Search

Oceanic crust west of North America at the beginning of the Jurassic belonged to the Kula plate. The development of the western margin of North America since the Jurassic reflects interaction with the Kula plate, the Kula-Farallon spreading center and the Farallon plate. The Kula plate ceased to exist in the Paleocene and later developments were caused by interaction of the Farallon plate and, subsequently, collision with the East Pacific Rise. At the beginning of the Jurassic, when spreading between North and South America began, the Kula-Farallon-Pacific triple junction moved to the north relative to North America, and the eastern end of the Kula-Farallon spreading center swept northwards along the continental margin. During the Paleocene, Kula-Pacific spreading ceased and the Kula plate fused to the Pacific plate. Throughout the Mesozoic, subduction of the Kula plate took place along the Alaskan continental margin. When the Kula plate joined the Pacific plate a new subduction zone formed along the line of the present Aleutian chain. Wrangellia and Stikinia, anomalous terrains in Alaska and northwestern Canada respectively, were emplaced by transport on the Kula plate from lower latitudes. Hypotheses which require transport of these plates in the Mesozoic from the “far reaches of the Pacific” ignore the problem of transport across either the Kula-Pacific or Kula-Farallon spreading centers. The interaction of the Kula plate and western North America throughout the Jurassic and the Cretaceous should result in emplacement of these terrains by motion oblique to the continental margin. Tethyan faunas in Stikinia must come from the western end of Tethys between North and South America, not the Indonesian region at the eastern end of Tethys. As the northeastern end of the Kula-Farallon ridge moved northward, the sense of motion changed from right lateral shear between the Kula and North American plates to collision or left lateral shear between the Farallon and North American plates. Left lateral shear along zones analogous to the Mojave-Sonora megashear may have been the means by which anomalous terrains were transported to the southeast into the gap between North and South America forming present day Central America. Such a model overcomes the overlap difficulties suffered in previous attempts to reconstruct the Mesozoic paleogeography of Central America.

A tomographic back projection scheme has been applied to S and SS travel times to invert for shear velocity below the North American plate. The data range in distance from 8° to 80°, and a total of 3923 arrival times were used. First arrivals were measured directly off the seismograms, while the arrival times of later arrivals were found by a waveform correlation technique using synthetic seismograms. The starting model was laterally heterogeneous in the upper 400 km to account for the first-order differences in ray paths already known. The model was divided into blocks with horizontal dimensions of 500 km by 500 km and varying vertical thicknesses. Good resolution was obtained for structure from just below the crust to about 1700 km depth in the mantle. In the upper mantle a high-velocity root was found directly beneath the Canadian shield to about 400 km depth with the Superior province having the highest velocity and deepest root. The east coast of the United States was found to have intermediate velocities from 100 to 350 km depth and the western United States the slowest velocities at these depths. Below 400 km depth the most significant structure found is a slab-shaped high-velocity anomaly from the eastern Carribean to the northern United States. Beneath the Carribean this anomaly is almost vertical and extends from about 700 km to 1700 km depth. Further to the north, the anomaly dips to the east with high velocities at 700 km depth in the central United States and high velocities below 1100 km depth beneath the east coast. The anomaly is about 1% in magnitude. This lower-mantle anomaly may be associated with past subduction of the Farallon plate beneath North America.

Summary. If intermediate and deep earthquakes occur in the coldest portions of the downgoing slabs of lithosphere, then different lengths of seismic zones represent different temperatures in the slabs. As the slab descends through the aesthenosphere, it warms primarily by conduction of heat through its upper and lower surfaces. Isotherms are advected downwards to distances approximately proportional both to the rate of subduction and to the square of the thickness of the lithosphere. Consequently, the lengths of seismic zones should be approximately proportional to the product of the rates and the squares of these thicknesses. As these thicknesses are approximately proportional to the square root of the age of the lithosphere, the lengths ought therefore to be approximately proportional to the product of the convergence rates times the ages. Although there is considerable scatter, observed lengths are approximately proportional to such products, and are not simply related to the rate, the age or the thickness alone. The data crudely fit the relationship: length = rate × age/10. Using this relationship, we infer that the Philippine sea and Pacific plates move slowly, if at all, with respect to one another and that the Farallon plate may have been too young to be subducted to a great distance beneath western North America in the Palaeogene. Calculations of temperatures at the depths of the deepest events suggest that these cut-off temperatures increase from about 600 ± 100°C at 200 km to 830 ± 50°C at 650 km depth, but the cut-off potential temperature is approximately constant. Assuming that the strength is a thermally activated parameter, and using the activation energy for olivine, a crude estimate of activation volume is obtained from the dependence of the cut-off temperature on depth.

The age and paleogeographic setting of middle to late Tertiary fluvial deposits in southern California are consistent with the theory that subduction of the Mendocino fracture zone produced a north-facing, north-moving slope and zone of deformation in the North American plate. The Mendocino fracture zone was a 1 km high, north-facing scarp in the Farallon plate that moved northward at about 3 cm yr −1 . Subduction of this scarp produced isostatic stress in the overlying North American plate. If this stress were expressed in the crust as a north-facing slope, then a given area should have experienced the following sedimentary evolution: (1) normal (probably oceanward) drainage when the slope was far to the south; (2) reorientation to northward drainage as the slope moved to a position immediately south of the area; (3) deposition of coarse sedimentary units, possibly accompanied by down-to-the-north faulting, when the slope reached the area; (4) uplift, deformation, and erosion of previously deposited sediments as the area was uplifted along the slope; and (5) reestablishment of normal drainage. Compilation of data from sedimentary sequences in southern California supports this model. In particular, the time of fracture-zone passage was marked by strong dominance of northward stream flow. These relationships suggest that middle Tertiary fluvial sedimentation in the southern California area was strongly influenced by passage of the fracture zone.

SummaryThe Wagwater Belt is a fault-bounded structural unit in which more than 3000 m of early Tertiary sedimentary and volcanic rocks are exposed. Geochemical analyses of the volcanic rocks show that they comprise a bimodal suite of plateau-type tholeiitic basalts and calcalkaline dacites. The extrusion of these volcanics is associated with the development of an interarc basin, the Wagwater Basin, at the beginning of the Cenozoic Era. The Wagwater Basin formed as a result of the splitting of a mature late Cretaceous volcanic arc into a frontal and third arc represented by the Blue Mountain Massif and the Clarendon Block respectively. This model for Jamaica can be correlated with tectonic events occurring in the NW Caribbean during the early Tertiary. The creation of the Wagwater Basin and the eruption of the basalts is related to the initial opening of the Cayman Trough. The cessation of dacite volcanic activity in Jamaica signified the separation of the Caribbean Plate from the East Pacific Farallon Plate.

Large Cenozoic clockwise rotations defined by paleomagnetic data are an established fact in the Pacific Northwest, and many tectonic models have been proposed to explain them, including (1) rotation of accreted oceanic microplates during docking, (2) dextral shear between North America and northward-moving oceanic plates to the west, and (3) microplate rotation in front of an expanding Basin and Range province. Stratigraphic onlap relations and local structure indicate that microplate rotation during docking was not a major contributor to the observed rotations. Coast Range structures, Basin and Range extension, and paleomagnetic data from middle Miocene (15 Ma) Coast Range rocks indicate that dextral shear is responsible for at least 40% of the post-15 Ma rotation of the Coast Range and that Basin and Range extension is responsible for the remainder. Reconstructions based on extrapolation of this ratio back to 37 and 50 Ma are consistent with reconstructions based on paleomagnetic and stratigraphic relations in older rocks and suggest that dextral shear has, been a significant contributor to rotation during most of Tertiary time. Changes in the dextral-shear rotation rate over the past 50 m.y. correlate directly with changes in the velocity of the Farallon plate parallel to the coast and provide a strong argument for oblique subduction as the driving mechanism. Continental reconstructions incorporating shear may provide constraints on the rate of extension in the northernmost Basin and Range region and suggest 17% extension since 15 Ma, 39% since 37 Ma, and 72% since 50 Ma near latitude 42°N.

