Continental Tectonics (1980)

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.

v CONTINENTAL EVOLUTION

Cenozoic Volcanism in the Western United States: Implications for Continental Tectonics INTROD UCTI ON PETER W. LIPMAN U.S. Geological Survey Volcanic rocks can provide insights into major features of continental tectonics, for which in some cases little other record may exist. Thus, study of Cenozoic volcanism in He western United States may reveal structural and com- positional constraints on the lithosphere of the American plate and on the geometry of past interactions with other plates. This chapter focuses on four features of conti- nental tectonics: (1) the composition and geometry of underlying and, as yet unexposed, cogenetic magma chambers and batholiths; (2) lateral compositional varia- tions in the lithosphere of the American plate; (3) struc- tural discontinuities in this lithosphere; and (4) the geometry of past interactions between the American plate and various Pacific oceanic plates. SUBVOLCANIC INTRUSIONS The study of Cenozoic volcanic activity in the western United States provides information on the distribution 161 and composition of young intrusions at shallow depth. Such intrusions constitute significant upper-crustal conti- nental-tectonic features of major potential economic sig- nificance for mineral and geothermal resources. The in- trusions are also products of evolving magmatic systems that may have caused long-lived perturbations of the lower crust and upper mantle during processes of partial melting, magma migration, and fractionation. The close relationship between continental volcanism and emplacement of shallow granitic intrusions was dis- cussed by Hamilton and Myers (1967~; they concluded that major silicic volcanism is typically the surface mani- festation of batholithic intrusive activity at high crustal levels and that many large batholithic complexes are shal- low, grossly lenticular bodies, roofed largely by cogenetic volcanic rocks erupted during immediately preceding stages in the rise and emplacement ofthe batholith. Publi- cations by these authors and others in the past 10 years have developed generally similar interpretations for volcano-plutonic associations in the western United States, especially for the Boulder batholith of Montana (Hamilton and Myers, 1974; Klepper et al., 1971) and

162 large parts of the Siesta Nevada batholith in California (Schweikert, 1976; Fiske et al., 1977~. By analogous rea- soning, the distribution and compositions of volcanic rocks, together with pertinent geophysical data, can be used to infer the geometry of subvolcanic intrusions or even magma chambers beneath young volcanic fields into which erosion has not exhumed widespread intrusive rocks. Oligocene volcanic activity in the San Juan field, south- westem Colorado, is interpreted as recording the rise, differentiation, and consolidation of a composite batho- li~, largely of intermediate composition and covering an area of roughly SO km by 100 km (Steven and Lipman, 1976; Lipman et al., 1978~. Early eruptions that fanned stratovolcanoes of intermediate composition in the San Juan field were accompanied by intrusion of small stocks into the cores of these volcanoes (Figure 14.1A). No evi- dence exists for any shallow magma bodies of batholithic dimensions at this stage, although some of the early volca- noes are clustered, apparently marking sites of concen- trated accumulations of magma. Within several of these PETER W. LIPMAN clusters, sufficient magma subsequently accumulated at shallow depth to permit formation of silicic differentiates, eruption of ash-flow tufts, and resulting caldera collapses (Figure 14.1B). At least 15 ash-flow calderas, mostly formed within a 2-million-year (m.y.) interval (29-27 m.y. ago), are clustered within terrane characterized by a large, stee~sided, flat-floored, Bouguer gravity low (Figure 14.2) that is interpreted as defining the final shape of the batholith, win the calderas marking isolated higher cu- polas. The dimensions of the Bouguer gravity anomaly in the San Juan Mountains (Figure 14.2) and, by inference, those of the concealed batholith are comparable with those of the Upper Cretaceous Boulder batholith of Montana. In Montana the gross stratigraphy of the Elkhom Mountains VoIcanics, erupted penecontemporaneously with em- placement of Me Boulder batholith, is also similar to that of volcanics in the San Juan field; a thick sequence of intermediate~omposition lavas and volcaniclastic rocks is capped by culminating more silicic ash-flow eruptions. Ash-flow-relate<1 caldera collapses must also have accom- v s_ ~ ~ ,~ _~y~~S s—~~ EX PI~ANA T~ ON ~~~ 1 a. 35-30 m.y. i,,,,,/, j~'~ v ~ -~ B. 30 - 26.5 m.y. 1 sit 1 . , ~ . note intermediate -composition rocks j~ vl, volconi class ic sedimentory rocks ~3 \2 sl, strotovoiconoes f.,2,ei`''.2 Lote intrusions (Numbered in order of emplacement) Silicic ash-flow sheets Eorly intrusions ~ ~ s 1 Eorly intermediote-compositian rocks v, volconiclastic sedimentary rocks s, strotovolcanoes FIGURE 14.1 Schematic model for evolution of Oligocene subvolcanic batholith in San Juan Mountains (Lipman et al., 1978, courtesy the Geological Society of America). A, Time of early intermediate volcanoes. Clusters of interrnediate~omposition stratovolcanoes are surrounded by interfingering aprons of volcaniclastic debns. Small intrusions fonn at shallow levels in volcanic pile, but it is uncertain whether a large high-level intrusive complex has developed by this time. B. Time of ash-flow eruptions and caldera Connation. Eruption of complex sequence of ash flows and associated caldera collapses is triggered by accumulation at shallow depth of batholithic-size magma bodies of intermediate to silicic composition. Many of these shallow accumulations are localized within clusters of earlier stratovolcanoes. Some calderas are composite, with younger activity nested within older collapse structures, and many caldera collapses are followed by resurgent doming, indicating renewed upwelling of silicic magma.

Cenozoic Volcanism in the Western United States ~- 3Y - 163 10 , ( in, fault—Bar and ball on downthrown side (3 Caldera __ ,` ,, Buried or inferred caidera ~ J (~ ~-;^ / _ _ I 0 10 20 30 40 50 K'L - ETRES 1 1 1 1 1 ~ —-300 - Elouguer gravity co ntours—Hachures point toward closed gravity lows. Contour interval 10 milligals. Gravity data from Plouff and Packer {1972, fig. 3) , . . DE~VE Ro ~ CoLORADO l it S.A Jeer ~ . _>'-~-'Ce^" "-'6 ~ LOC^710N Of 5~N ~~N VOLCANIC flELD CALDE RAS (in order of increasing age) Lake City Creede Coct~etopa Park San Luis Bachelor La Garita Mount Hope S. Iverton San Juan U ncompaNgre Lost Lake Ute Creek Summ.tv~lle Platoro Bonanza FIGURE 14.2 Calderas in the San Juan volcanic field in relation to Bouguer gravity field (Steven and Lipman, 1976). panted emplacement of the Boulder batholith, perhaps within the region where only intrusive rocks are currently exposed. At another major but much younger caldera complex, centered in Yellowstone National Park and related to large-scale ash-flow eruptions about 0.6 m.y. ago, exten- sive geological and geophysical evidence indicates the presence of a large shallow body of silicic magma (Eaton et al., 1975~. This body is thought to underlie an area more than 85 km long and 55 km wide, as indicated by surface geology and by gravity studies. Seismic-attenuation and P-wave-residual studies suggest that the top of the magma body is only a few kilometers below the surface and is still molten. The molten magma is interpreted to be underlain by an even larger volume of mechanically and thermally disturbed crustal and mantle rocks that contains pods of basaltic and silicic magma and extends at least 50 km into the mantle—probably through the lithosphere of the American plate (Eaton et al., 1975~. These features have been interpreted as the geophysical expression of a "grav- itational anchor" (Straw and Jackson, 1973), which re- sulted from sinking of dense residue from partial melting episodes during formation of the Yellowstone magmas (Christiansen and McKee, 1978~. Here, volcanism is a sur- face manifestation of geophysical discontinuities that ex- tend to great depth. Questions as yet unaddressed are whether such discontinuities are preserved under older volcanic fields and whether such discontinuities can af- fect later tectonic and igneous events. Putting together fragments of the magmatic record in this way, we can determine the main loci of relatively recent platonic activity, for which erosion has not yet provided direct exposure. Much more should be determinable about such economically important en-

164 vironments by combined geological, geochemical, and geophysical studies of transitional volcano~lutonic com- plexes, especially where interpretation of ~ird- dimensional relations can be augmented by drill-hole data. COMPOS ITIONAL VARIATIONS IN THE CONTINENTAL LITHOSPHERE Recent volcanic studies using a plate-tectonic framework have emphasized the importance of magma sources below the lithosphere in extensional environments, from ups welling asthenosphere along spreading oceanic ridges; in convergent environments, from descending slabs along subduction systems below magnetic arcs; and in intra- plate settings, Tom upwelling deep "mantle plumes" (Wilson, 1965; Morgan, 1972; Martin and Piwinskii, 1972~. Geochemical studies, especially Sr isotopic data, have tended to emphasize the importance of mantle sources for silicic as well as for magic volcanic rocks and to rule out major contributions from upper~rustal ma- tenals (Hurley et al., 1962; Peterman et al., 1970; Kistler and Peterman, 1973~. Such conclusions, initially drawn largely Tom studies of volcanism in ocean basins. have also been widely held to be generally applicable to continental igneous activity. However, long-recognized contrasts in the nature of volcanic activity between conti- nental and oceanic regions (Gilluly, 1955), as well as addi- tional arguments outlined below, indicate that conti- nental volcanism is strongly influenced by structural and compositional features of He immediately underlying lithosphere, even though the activity may ultimately have been initiated at greater depths. In particular, continental volcanism seems sensitive to the age and composition of the immediately underlying lithosphere, as well as to the geometry of major structural flaws. Implications of the compositions of continental volcanics for compositions of the lower crust and attached lithospheric upper mantle are discussed elsewhere (see Chapter 12~. Accordingly, aspects of compositional relations are outlined here only to the degree necessary to support other interpretations of relations between continental volcanism and tectonics. Recent Pb and Sr isotopic studies of continental vol- canic rocks in the western United States indicate that the isotopic composition of volcanic rocks reflects the age and composition ofthe underlying lithosphere, demonstrating major compositional control by the lower crust or the li~- osphenc upper mantle. For example, isotopic composi- dons of both Tertiary volcanic rocks of the Absaroka vol- canic field and Quatemary volcanics of the Yellowstone Plateau in the same region (Petennan et al., 1970) lie along a well-defined secondary isochron that defines an apparent age of the source region of 2.8 billion years (b.y. - he same as that of the underlying Precambrian basement (Figure 14.3~. In the San Juan Mountains and adjacent parts of the Rio Grande rifle in sounded Colorado and northem New Mexico, Pb isotopic data for rocks rang- PETER w. LIPMAN . 7 1 5 8 Q CL o Cal CL o 150 Cal 15.4 o ~_. ~~ / 46- 35 ,4.5 155 165 206pb/2o~pb 17 5 ,8 s FIGURE 14.3 Lead isotope relations for the Tertiary Absaroka volcanic field (solid circles) and Quatemary basalts and rhyolites (open circles) Mom Yellowstone National Park (Peterman et al., 1970). ing in age from Oligocene to Quaterna~y yield a secondary isochron that indicates an apparent source age of about 1.7-1.8 m.y.—again the dominant age of the underlying Precambrian basement (Lipman et al., 1978~. In these two regions the rocks analyzed are from both compressional and extensional tectonic environments and range from basalt to rhyolite. These relations strongly suggest that, whatever He ultimate origin of the volcanism and its thermal requirements, the compositions of the rocks are dominated, at least for Pb, by contributions from the lith- osphere of the American plate. In He case of the middle Tertiary rocks of intermediate composition in both the Absaroka and San Juan volcanic fields, for which complex subduction models have been proposed (Lipman et al., 1971; 1972), any initial compositional signature from melting of the subducted slab or the immediately overly- ing asthenospheric mantle has been masked by interac- tions of the rising magmas with the American plate. For the San Juan field, at least, detailed consideration of the isotopic data suggests a major contribution from lower- crustal sources; only minor interactions with upper- cn~stal rocks are compatible with the Sr isotopic data. Isotopic compositions of upper Tertiary and Quatemary basaltic rocks in the Yellowstone Plateau and Rio Grande no indicate dominant sources of Pb from mantle regions that have been a part of the American plate since forma- tion of the associated parts of the continental craton, re- spectively, about 2.8 b.y. and 1.7-1.8 b.y. ago. If the vol- canism of the Yellowstone Plateau~nake River Plain trend represents the trace of a deep mantle plume, as Morgan (1972) and Suppe et al. (1975) conjectured, then isotopic compositional identity of the deep source has been lost. Variations in Pb isotopic compositions of Cenozoic ig- neous rocks across the western United States delimit an abrupt discontinuity between a West Coast region and the eastern Cordillera; this discontinuity is interpreted as masking the limit of Precambrian crust (Doe, 1967; Zart- man, 1974~. Analogous Sr isotopic studies also show strong regional gradients and boundaries that have been interpreted as reflecting major structural boundaries in the underlying continental crust (Kistler and Peterrnan, 1973; Armstrong et al., 1977~. In some regions, however, interpretation of the isotopic data is age-dependent; in

Cenozoic Volcanism in the Western United States southern Arizona and New Mexico, mafic Oligocene vol- canic rocks are relatively radiogenic (87Sr/86Sr - 0.706- 0.708), whereas Miocene and Pliocene basaltic volcanics are significantly less so (87Sr/86Sr = 0.703~.7041. This compositional difference may reflect a change in the com- position of the source region, perhaps related to upwell- ing of deeper mantle material in late Tertiary time, during major lithosphere extension and development of Basin- Range structure (Lipman and Mehnert, 1975~. Interpretation of this sort eventually should increase our understanding of the complex structure and composi- tion of the continental crust and upper mantle of the west- em United States, where, during Cenozoic time, a plate- convergent regime was superceded by an extensional one as a result of changing plate boundaries (Atwater, 1970~. In this region the compositional and structural zones, with which rising and evolving Cenozoic magmas could have interacted, include (1) sialic upper crust that ranges re- gionally in age from Phanerozoic to as much as 3 b.y. old and that is relatively radiogenic both in Pb and Sr; (2) mor mafic lower crust, probably of less radiogenic iso- topic character; (3) lithospheric upper mantle ofthe Amer- ican plate, having a depth of about 50-100 km between the Moho and the top of the low-velocity zone, that may have been little modified chemically since formation of the overlying craton; (4) asthenospheric mantle, below about 100 km, of relatively worldwide compositional homogeneity; and (5) at various times and places (as dis- cussed later), a subclucted plate descending slowly to the east through the asthenosphere to a depth of several hundred kilometers. This inferred subducted plate con- sisted of oceanic lithosphere, generated at the East Pacific Rise, and was subjected to little-understood processes and interactions with asthenosphere beneath the Ameri- can plate for 10-20 m.y. before thennally equilibrating and losing its geophysical identity. STRUCTURAL DISCONTINUITIES IN THE CONTINENTAL LITHOSPHERE Structural discontinuities in the lithosphere also seem im- portant in controlling the distribution and type of Ceno- zoic volcanism in the western United States. Especially conspicuous as controls for Cenozoic volcanism seem to be several no~east-trending zones of different ages and origins: the Snake River-Yellowstone zone, the Springer- ville-Raton zone, and the Colorado mineral belt (Figure 14.41. Other possible crustal flaws, not discussed here, may also be important in interpreting aspects of conti- nentaI tectonics. Proposed interpretation of the Snake River Plain- Yellowstone zone as the trace of a mantle plume or melt- ing anomaly, mentioned earlier, is based largely on the age progression of volcanism during the last 15 m.y., from the Idaho-Nevada border to the Yellowstone caldera (Armstrong et al., 1975) and on agreement between this vector and the inferred absolute motion of the American 165 ~ ,°sO ,~_,-~ sit ~ .0 , . . . ,' .~1/ oU I ~ \ :' ,, ~ \ I'm! / ~ as. O / By; vale' _ of\/ \ ~ I / IF \ , ~ So of `q 'a Cot FIGURE 14.4 Major northeast-trending Cenozoic volcano- tectonic zones in the western United States (shown by solid lines where best constrained, dashed lines where less certain). "Rail- way track" indicates middle Cenozoic subduction boundary be- tween American and Farallon plates (base map adapted from Atwater, 1970). plate (Minster et al., 1974). However, geological and geo- physical studies in the Yellowstone~nake River region cast doubt on the mantle-plume model. The Pb isotopic evidence, already mentioned, apparently requires a man- tle source for the basaltic magmas that has been partofthe American plate for the past 2.8 b.y. Recent geological and geophysical studies in the region have documented other complications. In addition to the nor~east-trending age progression during the past 15 m.y., there is an opposing northwest-trending progression—toward Newberry cal- dera in Oregon (MacLeod et al., 1975~. The Yellowstone- Snake River zone lies along a regional northeast-trending structural and aeromagnetic lineament that extends from northeastern Nevada, beyond the Yellowstone caldera, into Canada. According to Eaton et al. (1975), "if the Yel- lowstone magma body marks a contemporary deep mantle plume, this plume, in its motion relative to the American plate, would appear to be 'navigating' along a funda- mental structure in the relatively shallow and brittle lith- osphere overhead. The concept that a northeastward- propagating major crustal fracture controls the migration path of the major foci of volcanism is at least equally favored by existing data." Another major northeast-trending igneous zone is the long-recognized Colorado mineral belt, which consists of aligned Upper Cretaceous~ower Tertiary (Laramide) in- trusives and locally preserved extTusives. No age progres- sion is evident along the Colorado mineral belt, but, as first pointed out by Tweto and Sims (1963), this belt fol- lows structural trends of Precambrian ancestry. The Colo- rado mineral belt has been proposed to constitute only part of a very long northeast-trending structural zone that extends from Arizona through Colorado, across the North- er;1 Plains, to join a major boundary between Precambrian age provinces in the Great Lakes region (Warner, 1978~. Farther to the south, another northeast-trending late )

166 FIGURE 14.5 Space time relations of Cenozoic volcanism along the Spnngervill - Raton zone. A, Distribu- tion of Pliocene and Pleistocene vol- canic field along the Spungervill~ Raton zone (stippled). B. Plot of age of volcanic activity versus position along the Spnngervill - Raton zone. Sloping line indicates a migration rate of 2.5 ctn/year, the observed migration rate for inception of silicic activity along the Snake River Plain-Yellowstone zone; no such age - Stance trend is evident for Me Spnngerville-Raton zone. Cenozoic volcanic feature (Figure 14.SA), referred to venously as the Spnngerville~aton zone, the Jemez zone, or the Raton volcanic chain, has also been proposed as marking the trace of a mantle plume (Suppe et al., 1975~. This major locus of late Cenozoic activity includes the Pinacate basalt field in northwest Mexico, the Spnngerville-White Mountain volcanic zone in eastern Arizona, the Mount Taylor and lemez volcanic fields in PETER W. LIPMAN 37o A 35r 1 1 4° 1 09° or / ll §F _r _ . 2 / ~ / '—V/ COLORADO . PLATEAU l 31 ! ~ ~ ~ ~. . .-. . . . . BASIN AND I RANGE C Capulin area CM Cimmaron Mountains G Gila Bend ~ Jemez Mountains O Ortiz Mountains P Pinacate volcanic field Q Ouesta area R Rabbit Ears area SF San Francisco volcanic field S S~ringerville area 0 O _i l MD Mogollon-Datit volcanic field MT Mount Taylor T Taos Plaueau W White Mountains Z Zuni Mountains l J ~ HIGH PLAINS j _1 ~ 11 \~ MEXICO ~ too 200 300 KILOMETERS 1 1 1 1 EXPLANATION 1 ~ 1 ! 1 ~ 1 c,0 30 c U" 0-25 z o ~ 20 z z '0 1 5 t _ ,,1 ol :~4 1 000 800 600 400 200 0 NE \ 800 600 400 SW DISTANCE ALONG VOLCANIC TREND. IN KILOMETERS B I- New Mexico, volcanics in the Rio Grande rip on the Taos Plateau and at the Cerros del Rio, and basalts of the High Plains of nor~eastem New Mexico. In contrast to the Yellowstone zone, no age progression is evident, either for Me upper Cenozoic basaltic rocks or for more silicic earlier Cenozoic activity along this zone (Figure 14.5B); and a mantle-plume origin seems less probable than con- trol by a crustal flaw (Mayo, 1958; Laughlin, 1976~. The

