University of Rhode Island DigitalCommons@URI Graduate School of Oceanography Faculty Graduate School of Oceanography Publications

1997 Peninsula, : A single suite of Caribbean oceanic plateau magmas Christopher W. Sinton University of Rhode Island

Robert A. Duncan

See next page for additional authors

Follow this and additional works at: https://digitalcommons.uri.edu/gsofacpubs

Terms of Use All rights reserved under copyright.

Citation/Publisher Attribution Sinton, C. W., R. A. Duncan, and P. Denyer (1997), , Costa Rica: A single suite of Caribbean oceanic plateau magmas, J. Geophys. Res., 102(B7), 15507–15520, doi: 10.1029/97JB00681. Available at: https://doi.org/10.1029/97JB00681

This Article is brought to you for free and open access by the Graduate School of Oceanography at DigitalCommons@URI. It has been accepted for inclusion in Graduate School of Oceanography Faculty Publications by an authorized administrator of DigitalCommons@URI. For more information, please contact [email protected]. Authors Christopher W. Sinton, Robert A. Duncan, and Percy Denyer

This article is available at DigitalCommons@URI: https://digitalcommons.uri.edu/gsofacpubs/254 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 102, NO. B7, PAGES 15,507-15,520,JULY 10, 1997

Nicoya Peninsula, Costa Rica: A single suite of Caribbean oceanic plateau magmas

ChristopherW. Sinton1 andRobert A. Duncan Collegeof Oceanicand Atmospheric Sciences, Oregon State University, Corvallis

Percy Denyer EscuelaCentroamericana de Geologia,Universidad de CostaRica, SanJose, Costa Rica

Abstract. The pre-Tertiaryoceanic crust exposed on the westcoast of CostaRica hasbeen broadlyreferred to asthe Nicoya Complex. This study was designed to determinethe age of the NicoyaComplex in theNicoya Peninsula, Playa Jac6, and the Quepos Peninsula using 4øAr-39Ar radiometricdating and to assessthe petrologic relationships between the different localities using majorelement, trace element, and Sr, Nd, Pb isotopicdata. Radiometricages of basaltsand diabasesfrom theNicoya Peninsula are 88-90 Ma (with a weightedmean of 88.5 Ma), andthose of two intrusiverocks (a gabbroand plagiogranite) are both 83-84 Ma. The combinedgeochemical dataindicate that the sampled Nicoya Peninsula rocks belong to a singlesuite related by fractional crystallizationof similarparental magmas. Nd andPb isotopic ratios indicate a commonmantle sourcedistinct from thatof mid-oceanridge basalts. Both the age and composition of theNicoya rocksare consistent with theidea that they are a partof the CaribbeanCretaceous oceanic plateau [Donnelly,1994]. TheJac6 lavas are geochemically similar to theNicoya Peninsula suite, and a singleage of 84 Ma is identicalto theage of theNicoya Peninsula intrusives. The one analyzed Queposbasalt has a radiometricage of ~64Ma, andit is enrichedin incompatibleelements relative to theNicoya rocks. Similarities in Nd andPb isotopicratios indicate that the Quepos and Nicoya/Jac6lavas were derived from a similarmantle source to thatwhich produced the Nicoya rocks,possibly the Galapagosplume.

Introduction was described as normal oceanic crust of Jurassic to Early Cretaceous age, and the upper unit was subdivided, based on Obductedoceanic crust exposed on the west coastof Costa composition, into a suite of Late Cretaceous oceanic rocks Rica is broadly referred to as the Nicoya Complex [Dengo, associatedwith the Caribbeanoceanic plateau and a primitive 1962] (Figure 1). Theserocks are composedof intermingled arc series. However, our re-examination of contacts between basalts (massive and pillowed flows, dikes, and breccias), chert and basalt demonstrated only intrusive (basalt into gabbros and plagiograniteswith interspersedpieces of sediment) rather than depositional (sediment on basalt) radiolarianchert [e.g., Kuijpers, 1980;Alvarado et al., 1993; relationships,so the sedimentage cannot be used to constrain Denyer and Arias , 1993]. Outcropsare generallylimited to the basalt age. The division of the igneous rocks into two coastalpeninsulas with the most extensiveexposures on the seriesof different ages is thereforeopen to question. Nicoya Peninsula.The full extentof the obductedcrust is not The Caribbean oceanic plateau (also called the Caribbean preciselyknown, but it is possiblethat the Nicoya Complex Cretaceous Basalt Province by Donnelly et al. [1990]) is is part of a belt of accretedMesozoic oceaniccrust that composedof the thick oceaniccrust that makesup the core of extendssoutheastward through Panama into westernColombia the Caribbeanplate and numerouspieces of obductedoceanic and Ecuador[e.g., Escalante, 1990]. crust along the plate margin [Donnelly et al., 1990]. On the basis of the fossil ages of the radiolarian chert Debatable aspects of the Caribbean plateau include its associatedwith the igneous rocks, most previous workers formation location (Pacific or in situ) and the plate tectonic have divided the oceanic igneous rocks of the Nicoya setting of origin (i.e., intraplate, ridge centered, back arc). Peninsulainto an upper and a lower unit [Schrnidt-Effing, The preexistingoceanic lithospherewhich, based on seismic 1979; Kuijpers, 1980; Bourgoiset al., 1984; Gurskyet al., reflection profiles [Diebold et al., 1981] and plate 1984; Wildberg, 1984; Frisch et al., 1992]. The lower unit reconstructions[Duncan and Hargraves, 1984], purportedly lies beneath the plateau has not yet been definitively identified in any of the on-land exposures nor has it been 1Now at GraduateSchool of Oceanography,University of Rhode sampledby deep sea drilling. The exposedoceanic crust in Island, Narragansett. western Costa Rica may play a very important role in understandingthe nature of the plateauif, indeed, somepart of Copyright1997 by theAmerican Geophysical Union. the Nicoya Complex representsthis preexistingoceanic crust. Papernumber 97JB00681. This possibility can be evaluatedby determiningthe age and 0148-0227/97/97 JB-00681 $09.00 compositionof the lavas.

15,507 15,508 SINTONET AL.' THE NICOYA PENINSULA

83.2 _+1.3 Peninsula

92.5 _+ 5.4 10030 ' .2 Nico' 1el0 • 90.3_+3.2 Peninsula PI. Jac6! ;00 PI. Herradura 89.7 + 1.7: 19 / •-• 18 83.8_+1.1 88.0_+0.7 84.2 _+ 1.8 1-33 63.9 _+ 0.5 Sta. Cruz 22

21 Nicoya

89.9 _+ 2.8

10 ø 23 Carmona

85 ø 30' Figure 1. Schematicmap of the Nicoya Peninsulawith inset showinglocation of Playa Jac6 and Quepos. Solid circlesare sample locations. The 4øAr-39Ar plateau ages of basaltsand diabases are given, with the ages of the intrusive rocks in italics. Thin lines on the Nicoya Peninsulaare major roads.

