Laterita niquelífera

Ophiolite-Related Ultramafic Rocks (Serpentinites) in theCaribbean Region: A Review of their Occurrence, Composition,

Origin, Emplacement and Ni-Laterite Soil Formation

Ultramafic rocks, mainly serpentinized peridotites of mantle origin, are mostly associated with the ophiolites ofMesozoic age that occur in belts along three of the margins of the Caribbean plate. The most extensive expo-sures are in Cuba. The ultramafic-mafic association (ophiolites) were formed and emplaced in several differenttectonic environments. Mineralogical studies of the ultramafic rocks and the chemistry of the associated maficrocks indicate that most of the ultramafic-mafic associations in both the northern and southern margins of theplate were formed in arc-related environments. There is little mantle peridotite exposed in the ophiolitic associa-tions of the west coast of Central America, in the south Caribbean in Curacao and in the Andean belts in Colom-bia. In these occurrences the chemistry and age of the mafic rocks indicates that this association is mainly partof the 89 Ma Caribbean plateau province. The age of the mantle peridotites and associated ophiolites is proba-bly mainly late Jurassic or Early Cretaceous. Emplacement of the ophiolites possibly began in the Early Creta-ceous in Hispaniola and Puerto Rico, but most emplacement took place in the Late Cretaceous to Eocene (e.g.Cuba). Along the northern South America plate margin, in the Caribbean mountain belt, emplacement was bymajor thrusting and probably was not completed until the Oligocene or even the early Miocene. Caribbean man-tle peridotites, before serpentinization, were mainly harzburgites, but dunites and lherzolites are also present. Indetail, the mineralogical and chemical composition varies even within one ultramafic body, reflecting meltingprocesses and peridotite/melt interaction in the upper mantle. At least for the northern Caribbean, uplift (post-emplacement tectonics) exposed the ultramafic massifs as a land surface to effective laterization in the begin-

Geologica Acta, Vol .4 , N1-2, 2006, 237-263

Avai lable onl ine at www.geologica-acta.com

UB-ICTJA 237

A B S T R A C T

J.F. LEWIS G. DRAPER J.A. PROENZA J. ESPAILLAT and J. JIMNEZ

Dept. of Earth and Environmental Sciences, The George Washington UniversityWashington DC 20052. E-mail: [email protected]

Dept. of Earth Sciences, Florida International UniversityMiami, FL 33199. E-mail: [email protected]

Departament de Cristalografia, Mineralogia i Dipsits Minerals, Facultat de Geologia, Universitat de Barcelona, Mart i Franqus s/n 08028 Barcelona, Spain.

E-mail: [email protected]

Corporacin Minera DominicanaMximo Gmez s/n, Dominican Republic. E-mail: [email protected]

Falconbridge DominicanaBonao, Dominican Republic. E-mail : [email protected]

1 2 3 4 5

1

2

3

4

5

INTRODUCTION

Ophiolite-related ultramafic rocks, mainly serpen-tinites, crop out along three of the margins to theCaribbean Plate (Fig. 1). They are most abundant alongthe northern plate margin, and constitute about 7 per cent(7500 km2) of the land surface of Cuba. Most of theseultramafic rocks were originally peridotites, formed in the

upper mantle, and were then altered to serpentinites, com-pletely or partly, by crustal fluids during their passage totheir present tectonic position. When uplifted and exposedto weathering, these partially to completely serpentinizedperidotites were altered to form a soil that has a composi-tion determined, not only by the composition of the ultra-mafic rocks, but by many other factors, the main one ofwhich is climate.

Ophiolite-related serpentinites in the Caribbean regionJ.F. LEWIS et al.

238Geolog ica Acta , Vo l .4 , N1-2, 2006, 237-263

FIGURE 1 Distribution of ophiolite-related ultramafic rocks around the margins of the Caribbean Plate: 1: Sierra de Santa Cruz; 2: Baja Verapaz Unit;3 Juan de Paz; 4: El Tambor Group (South Motagua and North Motagua); 5: Cajlbana; 6: Habana-Matanza; 7: Villa Clara; 8: Escambray; 9: Cam-agey; 10: Holgun; 11: Mayar-Cristal; 12: Alto de La Corea; 13: Moa-Baracoa; 14: Sierra del Convento; 15: Arntully; 16: North Coast Belt; 17:Loma Caribe; 18: Monte del Estado; 19: Ro Guanajibo; 20: Bermeja; 21: La Franja Costera; 22: Loma de Hiero; 23: Villa de Cura; 24: Guajira Penin-sula (serpentinites of Cabo de la Vela); 25: Dunita de Medelln (Aburr ophiolite); 26: Santa Elena; 27: Rio San Juan; 28: Siuna. Main localities arenumbered in italics.

Peridotite. Serpentinite. Ni-Laterite. Ophiolite. CaribbeanKEYWORDS

ning of the Miocene. Tectonic factors, determining the uplift, exposing the peridotites to weathering varied. Inthe northern Caribbean, in Guatemala, Jamaica, and Hispaniola, uplift occurred as a result of transpresionalmovement along pre-existing major faults. In Cuba, uplift occurred on a regional scale, determined by isostaticadjustment. In the south Caribbean, uplift of the Cordillera de la Costa and Serrania del Interior exposing theperidotites, also appears to be related to strike-slip movement along the El Pilar fault system. In the Caribbean,Ni-laterite deposits are currently being mined in the central Dominican Republic, eastern Cuba, northernVenezuela and northwest Colombia. Although apparently formed over ultramafic rocks of similar compositionand under similar climatic conditions, the composition of the lateritic soils varies. Factors that probably deter-mined these differences in laterite composition are geomorphology, topography, drainage and tectonics. Accor-ding to the mineralogy of principal ore-bearing phases, Dominican Ni-laterite deposits are classified as thehydrous silicate-type. The main Ni-bearing minerals are hydrated Mg-Ni silicates (serpentine and garnierite)occurring deeper in the profile (saprolite horizon). In contrast, in the deposits of eastern Cuba, the Ni and Cooccurs mainly in the limonite zone composed of Fe hydroxides and oxides as the dominant mineralogy in theupper part of the profile, and are classified as the oxide-type.

In the Caribbean, as a result of its tropical climate,thick, nickel-rich laterite soils are developed over theexposures of peridotites and serpentinites. The soilsdeveloped are characterized by variable contents of mag-nesium and iron, with relatively high contents of nickel,chromium and cobalt. Concentrations of nutrient ele-ments calcium, potassium and phosphorus are low. TheseNi-laterites are currently being mined in the centralDominican Republic (Loma Caribe), eastern Cuba(Mayari, Nicaro, Moa Bay, Punta Gorda), northernVenezuela (Loma de Nquel) and northwest Colombia(Cerro Matoso). Only the Dominican and eastern Cubandeposits are reviewed in this report.

A characteristic flora is developed on these lateriticsoils. Brooks (1987) stated that the Cuban serpentiniteflora is the third richest (in number of species) in theworld. A wide variety of flora are developed over differ-ent areas of serpentinized peridotite in the Caribbean butlittle has been done to understand the relation between theflora and the composition of the laterite soils.

In this brief report we summarize the present knowl-edge of the distribution of ultramafic rocks (mainly ser-pentinites and serpentinized peridotites) the Caribbean,their geological associations, the processes by whichthey were formed, tectonically emplaced, uplifted andexposed to weathering. Emphasis is placed on theprocesses of laterite soil formation from the ultramaficrocks. Some geological factors are known better thanothers, but it is hoped that this review will give a betterappreciation of our present knowledge of the processesinvolved. This is the first overall review of Caribbeanserpentinites since those of Dengo (1972) and Wadge etal. (1984).

NATURE AND ORIGIN OF CARIBBEAN OPHIOLITE-RELATED ULTRAMAFIC ROCKS: A SUMMARY

Most of the ultramafic rocks in the Caribbean occur inthe Mesozoic orogenic belts along the margins of theCaribbean Plate and are alpine-type peridotites (mantle-tectonites), partly or completely altered to serpentinite. Inmany of the occurrences, the alpine-type peridotites areclosely associated with layered and massive gabbros,basalts, mafic dykes, minor plagiogranites, chromititesand sedimentary rocks, which form the ophiolite associa-tion. An idealized cross section of the Moa-Baracoa ophi-olite, one of the few more complete ophiolites in thenorthen Caribbean, is shown in Fig. 2. However, general-ly, the ophiolites are highly dismembered; not all of themembers of the ophiolite suite are present, nor is theophiolite sequence as defined by the 1972 Penrose Con-ference (Anonymous, 1972). Although this problem is

common in many ophiolitic occurrences in the world, ithas led to considerable controversy and misunderstandingof the use of the term ophiolite, and this applies verymuch to the Caribbean.

Detailed field, petrological and geochemical studies inthe past 25 years, on rocks of the ophiolitic associations inmany areas, have shown that ophiolites form in a variety oftectonic environments, involving different types of oceaniccrust and underlying mantle. Dilek (2003) has developed aclassification of ophiolites based more on the tectonic rela-tions, and this classification has been adopted for this review.



0 km

0.5

1

1.5

2

2.5

Pillowedbasalts(MorelFm.,BABB)

Mantletectonite(mainlyharzburgite)

Plagioclaseperidotite

Sillsofgabbro

Gabbroandpegmatiticgabbro

dikes

Chromititebodiessurrounded

byduniteenvelope

Layeredgabbros(gabbros,troctolites

andnorite)Moho

MOA BARACOA OPHIOLITE

FIGURE 2 An idealized lithostratigraphic column of Moa-Baracoaophiolitic massif, which is a representative ophiolite from the north-ern Caribbean. BABB: back arc basin basalts (Modified from Proenzaet al., 2003b; Marchesi et al., in press).

Based on the field occurrence, the ultramafic rocks ofthe alpine-type in the Caribbean occupy two importantassociations that can be matched with similar occurrencesin other parts of the world. The first is the occurrence of atectonised melange of which the essential components arehigh pressure metamorphic rocks characteristic of sub-duction zones. These associations form a significant com-ponent of the accreting orogenic belts of the Pacific rimand are the San Francisco ophiolites of Dilek (2003). Inthe Caribbean, examples of this ophiolite type occur inthe Motagua Fault Zone, Guatemala, in central Cuba (Vil-la Clara), the northern Hispaniola coast belt and in theCordillera de la Costa in Venezuela (Table 1).

The second important type are the more complete, andsomewhat less dismembered ophiolites, exposed as sub-horizontal sheets, such as those in eastern Cuba. Themantle tectonites are mainly harzburgite and dunite withsignicant bodies of chromitite in the upper part of themantle sequence. High pressure xenoliths or tectonicblocks are absent. These occurrences can be equated withthe Mediterranean-type ophiolite of Dilek (2003).

