes common within continental crust. The transitions from arc collisions are not easy to interpret, and some of the volcanic rock associations developed following collision in New Guinea are difficult to relate to plate-tectonic processes Johnson et al., 1978~. Ancient mountain belts abound with evidence of arc- continent collisions and overprinting by younger plate- boundary systems (see Chapters 3, 6, and 16~. Because of the common overprinting of collided arcs, the reconstruc- tion of the plate-bounda~y systems related to the arc, its collisional history, and overprinting episodes are complex indeed. Study of older and more deeply eroded collided arc tenanes gives us the opportunity to examine the de- tails of the structure, geochemistry, metamorphism, distri- bution of rock types, and evolution of arcs and collisional margins that lie at depth below modern systems where the details remain obscure. Studies of these arc terranes will clearly extend back into Archean rocks of continents where arc systems have been identified. There are, however, significant differences between modem colli- siona] systems and Archean ones (see Chapters 2 and 15~. These differences must be understood, because they relate to the evolution of the earth, its lithosphere, and its continental regions. Continent~ontinent collisions occur where oceanic FIGURE 1.3 Distribution of tectonic styles in Asia along the Indian-Asian collisional system (Mom Tapponnier and .Molaar, 1976). Dark area, region of thrusting and crustal thickening; domed area, regions of strike-slip fault- ing; lined area, region of nonnal faulting and crustal thinning. Cor- respond~ng stress states are also indi- cated. " ~1 lithosphere between continents is completely subducted in one or more subduction zones. Remnants of the oceanic lithosphere with or without associated island-arc terranes may be preserved within the collisional boundary. Active systems of continent~ontinent collision are present in the Alpine-Himalayan mountain belt. Studies of this modem collisional system have shown that once collision has taken place, but convergence between the to o plates continues. geological activity w ithin the system broadens throughout both plates and becomes very diffuse. In fact. the geological activity may become developed in so many local structures that it is difficult, if not impossible. to define a single plate boundary. Nowhere is this type of activity better exemplified than in Asia. Molnar and Tapponnier (1975) have traced the collisional history between India and Asia from collision about 50 m.y. ago to the present. Since collision, India and Asia have converged by more than 2000 km. This postcol- lisional convergence has caused deformation within the Asian plate for a distance of 3000 km from the collisional boundary (Figure 1.3~- In addition, the Indian plate was disrupted by south-directed thrust faults. Slolnar and Tapponnier have suggested that the faults within the fanner Asian plate form a pattern that is explainable by indentation mechanics (Figure 1.~. India is displacing large parts of Asia laterally as it continues northward. Many ofthe active faults follow structural lines fanned by older collisional events within Asia (see Chapter 16~. Thus, it appears that such features as the extensional opening of the Baikal Rift are part of a convergent system whose initial collisional boundary was 3000 km away. Within this plate-boundary system complex patterns of volcanism' crystal defor~nation, and lithospheric modifi- cation are taking place. Once such a system becomes inac- tive, it is obvious that detailed geological, geophysical, and geochemical studies will be necessary to establish .~9 ~ e.. my' I' . ~j -~/W 4~=, ~~' ~ ~ as. ~~- 1
22 .L ':~ /~ ~~ as / b A, ~~ in/ iN; FIGURE 1.4 Plane indentation of semi-infinite rigid-plastic media by different rigid dies (from Tapponnier and !vlolnar, 1976). Allows indicate sense of shear along slip lines. Principal stresses as, and a3 bisect small quadniaterals delineated by slip lines. The names of corresponding major tectonic features are indicated. (a) Flat rigid die; (b) rigid wedge, similar to the situa- tion that occurs at the Pamirs (western end of the Himalayas); (c) flat triangular indenter and hallowed~ut medium, similar to situation that arises at the Himalaya-Bunnan syntaxis (eastern end of the Himalayas). B. CLARK BURCHFIEL contemporaneous relations between widely diverse geo- logical features and events. Ancient examples of continent~ontinent collisions are common in the geological record. Huang (Chapter 16) has described the tectonic evolution of China in such a way that it is obvious that China has been built by a complex series of arc and continent collision events that began in the late Precambrian, continued through the Paleozoic and Mesozoic, to the most recent collision of India during the Cenozoic. The ancient sutures can be identified in many places, but the far-reaching effects of convergent systems can produce a complex paKem of extensional, transla- tional, and compressional