The P^T History of Anatectic Pelites of the Northern East Humboldt Range, Nevada: Evidence for Tectonic Loading, Decompression

Home , Anatexis

JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 PAGES1^34 2013 doi:10.1093/petrology/egt057 Journal of Petrology Advance Access published November 13, 2013

The P^T History of Anatectic Pelites of the Northern East Humboldt Range, Nevada: Evidence for Tectonic Loading, Decompression, and Anatexis Downloaded from BENJAMIN W. HALLETT* AND FRANK S. SPEAR

DEPARTMENT OF EARTH AND ENVIRONMENTAL SCIENCES, RENSSELAER POLYTECHNIC INSTITUTE, JONSSON^ ROWLAND SCIENCE CENTER, 1W19, 110 8TH STREET, TROY, NY 12180, USA http://petrology.oxfordjournals.org/

RECEIVED NOVEMBER 1, 2012; ACCEPTED SEPTEMBER 18, 2013

The migmatitic lower plate of the Ruby Mountains^East Hum- INTRODUCTION boldt Range (RM^EHR) metamorphic core complex of northeast- In the lower to middle crust, partial melting of metasedi- ern Nevada represents a deep section of the hinterland of the Sevier mentary rocks dominantly occurs as a result of dehydra- orogenic belt. New major and trace element (Zr-in-rutile, Ti-in-

tion melting reactions of hydrous phases, most commonly at Rensselaer Polytechnic Institute on November 14, 2013 biotite) thermobarometry is aided by garnet zoning analysis and a micas, in conjunction with quartz and plagioclase. thermodynamic modeling approach involving melt reintegration.The Anatexis and migmatization require some combination of P^T history for rocks from the Winchell Lake nappe (WLN) is heating and/or decompression across the effective solidus characterized by a phase of high-temperature, nearly isothermal, (e.g. Groppo et al., 2012), which has a steeply positive slope probably tectonic loading followed by decompression and melting in pressure^temperature (P^T) space. The slope of the during continued heating from 675 to 7408C. Rocks of the Liz- P^T path crossing this solidus is critical to the interpret- zies Basin Block (LBB) below the emplacement fault for the ation of the tectonic evolution that led to anatexis, as well WLN allochthon record no evidence of a nearly isothermal phase of as the tectonic implications of the production of significant tectonic loading and associated metamorphism. Hence a different volumes of partial melt. P^T path characterized by more widespread melting during pro- The presence of even a small volume of distributed par- grade heating and contraction is compiled for the LBB. Different in- terpretations are presented for the exhumation history and the tial melt (several per cent) can have a profound effect on context of partial melting in the WLN versus the LBB. Anatexis in the bulk rheological properties of a crustal section of rock pelites of the WLN occurred as a result of dehydration melting reac- (Arzi, 1978; Paquet et al., 1981; Hollister & Crawford, 1986; tions that took place in a decompressional phase of the prograde Rushmer, 1996; Rosenberg & Handy, 2005; Vanderhaeghe, P^T path, in contrast to dehydration melting during contraction in 2009). Anatexis and leucogranite generation are demon- the LBB.The emplacement of the WLN must have been a signifi- strated to play a significant role in crustal differentiation cant event in the tectonic history of the RM^EHR. These results (e.g. Brown,1994), in addition to contributing to a decrease have important implications regarding the relationship between crus- in bulk-rock strength. Studies in the Himalaya have tal thickening, syncontractional exhumation, and anatexis within pointed to anatexis and leucogranite magmatism as both continental orogenic systems. a trigger (e.g. Searle et al., 2010) and a result of exhumation of a mid-crustal section or ‘channel’ (Patin‹ o Douce & Harris, 1998; Harris et al., 2004). Decompression of hot KEY WORDS: migmatite; garnet; partial melting; metamorphic core crust was, at least in part, a result of syncontractional complex; zirconium; rutile normal-sense motion on the South Tibet Detachment

The Author 2013. Published by Oxford University Press. All *Corresponding author. Telephone: 518-276-3358. Fax: 518-276-2012. rights reserved. For Permissions, please e-mail: journals.permissions@ E-mail: [email protected] oup.com JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

system (Harris & Massey, 1994; Patin‹ o Douce & Harris, inherently linked (Batum, 1999; McGrew et al., 2000; Lee 1998; Harris et al., 2004; Searle et al., 2010). et al., 2003). Rocks of the hinterland of the Sevier Orogenic belt in Protoliths of the RMÊHR metamorphic infrastructure western North America record tectonism related to crustal include Archean^Paleoproterozoic basement gneisses and thickening and contraction along an Andean-style contin- Neoproterozoic^Paleozoic miogeosynclinal sedimentary ental margin. The nexus of crustal thickening occurred rocks including siliciclastic, calcareous, and pelitic rocks west of the preserved Sevier fold and thrust belt at the (Howard, 1980; Lush et al., 1988; McGrew et al., 2000). western edge of the North American craton, in a region Two Jurassic granitoid plutons are preserved in the Ruby that has since experienced protracted Cenozoic magma- Mountains (Hudec, 1992; Jones, 1999b; Howard et al., 2011), tism and extension. Exposures of the Sevier hinterland yet no similar intrusions are recognized in the East are limited to exhumed footwall blocks in a series of meta- Humboldt Range. A suite of granitoid intrusive rocks of morphic core complexes stretching from Mexico to the Cretaceous to Early Oligocene age is ubiquitous through- Downloaded from Canadian Cordillera (Fig. 1; Armstrong, 1982). Mo de r n out the RMÊHR infrastructure. In the Ruby Mountains, surface expressions of this mid-crustal belt are much more these rocks include possibly five generations of leucogra- limited south of the Snake River Plain. Miocene to present nites (c. 92 Ma, possibly 83 Ma, 69 Ma, 38 Ma, and 29 Basin and Range extension has exposed the footwall Ma), an Eocene biotite^hornblende^quartz diorite, and block of the Ruby MountainsÊast Humboldt Range Oligocene biotite monzogranites (U^Pb zircon and mona- http://petrology.oxfordjournals.org/ (RMÊHR) metamorphic core complex in northeastern zite ages; Wright & Snoke, 1993; MacCready et al., 1997; Nevada. The RMÊHR represents a section of the middle Howard et al., 2011). All of these intrusions are penetra- to lower crust affected by Sevier tectonism and associated tively deformed by top-to-the-WNW mylonitic shearing magmatism (Fig. 2). The P^T and melting history of these rocks has significant bearing on the interplay of major deformation and anatexis as they relate to the evolution of batholith an orogenic system. metamorphic This study presents new petrological data as well as core complex Canada prominent major and accessory phase thermobarometric analysis to U.S.A. Laramide fault improve constraints on P^T paths for migmatites of the Cordilleran/ northern East Humboldt Range. In addition, the prograde Sevier thrust belt at Rensselaer Polytechnic Institute on November 14, 2013 and retrograde reaction history based on complex major and trace element zoning in garnet and thermodynamic modeling is presented. This contribution addresses the cause and significance of partial melting as it relates to decompression and heating. The results have implications regarding the nature of thickening and the mechanisms of exhumation of continental crust with parallels to active contractional orogenic systems such as the Himalaya and Snake RiverARG Plain Andes.

GEOLOGICAL SETTING

The Ruby MountainsÊast Humboldt Range meta- RM–EHR SR morphic core complex (Davis & Coney, 1979), like other Cordilleran core complexes, is characterized by unmeta- morphosed supracrustal rocks separated from metamorphic and intrusive rocks by a low-angle, top-to-the-NNW, mylonitic ductile shear zone and associated brittle normal fault system. The infrastructure, or footwall of the core complex shear zone, contains abundant leucogranitic N dikes and sills among other Jurassic to Tertiary intrusions. 200 km The leucogranite dikes and sills are concordant with, and cross cut the preserved subhorizontal metamorphic fabric. Fig. 1. Sketch map of part of the western North American Cordillera At least one generation of leucogranites has been showing the Ruby MountainsÊast Humboldt Range (RMÊHR) interpreted to be genetically linked with distinct in situ metamorphic core complex (black box), one of few exposures of the root of the Sevier orogenic belt in the Great Basin; ARG, Albion^ leucosomes in gneiss and metapelite, suggesting that the Raft River^Grouse Creek complex; SR, Snake Range complex (after magmatic and metamorphic history of this range are Davis & Coney,1979; Foster et al., 2007).

2 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE 115°30’ Clover Hill 91–72Ma z RM-9 published ge U-Pb sample ogn leucosome an R A’ 29±0.5Ma m,z RM-5 t ld 41° bt monzogr ogn o Wood b Structural Symbols m Hills Normal fault u H

t Winchell Low-angle s detachment fault a A Lake nappe E z

Normal-sense 40±3Ma RM-19 Downloaded from

mylonitic fabric hbl-bt-qtz

dioritic ogn Trend and direction of plunge of major fold z 32Ma z RM-11 84.8±2.8Ma RM-18 leucogr http://petrology.oxfordjournals.org/ 40°45’ bt-ms ogn m 37Ma RM-21 38±2 Ma z RM-13 leucogr ogn bt-hbl-qtz diorite 39Ma m RM-24 29–33Ma z RM-20 leucogr ogn bt monz ogn 84 Ma m RM-12 35±3 Ma z RM-7 grt-ms leucogr 39±1Ma m RM-3 bt granodio gneiss grt-ms leucogn

Lamoille Rock Units at Rensselaer Polytechnic Institute on November 14, 2013 30±4Ma z RM-15 Canyon Hanging wall bt monzogranite Tertiary sedimentary and volcanic rocks 40°30’ 83Ma(?) m RM-4 69 Ma z WRP02-7 Paleozoic to Triassic grt-bt-ms sedimentary rocks leucogr Footwall 26±8Ma z RM-17 ~36 Ma Harrison Northernbt Ruby monz Mountains ogn Pass pluton Igneous and high-grade metamorphic rocks 10 km Mixed Jurassic and N 115°15’ Cretaceous granites 115°

Fig. 2. Geological map of the Ruby MountainsÊast Humboldt Range (RMÊHR) after Sullivan & Snoke (2007). White line with black out- line labelled AÂ’ is the location of the cross-section show in Fig. 3. U^Pb geochronology (Phanerozoic rocks) of Wright & Snoke (1993), MacCready et al.(1997), McGrew et al. (2000), Premo et al.(2008, 2010), and Howard et al.(2011). z, zircon; m, monazite; monzogr or monz, monzogranite; ogn, orthogneiss; granodio, granodiorite; leucogr, leucogranite; leucogn, leucogneiss; other mineral abbreviations after Kretz (1983). associated with the main core complex shear zone, suggest- isoclinal folds (i.e. fold nappes) and a penetrative axial ing that plutonism preceded or was synchronous with core planar foliation (Snoke et al., 1979; Howard, 1980; Lush complex extension. et al., 1988; Snoke & Miller, 1988; MacCready et al., 1997; Contractional deformation is widespread in RMÊHR McGrew et al., 2000). Upper amphibolite- to granulite- infrastructure, manifested as large-scale, recumbent, facies metamorphism, and possibly coeval intrusion of