Cenozoic Volcanism in the Western United States Springerville-Raton zone follows general Precambrian structural trends of the region, such as the northeast- trending basement trends Great length on the Colorado Plateau (Shoemaker et al., 1974), and coincides in a gen- eral way with a major boundary between Precambrian age provinces. Analysis of patterns of Cenozoic volcanic activity sug- gests that these northeast-trending zones are key features in interpreting Cenozoic tectonics of the western United States- features identifiable only by the presence of vol- canic activity along them. Another aspect of tectonic control of volcanism that has implications for structure of the lithosphere is the clear pattern of recurrence of igneous (and tectonic) activity in the same places in the western United States through time. Such recurrence of activity was documented by Snyder et al. (1976), who delimited five major "magmatic loci" in the western United States. The distinction among several of these loci is not entirely convincing, for exam- ple, Weir Nevada locus is separated from the Idaho- Montana locus to the north only by younger rift-related volcanic cover along the Snake River Plain (Snyder et al., 1976~. In contrast, the Colorado Plateau was a stable "miniplate" during late Cretaceous-early Tertiary Lara- mide deformation and igneous activity, during the middle Cenozoic predominantly andesitic volcanism, and during the late Cenozoic fundamentally basaltic volcanism and associated extensional block faulting. The recurrence and general confinement of Cenozoic volcanism to regions where previous episodes of magnmatism and tectonism may have healed, annealed, and weakened the litho- sphere seems an important consideration in interpreting the distribution of Cenozoic volcanic activity in the west- em United States. PLATE-TECTONIC INTERACTIONS Continental volcanism also provides evidence for the na- ture of plate interactions in the western United States during the past 100 m.y. For relatively young interactions the seafloor-spreading record is sufficiently complete to permit fairly reliable reconstructions, but for the western Cordillera of North America the plate-tectonic history is complex, the seafloor record is largely missing, and many aspects can be reconstructed only from igneous and tec- tonic activity recorded on the continental plate. Interpre- tation of this record is still very incomplete and uncertain. A contrast exists in the western United Sates between earlier Cenozoic volcanic assemblages, which are pre- dominantly andesitic in composition and are inferred to be related to convergence and subduction along the west- em edge of the American plate, and a younger assemblage of fundamentally basaltic volcanism, in places including bimodal basalt-rhyol~te associations, that appears related to extensional tectonics within the American plate (Lip- man et al., 1971, 1972; Christiansen and Lipman, 1972; Snyder et al., 1976~. The change Tom predominantly an- desitic to fundamentally basaltic volcanism has been cor- 167 related in a general way with changing boundaries be- tween the American and various Pacific plates (Atwater, 1970~. Changes in volcanism accompanying the initial in- tersection of the Pacific and American plates are thought to be especially significant; these changes terminated the middle Tertiary and earlier subduction system and ini- tiated the evolving transfo~ boundary ofthe San Andreas Fault system, bounded at both ends by migrating triple junctions. Since first application of this plate model to changing patterns of volcanism in the western United States (Lip- man et al., 1971, 1972; Christiansen and Lipman, 1972), new data have accumulated, and furler evaluation is ap- propriate. Refinements of Cenozoic volcano tectonic pat- ten~s in terms of generally similar plate models for the western United States include those of Snyder et al. (1976), Coney and Reynolds (1977), Cross and Pilger (1978), Dickinson and Snyder (1979~. The re-examination presented here focuses on the predominantly andesitic volcanic suite, inferred to reflect subduction-related re- gimes. Aspects of the succeeding fundamentally basaltic volcanism and associated extensional tectonics are dis- cussed elsewhere (see Chapter 9~. Important variables that could affect distribution and compositional variation in subduction-related andesitic suites include dip of the Benioff zone, rate of plate con- vergence, angle between plate boundaries and conver- gence direction, temperature and thickness of descending slab, warps or breaks in the descending slab, depth of magma generation, composition of andesitic magma, dis- tance from the trench to volcanic front, and width of the volcanic belt. Compositions of convergent volcanic suites correlate with depth to the Benioff zone, becoming more potassic and alkalic with increasing depth, and vary also from intracontinental to intraoceanic settings (Dickinson, 1975~. Volcanoes occur at present above active Benioff zones over a depth range of about 75 300 km. Relatively rapid convergence favors more gently dipping Benioff zones (Luyendyk, 1970) and delayed heating of the de- scending slab; accordingly, such environments favor in- creased distances between trench and volcanic front, broader volcanic belts, and more alkalic volcanism. Oblique convergence, transitional toward transform mo- tion, would reduce the effective convergence rate. In ac- tive convergent systems, offsets of the volcanic front cor- relate with segmentation of the Benioff zone, requiring transverse breaks or flexures in the descending slab (Carr et al., 1973; Stoiber and Carr, 1974~. The behavior of the descending slab may also reflect the age of the oceanic lithosphere, which becomes cooler, thicker, and more dense with age and distance from the spreading ridge; older lithosphere should therefore sink more rapidly and preserve its physical identity to greater depths in a sub- duction environment than thin, hot, buoyant, young lith- osphere (Molnar and Atwater, 1978~. Complex, and as yet imperfectly understood, interplay of these variables seems capable of accounting for much of the diversity of arc volcanism. Some possible effects of

168 change in rate of plate convergence are illustrated dia- grammatically in Figure 14.6. In a subcontinental en- vironment, the structural and compositional complexities of Me continental lithosphere, discussed earlier, also seem capable of fiercer complicating the volcanic pattern, as illustrated by relations from the western United States. The following discussion is based on a series of figures Mat show in generalized fashion Me distribution of ig- neous activity Trough Cenozoic time in 10-m.y. incre- ments (Figure 14.7~. Through most of Mesozoic time and continuing to about 80 m.y. ago, igneous activity was confined to near the western margin of Me American plate (Figure 14.7A). Gra- nitic and cogenetic volcanic rocks as young as about 8() m.y. occur in Me Sier a Nevada batholith, but Mesozoic igneous rocks are sparse east of western Nevada. The bend in the igneous trend in the Pacific Northwest is probably due largely to Cenozoic deformation (Hamilton and Myers, 1966; Simpson and Cox, 1977~. Local east- A TYPICAL CONVERGENCE B INCREASED CONVERGENCE 1 ;~ c DECREASED CONVERGENCE - W~ i so FIGURE 14.6 Possible effects of change in convergence rate on condnen~l-margin subduction and related volcanism. A, Typical convergence rate: subduction zone dips about 30 degrees, and related volcanoes are in a relatively narrow arc, are low in potas- sium content, and are located relatively close to trench. B. In- creased convergence rate: continental plate ovemdes descend- ing slab at a lower angle, resulting~in a broader, more diffuse volcanic arc located farther from the Bench. Depending on depth to subduction zone, volcanic rocks can vary widely in composi- tion from low to high potassium. If the dip of the subduction system becomes very shallow, volcanism may cease entirely, as has probably happened in modern southern Peru (Barazangi and Isacks, 1976). C, Decreased rate of convergence: downgoing slab sinks gravitationally or breaks and reestablishes itself in a steeper orientation, causing arc volcanism to migrate toward the trench. Increased sinking of the descending slab requires coun- tedlow of asthenosphenc mantle into the region above the slab, possibly heating the base of the continental lithosphere and driv- ~ng extensional rifting and basaltic volcanism within the region of earlier arc volcanism. PETER W. LIPMAN west variations in distribution and composition of granitic rocks in the central Sierra Nevada are compatible with shifting depths of magma generation along a subduction system of relatively constant geometry (Dickinson, 19701. More limited, similar data along over transects across the batholithic belt suggest possibly more complex shifts in geometry of the subduction system, especially in the Klamath Mountains, where low-potassium quartz diorites and trondhjemites tend to be concentrated on the eastern side of the plutonic belt (Hotz, 1971), and also in the Peninsular Ranges batholith to the south (Silver et al., 1975~. Nevertheless, through late Mesozoic time, a long- lived, relatively steeply dipping, subduction system ap- parently was maintained, along the western margin of the American plate, concurrently with consumption of an eastern Pacific plate. Starting at about 80 m.y. ago, concurrently with inferred accelerated convergence between the American and east- em Pacific plates (Coney, 1972), igneous activity mi- grated eastward in the western United States (Figure 14.7B), as recognized long ago by Lindgren (1915~. The most dramatic eastward shift between 80 m.y. and 70 m.y. ago- (Figure 14.7B) occurred in the northwest, where plu- tonic rocks of the Boulder batholith and associated Elk- hom Mountains volcanics were emplaced as far east as western Montana; however, igneous activity of this age also moved eastward in Nevada and Arizona. Pattems of igneous activity continued to evolve rapidly and had changed notably by 70 m.y. to 60 m.y. ago (Figure 14.7C). The locus of intense activity associated with the Boulder batholith in western Montana extinguished abruptly about 70 m.y. ago. Many K-Ar ages in the Idaho batholith could indicate continuation of igneous activity; more likely, they reflect reheating by pervasive Eocene activity (Armstrong et al., 1977~. To the south, major activ- ity of Laramide age continued in southem Arizona and began to extend into southwestern New Mexico, consti- tuting the peak connation of porphyry copper deposits in this region. A southeasterly age progression of these eco- nomically important intrusives, suggested as a mantle- plume trace (Livingston, 1973), seems more likely part of the regional eastward migration of igneous activity when considered with distributions of other intrusions of roughly similar age in the region. Between northern Idaho and southem Arizona, He only other significant locus of igneous activity at this time was the northeast- trending zone of intrusions and associated volcanics along the Colorado mineral belt. These represent the most east- em sweep of early Tertiary igneous activity in the central and southern Rocky Mountain region. Thus, early Tertiary igneous activity in the we stem United States covered a broad region but was defuse, discontinuous, and nonsynchronous in time, reaching the eastern margins of the Cordillera earlier in the northern Rockies than farmer south. Even with abundant new data, these patterns reasonably seem related to effects of flat- tening of He subduction system during the early Tertiary (Lipman et al., 1971; Coney and Reynolds, 19771. Flatten-

Cenozoic Volcanism in the Western United States ,~ ~ ..2'.,'.~2 '. ,'~'..2 . ,2 ,~ ' ':: .' .' " ." ~: '''' . 2. i:: .:...2 .., W: ~~: /'' }/ / \// ~ FIGURE 14.7 Generalized distribution in the western United States of predominantly andesitic volcanic suites, inferred to be related to subduction. Distributions are based on compilations (Lipman et al.' 1972; Snyder et al., 1976; Stewart and Carlson, 1976; Armstrong et al., 1977; Cross and Pilger, 1978) and on descriptions of local areas too numerous to cite individually. The base maps and diagrammatic plate geometry are from Atwater (1970) and Atwater and Molnar (1973). No attempt has been made to remove effects of late Cenozoic extensional and rotational deformation, even though such effects are probably large (Hamilton and Myers, 1966). Northeast-t~ending lines mark approximate traces of the Snake River-Yellowstone zone, the Colorado mineral belt, and the Springenrill~Raton zone. 169

170 ing of the subduction zone may also account for the onset of Laramide foreland deflation (Lipman et al., 1971; Lowell, 1974; Coney, 1976; Dickinson, 1977~. Early extinction of igneous activity in the northern Rockies in the very region where activity first migrated eastward may have been a consequence of further {lat- tening of the subduction system, to such a low angle that igneous activity could no longer be sustained. A modern analog would be the southern Peruvian Andes, where a currently inactive segment of the Andean volcanic chain is underlain by a Benioff zone that dips only about 20° (Stauder, 1975; Barazangi and Isacks, 1976), although the geometry of even this modern plate boundary is currently controversial (Iarnes, 1978~. The confinement of early Tertiary igneous activity in the southern Rockies to the Colorado mineral belt, despite apparently continuous subduction at Me western margin of the plate, suggests the presence in this region of lithosphere that is too thick, cold, or rigid to permit penetration of subduction-related magmas, except along major structural flaws. Any igneous expression of passage of a Kula-Farallon-American plate triple junction, infected to have swept northward in the early Tertiary (Atwater, 1970), is obscure. The distribu- tion patterns (Figure 14.7C) do suggest, somewhat incon- clusively, a possible offset in the downgoing slab in the region between the Colorado mineral belt and the Springerville-Raton zone. The significance of such offsets in controlling distribution of igneous activity in the west- em United States, inferred also by Cross and Pilger (1978), is more apparent for later Cenozoic volcanism. Examination of compositional data for reliably analyzed suites of Laramide rocks should be useful in evaluating the validity ofthis possible early Tertiary offset in igneous pattems. For the interval 60~0 m.y. ago, the most notable changes in the igneous patterns (Figure 14.7D) are a flareup of volcanic activity in the northern Rockies, the virtual extinction of activity in the southern Rockies, and the continued slow, eastward migration of activity in Ari- zona and New Mexico. In the northern Rockies, volcanic activity broke out widely during this interval (mostly about 55 m.y. ago), including the Lowland Creek vol- canics in western Montana, the Absaroka volcanics in northwestern Montana, and the Challis volcanics in cen- tTa1 Idaho. Concurrent alkalic activity was also occulting in Montana. The reason for this increased activity after a 10-m.y. to 15-m.y. pause is not clear. Perhaps it reflects the presence of a more steeply dipping subduction system due to a decreased rate of plate convergence. Altemative- ly, a new segment of subducted slab may have been em- placed with a steeper dip. With Farallon-American plate convergence rates estimated at 10 cm/yr or more for this period (Cagney, 1976), newly subducted seafloor could have arrived beneath the eastern Cordillera within 10 m.y. Thus, the geometry of the subduction system could have shifted rapidly, perhaps more rapacity than can be resolved from radiometric dates on middle Tertiary vol- canic rocks, especially if changes in the rocks lag several PETER W. LIPMAN million years after a shift in plate geometry (Christiansen and Lipman, 1972; Snyder et al., 1976~. Farther south, in the Colorado sector, little igneous ac- tivity seems to have occurred in the interval 6040 m.y. ago, although a few intrusive dates suggest sparse con- tinued activity. Along the trend of the Springerville- Raton zone, a major northeast-trending discontinuity with the subduction system to the south is evident (Figure 14.7D). In the southwest, dominant activity moved farther eastward, occurring mainly in New Mexico, and tapered off in intensity after about 55 m.y. ago. This pattern is interpreted, following Coney and Reynolds (1977), as pri- marily reflecting continued flattening of the subduction zone in that sector, although their analysis is simplified by assumption of a constant 150-km depth of magma genera- tion. The ~ramide and Tertiary igneous rocks in this region become more alkalic to the east, indicating genera- tion from progressively greater depth as well (Lipman et al., 1971~. During the interval 50~0 m.y. ago, changes in the pat- tems of volcanic activity were relatively minor (Figure 14.7E). In the Northwest, continuing activity in the Chal- lis and Absaroka fields in Idaho and Wyoming began to wind down about 45 m.y. ago, but similar activity flared up farther south in northern Nevada (StewaIt et al., 1977) and in such parts of northern Utah as the Paris City district (Bromfield et al., 1977~. The nonlinear trend in volcanic activity in the Northwest at the end of this interval seemingly requires a flexure or discontinuity in the sub- ducted slab, changing from relatively steeply dipping in the Northwest to more gently dipping for the Nevada- Utah~outhern Idaho sector. The virtual lack of activity in the Colorado sector during this interval suggests the pos- sible continued presence of an even more gently dipping subducted slab under this region, and the east-west- trending zone of activity in the Idaho-Utah-Nevada re- gion may have been localized along a slab flexure (Stewart et al., 1977~. Initial development of this slab flex- ure and associated complex geometry of volcanic activity in the northwestern United States occurred near the site of the fixture Snake River-Yellowstone zone at about 45~0 m.y. ago, roughly coincident with changed motions of the major tectonic plates, as reflected in the Pacific Basin by the bend in the Emperor-Hawaii seamount chain. The effect of this reorganization was probably to reduce the convergence rate between the American and Farallon plates (Coney, 1972; 1976~. Major changes occurred during the interval 40~0 m.y. ago (Figure 14.7F). Andesitic activity began along the Cascades in western Washington and Oregon, and activity terminated to the east in Idaho and Wyoming. Inception of Cascade volcanism may in part be related to foundering and steepening of the preceding low-angle subduction system, but the position of the trench and subducted slab also probably shifted westward, concurrently with attach- ment of the Eocene submarine basaltic volcanics of the Coast Ranges of Oregon and Washington (Snavely et al., 1968) to the American plate. Concurrently, volcanic activ-