In the Caribbean, the term ophiolite has commonlybeen applied, in a general way, to include associations ofmafic rocks that often do not consist of any ultramaficrocks (or only very little). These associations are oftenconsidered to be part of the 89 Ma Caribbean LargeIgneous Plateau province (CLIP, Hauff et al., 2000a; Kerret al., 1997b). Examples are the eastern margin of CentralAmerica in Costa Rica, Curacao, western Colombia andthe southern Peninsula of Hispaniola in Haiti. Dilek(2003) has considered this type of occurrence to be a par-ticular type of ophiolite association, which he has termedthe Caribbean-type ophiolite.

Because of the similarity of ophiolite components withthose of the ocean floor, it has long been considered thatophiolites originated at mid-ocean ridges. However, it isnow considered more likely that many ophiolites wereformed at convergent plate boundaries as part of the islandarc structure (see discussion in Shervais, 2001; Robertson,2002; Dilek 2003; Pearce, 2003). The term supra-subduc-tion zone (SSZ) ophiolites is often used for these ophiolitesand they may originate, within the fore-arc, the arc proper,or back-arc spreading area. Based on the composition ofthe associated basalts and gabbros, and the mineral compo-sitions of both the peridotites and associated chromitedeposits, at least some of the Caribbean ophiolites haveinterpreted as the SSZ type (Beccaluva et al., 1996; Kerr etal., 1999; Proenza et al., 1999a, b, c; Giunta et al., 2002a;Marchesi et al., 2003, 2006; Garca-Casco et al., 2003).

The composition of the mafic rocks in the ophiolitecomplexes, including the CLIP complexes, has been fairly

extensively studied, in recent years, in order to determinetheir petro-tectonic affinity. Beccaluva et al. (1995) con-cluded that mid-ocean ridge (MOR) magmatism is themost widespread represented in the ophiolitic units alongthe three margins where these occur. They concluded thatthe oceanic crust was initially generated from multiplespreading centers (LREE depleted MORB compositionsin Venezuela, Costa Rica, Guatemala and Hispaniola) andthickened oceanic plateau-related basalts (flat REEMORB), often associated with picrites in Costa Rica, theSouthern Peninsula of Haiti, and the Dutch and Venezue-lan islands along the southern margin. Island arc magma-tism (island arc tholeiites, IAT) have been recognized inseveral ophiolite complexes in Guatemala, Cuba andVenezuela (Kerr et al., 1999; Giunta et al., 2002a, b;Marchesi et al., 2005). Detailed studies recently made ofthe chemistry of basaltic (extrusive) rocks from CostaRica, Colombia and Curaao as well as ODSP sites fromthe Caribbean basins (Hauff et al., 2000a) show that theserocks have a strikingly uniform major and trace elementand isotopic composition. All these rocks are consideredto belong to the Caribbean Plateau or CLIP and have acommon mantle-plume source (Sinton et al., 1998).Reliable 40Ar/39Ar ages suggest that the main pulse ofmagmatism forming the CLIP occurred mainly between92 and 88Ma, but continued to ~74 Ma (Sinton et al.,1998). Some geophysical evidence suggests the plateaucould be as old as 115 Ma (Mauffret and Leroy, 1997)and has been built in several phases.

It should be pointed out that a few of the ultramafic bo-dies in the Caribbean are not ophiolite-related, have origi-nated in a different way, and belong to a different associa-tion. These are peridotites formed by crystallization in acrustal magma chamber and usually exhibit layering as aresult of crystal accumulation processes. Those in which themain rock type is websterite and that exhibit zoned struc-tures, are termed zoned ultramafic plutons of the Alaskan-type (e.g. ultramafic rocks of Tobago; Snoke et al., 2001;Scott et al., 1999). The emplacement of these peridotitesinto the upper crust does not present a difficult problem.

Serpentinized peridotites of mantle origin, along withmafic rocks, have commonly been dredged from subma-rine scarps on the floor of the Caribbean Sea. Interestingoccurrences include the plagioclase-bearing peridotitesfrom the Cayman Trough spreading center (Perfit andHeezen, 1978; Dick and Bullen, 1984) and from the Puer-to Rico Trench (Bowin et al., 1966; Perfit et al., 1980).These occurrences indicate that mantle peridotites lieimmediately below the oceanic crustal rocks that formmost of the Caribbean basin. On the reasonable assump-tion that the allocthonous ophiolitic bodies of Mesozoicage now found around the Caribbean margins originatedas oceanic lithosphere from the area that is now the



Caribbean basin it is important to examine its composi-tion which can be deduced from geophysical studies,dredging and drilling. As reviewed by Mauffret and Leroy(1997) the Caribbean basin is composed of several thickvolcanic plateaus (e.g. Beata Ridge) separated by deepbasins with thin crust (e.g. Venezuela basin composed ofoceanic crust < 5 km thick).

The eastern lower Nicaragua Rise, Beata Ridge andColombia basin are composed of thickened crust (15-20km). The upper part of layer 2V is probably mainly basaltsills interlayered with sediment that overlies about 3 kmof oceanic crust. The upper part of Layer 3V consists ofabout 6 km of gabbro that outcrops along scarps on theBeata ridge. The lower part of layer 3V is interpreted tobe composed of 10 km of picrite and mafic cumulatesrather than a type of modified mantle. The rock types,which are interpreted as cumulates, not alpine-type peri-dotites, can be matched with the picrite types on Curaaoand Gorgona Islands. Mauffret and Leroy (1997) considerthat these crustal features correspond to the Kerguelenand Ontong Java plateaus (Mahoney and Coffin, 1997).Determination of the nature and composition of Layer 3Vis important in interpreting the obducted ultramafic rocksaround the Caribbean margin and determining how theplateau was formed. Mauffret and Leroy (1997) suggestedthat the thickening of the Caribbean crust is due largely tounderplating.

The question of emplacement of the ophiolitic rocksin the Caribbean, into the upper part of the crust, is a con-tinuing problem, as in other parts of the world. Theemplacement of the Franciscan-type ophiolites is prob-ably the easiest to comprehend. During subduction, faultbounded slices of ocean crust and some of the underlyingmantle, are removed from the descending oceanic lithos-phere and are incorporated into the overlying subductioncomplex that develops at the interface of the descendingand over-riding plates, forming serpentinite matrixmlanges (Gerya et al., 2002). The subduction complexesmay be developed to considerable depth (at least 30 km),and consequently the part of the ocean crust associatedwith the peridotites will recrystallize to form high pres-sure metamorphic rocks such as blueschists and eclogites.Subduction-related, high pressure/low temperature blocks(mainly garnet-bearing amphibolite, eclogites andblueschists) are common in the Caribbeans Franciscan-type ophiolite associations (Perfit et al., 1982; Draperand Nagle, 1991; Garca-Casco et al., 2002, this volume).

Although not yet well investigated, the blocks in themlanges are probably mainly mid-ocean ridge basalts orgabbro (MORB), from the descending plate. However, inthe mlanges of the Rio San Juan complex, in northernHispaniola, blocks of metabasalt of both MORB and calc-

alkaline (?SSZ) affinities are present (Anam, 1994) inaddition to rarer blocks of metagranitoids.

The ultimate emplacement or exhumation of thesehigh-pressure Franciscan-type ophiolites occurs when acontinent-facing arc collides, either directly or obliquely,with a continental margin (Moores, 1982, 1998) or otherbuoyant crustal mass, island arc, or even an oceanicplateau. Arc-continent collisions of this type seem to beresponsible for the final emplacement of the Franciscan-type ophiolites in both the northern and southernCaribbean (Pindell and Barrett, 1990; Pindell et al., thisvolume; Garca Casco et al., this volume).

The emplacement of coherent Mediterranian-type(SSZ or MORB) ophiolites is more problematic. The low-er boundary of such ophiolites is thrust faults and likeFranciscan ophiolites their emplacement must be relat-ed to some crustal shortening event. This shortening canbe related to collision (as in eastern Cuba), but also toother tectonic events, such as subduction polarity reversal(as in central Hispaniola; Draper et al., 1996). The geo-metry of the situation, however, has to be one where theophiolite does not enter a subduction zone, but is thrustonto arc or continental crust at shallow levels.

As the Caribbean has been tectonically active sincethe early Cretaceous (the last 140 million years), therehave been several events at different places, and at differ-ent times, that have resulted in the emplacement of alpineperidotites and associated ophiolitic rocks of either theFranciscan or Mediterranean type.

The earliest orogenic event involving ophiolitic rocksseems to have been in mid-Cretaceous times (about 110-120 million years ago) when ophiolitic bodies wereemplaced onto the arc crust of Central Hispaniola andwestern Puerto Rico, perhaps associated with subductionpolarity reversal (Mattson, 1973; Mattson and Pessagno,1979; Draper et al., 1996). In Cuba, also, the earliestevent of ophiolite emplacement may have been Aptian-Albian (Kerr et al., 1999; Garca-Casco et al., 2002),although the details are unclear.

The oceanic plateau that occupies the present centralregion of the Caribbean sea was formed in the Late Creta-ceous about 89 million years ago (Kerr et al., 1997a, b;Sinton et al., 1998) or earlier, somewhere to the west ofits present position with respect to the middle Americaregion (Pindell and Barrett, 1990). The eastern, leadingedge of this plateau was marked by an island arc systemthat allowed convergence between the plateau and theAmerican continents. As the Americas moved westwardin the Late Cretaceous, parts of the plateau were subduct-ed, but other fragments, along with various Pacific ter-





OPHIOLITE UNIT NUMBER IN ULTRAMAFIC CHROMITITES TECTONIC PERIOD OF Ni LATERITE SELECTEDFIGURE 1 ROCKS SETTING EMPLACEMENT DEPOSIT REFERENCES

GUATEMALA Sierra de 1 Harzburgite SSZ Late Cretaceous- Yes Beccaluva et al.Santa Cruz (1995), Giunta et al.

(2002b, c), Valls (2003, 2004)

Baja Verapaz Unit 2 Harzburgite Yes SSZ Late Cretaceous- Beccaluva et al.Paleogene (1995), Wadge et al.

(1982), Giuntaet al. (2002b, c)

Juan de Paz 3 Harzburgite SSZ Late Cretaceous- Beccaluva et al.Paleogene (1995), Wadge et al.

