JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 96, NO. B8, PAGES 13,593-13,608,JULY 30, 1991

IsotopicComposition of OligoceneMafic VolcanicRocks in the Northern Rio GrandeRift: Evidencefor Contributionsof AncientIntraplate andSubduction Magmatism to Evolutionof the Lithosphere

CLARK M. JOHNSON

Departmentof Geologyand Geophysics,University of Wisconsin,Madison

REN A. THOMPSON

U.S. Geological Survey,Denver,

Mafic lavas eruptedduring initiation of regional extensionat 26 Ma in the northernRio Granderift werederived from at leasttwo isotopicallydistinct mantle sources. One is characterizedby 87Sr/86Sr= 0.70495,•md -- -4,2ø6pb/2ø4pb = 18.2, and the other by 87Sr/86Sr = 0.7044, •Nd -- 0, 2ø6pb/2ø4pb= 18.2. The 1OW-Smdvalue source (MANTLE 1) is interpretedto largelyreflect the isotopiccompositions of the lithosphericmantle in the region. Isotopiccompositions of Cenozoicmafic lavasand Proterozoicrocks areused to constrainmodels for evolutionof thelithosphere. The low 8Ndvalues of theMANTLE 1 componentwere probablyproduced by evolution of a light rare earth element-enrichedupper mantle or

inputof low-sNd material. during development of thelithosphere in theEarly . The 8Nd values and2ø6pb/2ø4pb ratios •n evolvedrocks are as low as-7 and 17.3,respectively, indicating interaction with lowercrust that had •Na -< -12 and 2ø6pb/2ø4pb _<17.0. Initial87Sr/86Sr ratios both increase and decrease slightly in evolved rocks, indicatinginteraction with lower crust that had Sr isotoperatios that were generallysimilar to thoseof themantle. The 8Nd value of themodern lithospheric mantle beneath the northernRio Grande rift is >18 units lower than the projectedmodern values of the asthenospheric mantlefrom which the Proterozoic crust was originally derived. The higher-ema value source (MANTLE 2) may reflect mixing of asthenosphereand lithospherecomponents. The mantle sourcefor most late Cenozoicmafic lavasin the northernRio Granderift regionlies at 2ø7pb/2ø4pbratios that are signifi- cantlyhigher than the 2ø6pb/2ø4pb- 2ø7pb/2ø4pb array defined by Proterozoiccrust in the region. Althoughthe crustalarray is alsodisplaced to higher2øTpb/2ø4pb ratios as comparedto the northern hemisphereoceanic mantle a•ray, indicating incorporationof Archean Pb during crust formation, the stillhigher 2øTpb/2ø4pb ratios in theRio Granderift mantlesource is interpretedto reflectcontinued input of ArcheanPb into the developingEarly Proterozoiclithospheric mantle after crust formation. These relations are strong evidence that the lithosphere became stabilized shortly after the major crust formation eventsin the Early Proterozoic.

INTRODUCTION (Figure 1) preservesthe largestvolume of early rift (26 Ma) Rift-related volcanism in the northern Rio Grande rift is volcanic rocks in the northern Rio Grande rift. These lavas largely basaltic in composition, in contrast to prerift provide an importantand previouslypoorly known view magmatismin the region, which was dominatedby inter- of volcanism that occurredduring inception of the Rio mediate- to silicic-compositionash flow tuffs and related Grande rift. lavas [Steven, 1975; Lipman and Mehnert, 1975]. Most We reportSr, Nd, andPb isotopedata for earlyrift mafic studies of rift-related volcanic rocks in the northern Rio lavas of the Hinsdale Formationthat are exposedat San Granderift have concentratedon relatively youngvolcanic Luis Hills. These data bear on mantle sources of volcanism fields that formed significantlyafter initiationof extension. associatedwith inceptionof rifting andmodels for develop- Theseinclude the TaosPlateau, Rayton-Clayton, and Ocate ment of the lithosphericmantle. Compositionalvariations fields (Figure 1) [Stormer, 1972a, 1972b; Lipman and and isotopic compositionsindicate that many San Luis Mehnert, 1979; O'Neill and Mehnert, 1988], as well as Hills magmasinteracted with lower crust,as hasbeen pro- lavas exposed at Los Mogotes , in the Tusas posedfor otherlavas in the region. Detailedconsideration Mountains, and near Amalia (Figure 1)[Lipman and of chemical and petrologic characteristicsof the lavas, Mehnert, 1975; Lipman et al., 1986]. Early rift lavas are however, allows identification of primitive, uncontami- exposedonly on intrarift horstsand on the flanks of the rift natedcompositions that can be usedto determineisotopic in the SanJuan and Latir volcanicfields (Figure 1) [Lipman compositionsof the mantle. The data define two mantle and Mehnert, 1975, 1979; Thompsonet al., 1986; Lipman sourceregions, at leastone of which is interpretedto reflect et al., 1986]. Late Cenozoic basaltic lavas in northern the lithosphericmantle that had beenenriched in light rare Coloradoare temporallycorrelative with rift-relatedvolca- earth elements(LREE) during stabilizationand growth of nic rocksbut were eruptedsignificantly outside the Rio the lithospherein the Proterozoic. Grande depression[Larson et al., 1975]. San Luis Hills

GEOLOGY AND PETROLOGY OF SAN LUIS HILLS LAVAS Copyright1991 by the AmericanGeophysical Union. San Luis Hills are the surface expressionof a major Papernumber 91JB00342. intrarift horst within the northern Rio Grande rift that is 0148-0227/91/91 JB-00342505.00 largelyburied beneath late Cenozoicsedimentary and vol-

13,593 13,594 JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS

400 ] I•'SUITE 1i A

' i ß SUITE ,.3 ß ß I- m

[ I ,,,,..."', ' , ...... • LATIR m ',,, .•. LOS ß -...... ,ooJ ,,,,,,, . mTAOS '•"•', -,,, ,.•' -" xl.•.AMALIA ...... © '-... © ,SAN LUIS o.... •*.,,,, ....•...... :•": ...... :..... :...... :..... :-*'--: .....:...... :..... •...... HILLS " B COLORADO .340• TAOS- ALKALIC ..•_.. 280 ...... -,i;,' HILLSSANLUIS •, 220 e•" ,...... © ...... • // ......

PROTEROZOIC BASEMENT 160ß1 -,- .•- - •/•'.•.•.;-",a,, ,,.•"•/ / AMALIA•F ...... MID-CENOZOIC ! '• ..'_.•.' ? ./ ...... • I I I I I 100 1_-:,'-..-".•; ß i ...... •!" ...... ß ...... o 5o IOOKM :1',, ' ...... •-X' •os - •.o.[,,•,•

Fig. 1. Generalized map of the northern Rio Grande rift region 4.0--i " ß...... ß ß ß --ß LOS ß MOGOTE$ß ß ß ß ß . ß ß ß 48 52 56 60 64 showing major Cenozoic volcanic areas, including San Luis Hills, the subjectof this report. SJVF, San Juanvolcanic field; TPVF, Taos wf%SiO 2 Plateau ; TBM, Timber Mountain and Brushy Moun- 400 • C tain; AL, Amalia lavas; LVF, Latir volcanic field; R-CVF, Rayton- = Clayton volcanic field; OVF, Ocate volcanic field; LM, Los Mogotes LOSMOGOTES • volcano.

,.300i ß ' ''' AMALIA canic rocks [Kleinkopf et al., 1970]. Other surfaceexpo- (• ...... -=/•'-...- -• i .... suresinclude Timber Mountain and BrushyMountain (Fig- ure 1) [Thompsonet al., 1986]. Early rift lavas (26.4-25.7 E200 Ma) of the lower Hinsdale Formation, the subject of this '-.•L.C... report, overlie intermediate-compositionlavas at San Luis 1O0 '"'•<.::... '- TAOS Hills that are temporally correlativewith precalderalavas of the Latir volcanic field [Thompson et al., this issue]. Silicic volcanic rocks are conspicuouslymissing from San 0 ß ß ß ß ß , ß , ß ß Luis Hills, in contrastto volcanic sectionsexposed on the 80 120 160 200 240 2•0 320 flanks of the rift in the San Juan and Latir volcanic fields. pprn Zr Particularly notable is the absence of the Amalia , Fig. 2. Chemicalvariation diagrams (a) Cr-SiO2, (b) Zr-SiO2, and which eruptedfrom the Questacaldera of the Latir volcanic (c) Cr-Zr. Symbols shown for lower Hinsdale Formation lavas field contemporaneouswith initiation of regional extension exposedat San Luis Hills, the subjectof this report;outlines for other lavasin the northernrift region also shown. The samesymbols and at 26 Ma [Lipman et al., 1986; Hagstrum and Lipman, outlinesused for Figures3, 4, 5, 6, 7, and 10. Data from Dungan et 1986]. These relations, in addition to the general lack of al. [1986],Johnsonand Lipman [1988], Thompsonet al. [this issue], age sedimentary rocks at San Luis andM. Dungan(unpublished data, 1988). Asterisk,estimated SiO 2 Hills [Thompsonand Machett, 1989], suggestthat the horst contentfor suite 3 sample205. was a topographichigh during early evolutionof the north- em Rio Grande rift. the [Lipman and Mehnert, 1975]. Early rift lavas at San Luis Hills are divided into four Suite 2 lavas contain 10-15 vol % phenocrystsof plagio- suitesbased on petrographiccharacteristics, chemical com- clase and subordinate clinopyroxene and minor olivine, positions,and stratigraphicposition. Suites 1-4 representa have slightly higher alkali contentsthan suite 1 rocks, at a general sequencefrom oldest to youngest. Suite 1 lavas given SiO2 content, and are trachybasaltsand trachyande- occur at the base of the sectionand representN60 vol % of sites. Suite 2 lavas are notable for their very high Rb and the lavas exposed at San Luis Hills. They erupted from Th contents,up to 120 and 14 ppm, respectively,as com- multiple centers and are characterizedby up to 20 vol % pared to most Hinsdale Formation lavas in the region [Lip- phenocrystsof olivine and clinopyroxene[Thompson et al., man et al., 1973]. The high Rb and Th contents,however, this issue]. These lavas are transitional between tholeiite are similar to those of precalderalavas of the nearby Latir and alkali ,as determinedby Na20+K20-SiO 2 varia- volcanic field [Johnsonand Lipman, 1988]. Suite 3 lavas tions, and are similar to other Hinsdale Formation lavas are tholeiitic basalts and . The basalts contain exposedat Los Mogotes volcano and in the easternparts of sparseand small olivine phenocrysts,whereas the andesites JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS 13,$95

TABLE 1. Sr, Nd, and Pb Isotope Data for San Luis Hills Lavas 87Sr/86Sr Rb, Sr, 87Sr/86Sr 143Nd/144Nd Sm, Nd, Measured ppm ppm Initial Measured ppm ppm t•Nd(T) 206pb/ 204pb 207pb/ 204pb 208pb/ 204pb Suite 1 T84-12 0.704890+ 9 46 706 0.704820+ 13 0.512404ñ 7 6.65 31.5 -3.97ñ0.15 17.626 15.468 37.195 72 46 971 0.512448ñ 6 8.13 37.7 -3.12ñ0.13 17.605 15.474 37.351 74 0.704882+ 7 46 780 0.704819+ 10 0.512408ñ 7 7.29 35.5 -3.88ñ0.15 17.677 15.458 37.222 88 0.705568+21 46 665 0.705494+ 25 0.512380ñ11 6.36 28.1 -4.46ñ0.23 17.988 15.535 37.585 98 0.704980+ 8 52 710 0.704902+ 12 0.512414ñ 7 6.98 33.5 -3.76ñ0.15 17.808 15.492 37.371 99 0.704853+ 8 45 901 0.704800+ 11 0.512392ñ 7 9.10 45.3 -4.18ñ0.15 17.580 15.447 37.127 152 0.705015+10 40 647 0.704949+ 13 0.512409ñ 7 7.48 36.0 -3.86ñ0.15 18.133 15.523 37.569 (0.705001+ 7) (0.704935+ 10)

Suite 2 T84-38 0.704761ñ 7 62 1179 0.704705ñ 10 0.512476ñ10 10.4 56.9 -1.91ñ0.21' 17.959 15.556 37.800 (0.704769ñ 8) (0.704713ñ 11) 126 0.704730ñ 8 48 954 0.704676ñ 11 0.512415ñ 6 8.20 42.4 -3.13ñ0.13' 17.697 15.549 37.504 (0.704716ñ10) (0.704662ñ 13) 163 0.704731ñ13 57 1187 0.704680ñ 16 0.512536ñ 7 8.92 47.7 -1.34ñ0.15 17.963 15.526 37.682 204 0.705549ñ10 109 1119 0.705445ñ 15 0.512267ñ 8 9.60 49.7 -6.60ñ0.17 17.698 15.491 37.920

Suite 3 T84-46 0.704363ñ 8 17 489 0.704326ñ 10 0.512560ñ 9 4.98 21.2 -0.97ñ0.19 17.862 15.511 37.470 89 0.704475ñ10 20 395 0.704421ñ 13 0.512593ñ 6 4.35 17.4 -0.36ñ0.14 17.784 15.493 37.353 125 0.704911ñ 8 33 971 0.704875ñ 10 0.512309ñ 8 8.10 41.4 -5.79ñ0.17 17.457 15.458 37.116 150 0.704609ñ 8 48 1014 0.704558ñ 11 0.512371ñ 7 6.53 33.3 -4.58ñ0.15 17.308 15.455 36.986 205 0.704344ñ 7 9 360 0.704317ñ 8 0.512596ñ10 4.60 18.8 +0.30ñ0.21' 18.234 15.562 37.761