Palinspastic and global plate tectonic reconstructions of the southwestern United States, combined with analysis of the distribution of Cenozoic igneous activity suggest that Neogene and younger volcanic rocks in the Continental Borderland and Coast, Transverse and Peninsular Ranges appear to reflect subduction of the Farallon plate south of the Mendocino Fracture Zone and overriding of the Pacific—Farallon ridge. The igneous activity of Neogene age in coastal California appears to be a continuation of fields of magmatism in adjacent areas of southeastern California, Arizona and Baja California. With the close approach of the Pacific—Farallon ridge to the North American continental margin, progressively younger, hotter and thinner Farallon plate underwent subduction beneath California, resulting in more elevated temperatures at shallower depths and closer to the trench. Consequently, the associated volcanic arc apparently shifted to the west. Volcanic activity in coastal California appears to have been most widespread shortly after overriding of the Pacific—Farallon ridge occurred— that is, at that time at which the oceanic lithosphere was thinnest, and hot, partially molten asthenosphere was closest to the subduction zone. The zone of near-coast igneous activity in California was probably limited on the north by the extension of the Mendocino Fracture Zone in the descending Farallon plate. South of the fracture zone the lithosphere was young and hot; to the north it was older and, therefore, colder. Thus, Neogene igneous activity in the Basin-Range Province was associated with more typical subduction of older lithosphere involving a deeper, down-dip locus of melting, resulting in a wider arc-trench gap. The abrupt narrowing of the Neogene arc-trench gap from north to south occurred near the present-day Garlock Fault. The role of oblique and en echelon strike-slip faulting in permitting and localizing middle Cenozoic magmatism in the areas west of the San Andreas Fault is uncertain. Discrepancies in timing of movement on the northern and southern San Andreas Fault seem to require oblique strike-slip faulting, and by association, igneous activity, in the Borderland and Transverse Ranges prior to 10 m.y. B.P. Oblique, divergent strike-slip faulting in the San Francisco Bay area does seem to be responsible for significant igneous activity between 14 m.y. B.P. and present. The Page Mill, Grizzly peak , Sonoma and Clear Lake volcanic fields occur in zones of divergence between faults of the San Andreas, Calaveras and Hayward systems. Further, the age of these volcanic rocks corresponds with the documented displacement history of the San Andreas Fault.

A compilation of 417 isotopic dates on mid and late Cenozoic igneous rocks from the southwestern United States shows that volcanism migrated northward with time. The principal locus of volcanism at any given time was an east-west band that corresponded closely with the calculated position of the subducted Mendocino fracture zone (MFZ) under the North American plate. This correspondence supports the theory that volcanism was triggered by subduction of the MFZ, which was a major (1 km) north-facing topographic step in the Farallon plate. Both volcanism and the MFZ moved northward at about 3.1 cm/year. Andesites and rhyolites show close correspondence to MFZ passage, but many basalts were erupted significantly later. Cooling dates on basement rocks in southern Arizona cluster at the time the MFZ passed under that area. Earlier models proposing rapid late Tertiary steepening of the subducted Farallon plate were based on a westward sweep of volcanism. Our compilation shows no evidence for such a westward sweep.

During the Oligocene epoch, California was marked by extensive nonmarine sedimentation, in contrast to its pre-Oligocene and post-Oligocene depositional history. The Oligocene continental deposits are especially widespread in southern California and fill a number of small and generally partly restricted basins. Fluvial facies in many basins prograded over previously deposited lower Tertiary turbidites. Volcanism, from widespread centers, was associated with the nonmarine sedimentation. However, some basins remained marine and a few contain Oligocene turbidites and pelagic sediments deposited at bathyal depths. The Oligocene redbeds of California do not form a post-orogenic molasse sequence comparable to the Old Red Sandstone or Alpine molasse. They are synorogenic and record local uplift of basins and surrounding source areas. Late Cretaceous to contemporary orogenesis in California has been generally characterized by the formation of small restricted basins of variable depth adjacent to small upland areas in response to strike-slip faulting. Deposition of Oligocene redbeds was associated with climatic change from warm and humid to cold and semiarid, and a global lowering of sea level. Oligocene tectonism occurred during the transition from subduction of the Farallon Plate to initiation of the modern San Andreas transform system. However, the major influence that caused uplift, formation of fault-bounded basins, and extensive redbed deposition, especially in southern California, was the approach of the Pacific—Farallon spreading ridge to the western margin of California.

The Late Cretaceous history of the northern Pacific Ocean has not been adequately deciphered, largely because a major plate reorganization occurred during the Cretaceous magnetic quiet interval. Using primary data to reconstruct plate motions from fracture zone trends and Late Cretaceous seafloor spreading magnetic anomalies allows formulation of a reasonable sequence of events that accounts for all the geologic features of that region, especially the Emperor and Chinook troughs. The primary event in our reconstruction is the subduction of the old northwest Pacific triple junction. New relative plate motions imposed by formation of convergent boundaries along both the northern Pacific and Farallon plates caused the Farallon plate to crack. This subdivision occurred 82 m.y. ago and resulted in the formation of the Kula and Chinook plates. The Chinook plate was bounded on the north by the Chinook-Kula ridge, the western arm of the Great Magnetic Bight, on the west by the southern Emperor trough, a slowly spreading rift valley, on the south by the Mendocino transform, and on the east by the Chinook-Farallon ridge. The northern Emperor trough formed the initial western boundary of the Kula plate. From 82 m.y. ago to 50 m.y. ago this configuration was stable: the Chinook plate expanded, the Chinook-Kula ridge migrated north, and the Kula plate was subducted. With the subduction of this ridge, the Chinook plate adhered to the Pacific plate and the Pacific-Farallon ridge became the north-south-trending feature reflected by anomalies 22 and younger. This reconstruction obviates the awkward Late Cretaceous shifting of triple junctions found in other proposed histories and provides reasonable explanations for the several structural features of the region. To achieve closure of vector triangles representing the relative Late Cretaceous plate motions some extension must occur, presumably across the eastern Mendocino fracture zone, between the northern and southern portions of the Farallon plate. These two plates apparently were independent units between 105 and 50 m.y. ago; the initiation and cessation of rapid North Farallon-North American convergence 82 and 50 m.y. ago may have determined the nature and duration of the Laramide orogeny.

Ages of late Mesozoic–early Cainozoic igneous rocks in south-western North America indicate a magmatic arc initiated near the continental margin, swept over 1,000 km northeastward, then swept back, all in 110 Myr. The sweep in and return are interpreted as due to flattening of a Benioff zone to < 15° followed by rapid collapse.

The results of a recent large scale three-dimensional study of P velocity beneath North America are analyzed from the point of view of the search for the fossil Farallon plate in the mantle beneath the western edge of the North American continent. The large scale velocity structure obtained in the western U.S. is compatible with independent finer scale seismic studies and suggests that the Farallon plate can be traced to depths of 300 to 450 km along the entire North-South extent of the area. At greater depth, no clear evidence is found for its presence: it has either not reached it or is no longer distinguishable from the surrounding mantle. On the basis of the 3-D velocity study we also venture to predict that, whereas in the central and southern part of the western U.S. the top of the plate seems to have sunk to depths in excess of 200 km, it should still be present at shallower depths beneath the Northwestern region. This is in agreement with the progressive migration to the North of the Mendocino transform-transform-trench triple junction and its present day position.

Spreading along the Cocos-Nazca plate boundary since the breakup of the Farallon plate in the Miocene has resulted in the formation of the Panama basin and a complex interaction of plates in and near northwestern South America. Current plate boundaries have been defined, and segments of subducted lithosphere identified through selection of hypocentral locations of earthquakes, considering only well-located events, and through focal mechanism determinations. The existence of relict plate boundaries, bathymetric features, and the Panamanian isthmus has affected the subduction process of the Nazca plate beneath South America and determined the present-day configuration of the subducting lithospheric plate. There is no single triple junction separating the Caribbean, South American, and Nazca plates. Instead, the Panamanian isthmus and surrounding areas are accommodating east-west compression (and a lesser degree of north-south compression) along a series of thrust faults striking NW to NE, and the Andean ranges of Ecuador, Colombia, and Venezeula are moving as a block NNE relative to the rest of the South American plates, along a system of faults following the front of the Eastern Cordillera. The subducted portions of the Panama basin and old Farallon plate have become segmented into three pieces recognized in this study. From north to south, they are (1) a ‘Bucaramanga’ segment continuous with the Caribbean seafloor northwest of Colombia, (2) a ‘Cauca’ segment continuous with oceanic crust (Nazca plate) currently being subducted beneath South America at the Colombia-Ecuador trench, and (3) an ‘Ecuador’ segment at the northern end of the subducted lithospheric plate which is dipping at a small angle to the east beneath northern Peru. The segmentation of the subducted plate can be explained by the buoyancy of bathymetric features which have been partially subducted.