(1982), Giunta et al.(2002b, c)

El Tambor Group 4 Peridotite MOR (proto- Late Cretaceous- Beccaluva et al. South Caribbean) Paleogene (1995), Wadge et al. Motagua (1982), Giunta et al. North (2002a, b) Motagua

CUBA Cajlbana 5 Harzburgite, Yes (Cr-rich) SSZ Paleocene-Early Yes Fonseca et al. (1985),dunite, lherzolite, Middle Eocene Murashko and pyroxenite (earliest event Lavandero (1989),

may be Late Iturralde-VinentCampanian?) (1996, 1998),

Cobiella-Reguera(2002, 2005)

Habana-Matanza 6 Harzburgite, Yes SSZ and Paleocene-Early Fonseca et al.Plagioclase- (Al- and Cr-rich) oceanic Middle Eocene (1985), Iturralde-bearing perido- plateau? Vinent (1996),tites, dunite Llanes et al. (2001)

Villa Clara 7 Harzburgite Yes SSZ Late Paleocene- Iturralde-Vinent and MOR Middle Eocene (1996, 1998),

(earliest event Garca-Casco et al.may be Aptian- (2002), Cobiella-Albian?) Reguera (2002,2005)

Escambray 8 Serpentinite MOR Late Cretaceous Somin et al. (1992),(Faralln Milln-Trujillo,oceanic (1996), Auzende et al.crust) (2002), Scheneider

et al. (2004), Garca-Casco et al. (this volume)

Camagey 9 Harzburgite, Yes (Al-rich) SSZ Paleocene-Upper Yes Fonseca et al. (1985)dunite, (back arc) Eocene Iturralde-Vinent (1996websterite 1998, 2001) Cobiella-websterite Reguera (2002, 2005)

Holgun 10 Harzburgite, Yes SSZ Maastrichtian Kozary (1968),dunite (fore arc) Early Middle Iturralde-Vinent

and MOR Eocene (earliest (1996, 1998), Andevent maybe et al. (1996),Aptian-Albian?) Garca-Casco et al.

(2002), Cobiella-Reguera (2002, 2005)

Mayar-Cristal 11 Harzburgite Yes SSZ Maastrichtian- Yes Iturralde-Vinent,(Cr- and Al-rich) (fore arc) Danian? (1996), Proenza et

al. (1999a, b, 2003),Marchesi et al. (2003,in press), Cobiella-Reguera (2002,2005)

Alto de La Corea 12 Serpentinite MOR? ? Somin et al. (1992),Milln-Trujillo,(1996), Garca-Casco et al.(this volume)

TABLE 1 Main characteristics of ophiolite-related ultramafic rocks in Caribbean region: SSZ = Supra subduction zone ophiolite; MOR = Mid-oceanic ridge.



OPHIOLITE UNIT NUMBER IN ULTRAMAFIC CHROMITITES TECTONIC PERIOD OF Ni LATERITE SELECTEDFIGURE 1 ROCKS SETTING EMPLACEMENT DEPOSIT REFERENCES

Moa-Baracoa 13 Harzburgite Yes (Al-rich) SSZ Maastrichtian- Yes Iturralde-Vinent, (1996)(back arc) Danian? Proenza et al. (1999a,

b, 2003), Marchesi etal. (2003, in press)Cobiella-Reguera (2002, 2005)

Sierra 14 Serpentinite MOR Late Cretaceous Somin et al. (1992),del Convento Milln-Trujillo,

(1996), Garca-Casco et al.(this volume)

JAMAICA Arntully 15 Dunite, MOR Maestrichtian- Wadge et al. (1982),harzburgite, Paleocene Scott et al. (1999),lherzolite Abbott et al. (1999)

HISPANIOLA North Coast Belt 16 Harzburgite, MOR Paleocene Draper and Nagledunite and SSZ? (1991), Anam (1994)

Loma Caribe 17 Harzburgite, Yes (Cr-rich) SSZ and/ Late Albian Yes Lewis (1981), Lewislherzolite, or mantle and Draper (1990),dunite plume Draper et al. (1996),

Lewis et al. (2003)

PUERTO RICO Monte del Estado 18 Lherzolite, SSZ? ? Evans et al (1997),harzburgite, Jolly et al. (1998a, b)dunite

Ro Guanajibo 19 Lherzolite, MOR? ? Jolly et al. (1998b)harzburgite,dunite

Bermeja 20 Serpentinite MOR? Albian? Mattson (1974),Mattson andPessagno (1979),Jolly et al. (1998a, b)

VENEZUELA La Franja Costera 21 Peridotites Yes MOR? Late Cretaceous?. Donovan (1994),Final thrust em- Giunta et al. (2002a,placement during c), Ave LallementEocene-Oligocene (1997)

Loma de Hiero 22 Peridotites, MOR Late Cretaceous?. Yes Donovan (1994),plagioclase- (proto- Final thrust Giunta et al. (2002a,bearing dunite Caribean) emplacement c), Ave Lallement

during Eocene- (1997)Oligocen

Villa de Cura 23 Harzburgites SSZ Late Cretaceous. Donovan (1994),Final thrust empla- Giunta et al. (2002a,cement during c), Ave Lallement Eocene-Oligocene (1997)

Guajira Peninsula 24 Serpentinte MOR (proto- ? Case and Macdonald(serpentinites of Caribbean) (1973), DonovanCabo de la Vela) (1994), Seplveda et

al. (2003)

COLOMBIA Cordillera Central 25 Dunita, Yes (Al-rich) SSZ Permic-Triasic lvarez (1987),Dunita de Medelln harzburgita (back arc) Restrepo (2003),(Aburr ophiolite) Correa and Nilson

(2003)

COSTA RICA Santa Elena 26 Harzburgite, Yes MOR in a Late Tertiary Frisch et al. (1992),clinopyroxene- mantle Beccaluva et alpoor lherzolite, plume (1999dunite region and Gazel et al.

/or SSZ (this volume)

COSTA RICA - Ro San Juan 27 Serpentinite, MOR? ? Astorga (1992),NICARAGUA lherzolite Tournon et al. (1995)

NICARAGUA Siuna 28 Serpentinized, MOR? ? Baumgartnerperidotite et al. (2004)

TABLE 1 Continued.

ranes were involved in collisions with the western parts ofthe North and South American continents (Montgomeryet al., 1994; Kerr et al., 2003).

In the southern Caribbean boundary, oblique conver-gence also resulted in the final thrust emplacement ofperidotites with Cretaceous high-pressure associations,but in this case the emplacement may have occurred pro-gressively from Eocene times (about 50-55 million yearsago) and was not completed until Oligocene or even earlyMiocene time (15-30 million years ago; Stockhert et al.,1995; Ave-Lallement, 1997).

DISTRIBUTION OF OPHIOLITE-RELATED ULTRAMAFIC ROCKS IN THE CARIBBEAN

The distribution of ophiolite-related ultramafic rocksaround the margins of the Caribbean Plate is shown inFig. 1. The more significant features concerning theserocks are included in Table 1.

Northern Caribbean margin

Guatemala

Ophiolitic rocks exposed along the Motagua FaultZone in Guatemala include some large bodies of harzbur-gite such as the Sierra de Santa Cruz and La Gloria mas-sifs (Donnelly et al., 1990). Recently Beccaluva et al.(1995) and Giunta et al. (2002b, c) have recognized fourmain ophiolite units containing peridotites (Table 1): (1)Sierra de Santa Cruz, (2) Baja Verapaz, (3) Juan de Paz,and (4) El Tambor group (South Motagua and NorthMotagua).

Sierra de Santa Cruz, Baja Verapaz and Juan de Pazunits mainly consist of serpentinized mantle harzbur-gites, and have been interpreted as SSZ-type ophiolites.In contrast, South Motagua and North Motagua units,which also include serpentinized mantle peridotites,represent remnants of a proto-Caribbean oceanic crust(Beccaluva et al., 1995; Giunta et al., 2002b, c).

It is generally considered that most of these rocks,including the Santa Cruz body, were obducted (over-thrust) onto the Maya continental block to the northwhereas the South Motagua serpentinites were thrust tothe south over the Chortis block. Uplift of these tecto-nic blocks resulted from sinistral transpressive move-ment between the Maya and Chortis blocks. Theharzburgites contain inclusions of rocks such as jadeiteand glaucophane-bearing schists and eclogites, whichwere formed under high pressure conditions at depthsof about 30 kilometers. The harzburgites were probably

first exposed to weathering in the Late Eocene at thetime of the strike-slip faulting. The larger massifs ofSierra de Santa Cruz have well-developed Ni-lateritesoils with a potential for mining (Giunta et al., 2002b;Valls, 2003, 2004).

Cuba

In Cuba, the largest bodies of ophiolite-related ultra-mafic-rocks (serpentinized peridotites and serpentinites)crop out to the north of the island, along the so-callednorthern ophiolite belt (Iturralde-Vinent, 1994, 1996,1998). The ophiolites of northern Cuba occur as sevenseparate massifs (ophiolite blocks) or suites exposedalong the entire length of the island, from west to east:Cajlbana, Habana-Matanza, Villa Clara, Camagey, Hol-gun, Mayar-Cristal, Moa-Baracoa. All ophiolite massifscontain massive chromitite bodies, a characteristic featureof SSZ ophiolites. The larger serpentinized peridotitemassifs that crop out with these ophiolites are identifiedin Table 1 and Fig. 1.

The available data (Iturralde-Vinent, 1996; 1998; Kerret al., 1999; Proenza et al., 1999a, 2001; Cobiella-Reguera, 2002, 2005; Garca-Casco et al., 2002, 2003)suggests that at least two Jurassic-Cretaceous mantle sec-tions exist within the northern ophiolite belt, whichunderwent two major events: mid-oceanic accretion, andsubduction. The northern belt includes both MOR andSSZ-type peridotites. In addition, ultramafic rocks occuras tectonic slabs of serpentinite (commonly associatedwith high-pressure rocks) embedded in a metasedimenta-ry and/or metabasite matrix in the terranes of Guaniguani-co and Escambray, and in the metamorphic complexes ofSiera del Convento and La Corea (Somin et al., 1992; Itu-ralde-Vinent, 1994, 1996; Milln-Trujillo, 1996; Auzendeet al., 2002; Schneider et al., 2004; Garcia-Casco et al.,this volume).

Ultramafic rocks in western Cuba (the Bahia Hondaultramafic complex of Fonseca et al., 1985; Fig. 1, Table1) are strongly dismembered. Except for the Cajlbanamassif, the serpentinite bodies in western Cuba are allsmall and most have tectonic boundaries against adjacentrock units. In Cajlbana, the ultramafic rocks consist ofmainly harzburgite with subordinate dunite, lherzolite,and pyroxenite (Fonseca et al., 1985; Murashko andLavandero, 1989). These serpentinized peridotites havedeveloped Ni-laterite soils (Sgalen et al., 1980). Accord-ing to Garca-Casco et al. (2003), the protoliths of themetabasites from Cajlbana have an island-arc tholeiiticsignature (SSZ ophiolite). These authors suggested thatCajlbana ophiolites represent a fragment of the Proto-Caribbean plate incorporated to the Caribbean plate dur-ing Aptian-Albian time. Also, in western Cuba, in the



Cordillera de Guaniguanico, blocks of serpentinite withassociated eclogitic rocks, occur as olistoliths in the Low-er Eocene olistostrome of Manacas Fm (Milln-Trujillo,1996 and references therein).