3 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

pegmatitic leucogranite, has been interpreted as syn-kine- cutting intrusive relationships with plutonic rocks both cut- matic with respect to emplacement of the Winchell Lake ting and affected by penetrative contractional deformation fold nappe (WLN) in the East Humboldt Range (Hodges are well documented in the RM^EHR (MacCready et al., et al., 1992; McGrew et al., 2000; Fig. 3). Late Cretaceous 1997; McGrew et al., 2000; Howard et al., 2011). These field U^Pb ages for zircon rims in the migmatitic Angel Lake relationships suggest that plutonism probably began in a orthogneiss from the core of the WLN have been inter- contractional regime, prior to the Late Cretaceous phase preted to suggest that migmatization and melt crystalliza- of exhumation, and continued, episodically, through the tion occurred at 91^72 Ma (Premo et al., 2008, 2010; initiation of the core complex phase of extension. McGrew & Snoke, 2010). The high-grade phase of tectonism is obscured by a lower grade overprint of biotite þ sillimanite-grade extensional tectonism, evidenced by PREVIOUS PETROLOGICAL localized fibrolitic sillimanite and biotite intergrowths STUDIES Downloaded from with dynamically recrystallized quartz in extensional Various thermobarometric methods have been applied in shear bands (McGrew et al., 2000). Most geothermometers efforts to constrain a P^T path for the RM^EHR and ad- were partially reset on the retrograde path, and, as a jacent ranges. Hodges et al.(1992)applied the Gibbs result, peak temperature and temperatures at which peak method (Spear & Selverstone, 1983) using biotite and pressure was attained are difficult values to constrain.

plagioclase inclusions in zoned garnet porphyroblasts to http://petrology.oxfordjournals.org/ A possible phase of Late Cretaceous decompression in model P^T paths for rocks from the northern Ruby the significantly thickened Sevier hinterland has been dis- Mountains, southern East Humboldt Range, and the cussed in previous studies of Sevier-related metamorphic Clover and Wood Hills areas (see Fig. 2). They proposed belts (e.g. Hodges & Walker, 1992; Camilleri & Chamber- distinctly different P^T paths for different tectonic units: lain, 1997; Hoisch et al., 2002). In the RMÊHR, some rocks adjacent to the RMÊHR core complex mylonite workers have suggested that Late Cretaceous decompres- zone reached 7508Cat6 kbar, whereas kyanite-bearing sion is due to a phase of syncontractional ‘tectonic denuda- rocks from the Clover and Wood Hills area are dominated tion’ (Hodges et al., 1992), perhaps representative of by high-pressure metamorphism and steep, nearly isother- gravitational collapse of overthickened crust (Hodges mal decompression from pressures near 10kbar at et al., 1992; McGrew & Snee, 1994; Camilleri & Chamber- 6008C(Hodges et al.,1992). P^Tconditions during exten- at Rensselaer Polytechnic Institute on November 14, 2013 lain, 1997; McGrew et al., 2000). Surface expressions of sional mylonitization were constrained by assuming Late Cretaceous extensional structures occur in several garnet porphyroclast rims were in equilibrium with recrys- parts of the Sevier hinterland (Camilleri & Chamberlain, tallized matrix biotite, plagioclase, and muscovite, yield- 1997; Druschke et al., 2009a, 2009b). Parts of the Sevier ing 6008Cand3·5 kbar (Hurlow et al., 1991). Estimates hinterland contain sedimentary basins with moderately of peak conditions from the Wood Hills area were inferred thin Late Cretaceous to Eocene deposits (Vandervoort & from theT^XCO2 relationships of calc-silicate assemblages, Schmitt, 1990) and buried normal faults with relatively suggesting highest grade peak temperatures of 620^6408C low-magnitude offset (Long, 2012). These are arguably at 6 kbar in diopside þ dolomite þ quartz-bearing rocks minor features and have been suggested by some workers (Camilleri & Chamberlain, 1997). The Clover Hill area to be consistent with only a small degree of Late Cret- (see Fig. 2) is generally considered a block that was con- aceous to Eocene upper crustal extension (Long, 2012; tinuous with, but fault separated from, the EHR. Peak Miller et al., 2012).The later, core complex phase of exhum- metamorphic conditions at Clover Hill were estimated at ation occurred primarily in the Oligocene as mylonitic 7·2 kbar and 720^7608C(Sicard et al., 2011), though extensional shearing of the main core complex shear abundant kyanite implies that prograde metamorphism zone (Hurlow et al., 1991; Hodges et al., 1992). The full passed through the kyanite field below peak conditions, exhumation history of the high-grade metamorphic infra- consistent with a higher pressure phase of metamorphism structure of the RMÊHR therefore includes two major (Hodges et al., 1992; Snoke, 1992). phases, (1) Late Cretaceous and (2) primarily Oligocene Thermobarometry and stable isotope geochemistry were core complex formation. performed on rocks from the Lizzies^Weeks Basin area The relationship between plutonism, contractional (Fig. 3) in the northern East Humboldt Range (Wickham deformation, and decompressionêxhumation is complex. & Peters, 1990, 1992; Peters & Wickham, 1994, 1995), Episodic Late Cretaceous to Oligocene magmatism focused mainly on marble and calc-silicate units. Peak con- (Wright & Snoke, 1993; Premo et al., 2008; Howard et al., ditions were estimated at 6 kbar and 600^7508C, with a 2011) overlaps 40Ar/39Ar thermochronological constraints retrograde overprint that resulted in amphibole þ epi- for exhumation related to extensional core complex forma- dote garnet assemblages replacing clinopyroxene and tion in the Late Eocene to Miocene (Dallmeyer et al., plagioclase in calc-silicate gneisses (Peters & Wickham, 1986; Dokka et al., 1986; McGrew & Snee, 1994). Cross- 1994). Thermobarometric determinations from metapelitic

4 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

(a) EH21,22 Winchell Lake A’ Nappe Lizzies Basin Block Lizzies Basin EH48,49 EH06,09,10 EH45,46 Weeks EH35 Basin 41ºN EH37

EH30,31 Sample Location Downloaded from Fault

A WLN emplacement fault

1 km 115.1ºW N WLN axial trace http://petrology.oxfordjournals.org/

Biotite Monzogranite Quartzite and schist Leucogranite rich Paragneiss (>65%) Hornblende biotite quartz diorite Marble of Lizzie’s Basin Paragneiss of Angel Lake Leucogranite Biotite-garnet schist of Lizzies Basin Orthogneiss of Angel Lake (Archean) Paleozoic marble with schist and quartzite Paragneiss sequence of Lizzies Basin (b) WLN N A Lizzies Basin Block emplacement ’A fault Winchell Lake Nappe at Rensselaer Polytechnic Institute on November 14, 2013 11000’

10000’

9000’ 8000’ 0 2 4 6 8 10 12 14 16 km

Fig. 3. (a) Oblique aerial view of the northern East Humboldt Range facing NW with geological map projected on topography and satellite imagery (from Google Earth). Image Digital Globe. Image USDA Farm Service Agency. Geological mapping of McGrew et al. (2000). Scale bar is approximate and applies to the foreground. (b) Generalized schematic cross-section A^A’from (a) after McGrew (1992).Vertical ex- aggeration is roughly 2. and metabasic rocks from the WLN and the LBB yield Petrogenetic modeling of these rocks suggests that they peak conditions of 49 kbar and 8008C(McGrew et al., are derived from partial melting of metasedimentary 2000). An array of internally consistent thermobarometry source rocks. Lee et al.(2003)modeled a pegmatitic leuco- results is interpreted to define the retrograde portion of granite prevalent in the Lamoille Canyon area of the the P^T path characterized by decompression from peak Ruby Mountains (Fig. 2), interpreted to be genetically conditions for the collective northern portion of the East similar to leucogranites in Lizzies^Weeks Basin (Fig. 3; Humboldt Range (McGrew et al., 2000). Batum, 1999). The models suggest that these leucogranites Geochemical studies of intrusive rocks in the RM^EHR formed primarily as a result of muscovite dehydration indicate that the Late Cretaceous peraluminous leucogra- melting of metapelitic rocks, possibly the Neoproterozoic nites have compositions that are enriched in 18O(whole- McCoy Creek group mapped in the area (Hudec, rock d18O 12ø; Hodges et al., 1992; McGrew et al., 2000) 1990; Jones, 1999a). This is consistent with evidence for and depleted in heavy rare earth elements (HREE) with fluid-absent anatexis of micaceous metapelites discussed little or no Eu anomaly (Batum, 1999; Lee et al., 2003). below.

