3. Origin and Morphology of Ocean Margins

It is likely that the term was coined before its correctness (in terms of crustal rocks) was realized. The ocean floor, as we have seen, is made of geologically young basalt, unlike the continents. The basaltic rock that forms the ocean floor basement is rather close in composition to the mantle rock it came from (see the section on rocks, in the Appendix). Basalt is somewhat heavier than continental rock largely due to its high iron content. The lesser specific weight of the continental mass makes it protrude above the surrounding seafloor (Fig. 3.1). A thick wedge of sediment forms at the resulting boundary between ocean and continent. (In this view much of the shelf is continent.) The sediments may be well layered or strongly deformed, largely depending on the tectonic forces active at the particular margin considered.

The great difference in character of the margin (collision or trailing) is readily seen on the North American continent (Fig. 3.2). The western margin, where new continental crust is being made, is rising and displays cliffs with old sediments or with igneous rocks. The trailing eastern margin is sinking and bears large areas of wetlands and lagoons. Of course, photos taken at the surface can only show surficial differences along the coast. Most of the margins are in fact deeply submerged. The shelf (typically up to 200 m deep and with a slope of less than one half of a degree) has proportions of the seafloor of between less than 2% (Pacific, dominantly collision margin) and 8% (Atlantic, dominantly trailing edge) of its host ocean, the continental slope from 5% to 8% (the slope being typically very gentle, less than 1.5°), and the continental rise from 1.6% to 6.2% (with an inclination commonly well below 1°). Despite the relatively modest extent of the ocean margin, its sediment wedge tends to be very substantial, thanks to the input of materials from erosion on nearby land and (in places) of explosive volcanism. Also, here at the margin we find much of this input mixed with marine products from the coastal ocean, as was long ago recognized by John Murray of the Challenger Expedition and all marine geologists since.

There is one type of seafloor that everyone is familiar with: the beach, at the shallow end of the shelf. In Southern California, it is typically a thin strip of sand on a narrow terrace cut into cliffs (Fig. 3.3). Elsewhere, on trailing margins, beaches can be much wider. Also, in between the Californian beaches, there are marshes and wetlands, commonly filled to create real estate for human use. Some are left in their natural state, sporting abundant marine benthic creatures living in and on the surface of dark mud. The mudflats are important sites of geochemical reactions (e.g., denitrification), and they are also important as sources of larval plankton to the ocean offshore. They occur at sea level in association with Holocene delta deposits piled into bays that formed as a result of the great melting at the end of the last ice age (Chap. 11).

The seafloor of the permanently submerged portion of the continental margin is much less well known than the exposed part, of course. Its morphology is accessible through sounding, notably using acoustics, and its cover is discerned by taking spot samples and reconstructing the resulting large-scale patterns. The continents, which serve as a prolific source of sediment from erosion of uplifted continental crust, tend to dump their debris first of all in the coastal zone from where it passes on to the great depths offshore. Unsurprisingly, the correct interpretation of well-traveled sediments proved to be a serious challenge.

Mangrove growth is commonly observed at tropical margins. In some subtropical bays of Baja California, the shallowest part of the seafloor bears dense mangrove vegetation (Fig. 3.4). Mangrove is a sea-level indicator. See Chap. 6.

The continental margins (or ocean margins) are the dumping sites for the debris coming from the continents, that is, the terrigenous sediments. The margins also harbor the most fertile parts of the ocean. Thus, much organic matter becomes buried within the continental debris and the added pelagic material. In addition, such burial takes place in reef debris. If conditions are right (mainly a matter of history of burial and of heating), this organic material can eventually develop into petroleum, over millions of years. This happened in several places, one of which (conspicuously) is the Gulf Coast area, where oil is found buried under immense masses of sediment (see Chap. 14). Both oil and tar are found in the coastal zone of Southern California, where a thick wedge of sediments accumulated in the Neogene, with plenty of organic matter from upwelling. From experience most Californians know that drilling for offshore resources of oil and gas comes with considerable risks for underwater blowout.