This projectwas designedto (1) determinethe radiometric similar mantle source and that all can be related by fractional age of rocksfrom the NicoyaComplex, (2) testthe idea of an crystallizationof similar parental melts. Nd and Pb isotopic upper and lower series within the Nicoya Peninsulausing ratios from the Nicoya Peninsula,Playa Jac6, and Queposare 4øAr-39Ardating and geochemistry, and (3) betterconstrain similar but are distinct from normal mid-ocean ridge basalts the origin of the Nicoya rockswith an improvedgeochemical eruptedon the East Pacific Rise. Altogether,these data do not data set. In addition, some samples were collected from support the division of the Nicoya Peninsula igneous rocks coastalexposures of oceanicbasalts at Playa Jac6 and the into two or more suites; rather, the rocks appear to be QueposPeninsula for comparisonto the Nicoya rocks. We genetically related and are part of the Caribbean oceanic report4øAr-39Ar radiometric ages as well as majorelement, plateau. trace element, and Sr, Nd, and Pb isotopic data from these three localities. Our data show that rocks from the Nicoya Previous Work Peninsula erupted at 88-90 Ma, which is contemporaneous with other on-land Caribbeanplateau sites in Haiti, Curaqao, Donnelly [1994] summarizedthe previous studiesof the and western Colombia [Sinton, 1996]. The geochemicaldata, Nicoya Complex and the reader is referred to his paper for a in conjunctionwith previouslypublished data, show that all more thorough description of these. The majority of the of the rocks collectedfrom the Nicoya Peninsulacome from a exposuresof igneousoceanic rocks and oceanicsediments are SINTONET AL.:THE NICOYAPENINSULA 15,509 on the Nicoya Peninsula, and smaller exposuresare found [Gradstein et al., 1994]) Sabana Grande sediments along the coast to the north, at the Santa Elena Peninsula,and conformably overlie the igneous rocks [Bourgois et al., to the south, at Playa Jac6/PlayaHerradura, and the Quepos 1984], placing a minimum age on eruptionand crystallization Peninsula(Figure 1). For clarity, the rocks will be referredto of the magmas.This formationis a deepwater assemblage of by their locality, so the term Nicoya in this paper will refer to a sedimentarybasalt/radiolarian chert breccia base overlain by the Nicoya Peninsula(samples designated NC in Table 1). thin siliceousmudstones grading into limestone[Bourgois et In general, previous authors have subdividedthe Nicoya al. , 1984]. Peninsula igneous rocks into two general units, termed the There are no reportedradiometric or fossil ages for the Matapalo and Esperanza units by Kuijpers [1980] and Playa Jac6/Herradurarocks. The basalts at Quepos are Bourgoiset al. [1984]. This division was originally basedon interbeddedwith Paleoceneforaminifera and are overlainby fossil age and field observations [Schrnidt-Effing, 1979; early Eocenesediments [Schmidt-Effing, 1979; Baumgartner Baurngartner et al., 1984], although no contact between the et al., 1984], somewhatolder than the K-Ar reportedage of two units was defined [Kuijpers, 1980]. Later studiesby 46.2 + 4.4 Ma [Alvarado et al., 1992 and referencestherein]. Wildberg [1984] and Gursky et al. [1984] referred to thesetwo units as the Lower and Upper Nicoya, with the units separated by a discontinuouslayer of radiolarianchert. However, there Sample Descriptionsand Mineralogy are no observeddepositional contacts between the radiolarian chert and the igneous rocks. This observation as well as Rock sampleswere collectedfrom the Nicoya Peninsula, considerable differences in the interpretation of the Playa Jac6,and Queposin March 1993. Samplelocations are stratigraphy and origin of the rocks led Donnelly [1994] to shownin Figure 1, and brief descriptionsare givenin Table 1. suggestthat the Nicoya Peninsula igneous rocks may be a Samplingwas concentratedin the northernNicoya Peninsula single unit. becausethe goal wasto identifyand characterize the proposed Previous radiometric ages of the Nicoya Peninsula rocks oldernormal oceanic crust in this area. The bestexposures are have not been useful in constrainingtheir magmatichistory. along the coast and road cuts. However, even in large Alvarado et al. [1993] compiledK-Ar agesthat rangefrom 30 exposures,bedding or paleohorizontaldirection is impossible to73 Ma, reflecting theeffects ofalteration and40Ar loss and to determineowing to faulting, surficialweathering, and the not the actual time span of volcanism. There have been often limitedoutcrop area. Someoutcrops show spheroidally several reports of fossil ages from the radiolarian chert weatheredrocks in which well-preserved,rounded clasts of assemblagesthat occur as inclusionswithin the igneousrocks diabaseare foundin a matrix of heavily weatheredand friable throughout the peninsula. These assemblageshave been material. Pillow structures are rarely observed in the assignedages of Early Jurassic(Callovian) to Late Cretaceous northwesternpeninsula. (early Santonian)[Galli-Olivier, 1979; Bourgoiset al., 1984; Fractured,massive basalt flows exposedat the northernand Baurngartner et al., 1984]. However, becauseof the intrusive southernends of Playa Jac6 were also sampledas well as at a relationshipsof the igneous rocks into the radiolarian chert, road cut on the coastroad just southof Playa Jac6 (Table 1). the fossil ages provide only a maximumcrystallization age of All the basaltsexamined along the coastat Queposare heavily the igneous rocks. The late Campanian (-75-70 Ma, altered and only three samples with the least apparent

Table 1. CostaRica Sample Descriptions

Sample LocationDescription RockType Condition

NC93-2 massive basalt in Rio Brasil, Pilas mediumgrain slight alteration,moderate alteration basalt near veins NC93-3 exposedhill nearroadside, 2 km westof medium-fine slightalteration, plagioclase 0-10% grainbasalt replacedby clays NC93-11 smallroad cut across from entrance to Hotel Playa mediumgrain slightalteration, plagioclase 10% Azucar, spheroidalblocks basalt replacedby clays NC93-13 roadcut 3 km northof diabase slightalteration, some plagioclase cores beginningto alterto clays NC93-14 northof El Ed6n, follow QuebradaSalto to hill whererock gabbro altered,plagioclase partially replaced by exposed,great variation in grainsize over shortdistance, clays,some primary twinning seen, with contact unknown graphicquartz and plagioclase NC93-18 quan'y2 km westof Coyolito,overlain by radiolarite,contact mediumgrain lightalteration, some quartz veins inaccessible basalt NC93-20 roadcut at top of Vista al Mar fine grainbasalt light to moderatealteration NC93-34 northside of PuntaSabana, flat-lying and cliff-exposed diabase lightalteration, plagioclase 0-25% basalt/diabaseand radiolarite, basalt is massive,withigneous replacedby clays,depending on andfault contact w/radiolarite, primary bedding of proximity to veins radiolarite(red and gray/white) preserved, cut by dikes NC93-26 Coastalexposure, south of Pl.ayaMonte del Barco(north of plagiogranite altered,plagioclase partially replaced by Playa Panama) albite+ clay, pyroxenewith fractured texturealtering to secondaryoxides JC93-1 roadcut 1.2 km southof lastentrance to PlayaJac6, massive, mediumgrain variablyaltered fracturedbasalt, no sediments,20 m high basalt JC93-3 northend of PlayaJac6 Basaltcrops out on cliffs andflat mediumgrain variablyaltered, heavy vein penetration lying on beach,similar to S end basalt QP93-1 southend of beach1 at Quepos,basalt in cliff andlarge fine grain basalt light to moderatealteration, some fresh "islands"on beach,massive with columnarjoints, some butpartially resorbed plagioclase alteration and veins phenocrysts 15,510 SINTONET AL.: THE NICOYA PENINSULA

