Petrogenesis of olivine-phyric shergottite Larkman Nunatak 06319

Available online at www.sciencedirect.com

www.elsevier.com/locate/gca

Geochimica et Cosmochimica Acta 73 (2009) 2190–2214

Petrogenesis of olivine-phyric shergottite Larkman Nunatak06319: Implications for enriched components in martian basalts

Amit Basu Sarbadhikari a, James M.D. Day b,*, Yang Liu a, Douglas Rumble III c,Lawrence A. Taylor a

a Planetary Geosciences Institute, Department of Earth and Planetary Sciences, University of Tennessee, Knoxville, TN 37996, USAb Department of Geology, University of Maryland, College Park, MD 20742, USA

c Geophysical Laboratory, Carnegie Institution of Washington, Washington, DC 20015, USA

Received 20 August 2008; accepted in revised form 7 January 2009; available online 23 January 2009

Abstract

We report on the petrography and geochemistry of the newly discovered olivine-phyric shergottite Larkman Nunatak(LAR) 06319. The meteorite is porphyritic, consisting of megacrysts of olivine (62.5 mm in length, Fo77–52) and prismaticzoned pyroxene crystals with Wo3En71 in the cores to Wo8-30En23-45 at the rims. The groundmass is composed of finer grainedolivine (<0.25 mm, Fo62-46), Fe-rich augite and pigeonite, maskelynite and minor quantities of chromite, ulvospinel, magne-tite, ilmenite, phosphates, sulfides and glass. Oxygen fugacity estimates, derived from the olivine–pyroxene-spinel geo-barom-eter, indicate that LAR 06319 formed under more oxidizing conditions (QFM -1.7) than for depleted shergottites. The whole-rock composition of LAR 06319 is also enriched in incompatible trace elements relative to depleted shergottites, with a trace-element pattern that is nearly identical to that of olivine-phyric shergottite NWA 1068. The oxygen isotope composition ofLAR 06319 (D17O = 0.29 ±0.03) confirms its martian origin.

Olivine megacrysts in LAR 06319 are phenocrystic, with the most Mg-rich megacryst olivine being close to equilibrium withthe bulk rock. A notable feature of LAR 06319 is that its olivine megacryst grains contain abundant melt inclusions hosted withinthe forsterite cores. These early-trapped melt inclusions have similar trace element abundances and patterns to that of the whole-rock, providing powerful evidence for closed-system magmatic behavior for LAR 06319. Calculation of the parental melt traceelement composition indicates a whole-rock composition for LAR 06319 that was controlled by pigeonite and augite during theearliest stages of crystallization and by apatite in the latest stages. Crystal size distribution and spatial distribution pattern anal-yses of olivine indicate at least two different crystal populations. This is most simply interpreted as crystallization of megacrystolivine in magma conduits, followed by eruption and subsequent crystallization of groundmass olivine.

LAR 06319 shows close affinity in mineral and whole-rock chemistry to olivine-phyric shergottite, NWA 1068 and thebasaltic shergottite NWA 4468. The remarkable features of these meteorites are that they have relatively similar quantitiesof mafic minerals compared with olivine-phyric shergottites (e.g., Y-980459, Dho 019), but flat and elevated rare earth elementpatterns more consistent with the LREE-enriched basaltic shergottites (e.g., Shergotty, Los Angeles). This relationship can beinterpreted as arising from partial melting of an enriched mantle source and subsequent crystal–liquid fractionation to formthe enriched olivine-phyric and basaltic shergottites, or by assimilation of incompatible-element enriched martian crust. Thesimilarity in the composition of early-trapped melt inclusions and the whole-rock for LAR 06319 indicates that any crustalassimilation must have occurred prior to crystallization of megacryst olivine, restricting such processes to the deeper portionsof the crust. Thus, we favor LAR06319 forming from partial melting of an ‘‘enriched” and oxidized mantle reservoir, withfractional crystallization of the parent melt upon leaving the mantle.� 2009 Elsevier Ltd. All rights reserved.

0016-7037/$ - see front matter � 2009 Elsevier Ltd. All rights reserved.

doi:10.1016/j.gca.2009.01.012

* Corresponding author.E-mail address: [email protected] (J.M.D. Day).

mailto:[email protected]

Petrogenesis of martian shergottite LAR 06319 2191

1. INTRODUCTION

Shergottites are the most abundant rock type of greaterthan forty recognized Shergotty–Nakhla–Chassigny (SNC)meteorites thought to derive from the red planet (McSweenand Treiman, 1998; Meyer, 2006). SNC meteorites have adistinct oxygen isotope composition that is consistent withtheir derivation from a single, differentiated parent body(Treiman et al., 1986; Clayton and Mayeda, 1996) and theyformed under a range of oxidation states, 0 to �4 log unitsbelow the quartz–fayalite–magnetite (QFM) buffer-curve(e.g., Herd et al., 2001a; Wadhwa, 2001; Herd, 2003; Good-rich et al., 2003). In addition, the crystallization ages ofSNC meteorites, including the martian orthopyroxenite Al-lan Hills (ALH or ALHA) 84001, span a notably largerange (165 Ma to 4.5 Ga: Nyquist et al., 2001; Bridgesand Warren, 2006) and they contain highly distinctivetrapped martian atmospheric gases – perhaps the most per-suasive evidence for a martian origin (Bogard and Johnson,1983; Treiman et al., 2000).

The shergottites are subdivided into three varieties:basaltic, lherzolitic and olivine-phyric (Goodrich, 2003).Basaltic shergottites (e.g., Shergotty, Zagami, Los Angelesand Queen Alexandra Range (QUE) 94201) are pyroxene-and plagioclase-bearing basalts, believed to be products ofnear-surface crystallization in flows or conduits. Lherzolit-ic shergottites (e.g., ALHA 77005, Lewis Cliff (LEW)88516, Yamato (Y)-793605) are considered to be olivine–pyroxene cumulates. Olivine-phyric shergottites (Y-980459, Sayh al Uhaymir (SaU) 005, Dar al Gani(DaG) 476, Elephant Moraine (EETA) 79001, Dhofar019 and North West Africa (NWA) 1068) belong to a dis-tinct group of shergottites characterized by the presence ofmegacrysts of olivine and minor amounts of orthopyrox-ene in a fine-grained groundmass of olivine, pyroxeneand maskelynite. There has been considerable disagree-ment about the petrogenesis of this sub-group. Onehypothesis is that they are the product of mixing of twoend-members, namely basaltic and lherzolitic shergottites(Boctor et al., 1998; Mittlefehldt et al., 1999; Barratet al., 2002a). The alternative explanation for their originis that the olivine represent re-entrained cumulates (Usuiet al., 2008). There is also controversy regarding the causeof enriched and depleted signatures in shergottites, withtwo competing hypotheses of crustal assimilation (Jones,1989; Wadhwa, 2001; Herd et al., 2002), or sampling ofheterogeneous mantle (Herd, 2003; Borg and Draper,2003; Symes et al., 2008) offered to explain Sr–Nd isotopecompositions of martian meteorites, which greatly exceedthe isotopic variation seen in terrestrial basalts.

In this study, we present a detailed petrologic and geo-chemical study of LAR 06319 (mass upon discov-ery = 78.6 g), a new olivine-phyric shergottite from theLarkman Nunatak Icefield region of Antarctica. When itwas discovered, the exterior of the meteorite had a darkbrown to black fusion crust with a fine-grained wrinkledtexture and a slight sheen that covered 60% of the meteor-ite, and an interior that was found to be composed of a fine-grained gray and black groundmass (Antarctic MeteoriteNewsletter, 2007). We have performed petrographic analy-

sis on two polished sections of LAR 06319 as well as whole-rock geochemical analysis of a �1.7 g rock chip, all ofwhich were provided by the Meteorite Working Group(MWG). Our analysis of this meteorite shows it to be anolivine-phyric shergottite with close affinities to NWA1068 and NWA 4468. We show that detailed petrology ofLAR 06319 provides insight into the nature of the olivinemegacryst population in olivine-phyric shergottites, as wellas the nature of the incompatible-element enriched compo-nent in martian basaltic magmas.

2. METHODS

2.1. Quantitative petrology and mineral chemistry

Two polished sections of LAR 06319 (28 and 37) wereobtained from the Meteorite Working Group for texturaland mineralogical characterization (Fig. 1). Surface areasfor these two polished sections were �0.54 cm2 (LAR06319, 37 – thin-section) and �1.64 cm2 (LAR 06319, 28– thick section). Analyses of major-element compositionsof minerals were made using a CAMECA SX-50 electronmicroprobe at the University of Tennessee. For pyroxenesand oxides, operating conditions involved a 15 keV acceler-ating voltage, a 20 nA beam current and a 2 lm beamdiameter. For plagioclase (maskelynite), phosphates, andglass, a defocused beam (5 lm) and a lower beam current(10 nA) was employed to minimize loss of volatile elements(K, Na and Cl). PAP matrix correction protocols were car-ried out for all analyses. Natural and synthetic silicateswere used as standards and for calibration of the EMPon a session-by-session basis. The modal mineralogy (invol.%) was measured using the X-ray digital imaging proce-dure of Taylor et al. (1996). This technique is an automatedpoint-counting method utilizing the Feature Scan Phase

Distribution software package of an Oxford instrument en-ergy-dispersive spectrometer interfaced to the Cameca elec-tron microprobe.

Crystal size distribution (CSD) and spatial distributionpattern (SDP) analyses were performed for olivine crystalsusing polished thin-section LAR 06319, 37 (Fig. 1). A ma-jor limitation of quantitative textural analysis for LAR06319 is the available surface area of allocated polished sec-tions to the relative size of the crystals, resulting in analysisof only 239 individual olivine grains. The protocol we usedhas been outlined in detail in Day et al. (2006a) and Dayand Taylor (2007). Briefly, this involves the individual iden-tification and tracing of crystals with the aid of a micro-scope, enabling touching crystals and glomerocrysts to beidentified and separated. The area, length, width and orien-tation of individual olivine grains and the phase abundance(mode %) of olivine were measured using image analysissoftware (ImageJ v.1.40) as outlined in Jerram et al.(2003). Since the above measurements were in 2D, correc-tions to 3D were performed using the CSDslice softwareof Morgan and Jerram (2006). Size distribution of the oliv-ine crystals was then calculated using the CSDcorrections(v.1.37) software of Higgins (2000). SDP was applied toquantitatively assess the spatial packing arrangement ofthe constituent olivine grains of LAR 06319, 37. The SDP

2192 A. Basu Sarbadhikari et al. / Geochimica et Cosmochimica Acta 73 (2009) 2190–2214

method involves a statistical pattern recognition approachand calculates the value of a key parameter, R, which isthe ratio between the mean nearest-neighbor distance forthe observed sample and the predicted value for a randomdistribution (Jerram et al., 1996, 2003). Mean nearest-neighbor distance is defined as Rr/N, where r is the near-est-neighbor distance measured by XY grain-centercoordinates, and N as the total number of individual grainsmeasured. The value of R reflects the ordering or clusteringof touching or non-touching crystals, compared to the ran-dom distribution of constituent crystals in a rock. PlottingR with the porosity (% groundmass) of a rock produces aline of reference (random sphere distribution line, RSDL)between clustered and ordered crystals (Jerram et al., 1996).

2.2. Laser ablation ICP-MS analysis

Concentrations of minor and trace elements weredetermined in minerals and melt inclusions using a NewWave Research UP213 (213 nm) laser-ablation systemcoupled to a ThermoFinnigan Element 2 ICP-MS atthe University of Maryland. Olivine, pyroxene and mask-elynite were analyzed using individual spots with a 55–80 lm-diameter and a laser repetition rate of 7 Hz anda photon fluence of 2–2.5 J/cm2. Phosphates weremeasured with a 25 lm-diameter beam size, and melt-inclusion phases (either crystallized products in largemelt-inclusions or whole, exposed inclusions) were mea-sured with 55–80 lm-diameter beam sizes. Th/ThO pro-duction was <0.07% for the analytical session. Theablation analysis took place in a 3 cm3 cell with a heliumatmosphere. The cell was flushed with a He gas flow of1 L min�1 to enhance production of fine aerosols andwas mixed with an Ar carrier gas flow of 0.4 L min�1 be-fore reaching the torch. Each analysis consisted of �60 sof data collection. Backgrounds on the ICP-MS samplegas were collected for approximately 20 s followed byapproximately 40 s of laser ablation of the sample. Fusedbeads (Section 2.3) were measured using the same proto-col as for minerals with a 100 lm-diameter beam size,100 s data collection, and raster paths across the beads.Washout time between spots was >2 min. Data were col-lected in time-resolved mode so that effects of inclusions,mineral zoning and possible penetration of the laser beamto underlying phases could be evaluated for each analy-sis. Plots of counts per second vs. time were examinedfor each analysis, and integration intervals for the gasbackground and the sample analysis were selected manu-ally using LAM-TRACE software. Each LA-ICP-MSanalysis was normalized to a major-element oxide, mea-sured previously by electron microprobe, as an internalstandard to account for variable ablation yield (Fe forolivine and Ca for other silicates and glass). This internalnormalization is supported by the electron microprobedata, which allowed no more than �2% relative differencein Fe or Ca content over a 80 lm spot and the LA-ICP-MS time-resolved data showing that the ablated volumeswere generally homogeneous. For all data, the NIST 610glass standard was used for calibration of relative ele-ment sensitivities and the BCR-2g glass standard was

analyzed as an unknown to assess accuracy and repro-ducibility (GeoReM preferred values). Reproducibilityof the BCR-2g glass standard run four times during eachanalytical session was better than 5% (RSD) for all traceand minor elements in minerals and glasses, with theexception of Cr, Ge and W for plagioclase, Ni for olivineand W, Yb and Tm for pyroxene; reproducibility forthese elements was better than 10%.

