Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Effect of subducting sea-floor roughness on fore-arc kinematics, Pacific coast, Costa Rica Donald M. Fisher Department of Geosciences, Pennsylvania State University, University Park, Pennsylvania 16802 Thomas W. Gardner Department of Geosciences, Trinity University, San Antonio, Texas 78212 Jeffrey S. Marshall Department of Geosciences, Pennsylvania State University, University Park, Pennsylvania 16802 Peter B. Sak Marino Protti Observatorio Vulcanológico y Sismológico de Costa Rica, Universidad Nacional, Apartado 86-3000, Heredia, Costa Rica ABSTRACT Fault kinematics and uplift in the Costa Rican fore arc of the Middle America convergent margin are controlled to a large extent by roughness on the subducting Cocos plate. Along the northwest flank of the incoming Cocos Ridge, seafloor is characterized by short wavelength roughness related to northeast-trending seamount chains. Onland projection of the rough subducting crust coincides with a system of active faults oriented at high angles to the margin that segment the fore-arc thrust belt and separate blocks with contrasting uplift rates. Trunk segments of Pacific slope fluvial systems typically follow these margin-perpendicular faults. Regionally developed marine and fluvial terraces are correlated between drainages and across faults along the Costa Rican Pacific coast. Terrace separations across block-bounding faults reveal a pattern of fore-arc uplift that coincides roughly with the distribution of incoming seamounts. Magnitude and distribution of Quaternary uplift along the Costa Rican Pacific coast suggests that, despite a thin incoming sediment pile, the inner fore arc shows an accumulation of mass—a characteristic that may be due to underplating of seamounts beneath the fore-arc high. INTRODUCTION It has long been recognized along convergent margins that subduction of large bathymetric features, such as ridges, leads to broadly distributed deformation in the fore arc (Gardner et al., 1992; Chung and Kanamori, 1978). In this paper, we examine the fore-arc response to subduction of shorter-wavelength roughness elements such as Figure 1. Two topographic data sets combined to form digital elevation model for Costa Rica and Middle America Trench (viewed to the north). Offshore data from von Huene et al. (1995). Onland data from National Imagery and Mapping Agency of U.S. Department of Defense (30 arc second grid). Symbols: EB—Esparza Block, OB—Orotina Block, HB—Herradura Block, PB—Parrita Block, QB—Quepos Block, FC—Fila Costeña Thrust Belt, CS—Cóbano Surface, RSB—rough-smooth boundary, FS—Fisher Seamount, QP—Quepos Plateau, SL—sea level. Geology; May 1998; v. 26; no. 5; p. 467–470; 3 figures. seamounts. Along the Middle America Trench, there is a relatively abrupt southeastward change from subduction of smooth Cocos plate created at the East Pacific Rise to rough crust created at the Galápagos Rift system (Hey, 1977). Herein, we describe features of the onland geology that provide insights into how the basement and sedimentary rocks in the inner fore arc deform in response to the incoming sea-floor roughness southeast of the rough-smooth boundary. We use marine and fluvial terraces to establish the distribution of Quaternary uplift along the Costa Rican Pacific coast. These data are combined with analyses of mesoscale fault populations to characterize the fore-arc kinematics. The pattern of uplift and faulting is compared with the offshore bathymetry to evaluate the relationship between the seamount distribution on the Cocos plate and both the kinematics and spacing of faults within the inner fore arc. OFFSHORE BATHYMETRY AND STRUCTURE OF THE COSTA RICAN FORE ARC Offshore of Costa Rica, the major bathymetric features on the Cocos plate are the Cocos Ridge, the Quepos Plateau, the Fisher Seamount, and a series of smaller seamount chains (von Huene et al., 1995) (Fig. 1). The Cocos Ridge and adjacent seamount chains intersect the trench at high angles and approximately parallel to the relativeplate-motion vector between the Cocos plate and the overriding fore arc. Thus, the position of incoming seamounts relative to the fore arc does not vary greatly through