Suprasubduction-Zone Ophiolites: Is There Really an Ophiolite Conundrum?

The Geological Society of America Special Paper 438 2008

Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum?

Rodney V. Metcalf† Department of Geoscience, University of Nevada–Las Vegas, Las Vegas, Nevada 89154-4010, USA

John W. Shervais‡ Department of Geology, Utah State University, Logan, Utah 84322-4505, USA

ABSTRACT

Suprasubduction-zone ophiolites have been recognized in the geologic record for over thirty years. These ophiolites are essentially intact structurally and stratigraphically, show evidence for synmagmatic extension, and contain lavas with geochemical characteristics of arc-volcanic rocks. They are now inferred to have formed by hinge retreat in the forearc of nascent or reconfi gured island arcs. Emplacement of these forearc assemblages onto the leading edge of partially subducted continental margins is a normal part of their evolution. A recent paper has challenged this interpretation. The authors assert that the “ophiolite conundrum” (seafl oor spreading shown by dike complexes versus arc geochemistry) can be resolved by a model called “historical contingency,” which holds that most ophiolites form at mid-ocean ridges that tap upper-mantle sources previously modifi ed by subduction. They support this model with examples of modern mid-ocean ridges where suprasubduction zone–like compositions have been detected (e.g., ridge-trench triple junctions). The historical contingency model is fl awed for several reasons: (1) the major- and trace-element compositions of magmatic rocks in suprasubduction-zone ophiolites strongly resemble rocks formed in primitive island-arc settings and exhibit distinct differences from rocks formed at mid-ocean-ridge spreading centers; (2) slab-infl uenced compositions reported from modern ridge-trench triple junctions and subduction reversals are subtle and/or do not compare favorably with either modern subduction zones or suprasubduction-zone ophiolites; (3) crystallization sequences, hydrous minerals, miarolitic cavities, and reaction textures in suprasubduction-zone ophiolites imply crystallization from magmas with high water activities, rather than mid-ocean-ridge systems; (4) models of whole Earth convection, subduction recycling, and ocean-island basalt isotopic compositions ignore the fact that these components represent the residue of slab melting, not the low fi eld strength element–enriched component found in active arc-volcanic suites and suprasubduction-zone ophiolites; and (5) isotopic components indicative of mantle heterogeneities (related to subduc-

†E-mail: [email protected]. ‡E-mail: [email protected].

Metcalf, R.V., and Shervais, J.W., 2008, Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum?, in Wright, J.E., and Shervais, J.W., eds., Ophiolites, Arcs, and Batholiths: A Tribute to Cliff Hopson: Geological Society of America Special Paper 438, p. 191–222, doi: 10.1130/2008.2438(07). For per- mission to copy, contact [email protected]. ©2008 The Geological Society of America. All rights reserved.

191 192 Metcalf and Shervais

tion recycling) are observed in modern mid-ocean-ridge basalts (MORB), but, in contrast to the prediction of the historical contingency model, these basalts do not exhibit suprasubduction zone–like geochemistry. The formation of suprasubduction-zone ophiolites in the upper plate of subduction zones favors intact preservation either by obduction onto a passive continental margin, or by accretionary uplift above a subduction zone. Ophiolites characterized by lavas with MORB geochemistry are typically disrupted and found as fragments in accretionary complexes (e.g., Franciscan), in contrast to suprasubduction-zone ophiolites. This must result from the fact that oceanic crust is unlikely to be obducted for mechanical reasons, but it may be preserved where it is scraped off of the subducting slab.

Keywords: ophiolite, suprasubduction zone, mid-ocean ridge, geochemistry mantle.

INTRODUCTION Moores et al. (2000) challenged the suprasubduction interpretation of ophiolite genesis. These authors assert that the “ophio- Ophiolites are distinct assemblages of ultramafi c, mafi c, and lite conundrum” (seafl oor spreading shown by dike complexes felsic igneous rocks, commonly associated with siliceous pelagic versus arc geochemistry) can be resolved by a model called “his- sediments (cherts), that have long been recognized as important torical contingency,” which holds that most ophiolites are formed components of mountain belts worldwide (Steinmann, 1906; Hess, at mid-ocean ridges that tap upper-mantle sources previously 1955). In the 1960s, this assemblage was proposed to represent modifi ed by subduction. They support this model with examples oceanic crust formed at mid-oceanic spreading centers, a concept of subduction-zone reversal, which place oceanic spreading that became central to the new theory of plate tectonics (Gass, centers above lithosphere previously modifi ed by subduction 1968). A compelling aspect of this proposal was the recognition of (i.e., the Woodlark basin), with examples of modern mid-ocean sheeted dike complexes in some ophiolites that implied formation ridges where suprasubduction zone–like compositions have been by 100% extension (e.g., Troodos; Oman), consistent with the new detected (e.g., ridge-trench-trench triple junctions), with models concept of seafl oor spreading in ocean basins (Moores and Vine, of mantle convection that show recycling of oceanic lithosphere 1971). Dedicated campaigns of deep-ocean drilling, dredging, on grand scale, and with a discussion of the isotopic components and seismic-refraction surveys confi rmed the similarity of oceanic found in ocean-island basalts (OIBs) (Moores et al., 2000). crust to ophiolites, although there were differences in detail. As a Moores et al. (2000) also suggest that differences observed result, this paradigm became entrenched within the scientifi c com- in the structural preservation of ophiolites result from distinct munity—especially among those who did not work on ophiolites. spreading environments, not from their subsequent emplace- Suprasubduction-zone ophiolites have been recognized in the ment. Thus, ocean crust and ophiolites formed at slow spread- geologic record for over three decades (Miyashiro, 1973; Pearce ing centers are highly faulted and commonly have volcanic rocks et al., 1984; Shervais and Kimbrough, 1985). These ophiolites are juxtaposed against serpentine, whereas ocean crust and ophio- made up of plutonic rocks and lavas with the mineralogical and lites formed at fast spreading centers tend to be stratigraphically geochemical characteristics of arc-plutonic and arc-volcanic rocks, intact and lack the extreme structural attenuation found in slow and they are petrologically and chemically distinct from igneous spreading ocean crust (Moores et al., 2000). Examples of rocks formed at modern spreading centers in the major ocean slow spreading ophiolites would include those in the Western basins. In general, suprasubduction-zone ophiolites are intact Mediterranean (Apennines); examples of fast spreading ophio- structurally and stratigraphically and show evidence for nearly lites would include Troodos and Oman. 100% extension. Such ophiolites are now inferred to have formed We suggest that the historical contingency model is fl awed primarily by hinge retreat in the forearc of nascent or reconfi gured for several reasons: (1) the major- and trace-element composi- island arcs, a model derived from studies of Cenozoic subduc- tions of magmatic rocks in suprasubduction-zone ophiolites tion systems in the western Pacifi c (Fig. 1; Hawkins et al., 1984; strongly resemble rocks formed in primitive island-arc settings Stern and Bloomer, 1992; Bloomer et al., 1995; Hawkins, 2003). and exhibit distinct, consistent differences from rocks formed at Emplacement of these forearc assemblages onto the leading edge mid-ocean-ridge spreading centers; (2) slab-infl uenced composi- of partially subducted continental margins (Tethyan ophiolites) tions reported from modern ridge-trench triple junctions and sub- or exposure by accretionary uplift along an active plate margin duction reversals are subtle and/or do not compare favorably with (Cordilleran ophiolites) is a normal part of their evolution (e.g., either modern subduction zones or suprasubduction-zone ophio- Shervais, 2001). Several recent papers have discussed the develop- lites; (3) crystallization sequences, hydrous minerals (hornblende), ment of suprasubduction-zone ophiolite models, their genesis, and miarolitic cavities, and reaction textures in suprasubduction-zone tectonic implications, most notably Shervais (2001), Dilek (2003), ophiolites imply crystallization from magmas with high water Pearce (2003), Hawkins (2003), and Flower (2003). activities, rather than mid-ocean-ridge magmatic systems; (4) Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 193

A oceanic crust the low ﬁ eld strength element–enriched component found in active arc-volcanic suites and suprasubduction-zone ophiolites; LM LM and (6) the isotopic components are indicative of mantle heterogeneities (related to subduction recycling) observed in modern AM AM mid-ocean-ridge basalts (MORBs), but, in contrast to the pre- B SSZ forearc spreading diction of the historical contingency model, these basalts do not exhibit suprasubduction zone–like geochemistry. In the following sections, we present a more comprehensive review of the historical contingency model and its implications, and then we present detailed evidence that rebuts this proposal. We conclude that several processes can account for ophiolites, C including formation at mid-ocean ridges, but that most large, active island arc structurally intact ophiolites with suprasubduction-zone geochemical signatures must have formed above active subduction zones. We further suggest that a more appropriate formulation of the “ophiolite conundrum” is this: Given that many, if not most, ophiolites have geochemical and petrologic signatures consistent with formation above active subduction zones, under what SSZ-like back arc spreading D circumstances does this setting result in rock associations and remnant arc active island arc structures consistent with those observed in ophiolites?

THE “OPHIOLITE CONUNDRUM” AND “HISTORICAL CONTINGENCY”

E MORB-like back arc spreading The “ophiolite conundrum” poised by Moores et al. (2000) remnant arc active island arc addresses the structural and stratigraphic evidence in ophiolites that suggests formation by nearly 100% extension, which they suggest is uniquely characteristic of oceanic spreading ridges, and the overwhelming geochemical and petrologic evidence that these same rocks formed above subduction zones. Or as they state it: “This geochemical and petrologic evidence stands Figure 1. Western Pacifi c intra-oceanic subduction zone model of in strong contrast with evidence—from both the structure within ophio lite formation by slab retreat and upper-plate extension (after Tethyan ophiolite complexes and the paleogeographic environ- Stern and Bloomer, 1992; Bloomer et al., 1995; Shervais, 2001). (A) ments inferred from surrounding and overlying sedimentary Subduction initiates along transform margin, juxtaposing young, thin deposits—that these ophiolites originated well away from any (hotter) oceanic lithosphere against older, thicker (cooler, more dense) type of subduction-related activity” (Moores et al., 2000, p. 4). oceanic lithosphere. (B) Initiation of a new subduction margin generates suprasubduction-zone (SSZ) oceanic crust in a nascent forearc by As stated, the conundrum implies that the structural and sedi- extension and magmatism in response to slab sinking and hinge roll- mentary associations found in these ophiolites are inconsistent back. (C) Stabilization of an island-arc crust by continued subduction. with any type of arc environment, and that this evidence is more (D) Rifting and reconfi guration of the arc generates suprasubduction- compelling than the geochemical and petrologic evidence (see zone oceanic crust by spreading in a narrow back-arc basin in response Pearce [2003] for a discussion Bayesian decision methods as to continued hinge rollback. (E) Back-arc basin widens as hinge rollback continues. Solid arrows are motion vectors of the subducting slab; applied to the ophiolite conundrum). dashed arrows denote migration of asthenosphere into space created by Ophiolites have long been used as natural laboratories for sinking slab and hinge rollback. In panels C–D, dashed outline denotes studying processes related to mid-ocean-ridge spreading and the regions of hotter mantle beneath the regions of active arc and back- generation of oceanic basalts. Thus, the ophiolite conundrum is arc magmatism (after Weins and Smith, 2003). AM—astheospheric an issue primarily to those who make direct correlations between mantle, LM—lithospheric mantle, MORB—mid-ocean-ridge basalt. the structural architecture of ophiolites and ocean crust formed at mid-ocean-ridge spreading centers. Nonetheless, the historical contingency hypothesis has implications beyond these correla- many suprasubduction-zone ophiolites are overlain by evolved tions (which may, in any event, provide robust models despite the lavas and volcaniclastic rocks that typically are not counted as different origins inferred for ophiolites and ridges). part of the ophiolite, (5) models of whole Earth convection, sub- What is at stake in resolving the ophiolite conundrum? The duction recycling, and OIB isotopic compositions ignore the fact heart of the issue is how we interpret the ophiolite record, in that these components represent the residue of slab melting, not particular, with regard to paleotectonic reconstructions. Moores 194 Metcalf and Shervais et al. (2000) specifi cally addressed Tethyan ophiolites in devel- ues prior to ridge subduction but stops once the ridge has been oping the historical contingency model; however, their model subducted. Formation of this slab window may allow communi- has gained attention not only with regard to Tethyan ophiolites cation between subduction-modifi ed asthenosphere and astheno- (Barth et al., 2003; Liati et al., 2004; Beccaluva et al., 2004) but sphere sources beneath the active spreading center prior to its also with regard to ophiolites from a much broader temporal subduction (Klein and Karsten, 1995; Cousens et al., 1995). This and geographic context (e.g., Moores, 2002, 2003; Saltus et al., communication may result in magmas derived from an active 2003; Maxeiner et al., 2005). spreading center that carry a subduction-like chemical and iso- The historical contingency model of Moores et al. (2000) topic compositional component (Klein and Karsten, 1995). Each consists of several disparate elements that must be discussed of these ridges (Chile, Woodlark, and Juan de Fuca) was cited by and understood separately. In its simplest proposition, Moores Moores et al. (2000) as an example of modern mid-ocean ridges et al. state, “We propose a model wherein the nature of mantle where suprasubduction-zone geochemical signatures have been tapped at mid-oceanic ridges has varied in the past in response detected. We address these examples and their signifi cance to the to prior tectonic history of a region and/or the mantle” (Moores suprasubduction-zone ophiolite debate later in the paper. et al. 2000, p. 4). They suggest several mechanisms that may affect mantle source regions such that magmas formed at mid- The Fate of Old Plates: Subducted Slab Component ocean ridges, both current and ancient, carry the geochemical and Long-Term Mantle Heterogeneities signature typical of modern subduction-zone magmas. These mechanisms include asthenosphere modifi ed by previous sub- Numerous studies have shown that the mantle sources for duction events, formation of slab windows where ridges are ocean island basalts (OIBs) are heterogeneous both in terms of subducted orthogonally or obliquely, the subducted slab com- trace-element enrichment and isotopic composition. Several ponent carried by old plates, and isotopically distinct compo- distinct isotopic components have been proposed (depleted nents found in ocean-island basalts. MORB mantle [DMM], high μ [HIMU], enriched mantle I [EMI], enriched mantle II [EMII], prevalent mantle [PREMA]) Suprasubduction Geochemistry in Basalts that must represent persistent, long-term trace-element hetero- from Active Mid-Ocean Ridges geneities in the mantle (Zindler and Hart, 1986; Hart, 1988; Hofman, 1997). PREMA and DMM represent, respectively, In support of their model, Moores et al. (2000) point to two the predominant isotopic composition of the mantle and the modern plate confi gurations where mantle modifi ed by recent depleted, MORB-source asthenosphere. HIMU, EM1, and or active subduction could contaminate the source region of EM2 isotopic compositions record the recycling of altered an active mid-ocean-ridge spreading center. One confi guration oceanic lithosphere (including continent-derived sediments) involves seafl oor spreading over a region of asthenospheric man- and/or continental lithosphere (including subduction-modifi ed tle previously modifi ed by subduction fl uids, e.g., asthenosphere subcontinental lithospheric mantle) into the mantle either that recently resided in the mantle wedge region of a now extinct directly via subduction or potentially during entrainment of subduction zone. The clearest recent example of this is the Wood- continental lithosphere during continental rifting. lark basin in the southwest Pacifi c, where collision of the Ontong Moores et al. (2000) presented numerical models of mantle Java plateau with the Solomon arc stalled subduction along the dynamics suggesting that trace chemical and isotopic hetero- NE margin of the arc and caused inception of a new, NE-dipping geneities related to subduction of oceanic lithosphere may per- subduction zone along the SW margin of the arc (Taylor and sist on extremely long time scales. After Kellogg et al. (1999), Exon, 1987; Perfi t et al., 1987; Staudigel et al., 1987; Crook and they suggest that this subducted slab material accumulates in Taylor, 1994; Johnson et al., 1987; Muenow et al., 1991). Thus, the mid-mantle region around 1500 km depth and that later it the Woodlark spreading center is a former back-arc basin that may be recycled and mixed into the overlying MORB-source is now being subducted beneath the Solomon arc. This setting asthenosphere, thus introducing a “subduction component” into is complicated by the fact that the Woodlark spreading center is normal mid-ocean-ridge basalts. Moores et al. (2000) further being subducted orthogonally, opening a slab window within the suggest that perturbations of the deeper mantle heterogeneities trench that could allow subduction components to migrate into may be linked to cycles of continental assembly and dispersion the upper plate, as discussed subsequently. (Wilson cycles), leading to periods during Earth history when A second confi guration is ridge-trench-trench triple junc- increased mantle plume activity carries subduction-modifi ed tions that mark the intersections of active mid-ocean-ridge mantle into the zone of MORB production. spreading centers with active subduction zones. Modern exam- An implication or unstated assumption of the historical ples include the Woodlark (Perfi t et al., 1987), Chile (Klein and contingency model is that when partial melting beneath mid- Karsten, 1995), and Juan de Fuca (Cousens et al., 1995) spread- ocean-ridge centers taps mantle regions that carry isotopic evi- ing ridges, and such confi gurations must have been common in dence of subduction recycling, the resulting basalts should carry the past (Klein and Karsten, 1995). Ridge subduction opens a suprasubduction-zone trace-element signatures. This assumption slab window in the subducted plate because spreading contin- is testable with geochemical and isotopic data from modern mid- Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 195 ocean-ridge basalts, much of it published over the last decade. Major-Element Compositions of Mid-Ocean-Ridge In the following sections, we present data from modern ocean and Suprasubduction-Zone Basalts basins, including both mid-ocean ridges and suprasubduction zones, from well-studied examples of the ophiolite record, and Oceanic basalts formed at normal or enriched mid-oceanic- from modern ridges with suprasubduction zone–like composi- ridge segments and ocean-island basalts formed within oceanic tions, and we use those data to evaluate the historical contingency plates or on plume-enhanced ridge segments are characterized by model and its impact on interpretations of the ophiolite record. a limited range in silica contents, moderate to high TiO2 concentrations, and by tholeiitic fractionation trends. Table 1 compares GEOCHEMISTRY OF MODERN MID-OCEAN-RIDGE the average compositions (along with minimum, maximum, and AND SUPRASUBDUCTION-ZONE MAGMATIC ROCKS standard deviation) of over 2499 mid-ocean-ridge and 545 back- arc basin volcanic rocks from the PETDB database (Lehnert An empirical relation between the composition of modern et al., 2000) to 1335 analyses from the Georoc database of basalts magma suites and tectonic setting was recognized soon after the from primitive arcs that may be analogous to ophiolites (Mariana emergence of the modern plate tectonic theory. Over the ensuing and Tonga arcs). The PETDB MORB database includes samples decades, much research has led to an understanding of the basic from slow (Indian ridge, Mid-Atlantic Ridge), intermediate (Juan petrologic processes underlying these relationships. Magma- de Fuca Ridge), and fast (East Pacifi c Rise) spreading centers. tism in the modern ocean basins can be considered to represent The contrasting compositions are highlighted in Figure 2, which four major types: normal mid-ocean-ridge basalts (N-MORB), compares the silica and TiO2 concentrations of these suites. enriched mid-ocean-ridge basalts (E-MORB), within-plate or A few lavas from the MORB database exhibit extreme ocean-island basalts (OIB), and suprasubduction-zone magmas. compositions, but, in general, the data defi ne groups with con- Basaltic compositions dominate N-MORB, E-MORB and OIB sistent geochemical characteristics. Silica is uniformly low ≈ ≈ suites; by contrast suprasubduction-zone suites are composition- (48–52 wt% SiO2), with little variation (mean 50 wt%; sigma ally more diverse, ranging from basalt to more evolved com- 1%); less than 2.5% of the data exceeds 52 wt% SiO2, and in no positions, and their plutonic equivalents, and they may include case does silica exceed 65 wt% SiO2 (Table 1; Fig. 2A). Basalts unusual lava compositions like boninites and adakites. Back-arc from mature back-arc basins have similar silica modes (Table 1; basin magmas form above subduction zones, but their compo- Fig. 2B). In contrast, volcanic rocks from primitive island arcs sitions are most similar to mid-ocean-ridge basalts when these have a wide range in silica contents, with modes near 52 wt% basins are relatively mature. SiO2; almost half of the data has silica >52 wt% SiO2 (Table 1;