2.3. Whole-rock analysis

A �1.7 g sub-sample of LAR 06319, 14 was provided bythe MWG for whole-rock geochemical study. The samplewas crushed (under better than class 1000 clean lab condi-tions) in a high-purity agate mortar and pestle used purelyfor the desegregation of martian materials. The agate wascarefully cleaned and abraded between each use and thiscrushing procedure contributes insignificantly to the blank.Major- and trace-element compositions were performed ontwo beads of 30–40 mg that were made from the larger ali-quot of well homogenized whole-rock powder. The fusedbeads were made using a Mo-strip heater, in a nitrogenatmosphere. Major-element concentrations in fused beadswere analyzed by the electron microprobe with protocolsfor glass analyses. Trace-element concentrations were mea-sured via LA-ICP-MS analysis of the beads as outlined inSection 2.2 and reproducibility was better than 5% for allelements listed.

2.4. Whole-rock oxygen isotope analysis

Oxygen isotope analyses were performed at the Geo-physical Laboratory, Carnegie Institution of Washington.They are reported in d18O, d17O (the permil [&] deviationof 18O/16O or 17O/16O in an unknown from the interna-tional standard [std], V-SMOW, given by the relationship:dXOn = [1000 � ((XO/16On)/(XO/16Ostd) � 1)] (where X =18O or 17O), and D17O notation, which represents devia-tions from the terrestrial fractionation line (k = 0.526;D17O = 1000 ln((d17O/1000) + 1) � 0.526 � 1000 ln((d18O/1000) + 1); after Rumble et al., 2007). Two sub-chips ofLAR 06319, 14 were crushed under ethanol in a boron ni-tride mortar and pestle, ultra-sonicated in dilute hydrochlo-ric acid, and magnetic material was removed with a handmagnet. Samples were loaded in a Sharp-type reactionchamber (Sharp, 1990). Successive, repeated blanks withBrF5 and vacuum pumping were carried out for 12 h untilthere was <150 lm non-condensable gas pressure remain-ing after a blank run. There was no detectable contamina-tion from CF4. Quantitative release of oxygen byfluorination reaction was performed by heating samplesindividually with a CO2 laser in the presence of BrF5. Stan-dardization of delta values was achieved by comparisonwith the Gore Mountain garnet standard, USNM 107144,analyzed during every analytical session.

3. RESULTS

The modal mineralogy, CSD and SDP results, represen-tative major-element compositions of main- and accessory-


mineral phases, trace element concentrations of mineralsand whole-rock compositions of LAR 06319 are presentedin Tables 1–5. The complete LA-ICP-MS data set and oxy-gen isotope compositions for splits of LAR 06319 are pre-sented in electronic annex Tables 1–5 of the online data set.

3.1. General petrography and shock features

LAR 06319 can be classified as a porphyritic olivinebasalt that contains olivine megacrysts (62.5 mm), similarto other olivine-phyric shergottites. LAR 06319 also con-tains pyroxene laths that are up to 1.0 mm in length(Fig. 1). Groundmass material consists of relatively fine-grained (<0.25 mm) olivine, pyroxene, maskelynite, phos-phates, oxides and sulfides. Impact-melt glass is also pres-ent. Portions of LAR 06319 that we investigated wereinterior samples and did not possess any fusion crustmaterial.

Shock features are pervasive in both sections studied.Plagioclase grains have been completely converted intomaskelynite. Megacrystic olivine, pyroxenes and chromitegrains are pervaded by fractures, as well as kink-bandstructures and planar deformation within olivine crystals(Fig. 2a). Significantly, a prominent shock-induced meltvein cuts through LAR 06319, 37 (Fig. 1). These featuresindicate LAR 06319 was involved in a major impact shockevent (S5), equivalent to �30–35 GPa of shock pressure(Stoffler et al., 1991; Fritz et al., 2005). This range ofshock pressure is similar to most other olivine-phyricand basaltic shergottites, but not to lherzolitic shergotites,which were affected by somewhat higher shock pressuresin the range of �40–45 GPa (Nyquist et al., 2001; Fritzet al., 2005).

A notable feature of LAR 06319 is the abundant melt-inclusions trapped within the megacryst olivine (Fig. 2b).These melt inclusions range in size (<50 to <400 lm) andshow variable degrees of crystallization with a spectrumof minerals including pyroxene, maskelynite, glass, aSiO2-rich phase, chromite, phosphate, sulfide and, in atleast one case, amphibole.

Table 1Modal abundance (vol.%) of minerals in LAR 06319 and other shergott

LAR 06319

Thin-section Thick-section Weighavera

Olivine 22.3 25.1 24.4Orthopyroxene 4.3 4.1 4.1Pigeonite 27.8 27.7 27.7Augite 22.9 22.0 22.2Maskelynite 19.1 17.4 17.8Minor phases 3.6 3.7 3.7

Phosphate 2.1 2.1 2.1Spinel* 0.7 0.9 0.9Ilmenite 0.5 0.4 0.4Sulfides 0.3 0.3 0.3

The ranges for other shergottites are from Taylor et al. (2002), Bridges atr. = trace.* Contains chromite1, chromite2 and titanomagnetite. See text for detail

3.2. Olivine

Olivine (�24 vol.%) in LAR 06319 occur as largegrains (megacrysts; 0.5–2.5 mm), or as smaller grains(<250 lm) within the groundmass. Olivine megacrystshave generally euhedral to subhedral habit with a few pos-sessing a ‘hopper’-like morphology (e.g., Fig. 1). Almostall of the olivine megacrysts investigated contained meltinclusions. A distinctive feature of the LAR 06319 olivinesis that they are brown in color, as also described by Mit-tlefehldt and Herrin (2008), and Kurihara et al. (2008).Similar brown olivines have also been reported in oliv-ine-phyric and lherzolitic shergottites ALHA 77005 (McS-ween et al., 1979; Ostertag et al., 1984; Treiman et al.,1994); Dhofar 019 (Taylor et al., 2002); LEW 88516 (Trei-man et al., 1994); SaU 150 (Walton et al., 2005); Y-793605(Mikouchi and Miyamoto, 1997); and dunite NWA 2737(Mikouchi, 2005; Beck et al., 2006; Treiman et al., 2006,2007; van de Moortele et al., 2007). The cause of thebrown coloration in these olivines is unclear but has pre-viously been attributed to conversion of Fe2+ to Fe3+ bystructural H (e.g., Dyar, 2003), to shock (e.g., Ostertaget al., 1984), or by the change from Fe2+ to Fe–Ni metal(Reynard et al., 2006; Treiman et al., 2006, 2007; Kuriharaet al., 2008). Broadly speaking, the megacrystic olivinegrains become more pronounced in color towards theirrims; all groundmass olivines are reddish-brown in colormaking them readily distinguishable in thin-section underplane-polarized light. Planar deformation or lineationswere observed in brown olivine grains (Fig. 2c). The coresof all megacrystic olivines are strained, similar to the lat-tice strain described by Treiman et al. (2007) and thesecores also contain micron to submicron chromite grains.Kurihara et al. (2008) found micron inclusions of symp-lectitic exsolution of chromites and augite in all olivinesin LAR 06139.

The total range of olivine compositions is Fo77–46, withmore-fayalitic compositions measured in the rims of largeolivine and in the groundmass olivine grains (Figs. 3a–c).Megacrystic olivine ranges from Fo77 in the core to Fo52

ites.

Shergottites

tedge

Basalticshergottites

Olivine-phyricshergottites

Lherzoliticshergottites

0.3–4 7–29 35–60– 1.5–7 tr.�2538–80 48–65 4–60

10–54 12–26 4–162–10 1–3 tr.�4– – –– – –– – –– – –

nd Warren (2006), Meyer (2006) and references therein.

s.

Table 2Representative electron microprobe analyses of minerals in LAR 06319.

Phases Olivine Pyroxene Maskelynite

Sk-BO Sk-RBO Megacryst Matrix I^Px Prismatic MI (Al-rich) Matrix MI

Core Mantle Rim Matrix LCP HCP

Core Rim Core Rim Aug Pig

SiO2 38.0 34.2 37.8 34.5 35.4 34.5 34.2 54.5 50.5 48.1 49.2 48.3 49.2 47.6 56.1 56.3TiO2 – – – – – – – 0.11 0.36 0.69 0.81 0.63 0.83 1.42 – –Al2O3 – – – – – – – 0.53 1.95 0.89 0.69 0.35 7.35 6.03 26.9 26.4Cr2O3 0.04 0.03 0.03 0.04 0.04 <0.03 0.04 0.49 0.69 <0.03 0.10 <0.03 0.27 0.30 – –FeO 24.9 40.0 23.5 38.6 35.7 41.2 41.6 16.2 14.9 27.1 31.2 36.1 15.9 12.5 0.65 0.87MnO 0.51 0.69 0.44 0.76 0.64 0.78 0.76 0.53 0.56 0.76 0.81 0.91 0.51 0.41 <0.03 <0.03MgO 36.3 24.0 37.4 24.8 27.2 22.9 23.6 26.0 15.3 9.67 11.6 10.1 23.4 12.2 0.10 0.16CaO 0.14 0.17 0.12 0.13 0.14 0.13 0.13 2.14 14.5 11.0 5.09 2.54 1.34 18.1 9.79 9.26Na2O – – – – – – – 0.06 0.19 0.12 0.03 0.04 <0.03 0.19 5.49 5.81K2O – – – – – – – – – – – – – – 0.64 0.45

Total 99.9 99.1 99.3 98.8 99.1 99.5 100.3 100.6 99.0 98.3 99.5 99.0 98.8 98.8 99.7 99.2

Oxygen 4 4 4 4 4 4 4 6 6 6 6 6 6 6 8 8Si 1.003 0.989 0.996 0.992 0.998 0.997 0.983 1.971 1.926 1.944 1.961 1.972 1.811 1.820 2.539 2.554Ti – – – – – – – 0.003 0.010 0.021 0.024 0.019 0.023 0.041 – –Al – – – – – – – 0.023 0.088 0.042 0.032 0.017 0.319 0.272 1.434 1.410Cr 0.001 0.001 0.001 0.001 0.001 0.000 0.001 0.014 0.021 0.000 0.003 0.000 0.008 0.009 – –Fe 0.549 0.965 0.519 0.928 0.841 0.995 0.999 0.491 0.474 0.915 1.040 1.234 0.489 0.400 0.025 0.033Mn 0.011 0.017 0.01 0.018 0.015 0.019 0.018 0.016 0.018 0.026 0.027 0.032 0.016 0.013 0.000 0.000Mg 1.426 1.032 1.472 1.062 1.141 0.987 1.009 1.404 0.872 0.583 0.691 0.612 1.283 0.695 0.007 0.011Ca 0.004 0.005 0.003 0.004 0.004 0.004 0.004 0.083 0.593 0.478 0.217 0.111 0.053 0.743 0.475 0.450Na – – – – – – – 0.004 0.014 0.009 0.002 0.003 0.000 0.014 0.482 0.511K – – – – – – – – – – – – – – 0.037 0.026

Total 2.994 3.009 3.001 3.005 3.000 3.002 3.014 4.009 4.016 4.018 3.997 4.000 4.002 4.007 4.999 4.995

Mg# 72.2 51.7 73.9 53.4 57.6 49.8 50.2 74.1 64.8 38.9 39.9 33.1 72.4 63.5 – –Wo – – – – – – – 4.20 30.6 24.2 11.1 5.67 2.90 40.4 – –En – – – – – – – 71.0 45.0 29.5 35.5 31.3 70.3 37.8 – –Fs – – – – – – – 24.8 24.4 46.3 53.4 63.1 26.8 21.8 – –Or – – – – – – – – – – – – – – 3.72 2.63Ab – – – – – – – – – – – – – – 48.5 51.8An – – – – – – – – – – – – – – 47.8 45.6

2194A

.B

asuS

arbad

hik

ariet

al./G

eoch

imica

etC

osm

och

imica

Acta

73(2009)

2190–2214

Phases Spinel Ilm Phosphate Silica* Glass

Chr1 Chr2 TM AAChr2 A^TM Mrl Apatite MI MI MP MV

I^BO Matrix I^RBO Matrix

SiO2 0.72 0.70 0.86 0.31 0.83 0.55 0.04 0.44 0.43 93.1 63.1 54.4 48.0TiO2 0.58 0.62 7.60 9.11 18.9 50.2 48.7 – – 0.36 0.39 0.53 1.01Al2O3 5.78 5.67 5.47 6.30 1.39 0.04 0.11 0.07 0.32 3.12 20.0 17.1 11.3Cr2O3 56.9 56.1 36.7 31.0 0.10 0.94 0.04 – – <0.03 0.06 0.07 0.43FeO 31.0 31.6 44.4 48.5 74.4 43.6 47.8 1.29 0.73 0.49 3.52 9.60 15.4MnO 0.50 0.48 0.33 0.49 0.43 0.72 0.69 0.09 0.07 <0.03 0.11 0.27 0.40MgO 3.72 3.62 2.29 2.21 0.69 3.20 0.77 3.36 0.21 <0.03 2.69 4.71 9.86CaO <0.03 <0.03 0.28 0.09 0.19 0.07 0.27 45.3 52.2 0.55 5.40 9.13 9.60Na2O – – – – – – – 2.38 0.05 0.23 1.80 2.98 1.97K2O – – – – – – – <0.03 <0.03 0.65 3.56 0.34 0.22V2O3 0.33 0.34 0.63 0.67 0.30 – – – – – – – –P2O5 – – – – – – – 47.1 43.3 0.01 0.37 0.27 1.40SO2 – – – – – – – – – 0.05 0.08 0.37 0.57H2O – – – – – – – – 0.83 – – – –F – – – – – – – – 0.64 – – – –Cl – – – – – – – – 1.35 – – – –

Total 99.5 99.1 98.6 98.7 97.2 99.3 98.4 100.0 100.1 98.6 101.1 99.8 100.2O = F, Cl 0.57