Which margins are likely to have especially thick sediment wedges? Presumably, the largest and most active drainage areas are likely to emerge with the thickest wedges. One such super wedge, for example, is created by the deep-sea fan in the Bay of Bengal, a fan fed by the erosion of the Himalayas, the highest terrestrial mountains on the planet, and linked to a large and high plateau. Among the regular margins of ocean basins, however, it is those of the Atlantic that have the thickest wedges of sediment bordering it, up to 10 km thick and more. The Atlantic also has the largest proportions of slope and continental rise areas of the major ocean basins. The reason for this is not only sediment supply but also the fact that the Atlantic margins are “trailing edges” that have been sinking (and collecting sediments) for tens of millions of years.

Continental margins differ greatly depending on their origin. As early as 1883, the Austrian geologist Edward Suess (1831–1914) coined the terms Atlantic margins and Pacific margins to emphasize the major differences (Figs. 1.​2 and 3.2). As mentioned, Atlantic-type margins are steadily sinking regions accumulating thick sequences of sediment in layer-cake fashion. In contrast, Pacific-type margins are rising on the whole and are associated with volcanism, folding, faulting, and various other mountain-building processes. Atlantic-type margins are now referred to as “passive” and Pacific-type margins as “active,” because of the differences in tectonic style and the intensity of earthquakes and the absence or presence of active volcanoes.

It is now obvious that the origin of the margins must be understood in the context of seafloor spreading and plate tectonics. In the Atlantic the continental margins originated from a tearing apart of an ancient continent, along a line of weakness or great stress, with the torn edges sinking and accumulating sediment over large flat areas. A modern example of the rifting process associated with sedimentation can be studied in the Red Sea, where mantle material pushes up and tears the Arabian Peninsula from Africa (Fig. 3.5).

The initial phase of the process of passive margin formation can be studied in the East African Rift Valley, where the sea has not yet entered the rift. The pulling apart (and therefore the thinning) of the continental crust opens a window for mantle material (Fig. 3.5, left), melt that intrudes from the upper mantle. Heat flow increases correspondingly, and there is a bulging upward from the rising material, much as at the mid-ocean ridge. An increasing gap develops between the separating edges of continental crust, and pieces of continental crust break off at the edges at “listric faults.” The process stretches the continental crust. Volcanism can be prominent in the early stages of rifting margins (as readily seen in thick volcanogenic margins of the Greenland-Iceland-Norway Sea, for instance). However, not all rifting margins start out in volcanic fashion. Non-volcanic types can be seen in the Northern Bay of Biscay and at Georges Bank, for example. Nevertheless, the outflow of lava is common at the time of initial rifting. If sufficiently extensive, such outflow can produce volcanic margins of impressive dimensions, as happened in East Greenland (DSDP Leg 38 and ODP Legs 104 and 152). Volcanogenic margins, of course, are likely to be especially prominent wherever hot spots contributed to the initial rifting.

In the Atlantic, coral reef margins are mainly found around the Caribbean Sea. The Atlantic margins off Africa bear excellent examples of trailing edges loaded with thick mixtures of terrestrial and pelagic sediments. The sediment stacks are commonly mapped using acoustics, that is, with a vessel pulling a sound-making device across the margin and recording the return, much as in echo sounding, but with the sound penetrating into the sediment (rather than being reflected from the seafloor only). In the example here shown (Fig. 3.6), the results were used in preparing for ODP Leg 175. (Prior to any ODP drilling, proposed drill sites had to be carefully surveyed, to prevent the choice of hazardous sites, that is, sites that could conceivably produce petroleum or large amounts of gas.)

Present and Neogene conditions, of course, are not necessarily typical for the geography prevailing during the evolution of the Atlantic. The early Atlantic saw the influence of restricted access to a newly rifted basin, as well as that of warm climate and a high sea level. Ancient salt deposits and reef ramparts are what we see along many of the Atlantic margins, although other sediments, of a more continental character, do exist also (Figs. 3.7 and 3.8). Salt deposits, of course, also are well known from the Gulf of Mexico, where they push up salt domes (“diapirs”), providing a path for petroleum migration (see Chap. 14). In the Atlantic proper, good evidence for salt deposits exist especially off Angola. The salt in the South Atlantic presumably accumulated when the ocean basin was narrow and closed to the north. Presumably it would have had restricted exchange with the world ocean due to the Walvis Ridge – Rio Grande Barrier to the south (now near 30°S). Large petroleum reserves may be associated with the salt deposits because the South Atlantic also was the site of deposition of organic-rich sediments during a lengthy period in the middle Cretaceous.

