| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology Volume 42 Number 5 Pages 901-929 2001
© Oxford University Press 2001
Petrology and Geochemistry of the Late Eocene Harrison Pass Pluton, Ruby Mountains Core Complex, Northeastern Nevada
1DEPARTMENT OF GEOSCIENCES, TEXAS TECH UNIVERSITY, LUBBOCK, TX 79409-1053, USA
2DEPARTMENT OF GEOLOGY AND GEOPHYSICS, UNIVERSITY OF WYOMING, LARAMIE, WY 82071-3006, USA
3DEPARTMENT OF GEOLOGY, RICE UNIVERSITY, HOUSTON, TX 77251-1892, USA
4DEPARTMENT OF CHEMISTRY AND BIOCHEMISTRY, TEXAS TECH UNIVERSITY, LUBBOCK, TX 79409-1053, USA
Received April 20, 1999; Revised typescript accepted August 2, 2000
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUMMARY OF FIELD RELATIONSPETROGRAPHYGEOCHEMISTRYDISCUSSIONSCHEMATIC MODEL OF MAGMA...SUMMARYAPPENDIX: METHODSREFERENCES |
|---|
The late Eocene Harrison Pass pluton was emplaced in the transition zone between the infrastructure and suprastructure of the Ruby Mountains core complex. Emplacement was at
3 kbar pressure and was in two stages: early stage tonalitic to monzogranitic magmas, followed by late-stage monzogranites and mafic dikes. The early stage began with emplacement of biotite ± hornblende granodiorite of Toyn Creek, followed by the biotite monzogranite of Corral Creek. Quenched equivalents of these units are preserved as porphyritic dikes in the roof. Leucocratic monzogranite that forms cupolas in the roof zone represents the product of fractional crystallization of the early magmas. Al-in-hornblende barometry and the presence of magmatic epidote suggest that early stage magmas resided at a pressure of 5-6 kbar before emplacement in the upper crust. Compositional variation in the Toyn Creek and Corral Creek units is essentially linear, and can be explained by mixing of a tonalitic end member with a monzogranitic end member such as evolved samples of the monzogranite of Corral Creek. The tonalitic end member was itself a hybrid that formed by interaction of mafic magma with crustal melts. The monzogranitic end member is a crustal melt that escaped hybridization. Nd isotopic compositions of the early stage are heterogeneous and do not correlate with degree of differentiation, which is consistent with a compositionally heterogeneous felsic end member. Elemental variation in the early stage of the pluton is unusual because of its essentially linear trends, which suggest a single mixing event before emplacement in the upper crust. Late-stage activity consisted of three pulses of monzogranitic magma plus sparse mafic dikes. The largest and youngest of these pulses, the two-mica monzogranite of Green Mountain Creek, is distinct from all other units in the presence of restitic enclaves, low 
Nd, and high initial 87Sr/86Sr. Pod-like bodies of leucocratic biotite ± amphibole monzogranite and sheets and dikes of leucocratic two-mica monzogranite make up the other late-stage granites. These rocks display deep negative Eu anomalies and show wide, non-systematic concentrations of high field strength elements; however, their isotopic compositions are identical to those of the early stage rocks. They are thought to be small melt fractions of the lower to middle crust. Their elemental compositions are thought to result from the effects of residual plagioclase and accessory minerals. The variable and non-systematic isotopic compositions of all granitic units in the pluton suggest a heterogeneous source region, such as the Proterozoic Mojave province. KEY WORDS: granite; mixing; crustal melting; Nevada
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF FIELD RELATIONSPETROGRAPHYGEOCHEMISTRYDISCUSSIONSCHEMATIC MODEL OF MAGMA...SUMMARYAPPENDIX: METHODSREFERENCES |
|---|
The Late Eocene (
36 Ma) Harrison Pass pluton is exposed in the Ruby Mountains core complex (Snoke, 1980
), one of the best-exposed core complexes in North America. The pluton was emplaced during a period of voluminous magmatism in the Great Basin (Stewart & Carlson, 1976; Gans et al., 1989; Feeley & Grunder, 1991) that followed 45 my of magmatic quiescence.
Detailed studies of the coeval volcanic rocks (Gans et al., 1989; Feeley & Grunder, 1991) showed the importance of mantle-derived magma, but also demonstrated the significance of crustal melting, mixing of mantle- and crust-derived magmas, and assimilation of crustal material (Grunder, 1992, 1995). Grunder (1992, 1995) specifically showed that hybridization was predominantly by crustal assimilation combined with fractional crystallization.
The Harrison Pass pluton is distinct from these volcanic rocks in at least two important ways. First, the greatest volume of the pluton consists of felsic rocks with a wide range of compositions. Second, hybrid rocks in the pluton show geochemical variations that are suggestive of a single magma mixing event followed by minor mingling.
Therefore, the Harrison Pass pluton provides an important contrast in petrologic process and magmatic history relative to coeval volcanic rocks. The petrologic history of the pluton is presented in this paper. The relationship between extension and magmatism and the detailed emplacement history has been presented elsewhere (Burton, 1997).
|
SUMMARY OF FIELD RELATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF FIELD RELATIONSPETROGRAPHYGEOCHEMISTRYDISCUSSIONSCHEMATIC MODEL OF MAGMA...SUMMARYAPPENDIX: METHODSREFERENCES |
|---|
The Harrison Pass pluton was emplaced in the transition zone between the infrastructure and suprastructure of the Ruby Mountains core complex. This zone separates a structurally lower sequence of plutonic and metamorphic rocks north of the pluton (infrastructure) from a structurally higher sequence of weakly metamorphosed to unmetamorphosed Paleozoic sedimentary strata east and south of the pluton (suprastructure) (Fig. 1; Howard et al., 1979
). The regional dip of sedimentary strata in the suprastructure is 30° to the east.
|
Mesozoic regional metamorphism in the infrastructure reached upper amphibolite grade (maximum pressure 4·5 kbar; Hudec, 1992; Jones, 1999). Peak metamorphism probably occurred in late Cretaceous time and was followed by exhumation before Tertiary development of the core complex (McGrew et al., 2000). In contrast, regional metamorphism of the suprastructure was no higher than greenschist grade (Willden & Kistler, 1979), and Burton (1997) showed that metamorphic grade in these rocks increases with stratigraphic depth.