Sedimentary rocks of late Eocene to early Oligocene age from the Aleutian Islands, have been paleomagnetically studied to determine the paleolatitude of this island arc. Samples were collected from gently dipping, thin-bedded, laminated siltstone, mudstone, and very fine grained sandstone of marine origin. After demagnetization, samples from each locality showed both polarities of a remanent magnetization that thermal, isothermal remanent magnetization, and Js-T experiments suggest is carried by magnetite. After tectonic correction, the α95 decreases in both localities, suggesting magnetization was acquired before folding of the beds. The characteristic magnetization shows that no significant poleward motion of the Aleutian arc has occurred with respect to the North American plate since the early Oligocene (Amlia, F =−4 °, ΔF=5 °; Umnak, F = 0.8 °, ΔF=3.5 °), but the declinations are rotated significantly clockwise (Amlia, R=70 °, ΔR=23 °; Umnak, R=36.7 °, ΔR=11.2 ° ). Using these paleomagnetic data, relevant Euler poles, and hotspot locations, we propose a new model for the formation of the Zodiac fan. In this model, the start of Zodiac fan deposition corresponds to accretionary, igneous, and erosional events in the Pacific Northwest and fragmentation of the Farallon plate along the margin of the North American plate into small plates. Fan deposition on the Pacific plate was controlled by the geometry of the Bowie and Cobb seamount chains, which prevented sediment from being dispersed over the bathymetrically lower oceanic crust to the west, and by the bathymetry of the Pacific plate, which focused sedimentation in the magnetic bight region of the plate. Our model suggests that the present-day Zodiac fan is only a remnant of a much larger fan system.

A slablike high-velocity region 1200–1600 km beneath the east coast of North America apparently has a half-width of 450 km and a lateral extent of 3000 km and a depth extent of more than 800 km. This slab lies beneath the boundary where the Farallon Plate had been subducting for up to 120 MA. The high-velocity anomaly has been inferred previously from travel-time studies; this study uses body-waveforms to verify and refine its structure. The anomalous region has a peak shear wave velocity 1.5% greater than the surrounding region and compressional velocity 0.6% greater. The width and smoothness of this anomaly is constrained by the complexity observed in SH waveforms from deep earthquakes beneath South America observed at GDSN stations in North America. This slab is about 5 times broader and 5 times weaker in its velocity anomaly than most slabs observed in the upper mantle. This observation implies that if this slab has passed through the 650 km discontinuity, which is the most likely explanation for the observation, it has been spread by more than simply thermal conductivity. The possibility that there is two-layer convection with thermal coupling is not ruled out, however, as thermal coupling could lead to a similar pattern.

An “extensional orogeny” deformed the Basin and Range province, probably beginning in the late Eocene (about 40 ± 3 Ma). Its characteristics include partial melting of the continental lithosphere during the “ignimbrite flareup,” massive ductile extension (including detachment faulting), and rise of metamorphic core complexes. The affected zone became about 1200 km wide, possibly double its original width. It rose an average of 1–2 km, despite crustal thinning. Locally, some of the highest mountains of North America, up to 4.3 km high, rose through resurgence of ignimbrite cauldrons and isostatic uplift of underlying plutons. The climax of extension occurred prior to the development of the present basin and range topography. Modeling of major and trace elements and Sr and Pb isotopes strongly suggests that mid-Tertiary volcanic magmas equilibrated, and probably originated, in the continental lithosphere. Components attributable to subducted oceanic lithosphere have not yet been identified. The rocks seem to belong to two provinces, separated by the quartz diorite boundary line of Moore (1959), which also marks the western limit of North America at the end of the late Paleozoic Sonoman orogeny. To the west, low-K rocks rest on a basement of predominantly oceanic accreted terranes; to the east, high-K rocks rest on an autochthonous sialic basement. Within the high-K province, potassium variations can be correlated with crustal thickness; there is no need to invoke a K-h relationship. Conventional models of plate convergence and back arc extension which involve subduction of old, rigid, cool, and dense oceanic lithosphere may not apply to the mid-Tertiary Basin and Range province. The overridden Farallon plate is more likely to have been young, hot, ductile, buoyant, and no denser than continental asthenosphere, having been generated in a spreading center close to North America. Under these conditions, motion of the subducting plate slows and slab-pull is likely to approach zero. Even prior to ridge-trench collision, overridden oceanic lithosphere may have become underplated beneath the continental lithosphere and ruptured by rising mantle diapirs. Subducted oceanic lithosphere no longer acted as a heat sink, which could partly account for the great width of the affected zone and the anomalous thermal gradients required for partial melting, extension, and metamorphism. Had these processes not died down, after ridge-trench collision, the western segment of the Cordillera might have separated from North America to form a Japanlike archipelago, while the Basin and Range province foundered into an analog to the Sea of Japan. Instead of rupturing completely, the Basin and Range province fractured into fault blocks.

An attempt is made to calculate the total volume of Middle Cenozoic to Recent igneous rocks in the Columbia—Snake River—Yellowstone region by adding geophysically-derived estimates of the amounts of mafic intrusion to published data on the volumes of volcanic rocks. Recently published radiometric ages allow comparison of rates of basaltic magmatism in these ensialic provinces with those along equivalent lengths of mid-oceanic ridge. The rates of total magmatic activity (extrusive and intrusive) during the genesis of the Columbia River Plateau, Western Snake River Plain and Eastern Snake River Plain—Yellowstone sub-provinces were equivalent to those of mid-oceanic ridges with spreading half-rates of 10, 3 and 0.3 cm per year, respectively. It is shown that the hypothesis that magmatism in this region is the product of a deep-mantle convective plume, situated at present beneath Yellowstone, does not explain adequately several features of the Snake River Plain and becomes untenable when applied to the province as a whole. It is proposed instead that all major post-Eocene tectonic and magmatic features of the western U.S.A., east of the Sierra Nevada and Cascades, are products of comparatively shallow-rooted diapiric upwelling in the mantle; triggered at about 40 m.y. by extraction of mafic silicate melt and volatiles from the Farallon lithospheric plate, subducted at a comparatively low angle beneath the North American plate. During the Oligocene the diapirism appears to have become concentrated beneath the present sites of the Great Basin and Columbia River Plateau. Subsequent annihilation of the Farallon plate by the northward-propagating San Andreas transform fault released the compressive stress field across the Great Basin diapir, so that its heat content could disperse largely through attenuation and melting of the sialic crust, rather than mantle partial fusion and basaltic magmatism. Conversely, constriction of the Columbia River Plateau diapir by the subduction zone to its west led to “run away” mantle fusion and massive production of basic magmas. At 10–13 m.y. the E-W zone of offset at 42–44°N between these diapirs was the sub-sialic analogue of a transform fault. Lateral shearing of uppermantle peridotite within this zone caused the partial fusion which was the source of Western Snake River Plain magmatism. Once established, this upper-mantle thermal disturbance became self-perpetuating and, as the North American plate drifted westward over it, generated the Eastern Snake River Plain—Yellowstone “hot spot track”. The constant position of this hot spot in the mantle, relative to that beneath Hawaii, during the last 10 m.y. or so may indicate that its roots now penetrate down into the mesosphere.