The end result of rifting, then, are continental margins consisting of thick sediment stacks piled both on the sinking blocks of a continental edge and on the adjacent oceanic crust. We can generalize these conclusions to all margins that originated by rifting and are now riding passively on a moving plate (that is, passive margins). Besides the bulk of the margins of the Atlantic there are the East African margin, the margins of India, much of the margin of Australia, and practically the entire Antarctic margin in this category of passive margin. In the Antarctic, of course, special conditions have prevailed with respect to erosion and deposition, at least since the formation of ice sheets there. Thick ice sheets have been present there for sure since the late Tertiary and quite probably much earlier, at least after the Eocene.

There are a large number of unresolved questions in the study of passive margins. The commonly used analogies for the evolution of rifting – from East Africa and Red Sea to Gulf of California to Atlantic Basin – provide guidance as to the processes that might be at work. However, such analogies rarely answer geologic questions arising when studying actual rifting history. How much stretching (if any) was there before rifting? How wide was the original rift valley before the sea entered? How does the sinking of the marine parts of the continental edge affect landward crustal blocks? What were the rates of uplift and subsidence through time? And what were the rates of associated erosion and sedimentation? With regard to the history of subsidence, what is the relative role of “floating in the mantle” (isostatic equilibrium) of the blocks, versus gravitational sliding? What are the forces responsible for the very long-lasting uplift of certain parts of the continental margin and for the formation of long deep-seated barrier ridges along some margins? What is the significance of the lack of sediments of a certain age, in many margins? Was the lack caused by erosion or by nondeposition or by huge landslides?

In each particular case, a mixture of causal factors usually applies. “Silver bullet” situations with only one factor are rare.

One question of fundamental interest is whether and where the thick sediment stacks piling up on the passive margins will eventually be found in the geologic record on land. After all, the Atlantic margins cannot just go on moving apart – sooner or later a widening Atlantic runs out of space. One suggestion (by the Canadian geologist J.T. Wilson, 1908–1993) is that a proto-Atlantic once was formed by rifting, presumably in the early Paleozoic, and then closed again, running the previously passive margins into each other, thus turning a rift into a collision zone. A presumed product of this process (the “Wilson cycle”) is the chain of (Paleozoic) mountains from Norway through Scotland and Newfoundland and on to the southern Appalachians. Checking the positions of the mountains on the Bullard fit (Fig. 1.14), we find that Wilson’s suggestion makes sense as far as geography.

“Pacific-type” margins are the product of collision of plates. There are at least three types of collision margins that we need to consider: the ones produced by continent-continent collision as in the Himalayas, the ones reflecting continent-ocean collision as at the Peru-Chile Trench, and the ones where subduction takes place along volcanic islands, as along the Marianas, for example (see Fig. 3.9). There are two fundamentally different types of the latter. As pointed out by the Scripps-trained marine geologist Dan Karig (then a graduate student) around 1970, one of these features is back-arc spreading (in Fig. 3.9 the two active margin types are compared schematically).

Among the most important characteristics of collision margins are folding and shearing of sediments along faults, as well as the addition of volcanic and plutonic material derived from mobilizing matter from the down-going lithosphere. The fractionation processes associated with partial melting on the descending slab and with hydrothermal reaction (seawater with hot basalt) can lead to local enrichment of deposits with heavy metals – and hence to the formation of ore deposits (see Chap. 14).

Based on deep-ocean drilling, we learned that as a rule the types of rocks that characterize the continental margins next to subduction zones are extremely varied. Among volcanogenic types we see solidified ashes, as well as altered basalts and gabbros. Among other igneous rocks, we see andesitic types along with highly deformed metamorphic rocks. Among sediments we can recognize both pelagic and shallow-water contributions. Certain rock types originating from the deep seafloor, when found on land, are referred to as ophiolites and are mapped in an effort to find ancient subduction zones. It is like hunting for lost oceans on land. Occasionally, the reward is the discovery of massive copper sulfides and other ores. The mechanism that allowed these observable ophiolites to escape subduction and to enter mountain building is a matter of research and discussion.