Intrusive contacts with both upper and lower structural sequences are sharp. Contact metamorphic assemblages diagnostic of emplacement pressures are lacking in rocks of the infrastructure; however, andalusite and sillimanite are present in semipelitic rocks of the suprastructure collected along the southern contact.
The regional dip of sedimentary rocks of the suprastructure (Burton, 1997) combined with younging of K–Ar cooling ages from east to west (Kistler et al., 1981) indicate that the central Ruby Mountains were tilted eastward after emplacement of the Harrison Pass pluton. Subsequent erosion has thus exposed a structural section of the pluton of 5 km thickness.
Emplacement of the plutonic magmas was in two stages (Figs 1–3). The early stage began with intrusion of broadly granodioritic magmas and so this stage is referred to as the ‘granodioritic stage’. It consists of three mappable subunits: granodiorite of Toyn Creek, monzogranite of Corral Creek, and roof dikes of two types: porphyritic and equigranular. The second, late-stage part of the pluton consists of leucocratic two-mica monzogranitic dikes and sheets, leucocratic biotite monzogranitic masses, two-mica monzogranite of Green Mountain Creek, and mafic to intermediate dikes.
|
Early stage
Rocks of the early stage underlie the eastern part of the pluton and are host to abundant late-stage units in the western part (Fig. 1). Except for the roof dikes, early stage rocks are coarse to very coarse grained.
The granodiorite of Toyn Creek crops out in the northeastern and central parts of the pluton. It ranges from biotite ± hornblende tonalite to granodiorite, with sparse quartz monzodiorite (Fig. 3). The distribution of rock types is not correlated with their geographic position. Swarms of mafic microgranular enclaves are locally present, as are less common zones of concentrated, flow-aligned alkali-feldspar megacrysts. Magmatic foliation is defined by alignment of alkali-feldspar megacrysts, local orientation of biotite, and flattening of mafic enclaves. Sparse synplutonic tonalitic dikes are also present.
The monzogranite of Corral Creek crops out in the southern part of the pluton (Fig. 1); an isolated body southwest of the pluton is also thought to belong to this unit. The contact with the granodiorite of Toyn Creek truncates the magmatic fabric of the latter unit (Burton, 1997). Mafic enclaves are sparse. In the southeastern part of the pluton, sparse miarolitic cavities are present in the Corral Creek unit.
Roof dikes crop out along the eastern and southeastern side of the pluton, where they are intrusive into Ordovician metacarbonate rocks. They encompass two types: porphyritic dikes of broadly dacitic composition and medium-grained equigranular biotite granite to granodiorite (Fig. 3) that grades structurally downward into a zone of heterogeneous granitic rocks that enclose abundant stope blocks.
Late stage
Two-mica monzogranitic sheets and dikes of variable grain size, texture, and orientation underlie much of the western part of the pluton (Figs 1 and 2). Intrusive bodies are termed dikes if they have steep (subvertical) dips, whereas the intrusive sheets have approximately north–south strike and 30° dip to the east, subparallel to the dip of sedimentary rocks of the suprastructure. Many of these shallowly dipping bodies are composite pegmatite–aplite intrusions in which line rock is locally present. Nearly all samples of this group are leucocratic.
|
The biotite monzogranite unit consists of leucocratic granitic bodies with sparse biotite or biotite and hornblende. They range in grain size from medium to coarse grained. The bodies are most commonly pod-like, but also occur as dikes and sheets with orientations similar to the two-mica monzogranites.
The monzogranite of Green Mountain Creek crops out in the northern part of the pluton (Fig. 1). It is a homogeneous (Fig. 3), medium- to coarse-grained, two-mica monzogranite. It contains sparse enclaves of two types: foliated biotite- and muscovite-rich clots (surmicaceous enclaves; Didier, 1973) and fist-sized masses of quartz. Other late-stage leucogranite types do not intrude the monzogranite of Green Mountain Creek, which indicates that it is the youngest granitic unit of the pluton.
Mafic to intermediate dikes were also emplaced during late-stage activity. The dikes range widely in grain size, but most are medium- to fine-grained, sparsely amphibole-phyric quartz diorite to quartz monzonite (Fig. 3). In the west–central part of the pluton, several composite dikes occur. They consist of commingled fine-grained quartz monzodiorite and medium-grained two-mica leucogranite.
|
PETROGRAPHY |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF FIELD RELATIONSPETROGRAPHYGEOCHEMISTRYDISCUSSIONSCHEMATIC MODEL OF MAGMA...SUMMARYAPPENDIX: METHODSREFERENCES |
|---|
Early stage
Rocks of the granodiorite of Toyn Creek are coarse grained and hypidiomorphic granular; they commonly contain alkali-feldspar phenocrysts (now microcline:
Or90) as much as 3 cm long. Biotite is the ubiquitous mafic mineral and amphibole is present in the more mafic samples; amphibole-bearing rocks crop out predominantly in the northeastern part of the unit. The rocks are variably strained; typically with weak subgrain development in quartz and minor bent biotite and bent or broken plagioclase. Plagioclase crystals show core and mantle compositions of An51–35 (Fig. 4), with one or two abrupt zoning reversals. Plagioclase rims are normally zoned (An33–25). Biotite occurs as dark to medium brown books and as clusters that rim amphibole. Inclusions and intergrowths of allanite-cored epidote are common. Amphibole is blue–green to olive ferropargasite to ferro-edenite [classification of Leake et al. (1997); Fig. 4b]. Amphibole grains are subhedral to ragged anhedral and are partly or completely rimmed by biotite. Partial replacement by biotite is common. A few amphibole grains show symplectite-like intergrowths with epidote. Accessory zircon and apatite are intergranular and form minute inclusions in amphibole and biotite. Other accessory minerals are titanite (as much as 8 mm long), red to brown allanite, rare magnetite, and very rare ilmenite.
|
Synplutonic dikes in the granodiorite of Toyn Creek are medium-grained epidote-bearing hornblende biotite tonalite with hypidiomorphic granular texture. Oscillatory-normal zoned plagioclase (An55–38) reaches 2 mm in length; quartz and sparse alkali-feldspar are interstitial. Amphibole is ragged and commonly partly rimmed by brown biotite; it also forms rare inclusions in epidote. Most biotite is fine grained and intergranular, but some books reach 2 mm in length. Subhedral to anhedral epidote is commonly cored by reddish brown allanite. Accessory minerals are allanite, euhedral to subhedral titanite, apatite, zircon, and sparse hematite after magnetite.