The Rocky Mountain foreland and Great Plains of the western United States were formerly part of a continental platform, adjusted by erosion and deposition in Cambrian through Jurassic time to near mean sea level, and therefore to a near-uniform crustal thickness of approximately 33 km. Today the region stands at regional elevations up to 2 km, isostatically supported by a crust exceeding 50 km in thickness. Reasonable estimates of Tertiary sedimentation and Laramide strain do not account for more than 15% of the implied thickening. However, from approximately 70-40 m.y. B.P., this region was underlain by a horizontally-subducting slab of Farallon plate lithosphere moving northeast; this slab could have been the cause of the thickening. Finite-difference thermal models with specified kinematics show only a temporary cooling of the base of the North American lithosphere by this slab. The excess weight of the slab would have depressed the region; plate-bending calculations show a quantitative agreement of predicted depression with upper Cretaceous isopachs. Since depression by this slab lasted until the Eocene at least, the latest-Cretaceous regression was probably caused by a Laramide crustal thickening event. The Farallon slab might have caused crustal thickening in two ways. Its excess weight would have drawn in ductile lower crust from surrounding regions. However, calculations show that this effect is too slow, too local, and too reversible to explain most of the crustal thickening. Therefore it seems likely that ductile lower crust was transported from SW to NE by shear stresses which the Farallon plate exerted on the base of the North American lithosphere. A preliminary finite-element calculation based on this hypothesis shows the correct general pattern of crustal thickening. An unexpected but encouraging result is that predicted principal compression directions are orthogonal to many Laramide basement uplifts.

At subduction zones that have only recently ceased to be active, the lithospheric slab may retain a seismic velocity greater than that of the surrounding mantle even after the slab becomes seismically dead. To seek the subduction zone thought to have been recently active along the western margin of North America, we examined the variation with propagation direction of P-wave travel time residuals from sources at various distances and azimuths to seismograph stations in Washington and California. The uncertainty in source location and origin time was removed by referring the travel-time delay to a nearby station overlying presumably more uniform mantle. An eastward-dipping band of anomalously early arrivals at several stations on the western flank of the Sierra Nevada and California Cascades may imply that a dead slab is present beneath northern California, though a definitive conclusion is premature at present because of a paucity of seismic sources in eastern North America. The position of the dead slab speculatively suggested by the travel-time data is roughly consistent with that predicted by others on the basis of heat flow and geochemistry in the Sierra Nevada, and the southward decrease in the magnitude of the travel-time advance associated with such a slab is in agreement with the history of subduction of the Farallon plate as reconstructed from ocean floor magnetic anomalies and continental tectonic activity.

In the absence of convincing magnetic anomaly information, topographic profiles have been used to infer the tectonic history of the East Pacific Rise between 5°S and 20°S. Profiles projected at right angles to the rise crest show a sharp drop in elevation at roughly the same distance on either side of the crest. Profiles to the east of the rise also show a second topographic high at 95°W. Comparison of these profiles with empirical depth versus age curves for the North Pacific suggests that this rise, the Galapagos Rise, is the fossil East Pacific Rise which terminated close to 6 mybp by the spreading center jumping 900 km to the west. The extreme youthfulness of the present East Pacific Rise, the step structure of its flanks, and the similarity in age of the top of this step and the crest of the Galapagos Rise substantiate this interpretation. This jump coincided with a similar readjustment involving the Mathematicians ridge at 5° to 20°N and the opening of the Gulf of California. This wholesale realignment of the active spreading centers in the eastern Pacific from 20°N to 20°S may have arisen because of the break-up of the Farallon Plate into several smaller plates and/or the complex interaction of the East Pacific Rise with the North American continent.

The Juan de Fuca, Rivera, Cocos and Nazca plates in the eastern Pacific are small fragments remaining from the fragmentation and partial subduction of the long, narrow Farallon Plate which began as early as 55 Myr ago1. This fragmentation apparently resulted from interaction of the very irregular Pacific–Farallon ridge with the uneven coastlines of the American Plate1–4 and involved the formation of ridge–trench transforms5, creation of new spreading centres6, ridge jumps1,7,8, deactivation of subduction zones2 and other phenomena1. We present here magnetic evidence that a fifth small fragment of the Farallon plate is preserved in the eastern Pacific off the coast of Baja, California. Its preservation is the result of an eastward jump of the Pacific–Farallon ridge ∼12.5 Myr ago which was preceded by asymmetric spreading and perhaps pivotal subduction1. Menard1 first suggested this idea and showed that anomaly 5B is symmetric and parallel to a linear trough trending southwards from Guadalupe Island. He named the small fragment of the Farallon plate preserved between this trough and the margin of Baja California to the east, the Guadalupe plate1. We show that both anomalies 5B and 5C symmetrically flank the Guadalupe trough. Our results and conclusions are not based on new magnetic data, but rather on reinterpretation of aeromagnetic profiles published by Taylor et al.9.

At Laytonville, California, about 230 km north-northwest of San Francisco, three blocks of pelagic limestone, each a few tens of metres long, are incorporated in the Franciscan melange. Bedding is well defined, but top indicators are lacking. Paleomagnetic study of two of the blocks (16 samples, 52 specimens) yielded directions of D: 183.9°, I: −22.6°, α_95: 21.6° (block 1), and D: 229.5°, I: −31.8°, α_95: 10.7° (block 2) relative to the present orientation of bedding. Inclinations were not significantly different, but the difference in declinations shows that magnetization preceded emplacement of the blocks in the melange. Study of foraminifera showed that the blocks are of Albian and Cenomanian age and were deposited during the Cretaceous Long Normal Polarity Interval. Details of the foraminiferal zonation show that both blocks are right side up. The observed inclinations imply deposition on the Farallon plate at 17° ± 7° South paleolatitude at about 96 m.y. B.P. (block 2; block 1 data agree but are less definitive). Consideration of reasonable paleolongitudes and possible times of incorporation of the blocks in the melange indicates a Farallon–North America convergence rate significantly higher than any observed today. The most commonly accepted emplacement time (latest Cretaceous) indicates convergence at about 38 cm/yr (range: 24 to 60 cm/yr). An alternate interpretation, that emplacement occurred at 30 ± 15 m.y. B.P., would indicate convergence at about 15 cm/yr (range: 9 to 24 cm/yr). The convergence rates obtained in the first case are astonishingly high, and they suggest that perhaps current ideas on the Franciscan should be re-evaluated, with consideration given to the possibility that tectonic mixing continued until the middle or late Tertiary. Even so, the Farallon–North American convergence rate was apparently higher than any convergence rate observed today. This rapid convergence coincides with the Cretaceous rapid spreading pulse, and it may be responsible for such features as the large volumes of Late Cretaceous batholiths and subduction complexes in western North America.

The dominance of volcanic processes and the importance of vertical tectonics in the geological evolution of the Pacific Basin has been recognised since the time of Charles Darwin. Data gathered on several legs of the Deep Sea Drilling Project (DSDP) and numerous marine expeditions in the past decade have confirmed Menard's postulate that the Pacific Basin was the scene of volcanism on an enormous scale in Mesozoic time. Widespread mid-plate volcanism between ∼110 and 70 m.y. B.P. characterised the area bounded by the Line Islands, the Mid-Pacific Mountains and the Nauru Basin-Marshall Islands. Heating of the Pacific lithospheric plate during this period of volcanism resulted in regional uplift and the bathymetric evolution of the area diverged significantly from a “normal” Parsons-Sclater subsidence curve. The Farallon plate, now almost entirely subducted, was also the scene of mid-plate volcanism that produced such features as the Nicoya Plateau now found as an allochthonous ophiolitic terrain landward of the middle America trench. Large, benthonic, reef-associated foraminifera comprising a pseudorbitoid fauna, hitherto considered to be largely restricted to Central America, have now been additionally recorded from DSDP Sites 165, 315 and 316 in the Line Islands, Site 462 in the Nauru Basin, and in New Guinea. The distribution of this fauna, of Campanian/Maastrichtian age, is interpreted as indicating “stepping stone” connections (aseismic ridges, plateaus and seamounts) between the Caribbean, Farallon, and Pacific plates 70–80 m.y. B.P. Similarities between the geology of the Nauru Basin and the Caribbean Ocean crust reinforce the interpretation of the latter as a former part of the Farallon plate. Estimates of the sea-level and continental freeboard change caused by the thermally induced uplift of the Pacific and Farallon plates, as well as substantial areas in the Atlantic and Indian Ocean Basins, indicate that such shallowing of the sea floor could have been the major factor in causing global Cretaceous transgressions.