features, which Is only now being recognized. Burke (Chapter 4) has suggested that the Keweenawan rift system, which fonns a major crustal structure of the central United States (see Chapter 11), may be related to a coIlisional system within the Gren- ville (lOOO-m.y.mld) deformed belt (see Chapter 15~. Compound convergent systems may contain charac- teristics of two types. Such plate boundanes are char- actenzed by relative motions that are oblique or change rapidly because of continual shifts of rotational poles. Along the western part of the Java subduction system, the relative motion is oblique to the boundary and under- thrusting takes place in the Bench, whereas strike-slip faulting talces place in the overdying plate and is super- posed on the volcanic arc terrane. Further north, along the same plate boundary, the motion becomes strike-slip, ant] extension occurs behind the boundary in the Andaman Sea. Similar types of compound boundaries can be rec- ognizec} in ancient systems within continents. Within the Cordilleran belt of Canada and Alaska, both collisional and noncollisional convergence during late Mesozoic time was associated with major strike-slip filults that were fanned within the continent several hundred kilometers from the plate boundary (Davis et al., 1978~. The evolu- tion of the Pyrenees has been interpreted as a combina- tion of transfo.~ and both extensional and convergent systems at different times (Mattauer and Seguret, 1971~. Complex combinations of boundary systems can also result Mom collisions that take place along initially i'Tegu- lar boundaries. Dewey and Burke (1974) hypothesized that continent~ontinent collisions along irregular bound- aries could result in fragmentation and complex relative motions of small fragments. McKenzie (1972), Mom stud- ies of active seismicity in the eastern Mediterranean re- gion, showed that dunng such fragmentation complex relative motions take place. It can be argued that the mod- em seismicity gives us only an instantaneous picture of the boundaries and relative motions of small fragments, and, given longer periods of geological time, the entire belt of defonnation could be regarded as a continuous or at best a semicontinuolls system of deformation. Analysis of the motions of small fragments suggests that their mo- tions may not be a reflection of motion at depth. Within the alpine belt, analysis of fragment motion suggests that the continental fragments are probably decoupled fat
Plate Tectonics and the Continents: A Reeve?* levels within the crust andlor within the lithosphere and their motion is not directly reflected by deep lithospheric or asthenospheric flow (Burchfiel, in press). Such studies as those of Hadley and Kanamon (1977) along the San Andreas Fault and Davis (Chapter 8) and Eaton (Chapter 9) for the Basin and Range province add evidence that such detachments are possible. Some geophysical evi- dence suggests that for these zones decoupling may be related to low-veloci~ crystal layers. A1I these data and interpretations suggest that lower lithosphere now found beneath such continental fragments is exotic and may have an evolution different from the crust that overlies it. These types of complexities are only now being rec- ognized, and the extent of such activity must be under- stood because it could modify greatly the models we de- velop for the evolution of continental crust. ON-PLATE-BOUNDARY SYSTEMS There are many structural and magmatic features within continents that have no apparent relation to plate- boundary activity. Some of the features regarded above as related to plate-boundary systems are so far removed from the boundary that they may be independent of such systems and should be included in the category of non- plate-boundary systems. Included in non-plate-boundary systems, but not discussed in this chapter, are features of the interiors of oceanic plates. INTRAPLATE MAGMATIC ACTIVITY Intracontinental magmatic activity that is not related to plate-boundary systems takes the forTn of small plutons, plugs, laccoliths, dikes, diatremes, kimberlites, and some volcanic rocks. They have a complete range of compos~- tions from ul~amafic to alkalic. Sources of material for these roclcs are highly varied. Some basaltic igneous rocks and radiogenic isotopic data suggest that they are devel- oped from a deep mantle source, possibly below the as- thenosphere, that was homogenized about 2.0 b.y. ago (see Chapter 13~. Other basaltic rocks such as those ofthe Absaroka volcanic rocks, Snake River plain, and at Yel- lowstone. are derived from subcontinental mantle that has been attached to the overlying continental crust since the last major thermal event to affect these rocks (see Chapter 14~. The data indicate the continent and upper mantle in this area have been together since 2.8 b.y. ago. AII these types of igneous activity require perturbations in the mantle to cause Ideal melting. \\ hat these perturba- tions may be is uncertain and clearly represents problems for considerable future study. Whatever the causes, these