In the Habana-Matanza area (Fig. 1), the ultramaficrocks consist of serpentinized harzburgites, minor plagio-clase-bearing peridotites, dunite and lherzolite (Fonsecaet al., 1985; Llanes et al., 2001). Volcanic rocks, tectoni-cally embedded in deformed serpentinites from theHabana area, have been described as boninites (Fonsecaet al., 1989; Kerr et al., 1999). In contrast, basalts fromthe volcanic-sedimentary sequence of the Margot Fm,that also occur within serpentinite in the Matanza area,have been shown to have an oceanic intraplate affinity(Kerr et al., 1999).

Large bodies of serpentinite and serpentinized peri-dotites, forming a serpentinized harzburgite mlange(Fonseca et al., 1985), characterize the ophiolitic rocks inthe Villa Clara massif (Fig. 1). The ophiolite mlange, asa mass lies over the Bahamas continental margin, anditself is covered tectonically by Cretaceous arc rocks(Iturralde-Vinent, 1996). High-pressure metamorphicrocks are found included within the serpentinite mlange(Milln-Trujillo, 1996; Garcia-Casco et al., 2002). Meta-basites from Villa Clara region (Iguar-Perea Complex)have a calc-alkaline magmatic signature, and have beeninterpreted as formed in a suprasubduction environmentduring the Late Cretaceous (Garca-Casco et al., 2003).

Small bodies of serpentinite occur as tectonic sliceswithin the Escambray metamorphic complex to the southof Villa Clara, particularly along its eastern margin (Fig.1; Somin et al., 1992; Milln-Trujillo, 1996; Auzende etal., 2002; Schneider et al., 2004). Serpentinites form len-ses associated with eclogitic metabasites, and are embed-ded in a metasedimentary matrix. The association of high-pressure rocks in this complex indicates a relatively deeporigin for the rocks of the complex including the serpen-tinized peridotites. According to Schneider et al. (2004),eclogite occurring as exotic blocks within the Escambrayserpentinites, originated from a N-MORB type protolith(Farallon oceanic crust). Emplacement of the Escambraycomplex and the relationship with the ophiolite rocks ofthe Villa Clara area to the north is one of the most contro-versial issues of Cuban geology (Iturralde-Vinent, 1994;Milln-Trujillo, 1996; Grafe et al., 2001; Schneider et al.,2004).

The Camagey massif (Fig. 1) mainly consists of ser-pentinized harzburgite, dunite, websterite and lherzolite(Fonseca et al., 1985; Iturralde-Vinent, 1989, 1996,2001). Laterite soils are well-developed and preservedover the San Felipe serpentinite meseta, an erosional rem-

nant near Camagey city, but in the areas of lower reliefthe soils over the serpentinite have been washed awayinto adjacent sedimentary basins. For example, the Maga-ntilla basin contains deposits of sedimentary magnesiteand conglomerates with serpentinite pebbles interbeddedwith shales, often reddish in color. This implies that theserpentinites in the Camagey area have been continu-ously uplifted to different levels and exposed to weathe-ring and erosion since the Miocene (Iturralde-Vinent,pers. com., 2001). According to Gleeson et al. (2003)the nickel resource of Meseta de San Felipe is 2-3 Mtwith 1.3% Ni.

In the Holgun massif (Fig. 1), the ultramafic rocksform part of a serpentinite mlange. The tectonic mixingand shearing of serpentinite in the Holgun block is high-er than in other areas of Cuba (Fonseca et al., 1985). Peri-dotites consist of mainly harzburgite, with minor dunite(And et al., 1996).

The largest area of exposed serpentinized peridotitesin Cuba, and in the whole Caribbean, occurs in northeastCuba in the Mayar-Cristal and Moa-Baracoa massifs(Fig. 1; Mayar-Baracoa Ophiolite Belt: MBOB). Recentfield and structural data indicate that the western part ofMayar-Cristal massif (Mayar zone) is essentially madeup of mantle tectonites (>5 km thick) with subordinategabbro-diabase dikes (Proenza et al., 2003b; Marchesi etal., 2003, in press). Two main domains have been differ-entiated in this massif: (1) the lower domain, which iscomposed mainly of porphyroclastic harzburgite tec-tonites contains pyroxenite veins. Dunites occur as irregu-lar patches and sub-concordand dunite layers (

On the other hand, the Moa-Baracoa massif is com-prised of mantle tectonites (>2.2 km thick) topped by athin crustal section made up of lower gabbros (ca. 300 mthick) and discordant basaltic rocks of the Morel Fm withpossible back arc basin affinity (Iturralde-Vinent, this vol-ume; Proenza et al., this volume). Isotropic gabbros and adiabase sheeted dike complex are lacking. In this massif,the Moho Transition Zone (MTZ) is made up of residualharzburgites, minor dunites, and impregnated peri-dotites with plagioclase and clinopyroxene. Concordantand sub-concordant bodies of dunite, sills of gabbro andchromitite bodies with a dunite envelope also occur inthis transition zone. The MTZ peridotites are cut bydikes of gabbro, pegmatite gabbro, olivine norite andpyroxenite (Proenza et al., 1999a, b, 2003b; Marchesi etal., 2003, in press).

The MBOB is a SSZ-type ophiolite, that has beeninterpreted to result from an initial stage of rifting andformation of a Protocaribbean oceanic lithospherefollowed by the development of a juvenile island-arc ina ?Late Cretaceous SSZ setting and later a back-arcbasin (Proenza et al., 1999b, 2001; Marchesi et al.,2003, in press).

The structural/stratigraphic features of the Mayar-Baracoa ophiolite allocthon are probably the best docu-mented of any in the Caribbean. Ideas as to the mode ofemplacement of the ophiolite (Cobiella, 1978) have recent-ly been summarized by Iturralde-Vinent (2003) andCobiella-Reguera (2005). As in western Cuba the source ofthe oceanic crust (ophiolite) was from the south but thetiming and mechanism of emplacement was apparently dif-ferent in eastern Cuba. In both areas there was apparentlyan early thrust event at the end of the Cretaceous. In cen-tral Cuba an element of oceanic crust and mantle wasemplaced directly over the Bahamian continental border-land and later over segments of the volcanic arc rocks.

In eastern Cuba, the earliest emplacement of oceaniccrust over the continental crust is represented by theGuira de Jauco amphibolites. Cretaceous volcano-sedi-mentary rocks of the Santo Domingo Fm and metavol-canic rocks of the Purial were emplaced over the amphi-bolites and are interpreted to directly overlie theBahamian continental platform to the north. The emplace-ment of the ophiolite was from the south and took placein the Maastrichtian to early Danian with the olis-tostromes advancing in front of the ophiolite mass andfilling the Sagua de Tnamo basin with clastic rocks.

The volume of nickel laterite developed over theMayar-Cristal and Moa-Baracoa massifs makes this areathe largest known resource of nickel laterite in the world.The Moa-Baracoa laterites were developed over a broad

erosion surface, or surfaces, that today vary in heightfrom about 550 m to 1,000 m above sea level. In the low-er elevations the serpentinies are covered by sedimentaryrocks of Miocene age indicating that the ultramafic rockswere exposed at this time.

The smaller serpentinite massif of the Sierra del Con-vento, located in the southwest part of the Sierra del Pur-ial, is separate from the Moa-Baracoa massif to the north(Fig. 1). A characteristic feature of the Sierra del Conven-to serpentinite is that it contains an abundant variety ofblocks of high-pressure metamorphic rocks including gar-net amphibolites, blueschists, metatrondhjemites and peg-matites suggesting a different origin, probably an upliftedsubduction zone (Somin et al., 1992; Milln-Trujillo,1996; Garca-Casco et al., this volume.).

Jamaica

A small faulted block of serpentinized peridotite(~0.25 km2 in area) known as the Arntully sepentiniteis located in the southern edge of the Blue Mountainsalong the Plantain Garden Fault in eastern Jamaica(Wadge et al., 1982; Robinson, 1994; Scott et al., 1992,1999; Abbott et al., 1999). This body is considered tobe related to the Bath Dunrobin complex of basalts andgabbros which would make this an ophiolite associa-tion (Wadge et al., 1982, 1984). The Arntully body iscomposed mainly of serpentinite with a core of partlyserpentinized harzburgite, lherzolite, dunite and abun-dant blocks of rodingite.

Hispaniola

Serpentinized peridotites of apparent ophiolitic affini-ty occur in two belts across Hispaniola, namely the NorthCoast belt and the Median Belt in the Cordillera Central(Bowin, 1975; Lewis, 1981; Lewis and Draper, 1990). Inboth belts there is evidence that these originated separate-ly as parts of ophiolite associations that are now highlydismembered and most of the original features have beendestroyed. In the North Coast belt, the serpentinites in theRio San Juan complex and the two small bodies in theSamana Peninsula are associated with high pressure/lowtemperature metamorphic rocks (Nagle, 1974; Draper andNagle, 1991; Joyce, 1991; Abbott et al., 2004). The RioSan Juan complex is interpreted as an uplifted subductionzone complex. The north coast serpentinites are small(only the Punta Gorda body exceeds 5 km across) andthere is little soil developed on these.

The main serpentinite belt, the Loma Caribe peri-dotite, in central Hispaniola in the Median Belt (Bowin,1966) is about 4-5 km wide and extends for 95 km fromLa Vega to Cerro Prieta north of Santo Domingo, but the



southeastern part of the peridotite is exposed as thin faultslices only. The body extends at depth to the coast whereit has been intersected by drilling east of Santo Domingo,Most of the body is serpentinized harzburgite, but dunites,lherzolites and pyroxenites also occur (Lewis et al.,2003). The relationship of the associated basaltic rockswith the peridotite is not clear since all the contacts arefaulted, leading to considerable controvery. Draper et al.(1996) have suggested that the Loma Caribe peridotitewas emplaced first as early as the late Albian. Ni-lateriteis well-developed over the Loma Caribe serpentinizedharzburgites. Nickel resources are estimated at 1-2Mt at agrade of 1.2%Ni (Falconbridge Dominicana annualreports).

In addition to the two belts of the ophiolitic associa-tion described above, intrusive (magmatic) bodies of ser-pentinized peridotites also occur in several places alongthe Cordillera Central of the Dominican Republic andMassif du Nord of Haiti (Lewis, 1980; Lewis and Draper,1990). These intrusions are composed of massive and lay-ered peridotites and pyroxenites with norites and gabbrosof cumulate origin. These ultramafic-mafic complexes areclosely associated with the large tonalite intrusive bodiesof late Cretaceous age and the metamorphosed maficrocks of the Duarte Complex (Draper and Lewis, 1989).

