Daniel E. Karig

Subduction and Accretion in Trenches

Although the reality of subduction has been greatly strengthened by recent investigations, there is little information dealing with the mechanisms by which material is subducted or accreted to the upper plate. An attempt to determine the gross evolution of subduction zones has been made, assuming that geographic variations in morphologic and geophysical characteristics of trenches can be transformed into temporal trends. Deformation associated with subduction extends across the lower trench slope, from the trench axis to the trench-slope break. This region is a rising tectonic element, but the upper slope is a subsiding region of sediment accumulation. An upper slope discontinuity separates this zone of subsidence from the rising frontal-arc block. Examination of very young trenches indicates that the upper-slope discontinuity marks the upper section of the continental or insular slope that existed before a subduction pulse began. As material is fed to the subduction zone, the distance between the upper slope discontinuity and the trench increases, and an accretionary prism develops, but its shape depends on the relative rates of sediment feed from the arc and from the offshore basin. The lower boundary of the accretionary prism is the upper section of the seismic zone, which apparently widens and flattens as one or more accretionary prisms accumulate. The sediment cover on the downgoing plate and some of the igneous crust appears to be stripped off the plate before it reaches a point beneath the volcanic chain. Turbiditic sediments deposited in the trench axis are preferentially sheared off the underlying pelagic sediments and are accreted to the lower trench wall. The pelagic sediments and crustal material are probably accreted at deeper structural levels. Where turbidites overlie pelagic sediments in the trench axis, the turbidites are stripped off in fold packets with axial surfaces having very low dips. These dewatered and rigidified structural units move up the lower slope, as subsequent packets are accreted. In trenches that subduct lithosphere carrying very thin pelagic sediment covers, accretion and uplift of crustal slabs seem to occur as topographic irregularities enter the trench.

On the applicability of a universal elastic trench profile

Abstract Using thin elastic plate theory and neglecting horizontal applied forces, a universal deflection profile applicable to many oceanic trenches is derived. This theoretical profile is compared with bathymetric profiles from the central Aleutian, Kuril, Bonin, and Mariana trench-outer rise regions. The profiles were corrected for sediment thickness and age variation of the lithosphere. Good agreement between theory and observation is found. The distance from the first point of zero deflection seaward of the trench to the point of maximum height of the outer rise is directly related to the flexural rigidity of the lithosphere. The thickness of the elastic lithosphere is found to vary between 20 and 29 km for the trench profiles considered. The good agreement obtained shows that horizontal forces may be neglected and that the bending lithosphere behaves elastically in the cases considered. The analysis shows that only unreasonably large horizontal forces would affect the universal deflection curve. It is concluded that although the near-surface lithosphere may be subject to brittle fracture, the deeper lithosphere is capable of transmitting elastic stresses as high as 9 kbar.

Makran of Iran and Pakistan as an active arc system

The character of convergence along the Arabian-Iranian plate boundary changes radically eastward from the Zagros ranges to the Makran region. This appears to be due to collision of continental crust on the west in contrast to subduction of oceanic crust on the east. The Makran ranges display progressively older and more highly deformed sedimentary units northward from the coast, together with an increase in elevation of the ranges. North of the Makran ranges are large subsiding basins, flanked to the north by active volcanoes. Published geologic data as well as Landsat images indicate that the Makran is a large sedimentary prism accreted during the Cenozoic. Almost all the characteristics of accretionary prisms observed in well-studied arcs can be identified or inferred in the Makran, which, however, is unique in its degree of exposure. These circumstances lead to a better understanding of Makran geology and of the mechanics of island arcs hi general.

