Mantle plume or slab window?: Physical and geochemical constraints on the origin of the Caribbean oceanic plateau

 Antikes Griechenland

 6 views
of 11

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Description
Mantle plume or slab window?: Physical and geochemical constraints on the origin of the Caribbean oceanic plateau
Share
Tags
Transcript
  Mantle plume or slab window?: Physical and geochemical constraints on the srcinof the Caribbean oceanic plateau Alan R. Hastie ⁎ , Andrew C. Kerr School of Earth and Ocean Sciences, Cardiff University, Main Building, Park Place, Cardiff, CF10 3YE, UK  a b s t r a c ta r t i c l e i n f o  Article history: Received 22 December 2008Accepted 10 November 2009Available online 17 November 2009 Keywords: Caribbean oceanic plateaumantle plumeslab windowprimary magmaCuraçao The Caribbean oceanic plateau formed in the Paci fi c realm when it erupted onto the Farallon plate from theGalapagos hotspot at ∼ 90 Ma. The plateau was subsequently transported to the northeast and collided withthe Great Arc of the Caribbean thus initiating subduction polarity reversal and the consequent tectonicemplacement of the Caribbean plate between the North and South American continents. The plateaurepresents a large outpouring of ma fi c volcanism, which has been interpreted as having formed by melting of a hot mantle plume. Conversely, some have suggested that a slab window could be involved in forming theplateau. However, the source regions of oceanic plateaus are distinct from N-MORB (the likely sourcecomposition for slab window ma fi c rocks). Furthermore, melt modelling using primitive (high MgO)Caribbean oceanic plateau lavas from Curaçao, shows that the primary magmas of the plateau contained ∼ 20 wt.% MgO and were derived from 30 to 32% partial melting of a fertile peridotite source region whichhad a potential temperature ( T  p ) of 1564 – 1614 °C. Thus, the Caribbean oceanic plateau lavas are derivedfrom decompression melting of a hot upwelling mantle plume with excess heat relative to ambient uppermantle. Extensional decompression partial melting of sub-slab asthenosphere in a slab window with anambient mantle  T  p  cannot produce enough melt to form a plateau. The formation of the Caribbean oceanicplateau by melting of ambient upper mantle in a slab window setting, is therefore, highly improbable.© 2009 Elsevier B.V. All rights reserved. Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2842. Geological background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2852.1. Mantle plumes and the formation of LIPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2852.2. Slab windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2853. Volcanic rock types associated with slab window and mantle plume (oceanic plateau) magmatism . . . . . . . . . . . . . . . . . . . . . 2863.1. Rock types and geochemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2863.2. Volume of volcanic products from a slab window and oceanic plateau . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2864. Physical and geochemical in fl uences on mantle melting during oceanic plateau formation . . . . . . . . . . . . . . . . . . . . . . . . . 2874.1. Anhydrous nature of plateau lavas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2874.2. Previous  T  p  calculations for the source of oceanic plateaus (mantle plumes) and slab windows (ambient upper mantle). . . . . . . . 2874.3. Depth of melting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2874.4. Oceanic plateau primary magma compositions and degrees of partial melting . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2884.4.1. Previous work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2884.4.2. Calculated primary magma compositions of primitive lavas from Curaçao, Dutch Antilles . . . . . . . . . . . . . . . . . . 2884.4.3. Calculated primary magmas from a slab window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2895. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2895.1. Melt generation in mantle plumes: a synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2895.2. Melt generation in a slab window environments: a synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2905.3. Could a mantle plume pass through a slab window? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2916. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291 Earth-Science Reviews 98 (2010) 283 – 293 ⁎  Corresponding author. E-mail address:  hastiear@cf.ac.uk (A.R. Hastie).0012-8252/$  –  see front matter © 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.earscirev.2009.11.001 Contents lists available at ScienceDirect Earth-Science Reviews  journal homepage: www.elsevier.com/locate/earscirev  1. Introduction Knowledge of the tectonomagmatic history of the Caribbean plate(Fig. 1) is important in helping us to understand the palaeogeographyof the inter-American region from the Jurassic. Of particularimportance is the development, and subsequent closure, of thepalaeogateway between North and South America and the associatedimpacts on the global climate (e.g. Droxler et al., 1998; Schneider andSchmittner, 2006).Most of the Caribbean plate consists of a 8 – 20-km-thick LateCretaceous oceanic plateau ( ∼ 6×10 5 km 2 ) that formed in the Paci fi c(e.g. Edgar et al., 1971; Mauffret and Leroy, 1997; Kerr et al., 2003)and is possibly derived from the initial plume head phase of theGalapagos hotspot (e.g. Hoernle et al., 2002; Geldmacher et al., 2003;Thompson et al., 2003). Although somewhat controversial (e.g.Pindell et al., 2006), many consider the Caribbean oceanic plateau tohave erupted onto the Farallon plate and subsequently transported tothe northeast tocollide withalargeintra-oceanic arc[theGreatArc of the Caribbean, Burke (1988)] in the late Cretaceous (Fig. 2a) (e.g. Duncan and Hargraves, 1984; Burke, 1988; White et al., 1999;Thompson et al., 2003; Kerr et al., 2003; Mann et al., 2007). TheGreat Arc was located at the western side of the oceanic gap (the Fig. 1.  (a) Map of the Caribbean and Central American region showing the location of the obducted portions of the Caribbean oceanic plateau. Guatemala (Gma), El Salvador (ES),Costa Rica (CR), Panama (Pna), Swan Islands Transform Fault Zone (SITFZ), Oriente Transform Fault Zone (OTFZ), Plantain Garden-Enriquillo Fault Zone (PG-EFZ), Nicoya Complex(NC), Herradura Complex (HC), Quepos Complex (QC), Osa Complex (OC), Sona Complex (SC), Azuero Complex (AC), Serrania de Baudo Complex (SdBC) (from Hastie et al., 2008)and (b) General geology map of Curaçao, Dutch Antilles showing the location of the Curaçao Lava Formation (modi fi ed from Kerr et al., 1996c).Locality numbers refer to locations inTable 1.284  A.R. Hastie, A.C. Kerr / Earth-Science Reviews 98 (2010) 283 –  293  Proto-Caribbean) which had been opening between the North andSouth American continents since the Jurassic (Fig. 2a).Inthis model, the Caribbean plateauwould have beentoo thick, hotand buoyant to subduct beneath the American or Proto-Caribbeansubductionzones(e.g.Saunderset al.,1996).Accordingly,thesouthernportionsoftheplateauwouldhavebeenobductedontothecontinentalmargin of South America forming extensive accreted sequences inColombiaandEcuador(Fig.2a)(Kerretal.,1996b,2002a,b).Incontrast, whenthenorthernportionoftheplateaucollidedwiththeGreatArcof the Caribbean it clogged the subduction zone and initiated subductionpolarity reversal and subduction back-step such that the Proto-Caribbean crust began subducting in a westerly direction beneath theoceanic plateau (Duncan and Hargraves, 1984; Burke, 1988; Kerr et al.2003).Overthelast ∼ 80MatheplateauandthesegmentedGreatArcof the Caribbean were tectonically emplaced between the westwardmoving North and South American continents to form the Caribbeanplate (Duncan and Hargraves, 1984; Burke, 1988; Sinton et al., 1997;Hauff et al., 2000a,b; Kerr et al., 2003; Mann et al., 2007).Conversely, others have proposed a signi fi cantly earlier Aptian/Albian(125 – 99.6 Ma)subductionpolarityreversal(Fig.2b)(e.g.Lebron and Per fi t, 1993, 1994; Kesler et al., 2005; Pindell et al., 2005; EscuderViruete et al., 2007; Marchesi et al., 2007). This has led Pindell et al.