The nature and history of active margins are the subjects of ongoing research. It still holds many surprises. An early simple concept of an origin from scraped-off material left behind by the down-going crust and lithosphere had to be substantially modified, for example. There is little transfer of material in places. In fact, there is tectonic erosion, whereby portions of the growing margin are swallowed by subduction. The swallowing trench may be fed by massive slumping (as is the case for portions of the Japan Trench), a process that delivers material that, after modification, can be used to build crust. The role of fluids in shaping margins has received increased attention. Tectonic motions (faulting, thrusting) and chemical reactions within the accretionary prism are both generally influenced by the presence of such fluids and their composition. The fluids are largely expelled from porous rocks by tectonic compaction and also stem from dehydration reactions involving mineralogy. Gases are important, too. In the Caribbean Barbados Ridge Complex, for example, the low-angle fault called “décollement” above the oceanic crust (Fig. 3.10) is greased by methane-bearing fluids within the fault zone. The presence of fluids aids in keeping the sedimentary wedge detached from the down-going slab.

Margins in zones of volcanic island chains are especially active. Spreading may co-occur with subduction where back-arc spreading is in evidence (Fig. 3.9, upper panel). Back-arc spreading occurs landward of volcanic arcs as, for example, in the Philippines or west of Guam. More than three fourths of such marginal basins are found in the western Pacific. That extension (by spreading, which lets magma rise) should be associated with collision is surely surprising. Complexity also arises from adding shear to active or to passive margins or from replacing trailing edges or collision margins with shear margins. Shear margins, like active ones, tend to have narrow shelves. Not all margins can be readily classified, of course. The processes characterizing them may be difficult to identify and categorize.

Active margins also are sediment traps, although perhaps less obviously so compared with passive margins. In active margins sediments are piled up into chaotic mixtures of various types of rock, unlike in the passive margins, where well-ordered layering is the rule. One must keep in mind that in active margins enormous masses of material simply disappear deep within the mantle. The scale of subduction activity is difficult to imagine – the lithospheric slab now entering the Japan Trench is more than 10,000 km long! At present rates it will vanish in about 100 million years. Fluids leaving the landward wedge piling up in the subduction zone make “cold seeps.” They may come from sediments, rather than from volcanogenic sources, depending on local conditions. Fluids travel along faults, mainly.

The shelf is the submerged part of a continent. Its typical maximum depth is between 100 and 150 m. It is rather flat, connecting the nearshore zone with the shelf break over some distance. Some shelves are quite wide, especially on passive margins (such as those in the Atlantic) (Fig. 3.11).

On the whole, the wide Atlantic shelves are depositional features, that is, they reflect sediment buildup on a sinking site of deposition. Narrow and rocky shelves, on the other hand, are common on active margins (as off California). Here erosional processes play an important role in shaping the shelf. Uplifted shelves (largely displaying alternation of erosion and deposition) and erosion-made nearshore areas once at sea level can make “marine terraces,” features that are familiar along the coastal zone of California. Here they commonly serve as platform for roads or as sites for housing.

Some shelves extend deeply into continents and harbor shelf seas such as Hudson Bay, the Baltic Sea, or the Persian Gulf. Most of the marine sediments found on land were originally deposited in shelf seas. Thus, to understand these sediments – which cover a large portion of the continents – one must study sedimentary processes in modern shelf seas (see Chaps. 4, 5, and 6). Hudson Bay and the Baltic Sea, in their morphology and their sediments, have a memory of the ice ages, though, a rather unusual condition within Earth history. Low latitude shelf seas, for example, on the rim of the Mediterranean Sea or in Indonesia, might be more useful in delivering analogs for ancient shelf sediments. Given the ice buildup in the Neogene and a concomitant drop of sea level, useful shelf seas might not be as abundant as they were before the Neogene. Analogs for ancient shelf sediments are sometimes difficult to find.

Quite generally, present shelf environments and sediment types show considerable variety even over short distances. In large part, the variety owes to the fact that the sea level stood much lower than now only 15,000 years ago. Conditions were entirely different then from those of today, and many portions of the shelf still reflect those historic conditions in topography and sediment cover.