Cenozoic Volcanism in the Western United States ity extended farther south in Nevada and Utah (Stewart et al., 1977), accentuating the L-shaped pattern of activity in the Northwest. Beginning at about 40 m.y. but only spreading widely by about 35 m.y. ago, renewed igneous activity also broke out in the southern Rocky Mountains in Colorado, New Mexico, and extending as far southeast as western Texas, as well as in the Sierra Madre Occidental in Mexico. This complex pattern of volcanism suggests the pres- ence of a steep, east-dipping subduction zone under the Cascades, an oblique-trending flexure in the subducting plate under northern Nevada and Utah, and a more gentle, east-dipping zone to the south. Southward migration of the oblique-trending flexure beneath Nevada and Utah in middle Tertiary time (Stewart et al., 1977) might have been due to oblique subduction in a southeasterly direc- tion of a descending slab of stable geometry, although reconstructions of plate motions requiring northwesterly convergence to account for disappearance of the Kula plate (Atwater and Molnar, 1973) do not readily fit such a pattem. Also, the flexure appears to have been generated relatively abruptly about 40 m.y. ago in the Pacific North- west. Perhaps the flexure migrated southward within the descending plate, independent of absolute plate motions, accommodating the change from low-angle to steep sub- duction. Apparent offsets along all three lithospheric structural zones at this time (Figure 14.7F) may in part be deceptive, especially for the Springerville-Raton zone. There the boundary is also largely coincident with the southeast margin of the Colorado Plateau, which appears to have been nearly impenetrable to magma through Cenozoic time. An important discontinuity in volcanic patterns am pears to have been virtually coincident with the Colorado mineral belt, however, separating the major activity in the southern Rocky ,SIountains—more than 1000 km from the plate margin from compositionally similar volcanic fields in Nevada and Utah. This offset is interpreted as marking a major tear in the subducting plate where the change in dip was too great to be accommodated by the oblique flexure under Nevada and Utah. Such a disconti- nuity In the descending slab seems a preferable interpre- tation of the complex compositional patterns of middle Tertiary volcanism, for which duplication of subduction zones in an imbricate geometry of subduction was pro- posed earlier (Lipman et al., 1971; 1972~. The period 30-20 m.y. ago was an especially complex time in terms of evolving volcano-tectonic patterns, al- though the patterns of ir~termediate-composition volcanic activity changed only slightly from those of the previous interval (Figure 14.7G). Activity continued to spread southward in the Great Basin, and the offset along the Colorado mineral belt is even more conspicuous. This interval also includes the time of initial interaction be- tween the Pacific and American plates, and the 10-m.y. time slice ~ too coarse to portray effectively several sig- nificant concurrent changes in volcano tectonic patterns on the American plate. Basaltic volcanism and associated 171 extensional deformation (Christiansen and Lipman, 1972) were occurring nowhere in the region 30 m.y. ago but were widespread by 20 m.y. ago. Extensional faulting and associated basaltic volcanism apparently began first along the Rio Grande rift system, about 29-26 m.y. ago (Lipman and Mehnert, 1975; Chapin and Seager, 1975), near the interior of the Cordilleran belt, rather than closer to the continental side of the young, growing San Andreas trans- form. In the period 29-21 m.y. ago, transitional, relatively alkalic volcanism occurred west of the Rio Grande rid in New Mexico and Arizona; this volcanism was not charac- terized by welI-defined features of either the predomi- nantly andesitic or the fundamentally basaltic types that were significant earlier and later in the region (Elston et al., 1976~. Concurrently, kimberlitic and undersaturated alkalic basaltic volcanism occurred on the Colorado Pla- teau. These transitional volcanic rocks seem most reason- ably interpreted as related to termination of subduction (Christiansen and Lipman, 1972), either by a rapid retro- grade steepening of a subduction system that is no longer being actively regenerated at the plate boundary (Coney and Reynolds, 1977) or by slow foundering of a relatively gently dipping slab, perhaps weakened by the developing '~slab window" to the west (Dickinson and Snyder, 1979~. In either case, counterflow of asthenospheric mantle would be required to accommodate the sinking slab (Figure 14.6C), a process that could account for initiation of extension and basaltic volcanism at the eastern side of the subduction system, as observed for the Rio Grande rift. Thus, the Rio Grande rift, which also follows the trends of preceding Cenozoic volcanism and tectonism, can be considered an intracont~nental analog of oceanic marginal basins, even though the cause of the extension here was termination of the subduction system. An anal- ogous, but less well-documented, association between extensional faulting, andesitic volcanism, and steepening of subduction, may also exist for the Eocene volcanic areas in the Pacific Northwest, as recognized by Davis (see Chapter 8~. As considered below, widespread younger extension in the southern Basin-Range province and in the Great Basin also is broadly associated with steepening and destruction of the subduction regimes in these regions. By the interval 20-10 m.y. ago (Figure 14.7~), andesitic volcanism in the western United States was restricted to a narrow belt near the margin of the American plate and was represented by a southem extension into California and Nevada of the Cascade volcanic arc. During this in- terval, the southern limit of andesitic volcanism migrated northward, following passage ofthe plate-boundary triple junction and termination of active subduction (Christian- sen and Lipman, 1972) but showing a delay of about 5 m.y. in the transition (Snyder et al., 1976~. This reduced region of subduction-related volcanism, especially in Nevada and Utah, probably requires some steepening of the sub- duction zone in this sector. An important problem at this time, for which present plate models do not offer ready

172 explanation, is the episodic nature of subduction-related volcanism (Gilluly, 1973~; peaks of activity in the Cascade Range appear to correlate with outbursts on the Columbia Plateau, the western Snake River Plain, the Great Basin, and even with other continental-margin arc systems around the Pacific (McBirney et al., 1974~. Abler about 17 m.y. ago, a pattern of extensional tec- tonics accompanied by fundamentally basaltic volcanism, characterized by ENE-WSW directions of extension normal to the plate boundary, became established the length of the Great Basin and extending as far north as the Columbia Plateau (Christiansen and McKee, 1978; see Chapters 8 and 9~. These tectonic patterns and major asso- ciated volcanism, including eruption of the Columbia River basalt and first volcanic activity along the Snake River-YelIowstone trend, are thought to have been ini- tiated, largely in a back-arc environment, when the sub- duction system of a Juan de Fuca plate reached a critically diminished size as a result of impingement from both south and norm by advancing triple junctions. Subse- quent evolution ofthe patterns of extensional defonnation and associated fun~nent~ally basaltic volcanism, includ- ing nearly concurrent reorientation of the direction of ex- tensional faulting (Zoback and Thompson, 1978), activa- tion of the Snake River-Yellowstone zone (Chnstiansen and McKee, 1978), and voluminous eruption of the Columbia River basalt group (Swanson et al., 1975), prob- ably reflect complex interplay between the enlarging traDsfonn boundaries, the diminishing Juan de Fuca plate, and structures of the American plate, discussed elsewhere in this volume (see Chapters 8 and 91. PERSPECTIVE S Complex space~me~omposition patterns of Cenozoic volcanism in the western United States offer infonnation on the composition and structure of Me continental litho- sphere, as well as provide some of the best available con- straints on geometry of past interactions between Me American plate and various Pacific plates. These volcanic rocks are most readily interpreted in terms of two broad assemblages: an earlier Cenozoic, predominantly ande- sitic assemblage inferred to result from geometrically complex processes of plate convergence and a later, fun- damentally basaltic assemblage associated with exten signal tectonic settings both in regions behind continuing subduction systems especially where slab dip has steed ened and in regions where subduction has te,~'inated. Adequate age and compositional data are still lacking for many areas, and tight brackets on changes in volcanic type relative to tectonic settings are especially sparse. Many problems also remain in evaluating the primary sources of subducdon-related magmas: lower crust, Iithospheric upper mantle, asthenosphere; or the subducted oceanic slab. The distribution of Cenozoic volcanic suites of inter- mediate composition is interpreted as due to changing PETER W. LIPMAN geometry of subducted slabs of venous Pacific oceanic plates, modified by major structural features ofthe overly- ing American plate. Possibly important variables of the subducted slab include changes in dip, physical continu- ity, temperature, and thickness. An important question is whether the apparent changes in dip of Tertiary subduc- tion systems under the western United States result from defonnation of an essentially continuous slab as a result of changing rates and directions of convergence, or whether changes in dip result Tom detachment of the downgoing slab and rapid re-establishment of the subduction system in a new orientation. Also, how significant is the apparent correlation between increased slab dip and back-arc ex- tension? Particularly problematic are mechanisms by which major structural features of the Amencan plate may have influenced the distribution of subduction-related vol- canism, especially the northeast-trending zones of Pre- carnbnan ancestry, "miniplates" such as the Colorado Plateau, and upper-crustal batholiths that may be asso- ciated with structurally disturbed zones in the deep litho- sphere. If"gravitational anchors" associated with batho- lith fortnation extend entirely through the continental lithosphere, as seems likely for the Yellowstone region, can they affect interactions between tectonic plates; Can such perturbations along the nor~east-trending structural flaws actually deform gently dipping subducted slabs, or, alternatively, are lithospheric discontinuities reactivated whenever they are appropriately oriented to be utilized by magmas ascending from subduction systems? Questions such as these should present prime problems for many years for geologists, geophysicists, and geo- chemists interested in relations between volcanism and continental tectonics. - REFERENCES Annstrong, R. L., W. P. Leeman, and H. E. Malde (1975). K-A r dating, Quaternary and Neogene volcanic rocks of the Snake River Plain, Idaho, Am. 1. Sci. 275, 225-251. Annstrong, R. L., W. H. Taubeneck, and P. O. Hales (1977). R~Sr and K-Ar geochronomet~y of Mesozoic granitic rocks and their Sr isotopic composition, Oregon, Washington, and Idaho, Geol. Soc. Am. Bull. 88, 397~11. Atwater, T. (1970). Implications of plate tectonics for the Ceno- zoic tectonics of western North Arnenca, Geol. Soc. Am. Bull. 81, 3513 3536. Atwater, T., and P. Molnar (1973). Relative motions ofthe Pacific and North American plates deduced from sea-floor spreading in the Atlantic, Indian, and South Pacific oceans, in Proceed- ings of the Conference on Tectonic Problems of the San An- dreas Fault System, K. L. Kovach and A. Nur, eds., Stanford U. Publ. Geol. Sci. 13, pp. 136-148. Balazan~i. M.. and B. L. Isacks (1976). Spatial distribution of earthquakes and subduction of the Nazca plate beneath South Amenca, Geology 4, 686~;92. Bromfield, C. S., A. J. Erickson, Jr., M. A. Haddadin, and H. H. MeLnert (197 7). Potassium argon ages of intrusion, extrusion, and associated ore deposits, Park City mining district, Utali, Econ. Geol. 72, 837 848.

Cenozoic Volcanism in the Western United States Carr, .~1. J., R. E. Stoiber, and C. L. Drake (1973). Discontinuities in the deep seismic zones under the Japanese arc, Geol. soc. Am. Bull. 84' 2917-2930. Chapin, C. E., and W. R. Seager (1975). Evolution of the Rio Grande rift in SOCOITO and Las Cruces areas, in New Mexico Geological Society Guidebook, 26th Field Conferenc',, pp. 297~21. Christiansen, B. L., and P. W. Lipman (1972). Cenozoic ~ol- canism and plate-tectonic evolution of the western United States. Part 2, late Cenozoic, Proc. Roy. Soc. London 271, 249-284. Christiansen, R. L., and E. H. McKee (1978). Late Cenozoic vol- canic and tectonic evolution of the Great Basin and Columbia intermontane region, in Cenozoic Tectonics enc. Regional Geophysics of the Western Cordillera, R. B. Smith and G. P. Eaton, eds., Geol. Soc. Am. Mem. 152, 28~312. Coney, P. J. (1972). Cc!rdilleran tectonics and North American plate motion, Am. J. Sci. 272, 603428. Coney, P. J. (1976). Plate tectonics and the Laramide orogeny, N. Men. Geol. Soc. Spec. Publ. 6, 5-10. Coney, P. J., and S. J. Reynolds (1977). Cordilleran Benioffzones, Nature 270, 403406. Cross, T. A., and R. H. Pilger (1978). Constraints on absolute motion and plate interaction inferred Mom Cenozoic igneous activity in the western United States, Am. J. Sci. 278, 86~902. Dickinson, W. R. (1970). Relations of andesitic volcanic chains, granitic batholith belts, and derivative graywacke-arkose facies to the tectonic framework of arc-trench system, Rev. Geo- phys. 8, 813~60. Dickinson, W. R. (1975). Potash~epth (K-h) relations in continental-margin and intraoceanic magmatic arcs, Geology 3, 53 56. Dickinson, W. R. (1977). Inferred plate tectonic setting of classic Amide orogeny, Ceol. Soc. Am. Abstr. Programs 9, 950. Dickinson, W. R., and W. S. Snyder (1979). Geometry of triple junctions and subducted slabs related to San Andreas trans- fonn, J. Geophys. Ftes. 84, 561~72. Doe, B. R. (1967). The bearing of lead isotopes on the source of granitic magma, J. Petrol. 8, 51 83. Eaton, G. P., R. L. Chnstiansen, H. M. Iyer, O. R. Mabey, H. R. Blank, Jr., I. Zietz, and M. E. Gettings (1975). Magma beneath Yellowstone National Park, Science 188, 787-796. Elston, W. E., R. C. Rhodes, P. J. Coney, and E. G. Deal (1976). Progress report on the Mogollon Plateau volcanic field, south- westem New Mexico, No. Surface expression of a pluton, N. Mex. Ceol. Soc. Spec. Publ. a, .~28. Fislce, B. S., O. T. Tobisch, R. A'. Kistler, T. W. Stern, and M. Tatsumoto (1977). Minarets caldera, a Cretaceous volcanic center in the Bitter Range pendant, central Sierra Nevada, California, Geol. Soc. Am. Abstr. Programs 9? 975. C.illuly, J. (1955). Geologic contrasts between continents and ocean basins, Geol. Soc. Am. Spec. Paper 62, 7-18. Gilluly, J. (1973). Steady plate motion and episodic orogeny and magmatism, Geol. Soc. Am. Bull. 84, 499~14. Hamilton, W., and W. B. Myers (1966). Cenozoic tectonics ofthe western United States, Be?;. Geophys. 4, 509 - so. Hamilton, W., and W. B. Myers (1967). The nature of batholiths, U.S. Geol. Sure. Prof. Paper 554-C, 30 pp. Hamilton, W., and W. B. Myers (1974). The nature ofthe Boulder batholith of Montana, Geol. Soc. Am. Bull. 85, 365~78. Hotz, P. E. (1971). Plutonic rocks of the Klamath Mountains, California and Oregon, L.S. Geol. Sure. Prof. Paper 684-B, 20 pp. Hurley, P. M., H. Hughes. G. Faure, H. W. Fairbaim, and W. H. 173 Pinson (1962). Radiogenic strontium-S 7 model of continent for- mation, J. Geophys. Res. 67, 5315~334. James, D. E. (1978). Subduction of the Nazca plate beneath central Peru, Geology 6. 174-178. Kistler, R. W., and Z. E. Peterman (1973). Variations in Sr, Rb, K, No, and initial Sr87/Sr86 in Mesozoic granitic rocks and intruded w allrocks in central Califomia, Ceol. Soc. Am. Bull. 84, 3489~512. Klepper, M. R., G. D. Robinson, and H. W. Smedes (1971). On the nature of the Boulder batholith of Montana, Geol. Soc. Am. Bull. 82, 1563-1580. Laughlin, A. W. (1976). Late Cenozoic basaltic volcanism along the Jemez zone of New Mexico and Arizona, Geol. Soc. Am. Abstr. Programs 8, 598. Lindgren, W. (1915). The igneous geology calf the Cordilleras and its problems, in Problems of American Geology, Yale tTniver- sity Press, New Haven, pp. 23~286. Lipma~n, P. W., and H. H. Mehnert ( 1975). Late Cenozoic basaltic volcanism and development of the Rio Grande depression in the southern Rocky Mountains, Geol. Soc. Am. Mem. 144, 1 19-154. Lipman, P. W., H. l. Prostka, and R. L. Christiansen (1971). Evolving subduction zones in the western United States, as interpreted from igneous rocks, Science 148, 821~25. Lipman, P. W., H. J. Prostka, and R. L. Christiansen (1972). Cenozoic volcanism and plate-tectonic evolution of the west- ern United States: 1. Early and middle Cenozoic, Phil. Trans. Roy. Soc. London 271, 217-248. Lipman, P. W., B. R. Doe, C. E. Hedge, and T. A. Steven (1978). Petrologic evolution of the San Juan volcanic field, southwest- em Colorado: Pb and Sr isotope evidence, Geol. Soc. Am. Bull. 89, 59~2. Livingston, D. E. (1973). A plate tectonic hypothesis for the gene- sis of porphyry copper deposits of the southern Basin and Range province, Earth Planet. Sci. Lett. 20, 171-179. Lowell, J. D. (1974). Plate tectonics and foreland basement defor- mation, Geology 2, 275-278. Luyendyk, B. P. (1970). Dips of downgoing lithospheric plates beneath island arcs, Geol. Soc. Am. Bull. B1, 3411~416. MacLeod, N. S., G. W. Walker, and E. H. McKee (1975). Geo- thermal significance of eastward increase in age of upper Cenozoic rhyolitic domes in southeastern Oregon, in Proceed- ings 2nd U.?`J. Symposium on the Development and Use of Geothermal Resources, Vol. 1, U.S. Government Printing Of- fice, Washington, D.C., pp. 46~74. Martin, R. F., and A. J. Piwinskii (1972). Pragmatism and tectonic setting,J. Geophys. Res. 77, 4966~975. Mayo, E. B. (1958). Lineament tectonics and some ore districts of the southwest, Olin. Eng. 10, 1169-1175. McBirney, A. R.. J F. Super, H. R. Naslund, K. G. Sutton, and C. M. White (1974). Episodic volcanism in the central Oregon Cascade Range, Geology 2, 585~89. Minster, J. B., T. H. Jordan, P. Molnar, and E. Haines (1974). Numerical modeling of instantaneous plate tectonics, J. Roy. Astron. Soc. Can. 36, 541~76. Molnar, P., and T. Atwater (1978). The age of subducted oceanic lithosphere: a possible cause of interarc spreading or Cor- dilleran tectonics, Geol. Soc. Am. Abstr. Programs 10, 138. Morgan, W. J. (1972'. Plate motions and deep mantle convection, Geol. Soc. Am. .Uem. 132, 7-22. Peterrnan, Z. E., B. R. Doe, and H. J. Prostka (1970). Lead and strontium isotopes in rocks of the Absaroka volcanic field, Wyoming, Contrib. Mineral. Petrol. 27, 121-130. Schweickert, R. A. (1976). Shallow-leve] platonic complexes in

174 the eastern Sierra Nevada, California, and Weir tectonic impli- cations, Geol. Soc. Am. Spec. Paper 176, 58 pp. Shaw, H. R., and E. D. Jackson (1973). Linear island chains in the Pacific: result of thermal plumes or gravitational anchors?, J. Geophys. Res. 78, 8634~652. Shoemaker, E. M., R. L. Squires, and M. J. Abrams (1974). The Bright Angel and Mesa Butte Fault systems of Noreen Ari- zona, in Geology of Northern Arizona: Part I, Regional Stud- ies, Karlstrom et al., eds., Geological Society of Amenca, Cor- dilleran Section Meeting, Flagstaff, Anz., pp. 3S[;~91. Silver, L. T., T. O. Early, and T. H. Anderson (1975). Petrological, geochemical, and geochronological asymmetries of the Penin- sular Ranges batholith, Geol. Soc. Am. Abstr. Programs 7, 375. Simpson, R. W., and A. Cox (1977~. Paleomagnetic evidence for tectonic rotation of the Oregon Coast Range, Geology 5, 585~89. Snavely, P. D., fir., N. S. MacLeod, and H. C. Wagner (1968). Tholeiitic and alkaIic basalts of the Eocene SiIetz River vol- canics, Am. J. Sci. 266, 454~81. Snyder, W. S., W. R. Dickinson, and M. L. Silberman (1976). Tectonic implications of space~ime patterns of Cenozoic mag- matism in the western United States, Earth Planet. Sci. Lett. 32, 91-106. Stander, W. (1975). Subduction of the Nazca plate under Peru, as evidenced by focal mechanisms and by seismicity,J. Geophys. Res. 80, 1053 1064. Steven, T. A., and P. W. Lipman (1976). Calderas of the San Juan volcanic field, southwestern Colorado, U.S. Geol. Sure. Prof. Paper 958, 35 pp. Stewart, J. H., and J. E. Carlson (1976). Cenozoic rocks of Ne- vada, Ned. Burl Mines Geol. Map 52. PETE R W . L IP MA N Stewart, J. H., W. J. Moore, and I. Zietz (1977). East-west pat- terns oiCenozoic igneous rocks, aeromagnetic anomalies, and mineral deposits, Nevada and Utah, Geol. Soc. Am. Bull. 8S, 67-77. Stoiber, R. E., and M. J. Carr (1974). Quaternary volcanic and tectonic segmentation of Central Amenca, Bull. Volcanol. 37, 304~25. Suppe, J., C. Powell, and R. Berry (1971;). Regional topography, seismicity, Quaternary volcanism, and the present-day tec- tonics of the western United States, Am. J. Sci. 275A, 397036. Swanson, D. A., T. L. Wright, and R. T. Heltz (1975). Linear vent systems and estimated rates of magma production and eruption for the Yakima basalt on the Columbia Plateau, Am. J. Sci. 275, 877~05. Tweto, O., and P. Sims (1963). Precambrian ancestry ofthe Colo- rado Mineral Belt, Geol. Soc. Am. Bull. 74, 991-1014. Warner, L. A. (1978). The Colorado lineament: a middle Pre- car;~-brian wrench fault system, Geol. Soc. Am. Bull. 89, 16 ~ -171. Wilson, J. T. (1965). Evidence from oceanic islands suggesting moo; ment in the earth, Phil. Trans. Roy. Soc. London 258, 145 - 6S. Zartman, R. E. (1974). Lead isotope provinces in the Cordillera of the western United States and their geologic significance, Econ. Wool. 69, 792~05. Zoback, HI. L., and G. A. Thompson (1978). Basin and Bange rifting in northern Nevada: clues from a mid-\liocene rift and its subsequent offsets, Geology 6, 111-116.