Boninites are high-Si, high-Mg andesites that are found only Fig. 2C). Data for TiO2 present a similar picture: MORB values in primitive or nascent arc terranes (e.g., Crawford and Falloon, range in TiO2 from ~0.5 wt% to 3.6 wt%, with a mode around

1989). Boninites are most commonly found, along with low-K 1.6 wt% TiO2 (Fig. 2D). Back-arc basin basalts have slightly ≈ tholeiites and felsic differentiates, in forearc terranes that repre- lower modes ( 1.3 wt% TiO2), while arc-volcanic rocks have sent the extended basement upon which some modern arcs are TiO2 modes around 0.9 wt% (Fig. 2E). More than 90% of arc- built, e.g., the Izu-Bonin arc, the Marianas, and the Cape Vogel volcanic rocks have TiO2 less than 1.4 wt%, whereas more than arc (Hickey and Frey, 1982; Bloomer and Hawkins, 1983; Walker 80% of all ridge basalts have TiO2 greater than 1.2 wt% (Fig. 2F). and Cameron, 1983; Crawford and Falloon, 1989; Stern et al., The major-element data reviewed here, along with those 1991). These formed during rapid extension of the crust over a contained in the databases but not discussed here, show that mid- nascent subduction zone, prior to the establishment of modern ocean-ridge basalts exhibit a restricted range in major-element arcs (Fig. 1B). Other boninites appear to form when a mantle concentrations, which are conﬁ ned almost exclusively to compo- plume or propagating back-arc basin rift extends into the forearc sitions that would be deﬁ ned as tholeiitic basalts. More evolved region of a modern arc (e.g., Deschamps and Lallemand, 2003). compositions do occur, e.g., the recently described dacite samples

In all cases, experimental data suggest that boninites represent on the Pacifi c-Antarctic ridge at 55–65 wt% SiO2, which repre- partial melts of highly depleted previously melted mantle, in sent extreme fractionation of MORB parent magmas (Stoffers response to high fl uid fl ux from the subducting slab (e.g., Falloon et al., 2002). However, these evolved compositions are rare and and Danyushevsky, 2000; Flower, 2003; Van der Laan et al., do not represent a signifi cant fraction of the magmatic activity at 1989; Umino and Kushiro, 1989). The high fl uid fl ux lowers the mid-ocean-ridge spreading centers. Thus, if the evolved magmas solidus of the refractory, enstatite-rich mantle; the resulting melts found in suprasubduction-zone ophiolites are to be interpreted as are rich in silica and MgO because enstatite (which is rich in both mid-ocean-ridge dacite, these magmatic rocks should represent elements) dominates the melting assemblage. As a result, subarc only a small fraction of the total volcanic record, which must be lithosphere is dominated by a residual refractory mantle of harz- dominated by lavas with normal or enriched MORB composi- burgite composition, not lherzolite. In contrast, suboceanic litho- tions. In contrast, if suprasubduction-zone ophiolites represent sphere is dominated by lherzolite (diopside-bearing peridotite), primitive-arc volcanism, evolved magmas should be more com- which represents smaller degrees of partial melting of the MORB mon (up to half of all lavas), and the associated basalts should asthenosphere source (Dick, 1989). have arc tholeiite or calc-alkaline affi nities. 196 Metcalf and Shervais

TABLE 1. MEAN COMPOSITIONS, STANDARD DEVIATIONS, MAXIMUMS, AND MINIMUMS FOR MID-OCEAN-RIDGE BASALT (MORB) (MID-ATLANTIC RIDGE, EAST PACIFIC RISE, JUAN DE FUCA, INDIAN RIDGE), BACK-ARC BASINS, AND PRIMITIVE ARCS (MARIANA, IZU-BONIN, VANUATU, TONGA, KERMADEC) Mid-Atlantic Ridge East Pacific Rise Juan de Fuca Indian Ridge Back-arc basins Mean Stddev Max Min Mean StddevMax Min Mean StddevMax Min MeanStddev Max Min Mean StddevMax Min SiO2 50.31 1.02 59.46 45.40 50.05 1.05 60.1946.63 49.82 1.31 51.87 40.43 50.28 1.03 60.40 47.12 51.15 2.5372.41 45.46 TiO2 1.47 0.34 2.86 0.45 1.76 0.41 3.320.68 1.66 0.43 3.53 0.97 1.61 0.51 3.82 0.52 1.30 0.432.77 0.35 Al2O3 15.43 0.95 22.10 10.80 14.75 1.30 20.43 0.13 14.96 1.16 17.53 10.69 15.70 1.25 21.30 12.68 15.98 1.29 25.53 7.23 FeOT 9.66 1.13 14.16 0.00 10.38 1.80 15.441.12 10.53 1.55 17.38 7.60 9.38 1.56 14.63 5.72 9.17 1.4516.55 3.15 MnO 0.17 0.03 0.34 0.05 0.189 0.027 0.310.1 0.19 0.03 0.30 0.10 0.17 0.03 0.32 0.09 0.17 0.03 0.30 0.06 MgO 7.87 1.15 22.60 4.75 7.38 1.18 15.231.59 7.49 1.12 12.48 3.78 7.56 1.09 10.22 3.93 6.90 1.9423.50 0.75 CaO 11.37 0.84 14.10 1.40 11. 37 0.90 13.354.48 11.61 0.79 13.32 8.38 11.00 0.84 13.84 7.76 10.94 1.5113.86 3.03 Na2O 2.67 0.36 3.97 1.10 2.68 0.35 5.24 1.41 2.62 0.27 3.30 1.79 3.04 0.42 5.03 1.95 2.82 0.66 5.26 0.72 K2O 0.23 0.19 1.27 0.01 0.16 0.16 2.200.01 0.21 0.14 0.60 0.02 0.27 0.32 1.77 0.03 0.35 0.313.25 0.01 P2O5 0.16 0.06 0.42 0.03 0.17 0.07 1.150.04 0.17 0.08 0.59 0.05 0.20 0.09 0.67 0.04 0.17 0.10 1.12 0.01 n = 1396 n = 703 n = 106 n = 294 n = 545

Mariana Izu-Bonin Vanuatu Tonga Kermadec Mean Stddev Max Min Mean StddevMax Min Mean StddevMax Min MeanStddev Max Min Mean StddevMax Min SiO2 51.55 4.77 79.20 42.30 56.01 7.01 78.3043.03 54.06 6.61 71.23 46.38 53.40 4.91 76.65 43.80 54.21 7.0773.53 44.69 TiO2 0.99 0.37 2.55 0.17 1.08 0.68 2.700.05 0.83 0.38 2.49 0.44 1.11 0.61 2.51 0.14 0.83 0.351.68 0.12 Al2O3 15.30 1.87 20.18 10.18 15.02 1.81 20.5711.43 15.54 2.17 19.66 8.84 15.61 2.34 25.20 0.16 16.31 2.33 20.57 5.50 FeOT 10.56 2.53 16.58 2.19 8.50 2.90 11.80 0.76 9.39 2.54 16.24 4.17 9.57 1.91 13.22 1.01 8.25 3.12 12.83 0.51 MnO 0.18 0.05 0.34 0.01 0.18 0.07 0.34 0.00 0.20 0.04 0.31 0.09 0.18 0.04 0.34 0.01 0.17 0.05 0.30 0.02 MgO 6.70 2.79 18.06 0.36 4.15 3.00 13.080.00 5.92 4.44 22.61 0.70 6.40 4.44 46.35 0.36 5.38 2.7514.75 0.31 CaO 0.22 0.91 14.00 0.03 0.18 0.07 0.34 0.00 0.20 0.04 0.31 0.09 0.18 0.04 0.34 0.01 0.17 0.05 0.30 0.02 Na2O 2.52 0.65 4.72 0.41 3.17 1.19 6.90 0.39 2.84 0.81 5.20 1.43 2.61 1.03 6.39 0.10 2.48 1.08 4.93 0.00 K2O 0.62 0.53 3.82 0.00 1.16 1.36 11.51 0.20 1.89 1.15 4.90 0.19 0.42 0.36 1.91 0.01 0.52 0.512.69 0.01 P2O5 0.12 0.09 0.76 0.01 0.23 0.22 0.900.00 0.26 0.10 0.46 0.09 0.16 0.14 1.10 0.01 0.11 0.05 0.30 0.01 n = 245 n = 241 n = 566 n = 198 n = 85 Note: Back-arc basins include Mariana trough, North Fuji basin, Lau basin, Pearce Vela basin, West Philippine basin, Shikoku basin, Woodlark basin, South Sandwich basin, Sulu Sea, and Bransfield Strait. FeOT —total Fe as FeO.

Trace-Element Signatures of Mid-Ocean Ridge ure 3A. Relative to N-MORB, OIB is systematically enriched and Suprasubduction-Zone Basalts in the more incompatible trace elements (smooth slope on left side of Fig. 3A). Ocean-island basalts are thought to be derived Potential problems with the stability of major-element from a deeper, trace element–enriched mantle source, and they concentrations during low-temperature hydrous metamorphism may refl ect a complex, multicomponent source that includes have long been recognized. As a result, most geochemical stud- oceanic lithosphere recycled into the mantle by subduction. ies of submarine volcanic rocks in ophiolites have focused on Average E-MORB is also enriched relative to N-MORB but to the trace composition of the basaltic rocks, which carry the a lesser extent than OIB. The relative enrichments and depletions most stable information regarding magma source and, by infer- in oceanic basalts seen in the spider diagrams (Fig. 3A) can be ence, tectonic setting. conveniently illustrated using ratio-ratio plots such as the Nb/Yb versus Th/Yb plot (Pearce, 1982; Pearce et al., 1995) shown in N-MORB, E-MORB, and OIB Figure 3B. Such plots show the ratio of a more incompatible ele- Average N-MORB is often used as a reference when dis- ment (Nb, Th) and a less incompatible element (Yb); previous cussing the trace-element composition of oceanic magmas, partial melting in a mantle source produces a decrease in both including those from subduction settings. Compositional data for the Nb/Yb and Th/Yb ratios, while enrichments related to mantle N-MORB rocks indicate a source depleted in incompatible trace plumes result in increases in both ratios. Thus, average N-MORB, elements relative to estimated primitive-mantle compositions. E-MORB, and OIB magmas form a depletion-enrichment array This source, referred to as depleted MORB mantle (DMM), is on Nb/Yb versus Th/Yb plots. thought to reside in the shallow asthenosphere. Nd-Sr-Pb isotopic data suggest that DMM was formed during melt extraction events Suprasubduction-Zone Basalts that occurred early in Earth history (Jacobsen and Wasserburg, The hallmark of suprasubduction-zone basaltic magma 1980; O’Nions et al., 1977; Hofmann, 1997). compositions is elevated concentrations of large ion litho- A standard method for looking at the trace-element compo- phile elements (LILE: Cs, Rb, Ba, Th, K, Sr, Pb) relative to sition of oceanic basalts is the N-MORB–normalized spider dia- high fi eld strength elements (HFSE: Nb, Ta, Hf, Zr, Ti) (Wood, gram (Fig. 3) where trace elements are arranged from right to left 1980; Saunders et al., 1980; Pearce, 1982; Pearce et al., 1984). in the order of increasing incompatibility with respect to mantle Subduction-zone models hold that fl uids and/or siliceous melts mineralogy. Average compositions for N-MORB, E-MORB, and derived from the subducting oceanic slab carry high concen- OIB, taken from Sun and McDonough (1989), are shown on Fig- trations of LILE (±light rare earth elements [LREEs]) that Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 197

1200 600

MORB SiO2 MORB TiO2 MAR MAR 1000 A 2469 analyses EPR 500 D 2469 analyses IR EPR JFR IR JFR 800 400

600 300

400 200 Number of Analyses Number of

200 Analyses Number of 100

0 0 40 44 48 52 56 60 64 68 72 76 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8

SiO2 (wt%) TiO2 (wt%)

50 140

Back Arc Basin SiO2 Back Arc Basins TiO2 BAB 106 analyses 120 E 103 analyses 40 B BAB 100

30 80

20 60

40 10 Number of Analyses Number of Number of Analyses Number of 20

0 0 40 44 48 52 56 60 64 68 72 76 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8

SiO2 (wt%) TiO2 (wt%)

250 400

Primitive Arcs SiO2 Arc TiO 350 F 2 1657 analyses Marianas 1657 analyses Tonga 200 C 300 Tonga Marianas

150 250

200 100 150

100 50 Number of Analyses Number of Number of Analyses Number of 50

0 0 40 44 48 52 56 60 64 68 72 76 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8

SiO2 (wt%) TiO2 (wt%)

Figure 2. Histograms illustrating SiO2 and TiO2 contents (wt%) in modern mid-ocean-ridge and subduction-zone basalts: (A) SiO2 in modern MORB, n = 2469; (B) SiO2 in modern back-arc basins, n = 106; (C) SiO2 in modern primitive arcs, n = 1657; (D) TiO2 in modern MORB, n = 2469; (E) TiO2 in modern back-arc basin basalts, n = 103; (F) TiO2 in modern primitive arcs, n = 1657. See Table 1 for complete major- element summary. BAB—Back Arc Basin, EPR—East Paciﬁ c Rise, IR—Indian Ridge, JFR—Juan de Fuca Ridge, MAR—mid-Atlantic Ridge.

metasomatize the overlying mantle wedge and aid in lowering somatism composition of the mantle wedge. Consequently, the solidus temperatures. In addition to contributions from the slab trace-element signature of subduction-related magmas appears itself (i.e., dehydration reactions in altered oceanic crust), sub- to be derived from three main sources: the subducted oceanic duction ﬂ uids may carry elemental contributions from a variety lithosphere, subducted sediment, and the mantle wedge over- of subducted sediments. HFSE and heavy (H) REE concentra- lying the subducting slab (Perﬁ t et al., 1980; Pearce, 1982; tions, on the other hand, are controlled primarily by the premeta- Arculus and Powell, 1986; Davidson, 1996). 198 Metcalf and Shervais

Rock/NMORB 10 100 A B

OIB OIB 10 1 EMORB NMORB Th/Yb 1 EMORB Figure 3. (A–B) Trace-element data for 0.1 average normal (N) mid-ocean-ridge 0.1 basalt (MORB), enriched (E) MORB, NMORB and ocean-island basalt (OIB) dis- 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 played as N-MORB–normalized spider 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb diagrams and Th/Yb-Nb/Yb ratio-ratio Rock/NMORB plots. (C–D) Trace-element data for 10 typical subduction-zone basalts dis- C 100 Calc- D played as N-MORB–normalized spider Tholeiite Calc- alkaline Tholeiite diagrams and Th/Yb-Nb/Yb ratio-ratio alkaline OIB 1 plots. 10 Low-K Th/Yb tholeiite 1 EMORB 0.1 0.1 Low-K tholeiite NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb

Three samples from the South Sandwich Island arc are arcs (the initiation of back-arc basin formation), and magmatic plotted (Pearce et al., 1995) in Figures 3C and 3D: a calc-alkaline “forearcs” that form above nascent, retreating intraoceanic basalt, a tholeiite basalt, and a low-K tholeiite basalt. Elevated trenches (Fig. 1). In Figure 4, we have plotted forearc, arc, and concentrations of slab-derived components (Cs, Rb, Ba, Th, U, K, back-arc trace-element data for magmatic rocks from three mod- La, Ce, Pb, Sr) are superimposed on conservative mantle wedge- ern oceanic subduction zones: Izu-Bonin-Mariana, New Britain– derived components (Nb, Zr, Sm, Eu, Ti, Dy, Y, Yb, Lu), pro- Manus, and Lau-Tonga. All of the samples in the data set have ducing the distinctive pattern of by “spikes” and “troughs” that basalt or basaltic andesite compositions and classify as low-K is characteristic of suprasubduction-zone basalts (Fig. 3C). For tholeiites or boninites. The Izu-Bonin-Mariana data set includes the calc-alkaline basalt, the mantle wedge-derived components Eocene-age forearc basalt and gabbro formed during initiation of approximate N-MORB compositions, suggesting a mantle wedge subduction (Bloomer et al., 1995), Quaternary basalts from the with DMM composition (N-MORB normalized values for Nb, active arc (Elliott et al., 1997), and Quaternary basalts from Nd-Lu ~ 1 in Fig. 3C). The tholeiite basalt and a low-K tholeiite the active back-arc spreading ridge (Mariana Trough: Tian et al., samples, however, have conservative mantle wedge-derived 2005; Pearce et al., 2005). The New Britain–Manus data set components more depleted than N-MORB (Fig. 3C). A particu- (Woodhead et al., 1998) includes trench proximal (forearc) and lar feature of these basalts is that the most incompatible mantle- arc basalts from New Britain and back-arc basalts of the active derived components (e.g., Nb) typically are more depleted than Manus spreading ridge, all Quaternary in age. The Lau-Tonga less incompatible mantle-derived components (e.g., Y, Yb). Such data set includes Miocene forearc basalt and gabbro formed patterns cannot be produced by simple variations in the degree during initiation of subduction, modern basalts from the active of partial melting and must represent a residual MORB mantle, Tonga arc, and Pliocene to Quaternary basalts from the western i.e., one that has experienced a previous MORB melting event Lau back-arc basin (Ewart et al., 1994). (Pearce and Parkinson, 1993). The Nb/Yb versus Th/Yb plot pro- Variations in the compositions of suprasubduction-zone vides a means for evaluating the mantle wedge contribution to basalts can be attributable to two effects: (1) differences in the suprasubduction-zone basalt trace-element budget (Pearce, the magnitude of the subduction fl ux of LILEs and (2) verti- 1982). Addition of subduction-derived fl uid to the mantle wedge cal and/or lateral variations in the composition of the mantle increases the Th/Yb ratio, but not the Nb/Yb ratio, producing the wedge that control HFSEs (Taylor et al., 1992). The data in subduction component vectors shown in Figure 3D. Extrapola- Figure 4 illustrate some of these variations. Although nearly tion back along a subduction component vector provides an indi- all of the samples show signifi cant evidence of subduction- cation of presubduction composition of the mantle source. related LILE fl ux (prominent spikes and troughs in Figs. 4A, Suggested tectonic settings for the generation of supra- 4C, 4E, and 4G; elevated Th/Yb ratios in Figs. 4B, 4D, 4F, subduction-zone ophiolites include back-arc basins, rifted island and 4H), some back-arc samples from each of the subduc- Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 199

Rock/NMORB 10 100 A B

OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 Forearc Forearc NMORB 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu 0.01 0.1 1 10 100 Rock/NMORB Nb/Yb 10 100 C D

OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 Arc Arc NMORB Figure 4. Trace-element variation for 0.01 normal (N) mid-ocean-ridge basalt Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu 0.01 0.1 1 10 100 (MORB), enriched (E) MORB, and Nb/Yb Rock/NMORB ocean-island basalt (OIB) observed in: 10 (A–B) Cenozoic forearcs, (C–D) Quater- 100 E F nary arcs, (E–F) nascent back-arc basins, and (G–H) mature back-arc basins. OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 Backarc Backarc NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb Rock/NMORB 10 G 100 H Izu - Bonin - Mariana New Britain - Manus OIB 10 1 Th/Yb Lau - Tonga 1 EMORB 0.1 0.1 Backarc Backarc NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb Rb Th Nb LaPb Nd Ti Y Lu 0.01 0.1 1 10 100 Sr Sm Nb/Yb

tion zones and three arc samples (two New Britain and one In general, basalts erupted in a forearc setting, including Tonga) plot on or near the mantle depletion-enrichment array boninites, are among the most depleted rocks on Earth. The on the Nb/Yb versus Th/Yb plot. Differences in the (presub- forearc basalts from all three subduction zones are depleted duction ﬂ ux) composition of the mantle wedge are seen on the in HFSEs relative to average N-MORB (NMORB–normal- Nb/Yb versus Th/Yb plot as a broad range of Nb/Yb ratios ized values for Nb << 1.0, for Nd-Lu < 1.0, Fig. 4A; typically and on the spider diagrams as a variable negative Nb anomaly Nb/Yb < 0.8, Fig. 4B), which is consistent with derivation from a and variable depletions in HFSEs (Nd through Lu) relative to residual MORB mantle (RMM) source. The data for the arc basalts N-MORB normalization. vary in HFSEs from depleted to enriched (Nb/Yb ≈ 0.15–1.1, 200 Metcalf and Shervais