(2006) to challenge the mantle plume model by proposing that theCaribbean oceanic plateau may have been largely formed by magma-tism associated with a slab window. Pindell and Kennan (2001) andPindelletal.(2006)arguefromplatereconstructionsthataslabwindowexisted in the Caribbean region from the Aptian/Albian to the earlyCampanian (83.5 – 70.6 Ma) when an active Proto-Caribbean sea fl oorspreading centre subducted beneath the Great Arc of the Caribbean(Fig. 2b). Pindell et al. (2006) proposed that partial asthenospheric melting facilitated by the slab window would form the thickened crustof the Caribbean oceanic plateau.The aims of this paper are to compare oceanic plateau rocks tovolcanism associated with slab windows and to determine if anoceanic plateau can be formed by magmatism resulting from a slabwindow. Furthermore, primitive lavas from the island of Curaçao,DutchAntilles(Fig.1)willbeassessedusingthePRIMELT2softwareof Herzberg and Asimow (2008) to determine the composition, degreeof partial melting, and the potential temperature of their primarymagmas in order to resolve the likely af  fi nity of the mantle sourceregion of the Caribbean oceanic plateau. This information will used toplace constraints on the tectonomagmatic evolution of the Caribbeanplate in the Late Cretaceous. 2. Geological background  2.1. Mantle plumes and the formation of LIPs In the oceanic environment large igneous provinces (LIPs) arerepresented by oceanic plateaus, oceanic basins and aseismic ridges(Cof  fi n and Eldholm, 1994). It has been demonstrated both geochem-ically (e.g. Kempton et al., 2000; Herzberg and O'Hara, 2002;Thompson et al., 2003; Fitton and Godard, 2004) and by physicaland computational modelling (Richards et al., 1989; Campbell andGrif  fi ths, 1990; Farnetani and Richards, 1995; Farnetani et al., 2002;Farnetani and Samuel, 2005; Campbell, 2007) that voluminousoceanic plateaus can be formed from the partial adiabatic decom-pression melting of a hot, ascending, deep-mantle-derived, compo-sitionally heterogeneous mantle plume head as it collides with thebase of the lithosphere.In characterising the thermal properties of a given mantle sourceregion McKenzie and Bickle (1988) de fi ned the term mantle potentialtemperature ( T  p ), which represents the temperature of a mass of convecting mantle on the Earth's surface if it was to ascend along anadiabat (i.e. it neither loses or gains heat to its surroundings duringascent)anddidnotmelt.Calculationofthe T  p enablesustodetermineif a primary magma is derived from a mantle source region withexcess heat relative to ambient upper mantle (e.g. Bown and White,1995).Decompression partial melting, combined with higher calculated T  p  values within a mantle plume head ( Δ T  =100 – 400 °C) relative tothe ambient upper mantle ( T  p  ∼ 1280 – 1475 °C) (e.g. McKenzie andBickle, 1988; Kinzler and Grove, 1992; Herzberg and O'Hara, 2002;Putirka,2005;Courtieretal.,2007;Herzbergetal.,2007;Putirkaetal.,2007), will rapidly produce large amounts of partial melt [15 – 30% e.g.Kerr et al. (2002b); Chazey and Neal (2004), Fitton and Godard(2004); Herzberg (2004)] that will eventually form a LIP, such as anoceanic plateau.  2.2. Slab windows A slab window is formed at a ridge – trench – trench triple boundarywhenanactiveoceanicspreadingcentresubductsbeneathanoverridingplate (Dickinson and Snyder, 1979). As the ridge subducts the trailingedgesofeitheroneorbothoftheplateswillcontinuetodivergetoforman ever-widening, slabless gap between the two subducting plates Fig. 2.  Diagrams representing the two models for the evolution of the Caribbean plate(a) modi fi ed from Burke (1988) and Kerr et al. (1999). Northeast dipping subduction continues until the Turonian-Santonian. Subduction polarity reversal occurs when theCaribbean oceanic plateau collides with the Great Arc of the Antilles. Plateau material isobductedontoSouthAmericaandtheGreatArc(b)diagrammodi fi edfromPindelletal.