More than 90% of the known petroleum accumulations west of the San Andreas fault in California are in strata deposited in areally restricted Neogene basins that formed during a major tectonic reorganization of western California. These deep, localized Neogene basins replaced broad, regionally persistent Paleogene depositional aprons, although some of the Neogene basins in northern and central California had Paleogene precursors. The evolution of each of the Neogene basins is complex, and aspects of the kinematics of each are unique; nonetheless, all can be considered products of an overall right-lateral shear system associated with a sliding margin between the Pacific and North American lithospheric plates. The sliding margin developed in western California about 29 m.y. ago, when the Pacific plate contacted North America after subduction of the intervening Farallon plate. The initial position of the common boundary between the Pacific and North American plates was along the continental margin. Right slip between the Pacific and North American plates gradually shifted eastward to right-slip faults, such as the San Andreas, located farther inland. This shift seems to be documented by relations in the southern California area. About 300 km of right slip has occurred along the San Andreas fault during the past 10 to 15 m.y., and at least several hundred additional kilometers along associated right-slip faults of the San Andreas system. The Neogene basins in southern California began to develop during the interval in which the boundary between the Pacific and North American plates shifted from the continental edge to the San Andreas fault, apparently because the step-by-step switch to different surfaces of weakness caused local extension and compression within a broad zone of right-lateral shear. A major phase of basin formation appears to have been synchronous with a change in azimuth of relative shear between the Pacific and North American plates to a more westerly direction, resulting in extensional strain. This change in motion initiated basin development in offshore central and northern California and affected the ongoing development of basins as a result of right slip along the San Andreas and related faults in other parts of California.

Using finite element methods and a reconstruction of regional plate boundaries we calculated the state of stress in the Farallon plate at 30 m.y.b.p. A key-role in the model is played by the dependence of the driving forces (slab pull and ridge push) on the age of the oceanic lithosphere involved. The results show a highly tensional stress field which we consider to have been the cause for the initiation of spreading along the Galapagos (or Cocos-Nazca) spreading centre and the fragmentation of the Farallon plate into the Cocos plate and the Nazca plate, which took place approximately 30 m.y.b.p.

The Laramide orogeny (80-40 Ma) in the southwestern United States is usually thought of as a single tectonic event with attention concentrated on its early stage because of its dramatic expression in the sedimentary record. The sharpest pulse of deformation, however, occurred in the latest Paleocene-early Eocene and was separated from the early stage by a tectonic lull and development of widespread lateritic weathering profiles, remnants of which are preserved in some early Tertiary basins. The first stage correlates plate tectonically with opening of the North Atlantic and Labrador Sea, the second stage with opening of the Norwegian Sea and Eurasian basin in the eastern Arctic. Rapid convergence between the North American and Farallon plates decreased the dip of the subd cting Farallon plate, until by early Eocene (55 Ma), strong viscous coupling was occurring between the Farallon plate and the overlying lithosphere. Change in direction of convergence from west-southwest-east-northeast during the early stage to southwest-northeast during the late stage brought the first-order shear direction into near parallelism with the north-northeast-trending southern Rocky Mountain deformed belt in New Mexico. This parallelism allowed the Colorado Plateau to decouple from the craton along right-lateral wrench faults. The Colorado Plateau was translated 100-130 km north-northeast with a comparable shortening across the Wyoming province. Both sedimentation rates and coarseness of synorogenic sediments increased dramatically from Wyoming to the Gulf Coast of Texas beginning in the early Eocene. Simultaneously, an en echelon series of strike-slip basins formed along the zone of decoupling from southern New Mexico to southern Wyoming. Deformation in the Wyoming province also increased sharply, resulting in as much as 21 km of overhang on range-front thrusts and up to 15 km of structural relief between adjacent uplifts and basins. As much as 2,500 m of Eocene orogenic sediments were deposited in rapidly subsiding basins. Deformation was so rapid that surface drainage was disrupted and runoff was impounded in huge lakes in which as much as 1,000 m of lacustrine sediments accumulated to form the oil shale province. Wrench faulting along the eastern margin of the Colorado Plateau was distributed across a 100-km wide belt that followed zones of weakness inherited from Precambrian and late Paleozoic deformation. Wrenching changed from nearly parallel in New Mexico to strongly convergent in Colorado and southern Wyoming because of a change in structural grain from north-northeast in New Mexico to north-northwest in Colorado. At least 25 km of shortening across the zone of decoupling during convergent wrenching caused conspicuous low-angle thrusting, which has tended to mask the lateral component of movement. Major north-northeast shortening across the Wyoming province and a lack of such shortening on the High Plains demand right-lateral wrench faulting. Offset of transverse aeromagnetic anomalies ac oss the zone of decoupling in Colorado and New Mexico, offset of Precambrian metavolcanic belts and age province boundaries, and offset of northeast-trending lineaments and distinctive rock types provide independent measures of the magnitude of right slip. Northward translation of the Colorado Plateau was a relatively minor part of regional right-lateral shear that extended through much of the North American cordillera in the early Tertiary. Recognition of the wrench-faulted nature of the eastern margin of the Colorado Plateau and Wyoming province creates new opportunities for petroleum exploration. These can be divided into: (1) subthrust plays beneath low-angle faults caused by convergent wrenching; (2) structural, stratigraphic, and fracture-controlled traps along wrench faults; and (3) fracture reservoirs along northeast-trending lineaments dilated during wrenching. Recognition of wrench faults paralleling the structural grain of the southern Rocky Mountains also means that isopach and facies maps drawn smoothly across the zone of decoupling should be reevaluated with careful attention to control points. End_of_Article - Last_Page 2044

From abstract: The Geysers-Clear Lake geothermal area lies within the central belt of the Franciscan assemblage in northern California. The structure of this terrane is characterized by northeast-dipping imbricate thrust slices that have been warped and cut by steeply dipping strike-slip and normal faults. Introduction of magma into the crust beneath the Geysers-Clear Lake area can be related to eastsoutheast extension accompanying northward propagation of the San Andreas transform system between the Clear Lake region and Cape Mendocino within the last 3 million years. The initiation of strike-slip faulting during this time terminated subduction of elements of the Farallon plate beneath North America as strike-slip motion was taken up along the Pacific-North American plate boundary. The mechanism for magma generation appears to require a heat source in the mantle that mixed mantle-derived melts with various crustal rocks. These crustal rocks may have included the Franciscan central and coastal belts, ophiolite, Great Valley sequence, and possibly middle and late Tertiary rocks subducted before initiation of strike-slip faulting.

Exotic, purple-red rhyolitic clasts are abundant constituents in upper Paleocene and Eocene conglomerates in the California Continental Borderland. Four families of rhyolitic rocks have been recognized. Owl Creek clasts have low percentages of phenocrysts that never include quartz; they are primarily found in northern localities such as the Santa Ana Mountains. Poway rhyolites are packed with phenocrysts and invariably include quartz; they dominate conglomerates in the San Diego and northern Channel Islands areas. Las Palmas clasts have a distinctive micropore-microgranular groundmass; they abound in Baja California strata. Black Rhyodacites are much darker and more brittle and were derived from a closer source terrane of bedrock and Upper Cretaceous conglomerates; they are most abund nt in the earlier deposited conglomerates of the San Diego and northern Channel Islands areas. The striking similarities between San Diego and northern Channel Islands conglomerates demonstrate that they were once part of an integrated, east-west-oriented depositional system that acts as a piercing point to document major post-Eocene strike slip in the offshore California Continental Borderland. The sudden appearance of exotic rhyolitic clasts in mixed suites in upper Paleocene conglomerates, and the subsequent increases in their abundance and decreases in their variety testify to the growth of fewer, longer, and less overlapping rivers through the Eocene. The requisite integration of fluvial drainages followed the eastward-moving front of the Laramide Orogeny, caused by a shallowing ang e of Farallon Plate subduction.