rocks record events that contribute to the evolution of already formed continental lithosphere. Many of these igneous rocks contain fragments, .xenol~ths (see Chapter 12), that give its a sampling of the rock types present ~ ithin the lithosphere and a~thenosphere. If their source can be properly placed, and their age determined, they 23 yield additional valuable evidence concerning the en ol~- tion of continental lithosphere. These xenoliths. however naust be interpreted cautiously as they come from per- turbed areas within the earth and may not be representa- tive of the mantle or lower crust (Irving, 19 ~ 6). INTRAPLATE VERTICAL MOVEMENTS lithe development of basins and arches within continental areas appears to be unrelated to plate-boundary activity. Basins contain sedimentary accumulations that record the history of subsidence. The arches, on the other hand, are uplifted areas, and much of the record has been removed by erosion. and the history of their movement is more difficult to decipher. Vertical movements of the crust are the dominant processes in the formation of basins and arches, but the nature and cause of these movements is virtually unknown. Intracratonic basins are superposed on already formed continental crust of varied age and structure. The Michigan Basin overlies an older late Pre- cambrian riR terrane, whereas the Illinois Basin overlies an older Proterozoic folded belt that was affected by later Precambrian rhyolitic magnnatism. Some basins are ac- companied by early igneous activity and nonnal faulting, whereas others are not. The variety of basin settings has made it difficult to arrive at an explanation for their forma- tion. A number of models have been proposed that rely on thermal, eustatic, tensional, and compressional mecha- nisms (see Chapter 7~. None of these, however. has of- fered a completely satisfactory explanation. Basins are an important feature of continental tectonics because many contain important accumulations of hydrocarbons. Basin and arch development within continents indi- cates late-stage modification of already formed conti- nental lithosphere. What types of modifications talce place at depth are unknown but must be understood before a complete evolution of the lithosphere can be deciphered. Recent studies of unconfonn~ties and the fluctuation of sea level within basins has suggested a similarity between basins within a single continent and on several continents (Sloes, 1972~. Studies on continental margins show similar histories, indicating that these features may be a world- wide phenomenon. If this proves to be the case, at least part of the sedimentological controls could be related to plate-tectonic activity. Broad epirogenic movements, as with their more re- stricted counterparts, basins and arches, may also lead to modification ofthe continental lithosphere. Broad vertical movements, such as the elevation of the Colorado Plateau and adjacent Rocky Mountain area in the Cenozoic. must involve changes within the mantle. What these changes are and how they affect the continental crust and litho- sphere are essentially unknown. INTRAPLATE DE FORMATION Even though continental lithosphere away from active plate-bounda~y systems is considered rigid. it is not en-
24 tirely inactive. Recent studies have shown diffuse seismic activity to be present throughout the eastern part of the United States (see Chapter 7~. Some of these earthquakes have been large, with magnitudes greater than 7. A1- though intraplate continental seismicity is generally at a very loci level, recent studies have shown that the seis- micity occurs in several linear belts. It can be demon- strated that some of these linear belts are controlled by older trends (see Chapter 11) related to Precambrian, Paleozoic, and Mesozoic structures. Thus. mapping of older structures, both at the surface and their continuation beneath younger sedimentary cover within the conti- nents, may be useful in understanding the trends of mod- em and potential seismicity. It is not entirely clear that intraplate continental seismic activity is unrelated to plate tectonics. Studies of earth- quake focal mechanism, in site stress measurements, and hydraulic fracturing studies hare produced data showing the present stress field within the eastern United States to have a general pattern with the maximum compressive stress trending northeast, but with local complexities. Sbar and Sykes (1973) suggested that the regional stress field is related to the motion of the North American plate and that the earthquake zones are controlled by the pres- ence of unhealed fault zones. If intraplate fault activity is related to plate motion and to the reactivation of older structures, perhaps some intraplate igneous activity is also related to these reactivated fault and structural lines of weakness in the crust. Lipman (Chapter 14) has sug- gested that the temporal progression of volcanism asso- ciated with the Yellowstone volcanic field as well as some older Mesozoic and Cenozoic volcanic trends follow Pre- cambrian structural trends. It is clear that the relations between plate tectonics and intraplate tectonic and ig- neous activity require continued research to determine how much intraplate activity is related to plate motions and how much is related to independent causes. CO?