time. To the northwest, where smooth Cocos crust is subducting, the fore arc marks the edge of the Caribbean plate. The rough crust, however, is subducting beneath the edge of the Panama block, a microplate that has been independent of the Caribbean plate since the late Tertiary or early Quaternary (Silver et al., 1990). The position of the rough-smooth boundary along the trench may have migrated southeast to its present position offshore of the tip of the Península de Nicoya (von Huene et al., 1995; Hinz et al., 1996). The incoming seamount province has caused considerable tectonic erosion in the offshore part of the fore arc; the trench slope is scalloped and indented around seamounts, and marginperpendicular scarps record seamount tracks beneath the fore arc (Fig. 1; von Huene et al., 467 1995). As these features indicate, the outer forearc kinematics are characterized by initial uplift in response to the underthrusting of a seamount followed by subsidence after the seamount is subducted (e.g., von Huene and Lallemand, 1991; von Huene and Scholl, 1991; von Huene et al., 1995). This deformation history leaves the characteristic signature within the lower slope of a scarp-bounded depression where a seamount has passed beneath the margin (von Huene and Scholl, 1991). In this paper, we describe the forearc response arcward of this scalloped or damaged accretionary prism. A prominent feature in seismic reflection profiles of the fore arc is a rough, irregular reflector that represents the boundary between lowvelocity slope sediments and an underlying higher-velocity wedge that extends out to the lower slope (Stoffa et al., 1991; Shipley et al., 1992; McIntosh et al., 1993; Hinz et al., 1996). Shipley et al. (1992) proposed that the high- velocity wedge consists of accreted Tertiary sedimentary rocks, whereas von Huene and Flüh (1994) argued that this material was more likely Cretaceous–early Tertiary basement rocks similar to onshore exposures of the Nicoya Complex (see also Donnelly, 1994). Leg 170 of the Ocean Drilling Program drilled through the upper-slope sediments offshore of the Península de Nicoya and encountered a breccia consisting of fragments similar to the Nicoya Complex within a Miocene matrix (Kimura et al., 1997). This finding suggests that the scattered exposures of Cretaceous–early Tertiary basement rocks found along coastal peninsulas and headlands may represent sections of a continuous fore-arc basement exposed within structural highs uplifted along faults that offset the rough reflector. To support this interpretation, we present the observation that the basement exposures coincide with the locations of the highest Quaternary uplift rates in the fore arc. QUATERNARY GEOLOGY AND STRUCTURE OF THE EXPOSED COSTA RICAN FORE ARC Trunk sections of major fluvial systems draining the Pacific side of the mountainous fore arc typically follow active, margin-perpendicular faults that bound a system of independent forearc blocks. Regionally, extensive marine and fluvial terraces can be correlated along the Costa Rican Pacific coast and across these faultbounded segments of the fore arc. Terrace correlations are variously constrained by radiometric dating of terrace deposits (14C) and interbedded volcanic rocks (K-Ar), and by soil chronosequences based on B-horizon development and weathering-rind thicknesses on basalt clasts in terrace gravels. The elevations of fluvial terraces vary significantly along the coast with distinct, measurable changes occurring across blockbounding, coast-orthogonal faults. Uplifted Holocene marine platforms show the same sense of Figure 2. Central-southern Costa Rican Pacific coast schematic cross sections (marginparallel) showing inner fore-arc fault blocks and thrust belt, outer fore-arc peninsulas, coastal rivers with basin asymmetry factors, El Diablo fluvial terrace, Cóbano marine terrace, late Quaternary marine-terrace uplift rates, and offshore bathymetry. Upper plot: Close-up of fault blocks within inner fore arc and on southern edge of Península de Nicoya within outer fore arc. Inner fore-arc blocks: Ez—Esparza, O—Orotina, H—Herradura, Es— Esterillos, P—Parrita, Q—Quepos. Outer forearc block: C—Cóbano. Solid squares—El