Puerto Rico

In southwestern Puerto Rico there are three rela-tively small serpentinized peridotite bodies, namelyMonte del Estado, Rio Guanajibo and the Bermeja inthe south. Tectonic relations are not clear but theBermeja serpentinized peridotite appears to be relatedto an ophiolite association (Mattson and Pessagno,1979). Detailed studies show that there is considerablevariation in the chemistry and petrography of thesebodies (Jolly et al., 1998b). Dunites, harzburgites andlherzolites are all present even over a small area.Whether or not they are all part of the one peridotitemass that has been dismembered is not understood(Schellekens et al., 1989; Schellekens, 1998).

These serpentinized peridotites have features similarto abyssal peridotites (mid-oceanic accretion setting) andwere the subject of extensive investigations by Hess andothers (Burk, 1964). The Bermeja peridotites and associ-ated amphibolites and metabasalts represent pre-arcoceanic crust and Matson and Pessagno (1979) proposedthat the peridotites and crustal rocks were emplaced as anappe. Jolly and Lidiak (pers. com., 2002) suggest all theperidotite masses appear to be emplaced into the uppercrust as diapiric bodies. Jolly et al. (1998a) concludedthat the Monte del Estado peridotite was in geochemicalequilibrium with sub-arc magma.

Although there is a signicant saprolite zone developedover the Puerto Rico serpentinites, it is unlikely to bebecome an economic resource for nicked.

Southern Caribbean margin and northernColombian Andes

Relatively small bodies of serpentinized ultramaficrocks are found along the Caribbean mountain system onthe northern coastal region of South America (Fig. 1,Table 1). This mountain system extends discontinuouslyfrom the Sierra Nevada de Santa Marta and GuajiraPeninsula in Colombia, eastward to the islands of Mar-garita and Tobago and the Northern Range of Trinidad.Small bodies of serpentinites that occur in the SierraNevada de Santa Marta are associated with mafic gneissesand schists of late Paleozoic age (Bellizia and Dengo,1990). In the Guajira Peninsula, the serpentinites areassociated with metagabbros, mafic and pelitic schists,phyllites, metavolcanics and marbles of late Mesozoic toearly Tertiary age (Seplveda et al., 2003). Some of theseserpentinites were considered to be of detrital origin(Lockwood, 1971). In addition, peridotites and associatedgabbros occur in the Paraguan Peninsula, east of Mara-caibo (Martin-Bellizia and Arozena, 1972).

The ultramafic-mafic complexes in the Caribbeanmountains in cental North Venezuela occur in at least fivestructural units, or belts, separated by major thrusts.These represent the imbrication of both continental andoceanic crustal units along with slices of mantle rock(Maresch, 1974; Stephan et al., 1980; Av-Lallement,1997; Av-Lallement and Sisson, 2005). The main occur-rences of serpentinized peridotites and associated rocksare shown in Fig. 1 and Table 1. Recently Giunta et al.(2002a, c; Table 1) have examined the chemistry of themajor complexes and summarized the tectonomagmaticsetting of the main ophiolite units. Relationships amongrock types are complex and it is apparent that not all ofthe occurrences are ophiolites.

For example, detailed studies of the Tinaquillo spinellherzolite, located in the Caucagua-El Tinaco nappe struc-ture in the western part of the Serrania del Interior, haveled to the interpretation that it is a piece of continentallithospheric mantle emplaced at the base of, or within,high grade felsic gneisses during the late Cretaceous(Seyler and Mattson 1989, 1993; Seyler et al., 1998). It isone of the few examples of the so-called orogenic lher-zolite massifs (Menzies and Dupuy, 1991). Ostos et al.,2005 have interpreted the Tinaquillo peridotite complexto be a fragment of Jurassic rift zone.

Small bodies of serpentinized peridotite along theCordillera de la Costa occur within the ophiolite mlange



of the Franja Costera Unit (Fig. 1, Table 1). In this unit,basalt and gabbros show geochemical characteristics withMORB affinity (Giunta et al., 2002a, c).

Serpentinite is found within the complex of high-pres-sure metamorphic rocks of the island of Margarita northof Venezuela (Stockhert et al., 1995). Giunta et al.(2002a) have suggested that these peridotites bodies(mainly lherzolites) may represent an ultramafic complexequivalent of Tinaquillo.

Serpentinized peridotites (including plagioclase-bear-ing dunite) along with cumulate gabbros of late Jurassicage occur as part of the Loma de Hierro Unit (Table 1,Fig. 1) within the Serrania del Interior. This is describedas an ophiolite association consisting of mainly basalticrocks and siliceous-carbonate sequences in addition to thegabbros and serpentinized peridotites. The Loma HierroUnit has been interpreted as a MOR type ophiolite (Giun-ta et al., 2002a, c). Ni-laterite has been mined from thisarea since 2000 but there is no published information onthis deposit.

Serpentinized mantle peridotites, interpreted as part ofsubarc mantle and subduction complex, occur in the VillaCura Group (Giunta et al. 2002a; Fig. 1, Table 1).Although a recent detailed study of the geochemistry ofthe Villa Cura Goup showed that all the metavolcanicrocks are arc-related (Stiles et al., 1999; Unger et al.,2005), Kerr et al. (2003) considered that at least some ofthe Villa de Cura volcanic rocks have an oceanic plateauorigin similar to the basalts and picrites of Curacao andAruba.

At least three of the peridotite massifs in the Cari-bbean mountain range do not belong to the ophiolite asso-ciation. The ultramafic rocks on the island of Tobago arein the form of a layered intrusion and are interpreted toform the basal part of an island arc complex (Frost andSnoke, 1989; Scott et al., 1999). According to theseauthors, ultramafic rocks in Tobago are part of a singlefractionated plutonic body made up largely of gabbro anddiorite with minor tonalite. Snoke et al., (2001) haveinterpreted the complex as a zoned ultramafic intrusion ofthe Alaskan-type. Murray (1972) interpreted two of theperidotite intrusions in the Caribbean mountains inVenezuela as zoned ultramafic complexes of the Alaskantype. The largest of these intrusions, about 8 km long by 6km wide, intrudes the southwestern part of the Villa deCura Group.

Mafic and ultramafic rock associations of Cretaceousage form north-south trending belts in Colombia. Thesebelts occur in the Cordillera Central, the main belt, theCordillera Occidental and the Serrana de Baud along

the Pacific Coast discussed below. Many of the ultramaficbodies are thin slices elongated along faults. An exampleis the Itanguo serpentinite which is 40 km long and only4000 m wide. In recent years, there has been much con-troversy on the tectonics and petrogenesis of these mafic-ultramafic associations. Only the northernmost bodiesare discussed here.

lvarez (1983) documented twenty-eight occurrencesof ultramafic-mafic rock bodies in the Cordillera Centraland Cordillera Occidental in the Colombian Andes. Heand others (see Bourgois et al., 1987 review) consideredthat most of these ultramafic-mafic associations in theColombian Andes showed ophiolitic affinities. Based ontheir composition of basaltic lavas around Cali McCourtet al. (1984) and Aspden et al. (1987) advocated a sub-duction related origin for the basalts. Analyses and agedeterminations on a larger number of the basaltic rocksled Kerr et al. (1997a, b) to conclude that the basaltic ter-ranes were formed as an oceanic plateau and are mostprobably part of the mid-Cretaceous Caribbean plateau.Reynaud et al. (1999) have shown that the belt of oceanicplateau basalts, along with Cretaceous arc rocks, extendssouthward into southern Ecuador. The presence andextent of the Cretacous oceanic plateau in the northernAndes of South America was also based on the domi-nance of basaltic rocks, the absence of sheeted dikesequences and the occurrence of dunite rather thanharzburgite. However, harzburgite is part of many of themafic-ultramafic complexes and the presence of podiformchromite in some dunites is typical of ophiolite sequences(e.g. the so-called Dunitas de Medelln; Alvarez, 1987;Correa and Nilson, 2003; Restrepo, 2003) exposed in thenorthernmost part of the Cordillera Central. Most of thebodies are serpentinites and serpentinized dunites, butserpentinized harzburgites also occur. These ultramaficrocks have tectonic contact with orthoamphibolites, andboth units are included in the Aburr Ophiolitic Complex.

Ni-laterite soils are developed on several of the rela-tively small bodies of serpentinized ultramafics locatedin the north of Medellin. The largest of these is the Ce-rro Matoso which is 2 x 1.7 km in area and has a reliefof 300 m. The thickness of the laterite varies from a fewm to 100 m. The Ni content of the saprolite zone ave-rages 2.5%. A study of the mineralogy and geochemistryof the Cerro Matoso deposit has recently been made byGleeson et al. (2004).

Western Central America and northwesternColombia

It is generally considered that the mafic-ultramaficcomplexes exposed along the west coast of Central Amer-ica are part of the western most margin of the Caribbean





Plate. These include the well-studied segments of theNicoya and Santa Elena Peninsulas and Quepos in CostaRica and the Azuero massif in northwest Panama.

There has been considerable debate as to the origin,age and mode of emplacement of the oceanic crustal seg-ment along the Pacific Coast of Costa Rica and Panama(see Hauff et al., 2000b; Hoernle and Hauff, submitted,for recent reviews). In the past the complexes have beentermed ophiolites (Frisch et al., 1992; Beccaluva et al.,1999) but most of the important elements of a typicalophiolite are missing, and peridotite is found only in thenorth in the Santa Elena exposure (Fig. 1, Table 1). Here,the peridotite bodies consist of harzburgite, dunite andclinopyroxene-poor lherzolite (Frisch et al., 1992; Becca-luva et al., 1999). The 124-109Ma tholeiitic portions ofthe Santa Elena complex formed in a primitive island arcsetting believed to be part of the Chortis subduction zone(Hauff et al., 2000b).