The Shape of North America cluring the Precambrian 1~ INTRODUCTION WILLIAM R. MUEHLBERGER The University of Texas Historically, the geological evolution of North America has been the subject of studies since they were first made on this continent. This largely led to the development of geosynclinal theory, which was rapidly applied to thick sedimentary sequences worldwide. In the first half of this century, continental drift was not widely accepted in North America, but the evidence suggested continental accretion. Continental accretion of North America was supported by the early isotopic studies of Precambrian rocks and the apparent progressive younging of Phanero- zoic erogenic belts toward the continental margins. Con- tinuing work showed the presence of older rocks in the cores of hinterlands of erogenic belts and thus led to the recognition that continental crust, which had been stabil- ized in some earlier event, could be reactivated, have extensive sedimental belts (geosynclines) form within it, be defonned, metamorphosed, intruded, and once again be a part of the stable interior of a continent. Currently being debated is whether most continental crust was formed early in geological time or whether it has 175 grown progressively through time. Engel and Engel (1964) presented a good summary of these opposing views. Their paper supports the latter view of continental accretion. This paper supports the early formation of much if not most of continental crust and continued accre- tion, but at a much slower rate. Hargraves (1976) presents a lucid discussion ofthe progressive growth of continental crust throughout geological time. The advent of plate tectonics has revolutionized geo- logical thinking. T he three end members rifts, trans- forms, and collisions (subduction~with their character- istic associations of rock types provide new insights into the interpretation of continental rocks older than the cycle of seafloor spreading preserved in the ocean floor. These concepts have been widely applied to the Phanerozoic with an ever-increasing recognition and understanding of the wide variety of events that can happen during the Wilson cycle of continental riding, drifting, closing, and suturing (e.g., Burke et al., 1977; Dewey, 1977~. Structural evolutionary rates must be gradually slowing down because of an overall cooling of the earth through geological time. Thus, convective "rollover," a character-

176 istic feature of oceanic lithospheric generation, cooling, and subduction should be slower today relative to earlier episodes. This implies that crustal generation should also be slower now relative to the early Precambrian. Figures 15.1-15.4* show, in a summary form, the pro- gressive stabilization and "growth" of the North Ameri- can craton during four intervals of the Precambrian. How far back in geological time can plate-tectonic processes be inferred? Late Paleozoic? Grenville? Hudsonian/Peno- kean? Archean? If plate-tectonics processes; can be invoked in the Precambrian we have yet to detennine most of the interrelationships of active and passive mar- gins, what was nEted off, what was accreted, or what was involved in major transform margins. Three divisions of geological time are commonly used in discussing plate tectonics and its applicability to tec- tonic features: Archean, Proterozoic, and Phanerozoic (Anhaeusser, 1975; Burke et al., 1977; Sutton, 1973; Windley, 1977~. These are diachronous in space and time but roughly correspond to the standard geological time scale. They represent major intervals of differing re- sponse of "stable" cratons to erogenic deformation. ARC HEAN The Archean had a style that was unique and that was essentially terminated about 2.5 billion years (b.y.) ago in North America, earlier in some continents, later in others. The development of extensive gray gneisses, terranes of soda-nch granitic composition, is unique to this period. These probably represent the early formation of conti- nental crust by processes that are still poorly understood (Glikson, 1976; Goldich and Hedge, 1974; Goodwin, 1974a, 1974b; Moorbath et al., 1972; Myers, 1976~. The study of lunar samples has accelerated attempts at un- raveling the earliest history of earth; comparative plane- tary studies are an obvious important topic needing major research to fill in the gap between the terrestrial rock record and the origin of the earn. The abundance of greenstone graywacke belts is an- other association that is nearly unique to the Archean. The peridotite/komatiite basal unit is also nearly unique- it requires mantle temperatures well above those available today for magma generation [two early Proterozoic ex- amples (Arndt et al., 1977) and a possible Ordovician example (Upadhyay, 1978) have been identified thus far]. Green (1975) also shows that with his preferred steep isotherm` model for the Archean, eclogite could not fonn and subduction would not occur; thus, the basaltic oceanic crust would be scraped off against primitive sickle nuclei. The absence of Archean alkaline complexes, *Principal sources for all figures in this paper are Baer, 1970; Bayley and Muehlberger, 1968; Drummond, 1974; Goodwin, 1974a; Jackson and Taylor, 1972; King and Beikman, 1974; Lidiak, 1971; Lidiak and Zietz, 1976; Muehlberger et al., 1967; Price and Douglas, 1972; Stewart, 1976; Tectonic map of Canada, 1969. WILLIAM R. MUEHLBERGER carbonatites, or kimberl~tes, all characteristic of stable cratons (Windley, 1977) also attest to the mobility of the Archean crust. Most workers do not invoke plate-tectonic regimes dur- ing the formation of the Archean l~igh-grade gneiss ter- ranes or the low-grade greenstone belts. The greenstone belts have been argued to be sufficiently distinctive (Anhaeusser et al., 1969; Douglas and Price, 1972; Engel and Kelm, 1972) that there is no comparable environment in present-day erogenic belts. The many theories that have attempted to explain the evolution of greenstone belts are summarized by Lindsey (19~`, pp; 4.~81. Structural style is unique and dominated by a developing gravitational instability as mafic volcanic piles and ad- jacent region were intruded by soda-rich melts and mig- matized. These piles subsided into a thin lithosphere with folding occurring as the volcanic piles slid toward the subsiding flysch basins. These Men were injected by potash-nch granites with later stabilization (cooling) of He developing continental mass. Vertical motions appear to dominate the tectonics of this environment, although initial undulations have been interpreted to be the result of the flexing of thin lithosphere plate on the overriding side of ~ subduction zone or back-arc spreading (Decry, 1977; Fyson, 1978~. Goodwin (1974b) proposed that the initial crustal units developed around an early Archean paleoplume centered in Hudson Bay. These units aggregated to form a large craton by 1.7 b.y., which underwent periods of epeir- ogenic uplift, rifling, and downwarping until it began a continuing period of downwarping during the Phanero- zoic because of movement of the craton off of its parent plume. This thesis, as others (e.g., HurIey and Rand, 1969), supports the concept that widespread ~agrnenta- tion and dispersal of continental crust, so obvious in Cenozoic plate-tectonic analyses, could not have occurred ire the Precambrian. However, these concepts do not pre- cIude large-scale horizontal movements of an entire shield nor marginal accretions throughout this interval of time. Archean rocks in the United States (Figure 15.1) are limited to two major regions: Wyoming and vicinity and the southem extension of the Canadian Shield into the northern tier of states west of Lake Michigan. Structural trends in the greenstone and gneiss belts of Archean rocks of Wyoming are north to northeast for the earlier Archean deformation and north to northwest for the later (Houston, 1971~. Much of the Precambrian in Wyoming has yet to be mapped, and the possible effects of I~nide rotation have not been assessed. From a structural geologist's point of view, it is intriguing that Wyoming, the only region with Archean basement, is the home of the Wyoming upthrust province, a unique struc- tural style and trends of Laramide foreland deformation. Northern Minnesota and the subcrop extension into the Dakotas constitute the southwestern extension of the Superior province, the largest block of unreworked Archean crust in North America. The granitic Weiss un-

The Shape of North America during the Precambrian \~<,\ ~6 -` ~ '~) 2.5~b.y. ~ , .N t' t; If_ , 0 scow,. O SOO Kit ,,~?1 ~ —~ FIGURE 15.1 Area of North Amenca known to be underlain by rocks 2.5 b.y. old or older. Principal areas (provinces) are named Slave, Nain, Superior, and Wyoming. Short lines and dots are outcrop belts or wells where rocks of this intenral have been identified; most of these areas have been involved in erogenic events shown on Figure 1~.2. derlying the greenstone-graywacke belts across the southern third of Me province is about 3.0 b.y. old, and each greenstone belt evolved rapidly about 70 million years (m.y.) between the start of greenstone deposition and the end of granitic intrusions (Krogh and Davis, 1971; Krogh et al., 1974, 1975~. Nunes et al. (1978) and Nunes and Thurston (1978), however, have shown that the vol- canic rocks of the Abitibi greenstone belt were erupted in about 25 m.y., whereas the Uchi Lake greenstone belt spanned 220 m.y. The cooling and stabilization of the Superior province occurred near 9.5 b.y. as shown by the extensive K-Ar dating program of the Geological Survey of Canada (Stockwell, 1961~. In contrast, the Archean of southern Minnesota, Wisconsin, and northern Michigan has been involved in younger erogenic events and was remetamorphosed in those events. The Precambrian of the Minnesota Valley contains rocks older than any yet recognized from the Superior province (3.8 b.y.; Goldich and Hedge, 1974~. The Great Lakes region has been an area of intense study for many decades, initially to analyze the mineral 177 deposits but gradually developing into a sturdy of the broader geological environments and the rise of isotopic determinations in unraveling the geological history of the region. Extensive cover of glacial debris and Phanerozoic rocks around the southern and western sides has made conventional geological mapping difficult. The availabil- ity of regional magnetic and gravity coverage speeded up these studies significantly and has made it possible to extrapolate successfully from the limited outcrop ciata (see Chapter 11~. These studies (see Sims, 1976; Morey and Sims, 1976, for a summary) have demonstrated that the Archean is divisible into two major terranes. Orate is an older ensial~c gneiss terrane that extends across southern and central Minnesota, northern Wisconsin, and the Upper Peninsula of Michigan with ages as great as 3.8 b.y. These rocks mere then sutured Clout 2.7-2.6 lay. ago to an ensimatic greenstone belt that may have been formed in island-arc or continental borderland environments with oldest ages about 3.0 b.y. (Superior province of the Canadian Shield) but ages ranging from about 2.~2.7 b.y. for the green- stone units and their intensive granitic plutons. EARLY PROTEROZOIC Sims (1976) shows that the Lake Superior Proterozoic basin overlies an Archean suture and that the younger greenstone terrane is the stable block during this interval and the major source of sedimentary materials. This basin is compared with the intracratonic Labrador Trough (Dimroth et al., 1970) and shown to be analogous in nearly all depositional and tectonic events including timing, in- sofar as they have been detennined. Thus the ancient gneiss terrane that rims the Archean greenstone belt on the south and east acted as the mobile basement until it was finally stabilized about 1.7 b.y. ago to form a larger cratonic nucleus. Some of this belt was again reactivated and now forms part of the Grenville province. Figure 15.2 shows this boundary and its continuation across the midcontinent as derived from Muehlberger et al. (1967), Bayley and 1\Iuehlberger (1968), and Lidiak (1971) and a southwestern continuation (dotted) as pro- posed by Warner (1978~. Warner (1978) proposed that it is part of the Colorado lineament, a middle Precambrian (Penokean orogeny: 2.~1.7 b.y.) wrench fault (leR- lateral?) system that resembles the San Andreas type. De- tailed geological mapping of the Mullen Creek-Nash Fork shear zone in southeastern Wyoming (Houston et al., 1968) shows that it must have been a continental margin, thus a possible Proterozoic plate boundary (Hills et al., 1975~. Peterman and Hildreth (1977) demonstrated a Pro- terozoic overprint using K-Ar ages on the Archean rocks over a band up to 150 km ~ ide north of the shear zone in Wyoming suggesting a collision boundary. The Colorado lineament is a major discontinuity and deserves consider-

WILL IAM B. M U E HL ~ E ROE R \~ ~ t o 300 u . - ·, ~ O BOO Kid `` ~1 `~_J-~\ {~ "_; 1.7-2.5 BY FIGURE 15.2 Area of North America underlain by rocks 1.7 b.y. Old or older. Archean provinces of Figure 15.1 are stable cratons of this episode. Dashed lines show location and Mend of fold belts active during the interval 1.7-2.5 b.y. ago. Southern and western dashed-line boundaries are limit of known rocks of this time interval. Western boundary in United States based on R~Sr isotopic data (Kisser and Peterrnan, 1978). Line of triangles in Wyoming craton shows northern limit of middle Proterozoic K-Ar age overpnnt. Colorado Lineament fonns prominent dis- continuity from Lake Superior southwest to western Arizona. able effort to determine its Rue nature and regional . .^ slgul~lcance. Figure 15.2 also shows the Proterozoic fold belts that transect the Archean of the Canadian Shield and that are now part of the Churchill province that was consolidated with the Archean elements about 1.7 b.y. ago. In contrast to the Archean, the Proterozoic initiated the era that still continues, of rigid continental platforms, crator~s that stood above sea level and thus produced the first recog- nizable continental margin, rifted margins and associated aulacogens, and continental-margin geosynclines. The Coronation geosyncline and Athapuscow aulacogen (HofEnan, 1973) that flank the Slave province of the northwestern Canadian Shield are among the earliest rec- ognized in the Proterozoic. Intracratonic basins (Dimroth et al., 1970) or small ocean basins (Kearey, 1976), such as the Labrador geosyn- cline, are dominant interpretations of the Proterozoic belts that are involved with reactivated Archean crust of the Churchill province. Wynne-Edwards (1976) has proposed a model to ex- plain the ensialic basins seemingly typical of Proterozoic orogenesis. It requires the thinning of hot ductile siaIic crust over mantle upwarps. These then become sites of sedimentary basins that close by migration ofthe upwarp. Broad zones of shear, rather than the present transform faults, divide the spreading areas into segments and are spatially related to base-metal deposits. The lines of plu- tonic rocks such as the anorthosite massifs in the Gren- ville province provide tracks for relative movement vec- tors. Martin and Porada (1977), using the same African examples as does Wynne-Edwards, proposed instead that gravitational instabilities caused the development and subsequent closing of a multiple aulacogen to explain Proterozoic ensialic orogens. Kroner (1977) expanded upon their concept and showed how linear zones of weak- ness may develop ensialic geosynclines from graben systems via aulacogens. He further proposed that the hori- zontal mobility of large crystal plates increased through geological time to develop the wide opening and closing of ocean basins characteristic of the Phanerozoic. Gibb (1975; 1977) recognizes a set of major faults that extends westward from the northern termination of the Kapuskasing Gneiss Belt in the southernmost James Bay resulting from the suturing of the Superior and Churchill plates. He depicts a northward-moving Superior plate suturing progressively westward from James Bay and sub- ducting the intervening oceanic lithosphere. He inter- prets this collision, following the Tibetan Plateau analog (Dewey and Burke, 1973), to have reactivated the Churchill basement. He points to the Wollaston Lake fold belt, 4SO km to the northwest, whose shape mimics the curvature of the suture boundary, as the possible limit of penetrative deformation and crustal reactivation. How- ever, this does not explain why all the metasedimentary belts of the Churchill province mimic the suture bound- ary nor why the entire province has a uniform K-Ar age. His concept should be expanded to include the entire Churchill province to be the result of the Superior prov- ince collision. Seyfert (1978) interprets paleomagnetic apparent polar wander curves to show that a Norm America~ondwanaland collision at 1.85 b.y. caused the Penokean orogeny. He further suggests that Europe joined Norm America at this time; this might resolve the apparent paradox of having a collision margin completely circumscribing the Superior province and Sunfishes an additional collision direction to assist in remobilizing the Churchill province. Cavanaugh and Seyfert (1977) sug- gest that the Slave province collided with the Superior province about 1.75 b.y. ago. Their proposed suture ex- tends from between the Foxe and Committee fold belts of Baton Island southwest across the shield. Another pro- posed Proterozoic suture, based on the recognition of a zone of high electrical conductivity in the buried shield (Camfield and Cough, 1976), extends southward from the Wollaston fold belt to southeastern Wyoming.

The Shape of North America during the Precambrian The curvature of the belt that surrounds the Supenor province suggests a flexibility (ductility9) of lithosphere that is greater than present-day plate tectonics will allow. However, the tight curve around the Ungava Peninsula of northern Quebec has the same radius of curvature as the present-day Banda Arc of the East Indies, the eastern Caribbean arc, or Scotia Arc between South America and Antarctica. An important difference for these modern examples, however, is that each has oceanic crust inside the arc rather than circumscribing a continental craton. Important questions yet unresolved include: how broad was the ocean or seaways in which these fold belts were deposited? Paleomagnetic control on rocks of this age (2.~1.7 b.y.) permits oceans limited to widths no greater than 500 1000 km (about the width of Baffin Bay) that must close so that their original relative positions are maintained (McElhinney and McWilliams, 1977~. What was the sequence of formation? Are they successive events or plasterings onto the Superior margin, or were they ductile extension zones in a larger Archean continent that subsided, filled with sediments, and were later com- pressed? How can we best account for the circum- Superior collision without having it reactivated as welI? How many of these proposed sutures have supporting geological evidence? LATE PROTEROZOIC The arcuate trends, extrapolated from the Canadian Shield along linear gravity anomalies and the Black Hills outcrop belt under the northern plains, appear to be trun- cated along the Colorado lineament. A southeast trend continues faintly across Nebraska and into northwestern Missouri, but geological confirmation is lacking. If it con- tinues, then it is evidence for the Penokean orogeny being earlier than the Churchill deformations or it marks the western margin of the Archean crustal block of the Great Lakes region. The southern limit of North America at about 1.7 b.y. appears to be as shown in Figure 15.3, although the steps in its evolution are only now being deciphered. Isotopic data in Colorado and southern Wyoming indicate that no Archean crust exists beneath the exposed Precambrian rocks (Hills and Armstrong, 1974~; thus crust evolved since the Archean. How many lithological/structural belts are present in this band or their sequence of development is unknown for most of this region. Proterozoic greenstone belts are recognized at many places in Arizona, Colorado, and New Mexico (Anderson and Silver, 1976; Barker, 1969; Robertson et al., 1978~. Tectonic foundering of a continental margin in northern New Mexico and southwestern Colorado has been proposed by Barker et al. (1976) to explain the southeast- trending belt of trondhJem~tes and the nearly contem- poraneous rhyolite~uartzite temne that lies to the southwest. Sedimentary structures (Barrett and Kirschner, 1979; Montgomery, 1953) in the quartzites and inter- 179 Nk ~ >,\ Aid// ~ ~ ~ If 5~''~ ~ of' ~ `n LIT- : Ad, .,,, ~ ;..,, an/ ~ If--'- ~ 1 ~; '- I.3-1.7 b.y. 0 300M! ~ ~ ~ O 500 KM IS _: FIGURE 15.3 Area of North Amenca underlain by rocks 1.3 b.y. Old or older. Craton now includes all of the Canadian Shield, except for its eastern border. The southeastemmost line marks probable southern limit of rocks 1.3 b.y. old or older. The central line is the southern limit of rocks 1.7 b.y. old or older. Between these lines is the anorogenic pluton belt of Silveret al. (1977) and Emslie (1977); plutons shown as dots at location of dated sample or dated pluton. The northern line marks the northern limit of intrusive or metamorphic activity giving ages in the 1.3- to 1.7-b.y. range. Western boundary is same as Figure 15.2. Cross- hatched areas are rhyolite/granite terraces; dotted areas are Sioux quartzite. calated schists, which lie southwest of the rhyolite terrane, show them to be continental-margin deposits. Field and isotopic studies thus far have not definitely established the time sequence, although considerable progress is being made (Denison et al., in press). From a resource point of view, it is important to know the se- quence and environment of deposition of each belt so that intelligent mineral exploration programs can be planned. Figure 15.3 shows that rocks in the interval between 1.7 b.y. and about 1.35 b.y. ago dominate the midcontinent region of the United States. This division is a relic from our 1967 paper (Muehlberger et al., 1967), but because of more recent studies some of the distinctions made in that paper have blurred; others have sharpened. We used it to mark the end of the Elsonian orogeny (Stockwell, 19611; it now marks the end of the anorogenic pluton episode

180 (Emslie, 1977; Bickford and Van Schmus, 1978; Silver et al., 1977~. The northern belt of Figure 15.3 consists of a 1.7-b.y. basement intruded by anorogenic plutons about 1.~1.4 b.y. The belt south of the heavy dashed line con- tains no rocks known to be older than 1.4 b.y. Much of Me southern belt is overlain by younger Precambrian rhyolite and associated granite sills or Pre- cambrian sedimentary units, which mask the older crustal geology (Muehlberger et al., 1967~; thus evidence for an older basement, if any, has yet to be determined. Rhyolite-granite terranes become progressively younger from northeast to southwest (Bickford and Van Schmus, 1978; Denison et al., 1977~. This younging of granite- rhyolite terranes continues on the next age interval map (Figure 15.4) across the Texas Panhandle and into the El Paso area. A prominent exception to the southwestward younging is the early Cambrian magmatism (Ham et al., 1964) associated with the development of the Southern Oklahoma Paleozoic aulacogen. Rifting events have been proposed for the origin of the 1.4-b.y. to 1.5-b.y. anorogenic pluton belt, the 1.2-b.y. Mackenzie dike swarm (shown schematically on Figure 15.4), and the Coppermine basalts on the northwest Cana- dian Shield. Apparently slightly younger but possibly related to the Mackenzie dikes is the extensive intra- continental rift system of the Keweenawan. Sims (1976) has shown that the Lake Supenor segment of Keweena- wan rifting is bounded on the soup by pre-existing conti- nental faults that had substantial right-lateral movement before Keweenawan time. Sawkins (1976) has shown that this ricing event was a major period of copper mineraliza- tion. This rifting was soon terminated by the Grenville collision as indicated by the near coincidence of isotopic ages, and the truncation of Keweenawan marks the first opening?. The location of the suture is obscured by late Paleozoic activity ofthe Appalachian belt. King and Zietz (1978) describe the New York-Alabama lineament that lies beneath the Appalachian basin and propose that it is a major strike-slip fault (possibly a suture) analo- gous to the currently active Altyn Tagh fault of Tibet, resulting from the India-Asia collision (Molnar and Tapponnier, 1975~. Across Texas we have evidence for suturing at Gren- ville time. The west-trending overd~rust belt of Me Van Hom region (King and Flawn, 1953) and the serpentinites of the southern Llano uplift in central Texas (Barnes, 1946; Garrison, 1978) both show movement of material northward (using present geographic coordinates) onto the Norm American continent. The suture has been proposed as the closing of an aulacogen or small ocean (Garrison and Rarnirez-Ramirez, 1978; Sengor and Budder, 1977~. This proposed suture is part of the zone that several investigators include in the Texas lineament, a major zone of recurrent tectonic activity that extends across the sou~- western United States. The origin of the Texas lineament is shrouded in mystery (oldest demonstrated offsets are WILLIAM R. MUEHLBERGER 'a >fir. ~ ~ ~ .3 b.y. O. . .300 tell ~ `& O S O O K M —IFFY ~ <~—7 l .~ iv-~" f,,i, ~ FIGURE 15.4 Area of North Amenca at about 1.0 b.y. old; showing major tectonic and igneous events between 1.3 b.y. and 1.0 b.y. of age. The western margin (essentially that shown in Figures 15.2 and 15.3) is now seen to be a riRed margin (former western limit is not yet identified) win two major entrants (aulacogens). Fragments of Precambrian continental crust have been recognized in the northwestern part of the Cordilleran belt. Southwestern margin is distorted by late Phanerozoic transfonn tectonics. Dotted areas are sedimentary rock of Precambnan age; cross-hatched areas, rhyolite-granite tenants; b, basalt; dashed lines inside a linear boundary, Keweenawan basalt- gabbro; dashed lines trending northwest across Canada, Mackenzie dike swarm; dashed zone along eastern Norm Amer- ica, Texas, and southern Mexico are results of the Grenville collision. Black dot in Colorado is Pike's Peak granite (~1.1 b.y. old). Peripheral to this craton are the successive latest Precam- bnan and Phanerozoic sedimentary wedges. about 1.4 b.y.; Swan, 1975), but that it has been the site of recurrent tectonic activity is a fact (Wiley and Muchl- berger, 1971~. Thus it and the Colorado lineament are similar in style, but the southern zone is younger. Could these be the sites of oblique collision, zones that produce such a pervasive grain and fracture system that they never properly weld to pre-existing shield and thus are continu- ing zones of activity? The Grenville-Appalachian belt appears to be a younger and more extensive analog. The paucity of information on the Precambrian of Mexico

The Shape of North America during the Precambrian because of extensive cover of Mesozoic and younger rocks leaves all reassemblies speculative; however, the "Gren- ville front" shown across west Texas is based on geo- physical, petrographic, and isotopic data as described In Muehlberger et al. (1967~. The rifted western and northwestern margins of North America are reasonably well located because of the pres- ence of the miogeoclinal wedge of Belt rocks and their equivalents. Where the margin was in earlier times is speculative. The recognition of Precambrian blocks in the western Cordillera of Canada and Alaska and extensive strike-slip displacements make reconstruction a major problem. Stewart (1976) has proposed a late Precambrian age (about 850 m.y.) for the rifting that made this region. The Belt and Uinta embayments (aulacogens.9 Burke and Dewey, 1973) were filling early In Belt time, and thus a margin was in existence by at least 1.35 b.y. ago. Sears and Price (;978) present evidence that the western and southwestern boundaries of North America (Cordilleran Belt margin and Texas lineament) are pre-Belt riPc mar- gins and that the matching segments can be found in the Siberian cratons. Seyfert (1978) from paleomagnetic data suggests that Asia joined the west coast of North America between 1.4 b.y. and 1.15 b.y. ago. The absence of a west- em source of Belt and Windermere strata (Maxwell, 1974) suggests that Asia may only have been close to North America. Isotopic identification of continental crust in western United States places a western limit on it that essentially coincides with the latest Precambrian sedi- mentary wedges (Kistler and Peterman, 1978; Stewart, 1976~. On the other hand, Badham (1978a) proposes that the western margin of North America has been both an active and passive margin but was never destroyed by rifting (the Athapuscow aulacogen is only a persistent fault zone; Badham, 1978b) or collision and thus has pro- gressively grown westward. Late Precambrian rifts, aulacogens, and other exten- sional phenomena (alkalic magrnatism, anorogenic plutons) have been proposed for many regions. Some of these same belts are zones of seismicity today, for exam- ple, St. Lawrence rift system (Kumarapeli, 1978), Reelfoot rift (Ervin and McGinnis, 1975), and sounded Oklahoma aulacogen (Denison, 1978; Wickham, 1978~. Each ofthese structures has its origins in the Precambnan, although most of their documented activity is in the Phanerozoic. An understanding of how old tectonic lines can be re- activated will furnish significant data on paleostress orientations (Sykes, 1978~. SUMMARY This brief description of the evolution of North America has attempted to show that we have very little detailed knowledge of how, where, or when successive belts of rocks were added or were rifted off or were reactivated (continental collisions?) or, for that matter, how well or 181 even whether plate tectonics works in the Precambrian. Every aspect of the evolution of continents is in a state of active debate, research, and ferment (see Chapter 2~. Broad trends, however, are visible: Archean (2.o b.y.~; early Proterozoic (2.~1.7 b.y.) with stable cratons and intracratonic mobile belts; middle Proterozoic (1.7-1.2 b.y.) with anorogenic plutons (incipient riRing) and wide- spread rhyolitic volcanism across the central midconti- nent region; late Proterozoic to Cambrian (1.~0.6 b.y.) with ricing and collision events of Phanerozoic style; and the Phanerozoic- the recognized realm of plate tectonics. Each of these intervals marks a period of fairly uniform structural style, evolving from more ductile to more rigid deformation styles that reflect an increasing rigidity and areal growth of the lithosphere and, presumably, a concur- rent thickening (Jordan, 1975~. RE FE RE JO E S Anderson, C. A., and L. T. Silver (1976). Yavapai Series a greenstone belt, Anz. Geol. Soc. Digest X, 1~26. Anhaeusser, C. R. (1975). Precambrian tectonic environments, Ann. Ret;. Earth Planet. Sc'. 3, 31~3. Anhaeusser, C. R., R. Mason, M. J. Viljoen, and R. P. Viljoen (1969). A reappraisal of some aspects of Precambrian shield geology, Geol. Soc. Am. Bull. 80, 217~00. Amdt, N. T., A. J. Naldrett, and D. R. Pyke (1977). Komatiitic and iron-nch tholeiitic lacunas of Munro township, northeast Ontano, J. Petrol. 18, 319 369. Badham, J. P. N. (1978a). Has there been an oceanic margin to western North America since Archean time?, Geology 6, 621~25. Badham, J. P. N. (1978b). The early history and tectonic signifi- cance of the East Ann graben, Great Slave Lake, Canada, Tec- tonophysics 45, 201-215. Baer, A. J., ed. (1970). Symposium on basins and geosynclines of the Canadian Shield, Geol. Sur1;. Can. Paper 7040, 265 pp. Barker, F. (1969). Precambrian geology of the Needle Mountains, southwestern Colorado, U.S. Geol. Sure. Prof. Paper 644-A, 35 PP. Barker, F., J. G. Arth, Z. E. Peterman, and I. Friedman (1976). The 1.7 to 1.8 b.y.mld trondhjemites of Southwestem Colorado and northern New Mexico; geochemistry and depths of gene- sis, Geol. Soc. Am. Bull. 87, 18~198. Bames, V. E. (1946). Soapstone and serpentine in the Central Mineral Region of Texas, in Texas Mineral Resources, Bureau of Economic Geology, U. of Texas Publ. 4301, pp. 5~91. Barrett, M., and C. Kirschner (1979). Depositional environments of He Precambrian Ortega and Rinconada formations, Picuris Mountains, northcentral New Mexico, in New Mexico Geo- logical Society Guidebook, Santa Fe Country, pp. 121-126. Bayley, R. W., and W. R. Muehlberger (1968). Basement rock map of the United States, exclusive of Alaska and Hawaii, scale 1:2,500,000, U.S. Geological Survey. Bickford, M. E., and W. R. Van Schmus (1978). Chronology of middle and late Precambrian crust-forming events in the mid- continent region of North Amenca, Geol. Soc. Am. Abstr. Pro- grams 10, 2. Burke, K., and J. F. Dewey (1973). Plume-generated triple junc-

82 tions: key indicators in applying plate tectonics to old rocks, J. Geol. 81, 40~33. Burke, K., J. F. Dewey, and W. S. F. Kidd (1977). World distnbu- tion of sutures the sites of former oceans, Tectono physics 40, 6~99. Camfield, P. A., and D. I. Cough (1976). A possible Proterozoic plate boundary in North America, Eos Trans. Am. Geophys. Union 57, 1003. Cavanaugh, M. D., and C. K. Seyfert (1977). Apparent polar wander paths and the joining of the Superior and Slave prov- inces during early Proterozoic time, Geology 5, 207-211. Denison, R. E. (1978). Basement rocks in the Arbuckle Moun- tains, in Structural Style of the Arbuckle Region, J. Wickham and R. E. Denison, field trip leaders, Geological Society of Amenca, South-Central Sect., Field Tnp 3, pp. 42~6. Denison, R. E., E. G. Lidiak, M. E. Bickford, and E. B. Kisvar- sanyi (1977). Precambrian geology and geochronology of the central interior province, North Amenca, Geol. Soc. Am. Abstr. Programs 9, 947-948. Denison, R. E., E. G. Lidiak, HI. E. Bickford, and E. B. Kisvar- sanyi (in press). Geology and geochronology of Precambrian rocks in the central interior region of the United States, Econ. Geol. Dewey, J. F. (1977). Suture zone complexities: a review, Tec- tonophysics 40, 5~67. Dewey, J. F., and K. Burke (1973). Tibetan, Variscan and Precam- brian basement reactivation: products of continental collision, J. Geol. 81, 68~92. Dimroth, E., W. R. A. Barager, R. Bergeron, and G. D. Jackson (1970). The filling ofthe circurn-Ungavageosyncline, in Basins and Geosynclines of the Canadian Shield, A. J. Baer, ea., Geol. Sun>. Can. Paper 7040, 4~142. Douglas, R. I. W., and B. A. Price (1972)..Nature and significance of variations in tectonic styles in Canada, in Vanations in Tem tonic Styles in Canada, R. A. Price and R. J. W. Douglas, eds., Geol. Assoc. Can. Spec. Paper 11, 62~83. Drummond, K. I. (1974). Paleozoic Arctic margin of North Amer- ica, in The Geology of Continental Margins, C. A. Bulk and C. L. Duke, eds., Springer-Verlag, Berlin, pp. 797~10. Drury, S. A. (1977). Structures induced by granite diapirs in the Archean greenstone belt at Yellowknife, Canada: implications for Archean geotectonics, J. Geol. 85, 34~358. Emslie, R. F. (1977). Ano~osite massifs, rapakivi granites, and late Proterozoic rifting of North Amenca, Precambrian Res. 7, 61-98. Engel, A. E. J., and C. G. Engel (1964). Continental accretion and He evolution of North Amenca, Advan. Frontiers Geol. Geophys., 17~7e. Engel, A. E. J., and D. L. Kelm (1972). Pre-Pennian global tec- tonics: a tectonic test, Geol. Soc. Am. Bull. 83, 232~2340. Ervin, C. P., and L. D. McGinnis (1975). Reelfoot riR: reactivated precursor to the Mississippi Embayment, Geol. Soc. Am. Bull. 86, 1287-1295. Fyson, W. K. (1978). Structures induced by granite diapirs in the Archean greenstone belt at Yellowknife, Canada: implications for Archean geotectonics: a discussion,! Geol. 86, 767-769. Garnson, J. R., Jr. (1978). A petrogenetic model for the south- eastem Llano uplift, Texas, Eos Trans. Am. Geophys. Union 59, 1211. Garrison, J. R., Jr., and C. Ramirez-Ramirez (1978). The Llano uplift, central Texas: evidence for a Precambrian triple spread- ing system, Geol. Soc. Am. Abstr. Programs 10, 106. Gibb, R. A. (1975). Collision tectonics in the Canadian Shield?, Earth Planet. Sci. Lett. 27, 378 382. W I ~ L I AM R. MU E H L B E RG E R Gibb, R. A. (1977). A gravity survey of James Bay and its bearing on the Kapuskasing Gneiss Belt, Ontano, Tectonophysics 45, T7-T13. Glikson, A. Y. (1976). Earliest Precambrian ultramafic-mafic vol- canic rocks: ancient oceanic crust or relic terrestrial maria?, Geology 4, 201-205. Goldich, S. S., and C. E. Hedge (1974). 3,800 m.yr. granitic gneiss in southwestern Minnesota, Nature 252, 467068. Goodwin, A. M. (1974a). The most ancient continental margins, in The Geology of Continental Margins, C. A. Burk and C. L. Drake, eds., Springer-Verlag, Berlin, pp. 767-780. Goodwin, A. M. (1974b). Precambrian belts, plumes, and shield development, Am. J. Sci. 274, 987-1028. Green, D. H. (1975). Genesis of Archean peridotitic magmas and constraints on Archean geothermal gradients and tectonics, Geology 3, 1~18. Harn, W. E., R. E. Denison, and C. A. Merritt (1964). Basement rocks and structural evolution of southern Oklahoma, Okla. Geol. Surv. Bull. 95, 302 pp. Hargraves, R. B. (1976). Precambrian geologic history, Science 193, 363~371. Hills, F. A., and R. L. Armstrong (1974). Geochronology of Pre- cambrian rocks in the Iasamie Range and implications for the tectonic framework of Precambrian soothing Wyoming, Pre- cambrian Res. 1, 21~5. Hills, F. A., R. S. Houston, and G. V. Subbarayudu (1975). Pos- sible Proterozoic plate boundary in southern Wyoming, Geol. Soc. Am. Abstr. Programs 7, 614. Hoffinan, P. (1973~. Evolution of an early Proterozoic continental man, Phil. Trans. Roy. Soc. London A273, 547~81. Houston, R. S. (1971). Regional Tectonics of the Precambrian Rocks of the Wyoming Province and Its Relationship to Lara- mide Structure, Wyoming Geological Association Guidebook Twenty-third Annual Field Conference, A. R. Renfro, ea., pp. 1~27. Houston, R. S., et al. (1968). A Regional Study of Rocks of Pre- carnbrian Age in That Part of the Medicine Bow M,` Lying in Southeastern Wyoming- With a Chapter on tionship Between Precambrian and Laramide St.. Mem. 1, Wyoming Geological Survey, 167 pp. Hurley, P. M., and J. R. Rand (1969). Pre-dri:Gc continental nuclei, Science 164, 1229~1242. Jackson, D. G., and F. C. Taylor (1972). Correlation of major Aphebian rock units in the northeastern Canadian Shield, Can. J. Earth Sci. 9, 1650 1669. Jordan, T. H. (1975). The continental tectosphere, Rel;. Geophys. S pace Phys. 13, 1-12. Kearey, P. (1976). A regional structural model of the Labrador Trough, northern Quebec, from gravity studies, and its rele- vance to continent collision in the Precambrian, Earth Planet. Sc:. Lett. 28, 371~78. King, E. R., and I. Zietz (1978). The New York-Alabama linea- ment: geophysical evidence for a major crustal break in the basement beneath the Appalachian basin, Geology 6, 312 318. King, P. B., and H. PI. Beikman (1974). Geologic map of the United States, scale 1:2,500,000, U.S. Geological Survey. King, P. B., and P. T. Flawn (1953). Geology and Mineral De- posits of Precambrian Rocks of the Van Horn Area, Texas, Publ. 5301, Bureau of Economic Geology, U. of Texas, 2.18 pp. Kistler, R. W., and Z. E. Petei~,~ar~ (1978). Reconstruction of crustal blocks of California on the basis of initial strontium isotopic compositions of Mesozoic granitic rocks, U.S. Wool. Sure. Prof. Paper 1071, 17 pp. Krogh, T. E., and G. L. Davis (1971). Zircon 13-Pb ages of Ar-