alteration were taken from the beachjust north of the entrance contamination(4øAr/36Ar = 295.5). From the heating steps, to Manuel Antonio National Park (Table 1). age spectrumdiagrams (age versus % 39Ar released)were Thin sectionswere made of all of the samplesin order to generated (Figure 2). Plateau ages were calculated from determine their mineralogy and suitability for geochemical consecutivesteps that are concordantwithin 1(• errorusing and geochronologicalanalyses. The rocks were divided into the proceduredescribed by Dalrympleet al. [1988], in which five categories based on petrographic observations:basalt, stepages were weightedby the inverseof their variance(Table basaltic breccia, diabase,gabbro, and plagiogranite.Basaltic 2). Isotopecorrelation diagrams (36Ar/4øAr versus 39Ar/4øAr) samplesare fine grained and generally aphyric and consistof were also made for each analysis,in which the slope is plagioclase, pyroxene, Fe-Ti oxides, and altered mesostasis. proportionalto the age (isochron)and the inverseof the y Only one basaltic breccia was collected (NC93-25), and it is interceptgives the initial 4øAr/36Arcomposition. The composed of a heavily altered matrix surrounding basalt isochronages are calculatedusing the sameheating steps as in clasts. Diabasic samples have the same mineralogy as the the plateau ages. basalts except that they are holocrystalline, are coatset We definea reliablecrystallization age based on the criteria grained, and have trace amounts of tabular apatite. of Pringle [1993]as follows: (1) a well-definedage spectrum Spheroidally weathered samples are usually diabasic in plateauof at least three concordantconsecutive steps that texture. Isotropic gabbroic samples have the same representsat leastof 50% of thetotal 39Ar released; (2) an mineralogy as the basaltsand diabasesbut have grain sizesup isochronage concordantwith the plateauage and with a value to 10 mm. The plagiogranitesall have coarsetextures and are for theZ2-statistic SUMS/N-2 [York, 1969] below the 95% composed of myrmeketic plagioclase and quartz with lesser confidencelevel; and (3) a 4øAr/36Arintercept that is amounts of green clinopyroxene (hedenbergite) and Fe-Ti statistically indistinguishablefrom the atmosphericvalue 295.5. oxides. A cumulategabbro (NC93-100) containslarge (-4 cm diameter) unzoned plagioclase (An85_86), clinopyroxene The SUMS/N-2 cutoffvalues are takenfrom Pringle [ 1993, (Ens1) [Sinton, 1996], and trace olivine (completelyreplaced Table 2]. For three-, four-, and five-point isochrons,the by iddingsite) set in a diabasic matrix of plagioclase, cutoff valuesare 3.84, 3.00, and 2.60, respectively. clinopyroxene, and Fe-Ti oxides. Many of the samplesfrom the NicoyaPeninsula displayed The rocks show varying degrees of alteration with most erratic step ages, from which no age plateausor isochrons could be discerned. This can be attributed to a combinationof samples having relatively pristine pyroxene and relatively fresh to partially altered plagioclase (to colorless clays). In alterationeffects, 37Ar and 39Ar recoil, and the low K20 the basalts, the mesostasis has altered to chlorite and fine- contentsof the rocks. Someof the plagioclaseseparates grained secondary oxides. Thin veins of prehnite and rare (NC93-36 and NC93-9) gave step ages that increasedwith calcite are found in most of the samples, and the density of increasingtemperature, up to 180 Ma. Evidentlythey veins generally correlateswith the degree of alteration. The containedsome component of excessradiogenic 4øAr of plagiogranitesare all moderatelyto heavily altered. whicheven small amounts can produce erroneous ages, given the low K20 content(<0.10%) of theserocks. Plagioclase separatesfrom all but one of the analyzedplagiogranite Results samplesgave nonconcordant step ages that were usually less than 80 Ma. In this section,we reportonly the resultsfrom Geochronology the samplesthat yieldedreliable crystallization ages (Table 2). Age determinations for whole rock and plagioclase Of 13 analyzedsamples from the Nicoya Peninsula, 7 gave separatesfrom the least altered sampleswere performed at reliablecrystallization ages (Figure 2). Five (four basaltsand OregonState University (OSU) usingstandard 4øAr-39Ar one diabase)gave plateauages ranging from 88.0 to 92.5 Ma incrementalheating techniques [Duncan and Hargraves, 1990; (a weightedmean of the agesis 88.5 Ma anda nonweighted Duncan and Hogan , 1994]. Whole rock sampleswere either mean is 90.1 Ma). Of particularinterest is the diabaseNC93- crushed in bulk in a ceramic jaw crusher and sieved to a 34 (PuntaSabana), which was taken near a largeinclusion (50 uniform 0.5-1 mm grain size or cut into mini cores. m) of radiolarianchert. The plateauage of 88.0 + 0.7 Ma Plagioclase crystals were magnetically separated from indicatesthe age of intrusionand confirmsthat magmatism diabase, gabbro, and plagiogranite samples, checked for occurredwell after the Jurassic-EarlyCretaceous radiolarian purity under a binocular microscope,and washedfor 30 s in chert. Plagioclaseseparated from a gabbro(NC93-14B) and a 5% hydrofluoricacid. All sampleswere ultrasonicallycleaned plagiogranite(NC93-26) gave statisticallyindistinguishable in distilled water. Sampleswere sealedin evacuatedquartz plateau ages of 83.8 + 1.1 Ma and 83.2 + 1.3 Ma, glassvials and irradiatedfor 6 hoursat the OSU TRIGA nuclear respectively. The significantlylower ages of the intrusive reactorfacility. Neutron flux during irradiationwas monitored rocksrelative to the analyzedbasalts and diabaseindicate that by FCT-3 biotite (27.7 Ma, [Hufforal and Hammerschmidt, somemagmatic activity occurredat least 5-7 m.y. after initial 1985]). Ar isotopic compositionsof the crushedwhole rock basalt eruptionsat 88-90 Ma. samplesand someof the plagioclaseseparates (-•500 mg) were One of two analyzedbasalts from Jac6(JC93-1) yielded a determined using an Associated Electrical Industries (AEI) reliablecrystallization age with a three-stepplateau of 84.2 + MS-10S mass spectrometer. The mini cores and the other 1.8 Ma, an isochron of 83.4 + 2.5 Ma, and an initial plagioclaseseparates (-•100 mg) were analyzedusing a Mass 4øAr/36Arratio of 295.6+ 6.9 (Figure2). Thelower initial AnalyzerProducts (MAP) massspectrometer, equipped with a step age indicatessome radiogenic4øAr loss from less low-blank, double-vacuum resistance furnace. retentive (alteration) sites. The consistentplateau and Individualages for eachof the4øAr-39Ar temperature steps isochron ages of this sample indicate a volcanic event were calculated after corrections for background, mass youngerthan the eruptionof the Nicoya Peninsulabasalts but fractionation, isotopic interferences,and atmosphericargon perhapsrelated to the intrusions. SINTON ET AL.: THE NICOYA PENINSULA 15,511

o

ß

, , . , o

•-• o o o o .1V½ /-•V9œ o ("IN)

'1 ' '

o o o o o o o

o.o JVo•/JV9œ ("IN) •V ("I•) •V

8 • o• o •o o

(•I•) •V 15,5!2 SINTONET AL.:THE NICOYA PENINSULA

Table2. The4øAr-39Ar Plateau and Isochron Ages for Basaltic Rocks From the Nicoya Complex, Costa Rica

Sample Material PlateauAge 39ArPercent IsochronAge, m.y. N 4øAr/36ArIntercept _+ SLIMS by 1/02, m.y ofTotal 10 N-2

Nicoya PeninsulaBasalt and Diabase NC93-2 whole rock 90.3 + 3.2 100 87.8 + 0.3 4 296.3 + 2.6 1.00 NC93-3 whole rock 92.5 + 5.4 100 86.8 + 7.4 5 302.0 + 6.6 0.52 NC93-18 whole rock 89.7 + 1.7 100 92.3 + 3.0 5 294.0 + 4.4 0.47 NC93-20 whole rock 89.9 + 2.8 100 85.9 + 3.3 5 299.7 + 6.7 0.16 NC93-34 whole rock 88.0 + 0.7 97 86.4 + 1.3 5 308.1 + 5.5 0.29

Nicoya PeninsulaIntrusives NC93-14B plagioclase 83.8 + 1.1 100 80.9 + 1.1 7 301.5 + 1.1 2.00 NC93-26 plagioclase 83.2 + 1.3 100 83.2 + 2.0 7 295.2 + 2.3 2.20 JC93-1' wholerock 84.2+ 1.8 92 83.4+ 2.5 3 295.6+ 6.9 0.08

QP93-1? wholerock 63.9+ 0.5 51 60.1+ 0.7 3 376.7+ 20.7 0.08

Ageswere calculated using the following decay constants: •.e= 0.581 x 10-l0 yr-1, •.• = 4.963x 10 -lø yr -1. Isotopic interferences forthe 36 37 -4 39 37 4 40 39 OregonState University TRIGA reactor are ( Ar/ Ar)ca = 2.64x 10 '' ( Ar/ 37Ar)c a = 6.73x 10-, and( Ar/ Ar)K=1.00 x 10-3. Errorsfor bothstep age plateaus and isochrons are reported to 2•. Agesare corrected for Ar decay,half-life = 35.1days. N refersto thenumber of steps usedin theisochron calcuation. SUMS/(N-2) is thegZ-statistic from York[1969]. *Samplefrom Playa Jac6. ?Samplefrom Quepos.

Only one Quepossample (QP93-1) was consideredsuitable powderedsamples gave higher Ta values. Accordingly,Ta for analysis,although it is moderatelyaltered and very fine concentrationsfor WC-powderedsamples are not reported.Pb grained. A three-step(of six) plateaugave an age of 63.9 +_ concentrationswere reliably measured in only some of the 0.5 Ma, although the isochron age derived from the same samples. ICP-MS analysesof the basaltstandard W-1 agree steps is slightly younger (60.1 +_0.7 Ma) and the initial Ar with acceptedvalues [Sinton, 1996]. isotopic composition is considerably higher than Some of the major and trace elementdata are listed in Table atmospheric(376.7) which is an artifact of fitting only three 3, and the completedataset is reportedby Sinton [ 1996]. The points(Figure 2). The youngerages of the highertemperature basalts,diabases, and gabbrosfrom the Nicoya Peninsulaand stepscan be attributedto recoilof 39Atduring irradiation, a Jac6 have tholeiitic compositions with MgO contents problem often encounteredwith very fine grained samples. ranging from 10.7 to 4.0%. The plagiogranites have We considerthe plateauage to be the best estimateof the time andesiticto dioritic compositions(52 - 69% SiO2). The two of crystallization of the Quepos basalts,and it is consistent analyzed Quepos rocks are transitionalto alkalic basalts. with the Paleocene age of sediments intercalated with the Our major and traceelement data and thosefrom previous basalts[Schrnidt-Effing, 1979; Baurngartncret al., 1984]. studies[Wildberg, 1984;Frisch et al., 1992] from the Nicoya Peninsulaare plottedon severalvariation diagrams of MgO Mineral Chemistry versus major elements and the relatively immobile trace elements(Figure 3). Elementsthat are affectedby alteration Analyses of pyroxene and plagioclase are reported by and are present in low abundances,such as K and Rb, show Sinton [1996]. The trend of the pyroxenesis roughly similar considerablescatter and are not plotted. These plots, in to that of the Skaergaardintrusion clinopyroxenes[e.g., Deer additionto illustratingthe wide range in MgO contentfor the et al. , 1992] and is consistent with the Nicoya Peninsula Nicoya rocks, demonstratea single coherent trend for all magmas being derived from fractional crystallization of samples. A first-order interpretationof this trend is that the tholeiitic parental melts. rocks are derived from similar parent magmasby fractional crystallization. The Jac6 rocks fall along the Nicoya trend Major and Trace Element Geochemistry while the Queposbasalts show distinctly different trendsfor All the sampleswere analyzed for major elementsand Ni, some of the elements [Sinton, 1996]. Cr, Sc, V, Sr, and Y by X ray fluorescence (XRF) at The role of fractional crystallizationin the evolution of the Washington State University using a Rigaku 3370 Nicoya Peninsularocks can be assessedusing the variation spectrometer[Hooper et al. , 1993]. A subsetof the samples diagrams. The decreasingconcentrations of certain elements was analyzed for Rb, Y, Zr, Nb, Ba, Hf, Ta, Pb, Th, U, and the or element ratios with decreasingMgO content reflect the rare earthelements (REE) usinga V.G. Plasmaquadinductively crystallizationand removal of particularminerals. All of the coupled plasma mass spectrometer(ICP-MS) at Oregon State major element variationdiagrams, with the exceptionof those University [Pyle et al. , 1995]. Sampleswere powderedfor the for FeO and TiO2, show continuous,roughly linear trends. XRF and some of the ICP-MS analyses using a tungsten FeO and TiO2 both increasewith decreasingMgO until -4% carbide (WC) mill. For some of the ICP-MS analyses,cleaned MgO, at which point they decrease.This inflectionis likely rock chips were powderedin a ceramicmortar. There were no due to the appearanceof Fe-Ti oxides on the liquidus. At apparentdifferences between the ICP-MS resultsfrom sample greater MgO contents, there are no obvious inflections in the duplicatesusing the two preparationmethods, except the WC- trend and therefore the fractionating mineral assemblage SINTON ET AL.: THE NICOYA PENINSULA 15,513