The Calera Limestone, part of the Franciscan Complex of northern California, may have formed in a palaeoenvironment similar to Hess and Shatsky Rises of the present north-west Pacific1. We report here new palaeomagnetic results, palaeontological data and recent plate-motion models that reinforce this assertion. The Calera Limestone may have formed on Farallon Plate plateaus, north of the Pacific–Farallon spreading centre as a counterpart to Hess or Shatsky Rises. In one model2, the plateaus were formed by hotspots close to the Farallon–Pacific ridge axis. On accretion to North America, plateau dissection in the late Cretaceous to Eocene (50–70 Myr) could explain the occurrence of large volumes of pillow basalt and exotic blocks of limestone in the Franciscan Complex. Partial subduction of the plateaus could have contributed to Laramide (70–40 Myr) compressional events3.

Large clockwise rotations (15–80°) are characteristic of Cenozoic volcanic and sedimentary rocks along the convergent margin of the northwestern United States. Abundant paleomagnetic data from 62–12 m.y. old rocks in forearc, arc, and backarc regions show that rotation increases with age and with proximity to the coast. Paleomagnetic and structural studies both support dextral shear as a significant contributor to tectonic rotation in the Pacific Northwest, with an added contribution from Basin-Range extension. Paleomagnetism of individual Miocene (15-12 Ma) flows of the Columbia River Basalt Group show a well-defined, progressive increase in rotation across the forearc region toward the trench which is most readily explained by shear. This progressive increase in rotation toward the coast is also seen in Oligocene and Eocene rocks, although with steeper gradients, indicating that shear rotation has been important throughout Cenozoic time. The dextral shear may be accommodated by abundant, small-scale strike-slip faults which are acting as Riedel shears. Coast Range faults, usually with well-developed subhorizontal slickensides, occur throughout the stratigraphie section and are most abundant in the oldest rocks. Displacements are small to moderate (102–103 m), and both dextral and sinistral faults are common. Fault trends are variable, with N20°–45°W faults showing dextral slip and N60°W-S70°W faults showing sinistral slip. The driving force for dextral shear presumably results from coupling of the overlying plate with the subducting Farallon plate, which has been moving northeast throughout most of Cenozoic time. The increase in rotation from north to south along the coast probably represents the contribution of extension in the Basin- Range region to the rotations in Oregon.

The existence of an intrinsic depleted component in mantle plumes has previously been proposed for several hotspots in the Pacific, Atlantic, and Indian Oceans. However, formation of these depleted basalts is often associated with unusual tectonomagmatic processes such as plume-ridge interaction or multistage melting at plume initiation, where depleted basalts could reflect entrainment and melting of depleted upper mantle. Late Cretaceous to middle Eocene seamounts that accreted in Costa Rica and are part of the early Galapagos hotspot track provide new insights into the occurrence and nature of intrinsic depleted components. The Paleocene (ca. 62 Ma) seamounts include unusually depleted basalts that erupted on the Farallon plate far from a mid-ocean ridge. These basalts closely resemble Gorgona komatiites in terms of trace element and radiogenic isotope composition, suggesting formation from a similar, refractory mantle source. We suggest that this source may be common to plumes, but is only rarely sampled due to excessive extents of melting required to extract melts from the most refractory parts of a heterogeneous mantle plume.

Plate reconstructions since the breakup of Pangaea are mostly based on the preserved spreading history of ocean basins, within absolute reference frames that are constrained by a combination of age‐progressive hotspot tracks and palaeomagnetic data. The evolution of destructive plate margins is difficult to constrain from surface observations as much of the evidence has been subducted. Seismic tomography can directly constrain palaeo‐trench locations by imaging subducted lithosphere in the mantle. This new evidence, combined with the geological surface record of subduction, suggests that several intra‐oceanic arcs existed between the Farallon Ocean and North America during late Mesozoic times – in contrast to existing quantitative models that typically show long‐lived subduction of the Farallon plate beneath the continental margin. We present a continuously closing plate model for the eastern Pacific basin from 170 Ma to present, constrained using ‘tomotectonic analysis’ – the integration of surface and subsurface data. During the Middle to Late Jurassic, we show simultaneous eastward and westward subduction of oceanic plates under an archipelago composed of Cordilleran arc terranes. As North America drifts westward, it diachronously overrides the archipelago and its arcs, beginning in the latest Jurassic. During and post‐accretion, Cordilleran terranes are translated thousands of kilometers along the continental margin, as constrained by palaeomagnetic evidence. Final accretions to North America occur during the Eocene, ending ~100 million years of archipelago override. This model provides a detailed, quantitative tectonic history for the eastern Pacific domain, paving the way for tomotectonic analysis to be used in other palaeo‐oceanic regions.

Se integro el marco estructural y el modelo tectonico de la Planicie Costera de Todos Santos, utilizando un conjunto de datos gravimetricos y magneticos tomados entre 1962 y 2000 en tierra, aire y mar. En el presente trabajo se integro toda la informacion disponible, se homogeneizo y se mantuvo a su altitud original. Se empleo una nueva tecnica de inversion conjunta de datos gravimetricos y magneticos para estructuras tridimensionales. Como resultado se obtuvo un mapa topografico del basamento granitico-andesitico, donde se visualizan rasgos geologicos desconocidos hasta este trabajo. Se hicieron algunas secciones a partir del modelo tridimensional, en los que se establece el patron estructural de la Planicie Costera de Todos Santos, el cual es tipico de una cuenca extensional (pull-apart basin) asociada con un frente de arco en un ambiente tectonico durante la subduccion de la placa Farallon a final del Mesozoico y principios del Cenozoico. A set of gravity and magnetic data taken within the Todos Santos Coastal Plain between 1962 and 2000 from land, sea and air is used to integrate the structural framework and tectonic model of this area. The present work collected all the available data, homogenized them, and maintained them at their original altitude. A new technique of joint inversion for gravity and magnetic data from three-dimensional structures was applied. As a result, a well-defined topographic map of the granitic-andesitic basement was obtained and previously unknown geological structures were depicted. Several cross-sections were made from the three-dimensional model obtained, establishing the Todos Santos Coastal Plain structural pattern as a typical pull-apart basin associated with a tectonic fore-arc during the subduction of the Farallon Plate in Late Mesozoic and Early Cenozoic time.

MUCH of the late Cainozoic history of the western United States is characterized by tensional rifting. The Basin and Range Province is a zone of rifting 300 to 600 km wide (Fig. 1). The Rio Grande depression and the Snake River downwarp are areas of rifting which are similar in time and style to the Basin and Range faulting. The rifting in the Rio Grande depression has usually been related to the same causative mechanism as in the Basin and Range. The Basin and Range type faulting has variously been explained as the result of a westward moving convection cell1; as part of a wide, soft boundary reflecting the transform motion of the American and Pacific plates2; or as the result of a large mantle diapir which came up under the crust following the complete subduction of the Farallon plate in that region3. Although most of these explanations adequately account for the Basin and Range faulting, they fall short in explaining the rifting in the Rio Grande depression. None of the three hypotheses adequately explains why extensive rifting occurs over a area 600 km wide in the Basin and Range abruptly stops for a length of 500 km in the Colorado Plateau area and then resumes again in a narrow zone in the Rio Grande region. The Snake River rifting transects the regional trend of the Basin and Range rifting and does not fit any of the proposed tectonic models for the Basin and Range.

A minimum of 75 km of right-lateral fault displacement occurred within the Franciscan assemblage of northern California, most likely in late Tertiary time. This displacement is inferred from the present position of two geologically unrelated units, the rocks near Black Lassie, California, and the Redwood Creek schist. The Black Lassie rocks show at least 75 km of northward displacement from their source terrane within the Coast Range ophiolite. Although no conclusive evidence of northward transport for the Redwood Creek schist is available, the geologic setting of the schist and the nature of the faults that bound it strongly suggest that the Redwood Creek schist has also been laterally transported many tens of kilometres northward. We propose that both these units moved northward by means of one or a few major north-northwest–trending right-lateral faults that were active in the forearc basin off the California coast during right-oblique subduction of the Farallon plate under the North American plate in late Tertiary time.