Suite 4 T84-119 0.704541ñ 8 54 706 0.704459ñ 12 0.512442ñ 8 8.22 48.1 -3.14ñ0.17 17.787 15.523 37.330 140 0.704478ñ 8 39 1135 0.704441ñ 10 0.512560ñ 6 8.13 41.7 -0.89ñ0.13 17.698 15.484 37.375 140 0.704461ñ10 39 1219 0.704427ñ 12 0.512460ñ 8 8.94 47.5 -2.24ñ0.17' 17.518 15.485 37.248 Duplicate analyses(separate dissolutions) are noted in parentheses. *Nd isotopeanalyses performed after repair of amplifierhousing; measured i43Nd/144Nd not corrected for shift,but /•md (r) valuescorrected by +0.59 units. Analytical proceduresnoted in text. contain more abundant olivine and clinopyroxene phe- tents of the four San Luis Hills suites, and the high-Cr nocrysts. Suite 3 lavas have Sr and Zr contentsand chon- samplesare interpretedas primitive, mantle-derivedlavas. drite-normalized La/Yb ratios ([La/Yb]m) that are substan- tially lower than those of the other San Luis Hills suites ANALYTICAL METHODS (Figure 2 and Table 1) [Thompsonet al., this issue]. Suite 3 lavas are compositionally similar to the Servilleta Whole rock powdersanalyzed for Sr, Nd, and Pb isotope and andesiteof the Taos Plateau volcanic field [Dungan et ratios are the same as those used for chemical analyses al., 1986]. Suite 4 lavas are the youngestHinsdale Forma- reportedby Thompsonet al. [this issue]. Strontiumand Nd tion rocks exposedat San Luis Hills and are characterized were separatedusing 2.5M HC1 for Sr separation,followed by large xenocrystsof partially resorbed quartz, plagio- by groupseparation of the rare earthelements (REE) using clase, and rare clinopyroxenecrystals in otherwise apha- 6M HC1 and separationof Nd using0.150M and 0.225M 2- nitic or sparselyporphyritic lavas. This suite containsthe methyllacticacid. Lead was separatedusing 0.6M HBr and highestalkali contents,at a given SiO2 content,of the San 6M HC1 on an anion exchangecolumn. Total procedural Luis Hills suite and are trachybasaltsand trachyandesites. blanks were 200-400 pg for Sr, 40-80 pg for Nd, and 0.6- They are compositionallysimilar to the most alkaline lavas 1.5 ng for Pb, which are negligible. of the Hinsdale Formation that are exposedin the western Strontium was massanalyzed on a VG InstrumentsSec- San Juanvolcanic field [Lipman and Mehnert, 1975]. This tor 54 6-collectormass spectrometer at-•3x10 -•1A 88Sr(10 TM suite also contains the highest Zr contentsof the rocks ohm resistors)using single Ta filaments and H3PO4and a exposedat San Luis Hills, consistentwith its alkaline na- three collectortriple-jump mode (dynamicmulticollection) ture (Figure 2) [Thompsonet al., this issue]. that removes all collector biasesand beam instability fac- Several samplesof suites2 and 3 contain MgO and Cr tors. Measured ratios were exponentially corrected for contentsthat are higher than thosepredicted for their SiO2 massfractionation using 86Sr/88Sr=0.1194. Within-run er- contentsif they evolved largely by crystal fractionation rors noted (Table 1) are +2-sigma standard error (2SE) (Figure 2) [Thompsonet al., this issue], suggestingthat using n=100 (number of ratios calculated from three they representmixtures of fractionatedand primitive mag- jumps). Long-termdrift (-1 year) in the 87Sr/a6Srratio of mas. These relations are similar to those observed for other NBS-987 is negligibleusing three collector dynamic analy- mafic- to intermediate-compositionlavas in the region, sis. The a?Sr/a6Srratio measuredfor NBS-987 duringthis most notably in the Latir and Taos Plateau volcanic fields studywas 0.710215 +5 2SE (23 analyses). The a7Sr/86Sr [Dungan et al., 1986; Thompsonet al., 1986; Johnsonand ratio measuredfor BCR-1 during this study was 0.704990 Lipman, 1988]. Suite 1 lavas contain the highest Cr con- +7 2SE (11 analyses). Rubidium and Sr contents were 13,596 JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS determined by energy-dispersive X-ray fluorescence Lead isotope ratios were determined using single Re [Thompsonet al., this issue]and are preciseto +5%. Errors filaments and a mixture of dissolvedsilica gel and H3PO4 in initial 87Sr/a6Srratios (hereafterreferred to as Isr ratios) and four collector static (non peak jumping) multi- are calculatedusing a squared-sumpartial derivative ex- collection. Long-term(~ 1 year) drift in collectorbiases, as pressionof the decayequation which propagateserrors for determined by a reference voltage comparison,or static the measured 8*Sr/86Sr ratio, Rb and Sr contents, and multicollectionanalysis of Sr and Nd, is a factor of 20 less sampleage (taken as 26 +1 Ma). thanthe estimatedprecision to which Pb isotoperatios may Neodymium was mass analyzed as Nd+ (~lx10 -• A be determined (_+0.10%). Lead isotope ratios were cor- •44Nd)using double Re filamentsand a five collectortriple- rected for massfractionation by +0.10% per mass (pooled jump mode (dynamic multicollection) with exponential error,_+0.02 2SE n=13), as determinedby 2ø7pb/2ø6pband mass fractionation correction of •46Nd/•44Nd = 0.7219. 2ø8pb/2ø6pbratios measured on NBS-981 and NBS-982 Within-run errors noted (Table 1) are +2SE using n=100 standards,respectively. Standardswere run at tempera- (number of ratios calculated from three jumps). Prior to turessimilar to thoseof the samples(1100ø-1400øC). Sig- work on the amplifier housingin July 1989, the following nificant excursions from this value are found in our labora- Nd isotoperatios on standardswere determined: BCR-1, tory only for rare analyseswhere large changesin filament •43Nd/•44Nd=0.512619+8 2SE (eight analyses);internal temperatureare requiredto maintain a ~2x10'11 A 2ø8pb laboratory normal standard (Ames National Laboratory signal; mass fractionationdid not correlate with absolute ultrapuremetal), 143Nd/144Nd=0.512143+4 2SE (16 analy- temperaturewithin the range 1000ø-1500øCfor analyses ses);La Jolla Nd, •43Nd/rand=0.511859_+4 2SE (10 analy- that maintained relatively constanttemperatures. Mea- ses). We takethe 143Nd/144Ndratio of BCR-1 asequivalent sured Pb isotope ratios for NBS-981 during this study to the present-dayratio for CHUR [e.g., Wasserburget al., were 2ø6pb/2ø4pb=16.904_+8, 2ø7pb/2ø4pb=15.449_+10, 1981], which producesan •mdvalue of our internal Ames 2ø8pb/2ø4pb=36.568 _+10,and 2ø7pb/2ø6pb=0.91395_+22 2SE Nd standardof-9.29 _+0.08.For comparison,143Nd/144Nd (four analyses). MeasuredPb isotoperatios for NBS-982 measured for BCR-1 and our internal Ames Nd standard at during this study were 2ø6pb/2ø4pb=36.642 _+20, the U.S. Geological Survey, Menlo Park, using a Finnigan- 2ø7pb/2ø4pb=17.094_+13, 2ø8pb/2ø4pb=36.563_+40, and MAT 261 single-collector mass spectrometer was 2ø8pb/2ø6pb=0.99784_+57, 2SE (nine analyses). Errorsfor 0.512618 _+62SE (11 analyses) and 0.512139 _+112SE samplesare approximately 1-2 times larger than those of (seven analyses),respectively, in 1985-1987 [Johnsonet standards,based on randomreplicate analyses using differ- al., 1990]. The pooled 145Nd/144Ndratio of 34 standard ent filament loads. MeasuredPb isotoperatios are taken as analysesat University of Wisconsinwas 0.348413 _+42SE. initial ratios, because the U contents of the San Luis Hills This ratio is identicalwithin errorto thepooled 145Nd/144Nd lavas are very low (R. Thompson,unpublished data, 1987). ratio measuredon 15 San Luis Hills samplesprior to work on the amplifier housing,which is 0.348410 _+52SE. RESULTS Following work on the amplifier housingin July 1989, the 143Nd/144Ndand •45Nd/•44Nd ratios measured for our Strontium,Nd, and Pb isotopeanalyses were performed internal Ames Nd standard decreased to 0.512113 6 _+2SE on samplesthat spanthe compositionalrange of the four and 0.348402 3 _+2SE(28 analyses),respectively. This is suites. Both primitive lavas and those that have petro- equivalent to a 0.0059 _+0.0014%and 0.0032 _+0.0014% graphiccharacteristics and chemicalcompositions which decrease,respectively, and these factors have remained indicate contamination with continental crust were ana- constantthrough the end of 1990. The 143Nd/144Ndratio lyzed. measured for BCR-1 during this time interval was 0.512611 4 _+2SE(four analyses);this shift is not as greatas that measuredfor our internal Ames Nd standard,although Sr Isotope Data only four analysesof BCR-1 were made. No shift in Initial 87Sr/86Sr(Isr) ratios for mostSan Luis Hills rocks 87Sr/a6Srratios within the averageexternal reproducibility are between 0.7042 and 0.7050, similar to younger, rift- of +0.000010 was observedduring this time for NBS-987 relatedlavas exposed at Los Mogotesvolcano and the Taos and BCR-1, consistentwith the repeatedanalysis of sample Plateauvolcanic field (Figure 3a). In contrast,Isr ratios for 152 (Table 1), which was donein February 1990. Four San preextensionmafic- and intermediate-compositionrocks at Luis Hills samples(38, 126, 141 and 204; Table 1) were the San Juan and Latir volcanic fields are significantly analyzedfor Nd isotoperatios following work on the am- higher than thoseat San Luis Hills (Figure 3a) [Lipman et plifier housing, and the •mdvalues for these have been al., 1978; Colucci et al., this issue]. Isr-SiO2 relations for increasedby 0.59 units to maintain a common reference suites 1 and 2 are scattered,although Isr ratios modestly (Table 1), althoughthe 143Nd/144Ndratio is reportedas measured(Table 1). The 145Nd/144Ndratio of thesefour increasewith increasingSiO 2 contentsfor suite 3 lavas. Despiteclear petrographic evidence for crustalassimilation samplesis 0.348406 +7 2SE. in suite 4 lavas, Isr ratios for these rocks are relatively Neodymium and Sm contentswere determinedby instru- constantwith increasingS iO 2 contents. mental neutron activation analysis (INAA) [Thompsonet al., this issue], and are precise to _+5%. Present-day 147Sm/144Ndratio for CHUR taken as 0.1967 [Jacobsenand Nd IsotopeData Wasserburg,1980]. Errors in •mdvalues are calculatedin a manner similar to that used for calculating errors in Isr The •mdvalues for suites 2, 3, and 4 decreasewith in- ratios. creasing SiO2 contents (and decreasingCr and Mg con- JOHNSONAND THOMPSON: ISOTOPIC COMPOSITION OFMAFIC VOLCANIC ROCKS 13,597

0.706 "• AMALIA LATIR A that are higher than those of late-Cenozoicmafic lavas in northwesternColorado, which vary from-4 to -10 [Leat et al., 1988, 1989, 1990; Thompsonet al., 1989], but are •o 0.705 generallylower than Miocene and younger mafic lavasin thecentral and southern Rio Grande rift, whichvary from 0 '"'/ '•© ' © ß .....1' to +8 [Menzieset al., 1983;Perry et al., 1987;Roden et al., -4,',';, ', ," ...... ==...... "..... ß,._ ...... , ...... / --, 1988]. • /•' AMALIA • 0.704 z Pb IsotopeData LOS MOGOTES • SUITE 1 ß SUITE 3 •) SUITE 2 ß SUITE 4 The 2ø6pb/2ø4pbratios for suites2 and 3 decreasewith 0.703 increasingSiO 2 contents,similar to rocks at the Taos Pla- 2 teauvolcanic field (Figure3c). 2ø6pb/2ø4pb-SiO2relations for suite 1 are scattered. Evolved rocks from suite 4 that containcrustal xenocrysts have 2ø6pb/2ø4pbratios that are