5 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

METAPELITES OF THE bands. A small number of thin stringers of quartz þ plagioclase are interpreted as leucosomes in EH30, suggesting NORTHERN EAST HUMBOLDT that some in situ melt also formed in the paragneiss. Felsic RANGE bands differ from the quartz þ plagioclase stringers in that Metapelitic schists, paragneisses, and migmatitic rocks they contain a significant amount of biotite and coarse feld- were collected from within the recumbent south-verging spar porphyroblasts. Biotite within the felsic bands appears Winchell Lake nappe (WLN) and structurally beneath to have a more random orientation than in the mafic the nappe in the Lizzies Basin block (LBB). Samples layers, which may indicate that it was late to grow and EH06ÊH10 are from within a graphitic schist and quartz- formed as a result of melt crystallization reactions. iteunitmappedbyA.J.McGrew(McGrew, 1992; McGrew et al., 2000), interpreted as the youngest pre-metamorphic metasedimentary unit exposed in the East MINERAL ASSEMBLAGES AND Downloaded from Humboldt Range. These samples are locally migmatitic METAMORPHIC TEXTURES (Fig. 4a) and contain pegmatitic leucogranite intrusions Winchell Lake nappe (Fig. 4b), with only EH09 and EH10 showing distinct leu- Metapelitic samples from the WLN (EH06, EH09, EH10, cosomes that are overprinted by penetrative shear fabrics and EH21) are characterized by a peak assemblage of gar- and a lower amphibolite-facies retrograde plagioclase þ net þbiotite þ sillimanite þ ilmenite þ plagioclase þ quartz http://petrology.oxfordjournals.org/ biotite þ sillimanite assemblage. Sample EH21is a garnet- þ apatite þ monazite þ zircon tourmaline graphite bearing mylonitic paragneiss with a restitic composition (þ melt). Relict kyanite porphyroblasts are present in sam- from lower in the stratigraphic section. The protolith of ples EH06ÊH10 (e.g. Fig. 5a). Staurolite inclusions in this rock is probably of Proterozoic age (Angel Lake garnet are present in samples EH10 and EH21 (Fig. 5b^d). Paragneiss of McGrew et al., 2000). Quartz ribbons and Rutile occurs as inclusions in garnet (Fig. 5e) and kyanite, prominent shear bands define a penetrative mylonitic foli- and as a few coarse grains in the matrix. ation. Sample EH22 was collected from the edge of a gar- Sample EH22 contains kyanite inclusions in grossular- netiferous vein rich in anorthitic plagioclase that is rich garnet (Fig. 5f) along the margin of a calc-silicate texturally similar to the retrograde calc-silicate assem- layer. This garnet may have formed during retrograde blages described by Peters & Wickham (1994). reactions involving leucogranite intrusions that led to the at Rensselaer Polytechnic Institute on November 14, 2013 Field mapping (McGrew, 1992; McGrew et al., 2000)of formation of epidote þ garnet amphibole assemblages in the lower limb of the WLN delineates a series of low-angle the calc-silicates (Peters & Wickham, 1994). Howe ve r, no fault contacts, which are concordant with the penetrative kyanite was found in adjacent metapelitic sample EH21 fabric and probably formed during Late Cretaceous tecton- and is not present in most other parts of the Winchell Lake ism. These structures are partly obscured by ubiquitous leu- nappe (McGrew et al., 2000), apart from the graphitic cogranite intrusions of Late Cretaceous to Tertiary age. schist unit (samples EH05ÊH10). No distinct post-meta- For this study, rocks that lie structurally below the mapped morphic structure separates samples EH09 and EH10 from fault^shear zone that cuts across the nose of the WLN EH21. Sample EH21 is characterized by a strong shear above Winchell Lake (the ‘main WLN emplacement fault’) fabric and dynamically recrystallized quartz ribbons, and are considered to be part of the LBB and rocks above this contains sillimanite and abundant biotite (50 % fault^shear zone are part of the WLN. Samples from the volume), suggesting that intense deformation may have LBB (EH30, EH31, EH37, EH45, EH46, and EH49) were helped to transform relict kyanite to sillimanite during collected from paragneiss units and from within a gar- enhanced retrograde metamorphism on the exhumation net þbiotite-rich, locally migmatitic, schist layer in the path. Matrix kyanite in the graphitic schist samples Lizzies^Weeks Basin (Fig. 3).This area has been the subject (EH06, EH09, EH10) in some places is separated from of several petrological studies (Wickham & Peters, 1990; quartz by a reaction rim of plagioclase (Fig. 6a). Other Peters & Wickham; 1994, 1995; Batum, 1999; McGrew matrix grains are surrounded by biotite and commonly et al., 2000). Samples EH45, EH46, and EH49 are from a overgrown by fibrous and prismatic sillimanite (Fig. 6b). garnet þbiotite schist unit in Weeks Basin that is locally Distinct zones or layers of nearly equigranular quartz migmatitic. The leucosome domains do not occur strictly and plagioclase ( tourmaline) are present in samples as foliation-parallel, sub-planar layers, but rather as irregu- EH09 and EH10 (Figs 4a and 6c, d).These are commonly lar, commonly elongate centimeter-scale bodies that are in- separated from the melanosome by biotite-rich selvages, terpreted as amalgamations of partial melt formed in situ and may represent leucosomes formed during partial melt- (Fig. 4c, d and e). At one outcrop the leucosome material ing. The rest of the rock has a more Fe- and Mg-rich bulk appears to form a web-like network (Fig. 4f). composition, rich in biotite, garnet, and sillimanite, which Samples EH30, EH31, and EH37 from garnet-bearing probably represent melanosomes and mesosome material paragneiss contain both mafic (biotite-rich) and felsic surrounding leucosomes. Significant reaction coronas of

6 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE Downloaded from http://petrology.oxfordjournals.org/

1 cm at Rensselaer Polytechnic Institute on November 14, 2013

Fig. 4. Field photographs and digital slab scans showing leucosome textures. (a) In situ leucosome material in graphitic schist on the upper limb of the WLN. (b) Muscovite-bearing pegmatitic leucogranite inWLN graphitic schist. (c) Amalgamation of leucosome (arrow) in garnet þbio- tite schist of Lizzies^Weeks Basin. (d) Lineation normal cut showing irregular leucosome body (arrow) in hand sample from Lizzies^Weeks Basin. (e) Elongate in situ leucosome, sample EH48 from Lizzies^Weeks Basin. (f) Web-like interconnected network of leucosome in garnet þbiotite schist, Lizzies^Weeks Basin.

plagioclase þ biotite þ sillimanite are present around por- biotite þ sillimanite þ plagioclase þ alkali feldspar þ quartz tions of nearly all melanosome garnet in EH09 and EH10 ( melt) þ ilmenite þ apatite þ monazite þ zircon. Pelitic (Fig. 6e). In some places, sillimanite þ biotite pseudo- schist samples EH45, EH46, and EH49 are richer in biotite morphs after garnet are observed. No muscovite is and sillimanite than paragneiss samples EH30, EH31, and observed in the graphite schist samples (EH06^EH10) or EH37, which exhibit a gneissic texture. the mylonitic paragneiss (EH21). Garnet porphyroblasts in LBB metapelites contain inclusions of the prograde phases biotite, quartz, plagioclase, Lizzies Basin Block rutile, apatite, monazite, ilmenite, and rare xenotime. Locally migmatitic metapelitic samples from the LBB Xenotime inclusions, where present, are limited to garnet contain the peak metamorphic assemblage garnet þ cores, although some xenotime grains are found along

7 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013 Downloaded from http://petrology.oxfordjournals.org/ at Rensselaer Polytechnic Institute on November 14, 2013

Fig. 5. Photomicrographs of relict phases within the WLN. Abbreviations after Kretz (1983). (a) Metastable kyanite porphyroblast in sample EH09 matrix. (b) Staurolite inclusion in garnet. (c) Crossed polarized light image showing staurolite and biotite inclusions in garnet. Biotite inclusions in garnet core define an earlier fabric (Si), discordant to the mylonitic foliation in the matrix. (d) Ilmenite, quartz, and staurolite inclusions in poikiloblastic garnet. (e) Relatively coarse rutile inclusions in garnet occur along with fine (510 mm thick) rutile needles (not visible at this scale). (f) Ilmenite and kyanite inclusions in garnet porphyroblast.

garnet rims, within plagioclase-rich zones that may repre- Kyanite and staurolite are not present in any of the sam- sent reaction rims. Some garnet porphyroblasts, particu- ples from the LBB. Rutile occurs mainly as inclusions in larly those adjacent to or within quartzofeldspathic layers garnet, but some matrix rutile is present along the edges of that may be leucosomes, contain an inclusion-poor mantle ilmenite associated mainly with biotite and muscovite. LBB zone surrounding an inclusion-rich core. schist units are dominated by domains rich in biotite and

8 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE Downloaded from http://petrology.oxfordjournals.org/ at Rensselaer Polytechnic Institute on November 14, 2013

Fig. 6. Photomicrographs and digital scan of partial melt and decompression textures within the WLN. (a) Metastable kyanite porphyroblast separated from quartz by a reaction rim of plagioclase that may have formed during melt crystallization. (b) Relict kyanite and reaction rim of plagioclase overgrown by coarse sillimanite. (c) Tourmaline-bearing leucosome overprinted by quartz ribbon textures formed during mylonitic shearing. (d) Digital scan of leucosome and melanosome texture in sample EH10. (e) Large garnet porphyroblast with embayed rim and biotite þ sillimanite þ plagioclase corona. (f) Close-up of strongly resorbed garnet surrounded by relatively coarse plagioclase þ biotite and fine sillimanite. This texture may represent a retrograde net transfer reaction resulting in significant garnet breakdown. sillimanite with quartzofeldspathic layers forming 10^20% muscovite is generally highest within 10mm of quartzo- by volume. These felsic layers are separated from the more feldspathic layers that may have been leucosomes. mafic assemblage by biotite- and muscovite-rich selvages in Reaction textures in LBB metapelites are generally con- onesample(EH30;Fig. 7a). Muscovite is present both as sistent with retrograde processes. Garnet porphyroblasts grains parallel to the dominant foliation and as grains that in some samples are irregularly shaped and strongly appear to cross-cut the fabric. The volumetric proportion of embayed (Fig. 7b and c), surrounded by zones of nearly

9 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013 Downloaded from http://petrology.oxfordjournals.org/ at Rensselaer Polytechnic Institute on November 14, 2013

Fig. 7. Photomicrographs of leucosome and retrograde textures from the LBB. (a) Sillimanite-bearing leucosome with biotite þ sillimanite selvages. (b) Leucosome garnet with quartz and plagioclase inclusions (circled) in core and inclusion-free rim, partially resorbed and surrounded by biotite þ sillimanite þ plagioclase. (c) Strongly resorbed garnet porphyroblast in paragneiss. (d) Biotite-rich melanosomes with ‘cross-cutting’ muscovite (top, right) and muscovite þ plagioclase filling pull-apart structures in prismatic sillimanite. (e) Relict alkali feldspar (kfs) in leucosome surrounded by myrmekite may represent melt crystallization reaction (10r). (f) Biotite þ sillimanite þ quartz pseudomorph after what may have been garnet(?) in melanosome suggests a higher modal amount of garnet in the peak assemblage.