<CLIJSIONS The plate-tectonics model offers an explanation for the present heterogeneity of the external part of the earth It explains the youth and simplicity of oceanic lithosphere and offers the potential to explain the antiquity, complex- itv, and evolution of the continental lithosphere. Progress can clearly be made if we ( 1) develop a three-dimensional understanding of the kinematics. dynamics. and thermal structure of modern plate-boundary systems and at the same time recognize those geological and geophysical features that are unrelated to plate interactions; hi) use this understanding to reconstruct the extent and evolution <>t ancient systems that fonn the major elements of the continental angst; and (3) determine the dynamics and evolution of systems that hare no modem analogs. The continents display an intemal heterogeneity that is still poorly understood and that represents an obvious B. CLARK BURCHFIEL challenge to earth scientists. They are complex but not incomprehensible. It is clear from the discussions above that no single continental terrace contains complete ~nfor- mation on plate-boundary and non-plate-bounda~ sys- tems, both modern and ancient. Leo understand these systems en c! continental development requires not only multidisciplinary studies but international cooperation. We are on the threshold of significant advances in our understanding of continents, and our future advances may rival those of plate tectonics as we understand it today. BE FERENCE S Annbn~ster. J., L. Seeber, and K. H. Jacob ( 1978). The northwestern tennination of the Himalayan mountain front: active tectonics from microearthquakes, J. Ceophys. Res. 83, 26~282. Blake. M. C., Jr., R. H. Campbell, T. W. Dibblee, Jr., O. G. Howell, T. H. .Nonnark, E. A. Silver, and J. C. Vedder (1978). Neogene basin connation in relation to plate-tectonic evolution of San Andreas Fault system, Califomia, Am. Assoc. Petrol. Ceol. Bull. 62, 34~373. Bott, A. H. P. (1971). Evolution of young continental margins and the connation of shelf basins, Tectonophysics 11, 31~327. Burchfiel, B. C. (in press). Eastern European Alpine system and the Carpathian Orocline as an example of collision tectonics, Tectonophysics. Coward, of. P., P. R. James, and L. Wright (1976). .Northem margin of the Limpopo mobile belt, southern Africa" Ceol. Soc. Am. Bull. 87, 601~11. CroweIl, J. C. (1974). Ongin of late Cenozoic basins in southern California, Soc. Econ. Paleontol. Mineral. Spec. Publ. 72, 191~204. Davis, C. A., 1. W. H. Monger, and B. C. Burchfiel (1978). Mesozoic construction of the Cordilleran "Collage," central British Columbia to central California, in Pacific Coast Paleogeography, Symposium 2, Pacific Section, Society of Economic Paleontologists and Mineralogists, pp. 1~2. Dewey, I. F.. and K. Burke (1974). Hot spots and continental break-up: Implications for collisional orogeny, Geology 2, 57 60. Hadley, D., and H. Kanarnori (1977~. Seismic structure of the Transverse Ranges, Califomia. Ceol. Soc. Am. Bull. ask. 1469~1478. Ining, A. J. (1976). On the ·.alidity of paleogeotherms de- tennined from xenoliths suites in basalts and kimberlites,.4m. .Uineral. 61, 638 642. Johnson, R. W., D. E. Mackenzie, and I. E. .~. Smith (lo. Yolcanic roclc associations at convergent plate boundaries: re-appraisal of the concept using case histories from Papua. New Guinea, Ceol. Soc. Am. Bull. 89, 9~1~. \1aKauer, .~1.. and A. Seguret ( 1971). Les relations entire la chine des Pyrenees et le Golfe de Gascogne. in [fistoire Structurale do Colfe de Cascogne, Vol. 1. pp. I\' ~1-1\ 4-24. \!cKenzie, D. P. (1972). Active tectonics of the ~Iediterranean region, Ceu Whys. J. Roy. .~s train. Soc. 18 1_32. .`l`>ln;lr, P. and P. T.~pponnier ( 19~). Cenozoi<. te< tonics `'f.~i.~: Effects off.` c`>ntinental collision. Science 189. 419~96. S`lleeby l. B. ( 1g, · ~ Fracture zone tectonics continental Marvin fragmentation find empl;.~cement of the King- Oh
Plate Tectonias and the Cvntinents: .A Rev~e2c `'phic~lite T'~lt i`'uth`` estern Sierra Ne~ <~da. C.~lit`'mi.~. in .~'rth.Arneric<`n Opl~i`,l~tes R. C. Cr~leman `~nd \~. P. lruin eds.. Orexon Dep Ge<'l and .Mineral Industries Bul l. 9~. pp. 141-1~. S bar. !~. L.. .ind L Svkes (19~3). Contemporarv c`'mpres!siv e stress and seismicity in eastern .North America: an example of intraplate tectonics. Ceot. Soc. Am. Bull. R54, 1861-1882. 25 Sloss L. L. ( 19 ~ _). S`nchron~ of Phaner(>zolc sedimentarv- tectonic e`ents <~t the .~,rth .American craton and the Russian platform. Int. Ge`31. C()ngr.. 24th .se.ssi`,n. sect. 4. E'P -~32. Tapponnier P.. and P. Slolnar ~ 19 ~ 6). Sl m-line field theow and large-scale continental tectonics .\iat'`re 264. .319-~. Warner L. .~. ( 19 ~). The Colorado lineament: ~ middle Precambrian wrench fault system Geol. Soc. .Am. B``ll. s~g 161-1~1.