Nicoya complex consists mainly of tholeiitic pillowlavas and sheet flows with gabbroic and plagiograniticintrusions, and fault bounded sequences of sedimentaryrocks. These units are overlain unconformably by a Cam-panian to Tertiary sedimentary sequence. The Nicoya,Herradura, Golfito, and Burica complexes have beeninterpreted as part of the Caribbean oceanic plateau andgenerated as a mantle plume (Sinton et al., 1996; Hauff etal., 1997), or at an oceanic spreading center in a mantleplume region, analogous to the present Galapagosridge/hot spot system (Beccaluva et al., 1999). Ar/Ar dat-ing has shown that at least part of the Nicoya Complex isthe age equivalent (89 Ma) of other magmatic plumebasalt centers in the Caribbean-Colombian oceanic basaltplateau (Sinton et al., 1998). However, radiolarian chertssuggest that the lower part of the Nicoya complex mightbe late Jurassic or earlier in age (Galli-Olivier, 1979;Escalante, 1990). Denyer and Baumgartner (2005) con-sider that the Jurassic-Cretaceous cherts are sedimentblocks disrupted from the original basement by multipleinjections of magma. The Quepos segment, composed oftholeiitic pillow lavas overlain by highly vesicular hyalo-clastites, breccias and conglomerates, is interpreted as aseamount (Hauff et al., 1997) or ocean island (Hauff etal., 2000b), whereas the Osa segment is considered anoceanic ridge. The structural, textural, and geochemicalcharacteristics of the mafic rocks of Azuero and SoniaPeninsulas in south Panama are typical of CLIP and con-sistent with an origin from the Galapagos hotspot (Hoern-le and Hauff, submitted).

Peridotites and serpentinites are reported from theSerrana de Baud mafic complex in northwest Colom-bia (Case et al., 1971; Etayo et al., 1983). Little isknown of this complex and that of the Darien area in

southeastern Panama, but from limited studies of themafic rocks both are interpreted as part of theCaribbean-Colombian oceanic basalt plateau (Kerr etal., 1997a; Hoernle and Hauff, submitted).

Small exposures of serpentinized peridotite of appar-ent mantle origin crop out near the Rio San Juan alongthe border between Nicaragua and Costa Rica (Vargasand Alfaro, 1992). These rocks were considered byTournon et al. (1995) to be the remnants of an east-westtrending suture zone. Serpentinized peridotites, maficrocks and radiolarites, which together appear to form partof an ophiolitic association, have been reported from theSiuna area in northeast Nicaragua (Baumgartner et al.,2004). This occurrence is puzzling since it lies within theChortis block composed of essentially Paleozoic rocks.

FORMATION OF NORTHERN CARIBBEAN Ni-LATERITEDEPOSITS: EASTERN CUBA AND CENTRAL DOMI-NICAN REPUBLIC

Once exposed to the surface, the serpentinized peri-dotites and serpentinites readily weather to give lateritesoils in which Ni and other elements are distributed andconcentrated into vertical zones within the soil (Golightly,1979, 1981; Brand et al., 1998; Elias, 2002; Gleeson etal., 2003). Lateritization involves the dissolution of theoriginal minerals in the rock and mobilization and preci-pitation of new chemical species from solution. Thisprocess depends largely on the movement of waterthrough the rock and soil, the composition of the waterand on the mineral solubilities. The solubility of mineralsis known to be affected by lichens and microorganismssuch as bacteria. It is probable that bacteria play a signifi-cant role in the weathering of rock, the formation of lat-erites and the migration of ions through the laterite soil(Barker et al., 1997). Lichens, lichen acids and organicacids, particularly oxalic acid, are also considered to beimportant in the weathering process (Jones, 1988; Drever,1994; Easton, 1994). However the details of theseprocesses are not yet well understood.

With more complex minerals, such as serpentine min-erals, clays, garnierite and chlorite, ion exchange reac-tions occur. It should be mentioned that garnierite is nota recognized specific mineral species by the Commissionon New Minerals and Mineral Names, but is a term formixed structure of hydrous Ni-Mg silicates of low crys-tallinity with affinities to serpentinie, talc and chlorite(Elias, 2002). The most important ion exchange reactionrelevant to the weathering of ultramafic rocks is theexchange of Ni for Mg between soil, water and serpenti-nite. The major overall factor determining the soil profileand element redistribution as a result of weathering is cli-

mate and this depends on the altitude at which the rock isweathering. On a large scale the redistribution of the ele-ments by the soil waters is accomplished by the leachingand downward percolation of the elements in solution.The rate and direction of flow of water through the soiland rock substrate depends on the rainfall, the seasonalvariation and the location of the water table. Locally thesoil profile will be determined by the local drainage con-ditions which in turn will be determined by the degree offracturing of the rock. All of these factors are seen in theNi-laterite profiles in eastern Cuba and central DominicanRepublic.The Caribbean has 7% of the worlds Ni-lateriteresources (distribution by contained nickel) (Dalvi et al.,2004). Most of this resource is in eastern Cuba andDominican Republic (Loma Caribe).

Composition of protolith

The detailed composition of the ultramafic rocks(alpine-type peridotites) making up eastern Cuba and

Dominican Republic will be discussed in a future paperbut some representative examples are given here. Thecompositions of the primary minerals olivine, orthopyrox-ene, clinopyroxene are very similar in the different rocktypes and the significance of this with respect to weather-ing and the composition of the laterites is discussed later.In detail there are small but significant differences incomposition within and among the peridotites.

In all peridotite samples analyzed from Mayar-Baracoa peridotite (eastern Cuba) and Loma Caribeperidotite (Dominican Republic), the olivine in dunitehas similar Ni contents but higher Fo number thanolivine in harzburgite (Fig. 3). Olivine in lherzolitefrom Loma Caribe has distinctly lower average Fo-number and similar Ni contents than olivine in duniteand harzburgite (Fig. 3). On the other hand, the Ni con-tent of orthopyroxene and clinopyroxene is very low(NiO < 0.1 wt.%), usually below the detection limit ofthe electron microprobe.

The chemical and mineral data suggest that the prima-ry source of the Ni-rich laterite is the olivine contained inharzburgite and dunite. These data also exclude orthopy-roxene and clinopyroxene as a major source of the Nicontained in the Ni-rich laterite ore.

The Cr# (Cr/Cr+Al) of accessory chromite fromharzburgites varies from 0.45 to 0.65 in Moa-Baracoa,whereas in Mayar-Cristal ranges between 0.56 and 0.69.The Cr# of accessory chromite from dunite varies from0.44-0.58 in Moa-Baracoa, and between 0.56 and 0.70 inMayar-Cristal (Proenza et al., 1999a, b, 2001). Accord-ing to these authors, peridotites from Mayar-Baracoa beltare subduction-related.

On the other hand, the Cr# in Cr-spinel from LomaCaribe peridotites vary from 0.30 to 0.88. These largecompositional variations indicate the occurrence of peri-dotites with very different melting histories (Lewis et al.,2003). Relatively fertile peridotites as found in LomaCaribe (e.g. Cr# ~ 0.3) have not been reported in easternCuba ophiolites, where they exhibit mostly Cr# > 0.5.

Serpentinization

All the ultramafic rocks around the Caribbean havebeen serpentinized to at least some degree. Serpen-tinization is a hydrothermal process that takes place,generally, at crustal levels, and probably at tempera-tures between 200-500C (Moody, 1976; OHanley,1995). The serpentine minerals and the serpentiniza-tion process have been little investigated in theCaribbean ultramafic rocks, but it is clear that all threeof the main serpentine minerals (antigorite, lizardite



89 90 91 92 93

0.30

0.40

0.50Dunite

Harzburgite

Mayar-Baracoaperidotites

NiO

(wt

%)

Fo(%mol.)

0.30

0.40

0.50

NiO

(wt

%)

89 90 91 92 93

DuniteHarzburgiteLherzolite

LomaCaribeperidotites

Fo(%mol.)

FIGURE 3 Ni (wt%) versus Fo content [Mg/(Mg+Fe)] in olivine fromMayar-Baracoa peridotites (eastern Cuba) and Loma Caribe peri-dotites (Dominican Republic).

and chrysotile) can be found in different localities in theCaribbean.

Serpentinite samples from the Dominican Republic,Guatemala, Costa Rica and Puerto Rico were included inthe classic isotopic studies on serpentinisation of Wennerand Taylor (1971, 1973, 1974). Most of the Central Americaand Caribbean serpentinite included in the isotopic studiesof Wenner and Taylor (1974) consisted predominantly oflizardite-chrysotile, with the only occurrence of antigoritefrom Guatemala. According to Wenner and Taylor (1974)the isotopic composition of lizardite and chrysotile serpenti-nite from Dominican Republic and Puerto Rico ranged fromd18O = +6.7 to +8.7 and dD = -71 to -59. Serpen-tine from Central America (Guatemala and Costa Rica) hasd18O = +3.3 to +8.3 and dD = -97 to -78.

Serpentinites from Dominican Republic plot withinof the eastern Cuba serpentine field (Proenza et al.,

2003a), and very close to the suprasubduction domaindefined by Sakai et al. (1990). In contrast, isotopiccomposition of serpentine from Guatemala shows con-siderable variations in dD, and Costa Rica serpentineshows strong variations in d18O values, as well as rela-tively low and narrow range in dD (-94 to -91)(Fig. 4).

Based on d18O and dD (Fig. 4) values on the serpen-tine (mainly lizardite) from serpentinized chromititesand dunite from the Mayar-Baracoa Ophiolitic Belt,eastern Cuba, Proenza et al. (2003a) concluded that ser-pentinization process took place in the subocean floorscenario at moderate temperatures, and that the Mayar-Baracoa serpentines represent an example of serpentineformed during interaction with seawater. These authorsconsider that the serpentinization of the mantlesequence in the Mayar-Baracoa ophiolite occured pre-obduction, and probably took place in a suprasubduc-



-4 0 4 8

-200

-160

-120

-80

-40

0

Izu-Ogasawara-Marianaforearcseamounts

OphioliticSerpentines

EasternCubaSerpentines

ContinentalLizarditeChrysotile

d

D(S

MO

W)

d O

18(SMOW)

CostaRica

Guatemala

DominicanRepublic

PuertoRico

PuertoRican Trench

Venezuela

OceanicSerpentines

FIGURE 4 a, dD versus d1188O plot of Caribbean serpentine. The isotopic values of serpentine from Venezuela, Costa Rica, Guatemala, DominicanRepublic, Puerto Rico and Puerto Rican Trench are taken from Wenner and Taylor (1973, 1974). The isotopic domains are from: oceanic serpen-tine and ophiolite serpentine (Wenner and Taylor, 1973), continental serpentine (Wenner and Taylor, 1974), Izu-Ogasawara-Mariana forearcserpentines (Sakai et al., 1990), eastern Cuba serpentine (Proenza et al., 2003a).

tion setting (forearc environment or spreading axis of aback arc basin).

The laterite profile in central Domincan Republic

The description of the Dominican laterites in Fig. 5 isbased on the criteria used by Falconbridge Dominicana intheir mining operations (Haldemann et al., 1979). Twomain zones (horizons) can be defined: the limonite zoneand the saprolite zone. These are further divided into sub-zones labeled Zones A to F (Fig. 5). It should be notedthat the profiles vary laterally (Fig. 6) as well as vertical-ly, and no outcrop profile is identical to another. However,the basic features can be identified and matched through-out the area.