Biotite monzogranite of Corral Creek is similar to samples of the granodiorite of Toyn Creek in grain size and in the presence of alkali-feldspar phenocrysts. Some alkali-feldspar is interstitial to poikilitic, and quartz grains are commonly euhedral against such feldspar. Plagioclase is oscillatory-normal zoned, with core compositions from An44 to An25 and rim compositions from An25 to An12. Brown biotite is partly replaced by chlorite. Accessory minerals are titanite (to 1·2 mm long), primary muscovite, rare amphibole inclusions in plagioclase, sparse magnetite, red–brown allanite, apatite, and zircon. Apatite and zircon are minute crystals, typically adjacent to allanite or titanite.
Enclaves
Mafic microgranular enclaves from the Toyn Creek and Corral Creek units range from fine- to coarse-grained quartz diorite to quartz monzonite. Tonalitic compositions are most common. The samples contain sparse large crystals of plagioclase ± hornblende ± biotite ± quartz ± alkali-feldspar in a groundmass of plagioclase, quartz, and biotite ± hornblende. Some hornblende-bearing enclaves are present in the hornblende-free monzogranite of Corral Creek. Plagioclase phenocrysts display oscillatory-normal zoned cores (An49–35) with one or two prominent zoning reversals and normally zoned rims to An24. Many samples display evidence for disequilibrium; in particular, ovoid quartz with amphibole rims, biotite rimmed by amphibole, and acicular apatite as much as 2 mm long. Apatite is less commonly present as equant grains. Other accessory minerals are euhedral allanite, subhedral to interstitial titanite, zircon as subhedral granular to acicular grains, and epidote cored by allanite.
Roof dikes
The porphyritic dacitic roof dikes contain 30–37% phenocrysts that reach at least 5 mm in length and consist, in order of decreasing abundance, of plagioclase (An47–38), quartz, and biotite (Table 1). Sparse alkali-feldspar phenocrysts are present, as is rare ragged relict amphibole. Accessory minerals are allanite, titanite, and zircon (all as microphenocrysts), and secondary minerals after Fe–Ti oxides. Primary groundmass textures were typically granular to granophyric, but are ubiquitously recrystallized.
|
The equigranular roof dikes consist of biotite granodiorite to granite (Fig. 3; Table 1). They are hypidiomorphic granular, with subhedral to anhedral, blocky, oscillatory-normal zoned plagioclase and interstitial perthitic microcline and quartz. Sparse Carlsbad-twinned alkali-feldspar crystals reach 1 mm in length. Plagioclase, quartz, and biotite are inclusions in alkali-feldspar. Quartz generally displays subgrain development, which obscures its original size and habit. Reddish brown or brown to yellow biotite forms small flakes and books. Accessory minerals are the same as in the Toyn Creek unit except that allanite lacks epidote overgrowths and a few, acicular apatite inclusions are were observed. Sparse myrmekite is present. Alteration is variable, and consists of sericitization of plagioclase cores and replacement of biotite by chlorite.
Late-stage granitic rocks and mafic dikes
Leucocratic two-mica monzogranite ranges from fine to very coarse grained, with hypidiomorphic granular, aplitic, and pegmatitic textures. Plagioclase (An20–12 cores and mantles) and alkali-feldspar (microcline: Or90) range from subhedral to anhedral; quartz is interstitial to intergranular. Subhedral muscovite and biotite books reach 2·5 mm long and both minerals are also interstitial. Accessory minerals are euhedral to subhedral zircon (and monazite?) and short apatite needles as inclusions in the silicates.
Late-stage biotite monzogranite ranges from hypidiomorphic granular to aplitic and from coarse to fine grained. Equant plagioclase is sodic, with An20 cores and An12–8 rims (Fig. 4). Alkali-feldspar (microcline) varies from blocky to interstitial and quartz is generally interstitial. In some samples, plagioclase, alkali-feldspar, and quartz show mutual inclusion relationships. Fresh biotite is present as inclusions in other minerals; interstitial biotite is typically altered to chlorite ± muscovite. Accessory minerals are titanite, allanite, fine-grained apatite and rare garnet. A few dikes of this type contain euhedral to ragged blue–green to olive–green amphibole.
The muscovite biotite monzogranite of Green Mountain Creek is medium to coarse grained and hypidiomorphic granular. Plagioclase is normally zoned, with An30–18 cores and An17–12 rims (Fig. 4). Perthitic alkali-feldspar (microcline: Or92) is euhedral to anhedral; it commonly contains inclusions of subhedral plagioclase. Muscovite and biotite are commonly intergranular. Quartz shows subgrain development and micas are commonly bent. Accessory minerals are zircon, monazite, and apatite. Sparse surmicaceous enclaves (see Didier, 1973) are medium grained, with lepidoblastic foliation formed by muscovite and biotite, and locally by oriented plagioclase. Some plagioclase shows distinct altered cores, with abrupt boundaries to mantles of An21–12. Rim compositions are An10. Microcline (Or90) and quartz are intergranular. Accessory minerals are apatite, zircon, and monazite (?).