Triple junctions as a cause for anomalously near-trench igneous activity between the trench and volcanic arc

In the eastern Aleutian arc, western Sumatra, and southwestern Japan, the distribution of igneous rocks cannot be explained by direct analogy with the igneous activity in modern island arcs. In these places, episodes of intrusion occurred within the accretionary prism, anomalously close to an active trench. For the eastern Aleutian arc and Sumatra, we speculate that the near-trench igneous rocks could be related to the subduction of a ridge that trended nearly perpendicular to the trench. The ridge-trench-trench triple junction so formed migrated along the trench. As the accretionary prism was under-thrust by the ridge, it would increase the heat flow sufficiently to cause partial melting of the sediment. In Japan, the near-trench igneous rocks could be related to the migration of a trench-trench-trench triple junction along the Japan Trench. The activity associated with this triple junction may be continuing today.

Sediment deformation and hydrogeology of the Nankai Trough accretionary prism: Synthesis of shipboard results of ODP Leg 131

The main objective of Leg 131 was to provide data on the deformational processes and associated hydrogeology of the Nankai prism toe. Drilling succeeded, for the first time in the history of ocean drilling, in penetrating the complete sedimentary sequence to basaltic basement, reaching 1327 mbsf (metres below seafloor) with good core recovery (55%). Excellent correlation of the lithology and structure, including the frontal thrust and the decollement, with seismic reflection images was also determined. Bedding dips, faults and shear bands analyzed in the cores confirm the pattern of deformation to be mainly due to NW-SE shortening, as expected from the plate tectonic convergence vector. Below the decollement, no significant deformation features were observed, indicating that the decollement is a sharp discontinuity in stress transmission. Physical properties data show major discontinuities at the decollement, notably an increase in porosity below the later. This may indicate excess pore pressure in the subducted section and decollement zone. A less marked increase in porosity below the frontal thrust may reflect the youthfulness of this feature. Attempts to make downhole measurements were severely hampered by unstable hole conditions, but useful constraints have been placed on the thermal regime, and some calibration of laboratory physical properties toin-situ conditions has been provided, andin-situ stress and pore pressure were measured in the uppermost sediments. Evidence of channelized fluid flows is inconclusive. No sharp geochemical signatures or unequivocal geochemical anomalies indicative of channelized fluid flow were found. Thermal measurements are not significantly different from those predicted by a purely conductive heat flow model. A signature of low chloride pore water near the decollement may partly be related to smectite diagenesis but may also be due to episodic fluid flow events. We conclude that dewatering probably occurred dominantly through diffuse flow throughout the accreted sediments at this site.

Nature and distribution of deformation across the Banda Arc–Australian collision zone at Timor

Recently acquired seismic-reflection and SeaMARC II (side-scan and swath bathymetry) profiles near Timor show that the Banda Arc–Australia collision zone has a tectonic framework similar to that of a typical oceanic subduction system. Deformation is occurring, at present, most intensely at the foot of the inner slope of the Timor Trough. This deformation front is discontinuously advancing southward as new thrust slices develop within the subducted Australian margin strata. In contrast, present deformation is apparently negligible in the Savu Basin, the complex fore-arc basin north of Timor. A possible significant exception is a postulated right-lateral, northeast-trending fault zone offsetting the outer-arc high between Savu and Roti. Although back-arc thrusting has been documented north of the volcanic arc, this component of convergence is minor compared with the scale of ongoing deformation in the Timor Trough. The detailed nature of these surveys has also led to the recognition of along-strike variations in deformation in the Timor Trough and in the Savu Basin. These variations may be related to the variable degree of involvement of the Australian continental margin along the arc.

Development of sedimentary basins on the lower trench slope

In arc systems where an oceanic plate with a thick cover of sediments is subducted, the sediments are scraped off and accreted to the base of the inner trench slope. The accreted sediments form ridges, behind which younger sediments are often ponded in basins. The width of these basins progressively increases from 2 to 3 km at the base of the lower slope to 10 km near the trench slope break. Sediments within the basins increase in thickness from nearly zero in the basins now at greatest depths to several kilometres in the basins on the shallowest part of the slope. Slope basins begin to form at the base of the lower slope, where sediments accumulate between adjacent thrust faults. Addition of more accreted material at the trench causes uplift and rotation of the thrust slices and of overlying slope sediments. As deformation proceeds, motion along some of the thrusts dies out, and the inactive thrusts become buried by slope sediments, thus increasing the size of the slope basins. In orogenic belts, sedimentary rocks that were deposited on a “basement” of melange and are now tectonically enclosed by melange are hypothesized to be ancient slope-basin deposits.