Oxygen 4 4 4 4 4 3 3 8 25Si 0.025 0.025 0.031 0.011 0.031 0.014 0.001 0.022 0.072Ti 0.015 0.016 0.205 0.245 0.527 0.932 0.929 – –Al 0.240 0.237 0.231 0.265 0.061 0.001 0.003 0.004 0.062Cr 1.586 1.571 1.040 0.875 0.003 0.018 0.001 – –Fe – – – – – – – 0.055 0.101Fe3+� 0.084 0.100 0.239 0.329 0.812 0.089 0.136 – –Fe2+� 0.830 0.835 1.093 1.120 1.499 0.811 0.879 – –Mn 0.015 0.014 0.010 0.015 0.014 0.015 0.015 0.004 0.01Mg 0.196 0.191 0.122 0.118 0.038 0.118 0.029 0.254 0.053Ca 0.000 0.000 0.011 0.003 0.008 0.002 0.007 2.462 9.314Na – – – – – – – 0.234 0.017K – – – – – – – 0.000 0.000V 0.009 0.010 0.018 0.019 0.009 – – –P – – – – – – – 2.023 6.110S

Total 3.000 2.999 3.000 3.000 3.002 2.000 2.000 5.058 15.739(continued on next page)

Petro

genesis

of

martian

shergo

ttiteL

AR

063192195

Phases Spinel Ilm Phosphate Silica* Glass

Chr1 Chr2 TM AAChr2 A^TM Mrl Apatite MI MI MP MV

I^BO Matrix I^RBO Matrix

Mg# 19.1 18.6 10.1 9.51 2.47 12.7 3.21 – –Cr# 86.9 86.9 81.8 76.7 4.64 – – – –Usp 1.58 1.70 21.3 25.0 54.6 – – – –Spl 12.4 12.2 12.0 13.5 3.16 – – – –Chr 81.7 80.9 54.2 44.7 0.15 – – – –Mag 4.33 5.17 12.5 16.8 42.1 – – – –

Phases Troilite Troilite Pyrrhotite PyriteA^Px–Mask A^Px–Mask–Ilm

MI Matrix Matrix Matrix

Chemical composition (wt%)

Fe 62.5 62.4 62.0 52.1Ni 0.05 0.54 1.30 1.66Co <0.03 0.03 0.26 0.28S 36.4 36.5 37.3 47.4

Total 99.0 99.5 100.9 101.4

Atomic formula

Fe 49.54 49.34 48.29 38.19Ni 0.040 0.410 0.966 1.158Co 0.000 0.020 0.190 0.192S 50.30 50.23 50.55 60.45Total 99.88 100.0 100.0 99.99Fe/S 0.985 0.982 0.955 0.632

Abbreviations: Sk = hopper olivine; BO = brown olivine; RBO = reddish-brown olivine; A^Px = inclusion within pyroxene; Aug = augite; Pig.= pigeonite; MI = melt-inclusion; LCP = low-Capyroxene; HCP = high-Ca pyroxene; Mg# = 100 * Mg/(Mg + Fe).Chr1 = chromite1; Chr2 = chromite2; TM = titanomagnetite; l^BO = inclusion within brown olivine; 1^RBO = inclusion within reddish-brown olivine; Ilm = ilmenite; A^Chr2 = associated withchromite2; A^TM = associated with titanomagnetite; Mrl = merrillite; * = silica-rich phase; MI = melt-inclusion; MP = melt pocket; MV = melt vein; � = calculated from stoichiometry;Mg# = 100 * Mg/(Mg + Fe2+); Cr# = 100 * Cr/(Cr + Al); Usp = 2Ti/(2Ti + Al + Cr + Fe3+); Spl = Al/(2Ti + Al + Cr + Fe3+); Chr = Cr/(2Ti + Al + Cr + Fe3+); Mag = Fe3+/(2Ti + Al + Cr + Fe3+).A^Px–Mask = associated with pyroxene, and maskelynite; A^Px–Mask–Ilm = associated with pyroxene, maskelynite and ilmenite.

Table 2 (continued)

2196A

.B

asuS

arbad

hik

ariet

al./G

eoch

imica

etC

osm

och

imica

Acta

73(2009)

2190–2214

Table 3Representative trace element abundances of mineral phases in LAR 06319.

Olivine Pyroxene Maskelynite Merrillite Melt incl.

Core Rim Ground mass Pig. core Aug. core Pig. rim Aug. rim P2O5-rich P2O5-Poor

P2O5 (wt%) 1.52 0.06Sc (ppm) 6.9 8.6 7.6 48.6 31.9 34.3 61.1 22.5 44.5 201 35.8V 33.6 12.2 29.3 283 210 194 130 64.9 50.4 233Cr 1166 227 1881 3710 3363 1678 574 222 90.0 5248Co 111 109 109Ni 671 407 358Cu 11.6 2.9 0.4Zn 93.2 152 153Ga 0.6 1.0 1.7 2.86 2.59 4.68 9.93 42.3 59.2 14.4Ge 0.8 0.3 0.4 2.00 2.83 0.89 3.06 1.56 3.63 2.37Rb <0.11 0.11 1.85 74.0 11.7 17.8 75.6Sr 76Sr <2.72 <1.22 7.59 5.33 135 137 123 35.7Y 2.59 1.33 3.92 18.9 23.0 4.10 718 13.5Zr 2.85 0.86 7.42 67.8 122 78.9 215 46.1Nb 0.07 0.02 0.28 0.69 9.62 0.67 4.79 2.71Ba 0.16 0.17 4.65 20.3 72.3 99.7 18.8La <0.01 0.02 0.21 2.56 3.94 0.13 127 1.90Ce 0.07 0.06 0.59 6.39 10.5 0.35 273 5.02Pr 0.02 0.01 0.10 0.77 1.20 0.04 50.6 0.58Nd 0.14 0.05 0.54 3.68 6.02 0.19 250 3.15Sm 0.13 <0.05 0.24 1.57 2.26 0.12 86.9 1.23Eu 0.05 0.02 0.12 0.58 1.36 0.46 18.3 0.45Gd 0.13 0.10 0.40 2.26 3.00 0.28 113 1.87Tb 0.05 0.02 0.11 0.46 0.65 0.06 20.2 0.36Dy 0.41 0.17 0.59 3.55 4.78 0.74 125 2.48Ho 0.08 0.04 0.16 0.73 0.96 0.14 26.8 0.55Er 0.31 0.15 0.47 2.01 2.68 0.60 67.5 1.39Tm 0.05 0.03 0.06 0.28 0.42 0.12 8.82 0.20Yb 0.41 0.19 0.44 2.26 3.21 0.81 57.2 1.20Lu 0.04 0.03 0.08 0.35 0.43 0.11 7.19 0.15Hf 0.12 0.03 0.24 2.71 3.90 3.85 <4.92 1.58Ta <0.01 <0.01 0.01 0.06 0.44 0.08 0.19 0.12W <0.02 0.02 0.05 0.39 1.25 0.07 0.33Pb 0.04 0.02 0.21 0.33 1.14 0.79 3.58 0.94Th <0.01 <0.01 0.03 0.41 1.27 <0.05 116 0.33U <0.01 0.01 0.01 0.06 0.32 0.01 2.40 0.10


in the rim and groundmass olivine range in compositionfrom Fo62 to Fo46 (Fig. 3c). The compositional range formegacrystic and groundmass olivines is similar to that de-scribed for LAR 06319 by Mittlefehldt and Herrin (2008).Olivine included within pyroxene grains also has low Fo-contents. Minimum cooling-rate estimates, based on Fozonation in olivine megacrysts, yield values of �0.2–0.5 �C/h, using the kinetic-modeling method of Tayloret al. (1977). Based on experimental results on MORB, per-idotites, lunar and eucritic basalt compositions by Donald-son (1990) and Lofgren (1983), hopper olivines are formedat a lower degree of supercooling (<60 �C) and a coolingrate of 2–75 �C/h for lunar basalts and at 0.1 MPa and<14 �C/h for eucritic basalts under hydrous conditions at500 MPa. By approximation, the few hopper olivine grainsin LAR 06319 likely formed at a cooling rate consistentwith that estimated from the Mg–Fe profile of megacrysticolivine (0.2–0.5 �C/h).

The olivine compositions for LAR 06319 are compara-ble with other olivine-phyric shergottites which generally

have more Fe-rich olivines than lherzolitic shergottites(Fig. 3c). The notable exception is Y-980459. Y-980459has the most forsteritic olivine compositions of all shergot-tites (Greshake et al., 2004; Usui et al., 2008). Olivine pop-ulations in olivine-phyric shergottites SaU 005 and DaG476 have restricted compositional ranges, but fall withinthe range of other olivine-phyric shergottites (e.g., EETA79001; Dhofar 019; NWA 1068). Conversely, olivines ofthe basaltic shergottites and nakhlites are fayalitic com-pared with those of the olivine-phyric shergottites(Fig. 3c).

Systematics of Ni–Co (Fig. 4a) and V–Cr (Fig. 4b) inolivines show a continuum from the megacryst corethrough to the rims and into the groundmass olivine (Table3). Like the other olivine-phyric and lherzolitic shergottites,the Ni–Co trend in LAR 06319 olivines has a relatively flatslope, yet Co contents are lower in LAR 06319 olivines thanfor other olivine-phyric and lherzolitic shergottites. The flatCo profile is consistent with the concept of changing parti-tion coefficients of Co in olivine during crystallization,

Table 4Chemical composition of LAR 06319 in comparison with other shergottites.

LAR 06319 NWA 1068� EETA 79001A� Dhofar 019# DAG 476$ SAU 005@ Y 980459 Basaltic shergottites Lherzolitic shergottites

SiO2 46.7 45.8 48.6–51.1 48.4–49.2 45.8 47.2 48.7–49.9 43.5–51.4 43.1–45.4TiO2 0.68 0.77 0.64–0.83 0.49–0.70 0.39 0.42 0.48–0.54 0.70–2.00 0.35–0.44Al2O3 6.00 5.75 5.37–6.52 6.40–7.01 4.37 4.53 5.17–6.00 5.67–15.2 2.32–3.45Cr2O3 0.56 0.63 0.53–0.66 0.50–0.61 0.78 0.78 0.84–1.03FeO* 20.4 20.5 17.6–24.3 18.4–19.9 16.1 18.3 15.8–18.2 14.1–24.2 19.7–20.9MnO 0.48 0.46 0.47–0.54 0.48–0.49 0.45 0.46 0.43–0.52 0.20–0.63 0.44–0.51MgO 15.8 16.5 13.3–16.9 14.6 19.4 20.5 18.1–19.6 3.53–12.1 23.7–27.7CaO 6.46 7.91 7.05–7.98 7.28–9.42 7.66 5.74 5.88–7.20 8.81–11.5 3.35–4.25Na2O 1.14 1.14 0.82–1.06 0.89–0.92 0.51 0.60 0.48–0.80 0.13–2.24 0.35–0.59K2O 0.14 0.16 0.03–0.04 0.10 0.04 0.02 0.02 0.04–0.51 0.02–0.03P2O5 0.61 0.54–0.66 0.40 0.32 0.31 0.29–0.31 0.46–3.40SO2 0.19Li 3.77 4.34 1.58–4.54 2.86 1.30 3.30–5.60 1.30–1.60Be 0.35 0.08 0.03 0.54Sc 30.2 37.0 32.0–38.0 31.0 29.0 29.9 34.9–36.4 26.0–72.0 21.0–25.0V 202 280 192–225 175 136 188 84.0–380 132–202Cr 3677 4317 3633–4525 3417–3900 4755–4858 95.0–2923 5672–6900Co 54.9 56.2 45.5–49.0 44.5 51.3 55.0 51.6–70.0 13.7–63.0 63.0–72.0Ni 182 232 143–158 65.0 230 310 203–270 6.60–150 250–370Cu 7.21 14.0 11.7 9.54 8.30 11.0–139 5.00–80.0Zn 41.3 49.0 69.0–81.0 62.0 49.0 61.0 76.0–81.1 54.0–130 47.0–90.0Ga 13.5 13.4 12.4–13.5 10.2 7.97 8.80 11.0 13.9–30.5 6.00–9.00Ge 1.02 0.69–1.95 0.04–0.50Rb 5.49 5.75 1.26 0.49 0.51 0.13–14.0 0.17Sr 44.9 67.0 20.0–57.0 363 47.0 8.00–81.0 8.00–30.0Y 11.3 17.2 11.6 6.78 8.37 10.0–31.2 5.70–6.20Zr 48.4 62.1 29.1 17.4 9.02 41.0–159 10.0–19.0Nb 3.41 4.37 0.83 0.30 0.13 0.68–22.0 0.50–0.60Cs 0.30 0.45 0.07 0.03 0.01 0.03–0.88 0.03–0.08Ba 24.5 127 5.00–10.0 19.0 56.0 1.54–30.0 6.00–46.8 2.30–6.00La 1.73 2.25 0.37–0.50 0.20–0.24 0.11 0.11 0.12–0.17 0.31–5.10 0.20–0.30Ce 4.25 5.38 1.28–1.42 0.68–1.45 0.29 0.35 0.43–0.70 1.00–9.84 0.90–1.10Pr 0.57 0.78 0.19 0.11 0.05 0.07 0.08 0.70–1.43 0.13Nd 2.94 3.82 1.17–1.40 0.69 0.42 0.47 0.57 1.48–7.07 0.80–1.10Sm 1.10 1.49 0.75–0.81 0.44–0.48 0.35 0.43 0.47–0.50 0.76–3.40 0.30–0.40Eu 0.47 0.55 0.35–0.39 0.08–0.20 0.18 0.18 0.25 0.43–1.30 0.10–0.20Gd 1.55 2.14 1.61 0.99 0.92 0.86 1.13 1.64–4.30 0.40–1.10Tb 0.30 0.41 0.30–0.34 0.20 0.20 0.18 0.24–0.25 0.22–1.03 0.10–0.20Dy 2.01 2.80 2.11–2.22 1.30 1.46 1.30 1.60–1.70 1.70–6.10 0.60–1.30

Line missing

2198A

.B

asuS

arbad

hik

ariet

al./G

eoch

imica

etC

osm

och

imica

Acta

73(2009)

2190–2214

Ho

0.42

0.59

0.45

–0.5

00.