Late-stage dioritic to quartz monzonitic mafic dikes are medium to fine grained and hypidiomorphic granular. Edenitic amphibole is present in the groundmass, as phenocrysts to 0·5 mm long, and as oikocrysts as much as 5 mm in diameter. Biotite is a groundmass mineral and replaces amphibole. Plagioclase (andesine cores with oligocene rims) forms sparse phenocrysts and is interstitial with quartz and alkali-feldspar in the groundmass. Accessory minerals are apatite, very fine-grained Fe–Ti oxides, titanite, and rare allanite.
Mafic mineral compositions
Amphibole
Amphibole compositions range from ferropargasite to edenite (Table 2; Fig. 4b; see Appendix for analytical methods). The amphiboles are relatively Fe rich, except for those from late-stage mafic dikes, which generally have the highest Mg/(Mg + Fe) values (Fig. 4b). Amphibole from enclaves in the Toyn Creek unit plots on a trend that is distinct from amphibole from the host rocks (Fig. 4b). Amphibole from late-stage leucogranitic biotite monzogranitic dikes is distinctly more alkali rich (at constant values of AlIV) than samples from the early stage granodiorite of Toyn Creek.
|
Biotite
In the early stage, biotite Mg/(Mg + Fe) values range from 0·45 to 0·32 (Table 3), except for one sample from the Corral Creek unit, with an Mg/(Mg + Fe) value of 0·22. Biotite from the late-stage monzogranite of Green Mountain Creek and from two-mica granite sheets and dikes has Mg/(Mg + Fe) values between 0·25 and 0·15 and is richer in Al (>3·1 Al p.f.u.) compared with early stage biotite (<3·1 Al p.f.u.). These samples, along with the relatively Fe-rich sample of the Corral Creek unit, are also enriched in fluorine relative to all other samples of the granodioritic unit.
|
Estimates of pressure and temperature
The Al-in-hornblende barometer was applied to nine samples in which the equilibrium assemblage amphibole, plagioclase, alkali-feldspar, Fe–Ti oxide, quartz, and titanite is present (Hammarstrom & Zen, 1986; Hollister et al., 1987). Amphibole rim compositions from five samples of the Toyn Creek unit yielded pressure estimates of 4·9–6·9 kbar (average 6·1 ± 0·7) according to the Schmidt (1992) calibration. The temperature-corrected calibration of Anderson & Smith (1995) using the second temperature calibration of Holland & Blundy (1994; compare Anderson, 1996) yielded P estimates from 4·6 to 6·4 kbar (average 5·4 ± 0·6) at temperatures from 711 to 758°C (average 736 ± 15°C).
Three late-stage samples (two mafic dikes and a leucocratic syenogranite) yielded pressures between 3·3 and 4·0 kbar (average 3·8 ± 0·3) according to the Schmidt (1992) calibration and between 3·4 and 4·0 kbar (average 3·6 ± 0·3) according to the Anderson & Smith calibration. Temperature estimates for this latter group were from 619 to 718°C (average 682 ± 45°C) (Holland & Blundy, 1994). A single late-stage leucocratic monzogranitic dike with centimeter-scale amphibole yielded pressure estimates of 5·6 kbar (Schmidt, 1992) and 5·2 kbar (Anderson & Smith, 1995).
Two mineral saturation geothermometers are applicable to the Harrison Pass pluton: apatite saturation (Harrison & Watson, 1984; Watson & Harrison, 1984) and zircon saturation (Harrison & Watson, 1983; Watson & Harrison, 1983). The former is strictly applicable only to metaluminous samples (e.g. mafic samples of the Toyn Creek unit). In fact, all samples from the Toyn Creek unit yielded apatite saturation temperatures of 950°C; progressively lower temperatures were obtained for more evolved, peraluminous, compositions to 820°C for equigranular roof zone granites. Zircon saturation temperatures are in the 800–840°C range for the Toyn Creek unit, 775°C for the monzogranite of Green Mountain Creek, and <770°C for all other monzogranitic samples. Interpretation of mineral saturation temperatures for the Toyn Creek unit is problematic, especially in view of the mixing history we propose (see below). It is unclear whether these temperatures represent true mineral saturation or whether they reflect inheritance from the end members. Zircon saturation temperatures for the late-stage granites are thought to represent near-solidus conditions. In fact, the estimate for the monzogranite of Green Mountain Creek is probably a maximum value in view of its know zircon inheritance (Wright & Snoke, 1993).
|
GEOCHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF FIELD RELATIONSPETROGRAPHYGEOCHEMISTRYDISCUSSIONSCHEMATIC MODEL OF MAGMA...SUMMARYAPPENDIX: METHODSREFERENCES |
|---|
Classification
Samples from the pluton range in SiO2 content (Table 4; see Appendix for analytical methods) from 50 to 78 wt % (Fig. 5). However, samples with <63 wt % SiO2 consist entirely of mafic microgranular enclaves and late-stage mafic dikes. These rocks constitute <1% of the volume of the pluton; therefore the bulk of the intrusion consists of rocks with >63% SiO2. The pluton is broadly calc-alkaline, and extrapolation of the trend for the granodioritic stage yields an alkali–lime index of
57 (calc-alkalic). Nearly all samples plot within the high-K field of subalkaline rocks (Fig. 6b).
|
|
The early stage and the late-stage biotite monzogranite are generally mildly peraluminous, with alumina saturation indices [ASI = molar Al2O3/(CaO + Na2O + K2O)] in the range from 0·99 to 1·07 (Fig. 6a). SiO2 and ASI are uncorrelated within this group. Late-stage two-mica granitic dikes and sheets are peraluminous and their ASI ranges from 1·01 to 1·2. The monzogranite of Green Mountain Creek is strongly peraluminous (Fig. 6a), and foliated, mica-rich enclaves (surmicaceous enclaves, Didier, 1973) from the unit have ASI > 1·3. This is consistent with the Al-rich nature of biotite from this unit. Thus, strongly peraluminous compositions are restricted to two of the late-stage units: the monzogranite of Green Mountain Creek and two-mica granitic dikes and sheets.