Figs. 4C and 4D). The Tonga arc basalts have uniformly depleted ter and the subducting slab; basalts erupted in proximal back- HFSE patterns, with Nb/Yb ratios below average N-MORB, arc basins typically have arc-like trace-element patterns with an consistent with a RMM source. Arc basalts from both the Mari- RMM or DMM mantle component, while more distal (wider) ana and New Britain systems show a range of HFSE patterns, back-arc basins show progressively less subduction inﬂ uence and from depleted to more enriched values, between those of aver- more N-MORB-like and E-MORB-like trace-element composi- age N-MORB and E-MORB. With the exception of a few sam- tions (cf. Figs. 4E and 4F with Figs. 4G and 4H). ples from the New Britain arc, the arc basalts exhibit signiﬁ cant subduction-related LILE enrichment. As a group, the back-arc OPHIOLITE GEOCHEMISTRY basalts show the greatest variability both in the magnitude of LILE enrichment (Th/Yb ratios in Figs. 4B, 4D, 4F, and 4H) and We have compiled whole-rock geochemical data from the range of mantle enrichment-depletion in HFSEs (Nb/Yb ratio eight well-studied ophiolites to illustrate the range of compo- in Figs. 4B, 4D, 4F, and 4H). In general terms, back-arc basin sitions expressed in the ophiolite record. Concentrations of basalt compositions vary with proximity to the spreading cen- the major elements are shown in Figure 5 and summarized in

25 70

A SSZ Ophiolite SiO2 B 60 SSZ Ophiolites TiO2 n = 195 20 n = 195 50 Trinity 15 Betts Cove 40 Troodos Trinity Betts Cove 30 10 Troodos

20 Number of Analyses Number of Number of Analyses 5 10

0 0 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6

SiO2 TiO2

30 40 C D Mixed Ophiolite SiO2 35 Mixed Ophiolites TiO2 Figure 5. Histograms illustrating SiO2 25 n = 185 n = 185 and TiO2 contents (wt%) in ophiolite 30 basalts: (A–B) SiO2 and TiO2 in true 20 Josephine 25 suprasubduction-zone (SSZ) ophiolites, Bay of Islands n = 195; (C–D) SiO and TiO in mixed Oman Josephine 2 2 15 20 suprasubduction zone–mid-ocean-ridge Bay of Islands Oman basalt (MORB) ophiolites, n = 185; 15 10 (E–F) SiO2 and TiO2 in MORB-only Number of Analyses ophiolites, n = 29. See Table 2 for com- Number of Analyses 10 5 plete summary of major elements in 5 these ophiolites. 0 0 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6

SiO2 TiO2 8 8 “MORB” Ophiolite SiO “MORB” Ophiolite TiO2 7 E 2 7 F n = 29 n = 29 6 6 Macquarie Island Macquarie Island Pindos 5 Pindos 5

4 4

3 3 Number of Analyses 2 Number of Analyses 2

1 1

0 0 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6

SiO2 TiO2 Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 201

Table 2, which reports their mean, minimum, maximum, and Island ophiolite (Kamenetsky et al., 2000; Varne et al., 2000) standard deviation. Trace-element geochemical data are plotted is from the southwest Paciﬁ c Ocean. The data set includes in Figures 6, 7 and 8; geologic and geochemical characteristics examples of ophiolites with clear suprasubduction zone com- of the ophiolites are summarized in Table 3. positions (Trinity, Betts Cove, Troodos), ophiolites with mixed Tethyan ophiolites are represented by the Cretaceous Troodos suprasubduction-zone and MORB compositions (Josephine, ophiolite (Robinson et al., 1983; Robinson and Malpas, 1990; Oman, Bay of Islands), and ophiolites with MORB compositions Malpas and Langdon, 1984; Cameron, 1985; Rautenschlein (Argolis, Macquarie Island). et al., 1985; Laurent, 1992; Baragar et al., 1990), the Cretaceous Oman ophiolite (Glennie et al., 1974; Alabaster et al., 1982; Suprasubduction-Zone Compositions in Ophiolites Lippard et al., 1986; Ernewein et al., 1988; Rochette et al., 1991; Nehlig, 1993; Ishikawa et al., 2002), and the Triassic Argolis The Trinity, Betts Cove, and Troodos ophiolites all show (Pindos basin) ophiolite (Saccani et al., 2003). Cor dillera ophio- similar suprasubduction-zone compositions. All three ophiolites are represented by the Silurian-Devonian Trinity ophiolite lites exhibit a broad range of silica compositions dominated by (Wallin and Metcalf, 1998; Metcalf et al., 2000) and the Jurassic basalt to basaltic andesite, but they also include more silica-rich Josephine ophiolite (Harper, 2003a, 2003b). The Ordovician Bay dacite and rhyolite compositions. The basaltic rocks are low in of Islands ophiolite (Jenner et al., 1991) and the Ordovician titanium—typically less than 1.2 wt% TiO2—and are classiﬁ ed Betts Cove ophiolite (Bedard et al., 1998; Bedard, 1999) repre- as low-K tholeiites to boninites (Fig. 5; Table 2). In general, the sent Appalachian (Iapetus) ophiolites. The Miocene Macquarie HFSEs are more depleted than N-MORB, with Nb/Yb ratios at

TABLE 2. MEAN COMPOSITIONS, STANDARD DEVIATIONS, MAXIMUMS, AND MINIMUMS FOR SUPRASUBDUCTION-ZONE (SSZ) OPHIOLITES (TRINITY, BETTS COVE, TROODOS), MIXED OPHIOLITES (JOSEPHINE, OMAN, BAY OF ISLANDS), AND MID-OCEAN-RIDGE BASALT (MORB) OPHIOLITES (ARGOLIS, MACQUARIE ISLAND) Trinity Betts Cove Troodos Mean Std-dev Max Min Mean Std-dev Max Min Mean Std-devMax Min

SiO2 58.32 10.27 78.35 46.8 54.09 6.10 79.76 43.42 56.47 3.21 66.61 51.34 TiO2 0.428 0.2476 1.273 0.063 1.10 0.79 2.94 0.07 0.31 0.41 1.93 0.01 Al2O3 16.38 1.9418 10.98 19.54 15.50 2.04 20.96 8.91 1.06 0.28 1.51 0.45 FeO* 5.282 2.3265 0 0 9.35 2.57 20.58 0.75 15.70 0.90 18.21 13.45 MnO 0.111 0.0783 0.6 0. 013 0.18 0.09 1.16 0.03 10.02 1.78 14.87 5.94 MgO 6.297 4.3338 15.26 0.023 7.67 3.71 26.70 0.22 0.16 0.06 0.37 0.06 CaO 8.828 4.0171 18.05 1.463 7.70 3.19 16.16 0.08 5.37 1.52 9.06 1.57 Na2O 2.666 1.6295 6.443 0.011 3.64 1.50 7.31 0.04 6.82 2.70 17.39 1.77 K2O 0.126 0.1797 1.112 0.01 0.68 1.04 5.40 0.00 4.01 1.42 6.91 1.43 P2O5 0.10 0.13 1.02 0.00 0.08 0.03 0.16 0.03 n = 63 n = 214 n = 96

Josephine Oman V1 (Geotimes) Oman V2 (Lasail-Alley) Mean Std-dev Max Min Mean Std-dev Max Min Mean Std-devMax Min

SiO2 53.15 4.45 66.61 41.87 54.66 4.58 69.21 45.97 58.02 9.58 81.57 40.84 TiO2 1.26 0.85 3.46 0.23 1.46 0.39 2.44 0.51 0.65 0.26 1.28 0.20 Al2O3 14.80 1.72 18.03 10.22 15.47 1.39 18.49 11.64 14.73 2.09 19.32 8.41 FeO* 9.73 2.93 17.71 2.67 9.46 1.47 12.83 4.94 7.37 2.21 12.36 1.57 MnO 0.19 0.07 0.35 0.04 0.19 0.07 0.50 0.10 0.15 0.07 0.47 0.03 MgO 8.05 3.79 18.75 1.61 3.98 1.50 7.70 0.70 5.21 3.19 15.29 0.14 CaO 7.86 2.35 13.16 3.09 6.97 3.32 20.95 1.91 8.29 5.08 24.36 0.69 Na2O 3.90 1.47 8.33 0.99 5.48 1.23 8.31 0.22 3.63 1.63 7.37 0.0 K2O 0.62 0.65 2.99 0.01 0.30 0.36 1.88 0.01 0.50 0.96 7.91 0.0 P2O5 0.13 0.07 0.34 0.02 0.19 0.07 0.37 0.04 0.10 0.14 1.10 0.0 n = 49 n = 102 n = 76

Bay of Islands Argolis Macquarie Island Mean Std-dev Max Min Mean Std-dev Max Min Mean Std-devMax Min

SiO2 61.56 10.99 79.90 43.83 48.77 3.43 58.55 43.53 49.34 0.73 51.14 47.37 TiO2 0.98 0.83 3.44 0.12 1.60 0.77 2.89 0.21 1.51 0.22 2.10 0.97 Al2O3 15.83 2.69 19.75 11.07 15.05 0.87 16.55 12.82 16.98 0.69 18.22 15.03 FeO* 5.73 3.76 14.47 0.58 10.22 1.84 13.58 7.76 7.92 0.67 10.17 6.81 MnO 0.12 0.11 0.60 0.01 0.23 0.13 0.48 0.10 7.34 0.69 8.75 5.65 MgO 3.49 3.03 8.48 0.21 6.76 1.11 10.43 4.09 0.13 0.03 0.18 0.07 CaO 5.30 3.75 13.77 0.32 10.23 1.68 13.12 4.90 11.43 0.92 13.53 9.81 Na2O 5.59 1.79 9.48 2.75 3.61 0.86 4.98 1.08 3.11 0.43 4.24 2.37 K2O 1.19 2.19 11.01 0.01 0.39 0.41 1.39 0.04 0.74 0.34 1.76 0.12 P2O5 0.32 0.12 0.66 0.08 n = 40 n = 23 n = 55 Note: Data are from: Metcalf et al. (2000), Bedard (1999), Robinson et al. (1983), Robinson and Malpas (1990), Malpas and Langdon (1984), Cameron (1985), Rautenschlein et al. (1985), Laurent (1992), Baragar et al. (1990), Harper (2003a, 2003b), Alabaster et al. (1982), Lippard et al. (1986), Einaudi et al. (2000), Jenner et al. (1991), Saccani et al. (2003), Kamenetsky et al. (2000). FeO*—total Fe as FeO. 202 Metcalf and Shervais or less than N-MORB values, while LILE/HFSE ratios are ele- settings, including a forearc, a back-arc, or a marginal basin vated (Figs. 6A–6F). Data from these three ophiolites are most similar to the modern Andaman Sea (McCulloch and Cameron, similar to modern forearcs and suggest melting of a residual 1983; Gass et al., 1984; Moores et al., 1984). Indeed, it was MORB mantle enriched by subduction ﬂ uids. Geologic con- the major-element geochemistry of Troodos volcanic rocks that straints for both the Betts Cove and Trinity ophiolites are consis- led Miyashiro (1973) to challenge the mid-ocean-ridge origin tent with generation in a suprasubduction-zone forearc setting of Troodos (see Cann [2003] and Robinson et al. [2003], for (Bedard, 1999; Metcalf et al., 2000). Generation of the Troodos discussions of the importance of Troodos in the development of ophiolite has been ascribed to various suprasubduction-zone the ophiolite concept).

Rock/NMORB 10 A B 100 Trinity ophiolite Trinity ophiolite

OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb Rock/NMORB 10 C D 100 Betts Cove ophiolite Betts Cove ophiolite

OIB 10 1 Th/Yb 1 EMORB 0.1 Figure 6. Trace-element data for normal 0.1 (N) mid-ocean-ridge basalt (MORB), NMORB enriched (E) MORB, and ocean-island 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb basalt (OIB) from ophiolites with supra- 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb subduction-zone trace-element chemis- Rock/NMORB 10 try: (A–B) Trinity, (C–D) Betts Cove, EF (E–F) Troodos; and (G–H) a mixed 100 Troodos ophiolite Troodos ophiolite suprasubduction zone–mid-ocean-ridge basalt (MORB) ophiolite: Bay of Islands. OIB 10 1 Th/Yb 1 EMORB 0.1

0.1 NMORB

0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb Rock/NMORB 10 G H 100 Bay of Islands ophiolite Bay of Islands ophiolite

OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 203

Mixed Suprasubduction Zone and Mid-Ocean Ridge rocks in the lower lava sequence have low TiO2 values (typically Compositions in Ophiolites < 1.2 wt%), HFSE concentrations generally less than average N-MORB, and elevated LILE/HFSE ratios. Basaltic rocks of the

The Josephine, Oman, and Bay of Islands ophiolites are all upper lava sequence have higher TiO2 values (typically > 1.2 wt%) examples of ophiolites that exhibit both suprasubduction-zone and and HFSE concentrations and LILE/HFSE ratios that approximate

MORB geochemical signatures (Fig. 5; Table 2). In the Josephine N-MORB. Overall SiO2 values of both the lower and upper lavas ophiolite, suprasubduction zone–like lavas are overlain by MORB- range between 47 wt% and 53 wt%, with a few samples at higher like lavas in the volcanic stratigraphy (Figs. 7A–7D). Basaltic values (up to 62 wt%). The Josephine ophiolite occupies a paleo-

Rock/NMORB 10 A Josephine ophiolite B Josephine ophiolite 100 Lower lavas & Lower lavas & intrusions intrusions OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb Rock/NMORB 10 100 C Josephine ophiolite D Josephine ophiolite Upper lavas Upper lavas OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 NMORB Figure 7. Trace-element data for normal (N) mid-ocean-ridge basalt (MORB), 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 enriched (E) MORB, and ocean-island Rb Th Nb LaPb Sr Nd Sm Ti Y Lu 0.01 0.1 1 10 100 Nb/Yb basalt (OIB) from ophiolites with mixed Rock/NMORB 10 mid-ocean-ridge and suprasubduction- E Oman ophiolite F Oman ophiolite zone trace-element chemistry: (A–D) 100 Josephine upper and lower lavas and Lower lavas Lower lavas (E–H) Oman upper and lower lavas. OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb Rock/NMORB 10 100 G Oman ophiolite H Oman ophiolite Upper lavas Upper lavas OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb 204 Metcalf and Shervais position between two segments of a rifted Jurassic arc system and In the Oman ophiolite, MORB-like lavas (Figs. 7E–7F) records spreading in an extensional back-arc setting where supra- are overlain by suprasubduction zone–like lavas (Figs. 7G–7H) subduction zone–like magmas give way to MORB-like magmas in the volcanic stratigraphy. Basaltic to andesitic rocks of the

(Harper, 2003a, 2003b), similar to that seen in the modern Lau lower lava have high TiO2 values (typically > 1.2 wt%), and Basin and Mariana Trough (Hawkins, 2003; Pearce et al., 2005). HFSE concentrations and LILE/HFSE ratios that are similar to

Rock/NMORB 10 AB 100 Pindos basin Pindos basin

OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 Figure 8. Trace-element data for normal NMORB (N) mid-ocean-ridge basalt (MORB), 0.01 0.01 enriched (E) MORB, and ocean-island Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 basalt (OIB) from ophiolites with mid- Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb ocean-ridge trace-element chemistry: Rock/NMORB 10 (A–B) Pindos and (C–D) Macquarie C D Island (actually an uplifted segment of 100 Macquarie Island ophiolite Macquarie Island ophiolite ridge, not a true ophiolite). OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb

TABLE 3. SUMMARY OF MAIN CHARACTERISTICS OF EIGHT OPHIOLITES DISCUSSED IN TEXT Ophiolite age Locality Reported Basalt Trace-element signature Mantle Frac. Cover sequence Paleotectonic (Ma) range of SiO2 geochemistry source sequence interpretation (wt%) Josephine Cordillera 46–58 Lower: Lower: Lower: B1 Volcaniclastic SSZ (162–164) Low-Ti thol. High LILE/HFSE RMM back arc Boninite HFSE < N-MORB Upper: Upper: Upper: DMM High-Ti thol. LILE/HFSE ~ N-MORB HFSE ~ N-MORB Trinity Cordillera 46–57 Low-Ti thol. High LILE/HFSE RMM B1 Volcaniclastic SSZ (431–398) 71–78 HFSE < N-MORB forearc Betts Cove Appalachia 46–59 Low-Ti thol. High LILE/HFSE RMM B1 Volcaniclastic SSZ (489) Boninite HFSE < N-MORB forearc Bay of Islands Appalachia 48–55 Low-Ti thol. Elevated LILE/HFSE FMM to EMM A1, B1 Clastic Mature back arc (484) 60–64 High-Ti thol. HFSE ~ N-MORB or 72–78 Mid-ocean ridge Troodos Tethys 49–65 Low-Ti thol. High LILE/HFSE RMM to FMM B1,2 Chert overlain by SSZ (92–90) Boninite HFSE < N-MORB to ~ N-MORB marine carbonate forearc Oman Tethys 45–77 Lower: Lower: DMM to RMM Primary: Ophiolite breccia, Mid-ocean ridge (97–94) High-Ti thol. Elevated LILE/HFSE B1,2 marine carbonate and Upper: HFSE ~ N-MORB Minor: SSZ forearc Low-Ti thol. Upper: A1 High LILE/HFSE HFSE < N-MORB Argolis Tethys High-Ti thol. LILE and HFSE DMM to EMM A Radiolarian cherts Mid-ocean ridge (Triassic) ~ N-MORB to E-MORB Macquarie Pacific 47–51 High-Ti thol. LILE ~ E-MORB to OIB EMM to OIB A1,2 Volcaniclastic w/ Mid-ocean ridge (9) HFSE ~ E-MORB to OIB ophiolite clasts; marine carbonate Note: MORB—mid-ocean-ridge basalt; LILE—large ion lithophile element; HFSE—high field strength element; N—normal; E—enriched; OIB—ocean-island basalt. Mantle source: DMM—depleted MORB mantle; RMM—residual MORB mantle; EMM—enriched MORB mantle. Fractionation sequence: A—olivine > plagioclase > clinopyroxene; B—olivine > clino/orthopyroxene > plagioclase > hornblende. 1—Based on cumulate sequences; 2—based on phenocryst assemblages. Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 205

N-MORB. Highly variable K contents in the lower lavas may Macquarie Island is a slice of Miocene ocean ﬂ oor that result from hydrothermal alteration. Basaltic to andesitic rocks lies above sea level in the Southern Ocean south of Tasmania

of the upper lava sequence have lower TiO2 values (typically (Varne et al., 2000). It can be argued that Macquarie Island is < 1.2 wt%), HFSE concentrations that are generally less than aver- not, strictly speaking, an ophiolite because it has not yet been age N-MORB, and elevated LILE/HFSE ratios. While a majority emplaced into continental or arc crust. Some workers, however,

of volcanic and plutonic rocks in the Oman ophiolite have SiO2 have regarded Macquarie Island as an example of a MORB-type values between 47 wt% and 58 wt%, a number of samples range ophiolite (e.g., Dilek, 2003); we include it here as such. Ocean

between 60 wt% and 78 wt% SiO2 (Fig. 5). The paleotectonic ﬂ oor exposed on Macquarie Island was generated by spreading setting of the lower lavas in the Oman ophiolite remains contro- on the Australia-Paciﬁ c spreading ridge at 9–14 Ma, and it was versial (e.g., Ernewein et al., 1988; Shervais, 2001). exposed during transpression that formed the Macquarie Ridge Although the Bay of Islands ophiolite exhibits both supra- in the last 10 m.y. (Varne et al., 2000; Kamenetsky et al., 2000). subduction zone–like and MORB-like geochemical features The Macquarie Island ophiolite is dominated by basaltic com-

(Figs. 6G–6H), there is no clear stratigraphic sequence to positions (48–51 wt% SiO2) with TiO2 contents of 1–2 wt%. the magma types (Jenner et al., 1991). Overall, the ophiolite Basalts from Macquarie Island are highly enriched in incom- records a broad range of silica values with three modes, 48–55, patible trace elements, ranging from compositions equivalent

60–64, and 72–78 wt% SiO2 (Table 3). Unlike the Josephine to average E-MORB to compositions more enriched than OIB, and Oman ophiolites, which show trace-element compositions and they plot along the depletion-enrichment array on the ratio- defi ning two distinct magma types, basaltic rocks of the Bay of ratio plot (Fig. 8D). Trace-element patterns on the spider dia- Islands ophiolite exhibit a continuum of compositions between grams closely parallel those of average E-MORB and average suprasubduction-zone and MORB types. For example, TiO2 val- OIB (Fig. 8C). ues span a broad range from low-Ti, suprasubduction zone–like values (<1.2 wt%) to high-Ti MORB-like values (1.2–2.2 wt%; ANOMALOUS “SUPRASUBDUCTION ZONE–LIKE” Tables 2 and 3). A majority of the basalt samples show vary- BASALTS AT ERUPTED MID-OCEAN RIDGES ing degrees of subduction enrichment (variably elevated LILE/HFSE ratios, Figs. 6G–6H) superimposed on a mantle- A major thrust of the historical contingency model focuses derived component (HFSE) that varies between N-MORB and on examples of modern mid-ocean ridges where “suprasub- E-MORB. Another subset of basalt samples shows little or no duction zone–like” trace-element signatures have been reported. subduction enrichment (Figs. 6G–6H), and several samples Each of these examples represents a ridge-trench-trench triple approximate E-MORB and OIB compositions. Generation of junction where an active spreading ridge intersects an active sub- the Bay of Islands ophiolite has been ascribed to a back-arc duction zone, potentially permitting communication of MORB basin (Jenner et al., 1991). and suprasubduction-zone mantle source regions via a slab window. Three such examples were discussed by Moores et al. Mid-Ocean-Ridge Compositions in Ophiolites (2000): the Chile Ridge (Klein and Karsten, 1995; Karsten et al., 1996; Sturm et al., 1999, 2000), the Juan de Fuca Ridge Ophiolites exhibiting strictly MORB geochemical sig- (Cousens et al., 1995), and the Woodlark spreading center (Perfi t natures are rare and are known largely from thrust slices and et al., 1987; Staudigel et al., 1987). mélange blocks in accretionary complexes. Tethyan ophiolites from the Eastern Mediterranean region provide the best record Chile Ridge of MORB compositions in the ophiolite record. Saccani et al. (2003) reported MORB-like compositions from the Triassic Klein and Karsten (1995) fi rst reported suprasubduction Argolis ophiolite in Greece (Figs. 8A–8B). The Argolis ophio- zone–like trace-element signatures for basalts collected from four lite is one of several ophiolitic massifs that provide a record of active segments of the Chile Ridge adjacent to the ridge-trench- the Mesozoic Pindos ocean basin, which formed during rift- trench with the Andean subduction zone. Subsequent papers ing along the northern margin of Gondwanaland. The Triassic (Karsten et al., 1996; Sturm et al., 1999, 2000) have reported Argolis ophiolite massif provides a record of the early develop- Sr-Nd-Pb isotopic data and discussed the occurrence in the con- ment of the Pindos basin. The Argolis ophiolite is dominated text of the ophiolite conundrum. Trace-element data for the Chile

by basaltic compositions (45–51 wt% SiO2, with a few samples Ridge are shown in Figures 9A and 9B. Enrichments of LILEs

at ~60 wt%) and TiO2 contents at 1.4–2.8 wt% (Saccani et al., (Cs, Rb, Ba, Th, K, Pb, and Sr) are evident on both the Th/Y-Nb/Y 2003). Trace-element ratios Th/Y-Nb/Y plot along the mantle ratio plot and the spider diagrams, superimposed on mantle com- depletion-enrichment array between N-MORB and E-MORB, ponents that vary from slightly depleted N-MORB to E-MORB. albeit with slightly elevated Th/Yb ratios (Figs. 8A–8B). Trace- A few samples show, relative to average N-MORB, a weak Nb element patterns on the spider diagram are transitional between negative anomaly and a slight enrichment of LILE. Samples at N-MORB and E-MORB; element mobility is apparent in a few the other end of the spectrum show, relative to E-MORB, slight elements (e.g., K, U). enrichments in LILEs that also produce a weak negative Nb 206 Metcalf and Shervais anomaly. Although present, the suprasubduction-zone signature zone–like component at an active mid-ocean-ridge spreading in Chile Ridge basalts is subtle when compared to that of either center because there is very little about the composition of these modern subduction zones or the suprasubduction-zone ophiolite basalts that are suprasubduction zone–like. The Juan de Fuca data record (cf. Figs. 9A and 9B to Figs. 6, 7, and 8). do provide additional evidence that mantle source regions modi- ﬁ ed by subduction recycling do not necessarily give rise to basalts Juan de Fuca Ridge with suprasubduction zone–like trace-element compositions.

Cousens et al. (1995) reported geochemical and isotopic data Woodlark Basin from the West Valley segment of the Juan de Fuca Ridge near the ridge-trench-trench junction with the North American plate. Perfi t et al. (1987) and Staudigel et al. (1987) reported Trace-element patterns on the spider diagrams parallel those of geochemical data for recent basalts collected from the active E-MORB and show no enrichment in LILEs relative to HFSEs Woodlark basin spreading ridge. In addition to its position at a (Fig. 9C). On the Th/Y-Nb/Y ratio plot, the Juan de Fuca data plot ridge-trench-trench triple junction, recent subduction reversal along the enrichment-depletion centered on E-MORB (Fig. 9D). places the Woodlark spreading ridge over mantle that has for- Cousens et al. (1995) used Nd-Sr-Pb isotopic compositions to merly been modifi ed by subduction processes (Perfi t et al., 1987; evaluate mantle source regions and found evidence of a hetero- Staudigel et al., 1987). Woodlark basalts have Nb/Yb ratios simi- geneous source formed by mixing of DMM and HIMU compo- lar to N-MORB, and some samples show slightly elevated Th/Yb nents. This locality provides no evidence for a suprasubduction ratios; trace-element patterns on the spider diagram approximate

Rock/NMORB 10 100 A Chile Ridge B Chile Ridge

OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 NMORB 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb Rock/NMORB 10 C Juan de Fuca Ridge D Juan de Fuca Ridge Figure 9. Trace-element data for normal 100 (N) mid-ocean-ridge basalt (MORB), OIB enriched (E) MORB, and ocean-island 10 1 basalt (OIB) for basaltic rocks collected Th/Yb from modern mid-ocean ridges with re- 1 EMORB ported suprasubduction zone–like com- 0.1 positions: (A–B) Chile Ridge, (C–D) 0.1 Juan de Fuca Ridge, and (F–G) Wood- NMORB lark basin. 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb Rock/NMORB 10 100 E Woodlark basin F Woodlark basin

N-MORB (Figs. 9E–9F). The Woodlark basalts carry a weak sub- scales. Isotopic data point to subduction recycling as a major duction signature similar to that seen in mature back-arc basins contributor to MORB source heterogeneity, including most of (cf. Figs. 9E–9F with Figs. 4E–4H). the isotopic components identifi ed in the historical contingency model (e.g., HIMU, EMI, EMII). Thus, a test of the historical MANTLE HETEROGENEITY: TRACE-ELEMENT contingency model can be found in trace-element data from AND ISOTOPIC VARIATIONS IN MODERN modern mid-ocean-ridge basalts that carry isotopic evidence of MID-OCEAN-RIDGE BASALTS subduction contamination of their mantle source regions. In the following sections, we review data from several recent trace- A main thesis of the historical contingency model is that element and isotopic studies of mid-ocean-ridge basalts. Despite mantle source regions exhibit long-lived heterogeneities related trace-element and isotopic evidence for a heterogeneous source to prior history, specifi cally the recycling of oceanic lithosphere related to recycled (subducted) oceanic lithosphere, these basalts into the mantle via subduction. The model further suggests that bear little resemblance to basalts from either modern subduction while modern mid-ocean ridges primarily tap DMM mantle zones or the suprasubduction-zone ophiolite record. sources, past mid-ocean ridges may have tapped subduction- modifi ed mantle heterogeneities, thus making geochemical North Chile Ridge data unreliable as an indicator of tectonic setting and by infer- ence ophiolite discrimination. Early recognition of subduction- Recent basalts from the North Chile ridge (latitude 37–39°S; modifi ed mantle heterogeneities was found largely in isotopic Bach et al., 1996) provide an example of mid-ocean-ridge basalts data from ocean-island and plume related basalts (Zindler and that have more depleted compositions than average N-MORB Hart, 1986; Hart, 1988) as noted by Moores et al. (2000). An (Fig. 10). These data show a depletion in the most incompatible unstated assumption of the historical contingency model is trace elements (Rb to Nb; Fig. 10A) and, in particular, a depletion that partial melting of mantle regions carrying isotopic evi- in Nb relative to other HFSEs (e.g., Zr, Hf, Y, Yb). The depleted dence of subduction recycling would produce basalts that carry nature of these basalts relative to average N-MORB is particu- suprasubduction-zone trace-element signatures. larly apparent in the Th/Yb-Nb/Yb ratio plot (Fig. 10B). Bach It is widely recognized that magmas erupted at modern mid- et al. (1996) attributed this depletion to prior removal of a low ocean ridges are quite variable in terms of both trace-element melt fraction from the N-MORB source mantle within the last and isotopic composition. These variations largely refl ect source few million years. Dynamic melting, i.e., episodic melt extrac- hetero geneities and are observed at both regional (ocean basin) tion from a common source during a single protracted melting and local (single ridge segment or adjacent ridge segments) event, is capable of producing magmas from a common source

North Chile Ridge Rock/NMORB 10 A B 100

OIB 10 1

Th / Y b 1 EMORB 0.1 0.1 Figure 10. Trace-element and isotopic NMORB data for normal (N) mid-ocean-ridge 0.01 0.01 basalt (MORB), enriched (E) MORB, Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu and ocean-island basalt (OIB) for ba- Nb/Yb saltic rocks from the North Chile mid- ocean ridge. DMM—depleted MORB C HIMU D μ 0.5134 mantle, HIMU—high , EMI—enriched 15.8 DMM EMII man tle I, EMII—enriched mantle II. 0.5132 15.7 Nd Pb

144 0.5130 204 15.6 Nd/ Pb/ 0.5128 HIMU 143 207 15.5 EMI 0.5126 15.4 0.5124 EMII DMM EMI 17.0 18.0 19.0 20.0 21.0 0.702 0.703 0.704 0.705 0.706 206Pb/204Pb 87Sr/86Sr 208 Metcalf and Shervais

that exhibit varying degrees of depletion (Pearce et al., 1995). Nd- broadly divided into two groups (Figs. 12A–12B), an N-MORB Sr-Pb isotopic data (Bach et al., 1996) suggest mixing of DMM, group (fi lled squares) and a group that trends toward more enriched EMII, and possibly HIMU mantle reservoirs (Figs. 10C–10D). compositions, overlapping E-MORB (open squares). Although both groups show slightly elevated Th/Yb ratios, more pronounced East Pacifi c Rise in the N-MORB group, the data generally plot along the mantle depletion-enrichment array on the Th/Y-Nb/Y ratio plot and Recent basalts from the northern East Pacifi c Rise (latitude exhibit clear N-MORB and E-MORB patterns on the spider dia- 10–11°N; Niu et al., 1999; Regelous et al., 1999) exhibit a com- grams. Nd-Sr-Pb isotope data for the South Atlantic Ridge basalts plete range of compositions between N-MORB and E-MORB provide evidence of a heterogeneous mantle source (le Roux et al., (Figs. 11A–11B). On the ratio-ratio plot, data plot clearly along 2002). These data suggest mixing of DMM, EMI, EMII, and/or the mantle depletion-enrichment array, where the majority of HIMU isotopic components in the mantle source the South Atlantic samples are slightly more enriched than N-MORB (Fig. 11B). Ridge (Figs. 12C and 12D). le Roux et al. (2002) interpreted these Correlations among trace-element ratios and Nd-Sr-Pb isotopic isotopic signatures to refl ect the infl uence of (1) altered oceanic ratios point to mixing of two mantle components, one more trace lithosphere and pelagic sediment recycled to the shallow mantle element–depleted and one more trace element–enriched (Niu via plumes, (2) remnants of delaminated subcontinental litho- et al., 1999). Both mantle components are linked to recycled spheric mantle, and (3) Mesozoic suprasubduction metasomatism oceanic lithosphere—the depleted source to subducted litho- of mantle beneath Gondwanaland prior to opening of the South spheric mantle and the enriched component to subducted oceanic Atlantic. Trace-element variations discussed previously correlate crust (Niu et al., 1999). The Nd-Sr-Pb isotopic compositions of closely with the observed isotopic variations, providing a link basalts along this section of the East Pacifi c Rise represent a between trace-element compositions and mantle heterogeneities mixture of DMM and EMII mantle reservoirs (Niu et al., 1999), related to subduction recycling (le Roux et al., 2002). although component HIMU mantle cannot be ruled out. Australian-Antarctica Discordance South Atlantic Ridge A section of the Southeast Indian Ridge between the rifted The South Atlantic Ridge was formed by rifting of continental margins of Australia and Antarctica marks the boundary between lithosphere. Trace-element data (le Roux et al., 2002) for a suite of Indian-type and Pacifi c-type mantle domains and has been referred basalts collected from several segments of the active South Atlantic as the Australian-Antarctica discordance. Pyle et al. (1992) used Ridge (latitude 40–52.5°S) show considerable variation but can be isotopic data to map the location of the Australian-Antarctica dis-

East Pacific Rise Rock/NMORB 10 A B 100

OIB 10 1 Th/Yb 1 EMORB 0.1 Figure 11. Trace-element and isotopic 0.1 data for normal (N) mid-ocean-ridge NMORB basalt (MORB), enriched (E) MORB, 0.01 0.01 Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 and ocean-island basalt (OIB) for basal- Rb Th Nb LaPb Sr Nd Sm Ti Y Lu tic rocks from the East Paciﬁ c Rise mid- Nb/Yb ocean ridge. DMM—depleted MORB HIMU mantle, HIMU—high μ, EMI—enriched C D 15.8 0.5134 mantle I, EMII—enriched mantle II. DMM EMII 0.5132 15.7 Pb Nd 204 144 0.5130 15.6 Pb/ Nd/ 0.5128 HIMU 207 143 15.5 EMI 0.5126

15.4 0.5124 EMII DMM EMI 17.0 18.0 19.0 20.0 21.0 0.702 0.703 0.704 0.705 0.706 206Pb/204Pb 87Sr/86Sr Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 209 cordance along the active spreading ridge and recognized several ures 14A and 14B and include samples from the axial rift zone components in the mantle sources. Subsequent work (Pyle et al., (Ramad seamount, Hanish-Zukir Islands) and from the fl anks 1995) used older ocean fl oor basalts to confi rm the existence of of the ocean basin (Hamdan and Jizan volcanic fi elds). Nd-Sr- the Australian-Antarctica discordance during the opening of the Pb isotope data provide evidence for a heterogeneous mantle ocean basin between Australia and Antarctica and to map its late source that includes mixing among several isotopic components Mesozoic-Cenozoic migration. The isotopic composition of Pacifi c- (Figs. 14C–14D). Volker et al. (1997) argued for mixing among a type MORB refl ects a mixture of DMM and HIMU mantle compo- MORB-type mantle (DMM), an Afar plume component (HIMU), nents (Pyle et al., 1992). The isotopic composition of Indian-type and an EMI-EMII hybrid component derived from continental MORB is more complex and requires mixing of DMM, HIMU, lithosphere. Trace-element data for the Red Sea samples show and EMI components (Pyle et al., 1992). Trace-element data mantle enrichments similar to E-MORB and OIB on both the (Pyle et al., 1992, 1995) for the Australian-Antarctica discordance spider diagram and the Th/Y-Nb/Y ratio plot (Figs. 14A–14B). are shown in Figures 13A and 13B, both for active ridge (fi lled crosses) and older ocean fl oor (open crosses) basalts. On the Th/Y- ADDITIONAL EVIDENCE FOR A Nb/Y ratio plot, the data plot along the entire mantle depletion- SUPRASUBDUCTION-ZONE ORIGIN enrichment array, from values more depleted than N-MORB to FOR OPHIOLITES values more enriched than E-MORB. One sample of young basalt has an elevated Th/Yb ratio similar to a subduction component. Workers who prefer a mid-ocean-ridge origin for most or Trace-element data generally exhibit N-MORB and E-MORB pat- all ophiolites attempt to counter the suprasubduction-zone inter- terns on the spider diagrams, and a few samples show positive U pretation of ophiolite genesis by pointing to “...the inadequacy (but not Pb) values. Isotopic data shown in Figures 13C and 13D, of geochemistry itself to determine the tectonic environment...” only for Holocene (active ridge) basalts, are consistent with mixing of individual an ophiolite (Moores, 2003, p. 26). Such criticism of DMM, HIMU, and EMI components (Pyle et al., 1992). attempts to cast reasonable doubt on the validity of geochemical interpretations. However, the suprasubduction-zone interpretation Red Sea is not based solely on geochemical data on volcanic rocks but rather on a range of petrologic and geologic data, of which geochemis- The Red Sea provides an example of seafl oor spreading in try is a major component. Next, we review three additional lines a nascent ocean basin (younger than 5 Ma) formed by continen- of evidence that support a suprasubduction-zone origin for much tal rifting (Volker et al., 1997). Trace-element and isotopic data of the ophiolite record: evidence of wet magmas, the sedimentary for basalts collected from the Red Sea region are shown in Fig- cover on ophiolites, and the issue of ophiolite preservation.