þ þ equigranular quartz plagioclase biotite. Prismatic silli- MINERAL CHEMISTRY AND manite in some samples shows pull-apart structures filled by muscovite þ plagioclase (Fig. 7d). Most alkali feldspar ZONING IN METAPELITE present occurs in leucosomes, where it is commonly asso- The results of geochemical analysis of key metamorphic ciated with muscovite and plagioclase showing myrmekitic phases, including mineral zoning, performed by electron intergrowths of quartz (Fig. 7e). microprobe, are discussed below. Detailed analytical

10 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

methods and results are summarized in Electronic Rayleigh fractionation-like XGrs peak or annulus would be Appendix 1 (downloadable from http://www.petrology. expected (Spear et al., 1999), whereas lower XGrs garnet oxfordjournals.org). growth would continue beyond this Ca peak. This is not observed, and in most garnets the highest XGrs zones are Garnet smooth. As noted above, significant shifts in the XSps and Metamorphic garnet is particularly useful in deciphering Fe/ ( Fe þ Mg) occur within the higher XGrs mantles. High- the petrogenetic history of complex metamorphic terranes XGrs mantles are also discontinuous, suggesting that sig- because it is refractory (Hollister,1969) and grows as a con- nificant resorption took place prior to growth of a lower sequence of a variety of metamorphic reactions (e.g. Tr a c y XGrs rim (see arrows in Fig. 8c). Lower XGrs (0·05^0·07) et al., 1976; Spear, 1995). At metamorphic temperatures, rims are cut by resorption textures (Fig. 9b and d). This major and trace elements in garnet diffuse at different garnet may be relict ‘anatectic’ garnet that grew during rates, allowing the mineral to preserve growth zoning in melting. Downloaded from some elements whereas other elements are diffusionally The near-rim major element zoning pattern in garnet modified owing to changes in intensive and extensive vari- from the graphitic migmatitic schist samples includes ables (e.g. Hickmott et al., 1987; Spear & Florence, 1992; troughs of 0·01^0·04 XSps and 0·80^0·82 Fe/(Fe þ Mg). Caddick et al., 2010). There is then a strong increase in XSps and Fe/(Fe þ Mg) toward the rim, with maximum values that vary between Winchell Lake nappe http://petrology.oxfordjournals.org/ grains, reaching as high as 0·125 and 0·875 respectively.The Garnet porphyroblasts in the graphitic migmatitic schist variation in X may be due to local kinetic factors during Sps samples (EH09 and EH10) are 2^10mm in diameter retrograde cooling and garnet resorption, possibly related and generally almandine rich (X ¼ 0·68^0·75). They Alm to variations in leucosome volume during retrogression. are characterized by variably preserved, diffusionally Mylonitic paragneiss EH21 contains two types of garnet ‘relaxed’ prograde growth zoning that reflects different porphyroblast. Poikiloblastic garnet shows a simple length scales for diffusion of major and trace elements. zoning pattern that is drastically modified by resorption, This is manifest in most grains as a high-X (0·04^0·07) Sps with a relatively low Ca content (0·05; Fig. 10a) and and high-Fe/(Fe þ Mg) (0·85^0·87) core and rimward sharp increases in X and Fe/(Fe þ Mg) at the rim decrease in these components (Fig. 8a and b). Garnet por- Sps (from 0·02 to 0·07 and from 0·87 to 0·92 respectively; at Rensselaer Polytechnic Institute on November 14, 2013 phyroblastswithnoorpoorlypreservedhigherX and Sps Fig. 10b and c). Resorption of poikiloblastic garnets is prob- Fe/ ( Fe þ Mg) cores also occur in samples EH09 and ably greatly affected by the increased surface area when EH10, generally restricted to what are interpreted as leuco- former inclusions and their grain boundaries are con- some domains. One garnet grain in sample EH10 contains nected to the matrix. Poikiloblastic garnet may represent a partially preserved relatively high-XSps core, but no equivalent high Fe/(Fe þ Mg) zone. Intra-sample vari- garnet grown during anatexis. ations in major element zoning magnitudes may in part Sample EH21 also contains large garnet porphyroblasts reflect thin-section cuts that do not sample the center of with fewer inclusions. Despite the fewer inclusions, these the garnet porphyroblasts. garnets contain an important inclusion suite that records Distinct mantles are observed in garnet Ca zoning in the prograde history of these rocks. Staurolite, ilmenite, biotite and rutile (Fig. 5c and d) inclusions indicate that EH09 and EH10, showing a fairly abrupt increase in XGrs content toward the rim (from 0·085to0·095), which is these garnets grew in the presence of staurolite and prob- more defined in some porphyroblasts (see Figs 8cand9d). ably rapidly during the breakdown of staurolite and possibly chlorite. A staurolite inclusion occurs in the garnet XGrs then continues to rise to a peak (0·11) towards the rim, which may reflect garnet growth at the peak pressure core, which is high in XGrs (0·095; Fig. 10d) and relatively experienced by these rocks. In one sample (EH10a-p, low in XSps (0·02; Fig. 10e), and Fe/(Fe þ Mg) (0·845; garnet 1), the garnet mantle is characterized by a discrete Fig. 10f). Porphyroblastic garnet rims show an important step in Ca content that is only partly preserved (Fig. 9b). and significant decrease in XGrs to 0·075 with a coinci- This feature represents the second significant increase in dent increase in XSps (to 0·06), and Fe/(Fe þ Mg) (to Ca content and may signal garnet growth during one of 0·93). This lower XGrs rim is interpreted to represent two scenarios: (1) a change in the bulk mineralogy of the garnet grown during anatexis. The Fe/(Fe þ Mg) increase rock, such as when melting reactions began to break down near the rim is not a function of the rim grain boundary plagioclase; or (2) a rapid increase in pressure during a phase (i.e. not higher along garnet þbiotite grain bound- phase of garnet growth resulting in a major shift in gar- aries). These patterns are similar to those in garnet from net þ plagioclase Ca partitioning. If the higher XGrs man- samples EH09 and EH10, although no lower XGrs cores tles represented garnet growth during anatexis, with Ca are preserved in EH21garnet. supplied to garnet during continuous breakdown of plagio- Garnet within the WLN preserves several different clase, then an abrupt jump in XGrs and a continued zoning patterns owing to two main processes: (1) garnet

11 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

(a) Fe/(Fe+Mg) (b) Mn (c) XAn ≈ 0.34 Ca

XSps ≈ 0.028 XAn X ≈ 0.105 Grs ≈ 0.41

FM ≈ 0.850

XSps ≈ 0.052 XGrs ≈ 0.087

FM ≈ 0.805 Downloaded from

FM rim XGrs rim XSps rim 1 mm ≈ 0.830–0.855 1 mm ≈ 0.060–0.070 1 mm ≈ 0.071–0.080

(d) (e) (f) rim Y YY

≈ 500 ppm http://petrology.oxfordjournals.org/

Y bdl quartz ilmenite K-feldspar xenotime 2000–4000 ppm Y rutile

zircon

garnet at Rensselaer Polytechnic Institute on November 14, 2013

50 μm 1 mm Fig. 8. X-ray maps of garnet from migmatitic graphitic schist within the WLN (sample EH09; a^e).Values given are mole fractions except Y, which is by weight. (a) Fe/(Fe þ Mg) map with nearly uniform core, trough and increase at the rim. FM ¼ Fe/ ( Fe þ Mg). (b) Mn map shows similar pattern to Fe/(Fe þ Mg) but the difference in the size of the nearly uniform garnet core should be noted. (c) Ca map shows a higher Ca mantle, the thickness of which is variable, and partially preserved lower Ca rims, indicated by white arrows. (d) Y map indicates well-preserved high-Ycore, with a Yannulus, overgrown by a lower Y mantle and higher Y rims. White rectangle indicates the area shown in (f). (e) Sc zoning shows a low-Sc core, moderate mantle, trough, and thin high-Sc rim. (f) Backscattered electron image of polymineralic inclusion in garnet core. Quartz, ilmenite and xenotime may have been stable during garnet growth whereas rutile, zircon and alkali feldspar may actually be detrital grains.

nucleation occurred at several times during the meta- increase in XGrs discussed above (Fig. 8c and d). The sig- morphic history of these rocks; (2) garnets in different tex- nificance of the Y annulus, which is not observed in other tural settings within the same sample preserve different garnets, may be due to (1) kinetic factors affecting the zoning patterns, indicative of the localized scale (less than supply of Y to the garnet rim during garnet growth or (2) thin-section scale) of metamorphic equilibration. The dif- progressive garnet growth in equilibrium with xenotime ference in the zoning patterns between the graphitic mig- during a relatively steep pressure increase. The drop-off matitic schist (EH09 and EH10) and the mylonitic in Y from an annulus as high as 4400 130 ppm (see paragneiss (EH21) can be interpreted to indicate that Fig. 8d) to values below the electron microprobe detection some garnet nucleation in EH21did not occur until stauro- limit (5100ppm) marks the disappearance of xenotime as lite breakdown at relatively high pressure, and hence a Y ‘buffer’ during garnet core growth (Pyle & Spear, EH21 garnet cores grew during the same higher pressure 1999). This assertion is consistent with the presence of conditions at which the high-XGrs garnet mantles grew in xenotime inclusions in the garnet core (see Fig. 8f), EH09 and EH10. but not in the low-Y mantle. An increase in Y along Dramatic Y zoning is seen in samples EH09 and EH10, garnet rims (to 500 ppm) may represent garnet growth characterized by a high-Y core that shows a Y increase to during melting reactions that broke down monazite. The a diffuse annulus and then a slight decrease followed by a higher Y rims correspond to the lower XGrs rims discussed sharp decrease that corresponds to the core^mantle above.

12 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

(a) bt (b) bt Ca

pl pl XGrs ≈ 0.098 XGrs ≈ 0.086 XGrs ≈ grt 0.121

bt Downloaded from pl + qtz pl 2 mm

XAn ≈ 0.40 bt http://petrology.oxfordjournals.org/ XGrsrim ≈ 0.081 qtz XGrs ≈ 0.110 grt XGrs max ≈ 0.117 XGrs ≈ 0.082 XAn ≈ qtz 0.34 + pl

XAn ≈ 0.41 1 mm 1 mm at Rensselaer Polytechnic Institute on November 14, 2013

Fig. 9. Partially resorbed garnet from migmatitic graphitic schist within the WLN (sample EH10). (a) Transmitted light thin-section scan of partially resorbed garnet shown in (b). The biotite þ sillimanite þ plagioclase corona texture surrounding the grain should be noted. (b) Ca X-ray map of garnet in (a) showing lower Ca garnet core, moderate Ca mantle, and higher Ca outer mantle zone. Low-Ca rims are embayed and only partially preserved. Higher Ca outer mantle is significantly higher than more moderate Ca garnet, and may represent garnet grown near peak pressure. (c) Digital scan of garnet shown in (d). Rectangle indicates the location of rutile inclusion discussed in text. (d) Ca map shows low-Ca core, higher Ca mantle that increases to a maximum, and a lower Ca rim that is truncated by resorption.