- / bothlower and higher than one relatively mafic sample. The highest2ø6pb/2ø4pb ratios for rocks at San Luis Hills (•-18.2)are similar to thoseof primitivelavas that crop out nearAmalia, Los Mogotesvolcano, and at the TaosPlateau volcanicfield (Figure 3c). The 2ø6pb/2ø4pbratios of San LuisHills rocksinclude ratios that are as low (-17.3) as those found in the San Juan and Latir volcanic fields [Lipmanet al., 1978;Johnson et al., 1990;Colucci et al., this issue],although they are not as low as thosefound in 19.0 some evolved rocks of the Taos Plateau volcanic field [Dungan et al., 1986]. Lead isotopecompositions of the San Luis Hills lavas 1•.5J AMALIAi generallyform a linear array for all suiteson 2ø7pb/2ø4pb- ß :.,...... 2ø6pb/2ø4pband2øSpb/2ø4pb-2ø6pb/2ø4pb diagrams,similar to other mafic lavas in the region (Figure 4). A Pb-Pb 18.0 ...... (2ø7pb*/2ø6pb*)isochron age of 2006+540 Ma is calculated for theSan Luis Hills lavas(excluding sample 126 of suite e -. .... 2, whichhas an anomalously high 2ø7pb/2ø4pb ratio), similar to thosethat may be calculatedfor othermafic lava suitesin +'.... , ...... thenorthern Rio Granderift: precalderaLatir, 1387340 Ma 17.o • -"'"'• [Johnsonet al., 1990];Amalia, 1439 +1200 Ma [Johnsonet 48 52 56 60 64 al., 1990];Taos Plateau, 1760 +240 Ma [Dunganet al., wt• SiO2 1986]; extension-related(<26 Ma) Hinsdale, Brazos, Rayton-Clayton,and Ocate,1402 +480 Ma (C. Johnsonet. Fig. 3. Variationsin (a) initial87Sr/86Sr (ISr) ratio, (b)/•Nd value, and al., unpublisheddata, 1989). Within the errorsnoted, these (c) 2ø6pb/2ø4pbratio with SiO2 contentsof San Luis Hills lavas (symbols) and other lavas in the region (outlines). Data from "ages"are the sameas the 2ø7pb*/2ø6pb*agecalculated for Dungan et al. [1986], Johnsonet al. [1990], M. Dungan and S. exposedProterozoic rocks in northernNew Mexico of 1710 Moorbath(unpublished data, 1990), and M. Dunganand J. Davidson +60 Ma (J. Woodenet. al., unpublisheddata, 1987; ages (unpublisheddata, 1988). Asterisk,estimated SiO 2 content for suite anderrors (2-sigma) calculated using ISOPLOT [Ludwig, 3 sample205. 1988]). All Cenozoicmafic lavas in thenorthern Rio Granderift, tents) but are relatively constantfor most suite 1 lavas with the exceptionof primitive(high-eNd) lavas from Los (Figure3b). The high Cr and Mg contentsof the mafic Mogotesvolcano, have Pb isotopecompositions that scat- suite1 lavassuggest that they are relatively primitive, and ter abouta 2ø7pb/2ø4pb-2ø6pb/2ø4pbarraythat generally lies we interpretthe averageend value of-4 as indicativeof a above that definedby exposedProterozoic rocks in the mantlevalue. The SanLuis Hills lavashave Nd isotope region(Figure 4; 2ø6pb/2ø4pb=17to 33 (J. Woodenet. al., ratios that overlapthose of lavas at Amalia and the Taos unpublisheddata, 1978)). The 2ø7pb/2ø4pb-2ø6pb/2ø4pbPro- Plateauvolcanic field (Figure3b). Primitivelavas at Los terozoiccrustal array in turn lies abovethe field for north- Mogotesvolcano have slightly higher end values (•- +1 to em hemisphereoceanic lavas (Figure 4b). The 2øSpb/2ø4pb +2) thanthe highest found at SanLuis Hills (Figure3), as ratiosof all lavasin theregion are significantly higher than do Quaternarylavas in theTusas Mountains (Brazos basalt thoseof theProterozoic crustal array (Figure 4). [Williams,1984]) and mafic feldspathoidallavas in the Rayton-Claytonfield [Phelps et al., 1983].The lowest end ORIGINS OF ISOTOPIC VARIATIONS valuesat San Luis Hills overlapthose of intermediate- compositionrocks at the San Juan and Latir volcanic fields Isotopiccompositions of SanLuis Hills lavasgenerally [Johnsonet al., 1990;Colucci et al., this issue]. Mafic overlapthose of othermafic lavas in theregion. Isotopic lavasof thenorthern Rio Grande rift regionhave end values variationsof theSan Luis Hills rocks most closely resemble 13,598 JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS

38.0 0.70475 0.70500 ' ' ' ' 18.2 I SUITE1 B;'21 A 19.0 37.8 IMAIN TREND E4Z.157t I I- ...... LO-;...... "I . 7eo..15e / 18.5 ...... MOG 0---TE$ ":l i, m• œ0.1706...155.1',•' , 17.6 37.6 13_ .-'"" ...... '.... i ! .z 1 0 ...... 12' 37.4 • 18.o- - -/._._ I AMALIA i 14, .....ß / I-ATIR 37.2 o 17.5 __,:• 4` '"',,,,,,, 37.0

15.7 ß 17.0 ß ß ß ß ß I 0.703 0.704 0.705 0.706 LOS 0 Mo MOGOTES INITIAL87 Sr/86 Sr

15.6 Fig. 5. Initial 87Sr/86Sr(/Sr) - 2ø6pb/2ø4pbvariations for SanLuis Hills and other lavas in the region. Inset for the main trend of suite 1 also shownwith expandedscale. Size of boxesin insetnote 2SE analyti- cal error. "A" refers to Sr isotopeanalysis of sample 152 prior to 15.5 work on amplifier housing, and "B" refers to repeated analysis (dotted line) following work on amplifier housing. Decreasing 87Sr/86Sr- 2ø6pb/2ø4pb trend for mostsuite 1 lavasis interpretedas ,• • • • • I•IORB/HA•'AII reflectingassimilation of lower crustthat had 87Sr/86Srratios lower than that of the mantle. Numbers in italics in inset note Sr, Zr 15.4 contents. "1", "2", and "3" next to open boxes in main figure note 17.0 17.5 18.0 18.5 19.0 MANTLE 1, 2, and 3 compositions,respectively. Data sourcescited 206 Pb/ 204 Pb in Figure 3. Fig. 4. (a) 2ø6pb/2ø4pb- 2ø8pb/2ø4pb and (b) 2ø6pb/2ø4pb- 2ø?pb/2ø4pb variationsfor San Luis Hills lavas and otherlavas in the region. Data sourcescited in Figure 3. "1,2" and "3" next to large open squares decreaseslightly during early fractionationdepending upon indicate MANTLE 1 and 2 and MANTLE 3 compositions,respec- tively, discussedin text. Stacey-Kramersaverage crustal Pb evolu- the phasesthat are fractionated,trace elements such as Sr tion curve shown in heavy line [Stacey and Kramers, 1975]. and Zr are better indicators of fractionation in the olivine- Proterozoic crustal array regressionline (1710 60 Ma) shown by and clinopyroxene-phyriclavas of suite 1. stippledbar. Low 2ø8pb/2ø4pbratios of the crustindicate a Th/U ratio Variationsin 2ø6pb/2ø4pbratios with SiO2 or traceelement of-2 (J. Wooden et al., unpublisheddata, 1987). Data for MORB and Hawaiian lavas from Tatsumoto [1978], Cohen et al. [1980], contentsfor the majority of San Luis Hills rocks indicate Duprg and All•gre [1980], Cohen and O'Nions [1982], Stille et al. interaction with crust that had 2ø6pb/2ø4pbratios less than [1983], Staudigel et al. [1984], and Hegner et al. [1986]. 17.3. A possibleexception is one evolved samplefrom suite4 (119), whichhas a relativelyhigh 2ø6pb/2ø4pb ratio of 17.8 (Table 1). The generally small variation in lsr ratios those of the Taos Plateau volcanic field, suggestingthat with SiO2 contents for suites2, 3, and4 suggestsinteraction evolution of the two magmatic systemsinvolved similar with crust that had 87Sr/g6Srratios that were similar to those endmembercomponents. An important exceptionis the of the parentalmagmas. One samplefrom suite2 that hasa occurrenceof petrologically primitive lavas at San Luis low /•mdvalue and 2ø6pb/2ø4pbratio (204), suggestiveof Hills that have low/•Ndvalues of ~ -4. crustalinteraction, has a relatively high lsr ratio, indicating interaction with crust that had an 87Sr/86Srratio > 0.706. Crustal Interaction Most samplesfrom suite 1 indicate interactionwith crust that had an 87Sr/g6Srratio < 0.7045. Decreasing/•Nd values with increasingSiO 2 contentsfor Although some variability in isotopic compositionsof suites2, 3, and 4 is interpretedto reflect coupledassimila- the crustalcomponents in the San Luis Hills lavas is indi- tion/fractional crystallization (AFC [Taylor, 1980; catedby the scatterin the data,the generalisotopic compo- DePaolo, 1981a]) involving crust that had low/•Ndvalues, sitionsof thesecomponents are remarkablysimilar to those such as Proterozoic crust in the region [DePaolo, 1981b; proposedas crustalcontaminants in several other volcanic Nelson and DePaolo, 1984, 1985]. The presenceof crustal sequencesin the northernRio Granderift region. Contami- componentsin suites2, 3, and 4 is also indicatedby de- nation by crust that has 87Sr/86Sr=0.7045to 0.7070, /•Nd creasing2ø6pb/2ø4pb ratios with increasingSiO 2 contents for <-10, and2ø6pb/2ø4pb < 17.3 has been proposed for evolved suites 2 and 3, in addition to the common occurrence of rocks near Amalia and at Timber Mountain and Brushy crustalxenocrysts in suite 4. Although isotopicvariations Mountain, as well as in the San Juan, Latir, and Taos relative to SiO2 contentsfor suite 1 are scattered,Sr and Pb Plateau volcanic fields [Doe et al., 1969; Lipman et al., isotope ratio and trace element variations indicate some 1978; Williams, 1984; Dungan et al., 1986; Thompsonet interaction with crustal rocks (Figure 5). Except for one al., 1986; Johnson et al., 1990; Riciputi and Johnson, sample (88) that has an anomalously high lsr ratio, 1990; Colucci et al., this issue]. Crust that has low /•Nd 2ø6pb/2ø4pband lsr ratios decreasewith increasingdifferen- valuesand 2ø6pb/2ø4pbratios has been invoked as a major tiation, as indicatedby increasingSr and Zr contents. Inas- componentin volcanic and plutonic rocks of the Colorado much as SiO 2 contents of mafic magmas can increase or Mineral Belt region north of San Luis Hills [Stein, 1985; JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS 13,599

TABLE 2. Assimilation/FractionalCrystallization Models Suite 1 Suite 2 Suite 3 Suite 4 Parental MANTLE 1 MANTLE 2 MANTLE 2 MANTLE 2 composition (sample152) (extrapolatedfrom (sample205) (extrapolatedfrom sample 163) sample 140) a?Sr/a6Sr 0.70495 0.7044* 0.7044 0.7044 ENd -4 0 0 0 2ø6pb/2ø4pb 18.2 18.2 18.2 18.2 Most contaminated Main trend (sample204) (samples125 (sample 119) composition (sample99) and 150) a?Sr/a6Sr 0.70480 0.7055 0.7049 0.7045 ENd -4 -7 -6 -3 2ø6pb/2ø4pb 17.6 17.7 17.3 17.5 Crustal 8?Sr/86Sr 0.7030-0.7040 0.7060-0.7080 0.7060-0.7080 R <1/10 R=l/4 to 1/2 R = 1/4 to 1/2 negligible 30-50% AFC 40-80% AFC 40-70% AFC assimilation Ma/Mo = 0.03-0.07 Ma/Mo = 0.27-0.70 Ma/Mo = 0.10-0.23 required CrustalENd negligible - 12to - 15 - 12to - 15 -12 to -15 assimilation R ~ 1/2 R = 1/4 to 1/2 R~l/4 required 50-60% AFC 40-70% AFC 50-70% AFC Ma/Mo = 0.50-0.60 Ma/Mo = 0.15-0.40 Ma/Mo = 0.17-0.25 Crustal 2ø6pb/2ø4pb 16.0-17.0 ~17.0 16.0-17.0 16.0-17.0 R ~ 1/10 R = 1/4 to 1/2 R ~ 1/4 R = 1/10 to 1/4 30-50% AFC 20-50% AFC 50-70% AFC 50-70% AFC Ma/Mo = 0.03-0.10 Ma/Mo > 0.15 Ma/Mo = 0.10-0.20 Ma/Mo = 0.06-0.20 R is massratio of assimilatedcrust to crystallizedminerals [DePaolo, 1981a]. Ma/Mo is massratio of assimilatedcrust to initial massof magma[Farmer and DePaolo, 1983]. AFC refersto coupledassimilation/fractional crystallization. * Choiceof 87Sr/86Sr= 0.7047 (ratiofor sample163) for initial compositiondecreases calculated amount of assimilation basedon Sr isotopevariations by ~10%.

Stein and Crock, 1990; Johnson and Fridrich, 1990]. Crust The majorityof samplesin suite1 haveisotopic compo- of this nature apparentlycomprised the major crustalcom- sitions which indicate minimal interaction with the crust. ponentin Cenozoicmagmatism in the northernRio Grande AFC calculations indicate that the most contaminated rift regionthat spannedmore than 40 m.y. andcovered over sample(99) of the main trend of decreasingIsr and 100,000 km2. 2ø6pb/2ø4pbratios contains <10 wt % crust (Table 2). AFC models summarized in Table 2 are constrained to Sample152 has a 2ø6pb/2ø4pbratio that is closeto that realistic percentagesof crystallization(_<50% for basaltic estimated for the mantle beneath the northern Rio Grande rocks; _<90%for andesiticrocks). Isotopic compositionsof rift (below), and this suggeststhat it containsvery little crustal contaminantsare those that provide the best fits to crust. the data and provide similar rangesof percentassimilation for severalisotope ratios. Although the Sr isotoperatios of Mantle Reservoirs crustal contaminants are not well constrained, ENdvalues and 2ø6pb/2ø4pbratios of the crust are considered to be Two distinct mantle sources can be identified in the San substantiallylower than the lowest values and ratios mea- Luis Hills lavas, based on extrapolationto noncontami- suredin the San Luis Hills rocks in order to producereason- natedcompositions (Figures 3 and5) andconsideration of able AFC models. AFC calculations indicate that the most correlatedisotopic compositions (Figures 4-7). contaminatedsample of suite 2 (204) contains30-50 wt % crust (Table 2). The proportionof crust is best constrained for suite 2 by Nd isotope variations becauseof the large 19.0 rangein Nd isotopecompositions. Although AFC calcula- LOS WOGOTES tions indicate that the tholeiite basalts of suite 3 contain ...... relatively little crust (<10 wt %), the low ENdvalues and ..o 18.5 MANTLE( ...... MANTLE MANTLE 2ø6pb/2ø4pbratios of the andesites(samples 125 and 150) 1 ...... 2 ...... indicatethat they contain 10-25 wt % crust (Table 2). The ...... El',,, I-I most contaminatedsample of suite 4 (119) contains 10-20 • 18.0 wt % crust, based on AFC calculations (Table 2). AFC •.,.* ' '...... --•-(•-ß --'•, .." ..,"AMALIA calculationsare best constrainedfor suite 4 by Nd and Pb isotope variations, which indicate interaction with crust 17.5 ...... -;?_'2 __ -, thathas low ENdvalues and 2ø6pb/2ø4pb ratios. The relatively ..-'" ß ....--'"' •,•___ ...../ restrictedrange of Isr ratios for suite 4 lavas may reflect ?'" ...... -...... ""TAo, AMALIA assimilation of crust that had 87Sr/86Sr ratios that were 17.0 -8 ß -'6 ' -'4 ' -'2 ' (• ' 2 similar to thoseof the parentalmagmas; AFC calculations based on Sr isotope ratio variations alone do not require ENd significantcrustal assimilation, although assimilation is in- Fig. 6. ENd- 2ø6pb/2ø4pb variations for SanLuis Hills andother lavas dicatedby other isotoperatios and petrographicevidence. in the region. Data sourcescited in Figure 3. 13,600 JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS

2 • :"-.."•-LOSIdOGOTES ing early volcanismat San Luis Hills (suite 1) affd is not found in any younger rift-related volcanic rocks. The ...... MANTLE 2 componentwas the sourcefor suites2, 3 and o. • i"'•...... I.J"-¾,. I • I •,..• '"•"•-,. probably 4 at San Luis Hills, as well as primitive lavas at ß / V •.•../ b, • •", /..:' © \",, the Taos Plateau volcanic field and alkali-olivine basalts "o ...... 2 } '"'",,, AMALIA near Amalia (Figures 5-7). A third mantle sourceis indi- Z AIdALIA cated for Los Mogotes volcano, which has 87Sr/86Sr-• 0.7038 and end'*+2 ("MANTLE 3"; Figures 5-7); this is similarto the sourcepostulated for the Brazosbasalts in the Tusas Mountains and mafic feldspathoidal lavas at the Raton-Clayton volcanic field [Phelps et al., 1983; Wil- liams, 1984]. A basanitenear Amalia has an anomalously t8 ß .' ß ß ß ß ß ß ß 0.703 '0.•04 '0.•05 '0.706 high Isr ratio (0.706), which may reflect an additional source(Figures 5 and 7). INITIAL87 St/ 86 Sr The MANTLE 1 componenthas Sr and Nd isotoperatios Fig. 7. Initial 87Sr/86Sr(/Sr) - •Ndvariations for San Luis Hills and that are similar to thoseof the EMI reservoirproposed by otherlavas in the region. "1.... 2", and "3" next to openboxes note Zindler and Hart [1986], Hart et al. [1986], and Hart MANTLE 1,2, and3 compositions,respectively. Data sourcescited [1988],although ihe 2ø6pb/2ø4pbratio of MANTLE 1 is in Figure 3. somewhathigher than that of EMI. The endvalues of lavas largely derived from the oceanicasthenosphere (MORB), Based on detailed study of primitive and cont•aminated as well as virtuallyall oceanislands, are significantly Hinsdale Formation lavas exposed in the San Juan and higher than the end value of MANTLE 1, leading us to Jemez volcanic fields, Doe et al. [1969] conclude that the ascribe MANTLE 1 as a lithosphericcomposition. The mantlesource region for the basaltshad a 2ø6pb/2ø4pbratio higher endvalues of MANTLE 2 and 3 componentsthat are of-* 18.2. This ratio is similar to that measuredfor the suite the sourcesfor younger lavas in the northern rift may 1 sample that has the highest Cr and lowest Sr and Zr reflect a mixture of asthenosphereand lithospheresources contents(152) and the suite 3 samplethat has the highest that occurredduring continued extension, as hasbeen inter- ENdvalue and lowest SiO2 content (205). The SiO2- pretedfor rift-relatedlavas in the centraland southernparts 2ø6pb/2ø4pband eNd-2ø6pb/2ø4pbvariations for suite2 lavas of the Rio Grande rift [Perry et al., 1987, 1988]. That the define similar trends as compared to other lavas in the 2ø7pb/2ø4pband 2ø6pb/2ø4pbratios of MANTLE 1 and 2 do region and at San Luis Hills, suggestingthat the mantle not lie in the field for oceanic lavas in the northern hemi- sourceregions for suite2 magmashad a 2ø6pb/2ø4pbratio of spherefurther supporta lithosphericcomponent for these 18.0-18.2 (Figures 3 and 6). Although the highest sources. 2ø6pb/2ø4pbratio for suite 4 lavas is 17.8, the contaminated nature of these lavas preclude estimationof a mantle Pb COMPOSITION OF THE LOWER CRUST isotopecomposition. The mantlesource regions for primi- tive lavas exposednear Amalia and at the Taos Plateau The isotopiccompositions of contaminatedlavas in the volcanicfield alsohad a 2ø6pb/2ø4pbratio of-•l 8.2 [Dungan northernRio Granderift region,including those at SanLuis et al., 1986; Johnson et al., 1990]. This ratio lies within the Hills, bear on current debatesregarding the isotopiccom- rangemeasured for mantlexenoliths from California,Ari- positionsof the lower crust. As notedabove, most lavas in zona, and New Mexico [Zartmanand Tera, 1973; Galer the northern rift region have Pb isotope compositions and O'Nions, 1989; Meijer et al., 1990]. An important which indicatea crustalcomponent that has 2ø6pb/2ø4pb characteristic of the mantle source for northern Rio Grande ratios that are less than 17.0. Recent studies of lower rift mafic lavas, with the exceptionof lavas at Los Mogotes crustal ,however, have suggestedthat the lower volcano, is its markedly higher 2ø7pb/2ø4pbratio, at crusthas relativelyhigh 2ø6pb/2ø4pbratios that lie to the 2ø6pb/2ø4pb-18.2,as comparedto northernhemisphere oce- rioht c•ftho ooc•chrcm [o o •'•n•,rrtnYt •,t rtl 1 QRR' anic basalts("MANTLE 1, 2"; Figure 4). Moreover, the and Goldstein, 1990; Kempton et al., 1990]. Many lower mantle source for the majority of nodhern rift lavas crustal suites have been interpreted to represent ("MANTLE 1, 2"; Figure4) lies at a 2ø7pb/2ø4pbratio that recent magmatic underplating [Rudnick et al., 1986; plots abovethe 2ø7pb/2ø4pb-2ø6pb/2ø4pbarray defined for Rudnick and Taylor, 1987; Rudnick and Williams, 1987; Early Proterozoiccrust in the region(Figure 4). In contrast, Kemptonet al., 1990; K.L. Cameronet al., Granite-facies primitive lavas at Los Mogotes volcano have the lowest xenolithsfrom north central Mexico' Evidencefor a major 2ø7pb/2ø4pbratios of primitivelavas in the northernrift, and pulse of mid-Cenozoiccrustal growth, submittedto Jour- thissource has Pb isotopecompositions that are similar to nal of GeophysicalResearch, 1991], which may indicate thoseof mid-oceanridge basalts(MORB) in the northern that lower crustwhich has high 2ø6pb/2ø4pbratios is more hemisphereand Hawaiian lavas ('ZMANTLE '3; Figure 4). commdnin regionsof recenttectonic or magmaticactivity Distinction of the two mantle sources for San Luis Hills [e.g., Rudnickand Goldstein,1990]. Some lower crustal lavas lies in their Sr and Nd isotopecompositions (Figures xenolithsuites that have high 2ø6pb/2ø4pbratios are inter- 5-7). Extrapolationto primitive compositionsindicate one pretedto representlower crustof Proterozoicage, suchas sourcethat has87Sr/S6Sr -• 0.70495 and end"• -4 ("MANTLE at CampCreek, [Esperanca et al., 1988], although 1; Figures5-7), and a secondsource that has 875r/a6Sr a Proterozoic age for these xenoliths has been debated 0.7044 and end'•' 0 ("MANTLE 2"; Figures 5-7). The [Esperancaet al., 1990; Johnson,1990]. MANTLE 1 sourceappears to have been tappedonly dur- The necessityof a nonradiogeniccrustal componentin JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANIC ROCKS 13,601 evolved igneousrocks of the northern Rio Grande rift re- isotopecompositions of mafic lavasin the northernRio gion highlights the importance of Precambrian crust that Grande rift region are discussedbelow in the context of haslow 2ø6pb/2ø4pbratios. Other studiesof evolved,young models for evolution and stabilizationof the lithospheric igneous rocks that were emplaced in Precambrian crust mantle. Moreover, the exceptionaldata base of Nd and Pb supportthe presenceof low 2ø6pb/2ø4pblower or middle isotoperatios and REE concentrationsthat is available for crust [e.g. Doe et al., 1968; Peterman et al., 1970; Dickin, Proterozoic rocks in Colorado and New Mexico provides 1981; Doe et al., 1982]. Many studieshave shown that U an important temporal framework for interpretationof the depletionis characteristicof granulite-graderocks, includ- isotopiccompositions of Cenozoiclavas. ing high-gradecrustal xenoliths [e.g., Dostal and Capedri, 1978; Fowler, 1986; Rudnickand Taylor, 1987; Reid et al., Nd Isotope Constraints 1989; Whitehouse, 1989; Kempton et al., 1990], and we endorsethe view that the lower crustshould have generally Tholeiite lavas that are LREE depleted are common in low U/Pb ratios [Zartman and Doe, 1981; Zartman and lower Proterozoic sections exposed in New Mexico and Haines, 1988]. The importanceof Precambriancrust that Colorado. These include the Tijeras, Pecos, and Dubois has low 2ø6pb/2ø4pbratios is demonstratedby the fact that greenstonesequences [Condie and Budding, 1979; Condie, suchcrust is found in high-gradeterranes on every conti- 1980; Condie and Nuter, 1981; Nelson and DePaolo, 1984; nent [e.g., Moorbath et al., 1969; Gray and Oversby, 1972; Knoper and Condie, 1988; Robertsonand Condie, 1989]. If Moorbath et al., 1975; Sobotovich et al., 1973; Leeman, theserocks are productsof large degreesof partial melting 1979; Griffin et al., 1980; Tilton and Barreiro, 1980; of mantle that did not contain garnet as a residual phase, DePaolo et al., 1982; Bernard-Griffiths et al., 1984; Cohen then the high measuredmSm/•44Nd ratios would be ap- et al., 1984; Ovchinnikova et al., 1987; van Calsteren et al., proximatelythe sameas thoseof the mantle sourceregions. 1988]. Assumingthat this mahtle was incorporated into the lith- The relatively low ENdvalues for volcanic rocks at San osphere and evolved as a closed system, the average Luis Hills that containa large crustalcomponent (Table 2), present-day /•mdvalue would be ~ +7 (Figure 8). If the in addition to lavas in the northern rift region that are LREE-depletedlavas were generatedby relatively small similarly contaminated[Johnson et al., 1990; Riciputi and degreesof partial melting of mantle that containedresidual Johnson,1990; Colucci et al., this issue], suggestinterac- garnet,the mSm/144Ndratio of the sourcewould be sub- tion with crustthat had present-day/•mdvalues that were less stantially higher than that measuredfor the lavas, and this than -12. If the low 2ø6pb/2ø4pbratios of many of these would requirea very high averagepresent-day/•md value for rocks is indicative of interaction with lower crust, then the the mantle (> +12; Figure 8). present-day /•mdvalues of the upper and lower crust in Many Early •Proterozoicmafic lavas have relatively high northernNew Mexico and southernColorado may be simi- LREE contents,however, and melting calculations predict lar [DePaolo, 1981b; Nelson and DePaolo, 1985]. If these that thesewere derivedfrom mantle sourceregions that had conclusionsare generallyapplicable to the crustin western relativelylow •47Sm/•44Ndratios (Figure 8). For example, North America, they standin contrastto recent Nd isotope assumingthat the calc-alkalinePecos lavas were derivedby studies of lower crustal xenoliths in western North 30% partial melting with garnet as a residual phase, the America. Several studiesof xenolithshave suggestedthat average 147Sm/144Ndratio of the mantle sourceregion the Proterozoiclower crust may have /•mdvalues that are would be 0.15-0.16, which would producea present-day significantlyhigher than thoseof exposedProterozoic crust /•mdvalue of-7 during1750 m.y. of evolution(Figure 8). [e.g., Esperanca et al., 1988; Ruiz et al., 1988]. Critical to Gabbroic parts of the Pikes Peak batholith in Colorado this interpretation,however, is the evidencethat the xeno- appearto havebeen derived from mantle source regions liths representlower crust that is of Proterozoic age, and that had low mSm/144Ndratios [Barker et al., 1976], which this debate continues [Cameron and Robinson, 1990; would producelow present-day/•mdvalues (Figure 8). Esperanca et al., 1990; Johnson, 1990; Ruiz et al., 1990]. Although the earliest Proterozoic mafic lavas in Colo- rado and New Mexico were derived from a depleted mantle, as indicatedby the high 147Sm/•44Ndratios mea- EVOLUTION AND STABILIZATION sured for some lavas, and the uniformly high /•mdvalues OF THE LITHOSPHERIC MANTLE calculatedfor all Proterozoiclavas (Figure 9a) [Nelsonand DePaolo, 1984], we proposethat continuedmafic mag- A large number of studiesof mafic lavas in the western matism became more LREE enriched with time. Geo- United Stateshave interpretedlow 143Nd/144Ndratios as graphicallyextensive 1700-1600 Ma plutonicrocks in New diagnosticof an ancientlithosphere mantle source,similar Mexico and Colorado [e.g., Condie and Budding, 1979] to our interpretationof the MANTLE 1 componentat San probably reflect significant magmatic thickening of the Luis Hills [e.g., Menzies et al., 1983; Vollmer et al., 1984; lithosphere,and LREE-enrichedmafic magmastend to be Fraser et al., 1985; Dudas et al., 1987; Carlson and Hart, most commonin regions that are underlainby relatively 1987; Fitton et al., 1988; Ormerod et al., !988; Farmer et thick lithosphere[e.g., Pearce, 1982, 1983). LREE enrich- al., 1989; Thompson et al., 1989; Kempton et al., this ment of the lithosphericmantle may have occurredby issue]. Although the mechanisms for generating low metasomatismassociated with Early and Middle Protero- •43Nd/•44Ndratios in the lithosphericmantle often are not zoic magmatism,in addition to crystallizationof LREE- clear, metasomaticand subductionprocesses are some of rich basaltic magmasin the uppermostmantle. It seems the more common mechanismsthat have been proposed likely that the abundantLREE-enriched crustal magmatism [e.g., Ormerod et al., 1988; Fitton et al., 1988; Menzies, that occurred shortly after initial crust formation in the 1989; Kempton et al., this issue]. Neodymium and Pb region was associatedwith an increasein LREE contents 13,602 JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS throughout the entire lithospheric column as it became als to the uppermantle [Arndt and Goldstein,1989; John- isolated from the convecting asthenosphericmantle (Fig- son et al., 1990]. We note, however, the importanceof ure 9b). carefullyassessing the effectsof crustalcontamination in Input of low endvalue crustal material through subduc- the younglavas. Severalstudies have suggestedthat de- tion in the Proterozoic would also generate an average creasingend values with increasingLa/Nb ratios may re- evolution of the lithosphericmantle toward low present- flect the presenceof an ancient subductioncomponent day endvalues (Figure 9c). Evolution of the lithospheric [e.g., Fitton et al., 1988; Kemptonet al., this issue],al- mantle to low end values may also occur if Proterozoic thoughthis sametrend is producedin the San Luis Hills mantle-derived magmas stalled at the crust-mantlebound- suiteby crustalcontamination (Figure 10). ary and assimilated lower crust. Crustal recycling may The low SNdvalues for late Cenozoic mafic lavas in occurby this mechanismthrough return of cumulateminer- northwest Colorado (-4 to -10) [Leat et al., 1988, 1989, 1990; Thompson et al., 1989] may reflect lithospheric mantle that has remained isolated from interaction with the EVOLVED FROM DEPLETED MANTLE AT 1750 Mo asthenospheresince the Proterozoic. The lowest endvalue for mafic lavas from the northern Rio Grande rift that (o) representmantle valuesis -4 (MANTLE 1 component,this -15 -5 5 15 25 study), and this is interpreted to largely reflect a lithos- I I I I I I I I I pheric mantle source, although it may also represent a mixture of Nd derived from lithospheric (low endvalue) and asthenospheric(high endvalue) mantle. 1000 Mo PIKES PEAK BATHOLITH 4 • MEASUREDSLL1½1½ PART 2 • MEASUREDMAFICPART Pb Isotope Constraints