(2005, 2006). Subduction polarity reversal occurs in the Aptian. Dark grey SouthAmerica represents the location of the continent at ∼ 120 Ma.285  A.R. Hastie, A.C. Kerr / Earth-Science Reviews 98 (2010) 283 –  293  (ThorkelsonandTaylor,1989;Thorkelson,1996).Itisthisslablessgapinthemantle thatis termed a slabwindow, anda  “ slab-free region ”  refersto the area on the overriding plate above the slab window (Dickinson,1997). As a slab window forms it is likely to promote rifting in theoverlying lithosphere (e.g. Thorkelson, 1996). This is because the twoseparate down-going plates are coupled to the overlying plate and asthey diverge during descent they produce extension in the overridingplate (Thorkelson, 1996). Thus, the lithosphere above a slab windowrepresents a passive extensional regime. 3. Volcanic rock types associated with slab window and mantleplume (oceanic plateau) magmatism  3.1. Rock types and geochemistry Magmatism in a slab-free region is both compositionally distinc-tive and diverse (e.g. Thorkelson, 1996) and is usually bordered by “ normal ”  arc volcanism. In general, slab windows are associated withrelatively small volumes of volcanism which includes: (1) silicic fore-arc magmatism (e.g. Breitsprecher et al., 2003); (2) adakites (meltsfrom a ma fi c protolith), magnesian andesites (bajaites) and Nb-enriched basalts (e.g. Aguillón-Robles et al., 2001; Calmuset al., 2003;Bellon et al., 2006); and (3) highly complex decompression partialmelting and mixing of sub-slab and/or supra-slab asthenosphere toform alkaline and tholeiitic ma fi c volcanism (e.g. Benoit et al., 2002;Gorring et al., 2003; Bellon et al., 2006; Pallares et al., 2007). Theseupwelling asthenospheric melts can subsequently interact withoverlying enriched lithospheric and crustal sources resulting inanatexis and mixing to form more enriched ma fi c magmas (e.g. Coleand Basu, 1992, 1995; Gorring et al., 2003).The adakites and magnesian andesites (including Nb-enrichedbasalts) are considered to be derived from partial melting of a ma fi cprotolith (i.e. the slab edges) and subsequent interaction and meltingoftheoverlyingmantlewedgerespectively(e.g.Aguillón-Roblesetal.,2001; Calmus et al., 2003). The third type of magmatism consists of both enriched alkaline and tholeiitic ocean island basalt (OIB)-type(Gorring et al., 1997, 2003; Pallares et al., 2007) and depleted alkalineand tholeiitic mid-ocean ridge basalt (MORB)-type (Cole and Basu,1992, 1995; Benoit et al., 2002; Bellon et al., 2006) mantlecomponents and lavas. Additionally,  “ normal ”  arc-derived volcanismcansporadicallyoccurintheslab-freeregion(e.g.Gorringetal.,1997;Pallaresetal.,2007)suggestingthatasthenosphericmantlewhichhasbeen metasomatised by slab-related  fl uids exists, and can be melted,in a slab window environment.In contrast,oceanic plateaussuchas the OntongJava (OJP) andtheCaribbean are composed of voluminous, predominantly tholeiiticbasalts with minor picritic and komatiitic successions (e.g. Kerr et al.,1996a,b,c; Neal et al., 1997; Arndt et al., 1997, 1998; Révillon et al.,1999; Hauff et al., 2000a,b; Tejada et al., 2002; Fitton and Godard,2004; Hastie et al., 2008). Oceanic plateau rocks are mostly tholeiiticwith predominantly  fl at, primitive mantle-normalised multi-elementpatterns (Fig. 3) and these characteristics contrast markedly with theenriched ma fi c alkaline, and slab-melt related rocks formed in slabwindow environments (Fig. 3).Furthermore, oceanic plateau rocks lack the negative Nb and Taanomalies that are present in slab window related adakites and high-magnesian andesites (e.g. Pallares et al., 2007) as well as othergeochemical characteristics of these igneous rocks e.g. SiO 2 N 56%,Al 2 O 3 N 15%, MgO generally  b 3%, low Y ( b 18ppm) and HREE (Yb b 1.9 ppm)foranadakiteorSrupto3000ppm,Ba N 1000 ppmandhighNa/K ratios for magnesian andesites (e.g. Saunders et al., 1987; Defantet al., 1992; Yogodzinski et al., 1995). Trace element and radiogenicisotope compositions (Figs. 3 and 4) also reveal that oceanic plateaubasaltsarederivedfrommantlesourceregionsdistinctfromdepletedN-MORB-source upper mantle, with predominantly higher LREE/HREEratios and lower (more enriched)  ε  Nd ( i ) values than present-day N-MORB (e.g. Kerr et al., 1996a; Kempton et al., 2000; Tejada et al., 2004;Hastie et al., 2007, 2008).Consequently, the ma fi c alkaline and tholeiitic N-MORB and OIB-type lavas and slab-related adakitic volcanismformed in slab windowenvironmentsarecompositionallyverydifferenttothemantleplume-derived tholeiitic basalts, picrites and komatiites of oceanic plateaus.  