The Cretaceous Quiet Period ended with a significant reorganization of the plate tectonics of the North Pacific. The Farallon plate broke into three primary pieces: (1) the Kula plate, (2) a large fragment transferred to the Pacific plate, and (3) the continued Farallon plate. The time of the rearrangement is constrained to just before magnetic anomaly 34 by its presence, contrasts in tectonic spreading fabric (TSF), the Emperor Trough, and the Chinook Trough. Our model is based on (1) a complete reevaluation of primary bathymetric and magnetic data, (2) the distribution of major troughs (Emperor, Chinook, Hokkaido, and Krusenstern), (3) mapping and analysis of TSF orientation and amplitude on Quiet Zone crust, and (4) available Deep Sea Drilling Project drill hole data. East-west trending magnetic reversal lineations and TSF south of the Aleutian Trench are bounded on the west by the Buldir Fracture Zone at 175°E. The major troughs mark the abandonment and initiation of spreading in the reorganization; their absence elsewhere precludes other long-distance boundary relocations. Emperor Trough is the abandoned Pacific-Farallon spreading center. Rift initiation and early slow spreading have created the high amplitude topography of the Chinook Trough. We believe that the Hokkaido Trough is the initiation site of a failed oceanic rift. The incorporation of large amounts of Farallon plate into the Pacific plate obviates the need for the very rapid Pacific-Farallon spreading rates of previous models and significantly alters the interpretation of mid-Pacific geology.

To investigate the mechanism underlying the break-up of the Farallon plate into the Cocos plate and Nazca plate, we analyze the state of stress in the Farallon plate at about 30 m.y. B.P., just prior to the fragmentation. To this purpose we use finite element methods and a reconstruction of regional plate boundaries appropriate for 30 m.y. B.P. A key role in the model is played by the dependence of two important plate-tectonics forces (slab pull and ridge push) on the age of oceanic lithosphere. The results show a highly tensional stress field, with maximum principal stresses of 5 to 6 kbar. North-south tension in the vicinity of present-day Panama is proposed to have been the cause for the fragmentation of the Farallon plate and the inception of spreading along the new Cocos-Nazca plate boundary. Because the kind of plate interaction that gave rise to the high level of tensional stresses in the Farallon plate is not restricted to the area of this study, the proposed mechanism seems to shed light on the problem of fragmentation of oceanic plates in general.

East-northeast trends of folds, thrust faults, foliation, and beds of late Mesozoic age are common in a 1100-km-long belt along the west coast of Mexico. The structures indicate north-northwest–directed crustal shortening, which is parallel to the present and probable late Mesozoic continental margin. The structures are best known in southern Sinaloa, where timing constraints indicate that the deformation most likely occurred shortly before 100 Ma. Similar occurrences are in Nayarit, northern Sinaloa, Sonora, and Baja California; timing of deformation is less constrained in these areas but also appears to be mid-Cretaceous. The deformation patterns suggest compression parallel to the continental margin and plate boundary, which seems incompatible with simple perpendicular convergence. All plausible explanations require oblique convergence between the North American and cither the Farallon plate or the Kula plate along a north-northwest–trending continental margin. The most likely explanation involves transmission of oblique compression across the plate boundary, but the structures seem to require a more tangential component than motion histories indicate. Accretion by strike-slip faulting from elsewhere along the North American continental margin of a single, large block that encompasses the entire area of east-northeast structures is less likely and simply displaces the origin of these structures without explaining them. An east-trending continental margin off southwestern Mexico during the late Mesozoic is also unlikely.

A WNW trending lineated topographic depression (Magellan Trough) and associated magnetic anomaly lineations (Magellan lineation set) were revealed in the basin northeast of the Magellan Rise. We identified symmetric magnetic anomaly lineations from M9 (121 m.y.) to M11 (126 m.y.) of Early Cretaceous age with a symmetric axis at the Magellan Trough and concluded that the Magellan Trough is a remanent spreading center which ceased its opening at the latter half of M9 time. The Magellan lineation set fans drastically to the southeast, and it represents an asymmetric spreading system. The Magellan spreading system has two separate rotation poles for the northern flank lineations and the southern flank ones, and both poles lie on the extension of the Magellan Trough. The amplitudes of the magnetic anomalies decrease toward the poles, possibly owing to the contamination of the magnetized blocks as a function of decreasing spreading rate. The Magellan spreading system shows unusually rapid rotation rates of 4.2°/m.y. for the northern flank and 3.1°/m.y. for the southern flank, which together with the proximity of the rotation poles suggest unstableness of the spreading system. This unstable spreading may have shortened the life of the Magellan spreading system. Such an unstable spreading system could have been generated in order to balance changes in movements of the Phoenix, Farallon, and Pacific plates. The spreading center may have jumped to the northeast at M9 time as part of its evolution as the Pacific-Farallon ridge. Thus the Magellan Trough may be a fossil Pacific-Farallon ridge and the northern flank of the Magellan set a piece of captured Farallon plate.

To study the rigidity of the Pacific and Farallon plates between 72 and 35 Ma, we analyzed the magnetic anomalies and fracture zones in the Pacific Ocean between longitudes 125°W–180° and latitudes 43°S–60°N. A set of consistent rotations indicates that the Pacific and Farallon plates between the Pioneer and Agassiz fracture zones were rigid plates between 72 and 35 Ma. Thus the location of a missing Tertiary boundary, within the Pacific or Antarctic plates, is restricted to be south of latitude 43°S. The Farallon plate broke into two plates at the time of the change in the Pacific-Farallon spreading direction north of the Pioneer and Mendocino fracture zones, sometime between 59 and 49 Ma. The northern fragment, termed the Vancouver plate, was subducted beneath most of North America. We present rotations and uncertainties for the histories of motion of the Pacific and Kula plates from anomaly 32 to anomaly 25, of the Pacific and Vancouver plates from anomaly 21 to anomaly 13, and of the Pacific and Farallon plates from anomaly 32 to anomaly 13.

Paleomagnetic data from Cretaceous igneous and sedimentary rocks of the Peninsular Ranges and Continental Borderland of the Baja California Peninsula yield an average pole position of 83°N, 344°E (A95 = 3.4°), suggesting that the peninsula as a whole has moved northward ∼11° and rotated 32° clockwise, with respect to the North American craton, since Cretaceous time. Only a small portion of this movement is due to the Neogene opening of the Gulf of California. Results from stratified rocks giving both positive and negative fold tests are included in the average Cretaceous pole. The rocks that fail fold tests have been deformed and remagnetized by nearby intrusives that are only slightly younger; unconnected directions from these sites closely match corrected directions from sites where the fold tests are positive. Oceanic rocks were also sampled in the Continental Borderland from Upper Triassic and Upper Jurassic tectonostratigraphic terranes which are apparently allochthonous with respect to the western Baja Peninsula and each other. These oceanic rocks show magnetizations indicating low paleolatitudes, although the hemisphere of origin is presently unknown. The Cretaceous rocks of the Baja California Peninsula may have originally formed adjacent to the now-truncated continental margin along the northern Middle America Trench. Northward transport of the peninsula probably occurred by transcurrent faulting associated with oblique subduction of the Farallon plate beneath western North America between Late Cretaceous and late Miocene time.

Blocks of pelagic limestone, tectonically incorporated into coastal terrane melange, contain planktonic foraminifers of early Campanian to middle Maestrichtian age (82-69 Ma) typical of low latitudes between 20°N and 20°S. Carbonate concretions from the sheared terrigenous matrix of the melange yield temperate, middle-latitude planktonic foraminifers and dinoflagellates of middle to late Eocene age (49-41 Ma). The pelagic limestone and the basaltic substrate presumably formed on the eastern flank of the Pacific-Farallon spreading ridge in a low-latitude zone of high productivity and migrated northward on the Farallon plate until about 40 Ma when they initially encountered terrigenous turbidites north of lat. 30°N. Northward translation and final accretion were completed by about 25 Ma.