Lizzies Basin Block occur within and adjacent to leucosome material. XGrs in Garnet porphyroblasts from paragneisses and migmatitic most garnet porphyroblasts from these samples is nearly pelitic schists from the LBB are generally almandine rich uniform at 0·038. One garnet from sample EH49 shows (XAlm ¼ 0·72^0·78). Increases in XSps a n d Fe/ ( Fe þ Mg) a very small and gradational shift to a slightly lower XGrs occur at the rim of all metapelitic garnet from the LBB, mantle (Fig. 11f). Most grains show a slight XGrs decrease suggesting that at least some garnet resorption occurred along garnet rims to as low as 0·031. along the retrograde path. Most grains show cores that Zoning patterns in garnet from paragneiss samples are nearly homogeneous in XSps and Fe/(Fe þ Mg) (EH30, EH31, EH37; Fig. 12a) resemble those found in gar- (Fig. 11a, b, d and e). Intergrain variation in XSps and net þbiotite schist (EH45, EH46, EH49). Similar to the Fe/ ( Fe þ Mg) core compositions is consistent with the in- schist samples, XGrs is nearly uniform (0·037^0·040) with terpretation that diffusion during both prograde and retro- a slight near-rim decrease to 0·035, consistent with diffu- grade metamorphism had a profound affect on the major sion during cooling. Paragneiss garnet is more common in element zoning. LBB garnet does not preserve the distinct melanosomes than in leucosomes or felsic bands, and con- lower XGrs cores that are found in WLN garnets. tains inclusions of predominantly quartz and biotite (some Garnet porphyroblasts in locally migmatitic gar- are poikiloblastic; Fig. 12c). Biotite inclusions in some net þbiotite schist (samples EH45, EH46, and EH49) grains preserve an earlier fabric that is discordant to the occur in both (1) quartz þ plagioclase-rich domains inter- predominant fabric of the matrix (Fig. 12a). preted as leucosome material and (2) biotite þ sillimanite Garnet from the migmatitic garnet þbiotite schist (EH45, domains. The coarsest and most numerous garnet grains EH46, EH49) has relatively high-Y cores (250^400 ppm)

13 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

(a) (b) Ca rim XSps ≈ Mn 0.070–0.074

XGrs ≈ XSps ≈ 0.02 0.070–0.075

XGrs ≈ 0.054 Downloaded from

0.5 mm 0.5 mm

XAn ≈ 0.41

rim FM ≈ 0.92 XAn ≈ 0.34 XGrs ≈ 0.095 FM ≈ 0.87 at Rensselaer Polytechnic Institute on November 14, 2013

XAn ≈ 0.38 XGrs ≈0.075 0.5 mm 1 mm

(e) (f) Fe/(Fe+Mg) Mn

XSps ≈ 0.02

FM ≈ 0.845

rim XSps ≈ 0.06 rim FM ≈ 0.92-0.93

1 mm 1 mm

Fig. 10. X-ray maps of garnets from mylonitic paragneiss (EH21) from near the axis of the WLN. (a) Ca map of fractured, partially resorbed poikiloblastic garnet with preserved higher Ca ‘core’ and lower Ca zone. (b) Mn map shows preserved low-Mn core and high-Mn rim owing to significant garnet resorption. (c) Fe/(Fe þ Mg) zoning follows that of Mn. (d) Ca map of staurolite-bearing (white arrow) garnet porphyroblast from Fig. 5c showing high-Ca core and low-Ca rim. White rectangle indicates rutile inclusion discussed in text. (e) Mn map of garnet shown in (d) showing rim increase in Mn. (f) Fe/(Fe þ Mg) map of garnet in (d) and (e) shows a relatively thick Fe-rich rim, which may reflect a combination of growth and diffusion zoning.

14 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

(a) Fe/(Fe+Mg) (b) Mn (c) P

rim XSps≈ 0.15-0.17 rim FM≈0.90-0.93

FM≈0.868 XSps≈0.065 Downloaded from http://petrology.oxfordjournals.org/ 500 μm 500 μm 500 μm

(d) (e) Mn (f) Ca rim FM≈ 500 μm rim XGrs≈0.038 0.90 rim XSps≈0.14 XGrs≈0.035

FM≈0.84 XSps≈0.04 XGrs≈0.038 at Rensselaer Polytechnic Institute on November 14, 2013

Fe/(Fe+Mg) 500 μm 500 μm

(g) Y (h) Sc (i) P rim <145 ppm Y Apatite Y bdl

~250 ppm Y

500 μm 500 μm 500 μm

Fig. 11. X-ray maps of garnet from migmatitic schist from the LBB: (a^c) aample EH46; (d^i) sample EH49. Values given are mole fractions except Y, which is by weight. (a) Fe/(Fe þ Mg) map showing core^rim increase owing to retrograde diffusion. (b) Mn map showing uniform core and rimward increase. (c) P map showing zones of higher P along garnet rims. (Note high P along grain boundaries and fractures, as well as high-P plagioclase along the garnet rims.) (d) Fe/(Fe þ Mg) map with lowest values in the garnet core. (e) Mn map showing variation in thickness of diffusion zoning, probably owing to variable amounts of garnet resorption along the rim. (f) Nearly uniform garnet Ca map with very minor lower Ca‘mantle’. (g) Y map showing a significantly lower magnitude Ycore compared withWLN garnet. Small, high-Y inclusions include xenotime, apatite, and monazite. (h) Sc map showing core^rim increase. (i) P map with partly preserved P annulus near the rim. Flushed out (white) inclusions in garnet are apatite. (Note high P along grain boundaries and fractures.)

15 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

(b) (a) Mg Ca

XPrp ≈0.140 rim XPrp ≈0.085 FM bt grt ≈0.54 grt XAn≈0.233

pl Downloaded from

1 mm XAn≈0.169 matrix bt FM ≈ 0.60–0.68 1 mm (d) qtz http://petrology.oxfordjournals.org/ (c) EH31 qtz ap rt bt sil bt

grt qtz mnz inclusions grt pl in grt qtz at Rensselaer Polytechnic Institute on November 14, 2013

1 mm 500 μm

Fig. 12. Textures and chemical zoning in LBB sample EH31. (a) Mg X-ray map of garnet and the surrounding matrix showing higher XPrp core and low-XPrp rim. A high-Mg biotite inclusion in garnet, shown by a circle, may be relict from an early phase of metamorphism. FM ¼ Fe/ (Fe þ Mg). (b) Ca X-ray map showing plagioclase zoning in what may have been a garnet-bearing leucosome. (c) Crossed polarized light photomicrograph of poikiloblastic garnet porphyroblast with numerous quartz and biotite inclusions. (d) BSE image of area in white rectangle in (a) showing accessory phase inclusions in garnet that record prograde metamorphism.

and lower Y mantles (below electron microprobe detection P-bearing fluid was present. In sample EH49, P is not uni- limits). Some rims show a slight increase in Yas well, to as form in the high-P zone of garnet, but shows a gradual high as 145 ppm (Fig. 11g). High-Ycores contain both mona- rimward increase with a peak before decreasing to low zite and xenotime inclusions, although no xenotime is values at the rim (Fig. 11i). This P zoning may be related found in the low-Y mantles. Garnet cores may therefore to partitioning with and/or breakdown of phosphate min- have grown in equilibrium with xenotime. Minor xenotime erals (apatite?) during garnet growth, although apatite in- also occurs along resorbed garnet rims. Scandium zoning is clusions are present in the high-P zone (see Fig. 11i). characterized by a relatively uniform low-Sc core (consist- Truncation of the high-P zone may indicate that all of a ent with the higher Y core domain) with a broad increase particular phosphate was consumed during this phase of through the mantle to the highest Sc values along some of garnet growth. Alternatively, the observation that high-P the rims (Fig. 11h). zones occur adjacent to plagioclase may indicate that P Phosphorus zoning is interesting in that near some rims partitions into garnet during garnet growth and plagio- where garnet is touching or nearly touching plagioclase, clase breakdown, perhaps as a result of melting. In either high-P zones occur (Fig. 11c). P in these zones is as high as case, the high-P zones could indicate garnet growth 375 33 ppm, whereas the rest of the garnet is nearly uni- owing to melting reactions involving plagioclase and/or form with 5100ppm P. Very high P is observed along phosphates. A slight increase in the XGrs content is coher- grain boundaries and fractures in samples EH46 and ent with the step in P, consistent with the anatectic EH49 (Fig. 11c and i), and may indicate a late-stage growth of garnet.

16 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

Biotite Sample EH49 shows the widest range in biotite compos- Biotite is abundant in the metapelitic samples of this study. ition (see Electronic Appendix 4), and contains a number The majority of matrix biotite from within the WLN that of biotite inclusions in garnet. Inclusion biotite must have was analyzed shows generally only minor variation of undergone retrograde Fe/Mg diffusive exchange with Fe/ ( Fe þ Mg) within each sample. Therefore it is inferred garnet, yet preserves the highest Ti content for this sample that retrograde (primarily net transfer) reactions probably (up to 0·47), probably representing relict prograde biotite have caused wholesale changes in matrix biotite Fe/Mg grown at near peak temperatures. Biotite along garnet compositions. rims shows a variable composition that forms a trend Brown matrix biotite from the locally migmatitic graph- from higher to lower values of both Ti and Fe/(Fe þ Mg). itic schist unit (EH06, EH09, EH10) has Fe/(Fe þ Mg) of This is probably due to biotite growth during continued 0·46^0·52 and 0·26^0·42 cations of Ti (on a 22 O basis). breakdown of garnet along the retrograde path (e.g. Henry et al., 2005).

In all WLN samples, biotite adjacent to resorbed garnet Downloaded from · · porphyroblasts is generally 0 02^0 05 lower in Fe/ Feldspar (Fe þ Mg). Green biotite shows a large variation of Ti content trending toward significantly lower values (50·10). Within the WLN, plagioclase occurs throughout paragneisses in both mafic and felsic bands. In more mylonitic This green biotite occurs as fine grains within garnet em- samples (e.g. EH21), felsic bands are almost exclusively bayments or reaction coronas. The trend toward lower Ti composed of ribbon quartz. In the graphitic migmatitic http://petrology.oxfordjournals.org/ contents for biotite along garnet rims is interpreted to rep- schist unit, melanosome versus leucosome plagioclase resent retrograde growth of biotite at conditions below shows different zoning patterns. Melanosome plagioclase peak metamorphism. Variations in the Ti content of biotite has a lower X core (0·33^0·36) with a slightly higher along garnet rims seems loosely tied to the composition of An X rim overgrowth (0·38^0·41; see Fig. 8c). Plagioclase the nearby garnet that is being resorbed. This may be due An formed in the leucosome (presumed to have initially crys- to the coincidence of rutile needles in garnet with higher tallized from a melt) has higher X cores with slightly X , suggesting a local kinetic effect on the availability of An Grs lower X rims. A 0·38 X plagioclase inclusion in Ti to substitute into the biotite structure during retrograde An An garnet in sample EH21 occurs in the lower X rim garnet breakdown below rutile stability. Grs domain observed in this sample, suggesting that this por-