Problems of Pre-Mesozoic Continental Evolution LEON T. SlLvER California lustitute of Technology INTRODUCTION By virtue of its remarkable ability to explain currently observed dynamic processes of ocean basins and conti- nental margins, and to account for much of the geological record of the last 200 million years (m.y.), the plate- tectonics model Is a rational starting point in interpreting the earlier evolutionary history of the earth. The model's numerous predictions facilitate continued testing and ups grading of its own validity. Since early Mesozoic time, modem continental margins have developed with distinc- tive physical and chemical characteristics that can serve as bases for comparative studies of older margins, which have been incorporated within the continents in the past. The recent progress of geological studies in North Amer- ica ant! the diversity of its geological endowments make the United States a particularly appropriate region for many of these essential investigations. There is, however, some risk in directing these research programs solely to the paradigms of plate tectonics. Our This paper is Contribution Number 3271. Division of Geological and Planetary Sciences, Califomia Institute of Technology, Pasa- dena, Califomia 91125. 26 current knowledge of the earth's evolutionary history in- cltldes evidence for many unidirectional changes in earth properties and suggests the need for a cautious approach to the extrapolation of modem geodynamics to earlier geo- Iogical eras. Earth energies, forces, processes, material distribu- tions, temperatures, gradients, rates, and other factors all combine to constrain earth dynamics. and all have changed with time. If we are to understand the past fully. Other versions of the plate-tectonics model and other models also must be considered. It is vital not to be seduced into allowing modern plate tectonics to assume premature dominance of interpretations of earth history, and therefore the analysis of the early geological and geo- chemical evolution of the continents deserves a signifi- cant fraction of our continental research efforts. The immense bibliography required to document all the assertions, suggestions, questions, and other pro~oca- tions contained in this brief chapter would outstrip it in length I omit them here (with one deferential exception) but suggest that the reader will find most of them con- tained in the references in the other papers contained in this volume, especially the papers by Hanson (Chapter
Problems of Pre--Ues~zv~c Continental Evolution 13), .\f uehlberger (Chapter (Chapter 12~. TIME CALIBRATION OF THE TECTONIC RECORD 15), and Kay and Kay It is the essential dimension of time, recorded in the paleontological time scale. in the various radiometric systems, and in the compelling paleomagnetic record, that revealed for plate tectonics its true complex dynamic character. Extension and testing of plate-tectonic con- cepts to the earlier geological record will continue to require time calibration. For the earlier Phanerozoic, fos- sil zonation is still the most precise time-resolving tool available. Its intercalibration with radiometric techniques is required for establishment of dimensions of physical time and for extension to unfossiliferous segments of the geological column. Continuing efforts in this direction deserve high priority. The paleomagnetic pole-reversal sequences and the apparent pole-migration paths have not yet been integrated into a continuous record that can be used for precise age assignments, but significant prog- ress is being made in this direction. The establishment of a comprehensive magnetostratigraphic record in the sedi- mentary and igneous columns of the Phanerozoic de- serves high priority. For the first 4 billion years (b.y.) of earth history, various radiogenic isotope systems in rocks and minerals, increas- ingly supplemented by biological, chemical, and rema- nent magnetic information, will be the principal tools for both general time assignments and precise time resolu- tion. The characterization of long-term evolutionary addi- tions of radiogenic 87Sr, 206Pb, 207Pb, Mob, and '43Nd to the major crust and mantle reservoirs provides important ge- netic and temporal information on the nature and history of the source regions of continental rocks. Current research in the precise measurement of ism topic systems promises time resolution on the order of ~ 0.1 percent in the oldest crustal rocks, under favorable conditions and with suitable lithologies. It is essential, however, that integrated petrological and geochemical re- search into the methods for interpreting the geochronol- ogy of high-temperature, high-pressure metamorphisms be continued. Among the most challenging problems in tectonics are the time constraints for erogenic evolution at depth in the lower crest and upper mantle leading to granulite and eclogite facies metamorphism. CONTIN E NTAL TE CTON IC S FROM I H E PR E-ARCH EA ~ TH RO U G H TH E PALE OZO IC THE INTERPLANETARY CONNECTIO N The stunning revelations of 20 years of space exploration in our 4.