The upper zone (the limonite or laterite zone) is divid-ed into Zone A and Zone B (Fig. 5). Zone A is a choco-late brown limonite which is readily distinguished by itscolour and earthy friable texture. The upper part may con-tain concretions or pisolites of hematite up to 1 cm indiameter but the main constituent is goethite with irregu-lar pockets of gibbsite. A thin zone of ferricrete is oftenfound near the surface. Zone B is ochre-brown limoniteand is usually much thicker than Zone A. It is composed

of soft clayey material which varies in color from reddishto yellowish brown. It often forms tongues penetratinginto the underlying soft or hard serpentinite of Zone C. Itschemical composition is similar to Zone A, but it has lessiron and a higher silica content reflecting less goethiteand a higher clay content. The saprolite Zone consists ofsomewhat variably altered bedrock in which most of theoriginal structure and textures in the original bed rock(substrate) is preserved. In the upper part (Zone C of Fig.5), the material is soft and somewhat clayey and has awide color range from yellow to brown to reddish-brownwith pockets of bright green garnierite. Serpentine is themajor crystalline phase occurring with highly variableamounts of goethite and quartz. With increasing propor-tion of hard serpentine blocks, it grades downward intoZone D of Fig 5. Zone D consists of predominantly hardfragments of serpentinite from 5-25 cm in diameter set ina matrix of soft serpentine minerals. The fragments are apale yellow ochre or dark grey color and often showconcentric alteration zones. The outer skins of the frag-ments contain up to 3% or more of Ni and there is agradual decrease in Ni toward the center of the frag-ment. The iron concentration can increase to about 8percent with a decrease in the Ni/Fe ratio. This seems toinvolve leaching of the Ni3Si2O5(OH)4 component of



ZONE A

ZONED

ZONEE

ZONEF

ZONEC

ZONEB

Limonite(Upperlaterite)

Limonite(Lowerlaterite)

Softserpentinite

Hardserpentinite

Serpentinizedperidotite

Peridotite

Hematiticcap

Nickeliferouslimonite

Unalteredperidotite

Sap

roli

tic

zone

LateriteProfile

Ni

the serpentine and formation of more goethite (Golight-ly, 1981).

In places where the saprolite is developed on unser-pentinized peridotite, a different mineralogy and che-mistry is found compared with that developed on com-pletely serpentinized peridotites (Golightly, 1981). Herethe blocks of peridotite have a thin crust or skin about 4cm thick that grades rapidly into saprolite. Olivinedecomposes first into amorphous ferrisilicate hydratesor a mixture of smectite and goethite. The saprolite,often penetrated by tongues of limonite, has a coarseblocky appearance. Joints are filled by veins of gar-nierite and quartz.

Figure 7 gives the concentrations of the key elementsfor the various zones or horizons through a soil profile onthe Loma Caribe peridotite of the Dominican Republic.Iron oxide and MgO are inversely proportional. MgO andSiO2 remain high through the saprolite and then decreaseslowly through the limonite zone. Ni concentrates in theserpentine minerals (lizardite-nepouite) and garnierite,and may show considerable concentration in parts of thesaprolitic zone. Concentrations of Ni in the bedrock areconsistent at about 0.25%. In some thick saprolitic zones,the concentration of Ni is as high as 3%. Co and Ni con-centrate independently. Co does not concentrate into thegarnierite mineral structures along with Ni, but migratesand concentrates in Mn oxides (asbolanes) along with Mnat the base of the limonite zone.

There is a certain amount of concentration of Ni in theasbolane zone, but much less than Co. The Ni concentra-tion in the limonite zone varies from about 0.5 to 1.5%.

Note that Ni also varies independently from Mg in theweathering zone in contrast to its behaviour in high tem-perature mafic igneous rocks.

The laterite profile in eastern Cuba

A large province of Ni-laterites, that includes exam-ples of world-class deposits, is formed over the serpen-tinized peridotites of eastern Cuba. From west to east themining districts are: Pinares de Mayar, Nicaro, Moa andPunta Gorda (Linchenat and Shirokova, 1964; Lavaut,1998). These are mature laterites estimated to be 10-30million years old, and the minimum reserves have beenestimated at 1200 Mt of 1.3 % Ni.

The average thickness of the lateritic crust formedover the peridotite bedrock is 10 m although it can reach50 m (Lavaut, 1998). The laterite profile of the easternCuban deposits has been divided into various zones andsubzones. The classification most widely used in theCuban laterite is that proposed by Lavaut (1998). In gen-eral terms one can recognize, in eastern Cuba, the samehorizons that have been described in other Ni-lateritedeposits of the world (Brand et al., 1998; Gleeson et al.,2003, 2004). The laterite profile is composed of 4 princi-pal horizons. From bottom to top these are: (1) serpen-tinized peridotite, (2) saprolite, (3) limonite, (4) ferricrete.The lowest part of the profile is represented by tectonizedserpentinized peridotite that represents the first stages ofweathering. The saprolite zone is characterized by thepreservation of the primary fabric, marked reduction inthe quantity of primary minerals and the exclusive fm ofalteration minerals in the most fractured zones. Theboundary between the saprolitic zone and the peridotitic



FIGURE 6 Generalized profile showing lateral variation in a nickel laterite, Loma Caribe, Dominican Republic. Courtesy of Falconbridge Dominicana.

substrate (weathering front) is extremely irregular. Thesaprolitic zone passes upwards in the section through aleaching zone to a limonite zone (Fig. 8). The limonitezone is defined by its dominant mineralogic composition(goethite and hematite), and two subzones can bedefined: lower limonite and upper limonite. Finally, allthe zones of the profile can be protected from erosionby a ferruginous level of predominantly ferricrete (resid-ual limonite).

Serpentinized peridotite constitutes the major sub-strate over which the laterites are developed. The princi-pal source of Ni is from olivine (up to 0.4 wt%, Proenzaet al., 2003c).

The laterites developed over the Moa-Baracoa peri-dotites that have associated bodies of gabbro give a differ-ent soil profile. These bodies correspond with the pres-ence of abundant dikes and sills of gabbro in theperidotites and are transformed to bauxite. In these soilzones, the content of Ni diminishes abruptly and they areconsidered sterile intercalations that can generate pro-blems in the metallurgical process.

In some of the Ni-laterite deposits (e.g. Punta Gordadistrict) typical structures characteristic of redepositedlaterite are observed (Formell, 1979; Formell and Oro,1980). The laterites formed on slopes were periodicallytransported to more depressed zones covering previouslyformed laterites. Later, lateritic processes modified the

layering of the re-deposited laterite and obliterated theprimary zoning of the laterite profile.

The principal minerals that contain Ni in the saprolitezone of the Cuban deposits are lizardite-nepouite[(Mg,Ni)3Si2O5(OH)4], along with lesser amounts ofclinochlore-nimite [(Mg,Ni,Al)6(Si,Al)4O10(OH)8]. In thelimonite zone, the Ni is associated principally withgoethite and a minor quantity with spinel and oxides andhydroxides of Mn (Rojas-Puron and Orozco, 1994;Oliveira et al., 2001; Rodriguez-Pacheco et al., 2003),whereas Co is associated with Mn minerals (Oliveira etal., 2001).

The Ni-Co laterite deposits of eastern Cuba, in agree-ment with the classification of Brand et al. (1998) are theoxide type. This classification is based on the mineralogyof the principal Ni-bearing phases. According to Gleesonet al. (2003) the best known example of Ni-lateritedeposits of oxide-type are those of Moa (Moa Bay in theEnglish literature).

Comparison of the laterite profiles in easternCuba and Central Dominican Republic

Although more units are recognized in the easternCuban laterite profiles and the nomenclature used differs(Lavaut, 1998), in general, the main features of the east-ern Cuba laterites are very similar to those in the Domini-can Republic.



A

B

C

D

E

F

0

10

20

Depth Orezone

(x10)CoFe MgO SiO2 Ni

10 20 30 40 50Wt.% 0.5 1.0 1.5 2.0 2.5

FIGURE 7 Variation in chemical composition through a nickel laterite from Loma Caribe, Dominican Republic. Courtesy of Falconbridge Dominicana.

In eastern Cuba, a significant fraction of the nickel isassociated with Fe hydroxides and oxides in the upperpart of the profile (limonite zone); these deposits areclassified as oxide-type (Glesson et al., 2003). In con-trast, in the Domincan laterites most of the nickel isfound in the saprolite zone and these deposits are classi-fied as the silicate-type deposits (highest-grade Ni lat-erite deposits). The main ore consists of hydrated Mg-Nisilicates (serpentine and garnierite) occurring deeperin the profile (zone C to D). These zones account for40% of the total ore reserves, while zone B (limoniteore) is 25% of the reserve.

The main difference seems to be that Zones C andD, i.e. the saprolitic horizons are less developed ineastern Cuba whereas the limonite horizons are betterdeveloped, at least in the Moa Bay area. The reasonfor this feature has not been properly investigated butcould likely be related to the fact that the Moa Baylaterites (oxide type) are developed over a broadplateau with a gentle slope to the coast. Under thesetopographical conditions, which are unusual accor-ding to Golightly (1981), laterite can develop com-pletely down to the water table. Given sufficient timeand no rejuvenation a limonite zone will developwithout significant saprolite. In contrast, Dominicanlaterites, as well as other silicate-type laterite, can beformed where there is slow continuos tectonic upliftand the water table is kept low in the profile. In thissituation, weathering over long periods can result indevelopment of a thick saprolite horizon (Golightly,1981; Elias, 2002).

One of the factors that must have influenced the devel-opment of soils is the contrast in the tectonic evolutionbetween Cuba and Hispaniola during the Neogene (Lewisand Draper, 1990; Mann et al., 1991). In the DominicanRepublic upward movement of the serpentinized peri-dotites occurred in the late Oligocene and it is generallyagreed that the serpentinites along with other older unitswere exposed to weathering and erosion in the earlyMiocene. Lateritization began at this time. This Mioceneland surface was subsequently broken into blocks by verti-cal movements associated with transpressional movementalong major faults. At least four physiographic cycles can berecognized corresponding to different surface levels (Halde-mann et al., 1979). Lateritization still continues today.

In Cuba, the situation is different. The fact that marinesediments of late Eocene and Miocene age overlie the ser-pentinites unconformably as seen along the northeastcoast indicates that the serpentinites were exposed at thesurface by this time. The extensive east-west strike-slipfaulting that affected most of Hispaniola throughout theNeogene was not developed in Cuba. Hence, althoughgradual regional uplift occurred which would have aidedthe lateritization process the peridotite massifs were notbroken into smaller blocks that would have undergonedifferential uplift as a result of strike-slip movement.