South Atlantic Ridge Rock/NMORB 10

100 A B

OIB 10 1 Th/Yb 1 EMORB 0.1

0.1 Figure 12. Trace-element and isotopic NMORB data for normal (N) mid-ocean-ridge 0.01 0.01 basalt (MORB), enriched (E) MORB, Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb and ocean-island basalt (OIB) for basaltic rocks from the South Atlantic mid- HIMU ocean ridge. DMM—depleted MORB C 0.5134 D mantle, HIMU—high μ, EMI—enriched 15.8 DMM mantle I, EMII—enriched mantle II. EMII 0.5132 15.7 Nd Pb 144 204 0.5130 15.6 Pb/ Nd/ 0.5128 HIMU 207 143 15.5 EMI 0.5126

15.4 0.5124 EMII DMM EMI 17.0 18.0 19.0 20.0 21.0 0.702 0.703 0.704 0.705 0.706 206Pb/204Pb 87Sr/86Sr 210 Metcalf and Shervais

Australia - Antarctica Discordance Rock/NMORB 10 A 100 B OIB 10 1 Th/Yb 1 EMORB Figure 13. Trace-element and isotopic 0.1 data for normal (N) mid-ocean-ridge 0.1 basalt (MORB), enriched (E) MORB, NMORB and ocean-island basalt (OIB) for basal- 0.01 0.01 tic rocks from the Australian-Antarctica Cs Ba U K Ce Pr P Zr Eu Dy Yb 0.01 0.1 1 10 100 Rb Th Nb LaPb Sr Nd Sm Ti Y Lu Nb/Yb discordance section of the Southeast In- dian mid-ocean ridge. DMM—depleted HIMU MORB mantle, HIMU—high μ, EMI— 15.8 C 0.5134 D enriched mantle I, EMII—enriched DMM EMII 0.5132 mantle II. 15.7 Nd Pb 144 204 0.5130 15.6 Pb/ Nd/ 0.5128 HIMU 207 143 15.5 EMI 0.5126

15.4 0.5124 EMII DMM EMI 17.0 18.0 19.0 20.0 21.0 0.702 0.703 0.704 0.705 0.706 206Pb/204Pb 87Sr/86Sr

Evidence of Hydrous Magmas clase. Reaction textures with resorbed olivine, clinopyroxene, and An-rich plagioclase enclosed in hornblende oikocrysts are Water plays a major role in petrologic models for the genesis evidence of the crystallization of wet basalt and are reported in and evolution of subduction-zone magmas, in contrast to water- subduction-related magma systems. poor environment at mid-ocean ridges. As discussed already, In the ophiolite record, crystallization sequences can aqueous ﬂ uids derived from the subducting slab carry LILEs into be determined from phenocryst assemblages in the volcanic the mantle wedge source of suprasubduction-zone basalts, lower section and from cumulate sequences in the plutonic sec- solidus temperatures in the source, and trigger melting. Fraction- tion. In suprasubduction-zone ophiolites, basal cumulate ation of these wet basalts in the lithosphere produces crystalliza- sequences are dominated by dunite, wehrlite, and clino- tion sequences in arc basalts that are in contrast to those observed pyroxenite overlain by pyroxene gabbro and hornblende in MORBs. In addition, the concentration of water vapor in the gabbro (Fig. 15B). Hawkins and Evans (1983) reported well- residual magma during fractionation may lead to the separation displayed suprasubduction-zone cumulate sequences from of a hydrous vapor phase (retrograde boiling) and the formation the Zambales Range in the Luzon ophiolite. Thus, ophiolites of miarolitic cavities in isotropic gabbros of the upper plutonic with suprasubduction-zone geochemistry exhibit crystalliza- series (Fig. 15A). tion sequences (pyroxene before plagioclase) indicative of In MORB, the typical observed crystallization sequence wet magmas (Table 3). Hornblende gabbro is a volumetri- is olivine/spinel > plagioclase > clinopyroxene ± ortho- cally important constituent of suprasubduction-zone ophio- pyroxene (Bryan, 1983; Pearce et al., 1984). In arc basalts, lites and requires water-bearing magmas; in some cases, the typical observed crystallization sequence is olivine/spinel large, decimeter-scale hornblende crystallizes in the isotropic > clinopyroxene/orthopyroxene > plagioclase > hornblende gabbros and diorites where water concentration is highest (Pearce et al., 1984; Cameron, 1985). Experimental results for (Figs. 15C and 15D). This evidence for hydrous magmas in low-pressure (~2 kbar) crystallization of wet tholeiite (Sisson and suprasubduction-zone ophiolites places the geochemical data Grove, 1993) conﬁ rm the role of water in the production of the in a broader petrologic context. This evidence is in sharp con- typical crystallization sequence in arc basalts. These experiments trast to crystallization sequences observed in modern MORB, show that hydrous basalts crystallize olivine, clinopyroxene, and where plagioclase appears before pyroxene and basal cumu- plagioclase (±spinel or magnetite); as magma chemistry evolves lates should be composed of dunite, troctolite, and anorthosite toward more siliceous composition, olivine, clinopyroxene, and with hornblende as only a minor constituent (Pearce et al., An-rich plagioclase become unstable and react with the melt 1984). The MORB ophiolites reviewed here exhibit dry crys- to form orthopyroxene, hornblende, and more Ab-rich plagio- tallization (plagioclase before pyroxene) sequences (Table 3). Red Sea Rock/NMORB 10 A 100 B

OIB 10 1 Th/Yb 1 EMORB 0.1 0.1 Figure 14. Trace-element and isotopic NMORB data for normal (N) mid-ocean-ridge 0.01 0.01 basalt (MORB), enriched (E) MORB, Cs Ba U K Ce Pr P Zr Eu Dy Yb Rb Th Nb LaPb Sr Nd Sm Ti Y Lu 0.01 0.1 1 10 100 and ocean-island basalt (OIB) for Nb/Yb basaltic rocks from the Red Sea mid- HIMU ocean ridge. DMM—depleted MORB mantle, HIMU—high μ, EMI—enriched 15.8 C 0.5134 D DMM mantle I, EMII—enriched mantle II. EMII 0.5132 15.7 Nd Pb 144 204 0.5130 15.6 Pb/ Nd/ 0.5128 HIMU 207 143 15.5 EMI 0.5126

15.4 0.5124 EMII DMM EMI 17.0 18.0 19.0 20.0 21.0 0.702 0.703 0.704 0.705 0.706 206Pb/204Pb 87Sr/86Sr

A B

C D

Figure 15. Field photographs of plutonic rocks from ophiolites documenting “wet” magmas: (A) miarolitic cavities, Point Sal ophiolite, California, (B) cumulate clinopyroxenite and wehrlite, Point Sal ophiolite, California, (C) decimeter-scale hornblendes in appinite dike, Trinity ophiolite, California, (D) photomicrograph of quartz diorite, showing abundant quartz (clear), feldspar (pale brown, low relief), and hornblende (dark brown, high relief), Elder Creek ophiolite, Califor- nia, ﬁ eld of view = 5.2 mm, plane light. Scale in A is 15 cm long; hammers in B and C are ~35 cm long. 212 Metcalf and Shervais

Peridotite Mineral Chemistry: Evidence for 1998; Pearce, 2003). The extremely high Cr# observed in many Hydrous Melting forearc peridotite spinels is also characteristic of spinels in high- Mg andesites and dacites, and other boninitic lavas. In general, Peridotites form a signifi cant fraction of most ophiolite spinels from peridotites associated with back-arc basins (e.g., assemblages, but they are seldom as well studied as the crustal Mariana Trough) have compositions similar to those from abys- sections. There now exists a large body of data on both abyssal sal peridotites (e.g., Ohara et al., 2002). peridotites, dredged from oceanic fracture zones and other base- Spinel compositional data are not widely available for ment exposures, and suprasubduction-zone (forearc) peridotites, many ophiolite mantle sections, but the data available suggest sampled from fault scarps on the leading edge of subduction com- that ophiolite peridotites fall into two groups: those with clear plexes. Abyssal peridotites are dominantly lherzolite, consisting suprasubduction-zone affi nities and those with mixed MORB– of olivine, enstatite, Cr-diopside, and aluminous spinel (e.g., suprasubduction-zone affi nities. Vourinos is dominated by Dick and Bullen, 1984; Dick, 1989; Johnson et al., 1990). In con- suprasubduction-zone spinel compositions (Cr# 45–85; Kon- trast, forearc peridotites consist largely of harzburgite, consisting stantopoulou and Economou-Eliopoulos, 1990). Troodos is of olivine, enstatite, and Cr-rich spinel, and dunite, consist ing of similarly dominated by suprasubduction-zone spinel compo- olivine plus chromite (Ishii et al., 1992; Parkinson and Pearce, sitions (Cr# 48–82; Hebert and Laurent, 1990; Georgiou and 1998; Pearce, 2003). In general, pyroxenes from abyssal perido- Xenophontos, 1990), but it also contains small domains of tites are relatively rich in incompatible elements compared to lherzolite with spinel Cr# of 22–28 (Batanova and Sobolev, pyroxenes from forearc peridotites, which have extremely low 2000). The Lewis Hills massif in the Bay of Islands complex incompatible element concentrations. resembles Troodos and is dominated by suprasubduction-zone Perhaps the most useful mineral, however, is spinel, spinels with Cr# of 50–78, but it contains a geographically which varies systematically in composition in response to melt small domain of abyssal peridotite with spinel Cr# of 15–30 extraction and is resistant to low-temperature alteration during (Suhr and Edwards, 2000). Oman shows the greatest variation, serpentini zation; it is commonly the only primary phase remain- with peridotite spinel Cr# ranging from 21 to 67, where most ing in highly serpentinized peridotites. Abyssal peridotites values are greater than 40 (Le Mee et al., 2004). are characterized by relatively aluminous spinels, with Cr# The signifi cance of the mixed domains is unclear because our (100 × Cr/[Cr + Al]) ranging from ∼10 to 59 (Fig. 16), indi- database for forearc peridotites is relatively small, but the conclu- cating limited melt extraction (Dick and Bullen, 1984; Dick, sions reached from peridotite spinel compositions are generally 1989). In contrast, peridotites dredged from forearc regions consistent with those derived from volcanic rock geochemistry. are characterized by relatively Cr-rich spinels, with Cr# rang- Ophiolites with volcanic rocks that indicate a suprasubduction- ing from ~38 to 60 in harzburgites and ∼60 to 84 in dunites zone origin are dominated by Cr-rich spinels with Cr# mostly >50 (Fig. 16), indicating more extensive melt extraction in response and few if any abyssal peridotite composition spinels. Ophiolites to hydrous melting (Ishii et al., 1992; Parkinson and Pearce, with mixed MORB–suprasubduction-zone volcanic sections are

100

Abyssal Peridotites Figure 16. Spinel composition plot for 80 abyssal peridotites, forearc peridotites, Boninites and boninites. Abyssal peridotites have low Cr# (100 × Cr/[Cr + Al]) compared to forearc peridotites and boninites. Data 60 are from: abyssal peridotites—Dick and Bullen (1984), Dick (1989), Juteau et al.

Mg# Abyssal Peridotites (1990), Komor et al. (1990), Hellebrand Forearc Lherzolite et al. (2002), Ohara et al. (2002), Arai 40 Forearc Harzburgite and Matsukage (1998); forearc perid- Forearc Dunite otites—Parkinson and Pearce (1998), Boninite Forearc Peridotites Ishii et al. (1992), Arai et al. (1990), Franz et al. (2002), Okamura et al. 20 (2006); boninites—Falloon et al. (1989), Van Der Laan et al. (1992).

0 0 102030405060708090100 Cr# Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 213 underlain by peridotites with a range of spinel compositions that sedimentation. There are two reasons for this. First, extensional refl ect both abyssal peridotite and forearc peridotite composi- forearcs associated with nascent subduction zones tend to be tions, but even in these examples, high-Cr spinels characteristic wide (150–300 km) with diffuse extension distributed across the of forearc peridotites are most common. width. The crust is thin and broken into a series of linear horsts and grabens parallel to the trench axis, which will trap clastic Sedimentary Cover sediment near its source and prevent its distribution throughout the forearc (Fryer et al., 2000; Hawkins, 2003). Second, nascent One major focus of the “ophiolite conundrum” is the cover arcs lack a distinct volcanic front and an emergent arc edifi ce. As of pelagic sediment, which is commonly, but not universally, long as most volcanic activity occurs underwater, the distribu- associated with ophiolites. The crux of this argument is that the tion of ash and coarse volcaniclastic materials will be limited. pelagic cover found on many ophiolites is inconsistent with formation of those ophiolites in or near an island-arc setting. The Preservation and Emplacement most common pelagic sediment associated with Mesozoic ophiolites is chert; indeed, chert was part of “Steinmann’s Trinity” and The structural preservation of ophiolites relates in large part until the 1972 Penrose conference on ophiolites was considered to their emplacement mechanics. Moores (1998) grouped ophio- an essential part of ophiolite stratigraphy. These cherts are typi- lites into two broad categories: Tethyan ophiolites and Cor dilleran cally rich in radiolaria and are thought to represent the slow accu- ophiolites. Tethyan-type ophiolites (e.g., Oman, Troodos , Pindos, mulation of radiolarian ooze on the seafl oor over several million Vourinos, Muslim Bagh) are emplaced onto passive continental years prior to obduction (e.g., Pessagno et al., 2000). margins and are typically overlain by sediments characteristic Pure radiolarian cherts, with little or no clastic component, of passive-margin settings (limestone, dolomite). In contrast, are generally restricted to the highly dismembered, incomplete Cordilleran-type ophiolites (e.g., Coast Range ophiolite of Cali- fragments of oceanic crust found in accretionary complexes or fornia, Trinity ophiolite, Cape Vogel) are associated with active some collision zones, for example, the Franciscan assemblage continental margins and typically are underlain by accretion- of California (Karl, 1984; Murchey, 1984; Murchey and Jones, ary complexes and overlain by clastic sediments deposited in a 1984) and ophiolites of the Western Mediterranean (Alps, Apen- forearc basin setting (turbidites, mudstones, conglomerates). nines; Bill et al., 2001). These cherts contain abundant radio- Tethyan- and Cordilleran-type ophiolites are similar in that laria and consist of nearly pure silica. They may also represent both types commonly display complete or near complete ophio- large time spans, e.g., Franciscan cherts of the Marin Headlands lite stratigraphy (as defi ned by Penrose Conference Participants, terrane in California, which represent ~30 m.y. of accumulation 1972), are relatively intact structurally, and are characterized by on the seafl oor (Murchey and Jones, 1984). Ophiolites in the suprasubduction-zone lava compositions (e.g., Shervais, 2001). northern Apennines apparently formed close to a rifted conti- Thus, we infer that both Tethyan- and Cordilleran-type ophiolites nental margin (Rampone and Piccardo, 2000), so only a limited form in the upper plates of subduction zones and differ largely age range is expected. in their mode of emplacement (Fig. 17). Tethyan-type ophiolites In contrast, cherts overlying many suprasubduction-zone represent obduction of forearc lithosphere onto a passive con- ophiolites are rich in volcanic ash, and many are essentially tinental margin during the attempted subduction of the passive altered tuffs. Cherts associated with the Coast Range ophiolite margin (Figs. 17A–17D). A variation on the normal Tethyan type in California contain up to 18% alumina and minor radiolaria of ophiolite is found in the Alps, where the upper-plate ophiolite (Hopson et al., 1981). In some Coast Range ophiolite locations, represents tectonically thinned continental lithosphere that has these altered tuffs are overlain by volcaniclastic sections up to not been signifi cantly modifi ed by arc volcanism (Frisch et al., 1.5 km thick with intercalated “radiolarian tuffs.” Similarly, 1994). Cordilleran-type ophiolites are emplaced by “accretionary radiolarites overlying the Troodos ophiolite contain common ash uplift” (Shervais, 2001), where continued growth of the under- layers and grade upward into calc-alkaline volcaniclastic strata of lying accretionary complex gradually lifts the overlying forearc the Kanaviou Formation (e.g., MacLeod et al., 1990). In Oman, ophiolite assemblage—there is typically no collision with a pas- tuffaceous chert overlies arc volcanics of the Alley unit and is sive continental margin (Figs. 17E–17H). overlain by ocean-island basalts of the Sahali volcanics. In the We commonly fi nd ancient rock assemblages with geochemi- Josephine ophiolite, calc-alkaline volcaniclastic detritus is found cal and petrologic characteristics that resemble true oceanic crust as interpillow sediment in the volcanic section, showing its clear formed at mid-ocean-ridge spreading centers or intraplate oceanic relationship to arc volcanism (Pessagno et al., 2000). In all of islands as dismembered, incomplete fragments within subduction- these cases, sediments deposited on the ophiolite contain signifi - zone accretionary complexes. Complete ophiolite sections are cant arc-derived detritus. unknown in these complexes, and gabbro is rare, but volcanic Recent work in the western Pacifi c (Fryer et al., 2000; rocks overlain by chert and associated with mantle-derived perido- Hawkins, 2003) has shown that the extensional forearc envi- tite (now serpentinite) are common. In all cases, the volcanic rocks ronments thought to characterize suprasubduction-zone ophio- associated with these complexes are geochemically equivalent to lite formation are generally not the locus of thick arc-derived N-MORB, E-MORB, or OIB; arc-like suprasubduction-zone vol- 214 Metcalf and Shervais

A SSZ ophiolite forms over sinking slab E SSZ ophiolite forms over sinking slab.

B Ophiolite collides with ridge crest, ophiolite formation stops F Ophiolite collides with ridge crest, ophiolite formation stops.

C Ophiolite encounters passive margin, begins to thrust over sediment wedge G Continued subduction; formation of accretionary complex uplifts ophiolite.

D Ophiolite is emplaced onto passive margin above schüppenzone of H Growth of accretionary complex exhumes ophiolite from beneath passive-margin sediments cover of forearc sediments.

Figure 17. Cross-section models of ophiolite emplacement by (A–D) obduction and (E–H) accretionary uplift. (A) Suprasubduction-zone (SSZ) ophiolite forms over sinking slab, which is separated from a passive continental margin by a spreading center. (B) Ophiolite encounters the spreading center; ophiolite formation stops, and the basal part of the ophiolite is thermally metamorphosed by high heat ﬂ ux from the thin lithosphere near the ridge crest. (C) Ophiolite encounters sedimentary wedge of the passive margin, which is depressed below the ophiolite and over- ridden by it; imbricate thrust sheets form in the passive-margin sediments. (D) Ophiolite is emplaced onto the passive continental margin above a schüppenzone of imbricate thrust sheets in the passive-margin sediments (e.g., Hawasina nappes in Oman, Mamonia complex in Cyprus). (E) Suprasubduction zone (SSZ) ophiolite forms over sinking slab; sinking of slab slows as spreading center is approached. (F) Ophiolite encounters the spreading center; ophiolite formation stops, and the basal part of the ophiolite is thermally metamorphosed by high heat ﬂ ux from the thin lithosphere near the ridge crest. (G) Sediments deposited in the subduction-zone trench are subducted to form an accretionary prism beneath the leading edge of the ophiolite; abyssal sediments and volcanic rocks scraped off the subducting oceanic plate may be included in the accretionary prism, which is dominated by juvenile detritus from the upper plate. (H) The accretionary prism continues to grow and thicken, exhuming the leading edge of the ophiolite.

canics are rare or nonexistent. Examples of this category include and emplaced within the accretionary complex. One characteris- fragments of oceanic crust found in the Franciscan assemblage tic of these Franciscan-style accretionary complexes is that they (e.g., Shervais, 1990, 2006; MacPherson et al., 1990) and in the sample material with a wide range in ages and are assembled Apennines (Rampone and Piccardo, 2000). over a prolonged time period during continuous subduction of These dismembered fragments represent ocean crust the subjacent oceanic lithosphere (Shervais, 2006). formed either at mid-ocean-ridge spreading centers or on A variation on the classic Franciscan-style accretionary com- off-axis seamounts or plateaus. Unlike suprasubduction- plex is found in Cyprus and Oman, where alkali basalt seamounts zone ophiolites, these ophiolite remnants are never complete and fringing reefs associated with rifting of the passive margin stratigraphically. The scarcity of gabbro in accretionary com- are scraped off the subducting plate during collision between the plexes suggests that they may preferentially sample oceanic ophio lite and the passive margin. These detached seamounts are crust formed near fracture zones, where mantle serpentine is mixed into the adjacent passive-margin sediments to form a schüp- exposed on the seafl oor and volcanic rocks may be erupted penzone beneath the ophiolite; examples include the Mamonia directly onto serpentine, with no other intervening crust (e.g., complex and Ayia Varvara Formation in Cyprus ( Malpas et al., Coleman, 2000). Off-axis seamounts or oceanic plateaus are 1992, 1993; Robertson and Xenophontos, 1993) and the Oman not well anchored structurally to the underlying seafl oor, and exotics within the Hawasina nappes (Robertson, 1986; Bechennec they may also be preferentially detached during subduction et al., 1988). These complexes sample seamounts that formed Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 215 over a limited time period during initial rifting of the passive mar- Pacifi c plate and forced the inception of a new NE-dipping subduc- gin; the complexes themselves were formed during closure of the tion zone that consumed the former back-arc basin. As discussed ocean basin that was created during this rifting event. previously, the geochemistry of basalts from the Woodlark basin is Sturm et al. (2000) proposed that oblique subduction of a typical of back-arc basin basalts: they are generally MORB-like in spreading center will result in suprasubduction zone–type enrich- composition but have a faint subduction-zone signature in the more ments of the spreading center magmas due to sublithospheric mobile large ion lithophile elements, such as Pb. fl ow through the “slab window,” as discussed earlier (ridge- In any case, basalts formed in such a setting still face the trench-trench junction model). They also suggested that oblique problem of emplacement: how do you move dense rocks from subduction of the two plates bordering the spreading center cre- the lower plate of a subduction zone onto a passive margin ates a setting favorable for ophiolite emplacement, as shown by (obduction) or place them above an accretionary prism (accre- the Taitao ophiolite complex (Stern, 1980). In this model, the fi rst tionary uplift) without violating the laws of physics? Like true slab to be subducted (on the trench side of the spreading center) mid-ocean-ridge basalts, back-arc basin basalts that are sub- is stranded in the subduction zone, and the second slab (from the ducted beneath their parent arc due to a subduction polarity opposite side of the spreading center) is subducted beneath it— fl ip are unlikely to be preserved, except as small fragments and effectively stepping the subduction zone farther outboard from slivers within the accretionary complex of the subduction zone. the arc. They further suggested that this model may apply to suprasubduction-zone ophiolites like Semail and Bay of Islands Formation of a Slab Window Where Ridges Are (Sturm et al., 2000). Subducted Orthogonally or Obliquely A geochemical assessment of this model has already been presented; we present here a geometric assessment. This model The Chile and Juan de Fuca spreading ridges are modern implies that the ophiolite will be emplaced structurally beneath a examples of this process. As discussed earlier, in both of these pre-existing island arc and may be separated from the overlying ridges, the dominant geochemical signature is that of E-MORB arc by an accretionary complex. Should this arc collide with and enrichment, which is unrelated to their position adjacent to a subduct a passive margin (e.g., Tethyan ophiolites), the ophio- subduction zone. Subduction enrichment in large ion litho- lite will be a small part of the total package; the older arc and phile elements is minimal and superimposed on the dominant its accretionary complex will dominate. This is not observed E-MORB enrichment. Thus, while this process may inject small in Semail, Troodos, or any other Tethyan-type ophiolites; in volumes of subduction-enriched mantle into the spreading cen- contrast, these ophiolites are overlain stratigraphically by post- ter, it is not suffi cient to create the dominant pattern documented collisional platform sediments (e.g., Glennie et al., 1974). There here of depletion in the more incompatible elements relative to is no evidence for an older arc complex that structurally overlies N-MORB, strong negative anomalies in Nb and the other high the “pseudo-suprasubduction-zone” ophiolite. fi eld strength elements, and signifi cant enrichments in the fl uid- The ridge-trench-trench model may apply to some mobilized large ion lithophiles/low fi eld strength elements. Cordilleran-type ophiolites, which often have older arc complexes Dynamically, there is some merit to the suggestion that sub- behind them. Unfortunately, most Cordilleran-type ophiolites duction of a spreading axis may allow ridge segments to be more are overlain by thick accumulations of forearc basin sediment, easily emplaced onto a continental margin: the spreading axis which obscures primary tectonic relationships. In this case, forms a discontinuity that allows the more distal plate to be thrust detailed chemical/petrologic studies must be applied. under the more proximal plate (which enters the subduction zone fi rst), potentially trapping the proximal plate above a newly con- HISTORICAL CONTINGENCY REDUX fi gured subduction boundary. Geometrically, this model implies that the trapped portion of the proximal plate will be preserved At the beginning of this paper, we presented a summary of beneath a volcanic arc and its previous accretionary complex. the main precepts of the historical contingency model, as pro- Few, if any, ophiolites preserve this geometry: for example, the posed by Moores et al. (2000). In this section, we assess the Coast Range ophiolite of California lies above the Franciscan applicability of each of these precepts to ophiolite generation, accretionary complex, not below it, and it is overlain deposition- guided by our exploration of modern tectonic settings and the ally by arc-derived sediments. Seismic tomography shows that rocks that form them. during most ridge-trench collisions, both sides of the spreading axis are subducted and sink into the mantle (e.g., Rogers et al., Asthenosphere Modifi ed by a Previous Subduction Event 2002). If fragments of oceanic crust are emplaced in this way, it is probable that they will be preserved within the accretionary This represents one of the central precepts of the historical complex as large mélange blocks. It is more likely that the slab contingency model. The only clear example we have of this process window allows MORB-source asthenosphere to affect the mantle in the recent geologic past is the Woodlark basin in the southwest wedge above the subduction zone, infl uencing the compositions Pacifi c. Collision of the Ontong Java plateau with the Solomon of the resulting arc magmas (Shervais et al., 2004, 2005a, 2005b; arc along a SW-dipping subduction zone stalled subduction of the Sisson et al., 2003, and papers therein). 216 Metcalf and Shervais