270.

330.

300.

38–0

.44

0.56

–1.3

50.

20–0

.30

Er

1.14

1.63

1.35

0.81

0.93

0.77

1.09

1.33

–2.7

60.

30–0

.70

Tm

0.16

0.15

–0.1

70.

25–0

.58

0.08

–0.1

2Y

b1.

151.

371.

12–1

.28

0.81

–1.2

90.

830.

790.

97–1

.05

0.98

–3.7

70.

40–0

.50

Lu

0.15

0.20

0.15

–0.1

70.

12–0

.26

0.12

0.11

0.15

–0.1

60.

13–0

.73

0.06

–0.0

8H

f1.

281.

580.

93–0

.98

0.48

0.40

0.43

0.49

1.50

–4.2

00.

40–0

.60

Ta

0.12

0.20

0.02

–0.0

50.

020.

010.

02–0

.44

0.02

–0.0

4W

0.93

0.52

0.08

0.03

0.02

0.38

–0.8

038

.0–2

50P

b0.

350.

690.

290.

920.

170.

24T

h0.

260.

410.

10–0

.15

0.04

0.01

0.01

0.02

–0.1

00.

05–0

.92

0.04

–0.0

6U

0.07

0.10

0.02

–0.0

60.

110.

080.

050.

01–0

.02

0.01

–0.2

00.

01–0

.04

Maj

or-

elem

ent

oxi

des

inw

t%;

trac

eel

emen

ts(L

io

nw

ard

s)in

pp

m.

*T

ota

lF

eas

FeO

.�

Dat

afr

om

this

stu

dy,

Bar

rat

etal

.,(2

002a

)an

dB

urg

ess

etal

.(2

006)

.�

Dat

afr

om

Bu

rgh

ele

etal

.(1

983)

,W

arre

net

al.

(199

9)an

dN

eal

etal

.(2

001)

.#

Tay

lor

etal

.(2

000)

and

Nea

let

al.

(200

1).

$D

ata

fro

mZ

ipfe

let

al.,

2000

and

Bar

rat

etal

.(2

001)

.@

Dat

afr

om

Dre

ibu

set

al.

(200

0).


which balances the depletion of Co in the melt as crystalli-zation proceeded (Herd et al., 2001b).

3.3. Pyroxene

Pyroxenes in LAR 06139 occur as long prismatic grains(up to 1 mm in length), as small grains in the groundmassand as grains within olivine-hosted melt inclusions. Allpyroxenes possess rim overgrowths (Fig. 2d). LAR 06319pyroxenes cover a wide spectrum of compositions, withincreasing iron contents from core to rim (Fig. 5a; Table2), as typically seen in basalts that experience extensive frac-tionation. Prismatic pyroxenes are zoned from core to rimfrom Mg-rich orthopyroxene (Wo3En71, Mg# 75), to Mg-rich augite (Wo36En42, Mg# 55–68), and to Fe-rich rims(Wo8–30En23–45, Mg# 32–55). Groundmass pyroxenes(Wo5–24En26–50 and Mg# 28–53) are similar in compositionto the rims of the large pyroxene grains. Pyroxenes in melt-inclusions within the large olivine grains contain high Al2O3

but range from low-Ca pyroxenes (LCP; Wo3–11En53–71,Mg# 58–74) to high-Ca pyroxene (HCP; Wo40–46En38–46,Mg# 63–75).

Plots of molar Ti vs. Al (Fig. 5b and c) demonstrate thatpyroxenes crystallized over an extended temperature inter-val. Pyroxenes in the melt inclusions have high Al-contents(up to 0.45 afu), whereas high-Ca pyroxenes (HCP) in themelt-inclusions have the highest Ti-content (up to 0.09afu) of all pyroxenes in LAR 06319. These parameters indi-cate the presence of substantial amounts of Ca-tschermak(CaVIAlIVAlSiO6) and Ti-tschermak (CaVITi4+IVAl2O6)components in melt-inclusion pyroxenes. Ti-contents in-crease from pigeonite to the Mg-rich augite in the core, rel-ative to Al, and then become constant, or increase, towardsthe rims of large grains of pyroxene or groundmass pyrox-ene. The changing Ti/Al ratio reflects the late onset of pla-gioclase crystallization, simultaneous to the rim andgroundmass pyroxene (Fig. 5b and c). Absence of VIAl inthe majority of pyroxenes, except for those in melt inclu-sions, are consistent with relatively slow cooling rates(<0.5 oC/h; Grove and Bence, 1977) and with cooling rateestimates from olivine Fe/Mg zonation.

Rare earth element (REE) patterns of pyroxenes fromcore-to-rim are shown in Fig. 6. Pyroxene cores areLREE-depleted, but to a lesser degree than for olivine-phy-ric shergottites DaG 476 (Wadhwa et al., 2001), Dhofar 019(Taylor et al., 2002), EETA 79001A (Wadhwa et al., 1994)and Y-980459 (Usui et al., 2008). Rim pyroxenes are moreREE- and LREE-enriched relative to pyroxene cores, indi-cating extensive fractional crystallization and varying parti-tion coefficients with respect to the LREE, as a function ofchanging melt composition. Core pyroxenes have high con-centrations of Cr and V compared to rim pyroxenes (Table3). Indeed, the V–Cr trend (Fig. 4b) shows a crude positivecorrelation from core to rim of prismatic pyroxenes and forgroundmass pyroxenes.

3.4. Maskelynite (plagioclase)

Plagioclase has been completely converted to maskely-nite in the groundmass of LAR 06319. Plagioclase grains

Table 5Crystal size distribution and spatial distribution pattern data for the LAR 06319 olivine population.

SampleID

Measurement Area(mm2)

No. ofgrains

Percentarea

± 1 SD Averagedimensions

± 1 SD Max. Min. Slope(mm�1)

Intercept R value Shape

LAR Length 53.9 239 22.6 2 0.26 0.23 2.48 0.06 �11.1 7.8606319 37 Width 0.16 0.12 1.25 0.04 �12.6 8.49 1.04 1.0:1.2:1.9

Fig. 1. Photomicrographs of LAR 06319, 37 (a) thin-section in transmitted light, and (b) corresponding digital image of traced olivine.Megacrystic olivine and prismatic pyroxenes are set in a groundmass of smaller olivine (ol), pyroxene (px), maskelynite (mask), phosphate andspinel grains. Also note the impact-melt vein running through the sample.


also occur in olivine-hosted melt inclusions, where it is notpossible to assess the degree of maskelinitization. Through-out the rock, individual groundmass maskelynites are un-zoned but inter-granular compositions exhibit a range ofanorthite contents (An51 to An37; Fig. 7a). Maskelyniteoccurring in the groundmass is more enriched (�1–4 at.%)in an orthoclase component relative to plagioclase in themelt inclusions. The restricted range in composition(An51–37) of maskelynites in LAR 06319 is comparable tothat of olivine-phyric shergottite NWA 1068 (Fig. 7b). Bothof these meteorites have higher Na than the majority ofolivine-phyric shergottites and have maskelynite composi-tions in common with basaltic shergottites (e.g., Shergotty,Zagami and Los Angeles).

The REE data for maskelynite are shown in Fig. 6.There is a correlation between P2O5 and REE in maskely-nite, indicating that those with elevated REE abundanceshave likely incorporated REE-rich phosphate duringshock-melting. No evidence for beam penetration intounderlying phases was observed in the time-resolved LA-ICP-MS data. Interaction with phosphate could have oc-curred via shock implantation processes, considering theclose proximity of phosphate to plagioclase in LAR06319. The maskelynite with the lowest P2O5 is consideredto represent original plagioclase compositions, which arecharacterized by large positive-Eu anomalies and low-REE abundances. Maskelynites with low P2O5 also haveelevated concentrations of plagiophile elements such as

Ba and Sr (Table 3). The magnitude of the Eu anomaly ismore limited in P2O5-rich maskelynites than in P2O5-poorequivalents. The LREE-depleted pattern for plagioclaseunaffected by phosphate is similar to that for other oliv-ine-phyric shergottites (e.g., DaG 476: Wadhwa et al.,2001).

3.5. Oxides

Oxide minerals are diverse in LAR 06319 and includechromite, ulvospinel, titano-magnetite and ilmenite. Theappearance of diverse spinel minerals has been previouslyreported for other olivine-phyric shergottites (e.g., EETA79001: Goodrich, 2003; Goodrich et al., 2003; Wanget al., 2004; Dhofar 019: Taylor et al., 2002; Goodrichet al., 2003; DaG 476: Zipfel et al., 2000; SaU 005: Good-rich et al., 2003; NWA 1110: Goodrich et al., 2003). Formost basaltic shergottites (NWA 480 being an exception,Barrat et al., 2002b) and nakhlites, only magnetites havebeen observed (e.g., Herd et al., 2001a; Sautter et al.,2002; Taylor et al., 2002; Day et al., 2006a), whereasALH 84001 (orthopyroxenite), the chassignites and thelherzolitic shergottites contain only chromite (e.g., Floranet al., 1978; Berkley and Boynton, 1992; Goodrich and Har-vey, 2002; Taylor et al., 2002). Euhedral to subhedral chr-omites occur in the megacrystic olivines (Fig. 2b) andpyroxenes in LAR 06319, as well as in olivine-hosted meltinclusions, at the boundaries between megacrysts and

Fig. 2. Photomicrographs of key textural features in LAR 06319. (a) Displacement of a brown colored olivine grain (Ol) by fractures (imagein plane-polarized light; PPL). (b) Back-scattered electron (BSE) image of a megacrystic olivine, illustrating the location of Cr-spinelinclusions (see text for details), and a trapped melt-inclusion composed of vitreous material, pyroxene, chromite and sulfide. (c) BSE image ofmicron-sized chromite grains in a brown colored olivine grain; (d) prismatic and granular pyroxene grains set in a groundmass of co-crystallized pyroxene, plagioclase (converted to maskelynite) and opaque phases (PPL); (e) BSE image of the opaque phases in thegroundmass including the association of ilmenite (Ilm) and magnetite (Mt) and interstitial pyrrhotite (Po) between maskelynite (Mask) andpyroxene.


groundmass, and within the groundmass. Titano-magne-tites occur as individual crystals, or as paired ilmenite–mag-netite grains within the groundmass (Fig. 2e). Ilmenite iscommonly associated with chromite or magnetite (Fig. 2e).

Stoichiometric compositions of the various spinel grainsare presented in a modified Johnson spinel prism (afterHaggerty, 1991; Fig. 8a and b). The compositions of chr-

omites show a distinctive enrichment in Ti and Fe3+. Basedon their textual characteristics and compositions, chromitescan be classified into two types: (1) a nearly pure chromiteend-member that typically occurs within the olivine mega-crysts (Fig. 2b), or in the cores of groundmass spinelsand; (2) Ti and Fe3+ rich chromites (Ti–chromite) thatare typically found at the rims of olivine megacrysts

Fig. 3. Summary of olivine compositions and zonation in LAR 06319. (a) Electron microprobe analysis traverse of an olivine megacryst; (b)distribution of Fo-content along the traverse A and B in (a); (c) comparison of olivine compositions with other martian meteorites (datasources from Taylor et al., 2002; Goodrich, 2003; Ikeda, 2004; Meyer, 2006; Beck et al., 2006; Day et al., 2006a; and references therein).


(Fig. 2b) or in the rims of groundmass spinels. Projectionsof the data on an Fe3+-free (Fig. 8c) or Ti-free basis(Fig. 8d) illustrate that the spinel minerals gradually be-come enriched in Ti and Fe3+; this trend is similar to terres-trial-analog mafic-ultramafic Fe–Ti trends (Barnes andRoeder, 2001). Spinels become increasingly Fe2+ enricheduntil an Mg# of �12, where they show a pronounced dropin Cr and a corresponding increase in Ti (Fig. 8e and f).Thus, a simple crystallization sequence for the oxides canbe derived from chromite (chromite1) to Ti–chromite (chro-

mite2) and, finally, titano-magnetite. Ilmenite appears con-temporaneously with the Ti–chromite and becomesprogressively Fe-rich when crystallized together with tit-ano-magnetite (Table 2).

3.6. Phosphates

LAR 06139 contains both merrillite and apatite. Merril-lite grains occur in the melt inclusions and within thegroundmass. Apatite is only found in the groundmass asso-

Fig. 4. Systematics of Co vs. Ni (a) and Cr vs. V (b) for olivine and pyroxene in LAR 06319. Arrows illustrate crystallization trajectoriesassuming simple fractional crystallization. Also included are zoning of olivines from core to rim for other shergottites in (a), where solidarrows represent olivine-phyric shergottites and dashed arrows represent lherzolitic shergottites (data sources: Herd et al., 2001b; Taylor et al.,2002; Shearer et al., 2008; Usui et al., 2008).

Fig. 5. Compositions of pyroxenes in LAR 06319. (a) Pyroxene quadrilateral plot showing range of compositions from core to rim in themeteorite. Arrow marks the evolutionary trends of pyroxenes from core to rim measured during electron microprobe traverses; (b) and (c) Ti–Al systematics of pyroxenes illustrating path of crystallization trends and onset of plagioclase crystallization (Pl-in).


ciated with pyroxene and maskelynite. Phosphates in thegroundmass are interstitial with respect to the earlier-crys-tallized plagioclase and pyroxene. Both merrillite and apa-

tite grains were analyzed for their trace elementabundances, which demonstrate that these grains are themajor carriers of the REE in LAR 06319, and possess gen-

Fig. 6. CI-chondrite normalized rare earth element (REE) patterns for representative pyroxene, maskelynite and merrillite compositions inLAR 06319. In the shaded field are data of pyroxenes from olivine-phyric and lherzolitic shergottites. Normalization values are from Andersand Grevesse (1989). Additional data sources: Wadhwa et al. (1994, 2001) and Taylor et al. (2002).