|
Compositional distinctions among intrusive units
Early stage
In binary variation diagrams (Figs 7–10), samples from the Toyn Creek and Corral Creek units stage plot in approximately linear trends, although some show considerable scatter. Samples of the porphyritic roof dikes plot near the center of this trend, and a synplutonic tonalitic dike in the granodiorite of Toyn Creek (sample BB-70-94) lies at the mafic end (Figs 7–10). The distinct trend of these samples is most apparent in Fig. 8a, where their mg-number [Mg/(Mg + FeT), atomic] can be seen to be nearly constant at
0·35. This is in contrast to late-stage mafic dikes, which display a general trend toward decreasing mg-number with decreasing CaO and with equigranular roof dikes and late-stage monzogranitic bodies, which display a wide range of mg-number, from >0·5 to <0·15 at low CaO contents.
|
|
Samples of the monzogranite of Corral Creek are richer in SiO2 than those of the granodiorite of Toyn Creek (Figs 6, 7, and 10) and contain lower concentrations of rare earth elements (REE; Table 4; Fig. 11). Within the early stage, REE concentrations decrease regularly as a function of increasing SiO2 content, and the size of the Eu anomaly decreases irregularly with increasing SiO2, such that two of three analyzed Corral Creek samples lack Eu anomalies (Fig. 11).
|
Equigranular early stage biotite monzogranitic roof zone dikes and subjacent leucocratic monzogranite are weakly peraluminous (Fig. 6a) and have lower mg-number, Y, Nb, Sr, Ba, and REE than the rest of the granodioritic stage (Figs 9–11). These early stage roof zone monzogranites can be distinguished from late-stage biotite monzogranites because the latter group has higher Nb, Y, Zr, Sr, and mg-number and displays prominent negative Eu anomalies (Figs 8, 9, and 11).
|
|
In general, the Rb concentrations in the early stage increase only slightly with differentiation; however, some samples of both the Toyn Creek and Corral Creek units show non-systematic variation. In particular, a group of samples of the granodiorite of Toyn Creek collected from the west–central part of the pluton are host to late-stage two-mica monzogranite sheets and dikes. These samples have Rb contents 10–40 ppm higher than other samples of the Toyn Creek unit.
Mafic to intermediate enclaves from the Toyn Creek and Corral Creek units show compositional variability in binary plots (Figs 5–9). Their CaO contents and mg-number values are similar to those of their host rocks; however, they have lower Al2O3, P2O5, Sr, and Ba contents (Figs 6 and 8).
Late-stage units
Late-stage mafic dikes generally plot in arrays that are commonly collinear with the compositions of the early stage. However, in the plot of Al2O3 vs SiO2 (Fig. 6), the trend defined by the mafic dikes is parallel to the trend defined by the granodioritic stage, not collinear with it.
The majority of late-stage granitic rocks can be distinguished from those of the early stage on the basis of their lower mg-number (Fig. 8a), and higher Rb, Y, and Nb contents (Figs 9 and 10). They can be distinguished from each other on the basis of their REE patterns. The slope of the REE pattern for monzogranite of Green Mountain Creek is similar to patterns of the Toyn Creek unit, but with a much deeper negative Eu anomaly (Fig. 11). A mafic surmicaceous enclave from the Green Mountain Creek body has an REE pattern parallel to that of its host rock; this sample has the highest REE abundances of any from the pluton.
Late-stage two-mica monzogranite sheets and dikes have lower light REE (LREE) abundances than the Green Mountain Creek unit and a wide range of heavy REE (HREE) abundances (Fig. 11), which is consistent with the range of Y values among these samples (Fig. 9). The late-stage biotite monzogranites are distinct because they display REE patterns with positive slopes, a consequence of their low abundances of LREE.
Isotopic data
The early stage shows Nd values from -10 to -18 over a range of initial 87Sr/86Sr from 0·7096 to 0·7142 (ratios calculated at 36 Ma; Fig. 11a; Table 5). Isotopic compositions of the granodiorite of Toyn Creek and monzogranite of Corral Creek overlap and are indistinguishable. Within this unit, Nd is not correlated with any other geochemical parameter. The late-stage biotite monzogranite and two-mica monzogranite sheets and dikes have the same range of Nd and Sr isotopic compositions as the early stage (Fig. 11a). The two-mica monzogranite of Green Mountain Creek is distinct in its higher initial 87Sr/86Sr (0·725) and lower Nd (average –21·3).
|
Oxygen isotopic compositions of quartz separates and whole rocks (Table 5) are shown in Fig. 11b. Quartz 
18O values of samples of the early, granodioritic stage cluster between +9·7 and +11·1
. With one exception, quartz from late-stage biotite monzogranites is similar. Whole-rock 18O values for these two units typically range from +8·7 to +10·1, but four samples have whole-rock values < +7·8. Whole-rock 18O values for late-stage two-mica monzogranitic sheets and dikes and the monzogranite of Green Mountain Creek are between +8·0 and +9·0.
Except for the four samples with relatively low 18O cited above, the differences between quartz–whole-rock pairs are from 1 to 2. Such differences suggest that the whole-rock values reflect magmatic 18O values. The four samples with anomalously low 18O have whole-rock values from 2·5 to 6·5 lower than coexisting quartz. Such values are suggestive of disequilibrium caused by exchange with 18O-depleted, presumably meteoric fluids (e.g. Taylor, 1974; Criss & Taylor, 1986; Criss et al., 1991). Such exchange has been documented adjacent to the Harrison Pass pluton and in the northern Ruby Mountains (Fricke et al., 1992) and is similar to exchange with meteoric fluids reported from other core complexes (e.g. Kerrich & Rehrig, 1987; Losh, 1997; Morrison & Anderson, 1998). The timing and structural controls on this exchange are the subject of continuing study in the Harrison Pass area (C. G. Barnes & H. R. Karlsson, unpublished data, 1997).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF FIELD RELATIONSPETROGRAPHYGEOCHEMISTRYDISCUSSIONSCHEMATIC MODEL OF MAGMA...SUMMARYAPPENDIX: METHODSREFERENCES |
|---|
Petrogenetic models for the Harrison Pass pluton magmas must take into account the following observations. The first is the overall linearity of the compositional trend of the Toyn Creek and Corral Creek units of the early stage and the distinct slope of this trend relative to more mafic and more felsic parts of the pluton. The second is the diversity of elemental compositions displayed by late-stage granites in the pluton. The third is the broad isotopic similarity of all units except the two-mica monzogranite of Green Mountain Creek. Lastly, the Harrison Pass pluton is distinct from coeval volcanic units in the Great Basin because of the comparative paucity of mafic rocks. The origins of these features are considered in the following sections. They will then be combined with information about the level of emplacement and the regional geologic setting to develop a model for the evolution of Harrison Pass magmas. To that end, it is first necessary to consider the depth of emplacement of the pluton.