Table 3. Major andTrace ElementAnalyses for the CostaRica IgneousRocks

NC93-2 NC93-3 NC93-11 NC93-13 NC93-14B NC93-18 NC93-20 NC93-26 NC93-34 JC93-1 JC93-3 QP93-1

SiO2 50.18 50.50 50.05 50.47 51.43 50.83 50.6 63.62 51.19 51.38 49.57 49.44 AI203 13.58 14.44 14.27 14.26 13.38 13.99 14.34 11.18 14.72 14.63 14.43 13.29 TiO2 1.68 1.04 1.00 1.26 1.87 1.03 0.98 1.02 1.01 1.13 1.23 2.94 FeO 13.31 10.55 10.38 10.77 13.46 10.25 9.65 13.51 9.44 9.86 11.56 11.77 MnO 0.23 0.21 0.18 0.19 0.21 0.19 0.19 0.16 0.20 0.18 0.22 0.19 CaO 10.57 12.66 12.11 12.35 9.11 11.66 12.54 4.06 12.01 12.19 10.89 7.68 MgO 6.59 8.82 8.93 8.26 5.51 8.82 8.73 0.39 8.55 8.79 8.68 7.33 K20 0.11 0.07 0.07 0.10 0.14 0.10 0.06 0.14 0.38 0.05 0.13 0.87 Na20 2.73 1.59 1.86 2.33 3.11 2.13 1.82 5.74 2.54 2.08 2.77 5.38 P205 0.13 0.07 0.07 0.09 0.14 0.07 0.07 0.29 0.07 0.08 0.09 0.27 L.O.I. 1.56 3.31 3.29 2.11 4.42 3.59 3.89 1.20 2.90 3.26 3.64 2.83 Total 100.67 103.26 102.20 102.19 102.79 102.66 102.87 101.31 103.01 103.62 103.21 102.00 Ni 48 91 84 87 18 93 106 0 100 88 97 120 Cr 70 288 200 242 8 284 209 I 296 311 202 269 $c 46 52 55 49 47 50 50 23 53 54 52 34 V 406 319 323 330 415 324 319 2 333 335 343 383 Sr 140 171 100 146 298 218 178 83 139 87 76 275 Rb 1.10 0.54 0.42 1.31 2.17 0.47 0.80 3.23 0.41 0.87 11.78 Y 34 22 15 21 34 16 15 17 18 21 36 Zr 107 58 31 43 69 49 44 39 50 55 194 Nb 5.8 3.4 3.0 4.3 7.2 3.3 3.0 3.5 4.5 3.4 19.0 Ba 20 10 11 66 41 31 7 66 12 22 159 La 3.99 2.31 1.98 3.20 5.16 2.16 2.15 2.43 2.73 2.73 13.54 Ce 11.43 6.89 5.40 8.50 13.73 5.87 5.59 6.44 7.46 7.26 34.14 Pr 1.93 1.13 0.91 1.43 2.24 0.96 0.91 1.07 1.21 1.20 4.92 Nd 9.99 5.83 4.52 7.13 11.42 4.95 4.76 5.30 5.95 6.24 22.58 Sm 3.39 1.92 1.55 2.33 3.68 1.69 1.60 1.77 1.89 2.00 6.00 Eu 1.20 0.72 0.59 0.87 1.34 0.65 0.61 0.68 0.71 0.75 1.99 Gd 4.19 2.65 2.16 3.19 4.59 2.26 2.15 2.30 2.56 2.64 6.77 Tb 0.89 0.51 0.41 0.60 0.95 0.46 0.43 0.46 0.47 0.52 1.21 Dy 5.35 3.27 2.65 3.71 5.64 2.79 2.62 2.82 3.11 3.34 6.53 Ho 1.22 0.75 0.59 0.82 1.19 0.62 0.56 0.62 0.68 0.76 1.28 Er 3.38 2.17 1.51 1.92 3.30 1.64 1.51 1.56 1.77 2.01 2.96 Tm 0.52 0.33 0.25 0.32 0.47 0.27 0.24 0.26 0.30 0.34 0.48 Yb 3.34 2.03 1.44 1.81 3.19 1.60 1.55 1.54 1.72 2.06 2.62 Lu 0.51 0.31 0.23 0.31 0.51 0.23 0.24 0.26 0.29 0.32 0.41 Hf 2.98 1.57 0.94 1.36 2.22 1.29 1.18 1.14 1.29 1.52 4.62 Ta 0.23 0.18 0.29 0.20 0.18 0.23 0.30 0.23 1.24 Pb 0.32 Th 0.38 0.27 0.11 0.18 0.31 0.14 0.16 0.20 0.28 0.26 1.38 U 0.11 0.04 0.03 0.04 0.09 0.05 0.06 0.05 0.07 0.07 0.34

Majorelements and Sc, Cr, Ni, V, andSr determinedby X rayfluorescence; allother trace elements were determined by inductivelycoupled plasma mass spectrometry.

remained relatively constant. The fractionation of NC93-100) are aphyric, and petrographicobservations are of clinopyroxene is evident from the continuous decrease in limited help in determining the fractionating mineral CaO, Sc, and Cr with decreasingMgO. The decreasein A1203 assemblage. However, the cumulate gabbro NC93-100 and Ni can be attributed to the removal of plagioclase and containslarge crystalsof clinopyroxene(Ens1), plagioclase olivine, respectively. Together, the variation diagrams are (An85) [Sinton, 1996], and what appear to be olivine consistent with the removal of clinopyroxene, plagioclase, pseudomorphs set in a basaltic matrix. This cumulate and olivine from --9% MgO to --4% MgO, at which point Fe-Ti assemblagemay representthe early crystallizing phasesof oxides join the fractionatingassemblage. the Nicoya suite. Petrographic observationsare generally consistentwith The major and trace elementdata indicatethat the Nicoya this interpretation. As previouslystated, all samples(except Peninsulaplagiogranites were derivedfrom the more primitive 15,514 SINTONET AL.: THE NICOYA PENINSULA

20 AI 2 (wt. %)

ß'16 I•

•2/ • • • 10••• I I• I TiO• (•t. %)

2 • •

6

4

2

[] 6O Sc (ppm) Cr (ppm) -. 500

50 400 .