Fig. 7. Maskelynite (plagioclase) compositions in LAR 06319 (a) and comparison with other martian meteorites (b). Data sources are fromTaylor et al. (2002), Goodrich (2003), Meyer (2006) and references therein.


erally flat-REE patterns and negative-Eu anomalies, consis-tent with late-stage crystallization after plagioclase forma-tion (Fig. 6 and Table 3). This is typical of phosphate,which is generally the main REE carrier in crystalline igne-ous rocks (Lundberg et al., 1988; Wadhwa et al., 1994; Jol-liff et al., 2006).

3.7. Sulfide minerals

LAR 06319 contains small (<20 lm) troilite, pyrrho-tite and pyrite grains (Table 2). Troilites occur in melt

inclusions and in the groundmass; pyrrhotites and pyritesare found in the groundmass, where the pyrites appear tobe late-stage crystallization products. Ni and Co contentsincrease from troilite to pyrrhotite through to pyrite. TheFe/S atomic values are 0.989 for troilite and 0.632 forpyrite. The Fe/S atomic values in pyrrhotites (Fe1�xS)range from 0.97 to 0.92, which indicate that the pyrrho-tites lie at the Fe-rich end of the pyrrhotite solid solutionseries (Fleet, 2006). Although pyrite only crystallizes frommelt at relatively low temperatures (<743 �C; Taylor,1970), it can be a sub-solidus phase formed from igneous

Fig. 8. Plots of LAR 06319 spinel compositions. Compositions are plotted in the divalent (a) and trivalent (b) prisms. Panels (c) and (d) showthe projection on Cr–2Ti–Al and Cr–Fe3+–Al. (e) Plots the variation of Cr/(Cr + Al) vs. Fe2+/(Fe2+ + Mg) and (f) plots 2Ti/(2Ti + Cr + Al)vs. Fe2+/(Fe2+ + Mg). In (f), LAR 06319 spinels are compared with other martian meteorites. Data sources: Floran et al. (1978), McSweenand Jarosewich (1983), Treiman et al. (1994), Folco et al., 2000; Zipfel et al., 2000 and Taylor et al. (2002).


pyrrhotite (Vaughan and Craig, 1997). The late-stage pyr-ites in this rock may indicate a change to greater degreesof oxidation and/or sulfurization of pyrrhotite (Taylor,1971). Association of pyrite with ilmenite also indicateschanging oxygen activity conditions during late-stagecrystallization and differentiation of the LAR 06319parental melt.

3.8. Glass and SiO2-rich phase

Glass occurs in devitrified melt inclusions in megacrysticolivine, in impact-melt veins cutting across the sample, andas impact-melt pockets that have locally melted pyroxene

and plagioclase. Compositions of these glasses vary signifi-cantly in terms of their major-element compositions (Table2). Glass within the melt inclusions contains higher Si, Aland K-contents and lower Fe, Mg and Ca-contents thanmelt pockets and melt veins. These melt inclusions also con-tain blebs of a SiO2-rich phase (Table 2). Melt vein glass hasa composition comparable with the whole-rock, but withlower Fe and Mg contents (Table 4); similar observationshave been made for lunar meteorites, where it has beenshown that fusion crust or melt vein glass cannot be usedto approximate the whole-rock composition (Day et al.,2006b). Ultimately, the melt veins and pockets reflect melt-ing of local mineralogies within LAR 06319, as shown for


other shergottites (e.g., Los Angeles: Walton and Spray,2003; ALHA 77005: Walton and Herd, 2007a; NWA1950: Walton and Herd, 2007b).

3.9. Whole-rock composition and oxygen isotopes

The average major- and trace-element abundances of theLAR 06319 whole-rock are presented in Table 4. LAR06319 has 15.8 wt% MgO and >0.55 wt% Cr2O3. TheNa2O + K2O and Rb concentrations of LAR 06319, alongwith NWA 1068, are high relative to other olivine-phyricshergottites, but are lower than for basaltic shergottites.In the case of NWA 1068, elevated fluid-mobile elementabundances may reflect some degree of hot desert alter-ation. Like other olivine-phyric shergottites, LAR 06319is richer in compatible elements (Cr, Co, Ni) than basalticshergottites. In addition, LAR 06319 has low plagiophileelement concentrations (e.g., Sr) and low Pb, U and Thcontents.

LAR 06319 is characterized by a relatively flat-REE pat-tern, �8–9 � CI-chondrite (Fig. 9) and a small Eu anomaly(Eu/Eu* = 1.09). The rock has more elevated LREE andLILE abundances when compared with the majority ofolivine-phyric shergottites (Fig. 9). Of all the shergottites,NWA 1068 and NWA 4468 show the closest affinity toLAR 06319 in terms of whole-rock REE abundances. Themulti-elemental patterns (Fig. 9) of LAR 06319, NWA1068 and NWA 4468 are generally confined to the lowerrange of the field of the basaltic shergottites for the fluidimmobile incompatible elements.

New oxygen isotope data for LAR 06319 shows thatLAR 06319 has a d17O value of +2.332 ± 0.003& and ad18O value of +3.89 ± 0.06& (based on two separate mea-surements). The D17O value (0.29 ± 0.03&; electronic an-nex Table 5) is within error of previously measured SNCmeteorites (Mars fractionation line � 0.321 ± 0.013&, de-fined by Franchi et al., 1999) and confirms LAR 06319 asbeing part of the SNC meteorite clan.

Fig. 9. Whole-rock trace element systematics of LAR 06319, NWA 1068,Different shades denote different shergottite sub-groups, as mentioned in treferences therein). CI chondrite normalization values are from Anders a

3.10. Crystal size distribution and spatial distribution pattern

analysis

Quantitative textural analysis is a powerful tool for con-straining the kinetics and dynamics of crystal nucleationand growth (Marsh, 2007). In this study, the texture ofolivine grains are quantified using crystal size distribution(CSD) and spatial distribution pattern (SDP) analysismethods to help constrain the kinetics of crystallizationand the integrated space-time crystallization of the LAR06319 parent magma. Results of CSD and SDP are plottedin Fig. 10a and b. The best matching habit for LAR 06319olivine grains from the CSDslice program of Morgan andJerram (2006) is an aspect ratio of 1.0:1.2:1.9 (broadly rect-angular habit). As expected, the resulting CSD plot showsa major kink in the width and length slopes consistent withthe difference in grain sizes visually observed in the meteor-ite. Lowering of the regress slope in the >0.96 mm crystalsize regime reflects crystal accumulation during the earlystage of crystallization. Downturns for smaller size-fractionregimes (<0.10 mm) indicate possible cessation of nucle-ation due to reaction of olivine with melt to form pyroxene(Fig. 10a). It is worth noting that there are larger uncer-tainties in measuring crystal sizes for these small grainsdue to pixilation and tracing considerations. Furthermore,the number of small crystals may be overestimated due tofragmentation by shock. With these caveats in mind, theCSD pattern of olivine in LAR 06319 (Fig. 10a) showsan overall continuum of decreasing grain-size in the range<0.96 mm. We suggest that the larger size fraction of oliv-ine (<5 vol.% of the total population) is due to the effect ofolivine growth and accumulation in a magma conduit orchamber. The slope of the CSD of the smallest crystals(0.1–0.95 mm) can be used to approximate the growth rateand residence time, if the cooling rate is known (Marsh,1988). Using cooling rates inferred from Fe–Mg profilesin olivine and Ti–Al trends in pyroxene (0.2 and 0.5 �C/h,average of �0.4 �C/h), and the crystallization temperature

NWA 4468 and QUE 94201 in comparison with other shergottites.he figure index (data from Barrat et al., 2002a; Usui et al., 2008 andnd Grevesse (1989).

Fig. 10. (a) Crystal size distribution of olivine grains. Both the corrected length and width are plotted. Dotted and dashed lines are regressionlines for length and width data for the size fraction of 0.12–0.96 mm. (b) Spatial distribution pattern of LAR 06319 olivines. ‘‘Porosity” is thatportion of the rock that is not olivine. See the definition of R in the text. LAR 06319 lies close to the Belingwe komatiite basal chill, whichcontains olivine megacrysts set in a fine-grained groundmass (after Jerram et al., 2003). RSDL = random sphere distribution line.


range of 300 �C estimated from the MELTS algorithm (seelater discussion for details), the growth rate of olivine inLAR 06319 is �3.1 � 10�8 mm/s (Table 5), which is com-parable to those estimated for minerals in terrestrial lavaflows (Marsh, 1998; Jerram et al., 2003) and in martianand lunar meteorites (Day et al., 2006a; Day and Taylor,2007). SDP analyses of LAR 06319 olivine grains plot closeto measurements of fast chilled komatiite olivine (Fig. 10b).Indeed, the LAR 06319 olivine crystal morphologies andgrain size distributions show close resemblance to chilledkomatiite olivine (c.f., Fig. 10 of Jerram et al., 2003). Thus,cluster analysis indicates that the olivine grains grew fromclusters of nuclei, or accumulated in clusters (Jerramet al., 2003).

4. DISCUSSION

4.1. Olivine megacrysts: phenocrysts, or xenocrysts?

Several lines of evidence support LAR 06319 megacry-stic olivines as being phenocrysts. Perhaps the most con-vincing demonstration of a phenocryst origin for themegacrysts comes from melt inclusions trapped in the mostMg-rich olivine crystals. The trace-element compositions ofthese trapped melt inclusions are similar to that of thewhole-rock (Table 3). The compositions of the olivinesthemselves also support their origin as phenocrysts. Using

an exchange coefficient, KOl� liq

D; Fe–Mg= 0.32 (Toplis, 2005),

olivine in equilibrium with the bulk rock (Mg# = 58) would

have Fo81. The most Mg-rich olivine in our LAR 06319 sec-tions is Fo77. The small difference between measured andcalculated compositions of the most Mg-rich olivine inLAR 06319 may reflect the choice of exchange coefficient,or may reflect partial olivine accumulation in thewhole-rock relative to the parental melt composition.Constant Co in all olivines observed in LAR 06319 alsosupports the hypothesis that the olivine megacrysts areindeed phenocrysts. Using an exchange coefficient,

KOl� Px

D; Fe�Mg= 1.2 (Longhi and Pan, 1989), the olivine

composition in equilibrium with the pyroxene core (Mg#75) is Fo70. It is also notable that, unlike xenocrysticolivines in NWA 1068, which possess weakly zoned cores,olivines in LAR 06319 are strongly zoned with respectto major and trace elements, from core-to-rim (Figs. 3band 4).

Finally, olivine crystal size distribution (CSD) plots forLAR 06319 show a size gap between megacrystic olivine(>1 mm) and groundmass olivine grains (Fig. 10a). This‘kink’ is notable, considering that similar CSD size gapshave been recorded for Y-980459 (Greshake et al., 2004;Lentz and McSween, 2005), EETA 79001A and SaU 005(Goodrich, 2003) and are also likely to be present in otherolivine-phyric shergottites, due to their porphyritic nature.The distinct linear slopes on a CSD plot reflect differentcooling regimes for the two main crystal populations. Con-sidering this important information, Usui et al. (2008) sug-gested a model for the megacryst olivine population in Y-980459 forming in a magma conduit where new batches


of primary magma continuously promoted growth of mega-crysts as cumulus crystals. Given that LAR 06319 olivinesare close to equilibrium with the whole-rock Mg/Fe compo-sition, it seems that LAR 06319 may also be consistent withsuch a model.

4.2. Crystallization sequence

The crystallization sequence of LAR 06319 is summa-rized in Fig. 11a from petrography and mineral chemicalrelationships. Further evidence comes from thermodynamicmodels. Given the KD calculations above, olivine withFo77–71 is likely an early-crystallizing phase and those withFo70 crystallized coevally with orthopyroxene/pigeonitecores (Mg# 75). With continuous crystallization, pyroxeneevolves toward high-Mg augite, whereas olivine becomesmore Fe rich. Decrease of Al with respect to the Ti-compo-nent, during high-Mg augite crystallization, marks the firstappearance of plagioclase. Plagioclase crystallization con-tinues, as a major constituent, until the end of silicate crys-tallization. The association of spinels with silicate mineralsand their compositional variations suggest that chromitescrystallized at about the same time as the Mg-rich pigeoniteand are followed by co-crystallization of Ti–chromite withMg-rich augite and then to titano-magnetites with Fe-richpigeonite. Chromite is likely to have crystallized slightly la-ter than olivine with PFo70. Ilmenite, sulfide and phos-phate occur in the groundmass in association with

Fig. 11. Qualitative crystallization sequence based on petrographic obseequilibrium (b) and fractional crystallization (c). Modeling parameters dAug, augite; Plag, plagioclase; Sp, spinel; Ilm, ilmenite; P, phosphate; S,MC, megacryst; GM, groundmass.

groundmass pyroxene, maskelynite and titano-magnetite.Therefore, they are late-stage crystallization products.

The compatibility of the observed crystallization se-quence in LAR 06319 and that predicted from the MELTSalgorithm (Ghiorso and Sack, 1995) is shown in Fig. 11band c. MELTS was chosen over similar computer modelingsoftware because it has been reported by Thompson et al.(2003) to successfully model martian meteorite composi-tions more faithfully than MAGPOX or COMAGMAT.We have also determined the oxygen fugacity of LAR06319 as QFM-1.7 ± 0.5 from the equilibrium early high-Mg olivine–early high-Mg pyroxene–chromite assemblageusing the oxygen barometer of Wood (1991), as outlinedby Herd et al. (2002) and Goodrich et al. (2003). Therefore,calculations of the bulk rock composition of LAR 06319were performed at log fO2 of QFM-2 and 1 bar, startingwith a liquidus temperature of 1390 �C.