3.2. Volumeofvolcanicproductsfromaslabwindowandoceanicplateau Oceanic plateaus are voluminous LIPs formed by massive out-pourings of ma fi c extrusive lavas on  “ normal ”  oceanic crust and areoften associated with extensive intrusive activity (e.g. Cof  fi n and Fig. 3.  Primitive mantle-normalised multi-element diagrams of   ∼ 90 Ma obductedCaribbean oceanic plateau material from Ecuador, Costa Rica and Jamaica. Normalisingvalues from McDonough and Sun (1995). Caribbean oceanic plateau data taken fromReynaudetal.(1999),Hauffetal.(2000b),Kerretal.(2002b),Mambertietal.(2003)andHastie et al. (2008). OJP data from Mahoney et al. (1993) and Fitton and Godard (2004), BajaCaliforniaalkalinebasaltandadakiteslabwindowlavasfromPallaresetal.(2007)andAguillón-Robles et al. (2001). N-MORB data taken from Sun and McDonough (1989). 286  A.R. Hastie, A.C. Kerr / Earth-Science Reviews 98 (2010) 283 –  293  Eldholm, 1994). The OJP (the largest known LIP) has an estimatedmelt volumeof 44 – 50×10 6 km 3 and asurface areaapproximately thesame size as Western Europe (Eldholm and Cof  fi n, 2000; Fitton andGodard, 2004), whereas the estimated srcinal melt volume of theCaribbean oceanic plateau is ∼ 4×10 6 km 3 (Kerr, 1998).Furthermore, Cof  fi n and Eldholm (1994) have shown that thevolumes of melt required to form oceanic plateaus and continental fl ood basalts with 5 – 30% partial melting, would be so large that if thevolumes were represented by a sphere it would extend from the baseof the lithosphere, through the entire asthenosphere and into thelower mantle below the 670 km discontinuity. Consequently, theseLIPs are likely formed by causal mechanisms distinct from normalupper mantle partial melting processes.Conversely, ma fi c volcanism above slab windows is much smallerin volume; Gorring et al. (1997) report minimum total melt volumesof slab window plateau volcanism in Southern Patagonia of  ∼ 1000 km 3 . Similarly small melt volumes have also been observedin other slab window environments, such as, the Antarctic Peninsula(e.g. Hole, 1990). 4. Physical and geochemical in fl uences on mantle melting during oceanic plateau formation When comparing the partial melting of a mantle plume and a slabwindow source region a number of basic questions need to beconsidered: (a) is the source hydrous or anhydrous? (b) are the meltsderivedfrommantlesourceregionswithsimilar T  p values?(c)arethemelts derived from similar depths? (d) are the melts formed bycomparable degrees of partial melting? and (e) are the mantle sourceregions compositionally similar? 4.1. Anhydrous nature of plateau lavas Kerr and Mahoney (2007) have shown that, apart from twolocalities in the Caribbean (gabbros and pegmatites in Bolivar,Colombia and one komatiite  fl ow on Gorgona), the source regionsresponsible for forming the Caribbean oceanic plateau magmas, areanhydrous mantle peridotite. Thus, if the oceanic plateau was derivedfromdecompressionmeltinginaslabwindow,ahydrated,arc-derivedsupra-slab mantle source region cannot be considered as a viablesource region. This point is further demonstrated by the fact that thelavas of the Caribbean oceanic plateau (even those with a hydratedsource)lackanysubduction-relatedgeochemicalsignal(e.g.La/Nb ≫ 1),whichalso argueagainst the involvementof a supra-slabsource region(Fig. 3)(Kerretal.,2004).Therefore, anyproposed slabwindowsource region for Caribbean oceanic plateau has to be completely devoid of any geochemical signature of mantle that has been compositionallymodi fi ed by subduction-related processes. 