40 _ 300 3O

. 200 20

10

o i i i i i 100

120•• Y(ppm) 400 lOO [] [] [] ß 300

Zr (ppm) •: •n [] 200

[] [] 100

0 I I I I I I I 0 0 2 4 6 8 l0 12 2 4 6 8 10 12

MgO (wt. %) MgO (wt. %) Figure3. Variationdiagrams of somemajor and trace elements versus MgO for theNicoya Peninsula rocks (solidcircles) using data compiled from this paper, Sinton [1996], Wildberg [1984], and Frisch et al. [1992]and EastPacific Rise MORB (opensquares) from 13øSto 5øN[Langrnuir, 1988].

tholeiitic basalts. This is consistent with a model described the Nicoya rocks, the MORB data show an inflection in CaO by Wildberg [1987], in which the plagiogranitesare derived andSc between6% and8% MgO,indicating the appearance of by a two-stage processof fractional crystallization and later clinopyroxene on the liquidus at that time and thus low- filter pressing of a tholeiitic magma. The latter process pressure fractionation. As discussed previously, leaves a residual Fe-rich diabase/gabbro,such as samples clinopyroxenein the Nicoyasuite appears to havebeen part NC93-7 and NC93-15. of the fractionatingmineral assemblagefrom at least -9% MgO. Becausethe primaryeffect of high-pressurefractional Comparison to Mid-Ocean Ridge Basalts crystallization is to increase the temperature at which The Nicoya Peninsula rocks are comparedto mid-ocean clinopyroxeneappears on the liquidus[e.g., Benderet al., ridge basalt (MORB) glassesfrom the East Pacific Rise from 1978; Grove et al., 1992;Langmuir et al., 1992], fractional 13øSto 5øN [Langrnuir, 1988] in Figure 3. The MORB data crystallizationof the Nicoyamagmas appear to haveoccurred show relatively coherenttrends for all elements;these trends at pressuressignificantly greater than that for MORB. are distinct from the Nicoya suite for most elements. Unlike To examine whether the differences between the MORB and SINTON ET AL.: THE NICOYA PENINSULA 15,515

100 ...... •...-z:•.'• ......

!aS•••ra•n•t•[abases

1 I I I I I I I I I I I I I La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Figure 4. C1 chondrite-normalizedrare earth element diagrams. Basalt and diabase data are from this paperand Sinton [1996]. Plagiogranitedata are from Wildberg [1984]. Chondritevalues are from McDonoughand Sun [1995].

Nicoya magmasare a resultof differencesin the depthof Summarizing the major and trace element results, the fractional crystallization (from parental melts of similar Nicoya Peninsula rocks appear to be related by fractional major elementcomposition) or of differencesin parentalmelt crystallizationfrom similar parental melts. The Jac6 basalts composition,the two data sets are comparedat 9% MgO. are compositionally similar to the Nicoya rocks while the Both data setshave comparableCaO and FeO, but A1203, Quepos basalts are distinct as well as significantly younger. which behavesincompatibly and is depletedfrom the mantle The Nicoya rocks are compositionally distinct from MORB during partial melting [e.g., Forsyth, 1992], is higher in the lavas in terms of some major and trace elements. These MORB magmas. Other incompatibleelements in the Nicoya differencesmay be partly attributedto a differencein the first suite, such as TiO2, Zr, and Y, are also slightly to appearanceof clinopyroxeneon the liquidus,possibly due to a significantly lower than in the MORB suite at 9% MgO differencein the pressureat which crystallizationoccurred, as (Figure 3). These differences cannot be attributed to well as differences in the degree of partial melting of the differences in the fractionating mineral assemblage. Rather, source mantle. The flat REE patterns of the Nicoya rocks they must reflect a depletionof the Nicoya parentalmagmas indicate a sourcethat less previous melting than the depleted relativeto MORB parentalmagmas. Sucha depletioncould be upper mantle that is the sourcefor normal MORB, and this is causedby higherdegrees of partialmelting of the mantle. confirmed by the isotopic analyses.

Rare Earth Elements Sr-Nd-Pb isotopic Ratios

Figure 4 showsthe chondrite-normalizedrare earth element Six of the freshestNicoya Peninsulasamples as well as two analyses. All show relatively flat, parallel patterns, Jac6 basaltsand one Queposbasalt were analyzedfor Sr, Nd, consistent with the Nicoya Peninsula rocks belonging to a and Pb isotopes,and the data are presentedin Table 4. Both single, cogeneticsuite. All of the Nicoya rocks(including the measuredand age-correctedinitial ratios are reported. Sr, Nd, matrix of the NC93-100 cumulate gabbro) have slightly light and Pb isotopic ratios were determinedby M. Gorring at the rare earth element(LREE) -depletedto slightly LREE-enriched Departmentof Geological Sciences,Cornell University, using chondrite-normalized REE patterns (La/Yb n = 0.77-1.21, a V.G. Sector 54 mass spectrometer[White and Patchett, Ce/Ybn = 0.85-1.23) with the absoluteabundances of the REE 1984; White and Duprd, 1986]. Approximately 500 mg increasingwith decreasingMgO. The two analyzedJac6 rocks aliquotsof rock were leachedfor 2 hoursin 6 N HC1 prior to also have relatively fiat REE patterns(LaJYb n = 0.90-1.08) dissolution. A separatealiquot of NC93-20 was not leached which, along with the variation diagrams, suggeststhat they for comparison. The leachedand unleachedaliquots of sample and the Nicoya Peninsula rocks were derived from similar NC93-20 did not give significantly different results (Table 4). parental magmas. The fiat patterns are different from the Replicate isotopic analysesfor Sr and Pb ratios of someof the LREE-depleted patterns displayed by most normal MORB. sampleswere performedon separatesample dissolutionsand The Quepos basalt has a LREE-enriched pattern (LaJYbn= show excellent reproducibility(Table 4). 3.51). This observation and the differences in the variation The 87Sr/86Srratios show considerablescatter, while the diagramsand the younger age demonstratesthat the Quepos 143Nd/144Ndratios are relatively constant for all of theCosta rocks were derived from different parental magmas than the Rica samples(Figure 5). It is possiblethat the original Sr Nicoya and Jac6 rocks. isotopic signature of the rocks has been skewed to more 15,516 SINTONET AL.: THE NICOYA PENINSULA

Table4. Sr,Nd, andPb Isotopic Analyses for BasalticRocks from the Nicoya Complex, Costa Rica

87Sr 87Sr 143Nd 143Nd 206pb 206pb 207pb 207pb 208pb 208pb Sample 86Srm 86Sri 144Ndm 144Ndi end/ 204pbm 204pbi 204pbm 204pbi 204pbrn 204Pb i

NC93-20u 0.70346 0.70345 ...... 19.23 19.16 15.56 15.56 38.87 38.81 NC93-20 0.70341 0.70340 0.512967 0.512843 6.26 19.28 19.21 15.59 15.58 38.97 38.90 NC93-18 0.70393 0.70392 0.512990 0.512863 6.65 19.28 19.02 15.58 15.57 38.93 38.71 NC93-34 0.70450 0.70442 0.512977 0.512853 6.45 19.62 19.12 15.63 15.61 39.39 38.75 NC93-13 0.70329 0.70326 0.512981 0.512860 6.59 19.34 19.15 15.63 15.62 39.14 38.84 NC93-11 0.70316 0.70314 0.513014 0.512886 7.10 19.68 19.22 15.60 15.58 39.47 38.89 NC93-3 0.70361 0.70360 0.512985 0.512863 6.65 19.33 19.20 15.60 15.60 39.08 38.81 NC93-3R 0.70346 0.70344 ...... 19.26 19.13 15.57 15.56 38.91 38.64 JC93-1 0.70347 0.70345 0.512955 0.512848 6.36 19.37 19.21 15.58 15.57 39.04 38.84 JC93-1R 0.70344 0.70342 ...... 19.34 19.19 15.56 15.55 38.98 38.78 JC93-3 0.70313 0.70310 0.512987 0.512878 6.94 19.38 19.06 15.60 15.58 39.06 38.69 JC93-3R ...... 19.40 19.09 15.60 15.59 39.12 38.75 QP93-1 0.70460 0.70450 0.512916 0.512850 6.40 19.43 19.21 15.61 15.60 39.06 38.77 QP93-1R ...... 19.40 19.18 15.58 15.57 38.94 38.66

All sampleswere acid leached prior to dissolutionexcept NC93-20u. The "m" subscript denotes measured values; the "i" subscript denotes valuesage corrected using concentrations from inductively coupled plasma mass spectrometer analyses and 90 Ma forthe Nicoya rocks, 84 Ma forthe Jaco rocks, and 64 Ma forthe Quepos sample. Replicates are denoted with "R" and are complete replicates, that is, they are analyses of separaterock sample aliquots.