Biotite inclusions from the higher X mantle of a garnet at Rensselaer Polytechnic Institute on November 14, 2013 Grs tion of the garnet equilibrated with the rims of matrix porphyroblast in sample EH09 generally record similar to plagioclase, which are of a similar composition. Higher slightly lower Fe/(Fe þ Mg) than matrix grains from this XAn (0·36^0·47) plagioclase forms reaction coronas sample. This is probably a result of (1) biotite ‘entrapment’ around WLN garnet. XAn in the garnet corona varies as a occurring at higher grade conditions, followed by (2) retro- function of the Ca content of the garnet that was replaced. grade exchange with the surrounding garnet resulting in In the LBB, plagioclase from paragneiss samples shows an Fe/(Fe þ Mg) shift toward that of the matrix biotite. more distinct preserved low-XAn cores (0·17 ^ 0 ·23) and Biotite inclusions are small (580 mm wide) and complemen- higher XAn rims (0·23^0·29; e.g. Fig. 12b). Migmatitic tary higher Fe/(Fe þ Mg) garnet halos formed as a result garnet þbiotite schist (samples EH46 and EH49) exhibits of diffusion are very subtle and thin, perhaps in part owing a range in plagioclase compositions near leucosome gar- to rapid cooling from higher temperature conditions. nets. A core^rim relationship is not observed, and XAn in In the LBB metapelites, biotite is brown, generally more the matrix plagioclase ranges from 0·18 to 0·22. Plagioclase Fe rich, and occurs in abundance within melanosome seg- inclusions in garnet (e.g. Fig. 7b) are slightly higher in Ca regations. Matrix biotite from both schist and paragneiss (XAn ¼ 0·24), suggesting that these grains were included in is almost compositionally uniform within each analyzed garnet prior to crystallization of the matrix grains. Plagio- sample, suggesting that biotite homogenized at some point clase in sample EH45, which contains quartz veins but on the P^T path, probably during retrograde meta- does not contain material interpreted as leucosome, is sig- morphism. Matrix biotite in locally migmatitic schists nificantly lower in Ca, with XAn ranging from 0·13 to 0 ·16. and paragneisses (EH30, EH31, EH45, EH46, EH49) has This may indicate that some plagioclase produced on the Fe/ ( Fe þ Mg) of 0·61^ 0·69 and Ti of 0·25^0·42 cations per prograde path did not react during partial melting, and 22 O. A high-Ti biotite outlier in a felsic segregation of melt extraction inhibited retrograde recrystallization or sample EH30, interpreted to represent leucosome, contains growth. 0·47 Ti cations, and may be preserved biotite that grew It is difficult to determine the stable mineral assemblage during melt crystallization. A non-migmatitic paragneiss present when the plagioclase cores grew. Cores of zoned with almost no sillimanite (EH37) contains generally plagioclase from the WLN samples are interpreted to have more magnesian biotite, with Fe/(Fe þ Mg) of 0·52^0·54, grown on some portion of the prograde path. Plagioclase which probably reflects very little retrograde metamorph- grains with lower XAn cores, assumed to have equilibrated ism by retrograde net transfer reactions (ReNTRs). with high-XGrs garnet, and higher XAn rims may record a

17 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

significant decrease in pressure from a phase of higher XGrs generally consistent with previous studies in the East garnet growth. The shift in plagioclase composition could Humboldt Range. However, coupled with zoning analysis, reflect a period of plagioclase breakdown during anatexis and guided by thermodynamic modeling (discussed on the decompression path followed by renewed growth below) of major melting reactions, prograde and retro- with a higher XAn content during melt crystallization. grade compositions were used to estimate conditions at dif- Leucosome plagioclase in migmatitic samples is inter- ferent points along a P^T path. Thermobarometric results preted to have crystallized from a melt phase during cool- were compiled graphically and are summarized in Fig. 13. ing, hence the core to rim anorthite content decreases Constraints on the maximum temperature at peak pres- slightly (observed in sample EH09). sure and the peak temperature were estimated using gar- Alkali feldspar occurs in paragneiss samples from the net þbiotite Fe/Mg exchange thermometry (Hodges & þ WLN as well as in migmatitic garnet biotite schist from Spear, 1982)basedonhigh-XGrs garnet mantles, Fe/ the LBB. Alkali feldspar from the WLN paragneiss is (Fe þ Mg) troughs, and garnet rims. Matrix biotite away Downloaded from mainly interpreted to be metamorphic, produced by mus- from garnet was used in samples interpreted to have covite and possibly continuous biotite dehydration melting experienced ReNTRs. Pressures were estimated by reactions. Metamorphic alkali feldspar analyzed is mainly garnet þ aluminosilicate þ plagioclase þ quartz (GASP) XOr 0·80^0·90, with rare slightly more sodic grains. barometry (Hodges & Crowley, 1985), ga r n et þ horn- Alkali feldspar is also present in many leucogranitic segre- blende þ plagioclase þ quartz barometry (GHPQ), and http://petrology.oxfordjournals.org/ gations throughout the crustal section, including leuco- garnet þ plagioclase þ muscovite þ biotite (GPMB) bar- somes interpreted to have formed in situ in rocks from the ometry (Hodges & Spear,1982; Powell & Holland, 1988). LBB. In one leucogranite gneiss sample (EH35), XOr is lim- Retrograde exchange reactions (ReERs) and ReNTRs ited to the higher values found in metamorphic grains, had a profound effect on the mineral compositions in this · · 0 87^0 90. suite of samples. Retrograde exchange is evidenced by an Rutile increased XPhlogopite component in biotite and a sharp increase in Fe/(Fe þ Mg) in garnet rims. ReNTRs have a Rutile occurs in nearly every sample studied as inclusions similar effect on garnet zoning, although garnet is con- in garnet. Most rutile inclusions are fine needles, although sumed by ReNTRs and shows embayed edges, a reaction coarser anhedral rutile inclusions are also present

corona in the surrounding matrix, and truncated prograde at Rensselaer Polytechnic Institute on November 14, 2013 (Fig. 5e). Rutile needles are generally absent from the low- zoning. ReNTRs have the opposite effect on biotite com- X cores and high-X rims inWLN rocks, and probably Grs Grs position from ReERs, causing a shift in biotite composition represent exsolution of rutile from higher Ti garnet from more phlogopite-rich to more annite-rich. formed at relatively high pressures. WLN graphitic schist samples also contain rutile inclusions in metastable kyanite Thermobarometry must account for retrograde changes porphyroblasts and/or metastable rutile in the matrix. in mineral chemistry and hence X-ray mapping is essential Matrix rutile typically shows partial retrogression to in interpreting the results of major phase thermobarome- ilmenite. Matrix rutile grains are considerably larger, sug- try (see Fig. 14). X-ray mapping of garnet (see Figs 8^12) gesting that finer grains may have been fully replaced by was used to distinguish core domains from mantle troughs ilmenite on the retrograde path. Some occurrences of and diffusionally modified garnet rims. Thin-section scale matrix rutile are noted in LBB paragneiss EH30 along variations in biotite chemistry were also recognized by X- rims of matrix ilmenite. These rutile grains are rare and ray mapping, and as a result biotite near garnet showing probably formed during retrograde exsolution from sur- higher XPhlogopite was interpreted to have been affected by rounding biotite. retrograde Fe/Mg exchange with garnet. Rutile was analyzed for zirconium by electron micro- For a rock with little or no evidence of melting (EH37), probe. X-ray mapping was performed to test for trace peak temperature was estimated using garnet core com- element (Zr, Nb, Cr, Y) zoning in the rutile that was ana- positions and matrix biotite away from garnets, inter- lyzed for Zr. Trace element zoning in rutile inclusions in preted to have not experienced ReERs. Because this garnet is generally minimal. Rutile in contact with ilmen- sample contains neither leucosomes nor sillimanite ite grains shows an increase in Nb, perhaps owing to Nb (except for one retrograde textural occurrence), melting diffusion driven by exclusion of Nb from the ilmenite struc- reactions that produce sillimanite probably did not occur. ture during retrograde rutile breakdown. Rocks in the upper limb of the WLN record a distinctly different history than rocks from the LBB. The retrograde assemblage suggests a similar exhumation path for these THERMOBAROMETRY two blocks. Although the WLN rocks record higher pres- Major elements sure, the emplacement of the WLN may have occurred Independent reaction thermobarometric calculations per- after partial decompression at high temperature and formed on rocks from both the WLN and LBB are during significant anatexis. This requires not only burial

18 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

(a) 12 EH10 (1) Grt rim + Mg Winchell Lake Nappe EH06 adjacent Bt EH09 FM(grt)≈0.896 635º (5 kbars) 10 maxima FM(bt)≈0.630 minima peak P maxima 8 retrograde rims FM≈0.84 EH10 (2) Grt core + matrix Bt 6 886º (8 kbars) EH09 EH21

Pressure (kbar) EH06 4 Zr in rutile, incl. in lower Ca

garnet rims Downloaded from (EH21) 2 Zr in rutile Approximate FM≈0.634 matrix + incl. in Ky and high Ca garnet Solidus EH49 0 500 600 700 800 Temperature (°C) 1 mm http://petrology.oxfordjournals.org/ 12 (b) Fig. 14. Mg X-ray map of garnet and surrounding matrix in LBB EH45 sample EH49. White circles are electron probe spots on garnet and Lizzies Basin Block biotite with FM ¼ Fe/ ( Fe þ Mg) given. Fe/Mg exchange thermometry 10 maxima results are given [calibration of Hodges & Spear (1982)], demonstrat- EH46 minima ing the effects of (1) retrograde Fe/Mg exchange to record retrograde, sub-peak temperatures along garnet rims and (2) retrograde and net retrograde rims 8 EH37 transfer reactions to record false temperatures above the thermal peak. EH31 EH30 6 EH45 at Rensselaer Polytechnic Institute on November 14, 2013 Rutile thermobarometry Pressure (kbar) 4 EH30 EH49 Rutile in EHR metapelites most probably formed as a result of prograde breakdown of ilmenite in the presence EH31 of aluminosilicate and quartz to grow garnet via the EH46 2 GRAIL reaction EH37 Zr in rutile incl. Approximate in garnet Solidus quartz þ aluminosilicate þ ilmenite ¼ almandine þ rutile: 0 ð1Þ 500 600 700 800 Temperature (°C) Although this reaction is a well-calibrated geobarometer (Bohlen et al., 1983), it requires knowledge of the reactive Fig. 13. Thermobarometry results. Reconstructed P^T paths are shown, as discussed in the text. Parallelograms represent geothermo- garnet and ilmenite compositions at the time of rutile meter þ geobarometer pairs calculated using measured compositional growth, and hence application to rocks in which garnet variation and calibrations discussed in the text. (a) WLN results and zoning was modified by diffusion is problematic. Samples inferred P^T path. Peak pressure maxima indicate maximum tem- from the upper limb of the WLN preserve zoning that peratures for peak P, based on high-XGrs garnet and modified matrix biotite. Dashed P^T constraint (sample EH21) is based on Zr-in- may be ‘reminiscent’ of prograde zoning, although garnet rutile thermometry (shown as gray bands for pre- and post-high-XGrs cores are largely homogeneous with respect to almandine, garnet growth). Dashed solidus is from thermodynamic modeling based on sample EH10 and applies to rocks of a similar pelitic bulk spessartine, and pyrope, suggesting modification by diffu- composition. (b) LBB thermobarometry results as in (a). Dashed sol- sion. Hence the XAlm preserved in the cores of these gar- idus is from thermodynamic modeling based on sample EH49. nets probably does not represent the initial garnet composition. Samples EH09 and EH10 contain abundant graphite, which acted to buffer the metamorphic fluid of Paleozoic sedimentary rocks in the WLN, but also high- composition. Assuming that all fluid in the system is due temperature loading prior to decompression and emplace- to dehydration reactions, the maximum XH2O is solely a ment in partially melted lower to middle crust (Fig. 13). function of temperature at fixed pressure for the graphite Whatever event ‘decompressed’ these rocks from 10 kbar present system (Ohmoto & Kerrick, 1977; Connolly & must have begun prior to and probably continued during Cesare, 1993). He nc e, t h e fO2 of the system is in effect buf- in situ partial melting. fered by the presence of graphite to be at close to common