As shown above, pressure estimates determined on the basis of the Al-in-hornblende barometer show discrepancies between early and late-stage units, with an 2·3 kbar difference using the Schmidt (1992) calibration and an 1·8 kbar difference using the Anderson & Smith (1995) calibration. This discrepancy can result from (1) systematic analytical errors or in the choice of samples for analysis, (2) synplutonic exhumation, with deeper-level emplacement of the early stage followed by exhumation and then by shallow emplacement of the late stage (e.g. Anderson, 1988; Anderson et al., 1988; Hill et al., 1995), or (3) partial crystallization of the early stage under higher-pressure conditions followed by emplacement as a crystal-rich magma at lower pressure.
Analytical errors large enough to explain the observed pressure differences are unlikely. All samples contained the appropriate buffer assemblage (Hollister et al., 1987), all analyses were carried out on the same instrument (Appendix) with the same standard set, and the compositions of coexisting plagioclase were measured.
Synplutonic exhumation is also unlikely for several reasons. First, within analytical uncertainty, the oldest and youngest units of the pluton yielded identical U–Pb ages of 36 Ma (Wright & Snoke, 1993). Second, the presence of andalusite in the aureole indicates emplacement at pressures <4 kbar. One could argue that the thermal effects of shallow-level late-stage magmatism overprinted deep-level contact metamorphic assemblages. However, the diagnostic aureole assemblages are developed in weakly metamorphosed rocks of the suprastructure. Thus, if the early stage were emplaced at deeper crustal levels, one would expect preservation of moderate-P regional metamorphic minerals in the suprastructure at some distance from the pluton. No evidence for such an event exists. Finally, miarolitic cavities are present in the structurally highest parts of the early stage monzogranite of Corral Creek, a clear indication of emplacement at shallow depths.
The porphyritic roof dikes are thought to be ‘quenched’ samples of the underlying Toyn Creek and Corral Creek magmas. These dikes contain >30% phenocrysts, which include biotite, quartz, and alkali-feldspar. This assemblage indicates magmatic temperatures µ 800°C (e.g. Johannes & Holtz, 1996). Emplacement of crystal-rich, low-temperature magma could result in preservation of amphiboles and epidote that crystallized in deeper environments (e.g. Brandon et al., 1996). These temperatures are similar to zircon saturation temperatures determined for the granodiorite of Toyn Creek (800–840°C).
Evolution of early stage magmas
Elemental variation among the granodiorite of Toyn Creek and monzogranite of Corral Creek is broadly linear (Figs 5–7), with a narrow range of Mg/(Mg + Fe) (Fig. 7). The most mafic sample that is consistently a member of this linear trend is a synplutonic tonalitic dike (labeled with a ‘T’ in Figs 4–9); the most evolved sample is sample HP-37 from the monzogranite of Corral Creek. This linearity is accompanied by the systematic decrease in REE from the synplutonic tonalite to the monzogranite [Figs 10 and 13 (below)].
|
This linear relationship is not perfect; there is considerable scatter, especially in K2O, Ba, Sr, and Rb. In view of the variable abundance of alkali-feldspar ± plagioclase phenocrysts, we ascribe the scatter of K2O, Ba, and Sr to variable accumulation of alkali-feldspar and plagioclase. Extreme examples of such accumulation are displayed in zones of the pluton that consist almost entirely of flow-aligned alkali-feldspar phenocrysts. Scatter of Rb concentrations may result from accumulation of biotite, but could also result from Rb mobility in deuteric fluids.
The compositional variation of the early stage was modeled as a fractional crystallization process, combined assimilation–fractional crystallization (AFC), and magma mixing. Fractional crystallization was modeled with major element mass balance calculations (Bryan et al., 1969) in which the composition of the synplutonic tonalitic dike (sample BB-70-94) was used as the parental composition. The results (Table 6) indicate that the major element compositional variation in the early stage can be explained by removal of plagioclase + ilmenite ± hornblende ± biotite. Although these calculations yield acceptable sum of squares of residuals (<0·4), the calculated cumulate assemblages are enriched in TiO2, FeO, and Al2O3 (Table 6). Such cumulate compositions are rather unusual and have not been reported from the area. Furthermore, the mass balance results are inconsistent with variations in the REE among the granodiorite unit. In particular, plagioclase typically constitutes >60% of the cumulus assemblage, which should result in an increase in the size of the Eu anomaly with fractionation. However, within the granodioritic unit, the Eu anomaly is constant or decreases slightly with differentiation.
|
Assimilation–fractional crystallization models suffer from the same limitations as fractional crystallization models. In particular, AFC that involved separation of plagioclase, as required by decreases in Sr abundances, should also result in an increase in the size of Eu anomalies with differentiation. Furthermore, AFC models require a correlation between elemental and isotopic compositions, which is lacking in the granodioritic unit.