Subducted Slab Component—The Fate of Old Plates heterogeneities in the mantle (Zindler and Hart, 1986; Hart, 1988) generated by subduction recycling. As discussed already, Numerical models of mantle dynamics suggest that chemical the long-term trace-element heterogeneities that are responsible and isotopic heterogeneities related to subduction of oceanic litho- for the isotopic components found in OIB and MORB result sphere may persist on extremely long time scales (e.g., Kellogg from the decoupling of incompatible trace elements that are et al., 1999). It has been known for some time that subducted slabs mobilized in hydrous slab-derived fl uids to fertilize the overly- may remain distinct parts of the mantle, with chemical and isotopic ing mantle wedge from those elements that are not mobilized in systematics that differ from the surrounding MORB-source astheno- high-temperature hydrous fl uids and thus remain behind in the sphere. These slabs have been imaged by seismic tomography as slab. Some of these components may represent sediments that cooler, higher-velocity regions in the mantle (Rogers et al., 2002). are carried deep into the subduction zone and recycled into the However, these subducted slabs do not represent the subduction mantle, but all differ from the short-term enrichments in silica, component that modifi es the source region of suprasubduction-zone alkalis, and low fi eld strength elements that are characteristic magma systems (arcs and ophiolites). Lithospheric slabs that are of suprasubduction magma systems. Thus, the fact that MORB subducted deep into the mantle represent the residues that remain and OIB preserve isotopic evidence for a range of trace-element after extraction of the fl uid-rich component that carries silica, alka- enrichment processes is irrelevant to any discussion of the origin lis, and the large ion lithophile/low fi eld strength elements into the of suprasubduction-zone magmas, except where these sources mantle wedge above the subduction zone. This residual slab com- may be trapped in the mantle wedge above a subduction zone ponent is enriched relatively in the incompatible elements that are and participate in the formation of arc-related magmas. not mobilized in high-temperature fl uids, i.e., the high fi eld strength elements such as Ti, Nb, Ta, Hf, and Zr. Models for Mid-Ocean-Ridge Processes Trace-element systematics suggest that many OIBs form largely by the remelting of recycled oceanic lithosphere and The historical contingency model requires us to believe that depleted MORB-source asthenosphere (e.g., Hofmann, 1982; the only oceanic crust preserved intact is that formed over previ- Weaver, 1991). In addition, ridge-centered oceanic islands like ously modifi ed lithosphere; ocean crust formed from normal or Iceland and the Azores infl uence the composition of spreading plume-enriched MORB asthenosphere that has not been modifi ed center basalts by introducing these same components into the by these cryptic subduction-like enrichment processes is not pre- melting zone by fl ow along sublithospheric conduits (Schilling, served as ophiolites, even though >98% of all oceanic crust today 1973). As a result, basalts erupted along mid-oceanic ridges vary is normal or plume-enriched MORB with less than 52 wt% silica. in chemical and isotopic composition from “normal” N-MORB The historical contingency model uses isotopic hetero- to “enriched” E-MORB to true OIB at the ridge-centered oceanic geneities in the modern mantle refl ected in data from OIB, coupled islands. In this context, the OIB-style “enriched” basalt refers to with models of mantle dynamics, to argue that during certain a general enrichment in incompatible trace elements including periods of Earth history, mid-ocean-ridge magma systems could both LILE and HFSE, not to the fl uid-mobilized, LILE enrich- have tapped subduction-modifi ed mantle, yielding basalts with ment seen in suprasubduction-zone processes. This distinction is supra subduction zone–like compositions. Implicit in this argument important, because much of the historical contingency model rests is the assumption that the present is not one of those periods of on the defi nition of enrichment. In subduction-zone enrichment, Earth history; in other words, modern mid-ocean ridges presently there is a strong decoupling between LILEs (which are mobilized do not tap the type of subduction-modifi ed mantle represented by by aqueous solution and transferred from the subducted slab to OIB isotopic data. Our review of recently published trace-element the mantle wedge) and HFSEs (which are insoluble in hydrous and isotopic data from active mid-ocean ridges, however, refute fl uids and remain in the slab). The OIB-style enrichment is mobi- this implicit assumption. Although the modern MORB data set lized by silicate melts; LILE and HFSE are not decoupled, but the refl ects the same isotopic variations observed in OIB, variations enrichment sources (old slabs) were previously depleted in LILE attributed to contributions from subducted oceanic lithosphere, (subduction) components. In effect, the enriched component these mid-ocean-ridge basalts do not exhibit the suprasubduction found in OIB and E-MORB is the result of subduction, but it rep- zone–like, LILE enrichments seen in either modern subduction- resents the complement to that found in the fl uid-fl uxed mantle zone basalts or in the suprasubduction-zone ophiolite record. What wedge above a subduction zone. The processes that affect the these basalts do show are OIB-style enrichments, i.e., a coupled mantle wedge defi ne what most geologists refer to as subduction- LILE-HFSE enrichment, as discussed already. zone enrichment. DISCUSSION Isotopic Components in Ocean-Island Basalts The issues raised by the hypothesis of “historical contin- Ocean-island basalts (OIBs) contain a number of distinct gency” and our discussion here are not trivial. What is at stake? isotopic components (DMM, HIMU, EM1, EM2, PREMA/ Our reconstructions of global tectonics before the current ocean FOZO) that represent the persistence of long-term trace-element basins formed depend critically on how we interpret the ophio- Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 217 lite record. Even suprasubduction-zone ophiolites require the sistent with formation at a mid-oceanic spreading center. More existence of an ocean basin for their formation, but the details importantly, the structural setting of ophiolites during emplace- of mountain-building events and relations between allochthonous ment requires that they formed in the upper plate of a convergent- sheets in complex orogens hinge on whether a given ophio lite margin system. In fact, based on their structural setting alone, we assemblage represents true “oceanic crust” formed at a mid- would be forced reach this conclusion (e.g., Gealey, 1977). ocean-ridge spreading center, back-arc basin crust formed behind an arc, or suprasubduction-zone crust formed over a nascent sub- ACKNOWLEDGMENTS duction zone. These issues are particularly important for people who are not involved in the debate on how ophiolites form, but The authors have benefi ted over the years from spirited dis- who use ophiolites to reconstruct the tectonic history of oro- cussions with many ophioliteologists, but none has challenged gens. How do they know which model of ophiolite formation us more than Cliff Hopson and Eldridge Moores, who forced is correct, and how will it affect their interpretation of ophiolite us to examine our assumptions and think clearly about what is assemblages in their fi eld area? Researchers who are not involved observed and what we infer. We are grateful to Jim Hawkins in this debate may be easily misled by hypotheses that appear and Bob Stern, who provided thoughtful reviews that aided us in sound, but which do not have the data to sustain them. The ability improving the manuscript, and editorial handling by Jim Wright. to use suprasubduction-zone ophiolites as natural laboratories for studying the initial composition of new subduction zones is also REFERENCES CITED at stake; much of that record is found in suprasubduction-zone ophiolites formed at nascent forearc settings (Stern, 2004). Alabaster, T., Pearce, J.A., and Malpas, J., 1982, The volcanic stratigraphy and petrogenesis of the Oman ophiolite complex: Contributions We must ask whether the term “ophiolite” is descriptive to Mineralogy and Petrology, v. 81, no. 3, p. 168–183, doi: 10.1007/ or genetic. If it is descriptive, then rock assemblages that fi t BF00371294. the description may form from a variety of processes, and the Arai, S., and Matsukage, K., 1998, Petrology of a chromitite micropod from the Hess Deep, equatorial Pacifi c: A comparison between abyssal and evidence for these processes can be applied to any assemblage alpine-type podiform chromitites: Lithos, v. 43, p. 1–14, doi: 10.1016/ that fi ts the description. If it is genetic, then we must thoroughly S0024-4937(98)00003-6. understand the origin of any given assemblage before we can con- Arai, S., Ito, M., Nakayama, N., and Masuda, F., 1990, A suspect serpentinite mass in the Tokyo Bay area: Petrology and provenance of serpentinite sider whether this term applies. If we take the genetic approach, pebbles in the Upper Cenozoic system in the Boso Peninsula, central then most of the rock suites we call “ophiolites” would have to Japan: Journal of Geological Society of Japan, v. 96, p. 171–179. be renamed, because their petrology, geochemistry, and struc- Arculus, R.J., and Powell, R., 1986, Source component mixing in the regions of arc magma generation: Journal of Geophysical Research, v. 91, tural setting are at odds with the interpretation that they represent p. 5913–5926. oceanic crust formed at a mid-ocean-ridge spreading center. Bach, W., Erzinger, J., Dosso, L., Bollinger, C., Bougault, H., Etoubleau, J., We return to our suggestion that a new formulation of the and Sauerwein, J., 1996, Unusually large Nb-Ta depletions in the North Chile ridge basalts at 36°50′ to 38°S: Major element, trace element and “ophiolite conundrum” is needed: Given that many, if not most, isotopic data: Earth and Planetary Science Letters, v. 142, p. 223–240, ophiolites have geochemical signatures consistent with formation doi: 10.1016/0012-821X(96)00095-7. above active subduction zones, under what circumstances does Baragar, W.R.A., Varga, R.J., Verosub, K.L., Baglow, N., and Gibson, I.L., 1990, The sheeted dyke zone in the Troodos ophiolite, in Malpas, J., this setting result in rock associations and structures consistent Moores, E.M., Panayiotou, A., and Xenophontos, C., eds., Ophiolites: with those observed in ophiolites? Recent work in volcanic arcs Oceanic Crustal Analogues: Proceedings of the Symposium “Troodos of the western Pacifi c has shown that intra-arc rifting of existing 1987”: Cyprus, Nicosia, Geological Survey Department, p. 37–52. Barth, M.G., Mason, P.R.D., Davies, G.D., Dijkstra, A.H., and Drury, primitive arcs (e.g., Vanuatu) and rifting associated with the onset M.R., 2003, Geochemistry of the Othris ophiolite, Greece: Evidence for of subduction (e.g., the Eocene Mariana-Izu-Bonin system) may refertilization?: Journal of Petrology, v. 44, p. 1759–1785. form rock assemblages similar to those observed in ophiolite Batanova, V.G., and Sobolev, A.V., 2000, Compositional heterogeneity in subduction-related mantle peridotites, Troodos massif, Cyprus: Geology, complexes (Bloomer et al., 1995; Hawkins, 2003; Fryer et al., v. 28, no. 1, p. 55–58, doi: 10.1130/0091-7613(2000)28<55:CHISMP> 2000). These assemblages may represent up to 100% extension 2.0.CO;2. of the pre-existing crust over broad areas and may be starved of Beccaluva, L., Coltorti, M., Giunta, G., and Siena, F., 2004, Tethyan vs. Cor- dilleran ophiolites: A reappraisal of distinctive tectono-magmatic features clastic sediments other than windborne volcanic ash. They also of supra-subduction complexes in relation to subduction mode: Tectono- consist of volcanic and plutonic assemblages that have the petro- physics, v. 393, p. 163–174. logic and geochemical characteristics of suprasubduction-zone Bechennec, F., Le Metour, J., Raby, D., Villey, M., and Beurrier, M., 1988, The Hawasina Basin: A fragment of a starved passive continental margin, thrust magmas, including boninites, which appear to form only within over the Arabian platform during obduction of the Semail Nappe: Tec- the forearc regions of highly extended arcs. tonophysics, v. 151, p. 323–343, doi: 10.1016/0040-1951(88)90251-X. Finally, we emphasize that it is not just the geochemistry of Bedard, J.H., 1999, Petrogenesis of boninites from the Betts Cove ophiolite, Newfoundland, Canada; identifi cation of subducted source com- ophiolite volcanic rocks that requires their formation in supra- ponents: Journal of Petrology, v. 40, p. 1853–1889, doi: 10.1093/ subduction environments. The evidence for hydrous magmatic petrology/40.12.1853. systems, hydrous melting, and the common occurrence of vol- Bedard, J.H., Lauziere, K., Tremblay, A., and Sangster, A., 1998, Evidence for forearc seafl oor-spreading from the Betts Cove ophiolite, Newfound- canic ash in overlying cherts are consistent with an original land: Oceanic crust of boninitic affi nity: Tectonophysics, v. 284, no. 3–4, suprasubduction-zone setting for these ophiolites and are incon- p. 233–245, doi: 10.1016/S0040-1951(97)00182-0. 218 Metcalf and Shervais

Bill, M., O’Dogherty, L., Guex, J., Baumgartner, P.O., and Masson, H., 2001, Parsons, L., Allan, J., et al., Proceedings of the Ocean Drilling Program, Radiolarite ages in Alpine-Mediterranean ophiolites: Constraints on the Scientiﬁ c Results, Volume 135: College Station, Texas, Ocean Drilling oceanic spreading and the Tethys-Atlantic connection: Geological Society Program, p. 385–425. of America Bulletin, v. 113, p. 129–143, doi: 10.1130/0016-7606(2001) Falloon, T.J., and Danyushevsky, L.V., 2000, Melting of refractory mantle at