19 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

crustal values at lower temperatures and increasingly equilibration with zircon. It is possible that the Zr-in- reducing relative to common fO2 buffers (e.g. magnetite^ rutile temperatures recorded by rutile inclusions in the 1 wu« stite, Fe3O4 ¼ Fe O þ 2 O2) with higher temperatures. garnet mantle represent the actual temperature of rutile Therefore, ferric iron is minimal in ilmenite and garnet growth. If the temperatures do reflect growth, then the (as suggested by calculation from stoichiometric assump- GRAIL barometer must have been grossly altered by dif- tions; see Electronic Appendix 6) in these rocks. GRAIL fusional alteration of the garnet core composition to sig- barometry using the modified garnet cores and pure nificantly lower XAlm values, giving an erroneous lower ilmenite yields unreliable pressures of 5^6 kbar. estimated pressure, well within the sillimanite field. This The interpretation that significant garnet growth in the would also imply that the rutile þ zircon þ quartz assem- WLN rocks took place in the kyanite field, partially owing blage remained open to higher temperatures, resulting in to the breakdown of staurolite, and the observation that significant diffusion of Zr into rutile during slow heating. rutile occurs as inclusions in garnet, suggests that some por- Alternatively, if rutile growth took place at the tempera- Downloaded from tion of the garnet porphyroblasts equilibrated with rutile tures recorded by the Zr thermometer, then slow heating and ilmenite in the kyanite field. No evidence of sillimanite is not necessary to explain the homogeneous Zr content of before kyanite is found in these rocks. Therefore, rutile prob- the rutile inclusions. Thermodynamic modeling of sample ably formed as part of the prograde assemblage garnet þ EH10 suggests that rutile stability is reached in this case kyanite þ quartz þ plagioclase þ ilmenite þ rutile þ zircon at 6758C and 7·5 kbar, and hence GRAIL barometry http://petrology.oxfordjournals.org/ ( muscovite biotite). Using calculated GRAIL equili- does not represent equilibrium. bria, limited to the kyanite field, with pure ilmenite in equi- The Zr-in-rutile temperatures provide an important librium with the modified garnet composition requires constraint on the temperatures of garnet growth periods rutile growth at temperatures lower than 6008C. Clearly, in the WLN. Rutile inclusions in garnet in samples EH09 garnet continued to grow following rutile crystallization and EH10 occur within the higher XGrs mantle zone and GRAIL equilibration, as the majority of the rutile pre- (Fig. 9f). Although no rutile is found in the low-X core, served in these rocks is as inclusions in garnet. Therefore, Grs the interpretation that the garnet core grew below rutile the garnet in the GRAIL assemblage may have been effect- stability is preferred. The garnet mantle grew during ively cut off from the matrix during continued garnet and/or after equilibration of the rutile inclusions with growth, and subsequently altered by major cation volume

matrix zircon and quartz. Therefore the Zr-in-rutile tem- at Rensselaer Polytechnic Institute on November 14, 2013 diffusion (in sample EH10 to lower X values) at elevated Alm peratures, assuming the system closed to rutile þ matrix temperatures. equilibration during garnet growth, represent a minimum Zirconium-in-rutile for garnet mantle growth. Although the temperatures are below estimates of peak temperature based on melting re- The Zr-in-rutile thermometer (Zack et al., 2004; Watson et al., 2006; Tomkins et al., 2007)wasappliedtorutileinclu- action textures, they suggest that most garnet mantle sions in garnets and matrix rutile from both the WLN and growth occurred along a steep P^T path segment (nearly the LBB (see Ta b l e 1; Fig. 15). X-ray maps of Zr, Nb, and isothermal loading) above the Zr-in-rutile temperature. Fe indicate minor zoning of these trace elements in some Assuming that the lower XGrs core grew in equilibrium rutile grains (Fig. 15b, c and e). Partial rutile replacement with plagioclase and quartz, the garnet mantle records an by ilmenite is recognized in matrix grains or grains con- increase in the XGrs component of 0·025^0·03, which cor- nected to the matrix via cracks in enclosing garnet por- responds to as much as 2^3 kbar of loading assuming equi- phyroblasts. Ilmenite ‘stringers’ and decorations on rutile librium with plagioclase cores. grains are common and probably represent both exsolution InWLN paragneiss (EH21), a rutile inclusion occurs in a of ilmenite from rutile as well as replacement of rutile by il- distinctly different garnet domain from those in EH09 menite during retrograde decompression (Fig. 15d and f). and EH10 (see Figs 10d and 15 f).The lower XGrs garnet sur- Zr-in-rutile temperatures (Ta b l e 1) were calculated using rounding this higher Zr rutile is interpreted to be garnet the calibration of Tomkins et al.(2007)and range from grown during biotite dehydration melting.The rutile inclu- 55008C in matrix grains to 7228C for a rutile inclusion sion is significantly higher in Zr (766 ppm) and gives a in garnet. The majority of rutile occurs as inclusions temperature of 722 þ6/^78C. Another inclusion in the in garnet or kyanite. Rutile inclusions in garnet or kyanite same garnet domain gives a consistent temperature of give temperatures from 675 to 7228C for rocks within 715 48C. There is no evidence to suggest that these com- the WLN (Fig. 15a, e and f). In the LBB, rutile inclusions positions have been subsequently modified by diffusive ex- in garnet give temperatures from 655 to 7008C(Fig. 15g change with garnet, but they may have experienced and h). diffusional homogenization since being enclosed in garnet. In the WLN samples, rutile occurs mainly as inclusions In the case of diffusional homogenization, the Zr-in-rutile in garnet mantles, suggesting that significant garnet results from EH21 represent a minimum temperature for growth proceeded after rutile crystallization and the prograde growth of the surrounding garnet.

20 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

Table 1: Zr in rutile analyses from East Humboldt range metapelites

Sample Grain Context n* Av. Zr 1s SDy exp. SDz 1s SE T (8C) 2s SE 2s SD T Notes (ppm) (ppm) (ppm) (at 8 kbar) T§(8C) (þ/– 8C)

Winchell Lake nappe EH09b-p 4 matrix 48 397 56 18 — 663 — 22/25 2 traverses low-Zr rims 6 matrix 38 461 24 18 — 677 — 9/9 2 traverses slightly lower Zr rims 3 incl in grt 25 452 29 18 — 675 — 11/11 1 traverse no rims EH09a-p 2 incl in ky 24 490 21 18 — 682 — 7/7 1 traverse no rims 4 incl in ky 17 461 25 18 — 677 — 9/9 1 traverse no rims EH10b-p 1 incl in grt 22 462 17 18 3·7 677 1·4 6/7 1 traverse no rims Downloaded from 2 incl in grt 19 528 15 18 3·4 688 1·2 5/5 1 traverse no rims 3 adj to grt 15 508 31 18 — 685 — 5/6 1 traverse slightly lower Zr rims EH21a-p 1 incl in grt 7 766 27 19 — 722 — 6/7 1 traverse slightly higher Zr rims 2 incl in grt 3 706 16 19 9·1 715 2·4 4/4 no traverse small grain 3 incl in grt 4 458 61 18 — 676 — 22/24 no traverse small grain 5 matrix 2 60 26 16·5 — 529 — 46/68 no traverse small grain http://petrology.oxfordjournals.org/ 9 matrix 2 37 5 17 3·550010·8 14/16 no traverse small grain — some matrix grains from EH21 are below electron microprobe detection limit for Zr

Sample Grain Context n* Av. Zr 1s SDy exp. SDz 1s SE T (8C) 2s SE 2s SD Notes (ppm) (ppm) (ppm) (at 8 kbar) T§(8C) T (8C)

Lizzies Basin block EH30b-p 1 incl in grt 3 356 15 18 8·5 655 3·8 7/7 no traverse small grain 3 incl in grt 2 362 28 17 — 656 — 12/13 no traverse small grain — matrix grains from EH30 are below electron microprobe detection limit for Zr at Rensselaer Polytechnic Institute on November 14, 2013 EH37b-p 1 incl in grt 4 601 13 18 4·3 700 1·3 3/3 no traverse small grain 2 incl in grt 4 525 13 18 6·6 688 2·2 4/4 no traverse small grain 3 incl in grt 7 417 31 18 — 668 — 12/13 4 incl in grt 6 606 21 18 — 701 — 6/6 5 incl in grt 4 545 18 18 9·1 691 3 6/6 6 incl in grt 7 578 37 18 — 696 — 11/12 7 incl in grt 8 515 47 18 — 686 — 16/17 EH45a-p 1 incl in grt 1 558 — 18 — 693 — — one pt. tiny grain EH46b-p 2 incl in grt 4 403 34 17 — 665 — 13/14 no traverse small grain 3 incl in grt 8 576 22 18 — 696 — 7/7 4 incl in grt 8 455 38 18 — 675 — 14/15 5 incl in grt 8 502 38 18 — 684 — 13/14 6 incl in grt 4 410 27 18 — 667 — 11/11 7 incl in grt 2 477 53 18 — 679 — 18/20 no traverse incl in ilm EH49a-p 1 incl in grt 7 491 42 18 — 682 — 14/15 2 incl in grt 6 484 23 18 — 681 — 8/8 3 incl in grt 6 468 34 18 — 678 — 12/13 4 incl in grt 6 493 24 18 — 682 — 8/9 5 incl in grt 4 517 15 18 7·4 686 2·6 5/5 6 incl in grt 6 520 52 18 — 687 — 17/18 EH49b-p 1 incl in grt 7 465 24 18 — 677 — 9/9 2 incl in grt 6 413 16 18 6·4 667 2·6 6/6 3 incl in grt 7 456 28 18 — 676 — 10/11 4 incl in grt 7 497 7 18 2·7 683 1 2/3

Analytical protocol is summarized in Electronic Appendix 1. *n, number of spots; total analyses ¼ n 4. yMeasured standard deviation for each grain. zExpected standard deviation based on counting statistics for analytical protocol. §Measured standard error of all temperatures for the grain, given where expected SD measured SD. Measured standard deviation propagated to temperature.