Both equilibrium and fractional crystallization MELTSmodeling (Fig. 11b and c), using the whole-rock composi-tion, provide reasonable matches with the crystallizationsequence assessed from petrographic observations(Fig. 11a). The general trends are similar: olivine ? orth-opyroxene ? clinopyroxene ? plagioclase. Both olivineand pyroxene become progressively Fe-enriched with time.In detail, however, both models failed to predict the cor-rect compositional variations for olivine, pyroxenes andplagioclase. The calculated composition for early olivineis more Mg-rich (Fo84) than observed (Fo77). Fractional

rvation (a) vs. model calculations using the MELTS algorithm forefined in the text. Ol, olivine; Opx, orthopyroxene; Pig, pigeonite;sulfide; Chr, chromite; Mt, magnetite; Whit, whitlockite/merrillitte;


crystallization by MELTS suggests more Fe-rich late-stageolivines (Fo46-6) than were actually measured. Also, com-positional variations for pyroxenes from MELTS followtwo different trends (orthopyroxene to Fe-rich pigeonite,and Mg-rich pigeonite to augite) instead of the continuousvariation from orthopyroxene to augite to Fe–augite, asobserved. Plagioclase compositions from MELTS model-ing are more An-rich than observed. These discrepanciesbetween the observed and calculated compositions maybe explained by excess olivine accumulation, which prob-ably causes depletion of Na in the parent melt composi-tion. Nevertheless, the overall crystallization sequence,both from petrographic observations and MELTS arebroadly consistent. These observations lend further sup-port to the olivine megacrysts being in equilibrium withthe host rock.

4.3. Parent-melt composition of LAR 06319

Parental melt REE compositions in equilibrium withprimary magmatic phases can be estimated using corre-sponding experimentally and empirically derived partitioncoefficients. We have used distribution coefficients DPx-Melt

for the REE derived from the work of Lundberg et al.(1988, 1990) to calculate the equilibrium melt compositionof LAR 06319 from pyroxene cores. It should be noted thatREE partition coefficients for low-Ca pyroxenes by Lund-berg et al. (1990) are larger than those of McKay et al.(1986) by <20% for most of the REE, and a factor of 3for La and Ce. The difference in calculated melts for low-Ca pyroxenes using these two sets of partition coefficientsis indiscernible except for La and Ce. REE partition coeffi-cients between pyroxene and melt for high-Ca pyroxenesfrom Lundberg et al. (1990) are less than those of McKayet al. (1986) by, at most, a factor of two. As a result, theestimated melt REE concentrations are also greater by afactor of two.

Fig. 12. Calculated REE abundances in the parent melt from early-crystais drawn with the whole-rock and a representative trapped melt-inclusioncrystallized phases and whole-rock compositions. Also note the close corrwhole-rock composition. Normalization values are from Anders and Gr

Calculated equilibrium melt compositions are shown inFig. 12. Equilibrium melt compositions, calculated fromthe most Mg-rich pigeonite and augite core compositions,are approximately parallel to the measured LAR 06319whole-rock composition (Fig. 12). We have also plotted asmall, partially re-crystallized and round melt-inclusionfrom the core of a megacrystic olivine (with the highestmeasured Fo-content), which has similar trace elementabundances and patterns to the whole-rock. LREE deple-tion of the pyroxene cores is not as pronounced as thosein the olivine-phyric shergottites DaG 476 (Wadhwaet al., 2001), EETA 79001A and 79001B (Wadhwa et al.,1994), Y-980459 (Usui et al., 2008), and Dhofar 019 (Tayloret al., 2002). This provides powerful evidence that theparental melt for LAR 06319 was more LREE-enrichedthan other olivine-phyric (with the exception of NWA1068) and lherzolitic shergottites.

4.4. LAR 06319 in the context of the shergottite clan

There are some close petrological and geochemical sim-ilarities between LAR 06319, olivine-phyric shergottiteNWA 1068 and olivine-basalt shergottite NWA 4468 (Figs.3, 7, and 9). Geochemically, these three meteorites are vir-tually indistinguishable with similar FeO bulk compositionsand Mg-rich olivine compositions for both NWA 1068 andNWA 4468 (Barrat et al., 2002a; Irving et al., 2007). Thissimilarity is remarkable given the small sample sizes typi-cally processed for these meteorites (see Spicuzza et al.,2007 for a discussion on the mode-effect in small sampleallocations). Recent analysis of the cosmic exposure ageof LAR 06319 gives an age of �3.3 Ma (Nagao and Park,2008); a result in close agreement with the cosmic exposureage for NWA 1068 (2.5–3.1 Ma; Nishiizumi and Caffee,2006). However, there are some subtle differences in themode-of-origin of megacrystic assemblages and in mineralchemistry. In contrast to the phenocrystic olivines in

llized mineral assemblages, pigeonite and augite cores. Comparisonfrom the olivine host. Note the close correspondence between early-espondence between melt-inclusion compositions and the measuredevesse (1989).


LAR 06319, megacrystic olivines in NWA 1068 are consid-ered to be xenocrysts based on the flat profiles of forsteritecompositions in their cores (Barrat et al., 2002a). Pyroxenesin LAR 06319 also differ from those in NWA 1068. InNWA 1068, pyroxenes are only present in the groundmassand their composition is comparable to those of pyroxeneswithin the basaltic shergottites (Barrat et al., 2002a),whereas pyroxenes in LAR 06319 tend to be more evolved(Fe-rich compositions) at their rims and more typical ofolivine-phyric shergottites (e.g., Y 980459, DaG 476, EETA79001A, Dhofar 019). NWA 1068 has a finer-grainedgroundmass than LAR 06319, which may indicate moder-ately faster cooling. Based on our results, it seems reason-able to suggest that LAR 06319 is close to a parentalliquid composition, modified by accumulation of a limitedproportion of olivine megacryst cumulate material. Con-versely, megacrystic and groundmass assemblages ofNWA 1068 represent two different melt compositions, asindicated by differences in oxygen fugacity (Herd, 2006)as well as xenocrystic olivines and associated chromites(Barrat et al., 2002a; Herd, 2006).

The calculated fO2 (QFM -1.7) derived from early crystal-lized assemblages of LAR 06319 is intermediate between theLREE-depleted olivine-phyric shergottites (�QFM-2 to -4)and LREE-enriched basaltic shergottites (�QFM) (Herdet al., 2002; Herd, 2003). Considering the similar quantitiesof mafic minerals in LAR 06319 and NWA 1068 comparedwith the sub-group of olivine-phyric shergottites, these mete-orites likely represent an important and distinctive magmatype within the shergottite meteorites.

4.5. Heterogeneous mantle source vs. crustal contamination

for the enriched component in shergottites; the perspective

from LAR 06319

The shergottites possess a diverse array of geochemicaland isotopic signatures. They also span a range in ages,from 575 to 165 Ma based on Rb–Sr and Sm–Nd isotopesystematics (Nyquist et al., 2001). We note that there issome controversy regarding these ages given that Pb iso-tope systematics of shergottites have yielded ages of�4.1 Ga (Bouvier et al., 2008), but this does not affect pet-rological arguments made here. Based on modal mineral-ogy and major-element geochemistry, shergottites havebeen classified into basaltic, olivine-phyric (with an oliv-ine–orthopyroxene-phyric sub-group) and lherzolitic sher-gottites. In general, lherzolitic shergottites are traceelement (especially the LREE) depleted, with low to moder-ate oxygen fugacities, and basaltic shergottites span a rangeof compositions from enriched (Shergotty, Zagami, etc.) todepleted (QUE 94201) with respect to trace elements andthe LREE, and cover a range of fO2 from QFM -4 toQFM (Borg et al., 1997; Wadhwa, 2001; Herd et al.,2002; Herd, 2003; Goodrich et al., 2003). Olivine-phyricshergottites typically lie somewhere between these twogroups of shergottites, although there are exceptions. Forexample, DaG 476 and SaU 005 are more depleted inLREE than the lherzolites and LAR 06319 and NWA1068 closely resemble basaltic shergottites with respect totheir REE abundances. This remarkable and relatively

complex variation in trace element compositions has ledto two major competing hypotheses for the cause of traceelement enrichment, correlations in isotopic compositionsand oxidation state for the shergottites. These are assimila-tion of enriched, oxidized crust by mantle-derived magmas(assimilation fractional crystallization, AFC; Jones, 1989;Wadhwa, 2001; Herd et al., 2002), or mixing of distinctmantle reservoirs during melting (Herd, 2003; Borg andDraper, 2003). Perhaps the strongest argument for a heter-ogeneous mantle source for the shergottites comes fromlack of correlations between mineralogical and geochemicalindices of differentiation and incompatible trace-elementabundances, ratios or isotopic compositions (Borg andDraper, 2003).

There are no clearly defined correlations or trends be-tween major- and trace-element compositions for shergot-tites. For example, basaltic shergottite QUE 94201 has arelatively LREE-depleted pattern compared with LAR06319, much like the majority of olivine-phyric and lherzo-litic shergottites. McSween et al. (1996) have argued thatQUE 94201 represents closed-system crystallization froma fractionated parental melt that could be consistent with(1) limited contamination from REE-enriched crustal mate-rials, or (2) derivation from a LREE-depleted source. Con-versely, LAR 06319 and NWA 1068 are olivine-phyricshergottites that have REE-enriched patterns more closelymatching basaltic shergottites. There are correlations be-tween trace elements and Nd isotopes, however. Variousauthors (e.g., Borg et al., 1997; Symes et al., 2008 and ref-erences therein) have demonstrated that LREE-depletedshergottites also possess radiogenic 143Nd/144Nd (nakhlitespossess similar incompatible-element systematics), whereasshergottites with flat REE patterns possess both radiogenic87Sr/86Sr and unradiogenic 143Nd/144Nd. This correlationclearly emphasizes the antiquity of the enriched and de-pleted reservoirs present in shergottites, such that variationsin REE-rich or -depleted lavas are unlikely to reflect partialmelting processes. Borg et al. (2003, 2005) and Symes et al.(2008) have used these relations to propose a model for thegenesis of shergottites whereby early-formed enriched anddepleted mantle reservoirs are tapped by multiple parentalmagmas that ultimately result in the generation of theshergottites.

The compositions of melt inclusions within the cores ofolivine megacrysts in LAR 06319 place some importantconstraints on the nature of the enriched component inshergottites. Our measurements of the melt inclusions dem-onstrate a composition that is identical to the whole-rockcomposition (Fig. 12). Critically, these inclusions show novisible evidence for interaction with surrounding ground-mass. Thus, if the model of early olivine crystallization inmagma-feeder conduits is correct, it infers that the parentalmelt composition of LAR 06319 already possessed a REE-enriched character at the time of olivine crystallization andconstrains any possible contamination to have occurred atdepths greater than the melt-inclusion trapping event. Pres-ence of amphibole (kaersutite) in some shergottite olivinemelt inclusions has been used as permissive evidence for rel-atively deep (10 km) crystallization of olivine (McSweenand Harvey, 1993; Treiman, 1997). Typically, assimilation


of crust occurs in conjunction with fractional crystalliza-tion, yet the melt inclusions within olivine attest to an en-riched character in the LAR 06319 parental melt prior tocrystal–liquid fractionation. Clearly, such an observationplaces important constraints on the nature of any enriched,oxidized component in the martian mantle or crust.

4.6. A petrogenetic model for LAR 06319

The agreement between the whole-rock REE data and cal-culated melt compositions from early mineral assemblages,Fe–Mg equilibrium between olivine, orthopyroxene andthe whole-rock suggests that LAR 06319 preserves aclosed-system melt composition that shows evidence forsome olivine accumulation (<5% based on quantitative tex-tural analysis). Possible processes that could generate LAR06319 have to satisfy the observations that the meteorite con-tains: (1) phenocrystic olivine and co-magmatic prismaticpyroxenes; (2) a relatively high-magnesian (Mg# 58) whole-rock composition; (3) a flat REE pattern and enriched incom-patible elements (e.g., K, La); (4) an intermediate oxidationstate during the early-stage of crystallization (�QFM -1.7)compared with the LREE-depleted and -enriched shergot-tites; and (5) increasing oxidation conditions at later stagesof crystallization, based on increasing Fe3+ in Fe–Ti oxides.

We consider that the parental magma of LAR 06319 wasmost likely derived from partial melting (polybaric?) of anancient LREE-enriched mantle source region, followingthe model of Symes et al. (2008). This requires that thereis more than one parental melt composition for shergottites.Magma migration occurred through feeder conduits in themartian mantle and crust to a mid-level (10 km?) crustalmagma chamber. At this stage, the enriched incompatible-element character of the LAR 06319 parental melt had al-ready been inherited. Olivines with chromite inclusionsstarted to accumulate at the walls of the magma chamberor conduit and incidentally incorporated trapped melt com-ponents (the melt inclusions). Entrainment of the pheno-crystic olivines and crystallization of prismatic pyroxenesoccurred during a period of steady-state nucleation andgrowth. The decreasing magnitude of the Eu-anomaly inthe maskelynite and correspondingly in the pigeonite andaugite, and Fe3+ enrichment in the spinel during crystalliza-tion indicates that groundmass olivine and pyroxene werecrystallized under steadily increasing conditions of oxygenfugacity. Finally the crystal-bearing magma was eruptedto the surface and as it flowed and cooled, crystallized thegroundmass material. This model is similar to that of Usuiet al. (2008) for Y-980459, but differs in one important re-spect: the LAR 06319 parental melt is incompatible-elementenriched relative to that of Y-980459. The different ground-mass textures for olivine-phyric shergottites (e.g., Y-980459vs. NWA 1068 vs. LAR 06319) also indicates that coolingrates in the lava flows differed significantly, as implied forthe proposed nakhlite flow model (e.g., Day et al., 2006a).To summarize, LAR 06319 (NWA 1068 and possiblyNWA 4468) are likely derived from an incompatible-ele-ment enriched source and were part of a martian magmaticsystem where crystallization and re-entrainment occurred inmagma conduits in the crust, followed by near-surface crys-

tal–liquid fractionation to generate their porphyritic charac-ter. The composition of LAR 06319 serves to demonstratethat the myriad compositions of shergottites preserve evi-dence for a strongly differentiated mantle, presumablyformed early in the history of the red planet.