The general nature of modern shelves, then, reflects tectonic factors (active versus passive) along with the recent rise of sea level by some 100 m, as ice on land melted during the transition from the last glacial maximum to the present. Thus, changes in sea level join tectonics as factors of prime importance in shaping shelves. To these sculptor processes must be added climate conditions (e.g., wave climate, storms, and various other local effects). In low latitudes, buildup by reef-forming organisms is (and was for millions of years) important for shelf morphology in many places (Fig. 3.12, right panel). Some of the low-latitude shelves may be legacies of a distant past, even a pre-ice past. In high latitudes ice has been an important agent of shaping shelves for the last several million years (Fig. 3.12, left panel). The growth of ice not only led to exposure of the shelves to erosion but it also dumped enormous amounts of debris in places. Such debris includes large amounts of moraines. Much of this type of material still sits on the shelves off Newfoundland and in the North Sea. Ice moving out onto shelves also actively carved deep ravines and depressions that have not yet been filled with sediment. Runoff from melting ice can result in valleys carved into shelves. The fjords of Norway, Greenland, and Western Canada are witnesses to the powerful direct carving action of glacier ice. At the end of fjords, one finds sills made of moraines. Uplift (from ice unloading) can make such sills into barriers.

Shelves formed by large deltas off river mouths (e.g., Amazon, Mississippi) can be very flat and monotonous, in striking contrast to both rugged ice-carved shelves and irregular coral reef shelves. A rich supply of fine sediment brought to the delta environment allows for redistribution and much smoothing by waves and coastal currents. Of course, waves and currents can also build dunes, barriers, beach berms, and sand waves, depending on circumstances (see Chap. 9). Examples of conditions where the supply of terrigenous sediment (from rivers) is high and the shelf is flat include the North Sea, the shelf off the northern shore of Siberia, and the shelf of the Yellow Sea. Another prime example is the Senegal delta (West Africa), where the shelf has less than a foot of relief over several miles!

Tsunamis chiefly are produced close to the trenches rimming the Pacific; they travel over thousands of miles within a few hours. The waves are extremely long and quite low in the open sea, so that they are not noticed on board a ship riding them. But when such a wave reaches a shelf, it gains height on slowing down when “feeling bottom” in shallow water. Heights of tens of meters can be reached, and correspondingly severe damage can be produced along the coast suffering the onslaught.

The shelf break, where the shelf ends and where the continental slope begins, is a prominent morphological feature of most continental margins. In principle, the break marks the depth below which the influence of sea level on erosion and deposition wanes rapidly. However, many details governing this feature remain to be discovered, especially since the origin of a break may involve a number of different factors. In many places off Antarctica, the break is uncommonly deep (near 400 m) compared with the usual depth near 130 m. Presumably, thick ice played a role in making the shelf break so deep. However, the break is equally deep off southwest Africa, where an explanation linked to ice action apparently would not work. Most commonly the shelf break is readily identified, especially where coral reefs are involved (Fig. 3.12, right panel). However, the transition between shelf and slope can be quite gradual in places.

The classic profile of Atlantic-type (passive) continental margins shows a steepening of the slope beyond the shelf break and a gradually diminishing descent toward the deep seafloor (Figs. 3.13 and 3.14). The relatively steep part beyond the shelf break is the upper continental slope. The slope transitions into the extremely gentle (virtually flat) part of the margin that leads into the deep sea, the continental rise. The boundary between the lower slope and the upper rise is actually ill defined. Perhaps, it is helpful to think of the slope as definitely part of the continental margin, whereas the continental rise is built on oceanic crust and is essentially part of the deep-sea environment.

Not all slopes and rises, of course, fit the textbook outline of the ideal Atlantic-type margin – not even in the Atlantic. Deep marginal ridges, as off Brazil, sheer walls of outcropping ancient sediments and deep-lying plateaus, such as the Blake Plateau off Florida, can interrupt the ideal sequence.

The collision margins off Peru and Chile (Fig. 3.9b) are characterized by a steep slope and are without a rise – the trench swallows the material that would normally build the rise. A descent in a series of steps is typical for the collision slope, marking the complicated tectonics associated with the collision between opposing plates. Rather complicated conditions, in fact, prevail off much of the western US coast also. Earthquakes are common both at the rim of California and at that of Chile.