Diablo terrace outcrops (6 km inland); dashed gray lines—projected El Diablo terrace surface; solid triangles—Cóbano marine terrace; open triangles—Holocene marine-platform uplift rates projected to expected elevation for a 125 ka terrace (oxygen isotope stage 5e sealevel highstand). Arrows indicate sense of vertical separation on block-bounding faults. Shaded blocks expose Cretaceous basement rocks (Nicoya Complex). Major coastal rivers shown above plot (AF—drainage basin asymmetry factor: AF = 50 indicates symmetric basin; AF < 50 indicates northwest-tilted block with larger basin area southeast of trunk river). Middle plot: Inner fore-arc fault blocks, Fila Costeña thrust belt, and outer fore-arc peninsulas (Nicoya and Osa). El Diablo terrace elevations (from detailed plot above) shown with range of late Quaternary coastal uplift rates for southern tips of Nicoya and Osa peninsulas (Marshall and Anderson, 1995; Gardner et al., 1992). Uplift rate scale (right side) corresponds with elevation scale (left side) for a 125 ka surface. Lower plot: Marginparallel bathymetric profile across seamount domain and Cocos Ridge showing position of roughness elements with respect to fore arc. 468 GEOLOGY, May 1998 offset as fluvial terraces across block boundaries and provide quantitative measurements of shortterm uplift rates within discrete fore-arc segments. In addition, the highest topography and oldest rock exposures occur within the uplifted blocks. Thus, continuous Cretaceous basement is only exposed in structural highs with the fastest uplift rates (Figs. 1 and 2). Coast-orthogonal faults mark the boundaries between seven major blocks along a 150 km stretch of the central Costa Rican Pacific coast (Fig. 1). Each block displays measurable northwest-down tilting. The highest measured uplift rates along the Pacific coast of Costa Rica are from the Península de Osa, directly inboard of the axis of the incoming Cocos Ridge (Gardner et al., 1992) (Fig. 2). The long-wavelength decrease in elevation on the northwest flank of the ridge is broadly consistent with the overall pattern of coastal uplift, with lower uplift rates to the northwest and northwest tilting of individual blocks. We suggest, however, that the shorter-wavelength roughness related to seamounts superimposes local variability on this overall uplift pattern. Below we report data used to evaluate the vertical separation across block-bounding faults, the tilting within blocks, and the Holocene uplift rates from northwest to southeast along a 150 km stretch of the central Costa Rican Pacific coast. The thickest and most regionally extensive aggradational fluvial terrace, the El Diablo terrace (Fisher et al., 1994), occurs near the coast along most major Pacific slope rivers (Fig. 2). Aggradational terraces commonly develop within the lower reaches of coastal rivers in response to rising base level during major sea-level highstands (e.g., Merritts et al., 1994). Dated stratigraphic relationships, modern elevation, and extrapolation of Holocene uplift rates indicate that the El Diablo terrace may record the last major Quaternary sea-level highstand at oxygen isotope stage 5e (ca. 125 ka). El Diablo gravels overlie a welded tuff dated at 240 ka (K-Ar) near the Río Tárcoles (Fisher et al., 1994). An estuarine clay horizon at the base of the El Diablo gravels near Quepos yields abundant wood that is in excess of 46 020 yr B.P. Furthermore, Holocene ages for elevated wave-cut benches have been obtained on two blocks that preserve the El Diablo terrace. The Holocene uplift rate obtained for these blocks is broadly consistent with the elevation of the El Diablo terrace if these gravels were deposited during the highstand at oxygen isotope stage 5e. In this paper, we use abrupt first-order variations in the elevation of the El Diablo terrace with respect to modern river elevation to characterize relative uplift between fault-bounded blocks. We recognize that these elevation differences may also include second-order effects related to differences in river gradient and position of the coastline through time. Northwest of the rough-smooth boundary, fluvial systems are symmetric with equal drainage GEOLOGY, May 1998 areas