by John McPhee
As with so many things that were obscure or mystifying before the arrival of plate tectonics, the discovery of the origin of ophiolites was something like the discovery that the rock you have been using for twenty-five years as a doorstop is actually the Stone of Scone. Suddenly, for example, the concept unmentionable in a Stanford doctoral thesis was helping to tell the story of California as it had not been understood before. Moores, reflecting on this, said to me once, “If the story of California sounds fantastic, with all its accreting arcs and melanges coming from the western sea, just look at a map of the southwest Pacific—look at the relationship between Australia and Indonesia right now.”
As fresh attention accrued to the ophiolitic suite where—in completeness—it was found on land, the actual dimensions of the ocean lithosphere could be measured part by part from top to bottom, clarifying the results of seismic refraction. After the sequence has formed at spreading centers—where it is in large part liquid and is much swollen with heat—it cools and thins as it moves away, and after travelling twenty million years and seven hundred miles is a deep cold slab:
A few tens of metres of ocean sediments drift down upon the deep cold slab, settling on top of
a kilometre or so of pillow lavas, under which is
a kilometre or so of sheeted dikes, under which is
a kilometre or so of plutonic rock (plagiogranite, gabbro), under which is
a kilometre or so of plutonic rock in which cumulate crystals settled in layers upon a distinct chamber bottom—
the Moho—
under which is a kilometre or so of mantle rock, some of which was melted in the spreading center and some of which is peridotite in its solid original form and if water has reached it is serpentine.
When Moores talks about ophiolites with fifth-graders in Davis, he sketches the ocean floor much as he has for me, simplifying the vertical sequence and presenting an idealized rock column that closely resembles the one above. It is more than just a useful model. In its general way, it is accurate. But as a description it is only somewhat more encompassing than to say that Herman Melville wrote a novel about a one-legged madman in vengeful pursuit of a whale. Remarkable as ocean lithosphere may be for its worldwide youth and repeated character, it is not nearly as simple as that summary outline. In Moores’ words, “Where ophiolites are made, in spreading centers, hot fluids are mixing, cooling, and so forth. We are not talking about clear sedimentary layers; ophiolite contacts are gradational. Moreover, there are various types of ophiolites: some are from the basins behind or in front of island arcs, some from the intersections of spreading centers and transform faults. There is undersea ‘weathering.’ Parts of the sequence erode, more sediment comes down, and as a result there are hiatuses in seafloor rock, just as there are in rock that accumulates on land. In Italian ophiolites, the diabase is missing, and so is the gabbro. The serpentine is full of calcite—it’s called ophicalcite, very beautiful white and green and red stone, expensive building stone. There is some gabbro in Elba. But there are no sheeted-dike complexes in Italy. Obviously, Italian ophiolites formed in a different ocean environment, and what that may have been is not well understood. Oceanic crust is not a simple three-layered thing, as geophysics is telling us now. Geophysicists are unable to produce a consistent model. Nature is messy.”
One day, on a field trip we made to Cyprus, Moores did a long and detailed inspection of an outcrop he had not studied before, and figured out a chain of magmatic events in which layered gabbro had come first and a plagiogranite sill had intruded below the gabbro—an inversion of the usual sequence. “This reminds us not to take the chart of the ophiolites simply,” he said. “Layered gabbro may be lower than plagiogranite in the master chart, but here we see a plagiogranite sill under the layered gabbro. It came in later. Things don’t always happen in the earth as they do on charts.”
Not to confuse me but just to give me a reality shot, he sketched out what he described as “an expanded ophiolitic assemblage”—an elaboration of the picture I had written on my palm. He repeated the generalized column with enough added detail to suggest the actual complexity of the rock that lies below the oceans. I could forget it as soon as I read it, but at least I ought to sense the tangle of nature, and thus the nature of science.
Where you find ophiolites on land, you might find at the top of the sequence the shallow-water limestones of the sea from which they emerged, or even laterites from soils formed in air as the ophiolite was lifted by the continental margin.
The deepwater sediments that drifted down upon the moving lithospheric plate may be chalk (as in Cyprus) or the product of volcanoes (as in the Smartville Block) or chert (as in Italy or Greece). They tell you something about the oceanic environment through which the lithosphere travelled.
Beneath the massive pillow lavas are
more pillow lavas, shot through with diabase dikes that came up in molten state with enough pressure to continue past the
zone of massive sheeted dikes. If the ophiolite were an animal, these originally vertical laminations would be its brain. Taken together, they are the tape measure and chronometer of the seafloor. As each new dike forced its way into the complex, the seafloor spread that much. New dikes would intrude every fifty to a hundred years, often splitting the previous dike up the middle. The average width of the split dikes would be about seventy centimetres. They recorded absolutely the episodic widening of the ocean lithosphere, in contrast to the arithmetical notion of geophysicists, who suggest that seafloors spread continuously at a quotient number of centimetres per year.
Among the plagiogranites are large diabase dikes (feeders of the sheeted complex)
and large diabase dikes are among the massive varitextured gabbros as well.
The stratiform gabbros are generally cyclic: plagioclase and pyroxene cumulate above olivine and pyroxene cumulate with intercumulate plagioclase (and sometimes olivine and plagioclase) above olivine cumulate with some chromite in it, a combination known as dunite. You will not at once recognize all these things where an ophiolite has made an appearance in a roadcut or a mountain cliff, but at least try to remember that somewhere within this zone is the geophysical Moho—
and at the bottom of this zone is the petrologic Moho.
I have to interrupt him. Two Mohos? How could there be two Mohos? The Moho, as fifth-graders can tell you, is where the crust ends and the mantle begins—about five kilometres down from the ocean bottom and thirty-five kilometres below most places on continents and as much as sixty kilometres below deeply floating mountains. “Moho” is a geophysical term, coined in honor of the Croatian seismologist Andrija Mohorivicic, who in 1909 discovered the crust-mantle boundary. When geophysicists examine their autodriven strip-chart recordings, which look like close-up photographs of matted gray hair, they see what Mohorivicic saw. They see seismic waves speeding up when they hit the olivine-rich cumulates, and they call that the change from crust to mantle. Geologists looking at ophiolites in geographical settings see mantle material below the olivine-rich cumulates and discern that the mantle supplied the olivine that went into the rock above it and now devilishly accelerates the geophysicists’ seismic waves. Below the olivine-rich cumulates geologists see what they regard as the true transition from crust to mantle, and they call it the petrologic Moho. Geophysicists insist their machines can’t be wrong, the nature of the rock notwithstanding. Moores says to remember, however, that Moho means “Mohorivii discontinuity” and the seismic discontinuity is decidedly where geophysicists say it is. The discontinuity is seismic and is recorded on paper, but the crust-mantle boundary, which is lower, is recorded in the rock. Nature, in this case, is not messy or confused. The science is—for the time being. Like the early cartographers piecing together the face of the globe, geologists and geophysicists are now trying to map places that no human being will ever see but which are features of the earth no less than Scotts Bluff or the Shetland Islands, and were features of the earth when Scotts Bluff a
nd the Shetland Islands were other rock in other places, and will be features of the earth when Scotts Bluff has totally disintegrated and the Shetland Islands are under the sea. The two Mohos are like a camera’s divided range finder trying to close. The two Mohos are an imperfectly mapped frontier
under which, in the ophiolitic sequence, the kilometre or so of mantle rock is peridotite, a general term for rock composed mainly of olivine and a little pyroxene. Depending on the kind of pyroxene and the amount of pyroxene that is in it, peridotite is also called harzburgite and lherzolite and dunite, and, if water has reached it, serpentine. Peridotite that has moved in its solid original form is called tectonite, and presumably extends to the bottom of the lithospheric plate. One of the thickest measurable sections of mantle rock emplaced on any continent today is in Macedonia, and is about seven kilometres from top to bottom, a considerable weight to lift from the mantle into the air.
In this young field within earth science, the two Mohos are scarcely the sole battleground. Because ophiolites develop not only at mid-ocean ridges but also in the small spreading centers associated with island arcs and sometimes along oceanic transform faults, arguments about their origins can be intense. There is considerable agreement that the Smartville ophiolites relate to the arrival of an exotic island arc, and that the Bay of Islands ophiolite complex, in Newfoundland, formed at a mid-ocean spreading center. But Moores thinks Cyprus formed in mid-ocean, and most ophiolitologists do not. Italy’s ophiolites, with their missing parts, seem to some, but not others, to be fragments of oceanic transform faults. The Papuan ophiolites of New Guinea are so complicated that they seem—to some workers—to have come not only from a mid-ocean spreading center but also from behind an island arc.
Moores says that ophiolites are more important as models for the mechanism of spreading than they are as relics of the environments where they were made. The elapsed time between formation and emplacement is measurable by various methods, and some people argue that emplacement seems to have happened too quickly—an average of thirty million years—for most of the world’s ophiolites to have derived from mid-ocean centers. Moores says he has no argument with people who hold that most ophiolites form near continents—in fore-arcs or back-arcs—but he insists that there is enough time to bring them in from mid-ocean ridges as well. Geologists argue about the chambers of magma under spreading centers, and whether the chambers were continuous over time or were punctuated units that crystallized and were followed by new chambers that in turn crystallized, and so forth. They wonder, above all, why only a few ophiolites are more than a thousand million years old, while the earth itself is four and a half times older. For the last thousand million years they can work out the tectonic history—the shifting shapes of continents, the rise of long-gone mountains—from the ophiolites that were left behind. Before that, in the early Proterozoic and the Archean Eon, what was going on? Was something different going on? Something other than plate tectonics?
The long argument over how ophiolites are emplaced has provided workers with a more immediate distraction. In 1971, R. G. Coleman, of the United States Geological Survey, proposed that where ocean crust slides into a trench and goes under a continent, a part of the crust—i.e., an ophiolite—is shaved off the top and ends up on the lip of the continent. He called this “obduction.” In 1976, David Elliott, of Johns Hopkins, decided that rocks could not stand the roughshod tectonics that Coleman had proposed—that the ophiolites would be shattered into countless parts and would just not make it up there. Elliott proposed gravity sliding—ophiolites as toboggans coming to rest. But something would have to lift the seafloor and break some off before it could slide. And what, for example, could have lifted the Macedonian mantle more than forty-five thousand vertical feet? The idea that enjoys the widest acceptance was around before either of the others, but no one much noticed, because it came from a graduate student and a postdoctoral fellow. In 1969, Peter Temple and Jay Zimmerman had proposed that the emplacement of an ophiolite might occur when a continental margin goes down under ocean crust, jams a trench, and then isostatically lifts the ocean crust.
Their proposal, derived from seismic data, recognized that the subduction of lithospheric plates was far more varied than people had supposed. Not only did ocean floor dive under continents but also—and much more commonly—it dived under other ocean floor, like two carpets overlapping. The lower slab, after melting, rose through the upper one as a volcanic-island arc. Now the island arc begins to move with the plate on which it rests. Plate motions shift. New trenches form. In back-arc basins, new ocean crust is made. Some island arcs go in one direction for a while and then reverse themselves. They choke a trench, say, and then go the other way, eating up their own crust as they go. Some of the crust might get emplaced as an ophiolite. The Marianas back-arc basin is spreading now, and so is the Lau-Havre Basin behind the Tonga-Kermadec arc, and so is the basin behind the South Sandwich Islands. Between Indonesia and the Philippines are two trenches that are eating their way toward each other and if nothing stops them will destroy each other. Some people—Moores among them—think that in similar fashion off Jurassic California two trenches were active simultaneously, the easterly one dipping to the east and the westerly one dipping west, and both were destroyed during the Smartville emplacement. In Geology for December, 1983, the volcanologist Alex McBirney, after dealing with the increasingly complicated attempts to relate igneous rocks to plate tectonics, closed with a vision of the decade to come. He said, “I predict that our present confusion about igneous rocks will rise to undreamed-of levels of sophistication.”
Within half a decade, it was decided that the Smartville arc had formed where a spreading center developed in a transform fault or a fracture zone, after which the vector of the plate changed. When plate motions change, transform faults may turn into subduction zones or spreading centers. Forty-three million years ago, for example, the Pacific Plate changed its heading from north to northwest. In the hot-spot track of the Emperor and Hawaiian seamounts, the change is recorded in a pronounced bend—from north-south to northwest-southeast—dating exactly to that time. As transform faults turned into trenches all over the ancestral Pacific, the Tonga-Kermadec arc was created, and the Aleutian Islands and the Marianas. Something like this seems to have happened a hundred and sixty million years ago, creating the Smartville island arc.
That, Moores said, is where things are today—in the understanding of an ophiolite and its history in the world. If I was confused, so were geologists, not to mention geophysicists. “It has taken people a lot of time to stop thinking of these things as locally derived igneous rocks, and to begin thinking of them as transported tectonic features.” If the Smartville Block was bewildering in the third and fourth dimension, I should at least reflect on the complexity of the surface, where Yuba City is the county seat of Sutter County, Marysville is the county seat of Yuba County, Auburn is the county seat of Placer County, Placerville is the county seat of El Dorado County, and El Dorado is the county seat of nowhere.
Heaping Pelion upon Ossa, Moores reminded me that there was a possible ophiolite higher in the Sierra and that in many ways it perplexed the relative simplicity of the story he had been telling. Known in geology as the Feather River peridotite, it was a large body of serpentine and related rocks that we had seen near the interstate just below Dutch Flat, where we went into a canyon one day whose walls, soft to the touch, were snakeskin black and green. The Feather River peridotite defied explanation, because it was much older than the terranes on either side of it. If this region of the western United States consisted of accreted terranes, progressively younger as you moved west, what was the Feather River peridotite doing there between Sonomia and Smartville? “The dates on it range from the Devonian through the Permian,” Moores said. “We don’t know how it got here. It was metamorphosed at six hundred degrees centigrade and twenty kilometres down. Must have been an extraordinary thrust that brought it up from twenty kilometres. Some of t
he rocks it deforms on its eastern contact are Triassic in age. It is older than Sonomia but was emplaced later. It is older than Smartville and lies east of Smartville, but—who knows how or why?—it was emplaced later. We have the story of several successive terranes neatly tying onto North America, and then we discover that the Feather River peridotite is between No. 2 and No. 3, that it is much older than any rock in No. 2, and that its emplacement is younger than the arrival of No. 3. The Feather River peridotite was about two hundred million years old when emplaced here. What does that mean? That it’s anomalously old? That it was picked up first by an island arc and ultimately emplaced twice? Is it, in the first place, an ophiolite? What else can it be? It includes serpentine, unserpentinized peridotite, metagabbros, other amphibolites of diverse parentage, and, possibly, deformed sheeted dikes. What does that suggest?”