~-b.y.-old solar system, capped by direct lunar 2` sampling and field studies and by sophisticated orbital probing of Nlars, Mercury. Venus, and ttie Jovian systems, have focused scientific attention on the first 5~00 m.y. Of history in the terrestrial planets. The phenomenology of initial planetary aggregation and differentiation has been developed with extraordinary results for the earth's moon. Intemal lunar differentiation must hare been com- pleted in the first 100 or 200 m.y. Catastrophic surficial modifications from external bombardments are inferred in the interval Mom the moon's formation to about 3.9 b.y. ago. Comparative studies of the surfaces of the other ter- restrial planets also reveal profouncl effects of early im- pact histories. On the earth, the oldest rocks, peculiarly enough, are identified at close to 3.8 b.y. old and represent early dif- ferentiated granitic crust. Is this timing simply fortuitous? Are there still older rocks to be found in the continents? Or is this a most significant time convergence implying a critical shared episode in the genesis of the earth-moon system? Is this a time of common cessation of plane- tesimal infalls? Planetary fission? Planetary near~olli- sion? Some other phenomenon? In these, the earth's oldest rocks, the record commonly appears to be a montage of multiple episodes of igneous and metamorphic activities. No direct evidence for impact features has been reported; but impact involvement can- not be precluded. Can various constructional elements of the montage be stepped away to unveil cataclysmic im- pact products? What is certain is that crustal dynamics was well established more than 3.5 b.y. ago but in fonns and with rates that are still obscure. Was this ngi&-plate tec- tonics as it is modeled today? THE ARCHEAN NEED NOT BE ARCANE The earliest geological era for which there exists a signifi- cant record is the Archean, extending from 3.8 to 2.5 b.y. ago. It is no longer quite so cryptic or mysterious as it was when first tentatively identified. In North America, Greenland, Eurasia, Africa, and Australia, the earliest continental records have been extended back to 3.3 3.8 b.y. Although the rocks approaching 3.8 b.y. of age invari- abIy are strongly metamorphosed and complex, they es- tablish a minimum age for planetary differentiation and for the production of continent-forming crustal materials. Such rocks are found in west Greenland and in ~linne- sota. In the Barberton Mountains of South Africa and in the Pilbara region of western Australia, there exist essentially unmetamorphosed sedimentary and volcanic strata that are clearly 3.2 to more than 3.4 b.y. old. In their existence is the earliest record ofthe initiation ofthe very-long-term crustal stability, which is the unique distinction of conti- nental structures. In the rocks of the Barberton ~loun- tains, there is evidence for primitive life forms. Clearly. these regions contain the best surviving records of the early surface environments of the earth. At these sites the
28 crust. the atmosphere, and the hydrosphere interacted to produce at their interfaces the premonitory biological, chemical, and physical environments in which biochem- ical assemblages have evolved into the life forms and the life environments that we know. Widely recognized features of Archean shields are the numerous great linear arrangements of mafic volcanic rocks called greenstone belts, commonly more than 1000 km in length and hundreds of kilometers in width. Be- cause immensely valuable massive sulfides and other types of ore deposits are associated with them. the details of their local structure and stratigraphy, petrology, and mineralogy hare received considerable attention. How- ever, their broad spatial distributions, their isolation in "seas" of ancient granitic gneisses, their spatial and tem- poral relations to granite crust development are still in- completely understood. Among a number of features that distinguish Archean greenstone belts are abundant.Ug- rich, Si-poor basaltic lavas, which are unlike most recent basalts. These ancient rocks, called komatiites, require much higher melting temperatures, probably reflecting the much steeper temperature gradients that prevailed in the crust and upper mantle of the Archean earth. The relationship of these rocks to ancient crust is not under- stood. Were they part of ancient global oceanic floors sub- sequently deformed into linearity? Are they part of ancient island arcs produced at convergent plate bound- aries? Do their geometry, age ranges, and chemistry com- pare directly with Phanerozaic systems? Although Archean continental crustal rocks are present in large areas of the northern United States, their com- plete distribution and range in ages are still not estate fished. Their lateral boundaries with younger rocks are invariably concealed or complicated by igneous intrusion or intense shearing. Their vertical distribution and litI2o- spheric roots (extension to or below the base of the crust) are unknown. Their relationship to ancient ocean basins is not clear. Were there once supercontinents, subse- ~luentlv tr`~gmented and disaggregated prior to the Pro- ter<'z<'ic? Whether rigid" c~us~1 plates, as we know ~em, existed in the Archean has not been established. This is a central question for understanding early continental tec- tonics. PROTEROZOIC PERIMETERS IN SPACE AND TIME It is in the Proterozoic era, e.