on both sides of the trunk segment (Fig. 2). However, all rivers to the southeast of the Río Naranjo show asymmetric drainage basins as northeast-trending faults allow northwest tilting about rotation axes parallel to the trunk streams. This northwest down tilting is generally consistent with the regional decrease in uplift rates along the coast northwest from the Cocos Ridge (Gardner et al., 1992). Just to the southeast of this change in drainage basin morphology, a fault offsets the El Diablo terrace along the Río Barranca with a stratigraphic separation of ~30 m measured 11 km upstream from the coast (Fig. 1) (Fisher et al., 1994). The Esparza block has been uplifted faster than the Puntarenas lowlands to the northwest. Wood collected from beneath a colluvial wedge on a raised wave-cut bench at Punta Carballo yields a radiocarbon age of ca. 3 ka, which indicates an uplift rate of 1–1.5 m/k.y. Extrapolation of this Holocene uplift rate to oxygen isotope stage 5e at ca. 125 ka would give an elevation for the El Diablo terrace of between 125 and 190 m. The observed elevation for the El Diablo terrace of 140 m above river level, 6 km upstream from the coast, falls within that range (Fig. 2). Along the Río Jesús María, a steep, northeaststriking fault separates the Esparza block from the Orotina block (Fig. 1). The contact between Miocene sedimentary rocks and an overlying lahar is offset across the Jesús María fault, with a stratigraphic separation of ~120 m. The El Diablo terrace on the Orotina block is projected along its longitudinal profile to an elevation of 90 m above the river 6 km inland from the coast (Fig. 2). A steep, northeast-striking fault along the Río Tárcoles juxtaposes Miocene sedimentary rocks along the southeast margin of the Orotina block, and Cretaceous pillow basalts of the Nicoya Complex on the Herradura block (Fig. 1). The Herradura block coincides with the highest topography in the fore arc, and the exposed basement has been stripped of most of its Tertiary and Quaternary cover. Another margin-perpendicular fault separates the Herradura block from the Esterillos block to the southeast, a block that also exposes Cretaceous–early Tertiary rocks but has undergone less uplift, so it preserves Miocene sedimentary cover and the El Diablo terrace (Fig. 1). The El Diablo terrace is found at an elevation of 170 m above river level 6 km inland on the Esterillos block. The western end of the low-lying Parrita block has an uplift rate an order of magnitude slower than the Herradura or Esterillos blocks on the basis of the 20–40 m elevation above river level for the El Diablo terrace 6 km inland. The Parrita, Esterillos, Orotina, and Esparza blocks all have basin asymmetries and an inferred tilt on the El Diablo surface that indicate down-to-the-northwest tilting. The wavelength of the uplift profile along the central Pacific coast is similar to that of the shortwavelength bathymetry of the subducting plate. The Esparza, Herradura, and Quepos blocks show anomalously high uplift rates and have a trench-parallel spacing and a width comparable to incoming seamounts (Fig. 2). Inboard of the incoming Fisher Seamount at the tip of the Península de Nicoya, a Holocene wave-cut bench records uplift rates of 1.7 to 4.5 m/k.y. (Marshall and Anderson, 1995) and the Pleistocene Cóbano terrace, which may represent oxygen isotope stage 5e highstand at 125 ka (Mora, 1985), is found at an elevation of 180–200 m (Fig. 2). Late Figure 3. Geologic map of Costa Rica showing fault plane solutions and kinematic axes (Marrett and Allmendinger, 1990) within Miocene sediments (Ms) and Quaternary fluvial deposits (Qt). Solid circles—P axes, open squares—T axes, PFZ—Panama fracture zone. Inset shows tectonic map of Central America. 