0 --! m ß ß ß m m m m ß m m m m m m m m mmm m The 2ø?pb*/2ø6pb*isochron "ages" of mafic lava suitesin lO B the westernUnited Stateshave beenused to infer the age of 8 the underlyinglithosphere [e.g., Doe et al., 1982; Dudas et 6 PECOS al., 1987]. Becauserocks from the northern Rio Grande rift

4 BASALT thathave 2ø6pb/2ø4pb < -• 18.2 are interpretedto haveassimi- lated crust that had non-radiogenicPb isotope ratios, the 2 • MEASURED 2ø?pb*/2ø6pb*isochron "ages" calculatedfor these rocks 0 ß ßm ß ß ß ß ß ß ß• ß ß ß• ß ß CALC-ALKALINEß ß ß ß ß ß m place no constraintson the age of the lithosphericmantle. That the 2ø?Pb*/2ø6pb*isochron "ages" calculated for most 15 m OTHER northern rift lavas are similar (within large errors) to the J • MAFICLAVAS age of the Proterozoicbasement in the region simply re- flects the combinationof a large differencein 2ø6pb/2ø4pb 5 ratios of the crust and mantle componentsand the fact that 0 ...... ;";"•"•, .... the mantle sourcehas Pb isotope ratios that plot near the 12 crustalarray (Figure 4). These featuresproduce contami- 10 m DUBOlS D nation trends that are sub-parallelto the crustal array. In 8 I GREENSTONE contrast,a steeptrend on a 2ø?pb/2ø4pb-2ø6pb/2ø4pbdiagram (i.e., very old 2ø?pb*/2ø6pb*"age") may be producedby 4 mixing between a mantle componentthat has Pb isotope 2 ratios that lie off the crustal array and crust that has 0 ß ß ß ß ...... ß ....

15 PECOS E THOLElITE > 0.3 lO LAVAS 10 • MEASURED Fig. 8. Histogramsof 147Sm/•44Ndratios measured for Early Prot- 5 erozoic mafic rocks exposedin New Mexico and Colorado(coarse diagonalpattern) and those calculated for their mantlesource regions 0 ...... ß ß ß ß ß ß ß ß assuming5% melting(fine diagonalpattern) and 30 % melting(solid F pattern). Silicic rocks of Pikes Peak batholith also shown (no pat- lO TIJERAS tern). Predictedpresent-day ENd values of mantle sourcesshown at GREENSTONE top, assumingderivation from depletedmantle at 1750 Ma [DePaolo, 1981b]. Melting calculationsassume modal equilibrium melting 5 • MEASURED with 10% garnet, 30% clinopyroxene,and 60% olivine. Mineral/ melt distribution coefficients from Grutzeck et al. [1974], Shimizu and Kushiro [1975], Irving and Frey [1978, 1984], Fujimaki et al. ß ß [1984], and Green and Pearson [1985]. Nonmodal melting involv- O.11 0 1• 0.21 0.26 0.31 ing garnet or melting without garnet will shift calculated source 147Srn/144 Nd compositionstoward lower 147S1TIf•44Nd ratios and would produce lower present-dayENd values. Data for Proterozoicrocks from Barker et al. [1976], Condieand Budding[1979], Condie[1980], MIINTL E SOURCE Condie and Nuter [1981], Condie and McCrink [1982], Nelson and MELTING m 30 % MELTING DePaolo [ 1984], Boardman and Condie [ 1986], Knoper and Condie [ 1988], and Robertson and Condie [ 1989]. JOHNSONAND THOMPSON' ISOTOPIC COMPOSITION OFMAFIC VOLCANIC ROCKS 13,603

^•E (•o) •.------' .... 0 500 1000 1500 2000 •' SUITE4 ß •'•-•

\ l• "-'•.,.• '•.• x \ -2

_•. -4 •.•.•__.•'•

-8 -6 '•'•.•.•.•.• -12 ...... - PLUTONICANDI ...... /- • I- OT...oc.sI 0.5 ' ' 2;5 ' s55 ' 455 ' 5.5 -16 La/Nb EVOLUTIONBY / • B Fig. 10. end- La/Nbrelations for SanLuis Hills lavas(fields added for emphasis). San Luis Hills data define similar trendson a z 2ø6pb/2ø4pb-La/Nbdiagram; these variations indicate that increases in La/Nb ratiosare largely due to crustalcontamination. The crustal componentprobably had high La/Nb ratios,suggesting that it was LREE rich. This is consistentwith the low endvalues that are inferredfor the crustalcontaminant. eNd-La/Nb variations at San EVOLUTION BY SUBDUCTION - Luis Hills overlapthose of othermafic lavasin the westernUnited States[e.g., Fitton et al., 1988].

z towardhigh 2ø7pb/2ø4pbratios of Cenozoicarc-related and behind-arc lavas in the western [Church, 1976; Carlson, 1984; Carlson and Hart, 1987]. The steep trenddefined by Early Proterozoicore leads from Colorado Fig. 9. eNaevolution diagrams showing (a) netevolution of conti- nentalcrust and lithosphericmantle, (b) interpretiveevolution by and New Mexico cannot representan isochronage, but episodicintrusion into the lithosphericmantle of LREE-enriched insteadprobably reflects mixing betweensubducted Pb of magmas,and (c) interpretiveevolution by episodiccontamination of --1750 Ma Stacey-Kramersaverage crust composition and lithosphericmantle by ancientsubduction. High endvalues for Early asthenosphericmantle (Figure 11). Proterozoicmafic lavas indicatethat the early crust was originally That the MANTLE 1 and 2 componentshave Pb isotope underlainby depletedmantle, which would have evolvedto very highpresent-day end values had this mantleremained isolated be- compositionswhich plot abovethe crustalarray (Figures 4 neaththe crust. Cenozoiclavas exposed in northwestColorado and and 11) placeimportant constraints on modelsfor develop- the lowestend lavas in the northernRio Granderift regionmay ment of the Proterozoiclithosphere. Subductionthat con- largelyrepresent a lithosphericmantle composition, whereas higher tinued after initial crust formation may have resulted in endlavas in the northernrift may representmixtures of asthenosphericand lithospheric mantle. Solidboxes in Figure9a at greater contaminationof the developing lithospheric end= -4, 0, and +2 representMANTLE 1, MANTLE 2, and mantle with 2ø7pbas compared to the crust. Extensive MANTLE 3 reservoirsfor the northernRio Granderift, respectively, contaminationof the lithosphericmantle could producea asdiscussed in text andprevious figures. Data for Proterozoicrocks lithosphericmantle compositionthat lies abovethe Prot- from DePaolo [1981b] andNelson and DePaolo [1984, 1985]. erozoiccrustal array on a 2ø7pb/2ø4pb-2ø6pb/2ø4pbdiagram (Figure11). If themajor flux of crustalPb occurredat 1750 Ma, the contaminated lithospheric mantle would have 2ø6pb/2ø4pbratios that are close to those of the mantle. This evolved with a 238U/2ø4pbratio of 8.1 ("MANTLE 1 and 2"; maybe the bestexplanation for the Pb isotopevariations at Figure 11). This is slightly lower than the ratio of 9.7 Los Mogotes volcano(Figure 4). calculatedfor the mantle sourcefor Los Mogotes volcano Assumingthat the Proterozoiccrust and lithospheric ("MANTLE 3"; Figure 11), which may representastheno- mantle in northern New Mexico and Colorado ("MANTLE sphericmantle. 1 and 2"; Figures4 and 11) were originallyderived from the asthenosphericmantle, such as that represented by oce- SUMMARY AND CONCLUSIONS anic lavasin the northernhemisphere, their relativelyhigh 2ø7pb/2ø4pbratios require explanation. We proposethat Pb Earlyrift volcanismin the axisof theRio Grandedepres- which had a high 2ø7pb/2ø4pbratio, similar to that of the sion in northern New Mexico and Colorado was more mafic Stacey-KramersPb evolutioncurve at 1750 Ma, wastrans- and contained significantly smaller crustal components portedinto the mantleduring arc-related magmatism that than volcanicrocks that were eruptedon the flanks of the occurredduring the Early Proterozoicin Coloradoand New rift immediatelyprior to extension.Restriction of mafic Mexico (Figures11 and 12) [e.g., Condie, 1982]. This is andisotopically primitive early rift lavasto theaxial part of similar to modelsproposed for Pb isotopeevolution of the rift zone suggeststhat the modemrift depressionwas Precambrian crust in other areas of the western United theprimary locus of mantleupwelling at 26 Ma. Primitive States[Wooden and Mueller, 1988; Woodenet al., 1988]. lavas at San Luis Hills indicatethat at least two isotopically The sources of this Pb could be subducted oceanic sedi- distinct mantle sources were involved in early rift mentsor crustalmaterial from the adjacentArchean craton magmatism. One source,which is characterizedby (Figure12), similarto sourcesproposed to explainshifts 87Sr/86Sr=0.70495,end = -4, and 2ø6pb/2ø4pb=18.2,has 13,604 JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS

15.7 a 1750 Ma MANTLE 1 2 15.6 LOW143Nd •t [ HIGHLOW207pb143Nd 15.5 HIGH207pbf •.

15.4 1750 Ma S-K COUP MANTLE3 HAWAII 15.3

$UBDUCTION ZONE 15.2 I ATMIXING 1750 INMe 1750 Me ASTHENOSPHERE 15.1 15.0 16.0 17.0 18.0 19.0

206 Pb/204 Pb ASTHENOSPHERE

Fig. 11. 2ø6pb/2ø4pb-2ø7pb/2ø4pbevolution diagram illustrating con- tamination(mixing) of subcontinentalmantle with Archean crustal Pb duringEarly Proterozoicsubduction, noted by arrowin lower left. b 1600 Ma Composition of 1750 Ma asthenospheric(oceanic) mantle from Plumbotectonicsmodel of Zartman and Doe [ 1981]. Compositionof LOW 143Nd Pb contaminantis thought to be similar to that of 1750 Ma Stacey- LOW143Nd •t HIGH207pb Kramers(S-K) averagecrust [Staceyand Kramers, 1975]. Evolution of asthenosphericmantle to MANTLE 3 composition requires a HIGH207pb 238U/2ø4pbratio of 9.7, whereas evolution of contaminated litho- HEAN sphericmantle that had Stacey-Kramerscomposition at 1750 Ma to present-dayMANTLE 1 and 2 compositionrequires a 238U/2ø4pb .- X,TM x• '• i•x x , ••"•,, ..._-'- ratio of 8.1. This later evolution is interpretedto representthat of the lithospheric mantle, following formation and stabilization in the • }-- - -,_Y_.2,T,,'?'.'" ------M9 HO Early Proterozoic,and its isochronand evolutionpath are shownin dashed lines. Array for Proterozoic crustal rocks (double heavy wr/_///?2)!• '"'• Z3eu _. lines) lies betweenthe lithosphericmantle isochronand the field for w•:•J• LoL• =8'1 oceanic lavas (J. Wooden, et. al., unpublisheddata, 1987). Field of Early Proterozoic ore leads in Colorado and New Mexico from Stacey et al. [1977] and Stacey and Hedlund [1983].