5. CONCLUSIONS

LAR 06319 is an olivine-phyric shergottite with an oxy-gen isotope composition consistent with derivation fromMars (D17O = 0.29 ± 0.03). LAR 06319 is relatively oxi-dized and incompatible-element enriched compared toother olivine-phyric shergottites, with the exception ofNWA 1068. The megacryst olivine in LAR 06319 is pheno-crystic and accumulative. Melt inclusions within thesemegacryst olivines have nearly identical trace element pat-terns to that of the whole-rock. The similar composition be-tween LAR 06319 bulk rock and early trapped meltinclusions provides powerful evidence for closed-systembehavior for LAR 06319 and the inheritance of incompati-ble-element enrichment prior to fractional crystallization.Crystal size distribution and spatial distribution patternanalyses of olivine indicate at least two different crystalpopulations. This is most simply interpreted as crystalliza-tion of megacryst olivine prior to eruption in magma con-duits followed by eruption and crystallization of smaller(<0.25 mm) ground mass olivine.

LAR 06319 shows close affinity in mineral and whole-rock chemistry to olivine-phyric shergottite, NWA 1068and basaltic shergottite, NWA 4468. The outstanding nat-ure of these meteorites is that they have relatively similarquantities of mafic minerals compared with olivine-phyricshergottites (e.g., Y-980459, Dho 019), but flat and elevatedrare earth element patterns more consistent with the LREE-enriched basaltic shergottites (e.g., Shergotty, Los Angeles).This relationship can be interpreted as arising from partialmelting of an enriched mantle source and subsequent crys-tal–liquid fractionation to form the enriched olivine-phyricand basaltic shergottites. Thus, we favor LAR 06319 form-ing from partial melting of an ‘‘enriched” and oxidizedmantle reservoir.

ACKNOWLEDGMENTS

We acknowledge the Meteorite Working Group and the ANS-MET 06/07 field team for provision and collection of LAR 06319.Allan Patchen assisted in microprobe analyses and preparation offused beads and Dougal Jerram is thanked for pointing out thesimilarity in olivine crystal populations from some terrestrial kom-atiite flows. We acknowledge Jean-Alix Barrat, John Bridges andChristopher Herd for providing constructive reviews of the paperand Sara Russell for editorial handling. This research has been sup-ported by Cosmochemistry Grant NNG05GG03G (to L.A. Tay-lor) and NNX07AI48G (to D. Rumble).

APPENDIX A. SUPPLEMENTARY DATA

Supplementary data associated with this article can befound, in the online version, at doi:10.1016/j.gca.2009.01.012.

http://dx.doi.org/10.1016/j.gca.2009.01.012

http://dx.doi.org/10.1016/j.gca.2009.01.012


REFERENCES

Anders E. and Grevesse N. (1989) Abundances of the elements:meteoritic and solar. Geochim. Cosmochim. Acta 53, 197–214.

Antarctic Meteorite Newsletter 2007, v. 30, No. 2, p. 32. Availablefrom: <http://www.curator.jsc.nasa.gov/antmet/amn/amn.cfm>.

Barnes S. J. and Roeder P. L. (2001) The range of spinelcompositions in terrestrial mafic and ultramafic rocks. J. Petrol.

42, 2279–2302.

Barrat J. A., Blichert-Toft J., Nesbitt R. W. and Keller F. (2001)Bulk chemistry of Saharan shergottite Dar al Gani 476. Meteor.

Planet. Sci. 36, 23–29.

Barrat J. A., Jambon A., Bohn M., Gillet P., Sautter V., Gopel C.,Lesourd M. and Keller F. (2002a) Petrology and chemistry ofthe Picritic Shergottite North West Africa 1068 (NWA 1068).Geochim. Cosmochim. Acta 66, 3505–3518.

Barrat J. A., Gillet Ph., Sautter V., Jambon A., Javoy M., GopelC., Lesourd M. and Petit E. (2002b) Petrology and geochem-istry of the basaltic shergottite NWA 480. Meteor. Planet. Sci.

37, 487–999.

Beck P., Gillet P., Barrat J. A., Wadhwa M., Greenwood R. C.,Franchi I. A., Bohn M., Cotten J., Moortele B. and van deReynard B. (2006) Petrography and geochemistry of thechassignite Northwest Africa 2737 (NWA 2737). Geochim.

Cosmochim. Acta 70, 2127–2139.

Berkley J. L. and Boynton N. J. (1992) Minor/major elementvariation within and among diogenite and howardite orthopy-roxenite groups. Meteoritics 27, 387–394.

Boctor, N. Z., Fei, Y., Bertka, C. Alexander, C. M. O. D. andHauri, E. (1998) Vitrification and high pressure phase transitionin olivine megacrysts from lithology A in Martian meteoriteEETA79001. Lunar Planet. Sci. 29th, #1492 (abstr.).

Bogard D. D. and Johnson P. (1983) Martian gases in an Antarcticmeteorite? Science 221, 651–654.

Borg L. E. and Draper D. A. (2003) A petrogenetic model for theorigin and compositional variation of the Martian basalticmeteorites. Meteor. Planet. Sci. 38, 1713–1731.

Borg L. E., Nyquist L. E., Taylor L. A., Wiesmann H. and ShihC.-Y. (1997) Constraints on Martian differentiation processesfrom Rb–Sr and Sm–Nd isotopic analyses of the basalticshergottite QUE 94201. Geochim. Cosmochim. Acta 61, 4915–

4931.

Borg L. E., Nyquist L. E., Wiesmann H., Shih C.-Y. and ReeseY. (2003) The age of Dar al Gani 476 and the differentiationhistory of the Martian meteorites inferred from their radio-genic isotopic systematics. Geochim. Cosmochim. Acta 67,

3519–3536.

Borg L. E., Edmunson J. E. and Asmerom Y. (2005) Constraintson the U–Pb isotopic systematics of Mars inferred from acombined U–Pb, Rb–Sr, and Sm–Nd isotopic study of theMartian meteorite Zagami. Geochim. Cosmochim. Acta 69,

5819–5830.

Bouvier A., Blichert-Toft J., Vervoort J. D., Gillet P. and AlbaredeF. (2008) The case for old basaltic shergottites. Earth Planet.

Sci. Lett. 266, 105–124.

Bridges J. C. and Warren P. H. (2006) The SNC meteorites:basaltic igneous processes on Mars. J. Geol. Soc. 163, 229–251.

Burgess, K. D., Musselwhite, D. S. and Treiman, A.H. (2006)Experimental petrology of olivine-phyric shergottite NWA1068: toward defining a parental melt. Lunar Planet Sci. 37th,#1972.

Burghele A., Dreibus G., Palme H., Rammensee W., Spettel B.,Weckwerth G. and Wanke H. (1983) Chemistry of shergottitesand the shergottite parent body (SPB): further evidence for thetwo components model for planet formation. Lunar Planet Sci.

14, 80–81.

Clayton R. N. and Mayeda T. K. (1996) Oxygen isotopic studies ofachondrites. Geochim. Cosmochim. Acta 60, 1999–2017.

Day J. M. D. and Taylor L. A. (2007) On the structure of marebasalt lava flows from textural analysis of the LaPaz Icefieldand Northwest Africa 032 lunar meteorites. Meteor. Planet. Sci.

42, 3–17.

Day J. M. D., Taylor L. A., Floss C. and McSween, Jr., H. Y.(2006a) Petrology and chemistry of MIL 03346 and itssignificance in understanding the petrogenesis of nakhlites onMars. Meteor. Planet. Sci. 41, 581–606.

Day J. M. D., Taylor L. A., Floss C., Patchen A. D., Schnare D.W. and Pearson D. G. (2006b) Comparative petrology,geochemistry and petrogenesis of evolved, low-Ti lunar marebasalt meteorites from the La Paz Icefield, Antarctica. Geochim.


Donaldson C. H. (1990) An experimental investigation of olivinemorphology. Contrib. Mineral. Petrol. 57, 187–213.

Dreibus G., Spettel B., Haubold R., Jochum K. P., Palme H., WolfD. and Zipfel J. (2000) Chemistry of a new shergottite: Sayh AlUhaymir 005. Meteor. Planet. Sci. 35, A49 (abst.).

Dyar M. D. (2003) Ferric iron in SNC meteorites as determined byMossbauer spectroscopy: implications of the Martian Landersand Martian oxygen fugacity. Meteor. Planet. Sci. 38, 1733–

1752.

Fleet M. E. (2006) Phase equilibria at high temperatures. Rev.

Mineral. Geochem. 61, 365–419.

Floran R. J., Prinz M., Hilava P. F., Keil K., Nehru C. E. andHinthorne J. R. (1978) The Chassigny meteorite: a cumulatedunite with hydrous amphibole-bearing melt inclusions. Geo-

chim. Cosmochim. Acta 42, 1213–1229.

Folco L., Franchi I. A., D’Orazio M., Rocchi S. and Schultz L.(2000) A new Martian meteorite from the Sahara: the shergot-tite Dar al Gani 489. Meteor. Planet. Sci. 35, 827–839.

Franchi I. A., Wright I. P., Sexton A. S. and Pillinger C. T. (1999)The oxygen-isotopic composition of Earth and Mars. Meteor.

Planet. Sci. 34, 657–661.

Fritz J., Artemieva N. and Greshake A. (2005) Ejection of Martianmeteorites. Meteor. Planet. Sci. 40, 1393–1411.

Ghiorso M. S. and Sack R. O. (1995) Chemical mass transfer inmagmatic processes. IV. A revised and internally consistentthermodynamic model for the interpolation and extrapolationof liquid–solid equilibria in magmatic systems at elevatedtemperatures and pressures. Contrib. Mineral. Petrol. 119, 197–

212.

Goodrich C. A. (2003) Petrogenesis of olivine-phyric shergottitesSayh al Uhaymir 005 and Elephant Moraine A79001 lithologyA. Geochim. Cosmochim. Acta 67, 3735–3771.

Goodrich C. A. and Harvey R. P. (2002) The parent magmas oflherzolitic shergottites ALHA77005 and LEW 88516: a reeval-uation from magmatic inclusions in olivine and chromite.Meteor. Planet. Sci. 37, A54.

Goodrich C. A., Herd C. D. K. and Taylor L. A. (2003) Spinelsand oxygen fugacity in olivine-phyric and lherzolitic shergot-tites. Meteor. Planet. Sci. 38, 1773–1792.

Greshake A., Fritz J. and Stoffler D. (2004) Petrology and shockmetamorphism of the olivine-phyric shergottite Yamato980459: evidence for a two-stage cooling and a single-stageejection history. Geochim. Cosmochim. Acta 68, 2359–2377.

Grove, T. L. and Bence, A. E. (1977) Experimental study ofpyroxene–liquid interaction in quartz-normative basalt 15597.Proc. Lunar Sci. Conf. 8th, pp. 1549–1579.

Haggerty S. E. (1991) Oxide mineralogy of the upper mantle. Rev.

Mineral. 25, 354–416.

Herd C. D. K. (2003) The oxygen fugacity of olivine-phyricMartian basalts and the components within the mantle andcrust of Mars. Meteor. Planet. Sci. 38, 1793–1806.

http://www.curator.jsc.nasa.gov/antmet/amn/amn.cfm


Herd C. D. K. (2006) Insights into the redox history of the NWA1068/1110 Martian basalt from mineral equilibria and vana-dium oxybarometry. Am. Mineral. 91, 1616–1627.

Herd C. D. K., Papike J. J. and Brearley A. J. (2001a) Oxygenfugacity of Martian basalts from electron microprobe oxygenand TEM-EELS analyses of Fe–Ti oxides. Am. Mineral. 86,

1015–1024.

Herd C. D. K., Karner J. M., Shearer C. K. and Papike J. J.(2001b) The effect of oxygen fugacity on Co and Ni partitioningin olivine: insights into Martian magmas. Meteor. Planet. Sci.

36, A37 (Abstract).

Herd C. D. K., Borg L. E., Jones J. and Papike J. J. (2002) Oxygenfugacity and geochemical variations in the Martian basalts:implications for Martian basalt petrogenesis and the oxidationstate of the upper mantle of Mars. Geochim. Cosmochim. Acta

66, 2025–2036.

Higgins M. D. (2000) Measurement of crystal size distributions.Am. Mineral. 85, 1105–1116.

Ikeda Y. (2004) Petrology of the Yamato 980459 shergottite.Antarct. Meteor. Res. 17, 35–54.

Irving, A. J., Kuehner, S. M., Korotev, R. L. and Hupe, G. M.(2007) Petrology and bulk composition of primitive enrichedolivine basaltic shergottite Northwest Africa 4468. Lunar

Planet. Sci. Conf. Abs. XXXVIII, 1526.Jerram D. A., Cheadle M. J., Hunter R. H. and Elliott M. T. (1996)

The spatial distribution of grains and crystals in rocks. Contrib.

Mineral. Petrol. 125, 60–74.

Jerram D. A., Cheadle M. J. and Philpotts A. J. (2003) Quantifyingthe building blocks of igneous rocks: are clustered crystalframeworks the foundation? J. Petrol. 44, 2033–2051.

Jolliff B. L., Wieczorek M. A., Shearer C. K. and Neal C. R. (2006)New views of the Moon. Rev. Mineral. Geochem. Mineral. Soc.

Am. 60, 721.

Jones, J. (1989) Isotopic relationships among the shergottites, thenakhlites and Chassigny. Proc. Lunar Planet. Sci. Conf. 19th,pp. 465–474.

Kurihara, T., Mikouchi, T., Saruwatari, K., Kameda, J. andMiyamoto, M. (2008) Transmission electron microscopy ofolivine in the LAR 06319 olivine-phyric shergottite. Meteor.