469 Quaternary surfaces record rapid uplift and arcward tilting, possibly in response to incoming seamounts (Marshall and Anderson, 1995). The magnitude and distribution of uplift along the Pacific Coast of Costa Rica suggests that, in contrast to behavior near the toe of the wedge and despite a thin incoming sediment pile, there is net accumulation of mass. A comparison of uplift patterns and offshore bathymetry raises the possibility that incoming seamounts are underplating beneath the fore-arc high. Regional faults that lie along major trunk streams (Barranca, Jesús María, Tárcoles, and Parrita) are steep and strike northeast, roughly parallel to both the Cocos Ridge and the relativemotion vector between the Cocos plate and the overriding Panama block. These faults segment the fore arc and separate blocks with different uplift rates. The largest faults (along the Río Tárcoles and Río Parrita) juxtapose Cretaceous– early Tertiary basement with both shallow marine–deltaic sedimentary rocks of late Tertiary age, and Quaternary fluvial and volcanic units. Vertical offsets along major faults over the past 100 k.y. range from tens of meters to several kilometers (total potential relief on the rough surface inferred from the absence of any slope cover on basement blocks). Mesoscale faults are concentrated near regional faults and are observed in all units from Miocene sedimentary rocks to gravels of late Quaternary age (Fig. 3). The fault kinematics are also similar in all units: Steep faults at a high angle to the margin display normal, strike, and oblique slip. Shortening axes based on the method of Marrett and Allmendinger (1990) are typically vertical or north trending, whereas extension axes are typically east trending. Thus, mesoscale faults display orientations, strain axes, and recent kinematics that are consistent with the vertical separations inferred from stratigraphic correlations across regional faults. Focal mechanisms for recent seismic events are also consistent with the regional fault orientations and kinematics (e.g., Güendel et al., 1989), suggesting that the deformation on observed faults is ongoing. CONCLUSIONS The distribution of coastal uplift along the Costa Rican fore arc is broadly consistent with long-wavelength roughness related to the incoming Cocos Ridge (Gardner et al., 1992) (Fig. 2). Superimposed on this pattern is more local uplift related to incoming seamounts. Relative longterm and short-term uplift rates determined from fluvial and marine terraces vary significantly between independent blocks. The corresponding uplift profile along the margin is thus corrugated, and the distribution of uplifted blocks is roughly similar to the small-wavelength roughness on the incoming Cocos plate. Faults that bound fore-arc blocks are always at high angles to the margin and display strike, 470 normal, and oblique slip. We find that the Mesozoic basement rocks are restricted to blocks with higher short-term and long-term uplift rates. The boundary between basement rocks and slope sediments that is observed in offshore seismic lines is presumably eroded off these fore-arc highs. These observations are consistent with a continuous basement wedge that has undergone laterally variable uplift in response to small-wavelength roughness on the subducting plate. This roughness results in segmentation of the fore-arc basement, a consequence that may limit the lateral propagation of faults during earthquakes and prevent large subduction events (Protti, 1991). Thus, fore-arc kinematics and seismicity along thinly sedimented margins may largely depend on roughness of the subducting crust. ACKNOWLEDGMENTS This research was funded by National Science Foundation grants EAR-9214832 and EAR-9526955. We thank R. von Huene for providing us with the GEOMAR bathymetry data offshore of Costa Rica, and D. Pope for invaluable assistance with Figure 1. We also appreciate helpful reviews of this manuscript by E. Silver and R. Bürgmann. REFERENCES CITED Chung, W., and Kanamori, H., 1978, A mechanical model for plate deformation associated with aseismic ridge subduction in the New Hebrides arc: Tectonophysics, v. 50, p. 29–40. Donnelly, T. W., 1994, The Caribbean Cretaceous basalt association: A vast igneous province that includes the Nicoya complex of Costa Rica, in Seyfried, H., and Hellmann, W., eds., Profil 7, Geology of an evolving island arc: Stuttgart, Universität Stuttgart, p. 17–45. Fisher, D. M., Gardner, T. W., Marshall, J. S., and Montero, W., 1994, Kinematics associated with late Tertiary and Quaternary deformation in central Costa Rica: Western boundary of the Panama microplate: Geology, v. 22, p. 