radiogenic values by Sr addition, consideringthat the most Discussion altered of the analyzed samples (QP93-1) has the highest 87Sr/86Sr.However, this relationship is not clear, given that A Single Magmatic Suite sample NC93-34 from the Nicoya Peninsula is virtually unalteredand its high 87Sr/86Srratio cannotbe readily Ideally,the interpretationof the petrogenetichistory of a explained by seafloor alteration. This sample was taken near suiteof igneousrocks should be basedon the integrated a large intrudedbody of radiolarianchert, so it is possiblethat observationsfrom fieldwork,petrography, geochronology, the Sr isotopic signature has been increased by the and chemistry. In westernCosta Rica, individual flow units assimilation of pelagic sediments. Considering these are virtually impossibleto identify and stratigraphic possible contaminationeffects, Sr isotopescannot be used as relationships between the different outcrops are reliable indicators of mantle sources in this suite. unconstrained,so we mustrely moreon the geochemicaland The contaminationfrom pelagic sedimentsthat is evident agedata in interpretingthe magmatic history of the suite. We from the Sr isotopesis not apparentin the Nd and Pb isotopes havemade a casethat the rocksof the NicoyaPeninsula are (nor, for that matter, in the trace element data). This is not part of a single suiteof variablydifferentiated rocks derived unexpected for Nd, as Nd concentrationsin seawater and from similarparental magmas. On the basisof the 4øAr-39Ar pelagic sediments are very low. The range of initial studies,the magmascrystallized at 88 to 90 Ma while some 143Nd/144Ndratios for the samplesis small (0.51285- intrusivemagmatic activity occurred at 84 Ma. The previous 0.51289). The Pb isotopic ratios also show a restrictedrange designationsof a lower unit of normalmid-ocean ridge and show no evidence of a mixing trend toward pelagic generatedcrust and an upperunit partiallymade of oceanic sediment Pb isotopic compositions[Doe , 1970] (Figure 5). plateau lavas [Schmidt-Effing, 1979; Kuijpers, 1980; On 2ø6pb/2ø4pb vs. 2ø7pb/2ø4pb and 2ø6pb/2ø4pb vs. Bourgoiset al., 1984;Gursky et al., 1984;Wildberg, 1984; 2ø8pb/2ø4pbdiagrams, the age-corrected Costa Rica rocks plot Frisch et al., 1992] appearto be untenable. in a group that is more radiogenic than present day East The primitivearc seriesproposed by Wildberg[1984] and Pacific Rise basalts. Consideringthe analyticalerror, the Pb Frisch et al. [1992] was based on the observationthat some and Nd isotopic ratios of all of the analyzed samplesare not rockshave relatively low concentrationsof certain high field appreciably different, although NC93-11 and NC93-34 have strengthelements (HFSE) (e.g., Zr andTi) andsome have high 2ø6pb/2ø4pband 2ø8pb/2ø4pb ratios considerablymore 87Sr/86Srratios, which can be a characteristicof arc lavas. radiogenicthan the others. Together, the Nd and Pb isotopic However, these studies did not note that the low HFSE lavas data show that the mantle sourcesfor the Nicoya Peninsula, alsohave relatively high MgO contentsand the depletionof Jac6, and Quepos magmas were similar. These sourcesare HFSE (as well as otherincompatible elements) is an effectof enriched relative to the depleted, upper mantle source for dilution (that is, less fractionatedsamples have lowest normal MORB from the East Pacific Rise. absoluteincompatible element concentrations). As we have SINTONET AL.: THE NICOYA PENINSULA 15,517

0.5132 ¸¸ o Nicoya 0.5131 • ß -{-Quepos 0 ß& ß Galapagos ßß•, ß o E PacificMORB seawater•teration 0.5130 • • or•lagic sediment ß • -- . 0.5129 •::::::•::••::•:::::::::::::::::::::::::::::::::: :•::::::::::•:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: :::::::•::::::::::•:::: :::::::::::::::::::::::::

0.5128 0.7020 0.7025 0.7030 0.7035 0.7•0 0.7•5 0.7050 87Sr/86Sr

0.5132

0.5131

0.5130 ß ß

0.5129

0.5128 I I I 17.8 18.3 18.8 19.3 19.8 206pb/204pb

15.70

15.65 -

15.60 -

15.55 - -

15.50 --

15.45 - -

15.40 I ' I I ß 17.8 18.3 18.8 19.3 19.8 2O6pb/2O4pb

39.4

38.6

38.2 ¸ o 37.8

37.4 I I I 17.8 18.3 18.8 19.3 19.8 206pb/204pb Figure5. Sr,Nd, andPb isotopic ratio correlation plots. Typical analytical uncertainties are indicated by the sizeof thesymbols. East Pacific MORB data are from White et al. [1987]and Prinzhofer et al. [1989]. Pelagic sedimentvalues from Doe [ 1970]. C is the commonplume composition reported by Hanan and Graham[1996]. NC93-34is an unaltereddiabase that may have incorporated radiogenic Sr frompelagic sediment contamination. 15,518 SINTONET AL.:THE NICOYAPENINSULA

shown, 87Sr/86Srratios cannotbe reliably used as a Relationship to the Caribbean Oceanic Plateau geochemicaldiscriminant with the Nicoya lavas and therefore conclusionsbased on theseare suspect. It has been previously suggestedthat a portion of the We cannot definitively state that all of the pre-Tertiary Nicoya Complex is related to the thick CretaceousCaribbean rocks of the Nicoya Peninsula are from the same Late crust [e.g., Wildberg, 1984], which is considered to be the Cretaceous suite. The radiolarian chert must have been remnant of an oceanic plateau produced by initial, massive depositedon oceaniccrust that is at leastof Jurassicage (H.-J. volcanismover the Galapagosplume [Duncan and Hargraves, Gursky, personal communication, 1994). What we have 1984]. This is supportedby the fact that the -90 Ma ages of shown is that the rocks that have thus far been analyzed are the Nicoya Peninsulabasalts fall within the narrow range of not Jurassic MORB. We consider that the blocks of radiometric ages from basaltic rocks in Haiti, Cura•jao, radiolarian chert were either entrained by massive submarine Colombia,and GorgonaIsland [Sinton, 1996]. The relatively lava flows or heavily penetratedby sills and dikes. Implicit flat REE patternsand Nd and Pb isotopicratios of the Nicoya in this conclusionis that the Late Cretaceousmagmas erupted and Jac6 rocks are also similar to some of the Caribbean in an intraplate setting through preexisting oceanic crust. plateau lavas [Sinton, 1996]. Thus we conclude that the Some of the older oceanic igneous crust over which the Nicoya Peninsula rocks are one of many pieces of the radiolarian chert was deposited may be found to be Caribbeanplateau exposed by collisionaltectonics around the cornmingled with the plateau lavas described here by marginsof the Caribbean plate. continued sampling and analysis. Sr, Nd, and Pb isotopicratios are excellenttracers of mantle Eruption through preexistinglithosphere provides a ready source composition because they remain unchanged during model to explain the greater depth of crystal fractionation, partial melting and fractional crystallization. Differences in comparedto that at which normalMORB magmasare pooled. these ratios among magmascan be ascribedto the mixing of Becausethe lithosphericthicknesses at mid-oceanridges are magmas or mantle sources. The Nd and Pb isotopic ratios negligible, melts pool at shallow depths (1-2 km below indicate that the Nicoya rocks and probably the Jac6 and seafloor). In contrast,the parentalmelts of the Nicoya suite Quepos basalts were derived from similar mantle sources may have pooled beneath the pre-existing, perhaps 20-50 which are more enrichedthan the depletedupper mantle from m.y. old oceanic lithosphere, resulting in fractional which normalMORB are derived. The isotopicratios plot crystallizationat depthsexceeding 20-30 km below seafloor. within the field of presentday Galapagoslavas [White et al., However, we cannot discount the possibility that the 1993], consistentwith derivationfrom the Galapagosmantle relatively early onset of clinopyroxene crystallization was plume. In termsof mantleend-members described by isotopic due to a greater diopsidecomponent in the Nicoya parental composition [e.g., Zindler and Hart, 1986], the sourcefor the magmasrelative to thoseof normal MORB. Costa Rica rocks could be interpretedas a mixturebetween depletedupper mantle (DMM) andan enriched(HIMU) source. Relationship of Jac6 and Quepos Lavas Alternatively,the dataplot nearor withinthe generalrange of a common mantle source(FOZO, [Hart et al., 1992]; PHEM, to the Nicoya Peninsula Suite [Farleyet al., 1992];C, [Hananand Graham, 1996])(Figure 5). According to Hanan and Graham [1996], this common In the field, the Jac6 rocks are identical to the Nicoya Peninsula suite; massive, faulted, and fractured basalt flows; mantle source may be located in the transitionzone between but the one radiometricage indicatesthat magmatismoccurred the upperand lower mantle or at the core-mantleboundary. at 84 Ma. It is not clear at this point whether volcanismin this location was continuous from 88 to 84 Ma, but it is Melting Conditions possiblethat the Jac6 basalt and 84 Ma Nicoya intrusivesare related to a common magmatic phase, based on their The mantle melting conditionsthat producedthe Nicoya similarity in age and composition. The compositional lavas can be constrainedusing the compositionaldata. similarity between all of the Nicoya and Jac6 rocks indicates However,most mantle melting studiesusing geochemistry that there was no change in mantle source or conditions have focusedon mid-oceanridge environments[e.g., Klein between 90 and 84 Ma. and Langmuir , 1987; Johnson et al. , 1990] and not on The Quepos basalts are considerably younger than the oceanic plateaus. Partial melting of the mantle below mid- Nicoya and Jac6 suite and have major and trace element oceanridges appears to be controlledby polybaricmelting of compositionsthat reflect smaller degreesand perhapsdeeper a columnof mantlewith poolingof discretemelt packets from partial melting. It is possiblethat the Queposlavas represent different depths[e.g., Klein and Langmuir, 1987; Johnsonet an accretedseamount (M. Meschede,personal communication, al. , 1990; Grove et al. , 1992]. It is likely that a similar 1994) that originatedon the subductingFarallon (now Cocos) process,but operatingon a largerscale, produced the Nicoya plate. As indicated by the isotopic ratios, the Queposlavas lavas and oceanicplateaus in general. Oceanicplateaus, appear to have been derived from a mantle sourcesimilar to which have characteristicsindicative of rapidly constructed that of the Nicoya and Jac6 rocks. If this is the case, then submarineflood basaltprovinces, are probablyproduced by there must have been a long-term, active thermal anomaly in relatively high degreesof partial melting [Storeyet al., the area. An attractivecandidate is the Galapagoshotspot (see 1991; Mahoneyet al. , 1993], and so the generalmelting following section). It is likely that the Quepos lavas were regimewas significantlylarger and moreephemeral than that part of an older, now subductedextension of the aseismic of mid-ocean ridge melting. The depletion of some Cocos Ridge, which was producedby the northeastmotion of incompatibleelements in the Nicoya rocksrelative to MORB the Farallon plate over the Galapagoshotspot [Hey et al., cannotbe attributedto a moredepleted mantle source, given 1977]. thatthe REE andisotopic data indicate a relativelyundepleted SINTONET AL.:THE NICOYA PENINSULA 15,519