Planet. Sci. 43, Annu. Met. Soc. Meeting 71st, #5177.Lentz R. C. F. and McSween H. Y. (2005) A textural examination

of the Yamato 9804569 and Los Angeles shergottites usingcrystal size distribution analysis. Antarct. Meteor. Res. 18, 66–

82.

Lofgren G. E. (1983) Effect of heterogeneous nucleation on basaltictextures: a dynamic crystallization study. J. Petrol. 24, 229–255.

Longhi, J. and Pan, V. (1989) The parent magmas of the SNCmeteorites. Proc. Lunar Planet. Sci. Conf. 19th, pp. 451–464.

Lundberg L. L., Crozaz G., McKay G. and Zinner E. (1988) Rareearth element carriers in the Shergotty meteorite and implica-tions for its chronology. Geochim. Cosmochim. Acta 52, 2147–

2163.

Lundberg L. L., Crozaz G. and McSween H. Y. (1990) Rare earthelements in minerals of the ALHA77005 shergottite andimplications for its parent magma and crystallization history.Geochim. Cosmochim. Acta 54, 2535–2547.

Marsh B. D. (1988) Crystal size distribution (CSD) in rocks and thekinetics and dynamics of crystallization: 1. Theory. Contrib.

Mineral. Petrol. 99, 277–291.

Marsh B. D. (1998) On the interpretation of crystal size distribu-tions in magmatic systems. J. Petrol. 39, 553–599.

Marsh B. D. (2007) Crystallization of silicate magmas decipheredusing crystal size distributions. J. Am. Ceram. Soc. 90, 746–757.

McKay G., Wagstaff J. and Yang S. R. (1986) Clinopyroxene REEdistribution coefficients for shergottites: the REE content ofShergotty melt. Geochim. Cosmochim. Acta 50, 927–937.

McSween H. Y. and Jarosewich E. (1983) Petrogenesis of theElephant Moraine A79001 meteorite: multiple magma pulseson the shergottite parent body. Geochim. Cosmochim. Acta 47,

1501–1513.

McSween H. Y. and Harvey R. P. (1993) Outgassed Water onMars: constraints from melt inclusions in SNC meteorites.Science 259, 1890–1892.

McSween, H. Y. and Treiman, A. (1998) Martian meteorites. InPlanet. Mater. Mineral. Soc. Am. Washington, DC, Rev.

Mineral. (ed. J. J. Papike). vol. 36, p. 53.McSween H. Y., Stolper E. M., Taylor L. A., Muntean R. A.,

O’Kelley G. D., Eldridge J. S., Biswas S., Ngo H. T. andLipschutz M. E. (1979) Petrogenetic relationship between AllanHills 77005 and other achondrites. Earth Planet. Sci. Lett. 45,

275–284.

McSween H. Y., Eisenhour D. D., Taylor L. A., Wadhwa M. andCrozaz G. (1996) QUE94201 shergottite: crystallization of aMartian basaltic magma. Geochim. Cosmochim. Acta 60, 4563–

4569.

Meyer, C. (2006) Introduction to Martian meteorites 2006. TheMars Meteorite Compendium, Astromaterials Research andExploration Science (ARES), (JSC #27672 Revision C), Lyn-don B. Johnson Space Center, Houston, Texas. Available from:<http://curator.jsc.nasa.gov/antmet/mmc/index.cfm>.

Mikouchi T. (2005) Comparative mineralogy of Chassigny andNWA2737: implications for the formation of Chassigniteigneous body(s). Meteor. Planet. Sci. 40, 5240.

Mikouchi T. and Miyamoto M. (1997) Yamoto-793605: a lherzoliteshergottite from the Japanese Antarctic meteorite collection.Antarct. Meteor. Res. Nat. Inst. Polar Res. Tokyo 10, 41–60.

Mittlefehldt, D. W. and Herrin, J. S. (2008) Petrology andgeochemistry of Martian meteorites LAR 06319, RBT 04261,and RBT 04262. Meeting Met. Soc. 71st, #5307.

Mittlefehldt D. W., Lindstrom D. J., Lindstrom M. M. andMartinez R. R. (1999) An impact-melt origin for lithology A ofMartian meteorite Elephant Moraine A79001. Meteor. Planet.

Sci. 34, 357–367.

Morgan D. and Jerram D. A. (2006) On estimating crystal shapefor crystal size distribution analysis. J. Volcan. Geotherm. Res.

154, 1–7.

Nagao, K. and Park, J. (2008) Noble gases and cosmic-rayexposure ages of two Martian Shergottites, RBT 04262 andLAR 06319, Recovered in Antarctica. Meteor. Planet. Sci. 43

Annu. Met. Soc. Meeting 71st, #5200.Neal, C.R., Taylor, L.A., Ely, J.C., Jain, J.C. & Nazarov, M.A.

(2001) Detailed geochemistry of new shergottite, Dhofar 019.Lunar Planet. Sci. Conf. 32nd, #1671.

Nishiizumi, K. and Caffee, M. W. (2006) Constraining the numberof lunar and martian meteorite falls. Meteor. Planet. Sci. 41,Annu. Met. Soc. Meeting 69th, A133.

Nyquist, L. E., Bogard, D. D., Shih, C. -Y., Greshake, A., Stoffler,D. and Eugster, O. (2001) Ages and histories of Martianmeteorites. In (Eds.) Chronology and Evolution of Mars

Springer (eds. R. Kallenbach, J. Geiss and W. K. Hartmann),pp. 105–164.

Ostertag R., Amthauer G., Rager H. and McSween H. Y. (1984)Fe3+ in shocked olivine crystals of the ALHA77005 meteorite.Earth Planet. Sci. Lett. 67, 162–166.

Reynard, B., Van de Mootele, B., Beck, P. & Gillet, P. (2006)Shock-induced transformations in olivine of the chassigniteNWA2737. Lunar Planet. Sci. Conf. 37th, #1837.

Rumble D., Miller M. F., Franchi I. A. and Greenwood R. C.(2007) Oxygen three-isotope fractionation lines in terrestrialsilicate minerals: an inter-laboratory comparison of hydrother-mal quartz and eclogitic garnet. Geochim. Cosmochim. Acta. 71,

3592–3600.

http://curator.jsc.nasa.gov/antmet/mmc/index.cfm


Sautter V., Barrat J. A., Jambon A., Lorand J. P., Gillet Ph., JavoyM., Joron J. L. and Lesourd M. (2002) A new Martianmeteorite from Morocco: the nakhlite North West Africa 817.Earth Planet. Sci. Lett. 195, 223–238.

Sharp Z. D. (1990) A laser-based microanalytical method for thein situ determination of oxygen isotope ratios of silicates andoxides. Geochim. Cosmochim. Acta 54, 1353–1357.

Shearer C., Burger P. V., Papike J. J., Borg L. E., Irving A. J. andHerd C. D. K. (2008) Petrogenetic linkages among Martianbasalts: implications based on trace element chemistry ofolivine. Meteor. Planet. Sci. 43, 1241–1258.

Spicuzza M. J., Day J. M. D., Taylor L. A. and Valley J. W. (2007)Oxygen isotope constraints on the origin and differentiation ofthe Moon. Earth Planet. Sci. Lett. 253, 254–265.

Stoffler D., Keil K. and Scott E. R. D. (1991) Shock metamorphism ofordinary chondrites. Geochim. Cosmochim. Acta 55, 3845–3867.

Symes S. J. K., Borg L. E., Shearer C. K. and Irving A. J. (2008)The age of the Martian meteorite Northwest Africa 1195 andthe differentiation history of the shergottites. Geochim. Cosmo-

chim. Acta 72, 1696–1710.

Taylor L. A. (1970) Low-temperature phase relations in the Fe–Ssystem. Carnegie Inst. Washington Yearbook 68, 259–270.

Taylor L. A. (1971) The Fe–S–O system: oxidation of pyrrhotite.Carnegie Inst. Washington Yearbook 70, 287–290.

Taylor, L. A., Onorato, P. I. K. and Uhlmann, D. R. (1977)Cooling rate estimations based on kinetic modeling of Fe–Mgdiffusion in olivine. Proc. Lunar Sci. Conf. 8th, pp. 1581–1592.

Taylor L. A., Patchen A., Taylor D. H. S., Chambers J. G. andMcKay D. S. (1996) X-ray digital imaging petrography of lunarmare soils: modal analyses of minerals and glasses. Icarus 124,

500–512.

Taylor L. A., Nazarov M. A., Ivanova M. A., Patchen A., ClaytonR. N. and Mayeda T. K. (2000) Petrology of the Dhofar 019shergottite. Meteor. Planet. Sci. 35, A155.

Taylor L. A., Nazarov M. A., Shearer C. K., McSween H. Y.,Cahill J., Neal C. R., Ivanova M. A., Barsukova L. D., LentzR. C., Clayton R. N. and Mayeda T. K. (2002) Martianmeteorite Dhofar 019: a new shergottite. Meteor. Planet. Sci.

37, 1107–1128.

Thompson, C. K, Slater, V. P., Stockstill, K. R., Anand, M.,Nettles, J., Milam, K., Cahill, J. and Taylor, L. A. (2003) Anevolution of the igneous crystallization programs – MELTS,MAGPOX, and COMAGMAT part I: does one size fit all?Lunar Planet. Sci. Conf. 34th, #1881.

Toplis M. J. (2005) The thermodynamics of iron and magnesiumpartitioning between olivine and liquid: criteria for assessingand predicting equilibrium in natural and experimental systems.Contrib. Mineral. Petrol. 149, 22–39.

Treiman A. H. (1997) Amphibole in Martian meteorite ElephantMoraine 79001. Meteor. Planet. Sci. 32, A129.

Treiman A. H., Drake M. J., Janssens M. J., Wolf R. and EbiharaM. (1986) Core formation in the Earth and shergottite parentbody (SPB): chemical evidence from basalts. Geochim. Cosmo-

chim. Acta 50, 1071–1091.

Treiman A. H., McKay G. A., Bogard D. D., Wang M. S.,Lipschultz M. E., Mittlefehldt D. W., Keller L., Lindstrom M.M. and Garrison D. (1994) Comparison of the LEW88516 andALHA77005 Martian meteorites: similar but distinct. Meteor-

itics 29, 581–592.

Treiman A. H., Gleason J. D. and Bogard D. D. (2000) The SNCmeteorites are from Mars. Planet. Space Sci. 48, 1213–1320.

Treiman, A.H., McCanta, M., Dyar, M.D., Pieters, C.M., Hiri, T.,Lane, M.D. and Bishop, J.L. (2006) Brown and clear olivine inchassignite NWA2737: water and deformation. Lunar Planet.

Sci. Conf. 37th, #1314.Treiman A. H., Dyar M. D., McCanta M., Noble S. K. and Pieters

C. M. (2007) Martian dunite NWA 2737: petrographicconstraints on geological history, shock events, and olivinecolor. J. Geophys. Res. 112, E04002.

Usui T., McSween H. Y. and Floss C. (2008) Petrogenesis ofolivine-phyric shergottite Yamato 980459, revisited. Geochim.


van de Moortele B., Reynard B., McMillan P. F., Wilson M., BeckP., Gillet P. and Jahn S. (2007) Shock-induced transformationof olivine to a new metastable (Mg, Fe)2 SiO4 polymorph inmartian meteorites. Earth Planet. Sci. Lett. 261, 469–475.

Vaughan, D. J. and Craig, J. R. (1997) Sulfide ore mineralstabilities, morphologies, and intergrowth textures. In Geo-

chemistry of Hydrothermal Ore Deposits (ed. H.L. Barnes).Wiley, pp. 367–434.

Wadhwa M. (2001) Redox state of Mars’ upper mantle and crustfrom Eu anomalies in shergottite. Science 291, 1527–1530.

Wadhwa M., McSween H. Y. and Crozaz G. (1994) Petrogenesisof shergottite meteorites inferred from minor and traceelement microdistributions. Geochim. Cosmochim. Acta 58,

4213–4229.

Wadhwa M., Lentz R. C. F., McSween H. Y. and Crozaz G. (2001)A petrologic and trace element study of Dar al Gani 476 andDar al Gani 489: twin meteorites with affinities to basaltic andlherzolitic shergottite. Meteor. Planet. Sci. 36, 195–208.

Walton E. L. and Herd C. D. K. (2007a) Dynamic crystallizationof shock melts in Allan Hills 77005: implications for meltpocket formation in Martian meteorites. Geochim. Cosmochim.

Acta 71, 5267–5285.

Walton E. L. and Herd C. D. K. (2007b) Localized shock meltingin lherzolitic shergottite Northwest Africa 1950: comparisonwith Allan Hills 77005. Meteor. Planet. Sci. 42, 63–80.

Walton E. L. and Spray J. G. (2003) Mineralogy, microtexturesand composition of shock-induced melt pockets in the LosAngeles basaltic shergottite. Meteor. Planet. Sci. 38, 1865–1875.

Walton E. L., Spray J. G. and Bartoschewitz R. (2005) A newMartian meteorite from Oman: mineralogy, petrology, andshock metamorphism of olivine-phyric basaltic shergottite Sayhal Uhaymir 150. Meteor. Planet. Sci. 40, 1195–1214.

Wang A., Kuebler K. E., Jolliff B. L. and Haskin L. A. (2004)Raman spectroscopy of Fe–Ti–Cr oxides, case study: Martianmeteorite EETA 79001. Am. Mineral. 89, 665–680.

Warren P. H., Kallemeyn G. W. and Kyte F. T. (1999) Origin ofplanetary cores: evidence from highly siderophile elements inMartian meteorites. Geochim. Cosmochim. Acta 63, 2105–2122.

Wood B. J. (1991) Oxygen barometry of spinel peridotites. Rev.

Mineral. 25, 417–431.

Zipfel J., Sherer P., Spettel B., Dreibus G. and Schultz L. (2000)Petrology and chemistry of the new shergottite Dar al Gani 476.Meteor. Planet. Sci. 35, 95–106.

Associate editor: Sara S. Russell

Documents

Petrogenesis of olivine-phyric shergottite Larkman Nunatak 06319