ASTHENOSPHERE ]43Nd/144Ndratios that are slightly higherthan thosemea- sured for late Cenozoic lavas in northwest Colorado that eruptedoutside the rift, and this sourceis thoughtto largely reflect a lithospheric mantle composition. The other 1000 TO 1500 Ma source,which is characterizedby 87Sr/a6Sr=0.7044, and2ø6pb/2ø4pb=l 8.2, may reflect a mixtureof lithospheric

238U -I0 and asthenosphericmantle compositions. " ARCHEAN Comparisonof the isotopic compositionsof Cenozoic 147Sm lavas and those of Proterozoic crust in the region provide 144N•=O'IZ !!?:••'.;;- CRATON i.u important constraints on evolution of the lithosphere. 238U X X X X Lo Lu• W 204pb- I X X ? MOHO Neodymium isotope compositionsof Early Proterozoic 147Sm 0 mafic lavas in New Mexico and Colorado indicate that 238144Nd U =0-12TO 0.16 !• 9 9 depleted mantle underlay the earliest crust in the region 204p•=8'1 •1 [Nelson and DePaolo, 1984], which presumablyreflects 147Sm the major mantle sourcefor magmatismthat existedprior to - =0.24 144Nd formation of a stable lithospheric mantle (Figure 12a). L,'gr..::;:.:! ASTHENOSPHERE 238U During the major period of Early Proterozoicsubduction in = 9.7 Lo Lu 204pb the region,progressively larger volumes of LREE-enriched magmas were emplaced into the developing lithosphere Fig. 12. Interpretive cross sectionsthrough the crust and upper [e.g., Condie and Budding, 1979, Condie, 1982, 1986; mantle illustrating formation and stabilizationof the Proterozoic Condie and McCrink, 1982; Knoper and Condie, 1988; lithosphericmantle in Colorado and New Mexico. (a) Period of Robertsonand Condie, 1989], producinga decreasein the initial crustformation in primitive arc setting,(b) period of intense average147Sm/]44Nd ratio of the lithosphericmantle (Figure orogenic plutonism in mature arc setting, and (c) anorogenic plutonism. Lithosphereformation and stabilizationis thoughtto be 12b). Melting calculations indicate that the average accompaniedby contaminationof the subcontinentalmantle with •47Sm/•44Ndratio of the mantlesource region for the LREE- Archean crustal Pb, either from subducted sediments or material enrichedmagmas may have been NO.15-0.16, which would from the adjacentArchean craton (Figures 12a and 12b). A net produce an average present-day/•md value for the lithos- increasein the LREE contentof the lithosphericmantle is thoughtto have occurredduring major episodesof LREE-enrichedorogenic pheric mantle of-5 to -7, similar to Cenozoic lavas in (---1600Ma) and anorogenic(1000-1500 Ma) plutonism. Magma northwest Colorado and rift-related lavas of suite 1 at San intrusionand generationshown in solid pattern,crustal melting in Luis Hills. Anorogenic,intraplate plutonism occurred be- crosspattern, and plutonicbodies in hatcheredpattern. JOHNSONAND THOMPSON: ISOTOPIC COMPOSITION OF MAFIC VOLCANIC ROCKS 13,605 tween 1500 and 1000 Ma in the region [e.g., Condie and volcanicfield. In addition,Nd modelages for the SanLuis Budding, 1979] and may have also led to a decreasein the Hills lavas, which vary from 718 to 1240 Ma (calculated average147Sm/144Nd ratio of thelithospheric mantle (Figure relativeto depletedmantle [De?aolo, 1981b]),probably do 12c), as indicatedby the presenceof LREE-enrichedmafic not recordmeaningful age information,due to the effectsof rocks in at least the Pikes Peak batholith [Barker et al., heterogeneousisotopic compositionsof the mantle and 1976). The •Sdvalue of the modem lithosphericmantle in fractionationin Sm/Nd ratiosduring melt generation. the region is >18 units lower than that of the projected Volumetrically subordinatelavas in suites2, 3, and 4 at modemvalue for the asthenosphericmantle from which the San Luis Hills assimilated substantial amounts of crust and Proterozoiccrest was originally derived. If the lithospheric suggest that the lower crust beneath the northern Rio mantle evolved toward low present-day•Sd values, the Granderift regionhas low 87Sr/86Srand 2ø6pb/2ø4pb ratios of amount of crustal recycling proposedfor the genesisof 0.703-0.704 and 16.0-17.0, respectively. The •Sdvalues of Middle Proterozoicplutons has been markedly overesti- the lower crust that was assimilated must be _<-12,indicat- mated [DePaolo, 1981b). ing that althoughthe crustin the region is stronglyzoned in Based on isotopic variations in mafic lavas, a number of Rb/Sr and U/Pb ratios, it is relatively homogeneousin workershave proposedthe existenceof low 2ø6pb/2ø4pbSm/Nd ratios and hencepresent-day •Sd values. mantle beneath the continents [e.g., Doe et al., 1982; Fraser et al., 1985; Peng et al., 1986; Dudt•s et al., 1987; Acknowledgements. Peter Lipman and Michael Dungan are Thompson et al., 1989; Hawkesworth et al., 1990a; thanked for discussionson rift-related magmatism and for use of Kempton et al., this issue]. Direct evidence for low unpublisheddata. Joe Wooden reviewed an early version of the manuscript,and Phil Lear and Rick Carlson provided journal re- 2ø6pb/2ø4pblithospheric mantle is providedby xenolithand views. JenniferCousins, Brian Beard, and Lee Riciputi assistedin kimberlite samples, although most studieshave been re- data acquisitionand contributedtoward keeping the University of strictedto regions of Archean lithosphere[e.g., Kramers, Wisconsinradiogenic isotope laboratory running. Michelyn Hass 1977, 1979; Smith, 1983; Cohen et al., 1984; Menzies et al., assistedin manuscriptpreparation, and Paul Dombrowski drafted 1987; Walker et al., 1989; Hawkesworth et al., 1990b]. Figures 1 and 12. Acknowledgmentis made by Johnsonto the Donors of The Petroleum Research Fund, administeredby the Analysesof clinopyroxenemineral separatesfrom mantle American Chemical Society, and the National ScienceFoundation xenolithsfrom the westernUnited Statesdo not supporta (EAR-861836 and EAR-8803892), for supportof this research. model for exceptionallylow 2ø6pb/2ø4pbmantle in the re- gion [Zartman and Tera, 1973; Galer and O'Nions, 1989; Meijer et al., 1990], although the presence of a sulfide REFERENCES phasemay be an importantlow 238U/2ø4pbcomponent in the Arndt, N.T., and S.L. Goldstein,An openboundary between lower mantle [Kramers, 1979; Meijer et al., 1990]. Many studies continental crust and mantle: Its role in crust formation and have highlightedthe extremesensitivity of Pb isotopecom- crustalrecycling, Tectonophysics,161,201-212, 1989. positionsof mafic magmasto modification by crustal con- Barker, F., C.E. Hedge, H.T. Millard, Jr., and J.R. O'Neil, PikesPeak batholith: Geochemistryof someminor elementsand isotopes, tamination [e.g., Doe et al., 1969; Dickin, 1981; Cortini and implicationsfor magma genesis,Prof. Contrib. Colo. Sch. a•id van Calsteren,1985], andwe urgecaution in interpret- Mines, 8, 44-53, 1976. ing the very low 2ø6pb/2ø4pbratios measured for somemafic Bernard-Griffiths, J., J.J. Peucat, B. Postaire, P. Vidal, J. Convert, lavas in the westernUnited Statesas unambiguouslyrepre- and B. Moreau, Isotopicdata (U-Pb, Rb-Sr, Pb-Pb and Sm-Nd) on mafic granulitesfrom Finnish Lapland, PrecambrianRes., 23, senting a nonradiogenic lithospheric mantle. Although 325-348, 1984. many lavas at San Luis Hills have near-primitive Sr and Nd Boardman,S.J., and K.C. Condie,Early Proterozoicbimodal volca- isotopecompositions, only two have Pb isotoperatios that nic rocksin centralColorado, U.S.A., Part II: Geochemistry, probablyreflect their mantle sourceregions. While com- petrogenesisand tectonic setting, Precambrian Res., 34, 37-68, pelling argumentsmay be made for the insensitivity to 1986. Cameron,K.L., and J.V. Robinson,Comments on "Nd-Sr isotopic crustalcontamination of Sr and Nd isotoperatios in alka- compositionsof lowercrustal xenoliths - evidencefor theorigin of line rocks,which may containup to 10 times greaterSr and mid-Tertiaryfelsic volcanics in Mexico",by J. Ruiz, P.J.Patchett, Nd concentrationsas compared to average crust [e.g., and R.J. Arculus, Contrib. Mineral. Petrol., 104, 609-614, 1990. Vollmer et al., 1984; Fraser et al., 1985; Dudt•s et al., Carlson,R.W., Isotopicconstraints on ColumbiaRiver flood basalt 1987], few mafic lavas contain Pb concentrations that are genesis and the nature of the subcontinentalmantle, Geochim. Cosmochim. Acta, 48, 2357-2372, 1984. as high as the crest, including stronglyalkaline rocks, mak- Carlson,R.W., and W.K. Hart, Crustalgenesis on the OregonPla- ing Pb isotoperatios susceptibleto modification by even teau, J. Geophys.Res., 92, 6191-6206, 1987. small amounts of crustal contamination. Church, S.E., The Cascade Mountains revisited: A re-evaluation in The only valid constrainton the age of stabilizationof the light of new lead isotopicdata, Earth Planet. Sci. Lett., 29, 175- 188, 1976. lithosphericmantle in the northernRio Granderift region is Cohen,R.S., andR.K. O'Nions,The lead,neodymium and strontium the fact that the Pb isotopecomposition of most primitive isotopicstructure of oceanridge basalts,J. Petrol., 23,299-324, lavas lies on a 2ø7pb/2ø4pb-2ø6pb/2ø4pbisochron that lies 1982. above the Proterozoiccrustal array and intersectsthe high- Cohen, R.S., N.M. Everson, P.J. Hamilton, and R.K. O'Nions, U-Pb, Sm-Nd and Rb-Sr systematicsof mid-oceanridge basaltglasses, est2ø7pb/2ø4pb ratios for EarlyProterozoic ore leads. The Nature, 283, 149-152, 1980. high 2ø7pb/2ø4pbratios are bestexplained through incorpo- Cohen,R.S., R.K. O'Nions,and J.B. Dawson,Isotope geochemistry ration of Archean Pb in a Proterozoic subduction zone. of xenolithsfrom East Africa: Implicationsfor developmentof Lead-leadisochrons of the mafic lavasdo not reflect "ages" mantle reservoirs and their interaction, Earth Planet. Sci. Lett., 68, in the senseof recording a time of U/Pb fractionation in 209-220, 1984. Colucci,M.T., M.A. Dungan,K.M. Ferguson,P.W. Lipman,and S. their sourceregions, but insteadare artifactsproduced by Moorbath, Precaldera lavas of the southeast San Juan volcanic assimilationof lower crust, a conclusionalso reachedby field: Parentmagmas and crustalinteractions, J. Geophys.Res., Dungan et al. [1986] in their study of the Taos Plateau this issue. 13,606 JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS

Condie, K.C., The Tijeras Greenstone:Evidence for depletedupper Cenozoic basic magmasin the westernUnited States,J. Petrol., mantle beneathNew Mexico during the Proterozoic,J. Geol., 88, Spec.Lithosphere Issue, 331-349, 1988. 603-609, 1980. Fowler, M.B., Large-ion lithophile element characteristicsof an Condie, K.C., Plate-tectonics model for continental accretion in the amphibolitefacies to granulitefacies transition at GruinardBay, southwesternUnited States,Geology, 10, 37-42, 1982. north-westScotland, J. Metamorph. Geol., 4, 345-359, 1986. Condie, K.C., Geochemistryand tectonic setting of Early Protero- Fraser, K.J., C.J. Hawkesworth, A.J. Erlank, R.H. Mitchel, and B.H. zoic supracrustalrocks in the southwesternUnited States,J. Geol., Scott-Smith, Sr, Nd, and Pb isotopeand minor element geochem- 94,845-864, 1986. istry of lamproitesand kimberlites, Earth Planet. Sci.Lett., 76, 57- Condie, K.C., and A.J. Budding, Geology and geochemistryof 70, 1985. Precambrian rocks, central and south-central New Mexico, Mem. Fujimaki, H., M. Tatsumoto,and K. Aoki, Partition coefficientsof N.M. Bur. Mines Miner. Resour., 35, 1-58, 1979. Hf, Zr and REE between phenocrysts and groundmass, J. Condie, K.C., andT.P. McCrink, Geochemistryof Proterozoicvolca- Geophys.Res., 89, 662-672, 1984. nic and granitic rocks from the Gold Hill-Wheeler Peak area, Galer, S.J.G., and R.K. O'Nions, Chemical and isotopic studiesof northern New Mexico, Precambrian Res., 19, 141-166, 1982. ultramafic inclusions from the San Carlos volcanic field, Arizona: Condie, K.C., and J.A. Nuter, Geochemistryof the Dubois green- A bearingon their petrogenesis,J. Petrol., 30, 1033-1064, 1989. stonesuccession: An Early Proterozoicbimodal volcanic associa- Gray, C.M., and V.M. Oversby,The behaviorof lead isotopesduring tion in west-central Colorado, Precambrian Res., 15, 131-155, granulitefacies metamorphism,Geochim. Cosmochim. Acta, 36, 1981. 939-952, 1972. Cortini, M., and P.W.C. van Calsteren, Lead isotope differences Green, T.H., and N.J. Pearson,Rare earth element partitioning be- betweenwhole-rock and phenocrysts in recentlavas from southern tween clinopyroxeneand silicate liquid at moderateto high pres- Italy, Nature, 341,343-345, 1985. sure, Contrib. Mineral. Petrol., 91, 24-36, 1985. DePaolo, D.J., Trace element and isotopic effects of combined Griffin, W.L., V.R. McGregor, A. Nutman, P.N. Taylor, and D. wallrock assimilationand fractional crystallization,Earth Planet. Bridgwater,Early Archeangranulite-facies metamorphism south Sci. Lett., 53, 189-202, 1981 a. of Ameralik, West Greenland, Earth Planet. Sci. Lett., 50, 59-74, DePaolo, D.J., Neodymium isotopesin the Colorado Front Range 1980. and crust-mantle evolution in the Proterozoic, Nature, 291, 193- Grutzeck,M., S. Kridelbaugh,and D. Weill, The distributionof Sr 196, 1981b. andREE betweendiopside and silicateliquid, Geophys.Res. Lett., DePaolo, D.J., W.I. Manton, E.S. Grew, and M. Halpern, Sm-Nd, 1,273-275, 1974. Rb-Sr and U-Th-Pb systematicsof granulite facies rocks from Hagstrum,J.T., and P.W. Lipman, Paleomagnetismof the structur- Fyfe Hills, EnderbyLand, Antarctica,Nature, 298, 614-618, 1982. ally deformedLatir volcanic field, northernNew Mexico: Rela- Dickin, A.P., Isotopegeochemistry of Tertiary igneousrocks from tions to formation of the Questa and developmentof the the Isle of Skye, N.W. Scotland,J. Petrol., 22, 155-189, 1981. Rio Granderift, J. Geophys.Res., 91, 7383-7402, 1986. Doe, B.R., R.I. Tilling, C.E. Hedge, and M.R. Klepper, Lead and Hart, S.R., Heterogeneousmantle domains: Signatures,genesis and strontiumisotopes studies of the Boulder Batholith, southwestern mixing chronologies,Earth Planet. Sci. Lett., 90, 273-296, 1988. Montana, Econ. Geol., 63, 884-906, 1968. Hart, S.R., D.C. Gerlach,and W.M. White, A possiblenew Sr-Nd-Pb Doe, B.R., P.W. Lipman, C.E. Hedge, and H. Kurasawa,Primitive mantle array and consequencesfor mantle mixing, Geochim. and contaminated basalts from the southern Rocky Mountains, Cosmochim. Acta, 50, 1551-1557, 1986. U.S.A., Contrib. Mineral. Petrol., 21,142-156, 1969. Hawkesworth,C.J., P.D. Kempton,N.W. Rogers,R.M. Ellam, and Doe, B.R., W.P. Leeman, R.L. Christiansen,and C.E. Hedge, Lead P.W. van Calsteren,Continental mantle lithosphere,and shallow and strontiumisotopes and related trace elementsas genetictrac- level enrichmentprocesses in the Earth's mantle, Earth Planet. ers in the upper Cenozoic -basalt associationof the Sci. Lett., 96, 256-268, 1990a. Yellowstone Plateau volcanic field, J. Geophys.Res., 87, 4785- Hawkesworth, C.J., A.J. Erlank, P.D. Kempton, and F.G. Waters, 4806, 1982. Mantle metasomatism:isotope and trace-elementtrends in xeno- Dostal, J., and S. Capedri, Uranium in metamorphicrocks, Contrib. liths from Kimberley, South Africa, Chem. Geol., 85, 19-34, Mineral. Petrol., 66, 409-414, 1978. 1990b. Dudils,F.O., R.W. Carlson,and D.H. Eggler,Regional middle Prot- Hegner, E., D. Unruh, and M. Tatsumoto,Nd-Sr-Pb isotopecon- erozoic enrichmentof the subcontinentalmantle sourceof igneous straints on the sources of West Maui volcano, Hawaii, Nature, rocks from central Montana, Geology, 15, 22-25, 1987. 319, 478-480, 1986. Dungan, M.A., M.M. Lindstrom, N.J. McMillan, S. Moorbath, J. Irving, A.J., and F.A. Frey, Distributionof trace elementsbetween Hoefs, and L.A. Haskin, Open systemmagmatic evolution of the garnet megacrystsand host volcanic liquids of kimberlitic to Taos Plateauvolcanic field, northernNew Mexico, 1, The petrol- rhyolitic composition,Geochim. Cosmochim. Acta, 42, 771-787, ogy and geochemistryof the Servilleta Basalts,J. Geophys.Res., 1978. 91, 5999-6028, 1986. Irving, A.J., andF.A. Frey, Trace elementabundances in megacrysts Dupr6, B., and C.J. All•gre, Pb-Sr-Nd isotopic correlationand the and their host basalts: Constraintson partition coefficientsand chemistryof the North Atlantic mantle,Nature, 286, 17-22, 1980. megacrystgenesis, Geochim. Cosmochim. Acta, 48, 1201-1221, Esperanca,S., R.W. Carlson,and S.B. Shirey,Lower crustalevolu- 1984. tion undercentral Arizona: Sr, Nd and Pb isotopicand geochemi- Jacobsen,S.B., and G.J. Wasserburg,Sm-Nd isotopicevolution of cal evidence from the mafic xenoliths of Camp Creek, Earth chondrites, Earth Planet. Sci. Lett., 50, 139-155, 1980. Planet. Sci. Lett., 90, 26-40, 1988. Johnson,C.M., Comment on "Lower crustal evolution under central Esperanca,S., R.W. Carlson, and S.B. Shirey, Reply to "Comment Arizona: Sr, Nd, and Pb isotopicand geochemicalevidence from on Lower crustal evolution under central Arizona: Sr, Nd and Pb the mafic xenoliths of Camp Creek," by S. Esperanca,R.W. isotopicand geochemicalevidence from the mafic xenolithsof Carlson, and S.B. Shirey, Earth Planet. Sci. Lett., 99, 400-405, Camp Creek" by C.M. Johnson,Earth Planet. Sci. Lett., 99, 406- 1990. 409, 1990. Johnson, C.M., and C.J. Fridrich, Non-monotonic chemical and O, Farmer, G.L., and D.J. DePaolo, Origin of Mesozoic and Tertiary Sr, Nd, andPb isotopezonations and heterogeneity in the mafic-to granite in the western United States and implications for pre- silicic-compositionmagma chamberof the Grizzly Peak Tuff, Mesozoic crustal structure,1, Nd and Sr isotopic studiesin the Colorado, Contrib. Mineral. Petrol., 105, 677-690, 1990. geoclineof the northernGreat Basin,J. Geophys.Res., 88, 3379- Johnson,C.M., and P.W. Lipman, Origin of metaluminousand alka- 3401, 1983. line volcanic rocks of the Latir volcanic field, northern Rio Grande Farmer, G.L., F.V. Perry, S. Semken,B. Crowe, D. Curtis, and D.J. rift, New Mexico, Contrib. Mineral. Petrol., 100, 107-128, 1988. DePaolo,Isotopic evidence on the structureand origin of subcon- Johnson,C.M., P.W. Lipman, and G.K. Czamanske,H, O, Sr, Nd, tinentallithospheric mantle in southernNevada, J. Geophys.Res., and Pb isotope geochemistryof the Latir volcanic field and 94, 7885-7898, 1989. cogeneticintrusions, New Mexico, and relationsbetween evolu- Fitton, J.G., D. James, P.D. Kempton, D.S. Ormerod, and W.P. tion of a continentalmagmatic center and modificationsof the Leeman, The role of lithosphericmantle in the generationof late lithosphere,Contrib. Mineral. Petrol., 104, 99-124, 1990. JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANIC ROCKS 13,607

Kempton,P.D., R.S. Harmon,C.J. Hawkesworth,and S. Moorbath, Moorbath, S., H. Welke, and N.H. Gale, The significanceof lead Petrologyand geochemistryof lower crustalgranulites from the isotope studies in ancient, high-grade metamorphic basement Geronimo volcanic field, southeastern Arizona, Geochim. complexes,as exemplifiedby the Lewisianrocks of northwest Cosmochim. Acta, 54, 3401-3426, 1990. Scotland,Earth Planet. Sci. Lett., 6, 245-256, 1969. Kempton,P.D., J.G. Fitton, C.J. Hawkesworth,and D.S. Ormerod, Moorbath, S., R.K. O'Nions, and R.J. Pankhurst,The evolution of Isotopicand trace elementconstraints on the compositionand early Precambriancrustal rocks at Isua, West Greenland-Geo- evolution of the lithospherebeneath the southwesternUnited chemicaland isotopic evidence, Earth Planet.Sci. Lett., 27, 229- States,J. Geophys.Res., this issue. 239, 1975. Kleinkopf,D.M., D.L. Peterson,and R.B. Johnson,Reconnaissance Nelson, B.K., and D.J. DePaolo, 1,700-Myr greenstonevolcanic geophysicalstudies of the Trinidad quadrangle,south-central successionsin southwesternNorth Americaand isotopic evolution Colorado,U.S. Geol. Surv. Prof. Pap., 700B, B78-B85, 1970. of Proterozoic mantle, Nature, 312, 143-146, 1984. Knoper,M.W., andK.C. Condie,Geochemistry and petrogenesis of Nelson, B.K., and D.J. DePaolo, Rapid productionof continental Early Proterozoicamphibolites, west-central Colorado, U.S.A., crust 1.7-1.9 b.y. ago: Nd and Sr isotopic evidence from the Chem. Geol., 67, 209-225, 1988. basement of the North American mid continent, Geol. Soc. Am. Kramers, J.D., Lead and strontium isotopes in Bull., 96, 746-754, 1985. kimberlites and mantle-derived xenoliths from southern Africa, O'Neill, J.M., and H.H. Mehnert, Petrology and physiographicevo- Earth Planet. Sci. Lett., 34, 419-431, 1977. lution of the Ocate volcanic field, north-central New Mexico, U.S. Kramers,J.D., Lead, uranium,strontium, potassium and rubidium in Geol. Surv. Prof. Pap., 1478, parts A and B, 45 pp., 1988. inclusion-bearingdiamonds and mantle-derivedxenoliths from Ormerod, D.S., C.J. Hawkesworth,N.W. Rogers,W.P. Leeman, and southern Africa, Earth Planet. Sci. Lett., 42, 58-70, 1979. M.A. Menzies, Tectonic and magmatic transitionsin the western Larson, E.E., M. Ozima, and W.C. Bradley, Late Cenozoic basic Great Basin, USA, Nature, 333,349-353, 1988. volcanismin northwesternColorado and its implicationsconcern- Ovchinnikova, G.V., C.Z. Yakovleva, A.N. Komarov, O.A. ing tectonismand the origin of the ColoradoRiver system,Mem. Levchankov, and N.I. Moskovchenko, Study of U-Pb isotopic Geol. Soc. Am., 144, 155-178, 1975. geochemical system of the Enderbit gneisses in the upper Leat, P.T., R.N. Thompson,M.A. Morrison, G.L. Hendry, and A.P. Nelarbinsky granulite block of the Stanov shelf region (in Rus- Dickin, Compositionally-diverseMiocene-Recent rift-related sian), in Geochronologyand Geochemistryof Isotopes,edited by magmatismin northwestColorado: Partial melting,and mixing of L.K. Levsky, and O.A. Levchenkov, pp. 171-186, Institute of mafic magmasfrom 3 different asthenosphericand lithospheric Geology and PrecambrianGeochronology, Academy of Sciences mantlesources, J. Petrol., Spec.Lithosphere Issue, 351-377, 1988. of USSR, Leningrad, 1987. Leat, P.T., R.N. Thompson,A.P. Dickin, M.A. Morrison, and G.L. Pearce, J.A., Trace element characteristics of lavas from destructive Hendry, Quaternaryvolcanism in northwesternColorado: Impli- plate boundaries,in Andesites,edited by R.S. Thorpe, pp. 525- cationsfor the roles of athenosphereand lithospherein the genesis 548, John Wiley, New York, 1982. of continental basalts, J. Volcanol. Geotherm. Res., 37, 291-310, Pearce,J.A., Role of sub-continentallithosphere in magmagenesis at 1989. active continental margins, in Continental Basalts and Mantle Leat, P.T., R.N. Thompson,M.A. Morrison,G.L. Hendry, and A.P. Xenoliths, edited by C.J. Hawkesworth and M.J. Norry, pp. 208- Dickin, Geochemistryof mafic lavas in the early Rio Granderift, 249, Shiva, Cambridge, Mass., 1983. Yarmony Mountain, Colorado,U.S.A., Chem. Geol., 81, 23-43, Peng, Z.C., R.E. Zartman, K. Futa, and D.G. Chen, Pb-, Sr- and Nd- 1990. isotopic systematics and chemical characteristics of Cenozoic Leeman, W.P., Primitive lead in deep crustal xenoliths from the basalts, eastern China, Chem. Geol., 59, 3-33, 1986. Snake River Plain, Idaho, Nature, 281,365-366, 1979. Perry, F.V., W.S. Baldridge, and D.J. DePaolo, Role of astheno- Lipman, P.W., and H.H. Mehnert,Late Cenozoicbasaltic volcanism sphere and lithospherein the genesisof late Cenozoic basaltic and developmentof the Rio Grande depressionin the southern rocks from the Rio Grande rift and adjacentregions of the south- Rocky Mountains,Mem. Geol. Soc.Am., 144, 119-154, 1975. westernUnited States,J. Geophys.Res., 92, 9193-9213, 1987. Lipman, P.W., and H.H. Mehnert, The Taos Plateauvolcanic field, Perry, F.V., W.S. Baldridge, and D.J. DePaolo, Chemical and isoto- northem Rio Grande rift, New Mexico, in Rio Grande Rift: Tec- pic evidencefor lithosphericthinning beneath the Rio Granderift, tonics and Magmatism, edited by R.E. Reicker, pp. 289-311, Nature, 332,432-434, 1988. AGU, Washington,D.C., 1979. Peterman, Z.E., B.R. Doe, and H.J. Prostka, Lead and strontium Lipman, P.W., C.M. Bunker, and C.A. Bush, Potassium,thorium, isotopes in rocks of the Absaroka volcanic field, Wyoming, and uraniumcontents of upper Cenozoicbasalts of the southern Contrib. Mineral. Petrol., 27, 121-130, 1970. Rocky Mountains region, and their relation to the Rio Grande Phelps,D.W., D.A. Gust, and J.L. Wooden,Petrogenesis of the mafic depression,U.S. Geol. Surv.J. Res., 1,387-401, 1973. feldspathoidal lavas of the Raton-Clayton volcanic field, New Lipman, P.W., B.R. Doe, C.E. Hedge, and T.A. Steven, Petrologic Mexico, Contrib. Mineral. Petrol., 84, 182-190, 1983. evolution of the San Juan volcanic field, southwestern Colorado: Reid, M.R., S.R. Hart, E.R. Padovani, and G.A. Wandless, Contribu- Pb and Sr isotopeevidence, Geol. Soc.Am. Bull., 89, 59-82, 1978. tion of metapelitic sedimentsto the composition,heat production Lipman, P.W., H.H. Mehnert, and C.W. Naeser, Evolution of the and seismicvelocity of the lower crust of southernNew Mexico, Latir volcanic field, northern New Mexico, and its relation to the U.S.A., Earth Planet. Sci. Lett., 95, 367-381, 1989. Rio Granderift, as indicatedby potassium-argonand fission track Riciputi, L.R., and C.M. Johnson,Nd- and Pb-isotopevariations in dating,J. Geophys.Res., 91, 6329-6345, 1986. the multicyclic central caldera cluster of the San Juan volcanic Ludwig, K.R., Isoplot for MS-DOS -- A plotting and regression field, Colorado, and implicationsfor crustal hybridization, Geol- program for radiogenic-isotopedata, for IBM-PC compatible ogy, 18, 975-978, 1990. computers,U.S. Geol. Surv. OpenFile Rep., 87-601, 35 pp., 1988. Robertson,J.M., and K.C. Condie, Geology and geochemistryof Meijer,A., T.-T. Kwon,and G.R. Tilton,U-Th-Pb partitioning be- Early Proterozoic volcanic and subvolcanicrocks of the Pecos haviorduring partial melting in theupper mantle: Implications for greenstonebelt, Sangre de Cristo Mountains, New Mexico, Spec. the origin of high Mu componentsand the "Pb Paradox,"J. Pap. Geol. Soc. Am., 235, 119-146, 1989. Geophys.Res., 95, 433-448, 1990. Roden, M.F., A.J. Irving, and V.R. Murthy, Isotopic and trace ele- ment compositionof the uppermantle beneath a youngcontinental Menzies,M.A., Cratonic,circumcratonic and oceanicmantle do- rift: Results from Kilbourne hole, New Mexico, Geochim. mains beneaththe westernUnited States,J. Geophys.Res., 94, Cosmochim. Acta, 52,461-473, 1988. 7899-7915, 1989. Rudnick, R.L., and S.L. Goldstein,The Pb isotopiccompositions of Menzies, M.A., W.P. Leeman, and C.J. Hawkesworth, Isotope geo- lower crustal xenoliths and the evolution of lower crustal Pb, chemistryof Cenozoicvolcanic rocks reveals mantle heterogene- Earth Planet. Sci. Lett., 98, 192-207, 1990. ity belowwestern USA, Nature,303,205-209, 1983. Rudnick,R.L., andS.R. Taylor, The compositionand petrogenesis of Menzies,M.A., A.N. Halliday,Z. Palacz,R.H. Hunter,B.G.J. Upton, the lower crust: A xenolith study,J. Geophys.Res., 92, 13,981- P. Aspen,and C.J. Hawkesworth, Evidence from mantle xenoliths 14,005, 1987. for an enrichedlithospheric keel underthe Outer Hebrides,Na- Rudnick,R.L., and I.S. Williams, Dating the lower crust by ion ture, 35, 44-47, 1987. microprobe,Earth Planet. Sci. Lett., 85,'145-161, 1987. 13,608 JOHNSONAND THOMPSON:ISOTOPIC COMPOSITION OF MAFIC VOLCANICROCKS