4.2. Previous T   p  calculations for the source of oceanic plateaus (mantle plumes) and slab windows (ambient upper mantle) Herzberg (2004) calculates that the primary magmas of the OJPhad a  T  p  of 1500 – 1560 °C. Based on Caribbean oceanic plateau lavasfrom Gorgona Island [located ∼ 60 km off the Colombian Paci fi c coast(Fig.1a)] Herzberg and O'Hara (2002) and Herzberg et al. (2007) calculate that the  T  p  of high-MgO komatiites is 1520 – 1570 °C whileassociated picrites have higher  T  p  values of 1600 – 1700 °C.In contrast, the most recent studies of  Herzberg et al. (2007) andPutirka et al. (2007) based on phase equilibria and olivine thermom-etry respectively, calculate the average  T  p  of upper mantle sourcesbeneath mid-ocean ridge systems to be  ∼ 1300 – 1454 °C, similar toprevious studies (e.g.  ∼ 1280 – 1475 °C, McKenzie and Bickle, 1988;Kinzler and Grove, 1992; Asimow et al., 2001; Herzberg and O'Hara,2002; Wang et al., 2002; Putirka, 2005). It should be noted, thatalthough this review uses the full range of calculated ambient mantletemperatures, many recent studies favour a  T  p  range of 1300 – 1400 °Cfor the upper mantle (e.g. McKenzie et al., 2005; Courtier et al., 2007;Herzberg et al., 2007). Nevertheless, although the  T  p  of the uppermantle is variable, even the highest estimated values are still not ashot as the estimated  T  p  of the Caribbean and OJP oceanic plateausourceregions.Consequently,mantleplumesourceregionsareclearlyassociated with excess  T  p  relative to the mantle sources of N-MORB. 4.3. Depth of melting  Herzberg and O'Hara (2002) calculated using forward and inversephase equilibria modelling that the Gorgona komatiites and picriteswere derived from mantle sources at depths of 115 – 140 km and240 km respectively. Herzberg (2004) also concluded that similarmeltsin theOJP werederivedfroma mantlesourceat adepthof 108 – 132 km. These high pressures ( N 2.5 GPa) indicate that the primarymagmas are derived from a garnet peridotite mantle source region.This further explains the high-MgO contents of the mantle plumeprimary magmas because not only do high degrees of partial meltingproduce high-MgO melts, but melts in equilibrium with garnetperidotite (high pressures) become signi fi cantly enriched in MgO(e.g.Herzberg,1992).HerzbergandO'Hara(2002)andHerzbergetal. (2007) have shown that primary magmas derived from oceanicplateaus and continental  fl ood basalt provinces e.g. from Gorgona,Baf  fi n Island and West Greenland, have higher MgO contents ( ∼ 17 – 19 wt.%) than primary magmas from N-MORB (e.g. 10 – 13 wt.% fromthe Siqueiros transform fault).McKenzieandBickle(1988)usedparameterisedexperimentalandmathematical temperature, pressure and melt volume data of garnetperidotite to determine several adiabatic decompression pathways(Fig. 5). According to Fig. 5  ∼ 1280 °C upper mantle would have toascend to  b 50 km depth to promote partial melting and upper mantlewith estimated high  T  p  values of 1400 °C and  ∼ 1475 °C (e.g. Kinzlerand Grove, 1992; Courtier et al., 2007) would have to ascend to lowerthan 75 and 100 km respectively. These depths are similar to thecalculated  b 100 km depth of decompression melting beneath  “ pas-sive ” mid-oceanridges(e.g. Johnsonetal.,1990;HerzbergandO'Hara,2002; Putirka et al., 2007). Fig. 4.  ε  Nd ( i ) – ε  Hf  ( i ) diagram demonstrating that the isotopic composition of oceanicplateau andMORB lavas aredistinct. Primary magmas fromtheOJP(Tejadaetal., 2004)plot within the more compositionally heterogeneous Caribbean plateau  fi eld close tolavas from the CLF. The large  “ star ”  represents the composition of the OJP theoreticalmantle source region (Tejada et al., 2004). The Caribbean oceanic plateau  fi eld isconstructed with data from Colombia, Costa Rica, Curaçao, DSDP Leg 15, Galapagos,Gorgona and Jamaica. The references for the plateau and MORB data are in Appendix A.287  A.R. Hastie, A.C. Kerr / Earth-Science Reviews 98 (2010) 283 –  293
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks