The study of oceanic lithosphere has been aided by investigations of characteristic rock sequences on land known as ophiolites (literally “snake rock”, referring to the similarity of the color and texture to snakeskin; see Nicolas, 1989, for a full treatment of this topic). Ophio-lites usually occur in collisional orogens (Section 10.4), and their association of deep-sea sediments, basalts, gabbros, and ultramafi c rocks suggests that they origi-nated as oceanic lithosphere and were subsequently thrust up into their continental setting by a process known as obduction (Dewey, 1976; Ben-Avraham et al ., 1982; Section 10.6.3). The complete ophiolite sequence (Gass, 1980) is shown in Table 2.3. The analogy of ophi-olites with oceanic lithosphere is supported by the gross similarity in chemistry (although there is considerable difference in detail), metamorphic grades correspond-ing to temperature gradients existing under spreading centers, the presence of similar ore minerals, and the observation that the sediments were formed in deep water (Moores, 1982). Salisbury & Christensen (1978) have compared the velocity structure of the oceanic lithosphere with seismic velocities measured in samples from the Bay of Islands ophiolite complex in New-foundland, and concluded that the determined velocity stratigraphies are identical.
Figure 2.19 shows the cor-relation between the oceanic lithosphere and three well-studied ophiolite bodies. At one time it seemed that investigations of the petrology and structure of the oceanic lithosphere could conveniently be accomplished by the study of ophiolite sequences on land. However, this simple analogy has been challenged, and it has been suggested that ophiolites do not represent typical oceanic litho-sphere, and were not emplaced exclusively during continental collision (Mason, 1985). Dating of events indicates that obduction of many ophiolites occurred very soon after their creation. Con-tinental collision, however, normally occurs a long time after the formation of a mid-ocean ridge, so that the age of the sea fl oor obducted should be considerably greater than that of the collisional orogeny. Ophiolites conse-quently represent lithosphere that was obducted while young and hot. Geochemical evidence (Pearce, 1980; Elthon, 1991) has suggested that the original sites of ophiolites were backarc basins (Section 9.10; Cawood & Suhr, 1992), Red Sea-type ocean basins, or the forearc region of subduction zones (Flower & Dilek, 2003). The latter setting seems at fi rst to be an unlikely one. However, the petrology and geochemistry of the igneous basement of forearcs, which is very distinctive, is very comparable to that of many ophiolites. Forma-tion in a forearc setting could also explain the short time interval between formation and emplacement, and the evidence for the “hot” emplacement of many ophiol-ites. A backarc or forearc origin is also supported by the detailed geochemistry of the lavas of most ophiolites, which indicates that they are derived from melts that formed above subduction zones.
There have been many different mechanisms pro-posed for ophiolite obduction, none of which can satis-factorily explain all cases. It must thus be recognized that there may be several operative mechanisms and
that, although certainly formed by some type of accre-tionary process, ophiolite sequences may differ signifi -cantly, notably in terms of their detailed geochemistry, from lithosphere created at mid-ocean ridge crests in the major ocean basins. Although many ophiolites are highly altered and tec-tonized, because of the way in which they are uplifted and emplaced in the upper crust, there are defi nite indi-cations that there is more than one type of ophiolite. Some have the complete suite of units listed in Table
2.3 and illustrated in Fig. 2.19, others consist solely of deep-sea sediments, pillow lavas, and serpentinized
peridotite, with or without minor amounts of gabbro. If present these gabbros often occur as intrusions within the serpentinized peridotite. These latter types are remarkably similar to the inferred nature of the thin oceanic crust that forms where magma supply rates are low. This type of crust is thought to form when the rate of formation of the crust is very low (Section 6.10), in the vicinity of transform faults at low accretion rates (Section 6.7), and in the initial stages of ocean crust formation at nonvolcanic passive continental margins (Section 7.7.2). It seems probable that Hess (1962), in suggesting that layer 3 of the oceanic crust is serpen-tinized mantle, was in part infl uenced by his experience and knowledge of ophiolites of this type in theAppalachian and Alpine mountain belts.