263–266. Gardner, T. W., Verdonck, D., Pinter, N., Slingerland, R., Furlong, K., Bullard, T. F., and Wells, S. G., 1992, Quaternary uplift astride the aseismic Cocos Ridge, Pacific coast of Costa Rica: Geological Society of America Bulletin, v. 104, p. 219–232. Güendel, F., Montero, C., Rojas, D., Brenes, J., Segura, J., Sáenz, R., and González, V., 1989, Informe sobre el enjambre sismico ocurrido cerca de la cuidad de Orotina entre el 21 y el 24 de Enero de 1989: Boletín Sismológico, Observatorio Vulcanológico y Sismológico de Costa Rica, v. 5, p. 9–16. Hey, R., 1977, Tectonic evolution of the Cocos-Nazca spreading center: Geological Society of America Bulletin, v. 88, p. 1404–1420. Hinz, K., von Huene, R., Ranero, C. R., and the PACOMAR working group, 1996, Tectonic structure of the convergent Pacific margin offshore Costa Rica from multichannel seismic reflection data: Tectonics, v. 15, p. 54–66. Kimura, G., Silver, E. A., Blum, P., et al., 1997, Proceedings of the Ocean Drilling Program, Initial Reports, v. 170: College Station, Texas, Ocean Drilling Program. Marrett, R., and Allmendinger, R. W., 1990, Kinematic analysis of fault-slip data: Journal or Structural Geology, v. 12, p. 973–986. Printed in U.S.A. Marshall, J. S., and Anderson, R. S., 1995, Quaternary uplift and seismic cycle deformation, Península de Nicoya, Costa Rica: Geological Society of America Bulletin, v. 107, p. 463–473. McIntosh, K., Silver, E., and Shipley, T., 1993, Evidence and mechanisms for fore-arc extension at the accretionary Costa Rica convergent margin: Tectonics, v. 12, p. 1380–1392. Merritts, D. J., Vincent, K. R., and Wohl, E. E., 1994, Long river profiles, tectonism, and eustasy: A guide to interpreting fluvial terraces: Journal of Geophysical Research, v. 99, p. 14031–14050. Mora, C. R., 1985, Sedimentología y geomorfología del sur de la Península de Nicoya (Provincia de Puntarenas, Costa Rica) [Tesis de Licenciatura]: San José, Costa Rica, Escuela Centroamericana de Geológia, Universidad de Costa Rica, 148 p. Protti, M., 1991, Correlation between the age of subducting Cocos plate and the geometry of the Wadati-Benioff zone under Nicaragua and Costa Rica [Master’s thesis]: Santa Cruz, University of California, 66 p. Shipley, T. H., McIntosh, K. D., Silver, E. A., and Stoffa, P. L., 1992, Three-dimensional seismic imaging of the Costa Rica accretionary prism: Structural diversity in a small volume of the lower slope: Journal of Geophysical Research, v. 97, p. 4439–4459. Silver, E. A., Reed, D. L., Tagudin, J. E., and Heil, D. J., 1990, Implications of the north and south Panama thrust belts for the origin of the Panama orocline: Tectonics, v. 9, p. 261–281. Stoffa, P. L., Shipley, T. H., Dean, D. F., Kessinger, W. K., Elde, R., Silver, E., Reed, D., and Aguilar, A., 1991, Three-dimensional seismic imaging of the Costa Rica accretionary wedge: Field program and migration examples: Journal of Geophysical Research, v. 96, p. 21693–21712. von Huene, R., and Flüh, E., 1994, A review of marine geophysical studies along the Middle America Trench off Costa Rica and the problematic seaward terminus of continental crust, in Seyfried, H., and Hellmann, W., eds., Profil 7, Geology of an evolving island arc: Stuttgart, Universität Stuttgart, p. 143–160. von Huene, R., and Lallemand, S., 1991, Tectonic erosion along the Japan and Peru convergent margins: Geological Society of America Bulletin, v. 102, p. 704–720. von Huene, R., and Scholl, D. W., 1991, Observations at convergent margins concerning sediment subduction, subduction erosion, and the growth of continental crust: Reviews of Geophysics, v. 29, p. 279–316. von Huene, R., Bialas, J., Flueh, E., Cropp, B., Csernok, T., Fabel, E., Hoffman, J., Emeis, K., Holler, P., Jeschke, G., Leandro M. C., Peréz Fernandéz, I., Chavarria S. J., Florez H. A., Escobedo Z. D., León, R., and Barrios L. O., 1995, Morphotectonics of the Pacific convergent margin of Costa Rica, in Mann, P., ed., Geologic and tectonic development of the Caribbean plate boundary: Geological Society of America Special Paper 295, p. 291–307. Manuscript received October 22, 1997 Revised manuscript received January 30, 1998 Manuscript accepted February 13, 1998 GEOLOGY, May 1998