Rudnick, R.L., W.F. McDonough, M.T. McCulloch, and S.R. Taylor, Thompson;R.A., M.A. Dungan, and P.W. Lipman, Multiple differ- Lower crustal xenoliths from Queensland, Australia: Evidence for entiation processesin early-rift calc-alkaline volcanics,northern deep crustalassimilation and fractionationof continentalbasalts, Rio Granderift, New Mexico, J. Geophys.Res., 91, 6046-6058, Geochim. Cosmochim. Acta., 50, 1099-1115, 1986. 1986. Ruiz, J., P.J. Patchett,and R.J. Arculus, Nd-Sr isotopecomposition Thompson,R.A., C.M. Johnson,and H.H. Mehnert, Oligocenebasal- of lower crustalxenoliths-evidence for the origin of mid-Tertiary tic volcanism of the northern Rio Grande rift: San Luis Hills, felsic volcanics in Mexico, Contrib. Mineral. Petrol., 99, 36-43, Colorado,J. Geophys.Res., this issue. 1988. Thompson, R.N., P.T. Leat, A.P. Dickin, M.A. Morisson, G.L. Ruiz, J., P.J. Patchett,and R.J. Arculus, Reply to "Commentson Nd- Hendry,S.A. Gibson,Strongly potassic mafic magmasfrom litho- Sr isotopic compositionsof lower crustal xenoliths-evidencefor sphericmantle sourcesduring continental extension and heating: the origin of mid-Tertiary felsic volcanics in Mexico" by K.L. Evidence from Miocene minettes of northwest Colorado, U.S.A., Cameron and J.V. Robinson, Contrib. Mineral. Petrol., 104, 615- Earth Planet. Sci. Lett., 98, !39-!53, !989. 618, 1990. Tilton, G.R., and B.A. Barreiro, Origin of lead in Andean calc- Shimizu, N., and I. Kushiro, The partitioningof rare-earthelements alkaline lavas, southern Peru, Science, 210, 1245-1247, 1980. and barium between garnet and liquid at high pressures:Prelimi- van Calsteren,P.W., P.D. Kempton, and C.J. Hawkesworth,Deple- nary experiments,Geophys. Res. Lett., 2, 413-416, 1975. tion of U in the lower crust: Evidence from granulite xenoliths Smith, C.B., Pb, Sr and Nd isotopicevidence for sourcesof southern from South Africa, Chern. Geol., 70, 74, 1988. African Cretaceous kimberlites, Nature, 304, 51-54, 1983. Vollmer, R., P. Ogden, J.G. Schilling, R.H. Kingsley, and D.G. Sobotovich, E.V., M.M. Shatz, A.V. Lovtzius, A.B. Remov, V.A. Waggoner,Nd andSr isotopesin ultrapotassicvolcanic rocks from Rudnick, and V.M. Terenmive, Isotopes of Pb in rocks in the the Leucite Hills, Wyoming, Contrib. Mineral. Petrol., 87, 359- Atdanskycrystalline shield (in Russian),in Geological-Radiologi- 368, 1984. cal Interpretation of the Meaning of Isotopic Ages, pp. 165-186, Walker, R.J., R.W. Carlson,S.B. Shirey, and F.R. Boyd, Os, St, Nd, Commissionon the Studyof AbsoluteAges of GeologicalForma- and Pb isotopesystematics of southernAfrican peridotitexeno- tions, Academy of Sciencesof USSR, Moscow, 1973. liths:Implications for the chemicalevolution of the subcontinental Stacey, J.S., and D.C. Hedlund, Lead-isotopiccompositions of di- mantle, Geochim. Cosmochim.Acta, 53, 1583-1595, 1989. verse igneous rocks and ore deposits from southwesternNew Wasserburg,G.J., S.B. Jacobsen,D.J. DePaolo, M.T. McCulloch, Mexico and their implicationsfor Early Proterozoiccrustal evolu- and T. Wen, Precise determinations of Sm/Nd ratios, Sm and Nd tion in the western United States, Geol. Soc. Am. Bull., 94, 43-57, isotopicabundances in standardsolutions, Geochim. Cosmochim. 1983. Acta, 45, 2311-2332, 1981. Stacey, J.S., and J.D. Kramers, Approximation of terrestrial lead Whitehouse,M.J., Pb-isotopicevidence for U-Th-Pb behaviourin a isotopeevolution by a two-stagemodel, Earth Planet. Sci. Lett., prograde amphibolite tO granulite facies transition from the 26, 207-221, 1975. Lewisiancomplex of north-westScotland: Implications for Pb-Pb Stacey, J.S., B.R. Doe, L.T. Silver, and R.E. Zartman, Plumbo- dating,Geochim. Cosmochim. Acta, 53,717-724, 1989. tectonicsIIA, Precambrianmassive sulfide depositsin geochro- Williams, S., Late Cenozoic volcanism in the Rio Grande rift: trace nology, U.S. Geol. Surv. Open File Rep., 76-476, 30 pp. 1977. element, strontiumisotopic and neodymiumisotopic geochemis- Staudigel, H., A. Zindler, S.R. Hart, T. Leslie, C.Y. Chen, and D. try of the TaosPlateau volcanics, Ph.D. thesis,202 pp., Univ. of Claugue,The isotopesystematics of a juvenile intraplatevolcano: Minn., Minneapolis, 1984. Pb, Nd, and Sr isotope ratios of basalts from Loihi Seamount, Wooden, J.L., and P.A. Mueller, Pb, St, and Nd isotopic composi- Hawaii, Earth Planet. Sci. Lett., 69, 13-29, 1984. tions of a suite of Late Archean igneousrocks, easternBeartooth Stein, H.J., A lead, strontium, and sulfur isotopestudy of Laramide- Mountains:Implications for crust-mantleevolution, Earth Planet. Tertiary intrusions and mineralization in the Colorado Mineral Sci. Lett., 87, 59-72, 1988. Belt with emphasison Climax-type porphyry molybdenum sys- Wooden,J.L., J.S. Stacey,K.A. Howard,B.R. Doe, andD.M. Miller, tems plus a summaryof newly acquiredisotopic and rare earth Pb isotopicevidence for the formationof Proterozoiccrust in the element data, Ph.D. thesis, 493 pp., Univ. of N.C., Chapel Hill, southwesternUnited States,in Metamorphismand Crustal Evolu- 1985. tion of the WesternUnited States, Rubey Volume 7, editedby W.G. Stein, H.J., and J.G. Crock, Late Cretaceous-Tertiarymagmatism in Ernst,pp. 68-86, Prentice-Hall,Englewood Cliffs, N.J., 1988. the Colorado Mineral Belt: Rare earth element and samarium- Zartman, R.E., and B.R. Doe, Plumbotectonics -- The model, neodymiumisotopic studies, Mere. Geol. Soc.Am., 174, 195-223, Tectonophysics,75, 135-162, 1981. 1990. Zartman, R.E., and S.M. Haines, The plumbotectonicmodel for Pb Steven, T.A., Middle Tertiary volcanic field in the southernRocky isotopicsystematics among major terrestrialreservoirs -- A case Mountains, Mere. Geol. Soc. Am., 144, 75-94, 1975. for bi-directionaltransport, Geochim. Cosmochim. Acta, 52, 1327- Stille, P., D.M. Unruh, and M. Tatsumoto,Pb, St, Nd andHf isotopic 1339, 1988. evidence of multiple sourcesfor Oahu, Hawaii basalts,Nature, Zartman,R.E., andF. Tera, Lead concentrationand isotopiccompo- 304, 25-29, 1983. sitionin five peridotireinclusions of probablemantle origin, Earth Stormer, J.C., Jr., Ages and nature of volcanic activity on the south- Planet. Sci. Lett., 20, 54-66, 1973. ern high plains, New Mexico and Colorado, Geol. Soc.Am. Bull., Zindler, A., and S. Hart, Chemical geodynamics,Annu. Rev. Earth 83, 2443-2448, 1972a. Planet. Sci., 14, 493-571, 1986. Stormer, J.C., Jr., Mineralogy and petrology of the Raton-Clayton volcanic field, northwestern New Mexico, Geol. Soc. Am. Bull., C.M. Johnson,Department of Geologyand Geophysics 83, 3299-3322, 1972b. Universityof Wisconsin,Madison, WI 53706. Tatsumoto,M., Isotopic compositionof lead in oceanicbasalt and its implicationto mantleevolution, Earth Planet. Sci. Lett., 38, 63- R.A. Thompson,U.S. GeologicalSurvey, Box 25046, Denver Federal Center, Denver, CO 80225. 67, 1978. Taylor, H.P., Jr., The effects of assimilation of country rocks by magmason 180/160and 87Sr/86Srsystematics in igneousrocks, Earth Planet. Sci. Lett., 47, 243-254, 1980. Thompson,R.A., and M.N. Machette, Geologicmap of the San Luis (ReceivedJuly 2, 1990; Hills area, Conejos and Costilla counties,Colorado, U.S. Geol. revised December 24, 1990; Surv. Misc. Field. Invest., I-1906, 1989. acceptedJanuary 28, 1991.)