While the present slope and rise off northern and central California can be described simply enough in terms of coalescing deep-sea fan deposits (Fig. 3.13), no such description is possible for the margin off Southern California. The Southern California Borderland looks much like an extension into the sea of the basin and range topography familiar from the Mojave Desert, even though much complicated horizontal motion is involved within the Borderland.

The various physiographic diagrams commonly depicting the rapid descent from shelf edge to the deep seafloor (whether across passive or active margin) are somewhat misleading in that the slopes are, in reality, quite gentle (one to six degrees). A one-degree slope, of course, would appear as a plain on land to anyone standing on it.

Basically, most slopes are the surfaces of thick accumulations of sediments washed off the continents and mixed with marine materials produced by organisms. Close to volcanoes, of course, there is a strong component of volcanic debris. The high rates of accumulation on many slopes result in a precarious situation in places. When there is not enough time to de-water and solidify slope sediments, immense submarine landslides can result from rather small disturbances (e.g., smallish earthquakes), even on very gentle slopes. The same is true if gas pressures within the sediment rise sufficiently. Such gases are commonly created by decay reactions of organic matter and may be unable to escape rapidly enough into bottom waters. They then can destabilize their host sediment.

Examples of slides on slopes are found readily along almost all continental margins. Off Cape Hatteras, for example, a tongue-shaped mass of displaced sediment on the upper continental rise is 60 km wide and over 190 km long, with a hummocky relief of up to 300 m in places. Another example of a large-scale slide is seen off Senegal (NW Africa) and off Dakar (Fig. 3.14). The jumbled slide masses at the foot of the slope now form part of the continental rise. The seismic profile shown in the figure was obtained by echo sounding into the seafloor with powerful “booms” of sound, rather than the “pings” used to define the surface of the floor below the moving ship. The method using booming sound is referred to as continuous seismic profiling. It is sketched as the inset in Fig. 3.14.

As mentioned, margins are not just places of deposition but also of erosion. Erosion on slopes and rises can be produced by submarine landslides involving earthquakes and sediment instability and by the action of strong deep currents flowing horizontally along the slope. Commonly such currents have been dubbed contour currents. They are distinct from another type of important current: turbidity currents, mud-laden bodies of water flowing downhill. It is now thought that much of the sediment on continental rises and in adjacent abyssal plains is carried there by turbidity currents following slides and slumps that started somewhere below the shelf break, with contour currents doing the redistribution and shaping of the material. Thus, slides and turbidity currents apparently are largely responsible for shaping sediment-covered margins (Fig. 3.15).

The continental slope is commonly cut by various types of incisions, ravines and valleys; the most spectacular of which are called submarine canyons. The origin of these impressive features has long puzzled marine geologists. Much has been learned concerning the matter, but the topic is still a matter of debate. The largest of such canyons off California is “Monterey Canyon” of Monterey Bay (Fig. 3.16, left). Its cross section has dimensions like those of the Grand Canyon of the Colorado.

One of the earliest of detailed studies of submarine canyons was off S.I.O., by Francis P. Shepard and his students and colleagues. By chance, S.I.O. has two of these features off its shores, named, appropriately enough, La Jolla Canyon and Scripps Canyon. The latter, a tributary to La Jolla Canyon, is somewhat smaller than the former. They join at a depth near 300 m, below the shelf break. The resulting La Jolla Canyon subsequently turns in a southerly offshore direction to end up in the San Diego Trough off the Point Loma Peninsula. The late geophysicist D. Inman, an expert on coastal processes, found out that beach sand sporadically moves down the canyons (Fig. 3.16, right). Indeed, such sand is found in the San Diego Trough, along with the remains of kelp, presumably delivered there by muddy gravity-driven flows down the canyons (that is, by turbidity currents). Note that the canyon heads shown (the largest nearby and the closest to S.I.O.) are on the shelf (which was exposed during the last ice age).

Many large submarine canyons are much like their counterparts on land: with tributary systems in the upper parts, and meandering thalwegs along much of their course (similar to river beds), and with steep sides in places (20° to 25°, even 45°). Overhanging walls also occur (Fig. 3.17). Walls may consist of hard rocks, including granitic rock in places. As in river canyons, there is a continuous descent of the valley axis, with maximum slopes of 15% near the coast and gentle slopes (about 1%) farther out to sea.