lithe Shape of North America during the Precambrian chean metavolcanic rocks in the Canadian Shield, Carnegie Inst. Washington Yearb. 70, 241-242. Krogh. T. E;., I. F. Ermanov~cs, and G. L. Davis (1974). Two episodes of metamorphism and deformation in the Archean rocks of the Canadian Shield, Carnegie Inst. Washington Yea*. 73, 573 57.5. Krogh, T. E., G. L Davis, N. B. W. Harris, and I. F. Errnanovics (1975). Isotopic ages in the eastern Lac Seul region of the English River gneiss belt, Carnegie last. Washington Yearb. 74, 62~625. Kroner, A. (1977). Precambrian mobile belts of southern and east- em Africa ancient sutures or sites of ensialic mobility? A case for crustal evolution towards plate tectonics, Tectonophysics 40, 101-135. Kumarapeli, P. S. (1978). The St. Lawrence paleo-rift system: comparative study, in Tectonics and Geophysics of Conti- nental Rifts, I. B. Ramberg and E. B. Neumann, eds., D. Reidel, Dordrecht, Holland, pp. 367~84. Lidiak, E. G. (1971). Buried Precambrian rocks of South Dalcota, Geol. Soc. Am. Bull. 82, 1411-1420. Lidiak, E. G., and I. Zietz ( 1976). Interpretation of aeromagnetic anomalies between latitudes 31°N and 3~3°N in the eastern and central United States, Geol. Soc. Am. Spec. Paper 167, 37 pp. Martin, H., and H. Porada ( 1977). The intracratonic branch of the Damara orogen in South West Africa, Precambriah Res. S. 31 1~38. Maxwell, J. C. (1974). Early western margin ofthe United States' in The Geology of Continental Margins, C. A. Burk and C. L. Drake, eds., Springer-Verlag, Berlin, pp. 831-851. McElhinney, M. W., and M. O. McWilliams (1977). Precambrian geodynamics—a paleomagnetic view, Tectonophysics 40, 137-159. Molnar, P., and P. Tapponnier (1975). Cenozoic tectonics of Asia: effects of a continental collision, Science 189, 419026. Montgomery, A. (1953). Precambrian geology of the Picuris Range, northcentral New Mexico, N. Mex. Burl Mines Afiner. Resour. Bull. 30, 89 pp. Moorbath, S., R. K. Options, R. J. Pankhurst, N. H. Gale, and V. R. McGregor (1972). Further rubidium-strontium age deter- minations on the very early Precambrian rocks of the Godthaab district, West Greenland, Nature Phys. Sci. 240, 78 82. Priorly, G. B., and P. K. Sims (1976). Boundary between two Precambrian W terranes in Minnesota and its geologic signif~- cance, Geol. Soc. Am. Bull. 87, 141-152. Muehlberger, W. R., R. E. Denison, and E. G. Lidiak (1967). Basement rocks in continental interior of IJnited States, Am. Assoc. Petrol. Geol. Bull. 51, 2351-2380. Myers, J. S. (1976). Granitoid sheets, thrusting, and Archean crustal thickening in West Greenland, Geology 5, 26~268. Nunes, P. D., and P. C. Thurston (1978). Evolution of a single greenstone belt over 220 million years a zircon study of the Uchi Lace area, northwestern Ontario, in Short Papers of the Fourth International Conference, Geochronology, Cosmochro- nology, Isotope Geology, R. E. Zartrnan, ea., U.S. Geol. Sur';. Open-file Report 78-701, pp. 313 315. Nunes, P. D., D. R. Pyke, and L. S. Jensen (1978). Toward an absolute age stratigraphy of the Abitibi greenstone belt— zircon ages from the Timmins and Kirkland Lake areas, Geol. Soc. Am. Abstr. Programs 10, 464 465. Petel~an, Z. E., and R A. Hildreth (1977). Reconnaissance ge- ology and geochronology of the Precambrian of the Granite ~3 Mountains, Wyoming, U.S. Geol. Surc. Open-File Report 77-140. Price, R. A., and R. J. W. Douglas, eds. (1972). Variations in tectonic styles in Canada, Geol. Assoc. Can. Spec. Paper 11, 688 pp. Robertson, J. M., K. C. Condie, and D. C. Mathewson ( 1978). The Pecos greenstone belt: a Proterozoic volcano-sedimentary sequence in the southern Sangre de Cristo Mountains, New Mexico, Geol. Soc. Am. Abstr. Programs 10, 479. Sawkins, F. J. (1976). Widespread continental rifting: some con- siderations of timing and mechanism, Geology 4, 427~30. Sears. 3. W., and R. A. Price (1978). The Siberian connection: a case for Precambrian separation of the North American and Siberian cratons, Geology 6, 267-270. Sengor, A. M. C., and J. C. Butler ( 1977). The Llano uplift, central Texas: a Proterozoic example of continental collision?, Geol. Soc. Am. Abstr. Programs 9, 72-73. Seyfert, C. K. (1978). Palaeomagnetic evidence for three Wilson cycles during the Precambrian, Geol. Soc. Am. Abstr. Pro- grams 10, 4DO. Silver, L. T., M. E. Bickford, W. R. Van Schmus, J. L. Anderson, T. H. Anderson, and L. G. Medaris (1977). The 1.~1.5 b.y. transcontinental anorogenic platonic perforation of North America, Geol. Soc. Am. Abstr. Programs 9, 117~1177. Sims, P. K. (1976). Precambrian tectonics and mineral deposits, Lalce Superior region, Econ. Ceol. 71, 1092~1127. Stewart, J. H. (1976). Late Precambrian evolution of North Amer- ica: plate tectonics implications, Geology 4, 11-15. Stockwell, C. H. ( 1961). Structural provinces, progenies, and time classification of rocks of the Canadian Precambrian Shield, Can. Geol. Sure. Paper 61-17, 10~118. Sutton, J. (1973). Some changes in continental structure since early Precambrian time, in Implications of Continental Drift to the Earth Sciences, D. H. Tarling and S. K. Runcom, eds., Academic, New York, pp. 1071-1081. Swan, M. M. (1975). The Texas lineament-tectonic expression of a Precambrian orogeny, Geol. Soc. Am. Abstr. Programs 7, 128~1289. Sykes, L. R. (1978). Intraplate seismicity, reactivation of preexist- ing zones of weakness, alkaline magrnatism, and other tec- tonism.postdating continental fragmentation, Ret;. Geophys. Space Phys. 16, 621~88. Tectonic Map of Canada (1969). Geological Survey Canada Map 1251A. Upadhyay, H. D. (1978). Phanerozoic, peridotitic, and pyroxeni- tic komatiites from Newfoundland, Science 202, 1192-1195. Wamer, L. A. (1978). The Colorado lineament: a middle Precam^ Brian wrench fault system, Geol Soc. Am. Bull. 89, 161-171. Wickharn, J. (1978). The southern Oklahoma aulacogen, in Struc- tural Style of the Arbuckle Region, J. Wickham and R. E. Denison, field trip leaders, Geolog cal Society of America, South-Central Sect., Field Trip 3, pp. 9~1. Wiley, M. A., and by. R. Muehlberger (1971). The Texas linea- ment, in The Geologic Framework of the Chihuahua Tectonic Belt, K. Seewald and D. Sundeen, eds., Publ. 71-S9, West Texas Geological Society, pp. 1~23. Windley, B. F. (1977). The E1;olt;ing Continents, John Wiley, New York, 385 pp. Wynne-F:dwards, H. R. (197S). Proterozoic ensialic orogenesis: the millipede model of ductile plate tectonics, Am. J. Sc:. 276, 927-953.

An Outline of the Tectonic Characteristics of China INTROD UCTION Earlier in this century, western scholars were introduced to the geology of China Mom the works of Richthofe, Loczy, Obruchev, Willis, Grabau, and others. Unfortu- nately, no comprehensive treatise on the geology of China was available until the Geology of China (Lee, 1939) am peered. Lee stressed the concept of "tectonic systems" that led to the development of his well-known theory of geomechanics. Huang (1945) assembled ant] analyzed all available geological data and attempted to elucidate systematically the tectonic characteristics of China in his book, On Major Tectonic Forms of China. Recently, greater attention has been paid by Chinese geologists to the theories of seafloor spreading and plate tectonics. As most of the Chinese works are published in Chinese, this paper intends to give a current summary of the tectonic characteristics of China. Additional background informa- This paper was originally published essentially in its entirety in Eclogae Geol. Helv. 71, 611-635, 1978. 184 T. K. HUANG Chinese Academy of Geological Sciences tion can be found in the papers of Huang (1945, 1960), Hung et al. (1974), Li (1975), and Terman (1973~. TECTONIC UNITS OF CHINA SINO- KOREAN PARAPLATFORM The Sino-Korean paraplatfonn (triangular in form) covers the entire temtory of North China, the northern part ofthe Yellow Seat arid the northern part of Korea (Figure 16.1~. This paraplatfonn is the oldest in China, its basement last affected by the Chungtiao* orogeny at about 1700 million years (m.y.) ago (Table 16.1~. The marginal parts of the *In the Luliang area, the Cambrian lies directly upon the lower Proterozoic, Me upper Proterozoic is lacking. The tenn "Luliang orogeny" is here replaced by the Chungtiao orogeny, with the type locality of Chungtiaoshan, where We upper Proterozoic is well developed and lies unconformably upon the lower Protero- zoic.

185 ,: - g , ,~ ~ . ~ Zoo . ~ ~~ ~~l . _ S ey Z. to IVY ° ._~/ ~ o moo o ~ ~- ~ o e cq 4_~ :r _ _ .' If_ C ~ I] i ill 8i, ~ ~ ~ -~: Id] ~ i

· - o c c; Geological Isotopic Chronology Age (m.y.) Quaternary 15 Tertiary 67 Cretaceous 137 'to o Jurassic ~ 190 ~ . 1 nasslc 230 Permian 280 Carboniferous . 350 Devonian 405 Silunan 440 Ordovician 550 Cambrian 570 Eocambrian . 700 Sinian System Chingpaikou 10()0 Chigsien 1400 Changchen 17 __ Huto 2000 Wutai 2500 _ Fuping 186 TABLE 16.1 Subdivision of Orogenic Cycles and Important Events of the Tectonic Development of China T. E. HUANG Subdivision of Orogenic Cycles and Important Events of the Tectonic Development of China . . Orogenic Cycles of Europe Himalayan Yenshanian Variscan ._ a N o - et c, ._ be: AS "0 ~ D _. ~ 2 'a c .= c us c, ._ ON O 0- e. c' . 0- et O Indosinian | Destruction and disintegration of parts ~ l of P-A; intensive activity of M-P and T-H | Consolidation of Central Asiatic-Mongolian geosynclines; cementation of Siberian plat- form with Sino-Korean and Tarim platfonns; connation of P-A Caledonian ~ Formation of South China platform Hsingkaiian Yangtzeian ? Disintegration of Chinese protoplatforrn; formation of Kunlun, Tsinling, Perishan, Tienshan and other geosynclines Fonnation of Yangtze and Tarim platfonns; combination of these platfonns with Sino- Korean paraplatform to fonn the Chinese proto-platfonn Chungtiaoian 1 Formation of Sino-Korean paraplatform | Wutaiian Fupingian - ct Pa o a 0 A. ~ 0 Alpine _. . ~ ~mmenan Variscan Caledonian Assyntian Dalslandian Svecofennian Karelian Belomorian

An Outline of the Tectonic Characteristics of China paraplatform, such as the Alashan region, were consoli- dated by the end of the Proterozoic through the Yangtze cycle. Geologists now agree that the main portion of the paraplatfortn developed through three stages, as revealed by three major unconformities (Ch'eng et al., 1973~. These are (1) the unconformity between the Archean Fuping Group and the lower Proterozoic Wutai Group, represent- irlg the Fuping orogeny about 235(~2550 m.y. ago*; (2) the unconformity between the Wutai Group and the upper part of the lower Proterozoic, or the Huto Group, representing the Wuta~ orogeny about 2000 m.y. ago; and (3) the unconformity between the Huto Group or its equivalents and the upper Proterozoic Sinian Suberathem,! representing the Chungtiao orogeny. The blanket of the platfonn consists of Sinian and Cambro- Ordivician neritic sediments (largely carbonates), Perrno- Carboniferous terrestrial sediments (with marine inter- calations), as well as purely terrestrial Meso-Cenozoic sediments. The Sinian System is present in a few regions, while Silurian, Devonian, and lower Carboniferous are absent. An abundance of Mesozoic continental volcanics and granitoid intrusions occurs in the eastern part of the paraplatfo~l~, especially in the Yenliao depression, Shangtung, and other regions. Cenozoic basalts are widespread. Tectonic disturbances of the Indosinian cY- cle were limited in Inner Mongolia, Liaotung, and the Yenliao depression. YANGTZE PARAPLATFORM The Yangtze paraplatform (Figure 16.~) comprises the greater part of the Yangtze Basin from eastern Yunnan to Kiangsu, also including the southern part of the Yellow Sea. Previously, the consolidation ofthe Yangtze paraplat- forrn was considered contemporaneous with that of the Sino-Korean paraplatfonn. Recent investigations using stromatolites, micropaleoflora, and isotopic geochronol- ogy of the basement rocks indicate that it was consoli- dated about 700 m.y. ago (Yangtze orogeny). The blanket of the paraplatforrn consists chiefly of carbonate and clas- tic deposits ranging in age Tom Sinian System to Triassic, with Devonian and Carboniferous normally absent in Szechwan and North Kweichow, whereas terrestrial Jurassic, Cretaceous, and younger deposits occur in Szechwan, Central Yunnan, Hupeh, Kiangsu, and over regions. Tectonism was strong chiefly in the Yenshanian cycle with the formation of well-developed blanket folds. The Kam-Yunnan Axis and the Lower Yangtze, however, suffered polycyclic deformation and magmatism; the *The oldest basement rocks are dated at 320(~3400 m.y. ago. IThe author uses the classification of the Sinian as given in the new Geological Map of China (1:4,()00,000). The tenn "Sinian System," with its standard section in the Yangtze Gorges, com- prises sediments ranging in age from 600 800 m.y. ago. The tend "Sinian Suberathem," with its standard section in ChihEsien near Peking, comprises sediments ranging in age from 600-1700 m.y. ago. The term sinian in this paper corresponds to Sinian Suberathem. ~7 former belongs essentially to the Variscan cycle, while the latter belongs to the Yenshanian and Indosinian cycle. Formerly, the Huaiyang .\Iassif was considered a portion of the Sino-Korean paraplatform; recent data show that it belongs to the Yang~e paraplatforrn. TARIM PLATFORM The Tarim platform (Figure 16.1) is defined by the Tienshan fold system in the north and by the Kunlun fold system in the south and is largely covered by Cenozoic deposits. Its basement, together with its Paleozoic blanket, crops out along the northern border as seen at Kelpin and Kuruk Tagh. In the laker region, late Protero- zoic (S~nian) tillites and stromatolites are found, showing that the Tarim platform took shape near the end of the Proterozoic just as the Yangtze paraplatforrn did. TIENSHAN—EHINGAN GEOSYNCLINAL FOLD SYSTEM AND ARGON FOLD SYSTEM These two systems are components of the great central Asiatic-Mongolian arcuate fold region, which extends be- tween the Sibenan platform, on the north, and the Sino- Korean paraplatfonn and the Tarim platform, on the south (Figure 16.1~. This fold region is separated into two halves by the Derbugan depth Eacture (eastern extension of the Mid-Mongolian depth fracture); the southem half is the Tienshan-Khingan geosynclinal fold system, while the northern haIfincludes the Argun fold system. The laker is Hsingkaiian, and the fondler consists, within the Chinese territory, of the Altai (Caledonides and Variscides), the Dzungarian (Vanscides), the Tienshan (Variscides), and the Kirin-Heilungkiang (Variscides) fold systems. All of them, except the southern Tienshan, are eugeosynclinal in nature. During the Caledonian cycle, volcanic activity, chiefly submarine, was widespread in the Ordovician and Silurian. Radiolarian cherts were found closely associated with spilites and ultrabasic intrusives forming ophiolitic suites, such as in the western Dzungaria. Dunng the Variscan cycle, the Devonian and Carboniferous were characterized by calc-alkaline volcanics (mainly an- desites). From Carboniferous to the end of the Permian,* the geosynclinal coronations accompanied by granitoid in- trusives were brought into intense and complicated folds, thus converting the geosyncline into a craton. During the Yenshanian cycle, a general tendency of increasing tec- tonic intensity from east to west can be recognized, es- pecially in the Kirin-Heilungkiang and Great Khingan regions, where strong remobilization and magmatism took place. During the Himalayan cycle, the general tendency of increasing intensity of tectonism was inverted. Fault- ing and uplift along the general strike of the Tienshan were strong and widespread, foxing lofty mountain ranges with sunken northern and southern foredeeps as *This is the Variscan orogeny. Caledonian erogenic: movements are also present but they are very limited in distribution.