Magma mixing should result in hyperbolic compositional variation in ratio–ratio plots, with linear variation in companion plots (Langmuir et al., 1978). Figure 12 shows two such ratio–ratio plots and their associated companion plots for early stage samples. The ratio–ratio plots (Fig 12a and d) are clearly curvilinear and the companion plots (Fig. 12b, c, e and f) are generally linear. The poor correlation seen in Fig. 12f results from poor correlation among Sc and Nb. This poor correlation is thought to result from variable abundances of mafic silicates and titanite in the least evolved members of the granodiorite of Toyn Creek.
|
Mixing was also modeled with major-element mass-balance calculations, with the synplutonic tonalitic dike as the ‘mafic’ end member. A number of granitic end members were used, including evolved members of the Corral Creek unit and a variety of late-stage monzogranitic compositions. An evolved sample of the Corral Creek unit (sample HP-37-95) provided the only consistent results for the entire range of granodioritic compositions (Table 6). As a test of the major element calculations, the calculated proportion of mafic end member was plotted against trace element concentrations (Fig. 13). Within the limits imposed by probable crystal accumulation, these plots show a linear relationship between trace element abundances and the calculated proportion of mafic component.
The ‘mafic’ end member of the mixing trend is a tonalite with relatively high SiO2 and low MgO contents. These features, combined with its relatively high K2O, Ba, and Rb contents, suggest that the tonalite was not primitive (adakitic; e.g. Drummond & Defant, 1990) but had undergone differentiation before the mixing event. This inference is supported by the high whole-rock 18O value of the tonalite (+8·8), which is identical to that of the host granodiorite and is suggestive of pre-mixing assimilation of high-18O crustal rocks in the tonalitic magma before mixing.
Although mafic microgranular enclaves are common in the early stage, and particularly in the granodiorite of Toyn Creek, their importance in the mixing process is difficult to assess. Their compositions vary widely (Figs 6–11) and they commonly do not plot at the mafic end of the trend defined by their host rocks (e.g. Fig. 8b). Furthermore, some enclaves are not in mineralogical equilibrium with their host rocks: amphibole-bearing enclaves are present in the biotite monzogranite of Corral Creek. Relative to the tonalitic mafic end member of the early stage trend, most enclaves have lower SiO2, Al2O3, Ba, and Sr and higher REE. Such modification of an original tonalitic magma could arise from mingling with the host magma (as exemplified by ocellar quartz and K-feldspar ‘phenocrysts’), diffusional exchange with, and infiltration of the host magma (Eburz & Nicholls, 1990; Allen, 1991; Barbarin & Didier, 1992), and fractional crystallization within the enclave magma (Eberz & Nicholls, 1990). Diffusional exchange may explain the scatter of K2O and Rb contents in the enclaves (Figs 6 and 9); however, some combination of mingling, infiltration, and fractional crystallization is required to explain derivation of all enclave compositions from a single parental magma. We find it more likely that the range of enclave compositions reflect local injection and mingling of diverse mafic magmas (e.g. Barnes et al., 1986; Tobisch et al., 1997) after the principal mixing event, followed by some combination of post-mingling processes. Detailed study of enclave compositions and mineralogy is required to adequately assess this idea.
The equigranular roof zone monzogranite cannot be part of the mixing trend discussed above because the mg-number of the roof granites is too low (Fig. 8a). An origin by fractional crystallization was modeled by mass balance calculations. The results (Table 6) indicate that the equigranular roof granites could have formed by fractional crystallization of evolved granodioritic magmas of the Toyn Creek unit or of monzogranitic magmas of the Corral Creek unit. Either parental magma is consistent with trace element data. We prefer the Toyn Creek parent magma because most of the equigranular granites crop out structurally above the Toyn Creek unit. Fractionation probably occurred by rise of residual melt from the underlying crystal mushes (Mahood & Cornejo, 1992). In support of a Toyn Creek parental magma, we found that the calculated cumulate compositions from a Toyn Creek parent are slightly more refractory than typical Toyn Creek granodiorite. In contrast, calculated cumulates from a Corral Creek parent are similar to siliceous tonalite, a rock type that was not observed in the pluton.
Origin of the granitic magmas
In the previous sections, we have shown that the early stage monzogranite of Corral Creek is not the result of fractional crystallization or AFC from more mafic magmas. An origin of late-stage monzogranites from early stage magmas by fractional crystallization or AFC is also unlikely in view of the general decrease in Na2O among samples from the early stage followed by the abrupt increase in Na2O among late-stage monzogranites (Fig. 6). Neither fractional crystallization nor AFC can explain the constant or decreasing negative Eu anomaly with increasing SiO2 among the rocks of the early stage followed by the appearance of large negative anomalies in late-stage granites (Fig. 10).
Within the late stage, the biotite monzogranites display slightly higher CaO contents (Fig. 6), lower ASI (Fig. 5), and lower Rb contents (Fig. 9), which suggests that they are less evolved that the two-mica monzogranites sheets and dikes. However, the late-stage biotite monzogranites have lower mg-number and more highly ‘evolved’ REE patterns, in terms of their positive slopes and deeper negative Eu anomalies (Fig. 10). Finally, the higher initial 87Sr/86Sr, lower Nd, and distinctive REE patterns of the two-mica monzogranite of Green Mountain Creek preclude a direct petrogenetic link with any other unit in the pluton.
Therefore, with the exception of the early stage roof-zone biotite monzogranites, none of the granitic units in the pluton shares a liquid line of descent with any other granitic unit. This suggests that the Harrison Pass pluton contains evidence for emplacement of at least four distinct types of granitic magma: the early stage monzogranite of Corral Creek, the late-stage biotite monzogranites, the late-stage two-mica monzogranite sheets and dikes, and the late-stage monzogranite of Green Mountain Creek. Apparently, each of these granite types originated as magma batches from compositionally heterogeneous crustal source rocks.