113<0129:RAIAMO>2.0.CO;2. 1.5, 2 and 2.5 GPa under anhydrous and H2O-undersaturated conditions; Bloomer, S.H., and Hawkins, J.W., 1983, Gabbroic and ultramafi c rocks from implications for the petrogenesis of high-Ca boninites and the infl uence the Mariana Trench, in Hayes, D.E., ed., The Tectonic Evolution of the of subduction components on mantle melting: Journal of Petrology, v. 41, Southwest Asian Seas and Islands, Part 2: American Geophysical Union no. 2, p. 257–283, doi: 10.1093/petrology/41.2.257. Memoir 27, p. 294–317. Falloon, T.J., Green, D.H., and McCulloch, M.T., 1989, Petrogenesis of high- Bloomer, S.H., Taylor, B., MacLeod, C.J., Stern, R.J., Fryer, P., Hawkins, J.W., Mg and associated lavas from the North Tonga Trench, in Crawford, and Johnson, L., 1995, Early arc volcanism and the ophiolite problem: A A.J., ed., Boninites and Related Rocks: London, UK, Unwin Hyman, perspective from drilling in the western Pacifi c, in Taylor, B., and Nat- p. 357–395. land, J., eds., Active Margins and Marginal Basins of the Western Pacifi c: Flower, M.F.J., 2003, Ophiolites, historical contingency and the Wilson cycle, American Geophysical Union Geophysical Monograph 88, p. 1–30. in Dilek, Y., and Newcomb, S., eds., Ophiolite Concept and the Evolution Bryan, W.G., 1983, Systematics of modal phenocryst assemblages in sub- of Geologic Thought: Geological Society of America Special Paper 373, marine basalts: Petrological implications: Contributions to Mineralogy p. 111–135. and Petrology, v. 83, p. 62–74, doi: 10.1007/BF00373080. Franz, L., Becker, K.-P., Kramer, W., and Herzig, P.M., 2002, Meta somatic Cameron, W.E., 1985, Petrology and origin of primitive lavas from the Troodos mantle xenoliths from the Bismarck microplate (Papua New Guinea)— ophiolite, Cyprus: Contributions to Mineralogy and Petrology, v. 89, Thermal evolution, geochemistry and extent of slab-induced meta- no. 2–3, p. 239–255, doi: 10.1007/BF00379457. somatism: Journal of Petrology, v. 43, p. 315–343, doi: 10.1093/ Cann, J.R., 2003, The Troodos ophiolite and the upper oceanic crust: A recipro- petrology/43.2.315. cal traffi c in scientifi c ideas, in Dilek, Y., and Newcomb, S., eds., Ophio- Frisch, W., Ring, U., Duerr, S., Borchert, S., and Biehler, D., 1994, The Arosa lite Concept and the Evolution of Geologic Thought: Geological Society zone and Platta nappe ophiolites (eastern Swiss Alps); geochemical char- of America Special Paper 373, p. 309–321. acteristics and their meaning for the evolution of the Penninic Ocean: Coleman, R.G., 2000, Prospecting for ophiolites along the California conti- Jahrbuch der Geologischen Bundesanstalt Wien, v. 137, no. 1, p. 19–33. nental margin, in Dilek, Y., Moores, E.M., Elthon, D., and Nicolas, A., Fryer, P., Lockwood, J.P., Becker, N., Phipps, S., and Todd, C.S., 2000, Signifi - eds., Ophiolites and Oceanic Crust: New Insights from Field Studies cance of serpentinite mud volcanism in convergent margins, in Dilek, Y., and the Ocean Drilling Program: Geological Society of America Special Moores, E.M., Elthon, D., and Nicolas, A., eds., Ophiolites and Oceanic Paper 349, p. 351–364. Crust: New Insights from Field Studies and the Ocean Drilling Program: Cousens, B.L., Allan, J.F., Leybourne, M.I., Chase, R.L., and Van Wagoner, N., Geological Society of America Special Paper 349, p. 35–52. 1995, Mixing of magmas from enriched and depleted mantle sources in Gass, I.G., 1968, Is the Troodos massif of Cyprus a fragment of Meosozoic the northwest Pacifi c: West Valley segment, Juan de Fuca Ridge: Contri- ocean fl oor?: Nature, v. 220, p. 32–42. butions to Mineralogy and Petrology, v. 120, p. 337–357. Gass, I.G., Lippard, S.J., and Shelton, A.W., eds., 1984, Ophiolites and Oceanic Crawford, A.J., and Falloon, T.J., 1989, Experimental determination of Lithosphere: Geological Society [London] Special Publication 13, 413 p. boninite petrogenesis, in Crawford, A.J., ed., Boninites: London, UK, Gealey, W.K., 1977, Ophiolite obduction and geologic evolution of the Oman Unwin Hyman, p. 100–120. Mountains and adjacent areas: Geological Society of America Bulletin, Crook, K.A.W., and Taylor, B., 1994, Structure and Quaternary tectonic history v. 88, p. 1183–1191, doi: 10.1130/0016-7606(1977)88<1183:OOAGEO> of the Woodlark triple junction region, Solomon Islands: Marine Geo- 2.0.CO;2. physical Researches, v. 16, no. 1, p. 65–89, doi: 10.1007/BF01812446. Georgiou, E., and Xenophontos, C., 1990, Chromite occurrences and associ- Davidson, J.P., 1996, Deciphering mantle and crustal signatures in subduction ated plutonic rocks in the Akapnou Forest, in Malpas, J., Moores, E.M., zones magmatism, in Bebout, G.E., School, D.W., Kirby, S.H., and Platt, Panayiotou, A., and Xenophontos, C., eds., Ophiolites: Oceanic Crustal J.P., eds., Subduction Top to Bottom: American Geophysical Union Geo- Analogues: Proceedings of the Symposium “Troodos 1987”: Cyprus, physical Monograph 96, p. 251–262. Nicosia, Geological Survey Department, p. 585–592. Deschamps, A., and Lallemand, S., 2003, Geodynamic setting of Izu-Bonin- Glennie, K.W., Boeuf, M.G., Clarke, M.W.H., Moody-Stuart, M., Pilaar, W.F.H., Mariana boninites, in Larter, R.D., and Leat, P.T., eds., Intra-Oceanic Sub- and Reinhardt, B.M., 1974, Geology of the Oman Mountains: Verhan- duction Systems: Tectonic and Magmatic Processes: Geological Society delingen van het Koninklijk Nederlands Geologisch Mijnbouwkundig [London] Special Publication 219, p. 163–186. Genootschap, v. 31, 423 p. Dick, H.J.B., 1989, Abyssal peridotites, very slow spreading ridges and ocean Harper, G.D., 2003a, Fe-Ti basalts and propagating-rift tectonics in the Jose- ridge magmatism, in Saunders, A.D., and Norry, M.J., eds., Magmatism phine ophiolite: Geological Society of America Bulletin, v. 115, no. 7, in the Ocean Basins: Geological Society [London] Special Publication p. 771–787. 42, p. 71–105. Harper, G.D., 2003b, Tectonic implications of boninite, arc tholeiite, and Dick, H.J.B., and Bullen, T., 1984, Chromian spinel as a petrogenetic indica- MORB magma types in the Josephine ophiolite, California, Oregon, in tor in abyssal and alpine-type peridotites and spatially associated lavas: Dilek, Y., and Robinson Paul, T., eds., Ophiolites in Earth History: Geo- Contributions to Mineralogy and Petrology, v. 86, p. 54–76, doi: 10.1007/ logical Society [London] Special Publication 218, p. 207–230. BF00373711. Hart, S.R., 1988, Heterogeneous mantle domains: Signatures, genesis and mix- Dilek, Y., 2003, Ophiolite concept and its evolution, in Dilek, Y., and Newcomb, ing chronologies: Earth and Planetary Science Letters, v. 90, p. 273–296. S., eds., Ophiolite Concept and the Evolution of Geologic Thought: Geo- Hawkins, J.W., 2003, Geology of suprasubduction zones—Implications for logical Society of America Special Paper 373, p. 1–16. the origin of ophiolites, in Dilek, Y., and Newcomb, S., eds., Ophiolite Einaudi, F., Pezard, P.A., Cocheme, J.J., Coulon, C., Laverne, C., and Concept and the Evolution of Geologic Thought: Geological Society of Godard, M., 2000, Petrography, geochemistry and physical properties America Special Paper 373, p. 227–268. of a continuous extrusive section from the Sarami Massif, Semail ophio- Hawkins, J.W., and Evans, C.A., 1983, Geology of the Zambales Range, lite: Marine Geophysical Researches, v. 21, no. 3–4, p. 387–407, doi: Luzon, Philippines: Ophiolite derived from an island arc-backarc basin 10.1023/A:1026752415989. pair, in Hayes, D.E., ed., The Tectonic Evolution of the Southwest Asian Elliott, T., Plank, T., Zindler, A., White, W., and Bourdon, B., 1997, Element Seas and Islands, Part 2: American Geophysical Union Memoir 27, transport from slab to volcanic front at Mariana arc: Journal of Geophysi- p. 124–138. cal Research, v. 102, p. 14,991–15,019, doi: 10.1029/97JB00788. Hawkins, J.W., Bloomer, S.H., Evans, C.A., and Melchior, J.T., 1984, Evolu- Ernewein, M., Pfl umio, C., and Whitechurch, H., 1988, The death of an accretion of intra-oceanic arc-trench systems, in Carlson, R.L., and Kobayashi, tion zone as evidenced by the magmatic history of the Sumail ophiolite K., eds., Geodynamics of Back-Arc Regions: Tectonophysics, v. 102 (Oman), in Boudier, F., and Nicolas, A., eds., The Ophiolites of Oman: p. 175–205. Tectonophysics, v. 151, p. 247–274. Hebert, R., and Laurent, R., 1990, Mineral chemistry of the plutonic section of Ewart, A., Bryan, W.B., Chappell, B.W., and Rudnick, R.L., 1994, Regional the Troodos ophiolite; new constraints for genesis of arc-related ophio- geochemistry of the Lau-Tonga arc and backarc systems, in Hawkins, J., lites, in Malpas, J., Moores, E.M., Panayiotou, A., and Xenophontos, C., Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 219

eds., Ophiolites: Oceanic Crustal Analogues: Proceedings of the Sympo- Kellogg, L.H., Hager, B.H., and van der Hilst, R.D., 1999, Compositional strati- sium “Troodos 1987”: Cyprus, Nicosia, Geological Survey Department, fi cation in the deep mantle: Science, v. 283, p. 1881–1884. p. 149–163. Klein, E.M., and Karsten, J.L., 1995, Ocean-ridge basalts with convergent- Hellebrand, E., Snow, J.E., and Mühe, R.K., 2002, Mantle melting margin geochemical affi nities from the Chile Ridge: Nature, v. 374, beneath Gakkel Ridge (Arctic Ocean), abyssal peridotite spinel com- no. 6517, p. 52–57, doi: 10.1038/374052a0. positions: Chemical Geology, v. 182, p. 227–235, doi: 10.1016/ Komor, S.C., Grove, T.L., and Hebert, R., 1990, Abyssal peridotites from ODP S0009-2541(01)00291-1. Hole 670a (21°10N, 45°02W): Residues of mantle melting exposed by Hess, H.H., 1955, Serpentinites, orogeny and epeirogeny, in Poldervaart, A., non-constructive axial divergence, in Detrick, R., Honnorez, J., Bryan, ed., Crust of the Earth—A Symposium: Geological Society of America W.B., and Juteau, T., eds., Proceedings of the Ocean Drilling Program: Special Paper 62, p. 391–408. Scientifi c Results, Volume 106/109: College Station, Texas, Ocean Drill- Hickey, R.L., and Frey, F.A., 1982, Geochemical characteristics of boninite series ing Program, p. 85–101. volcanics; implications for their source: Geochimica et Cosmochimica Konstantopoulou, G., and Economou-Eliopoulos, M., 1990, Geochemistry Acta, v. 46, no. 11, p. 2099–2116, doi: 10.1016/0016-7037(82)90188-0. of the Vourinos chromite ores, Greece, in Malpas, J., Moores, E.M., Hofmann, A.W., 1982, Mantle plumes from ancient oceanic crust: Earth Panayiotou, A., and Xenophontos, C., eds., Ophiolites: Oceanic Crustal and Planetary Science Letters, v. 57, p. 421–436, doi: 10.1016/0012-821X Analogues: Proceedings of the Symposium “Troodos 1987”: Cyprus, (82)90161-3. Nicosia, Geological Survey Department, p. 605–613. Hofmann, A.W., 1997, Early evolution of the continents: Science, v. 275, Laurent, R., 1992, Peridotite intrusions emplaced in the fossil suprasubduction no. 5229, p. 498–499. zone environment of Cyprus, in Parson, L.M., Murton, B.J., and Brown- Hopson, C.A., Mattinson, J.M., and Pessagno, E.A., Jr., 1981, Coast Range ing, P., eds., Ophiolites and Their Modern Oceanic Analogues: Geological ophiolite, western California, in Ernst, W.G., ed., The Geotectonic Devel- Society [London] Special Publication 60, p. 233–239. opment of California, Rubey Volume I: Englewood Cliffs, New Jersey, Lehnert, K., Su, Y., Langmuir, C.H., Sarbas, B., and Nohl, U., 2000, A global Prentice-Hall, p. 418–510. geochemical database structure for rocks: Geochemistry, Geophysics: Ishii, T., Robinson, P.T., Mackawa, H., and Fiske, R.S., 1992, Petrological Geosystems, v. 1, no. 5, doi: 10.1029/1999GC000026. studies of peridotites from diapiric serpentine seamounts in the Izu- Le Mee, L., Girardeau, J., and Monnier, C., 2004, Mantle segmentation along Ogasawara-Mariana forearc, Leg 125, in Fryer, P., Pearce, J.A., and the Oman ophiolite fossil mid-ocean ridge: Nature, v. 532, p. 167–172. Strokking, L.B., eds., Proceedings of the Ocean Drilling Program, Sci- le Roux, P.J., le Roux, A.P., Schilling, J.-G., Shimizu, N., Perkins, W.W., and entifi c Results, Volume 125: College Station, Texas, Ocean Drilling Pro- Pearce, N.J.G., 2002, Mantle heterogeneity beneath the southern Mid- gram, p. 445–486. Atlantic Ridge: Trace element evidence for contamination of ambient Ishikawa, T., Nagaishi, K., and Umino, S., 2002, Boninitic volcanism in the asthenospheric mantle: Earth and Planetary Science Letters, v. 203, Oman ophiolite: Implications for thermal condition during transition from p. 479–498, doi: 10.1016/S0012-821X(02)00832-4. spreading ridge to arc: Geology, v. 30, no. 10, p. 899–902, doi: 10.1130/ Liati, A., Gebauer, D., and Fanning, C.M., 2004, The age of ophiolitic rocks of 0091-7613(2002)030<0899:BVITOO>2.0.CO;2. the Hellenides (Vourinos, Pindos, Crete); fi rst ion probe (SHRIMP) zircon Jacobsen, S.B., and Wasserburg, G.J., 1980, Sm-Nd isotopic evolution of ages: Chemical Geology, v. 207, p. 171–188. chondrites: Earth and Planetary Science Letters, v. 50, p. 139–155, doi: Lippard, S.J., Shelton, A.W., and Gass, I.G., 1986, The Ophiolite of Northern 10.1016/0012-821X(80)90125-9. Oman: Geological Society of London Memoir 11, 178 p. Jenner, G.A., Dunning, G.R., Malpas, J., Brown, M., and Brace, T., 1991, Bay MacLeod, C.J., Allerton, S., Gass, I.G., and Xenophontos, C., 1990, Structure of Islands and Little Port complexes, revisited; age, geochemical and iso- of a fossil ridge-transform intersection in the Troodos ophiolite: Nature, topic evidence confi rm suprasubduction-zone origin: Canadian Journal of v. 348, p. 717–720. Earth Sciences, v. 28, no. 10, p. 1635–1652. MacPherson, G.J., Phipps, S.P., and Grossman, J.N., 1990, Diverse sources for Johnson, K.M., Dick, H.J.B., and Shimizu, N., 1990, Melting in the oceanic igneous blocks in Franciscan melanges, California Coast Ranges: The upper mantle; an ion microprobe study of diopside in abyssal peridotite: Journal of Geology, v. 98, no. 6, p. 845–862. Journal of Geophysical Research, v. 96, no. B3, p. 2661–2678. Malpas, J., and Langdon, G., 1984, Petrology of the Upper Pillow Lava suite, Johnson, R.W., Jaques, A.L., Langmuir, C.H., Perfi t, M.R., Staudigel, H., Troodos ophiolite, Cyprus, in Gass, I.G., Lippard, S.J., and Shelton, A.W., Dunkley, P.N., Chappell, B.W., Taylor, S.R., and Baekisapa, M., 1987, eds., Ophiolites and Oceanic Lithosphere: Geological Society [London] Ridge subduction and forearc volcanism; petrology and geochemistry Special Publication 60, p. 155–167. of rocks dredged from the western Solomon arc and Woodlark basin, in Malpas, J., Xenophontos, C., and Williams, D., 1992, The Ayia Varvara Formation Taylor, B., and Exon, N.F., eds., Marine Geology, Geophysics and Geo- of SW Cyprus; a product of complex collisional tectonics: Tectonophysics, chemistry of the Woodlark Basin–Solomon Islands: Menlo Park, Cali- v. 212, no. 3–4, p. 193–211, doi: 10.1016/0040-1951(92)90291-D. fornia, Circum-Pacifi c Council for Energy and Mineral Resources, Earth Malpas, J., Calon, T., and Squires, G., 1993, The development of a Late Science Series, p. 155–226. Cretaceous microplate suture zone in SW Cyprus, in Prichard, H.M., Juteau, T., Berger, E.T., and Cannat, M., 1990, Serpentinized, residual mantle Alabaster, T., Harris, N.B.W., and Neary, C.R., eds., Magmatic Processes peridotites from the Mar Median Valley, ODP Hole 670a (21°10N, and Plate Tectonics: Geological Society [London] Special Publication 76, 45°02W, Leg 109): Primary Mineralogy and Geothermometry, in Pro- p. 177–195. ceedings of the Ocean Drilling Program, Scientifi c Results, Volume 106/ Maxeiner, R.O., Corrigan, D., Harper, C.T., MacDougall, D.G., and Ansdell, K., 109: College Station, Texas, Ocean Drilling Program, p. 27–45. 2005, Paleoproterozoic arc and ophiolitic rocks of the northwest-margin Kamenetsky, V.S., Everard, J.L., Crawford, A.J., Varne, R., Eggins, S.M., of the Trans-Hudson orogen, Saskatchewan, Canada: Their contribution and Lanyon, R., 2000, Enriched end-member of primitive MORB to a revised tectonic framework for the orogen: Precambrian Research, melts: Petrology and geochemistry of glasses from Macquarie Island v. 136, p. 67–106. (SW Pacifi c): Journal of Petrology, v. 41, p. 411–430, doi: 10.1093/ McCulloch, M.T., and Cameron, W.E., 1983, Nd-Sr isotopic study of primi- petrology/41.3.411. tive lavas from the Troodos ophiolite, Cyprus; evidence for a subduction- Karl, S.M., 1984, Sedimentologic, diagenetic, and geochemical analysis of related setting: Geology, v. 11, no. 12, p. 727–731, doi: 10.1130/0091-7613 Upper Mesozoic ribbon cherts from the Franciscan assemblage at the (1983)11<727:NISOPL>2.0.CO;2. Marin Headlands, California, in Blake, C., Jr., ed., Franciscan Geology Metcalf, R.V., Wallin, E.T., Willse, K.R., and Muller, E.R., 2000, Geology and of Northern California: Los Angeles, California, Society of Economic geochemistry of the ophiolitic Trinity terrane, California; evidence of Paleon tologists and Mineralogists, Pacifi c Section, Field Trip Guide- middle Paleozoic depleted supra-subduction zone magmatism in a proto- book 43, p. 71–88. arc setting, in Dilek, Y., Moores, E.M., Elthon, D., and Nicolas, A., eds., Karsten, J.L., Klein, E.M., and Sherman, S.B., 1996, Subduction zone Ophiolites and Oceanic Crust: New Insights from Field Studies and the geochemical characteristics in ocean ridge basalts from the southern Ocean Drilling Program: Geological Society of America Special Paper Chile Ridge; implications of modern ridge subduction systems for the 349, p. 403–418. Archean, in Ludden, J.N., Arndt, N.T., and Francis, D., eds., Evolu- Miyashiro, A., 1973, The Troodos ophiolitic complex was probably formed in tion of Mafi c Magmatism through Time: St. Malo, France, Lithos, an island arc: Earth and Planetary Science Letters, v. 19, no. 2, p. 218–224, p. 143–161. doi: 10.1016/0012-821X(73)90118-0. 220 Metcalf and Shervais