188 well as intermontane depressions such as the famous Tur- fan Basin. KUNLUN—NANSHAN—TSINLING GEOSYNCLINAL FOLD SYSTEM AND TIBET—YUNNAN GEOSYNCLINAL SYSTEM These two tectonic units, separated from one another by the Chinshak-~ang-Red River depth fracture, occupy the vast territory south of the Tarim platform and Sino-Korean paraplatfonn, west of the Yangtze paraplatform, and north of the Tsangpo depth fracture (Figure 16.1~. The Kunlun-Nanshan-Tsinling geosynclinal fold system is a typical polycyclic geosynclinal fold system embracing three erogenic events. During the Caledonian cycle, a eugeosyncline came into existence mainly in the Nanshan, with well-developed ophiolite suites and glaucophane schist belts. During the Variscan cycle, a large portion of the system was miogeosynclinal, while eugeosyncIines were maintained only in Burkhanbuddha, Amnemachin, and along the Chinshakiang. During the Indosinian cycle, miogeosynclinal development con- tinued in ~e Sungpan-Kantze, Tsinling, and South Ko- konor Range, while the area along the Chinshakiang re- mained eugeosynclinal. Different parts of the system differ in age of folding. The Nanshan is Caledonian, the Kunlun is Variscan, while the Sungpan~antze anc] Tsin- ling are basically Indosinian. The Yenshanian and Hima- layan cycles are characterized by remobilization ac- companied by a variable amount of magrnatism. That Cambrian volcanics were over~rust upon late Tertiary red bed~s in the Lachishan of Chinghai and Jurassic sandstones were ove~rust upon Quatemary gravels at the Hungliu Gorge near Yumen of Kansu demonstrate that intensive horizontal compression has been a factor even in more recent geoIogical time. The Tibet-Yunnan geosynclinal fold system, situated to the south and west of the Chinshakiang-Red Biver depth fscture, is a Mesozoic arc uate fold system, consisting of ~e Sankiang* or Three-River (Indosinides), the Tangla (Yenshanides), and the Lhasa (Yenshanides) fold systems. The existence of the proposed Precambnan Tibetan Massif (Terman, 1973) is negated by recent observations. The Sankiang fold system includes western Yunnan and the Changtu district of Tibet and is characterized bY sreo- synclinal folds strictly controlled by depth Eactures among which the Chinshakiang depth fracture, the Lant- sangkiang depth fracture, and the Nukiang depth fracture are of prime importance. Along these depth fractures, eugeosynclinal activity took place, with polycyclic tec- tonism accompanied by intensive polycyclic magmatism. Not only the Paleozoic and Mesozoic, but also, in some localities, the Tertiary formations underwent tectonism and metamorphism in various grades and formed several tectono-magmato-metamorphic belts (the Ailaoshan, the *So named because it embraces the Chinshakiang, the Lantsang- lciang, and the Nukiang drainage systems. T. E. HUANG Lantsangkiang, and the Kaol~kungshan metamorphic belts). HIMALAYAN GEOSYNCLINAL FOLD SYSTEM This tectonic unit (Figure 16.1), situated to the south of the Tsanggo and Indus Rivers, consists of Sulaiman (in Pakistan), the Himalaya, and the Arakan Yoma (in Burma). It must be pointed out that true geosynclinal deposits are absent in the Himalaya proper but appear immediately to the south of the Tsangpo, where a more or less continuous belt of ophiolites is well exposed. The lofty ranges of the main Himalaya form in fact the northern border of the Indian craton and associated Hsingkaiides, the metamor- phic and sedimentary rocks of which were deeply in- volved in the H~malayan folding when the Indian craton collided with Tibet (Chang and Cheng, 1973; Gansser, 1964, 1966; Molnar and Tapponnier, 1975~. SOUTH CHINA GEOSYNCLINAL FOLD SYSTEM This tectonic unit is located to the south of the Yangtze paraplatfonn (Figure 16.1~. The author believes that this unit is not entirely Caledonian as had been prev~ously regar~led but consists of two parts of different character. The major part, still called the South China fold belt, is truly Caledonian, but the subordinate part, including coastal regions of Chekiang, F'ukien, and Kwangtung and the neighboring continental shelves as well as a greater part of the Hainan Island, belongs to the Vanscan, to be narned the Southeastern Maritime fold system. Because the latter was strongly transfonned by Mesozoic tec- tonism and widely covered with Jurassic and Cretaceous continental volcanics, its tectonic character has long been misjudged. NATANHATA EUGEOSYNCLINAL FOLD BELT AND UPPER HEILUNGliCIANG MIOGEOSYNCLINAL FOLD BELT These are components of the Mesozoic geosynclinal fold systems located to the east of the Siberian platform (Figure 16.1). The Natanhata belongs to the Sikhote-Alin fold system, while the Upper Heilungkiang belongs to the Mongolo~Okhotsk fold system. Both are Yenshanides. DEEP-SEATED STRUCTURES AND DEPTH FRACTURES IN CHINA DEEP- SEATED STRUCTURE From a comprehensive examination of available geo- physical and seismic data, it is preliminarily concluded that the lithosphere within China can be subdivided into several heterogeneous layers, as elsewhere in the world. This layering character is demonstrated by recent investi- gations (Hsi et al., 1974a; 1974b; 1975) of seismic sound-

An Outline of the Tectonic Characteristics of China FIGURE 16.2 Cross section of the Moho from Himalaya to Nanshan. 1, Tsangpo depth fracture; 2, Lantsang- kiang depth fracture; 3, Chinshakiang depth fracture; 4, East Kunlun depth fracture; 5, depth fracture along north- e~n margin of Tsaidam; 6, Altyn depth fracture; (by, Himalaya fold system; @), Lhasa fold system; 0), Sankiang fold system; (9, Tangla fold system; (I), Sungpan-Kantze fold system; 6), Kunlun fold system; by, Nanshan fold system. FIGURE 16.3 Cross section of He Moho from Lanchow to Talien. 7, depth fracture zone of North Nanshan; 8, depth fracture of western margin of Ordos; 9, depth fracture zone of Taihangshan; 10, Tancheng-Lukiang depth fracture zone; 6), Nanshan fold system; 0), Sino-Korean paraplatfo~. ,,,, ~~ - .so. ~ st. . ~ imalaya Tangla Kuniun Km' '~Lhasa 2 3 4 Gormu !_ _ ~ _ _ 30 - 1 50 - ~1 God / 8oi Lanchow Taiwan Km, .x ,; t: ~ 'A _ it) ~ I ing along a profile from Yuanshih to Tsinan in the Norm China Plain and another profile across the eastern section of the Tsaidarn Basin. In these regions the earth's crust possesses layers of velocity gradients in addition to low- velocity zones. A blocklike pattern is developed for the Chinese main- land, especially its eastern part, when the areal variations of the depth to the Moho (Mohorovicic discontinuity) is considered. Two.cross sections show the nature of the depth to the Moho (Figures 16.2 and 16.3~. Each block is confined generally by depth fractures of persistent devel- opment. For example, the Yinchuan-Liupanshan-Karh- Yunnan gravity gradient belt (Kunming-Yinchuan depth fracture system) separated eastern China from western China in the geological past. The Great Khingan-Tai- hangshan-Wulingshan* gravity gradient belt, a dividing line between the eastern and western belts of the Marginal-Pacif~c tectonic domain, came into existence long since the late Mesozoic. The northern section of the above-mentioned gravity gradient belt, i.e., the section from the Great Khingan to Taihangshan, coincides with a depth Eacture zone. The North Tsinling-North Huaiyang depth fracture zone is a geological dividing line between North and South China. The gravity gradient belt of West *No NO directed depth fractures are present along the Wulingshan. 189 N anshan Aksai s 6i 44. ~ 1~ ~ : ~ i ~ , ~ 3 ~> , (I) i ~ 1 ,. O It)o Con .,of coo s`3oA'o1 .9 Cohn _~ Talien ~ lij Kunlun-Altyn-North Nanshan essentially coincides with the depth fracture zones between the Tarim platforms and the Alashan Massif on the north and the geosynclinal fold belts on the south. Moreover, the present geomorphic features of China form the mirror image of the Moho, as shown in Figures 16.2, 16.3, and 16.4; mountains and plateaus correspond to the "downwards" of the Moho, while plains and basins agree with its "uplifts." This appears to be a result of a combined effect of the Pacific plate and the Indian plate acting against the continental block of China since the Indosinian, particularly since the Himalayan cycle. In these movements, not only the crust but also the upper mantle were involved. DEPTH FRACTURES OF CHINA The author classifies depth fractures into three classes according to their depth (Table 16.2~. Translithospheric fractures either came into being during a definite geo- logical period or continued in motion till today marked by a series of deep-focus earthquakes (Benioff zones). Lithospheric fractures are generally characterized by basic and ultrabasic complexes but without the develop ment of extensive ophiolites. Crustal fractures are of much

190 ~ ~— En -~ 'my ~ ~~- W-o ~ Ale-'- - - N :-? I: brow ~ ~. - all it  ~ --a At e,— If ~ ] q_f en. '--] To ~~ yak li tare P388~ ee ~ - - o ~ - c: ~ - o c) c) en Hi - et At; in 3 o _ V, et £ ._ ._ U. CO ~4 C) _

An Outline of the Tectonic Characteristics of China TABLE 16.2 Classification of Depth Fractures According to Depth 191 Examples Class Depth World China Sialic crusted fracture Simatic depth fracture Lithosphenc fracture Translithospheric fracture Dissecting the sial but not clearly extending into the sima Dissecting the whole crust but not extending into the mantle Dissecting the lithosphere but not extending into the asthenosphere Dissecting the lithosphere and extending deep into the asthenosphere Many Many RiR Valleys of East Africa Benioff zones along island arcs of He Pacific Heyuan depth fracture East Tsangchow depth fracture Tancheng-Lukiang fracture Tsangpo depth fracture smaller magnitude and are usually accompanied by acid and/or intermediate magmas (sialic fractures) or by basalts (simat~c fractures). Depth fractures can also be divided into three classes* according to their mechanical behavior: tension~epth fracture, compression-depth fracture, and shear-depth fracture. Fractures along midoceanic ridges and along the great rift valleys are tension-depth fractures. Fractures along the island arcs of the western Pacific (marked by deep-sea trenches) are examples of compression- or com- pression-shear-depth fractures. The San Andreas Fault of North America, the Altyn Depth Fracture, and the Tancheng-Lukiang Depth Fracture of China are ex- amples of shear-depth fractures. Depth fractures are not unchangeable; on the contrary, their formation, develop- ment, transformation, and extinction came one after another in succession during the long history of the evolu- tion of the earn. Numerous facts note that different sec- tions of the same depth-firacture zone might have ongi- nated in different geological ages, belong to different types, and possess different charactenstics. Based on the available geological and geophysical data, the main depth fares are listed in Table 16.3, and a brief description of them follows. Those belonging to the translithospherrc depth frac- tures are as follows: 1. The Tsangpo depth-~acture zone (Huang, 1960) ex- tends westward following the Indus River valley and was termed the Indus suture (Fitch, 1972~. It is now consid- ered as the suture line between the Indian plate and the Eurasian plate and is marked by well-developed ophio- lites extending about 2000 km within the confines of China (Chang and Cheng, 1973; Gansser, 1964~. 2. The longitudinal depth-fracture zone of Taiwan forms only a small section of the long depth fractures (BenioE zones) along the island arcs of the western Pacific. Its presence is marked by ophiolites, glaucophane schists and melanges, and frequent deep-focus earth- *ln fact, there are transitional types, such as compression-shear and tension-shear depth fractures, which are attributable respec- tively to compression, tension, and shear depth fractures accord- ing to their dominant mechanical behavior. quakes. This depth fracture is sinistral shear in mechani- cal behavior in contrast to the predominantly compres- sional Benioff zones. Available data indicate that the Norm Nanshan depth fracture, the Chinshakiang-Red River depth fracture, and the Derbugan (Mid-Mongolian) depth fracture are trans- lithospheric, or ancient plate sutures. In China a great majority of depth fractures in geo- synclinal regions and a pa* of those in platforms belong to lithosphenc depth fractures. Among them the domi- nant ones are the following: (a) The Tancheng-Lukiang depth fracture system is perhaps the major depth-fracture system of eastern Asia and is composed of the Tancheng-Lukiang, the Mishan- Tunhua, and the Yilan-Yitung depth fractures with a total length of about 2400 km. From south to north, it cuts through the Yangtze paraplatforrn, the Sino-Korean para- platforrn, and the Kirin-Heilungkiang fold system. It also appears to be an important volcanic, metallogenic, and seismic zone, especially during Meso-Cenozoic times. Sinistral shear occurred in the geological past, but dextral shear is apparent from analyses of modem earthquake mechanisms. Some geologists consider that this fracture system existed in Pre-Sinian times, while others suggest that it was formed in the Indosinian cycle. The fracture system continues northward into the Far East of the U.S.S.R. (b) The Kunlun-Tsinling depth fracture system includes the North Nanshan-Nor~ Tsinling-North Huai- yang fracture zone, the northern margin of Tsaidam- South Kokonor Range-North Tsinling-North Huaiyang fracture zone, and the east Kunlun-Tsinling fracture zone. These complicated fracture zones controlled the origi- nation and development of the Kunlun-Tsinling geo- synclinal fold system and constitute a geological dividing line between northern and southern China. This divid- ing line shifted from the line of North Nanshan-North Tsinling-North Huaiyang during the Caledonian cycle to the line of the northern margin of Tsaid~n~outh Koko- nor Range-North Tsinling-North Huaiyang during the early Variscan cycle, and finally to the line of East Kunlun-North Tsinling-North Huaiyang since the late Variscan cycle. It is important to note that these fracture zones deepen from east to west. Widespread submarine

192 TABLE 16.3 The Pnncipal Depth Fractures in Chinaa T. lc. HUANG Number Oepth Fracture Zone Depth Character Age of Activity Magmatism and Metamorphism 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 23 24 25 26 27 28 29 30 31 32 Darbut Irtish Karameili North Margin of Central Tienshan Derbugan Silamulun Cherchen No~ern Margin of Inner Mongolian Axis Altyn North Nanshem-Northem Margin of North Tsingling~orth Huaiyang Northen~ Margin of Tsaidam~outh- em Margin of No~ Tsingling- North Huaiyang East Kunlun Kantze-Litang Chinshakiang-Red River Lantsangkiang Nukiang Anningho Hsiaokiang Lungmenshan Lingshan Wuchuan-Szehwei Heynan Lishui-Haifeng Changle-Amoy Taihangshan East Tsangchow Liaocheng-Lankao Tancheng-Lukiang Yilan-Yitung Fushun-Mishan Tsang+Indus Longitudinal Valley of Taiwan aNumbers correspond to those circled on Figure 16.4 T. Translithosphenc fracture zone L, Lithosphenc fracture zone C, Crustal fracture zone c-s, Compression and compression-shear t-s, Tension and tension-shear s, Shear s-s, Sinistra1 shear ~ . L L L L L L L L L L L L L L L C L C C C C C C C C L C L T T s c-s c-s c-s c-s c-s s-s c-s s-s c-s c-s c-s c-s c-s c-s c-s c-s c-s c-s c-s a a c-s c-s c-s t-s t-s s-s t-s s-s c-s s-s Pz Pz Pz Pt?,Pz Pz Pz Pz Pt?,Pz Pz Pt,Pz Pz Pz Pz Pz Pz Pz Pz Pz Pz Pz Pz ~z Pz ~z Mz Mz Mz Pt?,Mz? ~z Pz Mz Mz OS I,gs OS,gs 5,m 5,m I,m y y y y OS,m OS,m,gs Pt, Active since Proterozoic Pz, Active since Paleozoic Mz3 Active since Mesozoic OS' Oph~olite suites I, Basic and ultrabasic complexes y, Gran~toid plutons ,B, Basalts a, Compression~hear and tension~hear, aItemating m, ~Ielanges as, Glaucophane schists volcanics occur along the westem section, where eugeo- synclinal conditions prevailed, and especially in No~ Nanshan, where remarkable zones of ophiolites 700 km long and zones of gIaucophane schists more than 100 km long were recently discovered, indicating the existence of an ancient plate suture. On the contrary, the eastern sections of these hactures, being characterized by flysch formations and scanty ultrabasic rocks, appear miogeosyn- clinal. This is perhaps the reason why the Tsinling geo- syncline became appendixlike toward the east between the Sino-Korean and the Yangtze paraplatfortas. Crustal fractures are numerous in China as repre- sented by large-scale acid-intermediate plutons with ex- tensive volcanics of the "Pacif~c" type, lack of ultrabasic rocks, and zones of strong compression followed by strong tension along the same zones, thus giving rise to a series of Cretaceous-Tertiary basins Blled w~th red beds along

An Outline of the Tectonic Characteristics of China growth faults* of depth-fracture nature. ARerwards, in late Tertiary and Quaternary, they were subjected to compres- sion again. In this respect, the Heynan depth fracture is one of the fairly good examples. Another type of crusty fracture in eastem China is represented by numerous tension fractures along which originated and developed Mesozoic and/or Tertiary fault basins, filled with terrestrial elastic deposits. Among them the East Tsangchow depth fracture, the Liaocheng- Lankao depth fracture, and the Weiho-Fenho graben (young Tertiary) are the most important. Unlike the Heyuan depth fracture, they originated as tension frac- tures in Cretaceous or old Tertiary time and are buried by young Tertiary and Quatemary deposits. Thus' they are typical growth faults, forming with elastic deposits, the so-called "dustpan-like basins" accompanied by basalt flows. Geophysical data show that these depth fractures dissect the entire earth's crust but do not extend deep into the upper mantle and therefore should belong to simatic fractures. THE TECTONIC DEVELOPMENT OF CHINA Two major stages oftectonic development of China before Paleozoic times can be distinguished. The first is the stage from Archean to early Proterozoic or the Pre-Sinian stage, through which the Sino-Korean paraplatform came into being about 1700 m.y. ago. The second is the late Proterozoic or Sinian stage, through which the Yangtze paraplatform and the Tarim platform came into being about 700~00 m.y. ago. The great importance of the Yangtze orogenic cycle should be emphasized here. Available data show that platform` areas formed through the Yangcze orogeny were more extensive than the pres- ent Tangle paraplatfiorm and Tarim platform. Platform- type fo~l~ations correlatable with the Sinian System lying unconformably upon metamorphosed Sinian Subera- them, are met with In East Kunlun, Tsinling, Altyn, aIld along the northern border of the Tsaidam Basin. More- over, phosphatic sediments of early to middle Cambrian age have been discovered in the Tienshan and Peishan. These facts suggest that a gigantic craton, to which the name of Chinese proto-platform is temporarily given, might have been created by and maintained through the Yangtze orogeny with a time span of about 200 m.y. (Sinian System to lower Cambrian). The tectonic development of China since Paleozoic time could be divided into two stages, the Paleozoic stage and the Meso-Cenozoic stage. In spatial distribution, three major units (tectonic domains) were developed dur- ing this long time period. They are the Pal-Asiatic (~) tectonic domain, the Marginal Pacific (~-P) tectonic do- main, and the Tethys-Himalayan (T-H) tectonic domain. The PEA tectonic domain took shape and developed *Also termed synsedimentary faults. 193 through the Hsingkaiian, the Caledonian, and the Varis- can Progenies. The process was roughly as follows: as soon as the Sayan-North Mongolian-Argun Geosynclinal System was folded and uplifted at the end of the early Cambrian (Hsiangkaiian), the Chinese proto-platform was disintegrated and partially transformed to form the Tienshan, the Nanshan, the Tsinling, and other geosyn- clines, all of which were again folded, transfol~ed, and consolidated at the end of the Variscan cycle. As a result, these newly consolidated geosynclinal folds joined and "cemented together" the pre-existing four platforms, the Siberian platform, the Tarim platform, the Sino-Korean paraplatforrn, and the Yangtze paraplatform, to form a gigantic craton called the Pal-Asia. During the Meso- Cenozoic stage, the tectonic development of China was under the control of the Marginal Pacific tectonic domain and the Tethys-Himalayan tectonic domain. The former may be subdivided into an inner or Cenozoic tectonic belt and an outer or Mesozoic tectonic belt. The Taiwan fold system is the only representative of the inner belt within the Chinese territory. The outer belt of great importance was superimposed upon older tectonic units of various ages. In addition to the Southeast Coast Variscides and the Mesozoic Northeast Asiatic geosynclinal fold system, they are, from north to south, the Inner Mongolian~reat Khingan Variscides, the Kirin-Heilungkiang Variscides, the Sino-Korean paraplatforrn, the Yangtze paraplatform, and the South China fold system (Caledonides). The belt is characterized by large swells and depressions trending NE or NNE, with Mesozoic blanket folds and faults, ex- tensive "Pacif~c-type" volcanics, and large-scale granitoid intrusives. THE POLYCYCLIC TECTONIC EVOLUTION OF CHINA For many years, Stille's concepts of tectono-magnetic cy- cles have been prevalent in models of geosynclinaI de- velopment: 1. A geosynclinal period with initial magmatism, mainly basic and ultrabasic rocks; 2. An erogenic period with synorogenic magmatism, mainly granitoid intrusions; 3. A quasicratonic phase with subsequent magmatism, mainly porphyries and andesites; and 4. A full cratonic period with final magmatism, mainly basalts. Since then, many well-known tectonic and economic geologists have accepted this concept (monocyclic concept) especially in the Soviet Union and Western Europe, for example, Beloussov (1962), Rittrnann (1960), deSitter (1964), and Aubouin (1965~. In the authors explorations of many key regions of the geology of China, he arrived at the conclusion that geo- synclines developed polycyclically, both in orogeny and