Tectonic complexities in the bonin arc system

Abstract The Bonin arc system is anomalous in that it does not appear to fit the tectonic pattern observed in most arc systems. Re-examination of this arc system, with a new bathymetric chart and against a background of recent studies in other arcs, leads to reasonable explanations for its anomalous characteristics. The frontal-arc volcanics on the Bonin Islands, which now form part of the trench slope break, can best be explained by the northward rifting of the Bonin Islands block from a position along the frontal arc under the influence of oblique subduction. The very large positive gravity anomaly over the islands results from the greater than normal density and volume of the volcanics compared to most trench slope breaks. The dominant northeast—southwest ridge and trough topography, into which the Iwo Jima Ridge (frontal arc) is broken may have resulted from compressions of the arc along its trend. This compression would be attributed to the southward movement of Japan as the Yamato Basin of the Sea of Japan opened in the Late Oligocene and Early Miocene. Recent extension is occurring in the Bonin arc system, as earlier suggested, but in an east—west direction. Features associated with extension can best be identified at the south end of the arc, but may persist for its entire length. This extension is either more rapid, or began first at the south end.

Late Cenozoic subduction and continental margin truncation along the northern Middle America Trench

The narrow inner trench slope and the truncated igneous and metamorphic terrane along the west coast of Mexico between Cabo Corrientes and the Gulf of Tehuantepec indicate that part of the continental margin has in some way been removed during the process of subduction. However, a detailed marine geophysical survey of the inner trench slope near Acapulco indicates that this removal is not occurring now. South-southwest–trending magnetic anomalies produced by the Xolapa metamorphic complex extend seaward only 20 to 30 km. Oceanic magnetic anomalies that trend N50°W extend as much as 30 km landward of the trench. The boundary between these two magnetic patterns lies landward of the trench-slope break and beneath the upper-slope sediment pile. The nonmagnetic material forming the acoustic basement trenchward of the metamorphic rocks is interpreted to consist of late Miocene to Holocene deformed trench-floor turbidites. Deformation associated with subduction has reversed the gradients of several submarine canyons and tilted the seaward edge of the upper-slope sediment pile away from the trench. The morphology and structure of the inner trench slope is typical of accreting trench-arc systems, although the morphotectonic units in this system are smaller than usual. Accretion since late Miocene time is suggested by the age of dredged slope sediments and by analysis of offshore magnetic anomalies, which indicate a change from right-lateral oblique to perpendicular subduction at that time. Removal of the continental margin probably occurred intermittently between Late Cretaceous and late Miocene time. Possible mechanisms include subduction of continental crust (tectonic erosion), left-lateral translation associated with the Caribbean–North American plate boundary, and right-lateral translation associated with oblique subduction between the Farallon or Cocos plates and the North American plate. Geological data favor right-lateral offset and suggest that some of the missing margin may be the slivers of subduction complex found along the west coast of Baja California and possibly even farther north.

Role of strike-slip faulting in the evolution of allochthonous terranes in the Philippines

Concepts of allochthonous terrane transport and emplacement are dominated by the assumption that most terranes originate on the subducting plate, collide with the upper plate, and are emplaced there. Movement of terranes along the convergent margin is recognized but is generally attributed to postcollision slip. In the northern Philippines, allochthonous terranes originate primarily within the arc system, have been translated along it by strike-slip faults, and were emplaced by cessation of that slip. The authors suggest that in the Philippines some originally vertical strike-slip boundaries may have evolved into shallow-dipping sutures marked by fold and thrust systems. This mode of terrane evolution may be more common than generally appreciated, particularly in orogenic belts developed in response to oblique convergence.