mantle source. Therefore the depletions are best explained by References higher degreesof partial melting of the mantle. Higher degrees of melting are also indicated by the high- Alvarado, G. E., S. Kussmaul, S. Chiesa, P.-Y. Gillot, H. Appel, G. MgO lavas that are found in the SantaElena Peninsula(low-Ti W6rner, and C. Rundle, Resumencronoestratigr•fico de las rocas igneas de Costa Rica basado en datacionesradiom6tricas, J. South basalt of Wildberg [1984]) and to the north of Playa Am. Earth Sci., 6, 151-168, 1993. Jac6/Herradura [Alvarado et al., 1997]. An 4øAr-39Ar Alvarado, G. E., P. Denyer, and C. W. Sinton, The 89 Ma Tortugal radiometricage of plagioclasefrom a lava in the latter of these komatiitic suite, Costa Rica: Implicationsfor a commongeological sections is 89.7 + 1.4 Ma [Alvarado et al., 1997], which origin in the Caribbean and Eastern Pacific region from a mantle plume, Geology,in press,1997. indicates anomalously high thermal conditions in the mantle Baumgartner,P. O., C. R. Mora, J. Butterlin, J. Sigal, G. Glagon, J. at this time. Az6ma, and J. Bourgois,Sedimentaci6n y paleografiadel Cret•cico Sinton [1996] used batch melting equationsto show that a y Cenozoico del litoral pacffico de Costa Rica, Rev. Geol. Am. potential primary magma for the Nicoya basalt suite could be Cent., 1, 57-136, 1984. produced by 30% melting of a mixed depleted and primitive Bender, J. F., F. N. Hodges, and A. E. Bence, Petrogenesisof basalts from the project FAMOUS area: Experimentalstudy from 0 to 15 mantle. Although it is possibleor even likely that fractional kbars, Earth Planet. Sci. Lett., 41,277-302, 1978. or near-fractional melting occurred during formation of the Bourgois,J., J. Azema, P.O. Baumgartner,J. Tournon,A. Desmet,and primary magmas, if the fractional melts accumulated before J. Auboin, The geologic history of the Caribbean-Cocosplate eruption, then the aggregate melt compositionswould have boundarywith special referenceto the Nicoya ophiolite complex been identical to batch melting [Shaw, 1970; Johnsonet al. , (Costa Rica) and DSDP results (Legs 67 and 84 off Guatemala): A 1990]. Such a high degree of melting was also reported by synthesis,Tectonophysics, 108, 1-32, 1984. Dalrymple,G.B., D. A. Clague,T. L. Vallier,and H. W. Menard, 4øAr- Mahoney et al. [1993] for the Ontong Java oceanic plateau 39Arage, petrology, and tectonic significance ofsome seamounts in lavas. the Gulf of Alaska, in Seamounts,Islands, and Atolls, Geophys. The pattern of QP93-1 indicates that either the mantle Monogr. Ser., vol. 43, edited by B. H. Keating, P. Fryer, R. Batiza sourcewas enriched in LREE or that garnet was residual in the and G. W. Boehlert,pp. 297-315, AGU, Washington,D.C., 1988. source mantle or both. Sinton [1996] showed that a calculated Deer, W. A., R. A. Howie, and J. Zussman,An Introduction to the Rock primary magmafor this samplecould have been generatedby Forming Minerals, 696 pp., Longman,White Plains,N.Y., 1992. Dengo, G., Tectonic igneoussequence in Costa Rica, in Petrologic mixing a 5% melt of a garnet-bearingmantle with a 10% melt Studies:A Volume to Honor A.F. Buddington,edited by A. E. J. of the spinel-bearingmantle. Engel, H. L. Jamesand B. F. Leonard,pp. 133-161,Geol. Soc. Am., Boulder, Colo., 1962. Conclusions Denyer, P., and O. Arias, Geologiadel norte de la Peninsulade Nicoya, Costa Rica, Rev. Geol. Am. Cent., 16, 69-84, 1993. Diebold, J. B., P. L. Stoffa, P. Buhl, and M. Truchan, Venezuela basin Our combined data indicate that the mafic oceanic rocks of crustalstructure, J. Geophys.Res., 86, 7901-7923, 1981. the Nicoya Peninsulabelong to a single magmatic suite, the Doe, B. R., Lead Isotopes,137 pp., Springer-Verlag,N.Y., 1970. vast majority of which formed within a shortperiod at 88-90 Donnelly, T. W., The CaribbeanCreataceous basalt association: A vast Ma. The data are not consistent with the division of the igneousprovince that includesthe Nicoya Complex of CostaRica, Nicoya Peninsula rocks into separatesuites. Two intrusive Profil, vol. 7, pp. 17-45, Univ. of Stuttgart,Stuttgart, Germany, 1994. Donnelly, T. W., et al., History and tectonic setting of Caribbean rocks (a gabbro and a ptagiogranite)are slightly youngerat magmatism,in The Caribbean Region, vol. H, edited by G. Dengo 84 Ma, suggestingintrusive activity occurredafter the initial and J. E. Case,pp. 339-374, Geo. Soc. Am., Boulder,Colo., 1990. volcanic eruptions. We cannot discountthe possibilitythat Duncan, R. A., and R. B. Hargraves, Plate tectonic evolution of the fragmentsof Jurassicto Early Cretaceousoceanic crust exist Caribbeanregion in the mantle referenceframe, in The Caribbean- in the Nicoya Peninsula and that these have not been SouthAmerican Plate Boundaryand RegionalTectonics, edited by W. E. Bonini, R. B. Hargravesand R. Shagam,pp. 81-93, Mem. identified. The coherenttrends in the variation diagrams,the Geo. Soc. Am., 16, 81-93, Boulder, Colo., 1984. similarity in the REE diagrams,and the similarity in Nd and Duncan,R. A., andR. B. Hargraves,4øAr/39Ar geochronology of Pb isotopesprovide strong evidencethat the rocks collected basementrocks from the MascarenePlateau, ChagosBank, and thus far from the Nicoya Peninsula are related to the same Maldives Ridge, Proceedings qf the Ocean Drilling Program, event. This event is most likely the formation of the Scientific Results, 115, edited by R. A. Duncan, J. Backman and L. C. Peterson,pp. 43-52, Ocean Drill. Program,College Station,Tex., Caribbean oceanic plateau, based on similarities in age and 1990. compositionsto exposed sectionsof the plateau around the Duncan,R. A., and L. G. Hogan,Radiometric dating of youngMORB Caribbean plate margins [Donnelly et al., 1990; Sinton, usingthe 4øAr-39Ar incremental heating method, Geophys. Res. 1996]. The Playa Jac6/Herradurabasalts are slightly younger Lett., 21, 1927-1930, 1994. (84 Ma) than the Nicoya Peninsulabasalts and diabasesbut are Escalante,G., The geology of southernCentral America and western Colombia,in The CaribbeanRegion, vol. H, editedby G. Dengoand compositionally identical. The Quepos basalts are J. E. Case,pp. 201-230, Geol. Soc. Am., Boulder,Colo., 1990. significantlyyounger (64 Ma) and formed by smaller degrees Farley, K. A., J. H. Natland,and H. Craig, Binary mixing of enriched of partial melting of a mantle sourcesimilar to the one from and undegassed(primitive?) mantle components (He, Sr, Nd, Pb) in which the Nicoya magmaswere derived. All of the obducted Samoanlavas, Earth Planet Sci. Lett., 111, 183-192, 1992. Forsyth, D. W., Geophysical constraints on mantle flow and melt mafic material in western Costa Rica rocks could be products generation beneath mid-ocean ridges, in Mantle Flow and Melt of the Galapagosmantle plume. Generationat Mid-Ocean Ridges,,Geophys. Monogr. Ser., vol. 71, editedby J. PhippsMorgan, D. K. Blackmanand J. M. Sinton,pp. 1- Acknowledgments. We would like to thankD. Christie,D. Graham, 65, AGU, Washington,D.C., 1992. and A. Grunderfor their helpful commentsand input and M. Gorring Frisch,W., M. Meschede,and M. Sick, Origin of the CentralAmerican for doing the isotopicanalyses. We are indebtedto K. Cameron,T.W. ophiolites: Evidence from paleomagneticresults, Geol. Soc. Am. Donnelly, and D. Elthon for helpful reviews. M. Meschedeand H.-J. Bull., 104, 1301-1314, 1992. Gursky providedvaluable advice during the planningstages of this Galli-Olivier, C., Ophiolite and island arc volcanism in Costa Rica, project. This researchhas been supportedby the National Science Geol. Soc. Am. Bull., 90, 444-452, 1979. Foundation. Gradstein,F. M., F. P. Agterberg,J. G. Ogg, J. Hardenbol,P. van Veer, 15,520 SINTONET AL.: THE NICOYA PENINSULA