194 in magmatism (Huang, 1945, 1960; Huang and Chiang 1962; Huang et al., 1965, 1974~. Van Bemmelen (1949) also advocated a polycyclic development for the Indo- sinian geosynclines. Recent geological mapping in the Tienshan and Nanshan again confirms the view of poly- cyclic development of geosynclinal fold belts, which is briefly summarized below. 1. The Tienshan fold system, especially the North Tienshan consists of eugeosynclinal Vanscides. The Nanshan fold system, especially the Norm Nanshan, con- sists of eugeosynclinal Caledonides. Both exhibit un- questionable polycyclic Progenies; seven Progenies, three being the most important, are found in the Tienshan, while four Progenies (Figure 16.5), of which the last is Me most important, are found in the Nanshan. Roughly coinciding with the Progenies, intrusion of grani- toid plutons took place, indicating that magrnatism gen- erally corresponds to important Progenies. 2. Submarine volcanic eruptions are also polycyclic. In the Tienshan, intermediate volcanics (andesites) prevail and are found in four cycles corresponding to four oro- genic cycles. The Nanshan volcanics, generally basic in nature (tholeiites and spilites), can also be divided into FIGURE 16.5 Diagram showing the polycyclic development of the Nashan eugeosyncline.,B: basic submarine exclusives, mainly ~oleiites and spi- lites; a: andesite; Ha: andesite having an acidic tendency; A: andesite having a basic tendency; p: rhyolite; F: fly- schoid deposits, often alternating win ,B and c'; M: molasse, divided into two stages, D'_2 and D3, in Nanshan; I: basic and ultrabasic rocks, generally serpentinized; OS: ophiolitic suites, formed of serpentinites, basic ex~u- sives (with pillow structures) and radi- olarites; G: basic intrusives; y: granitic plutons. The size of the letters repre- senting the different rocks corresponds to the degree of their importance. T. lC. HUANG four cycles corresponding to the four most important oro- genic cycles. 3. Basic and ultrabasic rock complexes are Iilcewise polycyclic. In the Nanshan they are best developed and divisible into five cycles, each corresponding to an oro- genic cycle (Figure 16.~), of which the second cycle (Each) is the most important. Basic and ultrabasic rocks, together with radiolarian chert, form three typical oph~- olitic suites in three different ages (Z. E2, Off. On the contrary, ultrabasic rocks are poorly developed in the Tienshan, where ophiolites are absent. 4. Marine flyschoid sediments are well developed in both regions, and together with submarine volcanics form polycyclic volcano-flyschoid formations. However, they are andesite~lyschoid in the Tienshan and basaltic- flyschoid in the Nanshan. Continental molasse is devel- oped in both regions at the close of the geosynclinal evo- lution. The Devonian molasse of the Nanshan is the most typical. It is important to emphasize that polycyclicity of geo- synclinal development must not be considered as simple repetitions of geological processes as some writers be- lieve them to be, but rather they are processes of vectorial Ophiolitic Suites _ _ . ,~ ~ J Bare-- F Oslo 1 ! - 1 ~ 1 1 I l l ~ 1 ,~,\ ~ I ! 1 Us I 1 Cycle Age Marine Sediments. S u b~narine Extrus ives . V D-C ~ ~ _ F U A' S ~e 2 F urn 03 ~ ~5 ~ ~ _ O. ~~ II2 =~ I C ~ ~ O 5 ~ ~ :'~I Z: ~ .~ 1 Basic-U Itrabasic G ran itic ~ Complexes Plu20n~ ~ Remarks -gags I j Basic- ultrabasic complex chiefly | in S. Nanshan ~ 1 r I 1 Possible break beta een (2 and (: _ I ram I r _ . ~ Major angular unconforll~it~ ~ .\ngular unconformity ~ Stratigraphic break

An Outline of the Tectonic Characteristics of China . , , Granitic Marine Sediments. i Basic and Ultrabasic Cycle Age Plutons Submarine Extrusives I [ntrusives Remarks 1 -- .~1 1 1 i ~~ !~-- ,,, 21 ! it' Y-1 1 P'1 1 ' i c l~1 , m ~ it ~ ~ (-', 1~ ! ~ 1~— 'J D3 I Y , > TT D2 ~ cr D ~ al ,c~ 4~ j c, I J slj _ol ?- FIGURE 16.6 Diagram showing the polycyclic development of the Tien- shan eugeosyncline. Symbols are as in Figure 16.5. Ash spiral-like evolution with a regular arrangement. In the Nanshan geosyncline, for instance, ophiolites began to develop in Cycle I, reaching their acme of development in Cycle II, especially in Subcycle II2.* In Cycle III, although basic and ultrabasic rocks are present, no ophi- ol~tes were formed. In Cycle IV, they dwindled away. On the contrary, flyschoid sedimentation, poorly developed in Cycle I and Subcycle II', strengthened in Subcycle II2 and Cycle III, and reached its acme of development in Cycle IV. Folding, faulting, and granitoid intrusion all played an important role in each cycle, but it is obvious that they possessed a tendency of increasing activity from Cycles II to III and reached their peak of development in Cycle IV. Cycle V indicates the close of geosynclinal sedi- mentation, when extensive molasse Finned. The polycyclic evolution toward a definite direction of the Tienshan geosyncline is even more prominent (Figure 16.61. There, granitoid magmatism was not im- portant in Cycles I and Il.? but was greatly strengthened in Cycle III, and became best developed in the later part of Cycle III, i.e., in late Carboniferous time. Granitoid mag- matism rapidly decreased in Cycle IV. Moreover, grani- toid plutons of Cycles I and lI show gneissic structures absent in those of Cycles III and IV. It must be pointed out that each of the five granitoid magmatisms is clearly The author divides an erogenic cycle into subcycles (see Fig- ure~, 16.5 and 16.6). 195 l JO a F F (Ida I Basic and intermediate j extrusion es continental I Cyclic acid-intermediate I exlrusives G: Basic intrusives mainly stocks Two movements of uplift in C: l- s G ~ G: Basic intrusives | with gneissic structure, maim, stocks p in southern Tienshan Andesites, rh$olites as intercalations connected with an orogeny, while the occurrence of large- scale granite batholiths is generally contemporaneous with principal Progenies. From the petrochemical point of view, granitoid intrusives of early Variscan abound in pla- gioclase granites and albite granites; those of middle Variscan are generally normal granites, while those of late Variscan are characterized by kali-granites and alaskites, and even syenites appear. In other words, the petro- chemical characteristics of granitoid intrusives change from acid-inte~ediate to acid, then to acid~lkaline, and finally to alkaline. From the above discussion, the author arrives at the preliminary conclusion that the Nanshan geosyncline is characterized by poIycyclic basic extrusives (tholeiites and spilites), which, together with ultrabasic rocks, form polycyclic ophiolitic suites. Thlls, the two types of geo- synclines are quite different from each other in character. Poly~yclic geosynclina1 development is also distinctly revealed in other geosynclines of China, among which the Tsinling geosyncline and the East Kunlun-Tangla geo- syncline are typical (Huang et al., 1974~. Many of the famous geosynclinal systems in the world, such as the Appalachian, the Cordilleran, the Uralian, and the Tasman geosynclines' are also characterized by poly- cyclic development. Consequently, polycyclic develop- r~ent of geosynclines is, without doubt, to be considered as the general rule. For the sake of simplicity, a preliminary model for the

196 development of geosynclinal fold belts as shown in Fig- ure 16.7 is proposed. It can be seen Tom Figure 16.7 that Me development of geosynclinal fold belts includes the following: 1. Early geosynclinal cycles, including Cycle I, Cycle II, and possibly Cycle III; 2. Principal geosynclinal cycles, including Cycles III and IV; and 3. Postgeosynclinal cycles, including Cycles I and II. It must be noted that each geosynclinal cycle may include initial magmatism (ophiolites), synorogenic magmatism (granites), and subsequent magmatism (porphyries and andesites) but usually without final magmatism (plateau basalts). Moreover, flysch and molasse formations are also polycyclic, while the postgeosynclinal cycles are charac- tenzed by block-faulting accompanied by folding, again with polycyclic granites and polycyclic molasse (such as in Eastern China). If we accept the theory of plate tectonics to interpret the origin and development of geosynclines, as quite a number of geoscientists are doing, we come to the conclu- sion that plate motions are likewise polycyclic in nature. This is particularly manifested in the case of the Tasman Geosyncline. Scheibner (1972) pictured the development of that geosyncline as a series of successively eastward- accreting continental blocks, subducted by a series of eastward-retreating oceanic crust. From his palinspastic maps, it appears clear that the Tasman fold belt was formed by polycyclic plate motions. FIGURE: 16.7 Diagram showing polycyclic development of geosyn- ,'0~O clinal foldbelts. ,6000 ~ '.ooo a — 1 2000 1~0 ~ 1 to) 6000 2000 T. E. HUANG PRELIMINARY OBSERVATIONS ON PLATE TECTONICS IN CHINA TSANGPO DEPTH FRACTURE ZONE Geological and geophysical data indicating the applica- bility of plate tectonics along the Tibetan part of the Himalayan geosyncline were collected and analyzed by Chinese geologists (Huang et al., 1974; Chang and Cheng, 1973~. Recent observations in the Tsangpo Valley reveal the occurrence of abundant ophioIites, and to the east of Shigatze typical melanges with exotic bloclcs of limestone containing Triassic and Jurassic fossils are found. It is thus probable that the Tsangpo depth-fracture zone is a subduction zone with the southern oceanic plate underthrusting the northern Asiatic plate. Such subduct- ing activities might be polycyclic in nature. NANSHAN—TSINLING REGION The eugeosynclinal character of the northern Nanshan was described in previous articles (Wang and Liu, 1976; Huang et al., 1965, 1974), while its polycyclic develop- ment has already been stressed. The glaucophane schist zone chiefly consists of quartz Muscovite glaucophane schist and garnet-epidote glaucophane schist, and the majority of He tholeiites of the ophiolitic suites are char- acterized by very low K2O contents (usually <0.3 butt). It is interesting to note that all of these features are similar to those of other paleogeosynclines in the world. The 20000 M East tics ' Principal cycles Po*~-c~les .. cop ~ _~ <2 i ads, coca 1 I C>c~ ~ I ~N 1 ~ >an pats - Sg - Worm I ~ ~ ! too ) ^" ~ f55~ q Posit 1 & - blazon ~ Cycle ~ . Cycle ~ ,Cyc~ 1 Cay ~ LEGEND Ale 0' Ophwliiic Suiles (hoot li7zes S~Ul7Z9 depth f7.aci~r.es) F`'Fz'F3 Bush: ull7abas~c comic:: ice of, go alone character Flinch formafio7`s I',M`,M~ ~,~,~37~ ~~ ~05' (g`,gz ~ems" ~ H.~l folding `iles3 early 7, ~ i7~ler~ncli - , ~~ Folding of secondary ~mp~- la~r /, f alkaline character 1 Sac ~Cl, ~2~3 Po - ~e. AL minor m`~ Ads, eat BY, Of band, la~ Mola~se forays \\ Back faming

An Outline of the Tectonic Characteristics of China ophiolites are polycyclic, with successive motions being from the south to the north (Wang and Liu, 1976~. The Northem Tsinling geosyncline is the continuation of the Northern Nanshan. Observations show that the Caledonides of the Tsinling extend from south of P~ochi to Nanyang Basin, probably including ophiolitic patches. Li (1975) advocates that the Tsinling is characterized by plate-tectonic interactions. It is possible that the eastern Tsinling fold belt was formed by the mutual approach between the Sino-Korean and the Yangtze paraplatfo~ms. CHINSHABIANG—RED RIVER DEPTH- FRACTURE ZONE Recent mapping in the Chinshakiang-Red River region discloses evidence for paleoplate motions. South of Batang, in the western side of the upper Chinshakiang fracture zone, occur typical ophiolitic melanges charac- terized by serpentinites, spilites with pillow structure, and various kinds of basic to ultrabasic rocks intercalated with radiolarian cherts, while exotic blocks of omphacite- eclogite; dillage~innamon stones; and particularly De- vonian, Carboniferous, and Permian limestones are found in the matrix. In the eastern side of the same zone, wild- flyschlilce deposits characterized by a matrix of argilla- ceous and arenaceous rocks with various exotic blocks yield many fossils ranging in age from Silurian to Per- mian, which are, without exception, older than those in the matrix. Judging from all the facts observed so far, it is probable that a western oceanic plate approached and collided, from Permian to Late Triassic time, with the eastern suboceanic plate (this belongs to the Indosinian orogeny). Along the Red River valley, the existence of ophiolitic melanges and glaucophane schists is also prob- able. Apparently, the Chinshakiang zone and the Red River zone belong to the same system of convergent plates. Since in western Yunnan and northern Burma there occur a series of lithospheric fractures, i.e., Chin- shakiang-Red River, Lantsangkiang, Nukiang, and Naga- Arakan Yoma, and since the principal ages of these Eacture zones seem, respectively, to be Indosinian, early Yenshanian, late Yenshanian, and Himalayan, the hy- pothesis is suggested that polycyclic plate subductions happened from east to west in a successively repeating manner. RE FE RE NC E S Aubouin, J. (1965). Geosynclines, in Development in Geotec- tonics, Elsevier, Amsterdam. Beloussov, V. V. (1962). Basic Problems in Geotectonics, McGraw-Hill, New York, 809 pp. Chang, C. F., and H. L. Cheng (1973). Some tectonic features of the Mt. Joimo Lungma area, southern Tibet, Sci. Sin. 16, 257-265. 197 Ch'eng, Y. C., F. D. Chung, and Y. C. Su (1973). The Chen-Tan series in northern and northwestern Chinese areas, Acta Geol. Sin. 1973, no. 1 (in Chinese). Chinese Academy of Geological Sciences (1976). The Geological Map ofthe Peoples Republic of China (1:4,000,000), Map Pub- lishers (in Chinese). deSitter, L. U. (1964). Structural Geology, 2nd ea., McGraw-Hill, New York, 530 pp. Fitch, T. J. (1972). Plate convergence, transcu~Tent faults and internal deformation adjacent to southeast Asia and the west- ern Pacific, 1 Geophys. Res. 77, 4432~460. Gansser, A. (1964). Geology of the Himalayas, Interscience, New York. Gansser, A. (1966). The Indian Ocean and the Himalayas: a geo- logical interpretation, Eclogae Geol. Hel?;. 59, 831 848. Hsi, C. W., S. F. Feng, C. S. Li, H. P. Chen, K. T. Wen, C. J. Chang, and C. C. Hsiung ( 1974a). The background of the deep structure in the middle of the Northern China Plain and the Hsing-Ta earthquake, Acta Geophys. Sin. 17, no. 4 (in Chinese). Hsi, C. by., S. F. Feng, C. S. Li, H. P. Chen, K. T. Wen, C. J. Chang, and C. C. Hsiung (1974b). The reflection wave of the deep earthquake in Chai-Ta-Mu Basin and crustal structure, Acta Geophys. Sin. 17, no. 2 (in Chinese). Hsi, C. W., S. F. Feng, C. S. Li, H. P. Chen, K. T. Wen, C. J. Chang, and C. C. Hsiung (1975). The background of the deep structure in the middle of the Northem China Plain and the Hsing-Ta earthquake, Acta Geophys. Sin. 18, no. 3 (in Chinese). [Iuang, T. K. (1945). On major tectonic fonns of China, Geol. Mem. Ser. A, no. 20, National Geological Survey of China. Huang, T. K. (1964)). The main characteristics of the structure of China: preliminary conclusions, Sci. Sin. 9, 494 544. Huang, T. K., and C. F. Chiang (1962). The primary study of the crust rules developed from the polycyclic point of view, Acta Ceol. Sin. 40, no. 1 (in Chinese). Huang, T. K., C. K. Chang, C. M. Chang, and K. M. Chen (1965). On eugeosynclines and miogeosynclines and their polycyclic development, Chin. Acad. Geol. Sci. Selec. Prof. Papers Reg. Tectonics, Sec. C, Peking (in Chinese). Huang, T. K., C. S. Jen, C. Chiang, C. M. Chang, and C. K. Chang (1974). The new knowledge of some characteristics for Chinese plate tectonics, Acta Geol. Sin. 1974 (in Chinese). Lee, J. S. (1939). The Geology of China, Thomas Murby & Co., London, 528 pp. Li, C. Y. (1975). The primary analysis for the development of some parts of Chinese regional structure using plate tectonic theory, Acta Geophys. Sin. 18, no. 1 (in Chinese). Molnar, P., and P. Tapponnier ( 1975). Cenozoic tectonics of Asia: effect of a continental collision, Science 189, 419~26. Rit~mann, A. (1960). Vulkane and ihre Tatigleit 2. umgearioeite- ten and erw., Add., Enke, Stuttgart, 336 pp. Scheibner, E. (1972). Actualistic models in tectonic mapping, 24th Sess. let. Geol. Congr. Can. 405~22. Terman, M.J. (principle compiler) (1973). Tectonic Map of China and Mongolia, (1:5,000,000), Geol. Soc. Am. van Bemmelen, R. W. (1949). The Geology of Indonesia, Govern- ment Printing Office, The Hague, 732 pp. Wang, C., and H. Y. Liu (1976). The ancient oceanic crust in western China's Ch'i-Lien mountains and its plate tectonic meanings, Geol. Sci. 1974, no. 1 (in Chinese).