On the basis of REE patterns, the source of the early stage monzogranite of Corral Creek would be expected to have small proportions of residual feldspar. In contrast, low abundances of Sr, Ba, Zr, and the REE (especially Eu) in the late-stage biotite monzogranites suggest a source with residual plagioclase, alkali-feldspar, zircon, and allanite (e.g. Cameron, 1984). Thus, if these two-biotite monzogranite units have a common source, then the older Corral Creek unit represents a much larger melt fraction. The weakly peraluminous nature and relatively high 18O values of both early and late-stage biotite monzogranite units are consistent either with primitive metasedimentary source rocks or with meta-igneous source rocks, such as metamorphosed granite, granodiorite, or tonalite, or their metavolcanic equivalents.
The strongly peraluminous monzogranite of Green Mountain Creek and two-mica monzogranite sheets and dikes are probably from metapelitic source rocks in which feldspars were residual (e.g. negative Eu anomalies and low Sr and Ba). However, the distinct Sr and Nd isotopic compositions of the monzogranite of Green Mountain Creek indicate that these two granite types had distinct sources. Finally, the surmicaceous and massive quartz enclaves in the monzogranite of Green Mountain Creek are probably refractory material from the source region (Chappell et al., 1987; Didier, 1973). They are consistent with a pelitic source and additionally suggest emplacement of the monzogranite of Green Mountain Creek as a relatively low-temperature, viscous magma.
Tertiary and Late Cretaceous granites in the northern Great Basin were divided into two distinct groups by Wright & Wooden (1991) and Wright & Snoke (1993). These workers showed that samples collected north of an east–west line between the East Humboldt Range and the Ruby Mountains display isotopic trends toward very low Nd and high 87Sr/86Sr values (‘lower array’ in Fig. 11a). South of that line, Tertiary and Late Cretaceous granitic rocks have a smaller range of initial 87Sr/86Sr and higher Nd (‘upper array’ in Fig. 11a). These distinctions were ascribed to the influence of Archean crustal rocks on magma genesis for the ‘lower array’ and Proterozoic rocks for the ‘upper array’. Schauble et al. (1997) noted that among Cenozoic granites in the northern Great Basin, 18O values are typically lower for samples of the ‘lower array’ (+7·2 to +8·4) and higher for samples of the ‘upper array’ (+8·9 to +10·2).
Samples of the Harrison Pass pluton plot between the two arrays, rather than within one or the other (Fig. 11a). This feature suggests that the source rocks contained both Proterozoic-like and Archean-like characteristics. The mixed isotopic signature of the source, combined with compositional heterogeneity observed is typical of the Proterozoic Mojave basement province of the western USA (Bennett & DePaolo, 1987; Wooden & Miller, 1990), which is thought to underlie the Harrison Pass pluton. It could also result from melting of late Proterozoic and early Paleozoic sediments in the region (Farmer & Ball, 1997), assuming that Mesozoic tectonism emplaced such rocks at depths appropriate for partial melting.
|
SCHEMATIC MODEL OF MAGMA GENERATION AND EMPLACEMENT |
|---|
|
TOPABSTRACTINTRODUCTIONSUMMARY OF FIELD RELATIONSPETROGRAPHYGEOCHEMISTRYDISCUSSIONSCHEMATIC MODEL OF MAGMA...SUMMARYAPPENDIX: METHODSREFERENCES |
|---|
An outline of the origin and evolution of the large number of magmatic components of the Harrison Pass pluton is schematically illustrated in Fig. 14. Activity began with emplacement of mafic magmas into heterogeneous lower crust that had isotopic signatures similar to the Mojave basement province. This flux of mafic magma caused crustal melting (e.g. Grunder, 1992
, 1995). Because of the deep crustal conditions, high crustal melt fractions and small proportions of residual feldspar could be achieved; these are characteristic of the felsic end member of the early stage. In addition to deep-seated crustal melting, some mantle-derived mafic magma hybridized with crustal rocks to form tonalitic magmas similar to the mafic end member of the early stage. Our data indicate that these two magmas then mixed to yield the early stage Toyn Creek and Corral Creek magmas.
|
The location and details of the mixing process can only be speculated on at present. Here we present two possible models (Fig. 14), both of which assume that mixing occurred in a middle crustal magma chamber. The basis for this assumption is the 5–6 kbar pressures estimated for amphibole crystallization in the granodioritic magma of Toyn Creek. This corresponds to crustal depths of 17–22 km. The models differ in the assumed composition of the initial magma in the chamber. Model I assumes the chamber was filled with monzogranitic magma of Corral Creek type. Hybrid tonalitic magma (the mafic end member) rose into this magma body and mixed with it in variable proportions to yield the granodioritic magma of Toyn Creek and the less evolved parts of the Corral Creek magma. Model II assumes that the middle crustal chamber was filled with hybrid tonalitic magma (the mafic end member) and that monzogranitic magmas of Corral Creek type entered the chamber from below. As in model I, mixing in variable proportions gave rise to the early stage magmas.
In model I, mafic enclaves in the early stage magmas resulted from mingling of small batches of mafic magma injected after the principal mixing event. In model II, mafic enclaves could have been entrained in the Corral Creek magma before mixing with the hybrid tonalite, or they could have formed after the principal mixing event.
Following the mixing event, the Toyn Creek magma rose to the level of emplacement. It began to solidify but was still a crystal–liquid mush when the Corral Creek magma was emplaced. Early stage magmas were emplaced at relatively low temperature, as indicated by the phenocryst assemblages in the porphyritic roof dikes. The paucity of alkali-feldspar phenocrysts or megacrysts in the roof dikes compared with their abundance in the underlying rocks suggests either that these crystals grew at the level of emplacement after intrusion of the roof dikes or that large alkali-feldspar phenocrysts were preferentially removed from the roof zone dike magmas by flow sorting. Local dike-like concentrations of alkali-feldspar megacrysts in the granodiorite of Toyn Creek may represent such preferential sorting.
As the early stage magmas solidified, post-emplacement fractional crystallization resulted in leucocratic biotite monzogranite magmas that collected along the roof of the magma chamber (e.g. Fig. 2). These rocks are the only clear-cut examples of in situ fractional crystallization in the pluton.
The early stage probably represents the thermal peak of magmatism in this