CONCLUSIONS

This review shows that major progress has been madein the last 25 years in our understanding as to the role of



0 1 2

Mn

Ni

Cr

0 10 20 30 40 50

Fe

Si

Mg

Al

Ferricrete

Limonite

Upper

Lower

Saprolite

Transitionzone(leachingzone)

Peridotite

(wt%)

FIGURE 8 Schematic laterite profile developed on eastern Cuba ultramafic rocks, showing the variation in chemical composition through each zone(from Rodrguez-Pacheco et al., 2003).

ophiolite-related ultramafic rocks in the evolution of theCaribbean. However many problems remain. The follow-ing comments relate to ultramafic rocks and their signifi-cance in the ophiolite concept as applied to theCaribbean.

1) It is important to retain the ophiolite concept in thestudies concerning Caribbean ophiolites and consider therocks (peridotites, gabbros, basalts, sedimentary rocks)not as isolated units, but as an association. For example,in considering melting processes and magma types, thecomposition of the rocks in the crust-mantle transitionzone (mantle peridotites and the intruding dike rocks andchromitites) must be examined together with crustalbasalts and gabbros from the same complex. Although thedifferences are small much can be learned about the com-positions of the peridotites through detailed mineralogicaland geochemical studies (Proenza et al., 1999b; Lewis etal., 2003; Marchesi et al., in press) and by interpretationof the nature of the mantle at the time of melting and themagma compositions.

2) Although SSZ ophiolites have been identified in anumber of localities in the Caribbean, based on the miner-alogy of the peridotites or on basalt chemistry, no exam-ple has yet been shown to be linked to the development ofthe axial part of an oceanic arc (Shervais, 2001). With fur-ther studies of mlange rock fragments, it should be pos-sible to better determine different types of subductionzones and the polarity of subduction. However, these bod-ies are often highly dismembered and metamorphosed sothere should be caution in the use of these data. Randomcollections can be misleading.

3) Caribbean ophiolites show many of the featuresseen in orogenic belts in other parts of the world but therecognition in the Caribbean basin itself of plateaus ofthickened basaltic crust formed by major eruptions ofbasalt over a short period of time is a particularly interest-ing feature. Parts of the oceanic plateau in westernColombia, western Central America, Southern Peninsulaof Haiti and possibly central Hispaniola (Duarte Com-plex) have been found accreted to the Caribbean marginsas expected for buoyant plateau crust, as opposed to nor-mal oceanic crust which should be subducted (Kerr et al.,2003). Dilek (2003) has termed these as ophiolites of theCaribbean-type even though ultramafic rocks are general-ly absent in these oceanic plateau complexes. This term asused by Dilek maybe confusing to readers not familiarwith the area since these plateau (CLIP) rocks are not ty-pical of the major ophiolite bodies in the northernCaribbean such as those of Cuba. The reason for theabsence of oceanic plateau crust in Cuba and elsewhere,particularly where mantle peridotites are dominant, maybe due to factors of crustal buoyancy, stress geometry or

simply other oceanic crust in the area at the time of colli-sion and accretion, but this question should be pursuedmore vigorously.

4) The Caribbean has the largest resources of Ni-Colaterite in the world formed by the weathering of serpen-tinized peridotites under special conditions of climate andtectonics of the Caribbean in the Tertiary. Unfortunately,these deposits have been studied in a rudimentary wayonly and many questions need to be answered. The rela-tion between laterite composition and composition of theserpentinite/peridotite protoliths is not known. There hasbeen no recent study of the mineralogy and geochemistryof the deposits in order to determine and compare quanti-tatively the variation in chemistry through the soil profilesof different deposits. Such studies are important not onlyto evaluate the deposits economically, but to investigatehow the deposits were formed. In addition a unique florais found on the Caribbean Ni-laterite soils (Brooks,1987), but the nature and formation of this flora and itsrelation to the composition of the soils has been littlestudied. Such studies should be multi-disciplinary.

ACKNOWLEDGEMENTS

We are indebted to our colleagues Edwin Devaux and Fran-cisco Longo at Falconbridge Dominicana, Bonao, for their con-tinued support in our research on the Loma Caribe peridotiteand laterites. We thank our Cuban colleagues, in particular thoseat the Empresa Geolgica Minera de Oriente and the InstitutoSuperior Minero Metalrgico in Moa, for sharing their knowl-edge of the Cuban nickel laterites. We are grateful to ManuelIturralde-Vinent for discussion and for sending us informationon the Camagey ophiolite. This paper originated with a talk byJFL on Caribbean serpentintites and laterite soils given at the7th Latin American Botany Congress held in Mexico City inOctober 1998. JFL thanks Julio Figueroa from Puerto Rico forinviting him to participate in the conference. The paper hasevolved considerably since then as a result of more detailedknowledge of the Northern Caribbean ophiolites and lateritesand the opportunity to exchange information through the IGCP433 project to which this paper is a contribution. This paper waspartially supported by the Spanish project BTE2001-3308. Con-structive reviews by Fernando Gervilla and an unknown refereeare greatly appreciated.

REFERENCES

Abbott, R.N., Jackson, T.A., Scott, P., 1999. The serpentiniza-tion of peridotite from Cedar Valley, Jamaica. InternationalGeology Review, 41, 836-844.

Abbott, R.N., Draper, G., Shantanu, K., 2004. UHP magma par-agenesis, garnet peridotite, and garnet clinopyroxenite: an



example from the Dominican Republic. International Geolo-gy Review, 47, 233-247.

lvarez, J., 1983. Rocas ultramficas-mficas en Colombia ydepsitos minerales asociados. Medellin, Informe del Insti-tuto Nacional de Investigaciones Geolgicas Mineras (Ingeo-minas), 67 pp.

lvarez, J., 1987. Mineraloga y qumica de los depsitos decromita podiforme de las dunitas de Medelln, Departamentode Antioquia, Colombia. Boletn Geolgico, 33(1-3), 34-46.

Anam, K.I., 1994. Petrology and geochemistry of some highpressure rocks from the northern part of the Ro San JuanComplex, Dominican Republic. MS thesis. Florida Interna-tional University, Miami, Florida, 127 pp. (Unpublished).

And, J., Harangi, Sz., Szkmny, By., Dosztly, L., 1996. Petrologade la asociacin ofioltica de Holgun. In: Iturralde-Vinent, M.A.(ed.). Ofiolitas y arcos volcnicos de Cuba. Miami, USA, IGCPProject 364, Special Contribution n.1, 154-178.

Anonymous, 1972. Penrose Field Conference on ophiolites.Geotimes, 17, 24-25.

Aspden J.A., McCourt, W.J., Brook, M., 1987. Geometrical con-trol of subduction-related magmatism: the Mesozoic andCenozoic plutonic history of western Colombia. Journal ofthe Geological Society of London, 144, 893-905.

Auzende, A.L., Devouard, B., Guillot, S., Daniel, I., Baronnet,A., Lardeaux, J.M., 2002. Serpentinites from Central Cuba:petrology and HRTEM study. European Journal of Minera-logy, 14, 905-914.

Av-Lallement, H.G., 1997. Transpression, displacement parti-tioning and exhumation in the eastern Caribbean/SouthAmerican plate boundary. Tectonics, 16, 272-289.

Av-Lallemant, H.G., Sisson, V.B. (eds.), 2005. Caribbean-South American plate interactions, Venezuela. GeologicalSociety of America Special Paper, 394, p. 346.

Barker, W., Welch, S., Banfield, J., 1997. Biogeochemical weath-ering of silicate minerals. In: Banfield, J.F and Nealson, K.H.(eds.). Geomicrobiology: Interactions between Microbes andMinerals. Reviews in Mineralogy, 35, 391-428.

Baumgartner, P.O., Flores, K.E., Denyer, P., 2004. Discovery ofocean remnants in the Siuna area (NE Nicaragua). II SwissGeoscience Meeting, Lausanne, Abstract, p. 80.

Beccaluva, L., Bellia, S., Coltorti, M., Dengo, G., Giunta, G.,Mendez, J., Romero, J., Rotolo, S., Siena, F., 1995. Thenorthwestern border of the Caribbean Plate in Guatemala;new geological and petrological data on the Motagua ophi-olitic belt. Ofioliti, 20, 1-15.

Beccaluva, L., Coltorti, M., Giunta G., Iturralde-Vinent M.,Navarro, E., Siena F., Urbani F., 1996. Cross sectionsthrough the ophiolitic units of the southern and northernmargins of the Caribbean Plate in Venezuela (NorthernCordilleras) and central Cuba. Ofioliti, 21, 85-103.

Beccaluva, L., Chinchila-Chaves, A.L., Coltorti, M., Giunta, G.,Siena, F., Vaccaro, C., 1999. Petrological and structural sig-nificance of the Santa Elena-Nicoya ophiolitic complex inCosta Rica and geodynamic implications. European Journalof Mineralogy, 11, 1091-1107.

Bellizia, A., Dengo, G., 1990. The Caribbean mountain system,northern South America: a summary. In: Dengo, G., Case,J.E. (eds.). The Geology of North America. The Caribbeanregion. Boulder Colorado, Geological Society of America,Volume H, 167-175.

Bourgois, J., Toussaint, J.F., Gonzales, H., Azema, J., Calle, B.,Desmet, A., Murcia, L.S., Acevedo, A.P., Parra, E., Tournon,J., 1987. Geological history of the Cretaceous ophioliticcomplexes of Northwestern South America (ColombianAndes). Tectonophysics, 143, 307-327.

Bowin, C., 1966. Geology of the Central Dominican Republic(case history of part of an island arc). In: Hess, H. (ed.).Caribbean Geological Studies. Geological Society of Ameri-ca Memoir 98, 11-84.

Bowin, C.O., 1975. The geology of Hispaniola. In: Nairn,A.E.M., Stehli, F.G. (eds.). The Ocean Basins and Margins.3. The Gulf of Mexico and the Caribbean. New York,Plenum ed., 500-552.

Bowin, C.O., Nalwark, A.J., Hersey, J.B. 1966. Serpentinizedperidotite from the north wall of the Puerto Rico Trench.Geological Society of America Bulletin 77, 257-270.

Brand, N.W., Butt, C.R.M., Elas, M., 1998. Nickel laterites:classification and features. Journal of Australian Geologyand Geophysics, 17, 81-88.

Brooks, R.R., 1987. Serpentine and its vegetation: A multidisci-plinary approach. Portland, Oregon. Dioscorides Press, 454pp.

Burk, C.A., (ed.), 1964. A study of serpentinite, the AMSOCcore hole near Mayaguez. Puerto Rico. National Academyof Sciences, National Research Council, Pub. 1188.

Case, J.E., Duran, S.L.G., Lpez, A., Moore, W.R., 1971. Tec-tonic investigations in western Colombia and eastern Pana-ma. Geological Society of America Bulletin, 82, 2685-27

Documents

Laterita niquelífera