Moores, E.M., 1998, Ophiolites, the Sierra Nevada, “Cordillera,” and orogeny Howells, M.F., eds., Marginal Basin Geology: Palo Alto, California, along the Pacifi c and Caribbean margins of North and South America: Blackwell Scientifi c Publications, p. 74–94. International Geology Review, v. 40, no. 1, p. 40–54. Pearce, J.A., Baker, P.E., Harvey, P.K., and Luff, I.W., 1995, Geochemical evi- Moores, E.M., 2002, Pre–1 Ga (pre-Rodinian) ophiolites: Their tectonic and dence for subduction fl uxes, mantle melting and fractional crystallization environmental implications: Geological Society of America Bulletin, beneath the South Sandwich Island arc: Journal of Petrology, v. 36, no. 4, v. 114, p. 80–95, doi: 10.1130/0016-7606(2002)114<0080:PGPROT> p. 1073–1109. 2.0.CO;2. Pearce, J.A., Stern, R.J., Bloomer, S.H., and Fryer, P., 2005, Geochemical map- Moores, E.M., 2003, A personal history of the ophiolite concept, in Dilek, Y., ping of the Mariana arc-basin system: Implications for the nature and and Newcomb, S., Ophiolite Concept and Evolution of Geologic Thought: distribution of subduction components: Geochemistry, Geophysics, Geo- Geological Society of America Special Paper 373, p. 17–29. systems (G3), v. 6, no. 7, doi: 10.1029/2004GC000895, Moores, E.M., and Vine, F.J., 1971, Troodos massif, Cyprus, and other ophio- Penrose Conference Participants, 1972, Penrose Field Conference on Ophio- lites as oceanic crust: Evaluation and implications: Philosophical Trans- lites: Geotimes, v. 17, p. 24–25. actions of the Royal Society of London, v. 268, p. 443–466. Perfi t, M.R., Gust, D.A., Bence, A.E., Arculus, R.J., and Taylor, S.R., 1980, Moores, E.M., Robinson, P.T., Malpas, J., and Xenophontos, C., 1984, Model Chemical characteristics of island-arc basalts: Implications for mantle for the origin of the Troodos Massif, Cyprus, and other Mideast ophio- sources: Chemical Geology, v. 30, p. 227–256. lites: Geology, v. 12, no. 8, p. 500–503, doi: 10.1130/0091-7613(1984) Perfi t, M.R., Langmuir, C.H., Baekisapa, M., Chappell, B.W., Johnson, R.W., 12<500:MFTOOT>2.0.CO;2. Staudigel, H., and Taylor, S.R., 1987, Geochemistry and petrology of Moores, E.M., Kellogg, L.H., and Dilek, Y., 2000, Tethyan ophiolites, man- volcanic rocks from the Woodlark basin; addressing questions of ridge tle convection, and tectonic “historical contingency”; a resolution of subduction, in Taylor, B., and Exon, N.F.E., eds., Marine Geology, Geo- the “ophiolite conundrum,” in Dilek, Y., Moores, E.M., Elthon, D., and physics and Geochemistry of the Woodlark Basin–Solomon Islands: Nicolas, A., eds., Ophiolites and Oceanic Crust: New Insights from Field Menlo Park, California, Circum-Pacifi c Council for Energy and Mineral Studies and the Ocean Drilling Program: Geological Society of America Resources, Earth Science Series, p. 113–154. Special Paper 349, p. 3–12. Pessagno, E.A., Jr., Hull, D.M., and Hopson, C.A., 2000, Tectonostratigraphic Muenow, D.W., Perfi t, M.R., and Aggrey, K.E., 1991, Abundances of volatiles signifi cance of sedimentary strata occurring within and above the Coast and genetic relationships among submarine basalts from the Woodlark Range ophiolite (California Coast Ranges) and the Josephine ophiolite basin: Southwest Pacifi c: Geochimica et Cosmochimica Acta, v. 55, no. 8, (Klamath Mountains), northwestern California, in Dilek, Y., Moores, p. 2231–2239, doi: 10.1016/0016-7037(91)90099-Q. E.M., Elthon, D., and Nicolas, A., eds., Ophiolites and Oceanic Crust: Murchey, B.L., 1984, Biostratigraphy and lithostratigraphy of chert in the New Insights from Field Studies and the Ocean Drilling Program: Geo- Franciscan Complex, Marin Headlands, California, in Blake, C., Jr., ed., logical Society of America Special Paper 349, p. 383–394. Franciscan Geology of Northern California: Los Angeles, Society of Pyle, D.G., Christie, D.M., and Mahoney, J.J., 1992, Resolving an iso topic Economic Paleontologists and Mineralogists, Pacifi c Section, Field Trip boundary within the Australian-Antarctica discordance: Earth and Plan- Guidebook 43, p. 51–70. etary Science Letters, v. 112, p. 161–178, doi: 10.1016/0012-821X(92) Murchey, B.L., and Jones, D.L., 1984, Age and signifi cance of chert in the 90014-M. Franciscan Complex, in the San Franciscan Bay region, in Blake, C., Jr., Pyle, D.G., Christie, D.M., Mahoney, J.J., and Duncan, R.A., 1995, Geo- ed., Franciscan Geology of Northern California: Los Angeles, Society of chemistry and geochronology of ancient southeast Indian and south- Economic Paleontologists and Mineralogists, Pacifi c Section, Field Trip west Pacifi c seafl oor: Journal of Geophysical Research, v. 100, no. B11, Guidebook 43, p. 23–30. p. 22,261–22,282, doi: 10.1029/95JB01424. Nehlig, P., 1993, Interactions between magma chambers and hydrothermal sys- Rampone, E., and Piccardo, G.B., 2000, The ophiolite-oceanic lithosphere tems; oceanic and ophiolitic constraints: Journal of Geophysical Research, analogue; new insights from the northern Apennines (Italy), in Dilek, Y., ser. B, Solid Earth and Planets, v. 98, no. 11, p. 19,621–19,633. Moores, E.M., Elthon, D., and Nicolas, A., eds., Ophiolites and Oceanic Niu, Y., Collerson, K.D., Batiza, R., Wendt, J.I., and Regelous, M., 1999, Origin Crust: New Insights from Field Studies and the Ocean Drilling Program: of enriched-type mid-ocean ridge basalt at ridges far from mantle plumes: Geological Society of America Special Paper 349, p. 21–34. The East Pacifi c Rise at 11°N: Journal of Geophysical Research, v. 104, Rautenschlein, M., Jenner, G.A., Hertogen, J., Hofmann, A.W., Kerrich, R., no. B4, p. 7067–7087, doi: 10.1029/1998JB900037. Schmincke, H.U., and White, W.M., 1985, Isotopic and trace element Ohara, Y., Stern, R.J., Ishii, T., Yurimoto, H., and Yamazaki, T., 2002, Peridotites composition of volcanic glasses from the Akaki Canyon, Cyprus; implica- from the Mariana Trough: First look at the mantle beneath an active back- tions for the origin of the Troodos ophiolite: Earth and Planetary Science arc basin: Contributions to Mineralogy and Petrology, v. 143, p. 1–18. Letters, v. 75, no. 4, p. 369–383, doi: 10.1016/0012-821X(85)90180-3. Okamura, H., Arai, S., and Kretschmar, L.M., 2006, Petrology of forearc Regelous, M., Nui, Y., Wendt, J.I., Batiza, R., Greig, A., and Collerson, K.D., peridotite from the Hahajima Seamount, the Izu-Bonin arc, with special 1999, Variations in the geochemistry of magmatism on the East Pacifi c reference to chemical characteristics of chromian spinel: Mineralogical Rise at 10°30N since 800 ka: Earth and Planetary Science Letters, v. 168, Magazine, v. 70, p. 15–26, doi: 10.1180/0026461067010310. p. 45–63, doi: 10.1016/S0012-821X(99)00048-5. O’Nions, R.K., Hamilton, P.J., and Evensen, N.M., 1977, Variation in 143Nd/144Nd Robertson, A.H.F., 1986, Geochemical evidence for the origin of Late Triassic and 87Sr/86Sr in oceanic basalts: Earth and Planetary Science Letters, v. 63, mélange units in the Oman Mountains as a small ocean basin formed by p. 229–240, doi: 10.1016/0012-821X(83)90039-0. continental rifting: Earth and Planetary Science Letters, v. 77, no. 3–4, Parkinson, I.J., and Pearce, J.A., 1998, Peridotites from the Izu-Bonin-Mariana p. 318–332, doi: 10.1016/0012-821X(86)90143-3. forearc (ODP Leg 125): Evidence for mantle melting and melt-mantle Robertson, A., and Xenophontos, C., 1993, Development of concepts concern- interaction in a supra-subduction zone setting: Journal of Petrology, v. 39, ing the Troodos ophiolite and adjacent units in Cyprus, in Prichard, H.M., p. 1577–1618. Alabaster, T., Harris, N.B.W., and Neary, C.R., eds., Magmatic Processes Pearce, J.A., 1982, Trace element characteristics of lavas from destructive and Plate Tectonics: Geological Society [London] Special Publication 76, plate boundaries, in Thorpe, R.S., ed., Andesites: Orogenic Andesites and p. 85–119. Related Rocks: Chichester, UK, John Wiley & Sons, p. 525–548. Robinson, P.T., and Malpas, J., 1990, The Troodos ophiolite: new perspective Pearce, J.A., 2003, Suprasubduction zone ophiolites: The search for modern on its origin and emplacement, in Malpas, J., Moores, E.M., Panayiotou, analogues, in Dilek, Y., and Newcomb, S., eds., Ophiolite Concept and the A., and Xenophontos, C., eds., Ophiolites: Oceanic Crustal Analogues: Evolution of Geologic Thought: Geological Society of America Special Proceedings of the Symposium “Troodos 1987”: Cyprus, Nicosia, Geo- Paper 373, p. 269–294. logical Survey Department, p. 13–26. Pearce, J.A., and Parkinson, I.J., 1993, Trace element models for mantle Robinson, P.T., Melson, W.G., O’Hearn, T., and Schminke, H.-U., 1983, Vol- melting; application to volcanic arc petrogenesis, in Prichard, H.M., canic glass compositions of the Troodos ophiolite, Cyprus: Geology, v. 11, Alabaster, T., Harris, N.B.W., and Neary, C.R., eds., Magmatic Processes p. 400–404, doi: 10.1130/0091-7613(1983)11<400:VGCOTT>2.0.CO;2. and Plate Tectonics: Geological Society [London] Special Publication 76, Robinson, P.T., Malpas, J., and Xenophontas, C., 2003, The Troodos Massif p. 373–403. of Cyprus: Its role in the evolution of the ophiolite concept, in Dilek, Y., Pearce, J.A., Lippard, S.J., and Roberts, S., 1984, Characteristics and tectonic and Newcomb, S., eds., Ophiolite Concept and the Evolution of Geologic signifi cance of supra-subduction zone ophiolites, in Kokelaar, B.P., and Thought: Geological Society of America Special Paper 373, p. 295–308. Suprasubduction-zone ophiolites: Is there really an ophiolite conundrum? 221

Rochette, P., Jenatton, L., Dupuy, C., Boudier, F., and Reuber, I., 1991, Dia- Stern, C.R., 1980, Geochemistry of Chilean ophiolites; evidence for the compo- base dikes emplacement in the Oman ophiolite; a magnetic fabric study sitional evolution of the mantle source of back-arc basin basalts: Journal with reference to geochemistry, in Peters, T., Nicolas, A., and Coleman, of Geophysical Research, v. 85, no. B2, p. 955–966. R.J., eds., Ophiolite Genesis and Evolution of the Oceanic Lithosphere: Stern, R.J., 2004, Subduction initiation: Spontaneous and induced: Earth and Muscat, Oman: Norwell, Massachusetts, Kluwer Academic Publisher, Planetary Science Letters, v. 226, p. 275–292, doi: 10.1016/S0012-821X p. 55–82. (04)00498-4. Rogers, R.D., Karason, H., and van der Hilst, R.D., 2002, Epeirogenic uplift above Stern, R.J., and Bloomer, S.H., 1992, Subduction zone infancy; examples from a detached slab in northern Central America: Geology, v. 30, p. 1031–1034, the Eocene Izu-Bonin-Mariana and Jurassic California arcs: Geological doi: 10.1130/0091-7613(2002)030<1031:EUAADS>2.0.CO;2. Society of America Bulletin, v. 104, no. 12, p. 1621–1636, doi: 10.1130/ Saccani, E., Padoa, E., and Photiades, A., 2003, Triassic mid-ocean ridge 0016-7606(1992)104<1621:SZIEFT>2.3.CO;2. basalts from the Argolis Peninsula (Greece); new constraints for the early Stern, R.J., Morris, J.D., Bloomer, S.H., and Hawkins, J.W., Jr., 1991, The oceanization phases of the Neo-Tethyan Pindos Basin, in Dilek, Y., and source of the subduction component in convergent margin magmas; trace Robinson Paul, T., eds., Ophiolites in Earth History: Geological Society element and radiogenic isotope evidence from Eocene boninites, Mariana [London] Special Publication 218, p. 109–128. forearc: Geochimica et Cosmochimica Acta, v. 55, no. 5, p. 1467–1481, Saltus, R.W., Hudson, T.L., Karl, S.M., and Morin, R.L., 2003, Rooted Brooks doi: 10.1016/0016-7037(91)90321-U. Range ophiolite; implications for Cordilleran terranes: Reply: Geology, Stoffers, P., Worthington, T., Hakinian, R., Petersen, S., Hannington, M.D., v. 31, no. 1, p. 92, doi: 10.1130/0091-7613(2003)031<0092:>2.0.CO;2. Türkay, M., and Scientifi c Shipboard Party, 2002, Silicic volcanism and Saunders, A.D., Tarney, J., Marsh, N.G., and Wood, D.A., 1980, Ophiolites as hydrothermal activity documented at Pacifi c-Antarctic Ridge: Eos (Trans- ocean crust or marginal basin crust; a geochemical approach, in Ophio- actions, American Geophysical Union), v. 83, no. 28, p. 303–304, doi: lites: Proceedings of the International Ophiolite Symposium: Cyprus, 10.1029/2002EO000215. Nicosia, p. 193–204. Sturm, M.E., Klein, E.M., Graham, D.W., and Karsten, J., 1999, Age con- Schilling, J.-G., 1973, Iceland mantle plume: Geochemical study of Reykjanes straints on crustal recycling to the mantle beneath the southern Chile Ridge: Nature, v. 242, p. 565–571, doi: 10.1038/242565a0. Ridge; He-Pb-Sr-Nd isotope systematics: Journal of Geophysical Shervais, J.W., 1990, Island arc and ocean crust ophiolites: Contrasts in the Research, ser. B, Solid Earth and Planets, v. 104, no. B3, p. 5097–5114, petrology, geochemistry, and tectonic style of ophiolite assemblages in doi: 10.1029/1998JB900107. the California Coast Ranges, in Malpas, J., Moores, E.M., Panayiotou, A., Sturm, M.E., Klein, E.M., Karsten, J.L., and Karson, J.A., 2000, Evidence and Xenophontos, C., eds., Ophiolites: Oceanic Crustal Analogues: Pro- for subduction-related contamination of the mantle beneath the south- ceedings of the Symposium “Troodos 1987”: Cyprus, Nicosia, Geological ern Chile Ridge; implications for ambiguous ophiolite compositions, in Survey Department, p. 507–520. Dilek, Y., Moores, E.M., Elthon, D., and Nicolas, A., eds., Ophiolites and Shervais, J.W., 2001, Birth, death, and resurrection: The life cycle of supra- Oceanic Crust: New Insights from Field Studies and the Ocean Drilling subduction zone ophiolites: Geochemistry, Geophysics, Geosystems, v. 2, Program: Geological Society of America Special Paper 349, p. 13–20. doi: 10.1029/2000GC000080. Suhr, G., and Edwards, S.J., 2000, Contrasting mantle sequences exposed in the Shervais, J.W., 2006, Signifi cance of subduction-related accretionary com- Lewis Hills Massif; evidence for the early arc-related history of the Bay plexes in early Earth processes, in Reimold, U., and Gibson, R., eds., of Islands ophiolite, in Dilek, Y., Moores, E.M., Elthon, D., and Nicolas, Early Earth Processes: Geological Society of America Special Paper 405, A., eds., Ophiolites and Oceanic Crust: New Insights from Field Studies p. 173–192. and the Ocean Drilling Program: Geological Society of America Special Shervais, J.W., and Kimbrough, D.L., 1985, Alkaline and transalkaline meta- Paper 349, p. 433–460. basalts in the Franciscan Complex mélange, California, in Morris, E.M., Sun, S.-S., and McDonough, W.F., 1989, Chemical and isotopic systematics of and Pasteris, J.D., eds., Mantle Metasomatism and Alkaline Magmatism: oceanic basalts: Implications for mantle composition and processes, in Geological Society of America Special Paper 215, p. 165–183. Saunders, A.D., and Norry, M.J., eds., Magmatism in the Ocean Basins: Shervais, J.W., Kimbrough, D.L., Renne, P.R., Hanan, B.B., Murchey, B., Snow, Geological Society [London] Special Publication 42, p. 313–345. C.A., Zoglman-Schuman, M.M., and Beaman, J., 2004, Multi-stage origin Taylor, B., and Exon, N.F., 1987, An investigation of ridge subduction in the of the Coast Range ophiolite, California: Implications for the life cycle of Woodlark-Solomons region: Introduction and overview, in Taylor, B., supra-subduction zone ophiolites: International Geology Review, v. 46, and Exon, N.F., eds., Marine Geology, Geophysics and Geochemistry of no. 4, p. 289–315. the Woodlark Basin–Solomon Islands: Menlo Park, California, Circum- Shervais, J.W., Murchey, B., Kimbrough, D.L., Renne, P., and Hanan, B.B., Pacifi c Council for Energy and Mineral Resources, Earth Science Series, 2005a, Radioisotopic and biostratigraphic age relations in the Coast p. 1–24. Range ophiolite, northern California: Implications for the tectonic evolu- Taylor, R.N., Murton, B.J., and Nesbitt, R.W., 1992, Chemical transects across tion of the western Cordillera: Geological Society of America Bulletin, intra-oceanic arcs: Implications for the tectonic setting of ophiolites, in v. 117, no. 5/6, p. 633–653, doi: 10.1130/B25443.1. Parson, L.M., Murton, B.J., Browning, P., eds., Ophiolites and Other Shervais, J.W., Zoglman-Schuman, M.M., and Hanan, B.B., 2005b, The Stony- Modern Oceanic Analogues: Geological Society [London] Special Pub- ford volcanic complex: A forearc seamount in the northern California lication 60, p. 117–132. Coast Ranges: Journal of Petrology, v. 46, no. 10, p. 2091–2128, doi: Tian, L., Zhao, G., Zhao, G., Shi, X., and Hailong, L., 2005, Geochemistry of 10.1093/petrology/egi048. basaltic lavas from the Mariana Trough: Evidence for infl uence of sub- Sisson, T.W., and Grove, T.L., 1993, Experimental investigations of the role duction component on generation of backarc basin magmas: International

of H2O in calc-alkaline differentiation and subduction zone magmatism: Geology Review, v. 47, p. 387–397. Contributions to Mineralogy and Petrology, v. 113, p. 143–166, doi: Umino, S., and Kushiro, I., 1989, Experimental studies on boninite petro- 10.1007/BF00283225. genesis, in Crawford, A.J., ed., Boninites: London, UK, Unwin Hyman, Sisson, V.B., Roeske, S.M., and Pavlis, T.L., eds., 2003, Geology of a Trans- p. 89–111. pressional Orogen Developed during Ridge-Trench Interaction along the Van der Laan, S.R., Flower, M.F.J., and Koster van Groos, A.F., 1989, Experi- North Pacifi c Margin: Geological Society of America Special Paper 371, mental evidence for the origin of boninites; near-liquidus phase rela- 375 p. tions to 7.5 kbar, in Crawford, A.J., ed., Boninites: London, UK, Unwin Staudigel, H., McCulloch, M.T., Zindler, A., and Perfi t, M.R., 1987, Complex Hyman, p. 112–147. ridge subduction and island arc magmatism; an isotopic study of the New Van Der Laan, S.R., Arculus, R.J., Pearce, J.A., and Murton, B.J., 1992, Georgia forearc and the Woodlark basin, in Taylor, B., and Exon, N.F., Petrography, mineral chemistry and phase relations of the basement eds., Marine Geology, Geophysics and Geochemistry of the Woodlark boninite series of Site 786, Izu-Bonin forearc, in Fryer, P., Pearce, J.A., Basin–Solomon Islands: Menlo Park, California, Circum-Pacifi c Council and Strokking, L.B., eds., Proceedings of the Ocean Drilling Program for Energy and Mineral Resources, Earth Science Series, p. 227–240. Scientifi c Results, Volume 125: College Station, Texas, Ocean Drilling Steinmann, G., 1906, Die ophiolitischen Zonen in den Mediterranean Kettenge- Program, p. 171–201. birgen (translated by Bernoulli, D., and Friedman, G.M.), in Dilek, Y., and Varne, R., Brown, A.V., and Falloon, T., 2000, Macquarie Island: Its geology, Newcomb, S., eds., Ophiolite Concept and the Evolution of Geological structural history, and the timing and tectonic setting of its N-MORB Thought: Geological Society of America Special Paper 373, p. 77–92. to E-MORB magmatism, in Dilek, Y., Moores, E.M., Elthon, D., and 222 Metcalf and Shervais

Nicolas, A., eds., Ophiolites and Oceanic Crust: New Insights from Field Subduction Factory: Active Margins and Marginal Basins of the West- Studies and the Ocean Drilling Program: Geological Society of America ern Pacifi c: American Geophysical Union Geophysical Monograph 138, Special Paper 349, p. 301–320. p. 59–82. Volker, F., Alther, R., Jochum, K.-P., and McCulloch, M.T., 1997, Quaternary Wood, D.A., 1980, The application of a Th-Hf-Ta diagram to problems of volcanic activity of the southern Red Sea: New data and assessment of tectonomagmatic classifi cation and to establishing the nature of crustal models on magma sources and Afar plume-lithosphere interaction: Tec- contamination of basaltic lavas of the British Tertiary volcanic prov- tonophysics, v. 278, p. 15–29, doi: 10.1016/S0040-1951(97)00092-9. ince: Earth and Planetary Science Letters, v. 50, no. 1, p. 11–30, doi: Walker, D.A., and Cameron, W.E., 1983, Boninite primary magmas; evidence 10.1016/0012-821X(80)90116-8. from the Cape Vogel Peninsula: PNG: Contributions to Mineralogy and Woodhead, J.D., Eggins, S.M., and Johnson, R.W., 1998, Magma genesis in Petrology, v. 83, no. 1–2, p. 150–158, doi: 10.1007/BF00373088. the New Britain island arc: Further insights into melting and mass trans- Wallin, E.T., and Metcalf, R.V., 1998, Supra-subduction zone ophiolite formed fer processes: Journal of Petrology, v. 39, p. 1641–1668, doi: 10.1093/ in an extensional forearc: Trinity terrane, Klamath Mountains, California: petrology/39.9.1641. The Journal of Geology, v. 106, p. 591–608. Zindler, A., and Hart, S., 1986, Chemical geodynamics: Annual Review of Weaver, B.L., 1991, The origin of ocean island basalt end-member compositions: Earth and Planetary Sciences, v. 14, p. 493–571, doi: 10.1146/annurev.ea. Trace element and isotopic constraints: Geology, v. 104, p. 381–397. 14.050186.002425. Weins, D.A., and Smith, G.P., 2003, Seismological constraints on structure and fl ow patterns within the mantle wedge, in Eiler, J., ed., Inside the MANUSCRIPT ACCEPTED BY THE SOCIETY 16 JULY 2007

Printed in the USA