The paleomagnetic record for the western edge of North America shows consistent, systematic discordance. Virtually none of roughly three dozen high-quality paleomagnetic studies of pre-Pliocene rock units located within several hundred kilometers of the continental margin are lying anywhere near their appropriate reference poles (constructed from high-quality paleomagnetic data from the craton). Nor do these discordant paleomagnetic studies show simple random scatter, as might be expected for rock bodies whose postmagnetization histories include reheating, complex tectonic deformation, or chemical change. Instead, their paleomagnetic poles are clearly displaced systematically away from the reference curve into the general area of the Atlantic Ocean. To accomplish this tectonically requires that the magnetized rock body move northward, rotate clockwise, or both, in relation to North America. The consistency of paleomagnetic record argues that most of the western edge of North America has undergone such block movement and therefore is allochthonous, at a scale at least as large as the area of a typical paleomagnetic study. It further argues that northward transport and clockwise rotation have been prime elements in shaping the Cordillera. Differences between paleomagnetic records for separate parts of the Cordillera probably reflect differences in specific platetectonic histories. For instance, although northward transport of allochthonous blocks is important south of Cape Mendocino and north of southern Vancouver Island, in the area between (the Pacific Northwest), only clockwise rotations are found. This is consistent with steady under-thrusting of the Pacific northwest coastline by the Farallon plate for most of the Tertiary, which is in contrast to the steady or intermittent transform activity to the north and south. Batholith belts, including the Peninsular Range batholith of Southern California and the Coast Plutonic Complex of British Columbia, also seem to have been involved in the general northward transport and clockwise rotation, although the possibility of undetected tectonic tilts clouds the record. However, consideration of the thermal behavior of large, tilted plutonic blocks suggests that post-magnetization tilts probably are small. If so, then most western Cordilleran Cretaceous batholiths (not including the Sierra Nevada) also are allochthonous. Northward transport and clockwise rotation could be accomplished by a variety of mechanisms, involving recognized plate tectonic processes. Some of these mechanisms are discussed. However, dextral shear between North America and plates to the west is central to each mechanism. It seems likely that shear also has disrupted and internally modified older accreted terranes, producing internal block rotations. Such older, disrupted allochthonous terranes appear to be drawn out into highly attenuated tectonic laminae plastered on the western edge of the continent.

Twenty-two sites from two small exposures of basalt of probable Eocene age immediately east of the Olympic Mountains have the following mean paleomagnetic direction: declination, 176.5°; inclination, −66.5°; circle of 95% confidence, 8.6°. This is the only paleomagnetic study of pre-Miocene rocks anywhere in western Washington or Oregon that matches the polar wandering curve for North America. However, a marked fanning of paleomagnetic declinations around the Olympic core suggests that concordance in this case is fortuitous and results from oroclinal bending during or immediately after Olympic orogenesis. Creation of the Olympic Mountains involved accretion of thickened basaltic crust followed by underthrusting of large quantities of oceanic sedimentary debris, as the Farallon plate pressed northeastward into a recess formed by the thick crystalline blocks of Vancouver Island and the North Cascades.

As the East Pacific Rise approached and impinged upon the western margin of North America, the Farallon plate began to break into multiple smaller plates, whose relative movements changed strongly from those before breakup. Movements of the northern Cocos plate, inferred from magnetic lineations, show rapidly changing rates and pole positions which may be characteristic of other small remnants of the Farallon plate. Finite difference poles and rates for Cocos-Pacific relative motion during the time intervals 2–5, 5–7, 7–10, and 10–14 m.y. have been determined from the geometry of magnetic isochrons. Prior to about 2 m.y. the Cocos-Pacific pole was 10° to 20° south of its present location, and angular rates were 2–4 times the present rate. From 10–7 m.y., we deduce a large degree of oblique convergence between the Cocos and North American plates, which may be an important constraint on the evolution of the structure of the continental margin of southwestern Mexico. The Tehuantepec Ridge originated as the ridge-ridge transform which generated the Clipperton fracture zone. Spreading rate estimates for the southern part of the Cocos plate suggest the possibility that the Tehuantepec Ridge was reactivated as a ridge-trench transform between the two Cocos plate fragments prior to about 7 m.y.

The Franciscan formation, the bedrock beneath most of San Francisco, is now recognized as an ocean floor assemblage that was probably scraped off the Farallon Plate and its ancestors as these plates were subducted beneath North America. During subduction, the ocean floor, together with slabs of ultramafic material from deeper sources, now serpentinized, was fragmented and stacked against the western continental margin. Excavations throughout San Francisco and in southern Marin County, and sea cliffs along the Golden Gate, provide superb exposures which exemplify much that marine geologists have learned of modern ocean floor processes, both sedimentary and volcanic. Geophysicists studying the physical properties of small specimens of serpentine in the laboratory should contemplate the homogeneity of serpentine en masse and in situ in the outcroppings on Duboce Street or near Fort Point.

A number of well- to relatively well-dated significant volcanic and tectonic events along the western margin of the North American plate began, ended, or increased intensity at 16 ± 1 m.y. B.P., 10 ± 2 m.y. B.P., and 5 ± 1 m.y. B.P. Continental events at 16 ± 1 m.y. B.P. were related to the upwelling of an elongated mantle diapir, which created a widespread thermal disturbance east of the subducting Farallon plate. Oceanic events at 16 ± 1 m.y. B.P. were largely related to the approach of the East Pacific Rise toward the North American plate. Continental and oceanic events at 10 ± 2 m.y. B.P. were related to the clockwise rotation of the Pacific plate spreading direction and to the increased rate of relative motion between the Pacific and North American plates. Most events at 5 ± 1 m.y. B.P. were related to the opening of the Gulf of California and to the oceanic plate reorganizations of which the gulf-opening was a part. Different styles of response reflect differences between intraplate volcanic and tectonic subdomains. Pulses of volcanic activity do not seem to be directly or solely related to changes of plate motion. Increased volcanic activity at arcs and some intraplate volcanic settings may be indicators or precursors of significant tectonic transition intervals. If so, we may presently be in the midst of such a transition interval.

The Pacific seafloor is littered with small fragments of lithosphere captured from adjacent plates by past plate boundary reorganizations. One of the clearest examples of such a reorganization is documented in the Mathematician Seamounts region, where a distinctive geomorphology and well-developed magnetic anomalies are present. This reorganization involved a short-lived microplate between the failing Mathematician Ridge and a new propagating spreading center: the East Pacific Rise. It produced a transfer of a fragment of lithosphere from the Farallon to the Pacific plate, and also created a number of landforms and magnetic patterns, within and on the margins of the captured fragment; these make up the Mathematician paleoplate. In many cases, two sides of a microplate are active spreading ridges. A microplate evolves into a paleoplate when dual spreading ceases and full spreading resumes at the prevailing spreading ridge. We define a paleoplate as the area of the seafloor, from the axis of a failed rift to the boundary of resumed, full spreading. It includes a fragment of captured lithosphere and the lithosphere slowly accreted to it during the period of dual spreading, prior to complete abandonment of the failed rift. The Mathematician paleoplate has the following boundaries and components from west to east: the axis of the Mathematician failed rift, the fragment of captured Farallon plate, a complex rift initiation site at the Moctezuma Trough, a zone of slow spreading, and an as yet ill-defined eastern boundary where dual spreading stopped and full spreading resumed. The northern boundary of the paleoplate is the Rivera fracture zone; its southeastern boundary a now-inactive transform fault, the West O'Gorman fracture zone. In this case, as well as in other more poorly documented ones, relict landforms and magnetic patterns are carried on the aging lithosphere, away from the spreading ridge, recording a former geometry.

In Washington County, southwestern Utah, a thick sequence of volcanic rocks ranges in age from latest Oligocene to Holocene. The Tertiary units consist predominantly of ash flows of intermediate calc-alkalic composition, whereas the youngest volcanic rocks are mostly basaltic andesites. Thermochemical calculations indicate that some of the mafic lavas could be derived from melting of spinel peridotite within the upper part of the low-velocity zone at depths of 50 to 65 km. Some basaltic andesites contain megacrysts of quartz; if the quartz is cognate, the magmas may have formed by partial fusion of quartz eclogite at depths of approximately 120 to 150 km. We suggest that this quartz eclogite is a remnant of the Farallon plate, which underwent subduction at the time of calc-alkalic Tertiary volcanism.