J. Thierry,and Z. Huang,A Mesozoictimescale, J. Geophys.Res., Prinzhofer, A., E. Lewin, and C. J. Allegre, Stochasticmelting of the 99, 24,051-24,074, 1994. marble cake mantle: Evidence from local study of the East Pacific Grove, T. L., R. J. Kinzler, and W. B. Bryan, Fractionationof mid- Rise, Earth Planet Sci. Lett., 92, 189-206, 1989. oceanridge basalt (MORB), in Mantle Flow and Melt Generationat Pyle, D. G., D. M. Christie, J. J. Mahoney, and R. A. Duncan, Mid-Ocean Ridges,, Geophys.Monogr. Ser., vol. 71, editedby J. Geochemistryand geochronologyof ancientsoutheast Indian and PhippsMorgan, D. K. Blackmanand J. M. Sinton,pp. 281-310, southwestPacific seafloor,J. Geophys.Res., 100, 22,261-22,282, AGU, Washington,D.C., 1992. 1995. Gursky,H.-J., M. Gursky,R. Schmidt-Effing,and H. Wildberg,Karten Schmidt-Effing, R., Alter und genese des Nicoya-Komplexeseiner zur Geologie von Nordwest Costa Rica (Mittelamerika) mit ozeanischen P/ileokruste (Oberjura bis Eoz/in) im siidlichen Efi•iuten, Geol. Palaontol., 18, 173-182, 1984. Zentralamerika, Geol. Rundsch., 68, 457-494, 1979. Hanan, B. B., and D. W. Graham, Lead and helium isotopicevidence Shaw, D. M., Trace element fractionation during anatexis, Geochim. from oceanicbasalts for a commondeep sourceof mantleplumes, Cosmochim. Acta, 34, 237-243, 1970. Science, 272, 991-995, 1996. Sinton, C. W. 1996. A tale of two large igneous provinces: Hart, S. R., E. H. Hauri, L. A. Oschmann,and J. A. Whitehead,Mantle Geochronologicaland geochemicalstudies of the North Atlantic plumeentrainment: Isotopic evidence, Science, 256, 517-519,1992. Volcanic Province and the Caribbean Oceanic Plateau, Ph.D. thesis, Hey, R., G. L. Johnsonand A. Lowrie,Recent plate motionsin the Oreg. StateUniv., Corvallis, 1996. Galfipagosarea, Geol. Soc. Am. Bull., 88, 1385-1403,1977. Storey, M., J. J. Mahoney, L. W. Kroenke, and A.D. Saunders,Are Hooper,P. R., D. M. Johnson,and R. M. Conrey,Major and trace oceanicplateaus site of komatiiteformation?, Geology, 19, 376-379, element analysis of rocks and minerals by X-ray spectrometry, 1991. OpenFile Report,Geol. Dept., Wash. State Univ., Pullman,Wash., White, W. M., and 13.Dupr6, Sedimentsubduction and magmagenesis 1993. in the Lesser Antilles: Isotopic and trace element constraints,J. Hurford,A. J.,and K. Hammerschmidt,4øAr/39Ar and K-Ar dating of Geophys.Res., 91, 5927-5941, 1986. theBishop and Fish Canyon Tuffs: Calibration ages for fission-track White, W. M., and P. J. Patchett,Hf-Nd-Sr and incompatibleelement datingstandards, Chem. Geol., 58, 23-32, 1985. abundancesin islandarcs: implications for magmaorigins and crust- Johnson,K. T., H. J. B. Dick and N. Shimizu,Melting in the oceanic mantle evolution, Earth Planet Sci. Lett., 67, 167-185, 1984. uppermantle: An ion microprobestudy of diopsidesin abyssal White, W. M., A. W. Hofmann, and H. Puchelt, Isotopegeochemistry peridotites,J. Geophys.Res., 95, 2661-2678,1990. of Pacific mid-oceanridge basalt,Earth Planet Sci. Lett., 92, 4,881- Klein, E. M., and C. H. Langmuir,Global correlationsof oceanbasalt 4,893, 1987. chemistrywith axial depthand crustalthickness, J. Geophys.Res., White, W. M., A. R. Mc13irney, and R. A. Duncan, Petrology and 92, 8089-8115, 1987. geochemistryof the GalfipagosIslands: Portrait of a pathological Kuijpers,E. P., The geologichistory of theNicoya Ophiolite Complex, mantleplume, J. Geophys.Res., 98, 19,533-19,563,1993. Costa Rica and its geotectonicsignificance, Tectonophysics, 68, Wildberg, H.,Der Nicoya-Komplex, Costa Rica, Zentralamerika: 233-255, 1980. Magmatisme und Genese polygenetischenOphiolith-Komplexes, Langmuir,C., EastPacific Rise Data Synthesis(petrology database), vol. 62, 123 pp., MiinsterischesForsch. Geol. undPal/iontol., 1984. vols. I and2, JointOceanogr. Inst., Washington, D.C., 1988. York, D., Least-squaresfitting of a straightline with correlatederrors, Langmuir,C. H., E. M. Klein,and T. Plank,Petrological systematics of Earth Planet Sci. Lett., 5, 320-324, 1969. mid-oceanridge basalts:constraints on melt generationbeneath Zindler, A., and S. R. Hart, Chemical geodynamics,Annu. Rev. Earth oceanridges, in Mantle Flow and Melt Generationat Mid-Ocean Planet. Sci., 14, 493-571, 1986. Ridges,,Geophys. Monogr. Ser., vol. 71, editedby J. PhippsMorgan, D. K. Blackmanand J. M. Sinton,pp. 183-280,AGU, Washington, D.C., 1992. P. Denyer, Escuela Centroamericanade Geologia, Universidadde Mahoney,J. J., M. Storey,R. A. Duncan,K. J. Spencer,and M. Pringle, CostaRica, Apdo. 35-2060, SanJose, Costa Rica. Geochemistryand age of the OntongJava Plateau, in TheMesozoic R.A. Duncan, College of Oceanic and Atmospheric Sciences, PacificGeology, Tectonics, and Volcanism,Geophys. Monogr. Ser., OregonState University, Corvallis, OR 97331 vol. 77, editedby M. Pringle,et al., pp. 233-261,AGU, Washington, C.W. Sinton, Graduate School of Oceanography,University of D.C., 1993. Rhode Island, South Ferry Road, Narragansett, RI 02882. (email: McDonough,W. F., andS.-s. Sun, The compositionof the Earth,Chem. csinton@ gsosun l.gso.uri.edu) Geol., 120, 223-253, 1995. Pringle,M., Age progressivevolcanism in the MusiciansSeamounts: A testof the hotspothypothesis for theLate Cretaceous Pacific, in The MesozoicPaco'i'c Geology, Tectonics, and Volcanism,Geophys. Monogr. Ser., vol. 77, editedby M. Pringle,et al., pp. 187-215, (ReceivedSeptember 3, 1996; revisedFebruary 25, 1997; AGU, Washington,D.C., 1993. acceptedFebruary 26, 1997.)