~ctending from 2500 to about 600 m.y. ago, that the opportunities for comparison of pre-Phanerozoic tectonics to plate tectonics are most favorable. The boundary in time between the Archean and Proterozoic eras is commonly set arbitrarily at about 2~00 m.y. ago. The boundaries in space between Archean and Proterozoic rocks in North America are far more diverse and potentially offer critical into,~.ation on tec- tonic processes. The last major magmatic culmination in the Archean appears to be recorded on all continents in the interval oO0-96~0 m.y. ago. In the Allowing hall aeon, 900~2500 LEON A. SILVER m.y. ago, well-documented evidence for significant oro- genic and magmatic arc den elopment is surprisingly sparse. Cratonic sedimentary sequences w ith plateau ba- salts (diabases) and felsic volcanic rocks are the dominant litholog~es of this period and are well preserv ed on all the continents. There is, therefore, a significant possibility that at approximately 2500 m.y. ago a global tTansforma- tion in the earth's dynamic systems occurred and was followed by 500 m.y. of comparative continental stability . This possibility is still incompletely seen, much less documented. The search for a more complete erogenic record of this seemingly quiet period needs new mo- mentum on all continents. A distinctive aspect of the sedimentary record found in the cnatonic and circumcratonic basins <'f the early Pro- terozoic is the abundance and diversity of sedimentary ore deposits. The banded iron formations, which are the prin- cipal world source of iron ores, reached maximum devel- opment between 2500 and 18()0 m.y. ago. The great uranium-rich conglomerates of the Canadian Huronian formations and the South African gold-uraniu~r. glomerates (e.g.' Witwatersrand) were formed in ;- interval. In each case, the sources of the metals - localization mechanisms are in need of better a. In the United States, iron, uranium, and gold ores are major targets for exploration; research into the early Pro- terozoic history of the continent conceivably may pros ide a substantial contribution in these important areas of re- source development. Broadly viewed, widespread Proterozoic orogenies in North America appear only in two significant time in- tervals, 200~1600 m.y. before the present (B.P. ) and 1300 900 m.y. B.P. In Canada, these are recognized as the Hudsonian and Grenvillian progenies, respectively. Most of the United States and Mexico appears to be underlain by crustal materials that seem to have been initiated in these two great episodes (see Chapter 15 for geographic distributions). Although their total distribution is incom- pletely known, especially under the Phanerozoic cover of modern continental margins, it commonly has been in- Erred that the major Proterozoic erogenic belts are sutured against the Archean crustal remnants. Suggested examples are along the Grenville front In Ontario and Quebec, the Nash Fork-.\lullen Creek shear zone in Wyoming, and in portions of Minnesota and Michigan. The geological style of each ofthese "sutured" perimeters appears to be sufficiently distinctive. however, to make all such boundaries worthy of specific study. As the proposed loci of continent-continent collisions. or transforms. they constitute a major class of targets for the study of Precam- brian crusta1 dynamics. Are there unique elements of the Proterozoic tectonic record that differ from more ancient or more recent tec- tonic products? Perhaps! At feast two possibilities may be considered. The midcontinent gravity high ofthe United States (see Chapter 11) is an anomalous geophysical feature with no precise analog recognized anywhere else on earth. It is the consequence ot a late Proterozoic continental rifting
Problems of Pre-`Uesozo~c Continental Evolution event in which older Proterozoic and Archean granitic crust was invaded in an arc uate belt by great masses of basaltic magma over a distance of about 2000 Ian, ~ idths of 50 150 km, and probably to depths of 20 30 km (base of granitic Christ). The timing of this catastrophic noting event is remarkable in that most of it occurred about 1100 1120 m.y. ago, essentially synchronous with a major plutonic magmatic are culmination recorded throughout the not too distant Grenville province. If the Grenville province was developing at a convergent plate boundary. Or was introduced by continent~ontinent collision at this time, the Keweenawan basaltic rip has no clear equiva- lent in comparable plate-tectonic settings developed in the last 200 m.y. Does this reflect a fundamental change in crustal or lithospheric strength and rigidity from the Proterozoic to the Mesozoic? Extensive geochronological studies have suggested a second remarkable Proterozoic phenomenon in North Amenca. In the interval from 140~1500 m.y. ago, in newly developed mid-Proterozoic crust, a magmatic event of enormous magnitude and with distinctive lithol- ogies occurred without identifiable relation to synchro- nous orogeny or sedimentation. From Labrador to Cali- fomia (and beyond?) hundreds, perhaps thousands, of ano~osite-syenite and rapakivi granite plutons with sur- face diameters up to 100 km perforated the continental crust in a belt at least 6000 km long and 1000 km wide. Within that belt these plutons became the most important voIumetnc constituent in the upper crust. The volume of magma is staggering for an anorogenic setting; the ma- tenal and energy sources are a first~rder problem in crust-mantle evolution; the state of crystal stress that ac- companied the emplacement is still speculative. FIas this type and magnitude of event been recorded in the Pha- nerozoic or in the Archean? If so, the record has not yet been read. In summing up some of the problems of Proterozoic dynamics in such cursory fashion, it is perhaps best to return to a first-order question. Can the nearly 2 b.y. of Proterozoic time essentially be represented by two great erogenic intervals, each approximately 300 400 m.y. long and separated from each other and We preceding Archean and succeeding Phanerozoic erogenic culminations by a quiet interval of comparative durations. In the detailed record, we know of some regionally significant orogenic events in the so-callecI quiet intervals, but the promi- nence of these two great erogenic periods, corresponding to the Hudsonian and Grenvillian, is clearly seen not only in North Amenca but on several other continents. What does this indicate for ep~sodicity, even penodicity, in global dynamics during this time interval? Is there possi- bly an extension of periodicity as well as episodicity beyond the Proterozoic? Were these orogenies shared events on the margins of supercontinents that subse- quently were fragmented and redistributed? Such propo- sitions represent some of the outstanding questions for those who would extend post-Paleozoic plate tectonics into the Precarnbnan. 29 ARE PALEOZOIC PLATES \fORE PALATABLE THA^N' PRECA\1 BRIA`NT ? In a summary discourse of this nature. c.~.sual pr<~ocati<'n is easier than thoughtful analysis. Current views `'f plate tectonics may not be adequate to explain even the P.`leo- zoic orogenies. \Vith the extraordinary [~,dv `~f Pale`'z<'ic faunal data on provinciality. ecology, and climatology. with increasingly detailed stratis~raphic and pale`'ma~z- netic records with over 100 years <'f intense ~e<'l`~gic`~l study of Paleozoic erogenic provinces on troth sides <'f the Atlantic Ocean. plus an essentially straightforward rec<'n- struction of Pangaea at the end of the Paleozoic. ~ critical analysis of the effectiveness of the plate-tectonics mode] in the Paleozoic seems possible. Indeed. an impressive number of rational interpretations of venous early Paleo- zoic plate-tectonic features have been made. They range from old marginal geosynclines and island arcs (northeast North America Great Bntain) to ocean floor (Bay of Islands) to closing oceans (Iapetus) to continental colli- sions (Appalachian and Caledonide orogenies) to a co- herent supercontinent (Pangaea). Given the immense complexity yet coherency of most of We diverse data in- puts, it would be difficult not to accept what I. Tuzo ``il- son ( 1966) suggested were the consequences of the riding, spreading, and closing of an ancient ocean in the early to mid-Paleozoic (the Wilson cycle). The Devonian to Permian history of the same region, however, involved a widespread, perplexing erogenic episode (the Hercvnian), which is found superimposed on the earlier orogenies, without the obvious plate-tectonic features Mat characterized He earlier events. Intense fold- ing, high-grade metamorphism, and granitic mobilization of a different chemistry are some of the consequences of the Hercynian (Allegheny) erogenic interval. .\ large number of studies attempt to relate the Her- cynian episode to hypothetical subduction and conver- gence, but almost always without broad acceptance. Many thoughtful students of plate tectonics recognize in this case that there are additional dimensions to the phenom- ena of continental dynamics that have not yet been elabo- rated in the existing models. One of the most exciting rewards of the study of pre- Mesozoic continental evolution will be. obviously. the ex- tension ;md improvement occurrent tectonics models. It its those workers ``ho exercise admiration for. and careful restraint in. the use of these models from whom me can expect the next generation of state-of-~e-art tectonics con- cepts. The greatest rewards, perhaps, may be found in our increased ability to approach the continental dynamics record as the key to the global dynamics record and, in turn, utilize terrestrial tectonic evolution as ~ key to the comparative tectonic history of the inner planets of our solar system. RE FE RE NC E Wilson, 1 T. (1966). Did the Atlantic close and then reopen? .`Jature 211, 676.