21 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

(a) ap EH09 (b) Nb (c) Zr 461±24 ppm 677±9°C zr n = 38 trav2 bt qtz trav1 grt ap bt 50μm Downloaded from 50μm 50μm

(d) zrn EH10 (e) Zr grt ilm 462±17 ppm 644 +6/-7°C http://petrology.oxfordjournals.org/ n = 22

trav1 rt grt rt qtz rt

zrn at Rensselaer Polytechnic Institute on November 14, 2013 200μm 50μm

(f) EH21 (g) EH46 (h) EH37 766±27 ppm 576±22 ppm 722+6/-7°C 696±7°C 606±21 ppm n = 7 n = 8 trav1 701±6°C n = 6 rut

rut rut bt

grt grt grt ap qtz

20μm 20μm 20μm

Fig. 15. Rutile from the East Humboldt Range: (a^f) are fromWLN; (g, h) are from LBB. White bands labeled ‘trav’ represent electron microprobe traverses. The average Zr content and corresponding temperature are given for each grain. (a) Backscattered electron image (BSE) of matrix rutile from sample EH09 showing Zr analysis locations and results. (b) Nb X-ray map of rutile in (a). (c) Zr X-ray map of rutile in (a). (d) BSE image of rutile and zircon inclusions in garnet from Fig. 9. (e) Zr X-ray map of rutile in white box in (d). (f) BSE image of rutile inclusion from Fig. 10d. (g) Rutile inclusion in garnet with biotite from sample EH46. (g) Rutile inclusion in garnet from sample EH37.

Thermodynamic modeling considerations suggest that the high-XGrs garnet core (mantle in EH09 and EH10) must temperatures of anatectic garnet growth may have been have grown at a temperature below 7408C. Therefore, slightly higher than this value, consistent with some diffu- Zr-in-rutile thermometry for these samples (EH09, EH10, sional modification. Based on this interpretation, the and EH21) indicates that a phase of tectonic loading and

22 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

subsequent decompression occurred in the WLN between therefore presented as an illustration of the interpretation 675 and 7408C. of petrological observations for the WLN. In particular, calculated isopleth maps are not taken as accurate with respect to absolute compositional predictions, but rather rep- THERMODYNAMIC MODELING resent guides with approximations of equilibrium Relict phases, garnet zoning and reaction texture observa- chemical gradients in P^Tspace. tions, combined with thermobarometry, are better inter- Both melt-reintegrated models yield peak pressure and/ preted in conjunction with appropriate equilibrium or peak temperature assemblages that are consistent with assemblage diagrams contoured for phase composition petrographic observations and major phase thermobaro- and abundance. Calculations were performed using the metry with one exception. Alkali feldspar is notably thermodynamic data of Holland & Powell (1998), White absent from the WLN migmatitic schist (EH05^EH10). It et al.(2001), the database of F. S. Spear et al. (unpublished is possible and likely that all alkali feldspar broke down data), and a combination of program Gibbs (Spear & owing to a retrograde net transfer reaction during melt Downloaded from Menard, 1989; Spear et al., 2001) and Theriak^Domino (de crystallization, the same reaction that resulted in signifi- Capitani & Brown, 1987; de Capitani & Petrakakis, 2010). cant garnet resorption and production of biotite and silli- Appropriate generalized bulk compositions in the system manite. LBB migmatitic schist samples (EH45, EH46, MnNCKFMASH( Ti) were calculated with program EH49) do contain alkali feldspar, interpreted to have been Gibbs from estimated volumetric proportions of peak part of the peak metamorphic assemblage. http://petrology.oxfordjournals.org/ phases and measured compositions of solution phases. As EHR migmatites show evidence of melt migration and seg- Modeled solution phase compositions regation from the reactive bulk composition, additional The prograde, pre-melt extraction model presented in melt in equilibrium with modeled solidus phases was Fig. 16 a predicts two major phases of garnet growth in added, using the melt reintegration method of Indares the subsolidus region and minor additional supersolidus et al. (2008). A melt composition was calculated at the mus- garnet growth for the WLN migmatitic schist (EH10). covite dehydration melting reaction for the proposed P^T According to the model, garnet initially began to grow path, at 7488C, 8·6 kbar for a model of WLN sample owing primarily to the breakdown of chlorite, followed by EH10, and at 722·58C, 7·5 kbar for a model of LBB sample little growth or slight resorption at the staurolite-in reac- EH49. This melt was then added to the system until the ef- tion (Fig. 16b). Garnet growth resumed in a staurolite- at Rensselaer Polytechnic Institute on November 14, 2013 fective muscovite dehydration melting reaction shifted to bearing assemblage and grew as a result of staurolite H2O-saturated conditions (equivalent to 10mol % melt breakdown and a continuous reaction in a subsolidus kyan- for EH10, 13mol % melt for EH49). This does not signifi- ite-bearing assemblage to peak pressure and XGrs content cantly change the temperature of the effective solidus (in- (Fig. 16b and c). At this point decompression accompanied terpreted as the muscovite-out^alkali feldspar-in reaction) resorption, although very little change in the XSps content at the expected pressure range for melting, but it does is depicted (Fig. 16d). Crossing the muscovite dehydration affect the stability fields of a number of subsolidus phases melting reaction results in some anatectic garnet growth interpreted to have been part of the prograde reaction to peak temperatures, followed by resorption during cool- history. ing and melt crystallization. Along the P^T path, major Composite equilibrium assemblage models were pro- resorption is predicted to occur on steeply decompressional duced (Figs 16 and 17) with the melt-reintegrated compos- segments below the solidus, and shallower cooling seg- ition, using a fixed H2O content based on the reintegrated ments in melt-bearing assemblages. melt composition for the suprasolidus portion (including The EH10 (WLN) garnet composition during growth the solidus), and an H2O-saturated composition for all of along the proposed P^T path is initially XSps-rich and the subsolidus fields. In subsolidus fields, H2O is treated with moderate XGrs (see Fig. 16c and d). Garnet inter- as a fractionated or isolated phase and hence this portion preted to have grown along the prograde P^T segment of the model was calculated using a melt-free database. below staurolite stability (garnet zone) may not have been Calculations were performed, and equilibrium assemblage in equilibrium with plagioclase, as the plagioclase present diagrams were constructed, with Theriak^Domino. A may have been detrital and not entirely reactive, acting to number of assumptions are required to apply thermo- sequester Ca from the growing garnet. More moderate dynamic modeling to the interpretation of petrological ob- XSps garnet is predicted to grow in the presence of stauro- servations. This method does not account for (1) fractional lite, followed by relatively low-XSps garnet in the kyanite- crystallization, (2) subsolidus fluid flux into or out of the bearing field. Garnet grown in the kyanite field along system, (3) the complete chemical composition including the steep loading segment of the P^T path will become 3þ Fe and minor elements, (4) possible multiple or pro- increasingly XGrs-rich. Finally, anatectic garnet is signifi- tracted melt loss events and (5) incorporation of Fe and cantly lower in XGrs than subsolidus garnet grown near Mg into the melt. Equilibrium assemblage diagrams are peak pressures. Model equilibrium XAn compositions

23 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

12 (a) (A) Grt Ky Grt Kfs WLN Ms Pl Qtz Ky Pl Bt Qtz L Chl Grt Chl Ms Qtz 10 Grt Bt Grt Grt Kfs Chl Cld Ms Ky Ms Pl Qtz Qtz Pl Qtz Sil L Grt Ms effective solidus Qtz +Pl 9

-St 8 Bt Grt Grt Pl +Ky Kfs Pl Qtz Sil Bt Qtz Sil L Grt L (B) (D) Ms Pl (C) 6 Qtz Sil Crd Grt 6 (E) Kfs Pl (F) (G) Chl Cld -Chl Qtz Sil Grt Ms L Pl Qtz Bt Grt Pl Downloaded from Qtz Sil +Bt (I)

Pressure (kbar) (H) Crd Grt Crd Grt 4 Bt Crd +Kfs Pl L Grt Pl -Sil Pl Kfs 3,5 Qtz Sil +Grt Qtz L Chl Chl -Grt Cld Ms And Bt Crd Ms Crd Bt Pl Qtz 2 Grt Pl (K) -Sil Pl Qtz Grt Pl Qtz Kfs

Qtz Crd http://petrology.oxfordjournals.org/ 2 (J) L Crd Pl L And Bt Crd Grt Ms Pl Qtz Bt Crd Pl Kfs Kfs Pl Qtz L +Grt -Bt Crd H O Pl L 0 2 400 500 600 700 800 900 1000 Temperature (°C)

(b) 12 vol% grt 20 40

17.5

35 15 22.5 at Rensselaer Polytechnic Institute on November 14, 2013

12.5

30 10 10 20

17.5 45

20 15

8 12.5 10 10 7.5

7.5 50

10 25 5 6 7.5 5 2.5 20 15 15 2.5

Pressure (kbar) 12.5 7.5 5 4 10 2.5 2.5

5 2 2.5 7.5

0 400 500 600 700 800 900 1000 Temperature (°C) Fig. 16. Equilibrium assemblage diagrams in the MnNCKFMASH system constructed using thermodynamic data of Holland & Powell (1998) and the melt model of White et al.(2001). (See Supplementary Data Electronic Appendix 5 for activity models and Ta b l e 2 for bulk compositions used.) L, liquid (melt). (a) Assemblage stability fields for prograde composition based on WLN sample EH10 with reintegrated melt. Subsolidus region is entirely vapor saturated, using an H2O-enriched bulk composition. (See text for discussion.) Dot pattern indicates peak assemblage. Shading darkness correlates to assemblage variance. Dashed thick gray line with arrow represents inferred P^T path, labeled with reaction numbers from text. Field labels: (A) Bt þ Grt þ Ms þ Qtz; (B) Chl þ Grt þ Ms þ Pl þ Qtz; (C) Chl þ Grt þ Ms þ Pl þ Qtz þ St; (D) Bt þ Grt þ Ms þ Pl þ Qtz þ St; (E) Bt þ Grt þ Pl þ Qtz þ Sil þ L; (F) Crd þ Grt þ Pl þ Qtz þ Sil þ L; (G) Crd þ Grt þ Pl þ Sil þ L; (H) Chl þ Grt þ Ms þ Pl þ Qtz; (I) Bt þ Crd þ Grt þ Pl þ Qtz þ Sil þ L; (J) Bt þ Chl þ Cld þ Pl þ Qtz; (K) And þ Bt þ Crd þ Kfs þ Pl þ Qtz. (b) As in (a) with volume per cent isopleths of garnet. Some isopleths are omitted for clarity. Black P^T path segments represent garnet growth; white segments are garnet resorption. (c) As in (a) with isopleths of XGrs. (d) As in (a) with isopleths of XSps. (e) As in (a) with isopleths of XAn in plagioclase. Some isopleths are omitted for clarity. (continued)

24 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

0.12 0.06

0.09 0.11 10 0.10

0.08

0.05

0.07

0.04

6 Downloaded from Pressure (kbar) 4

grt in http://petrology.oxfordjournals.org/ 2

0.05

0 400 500 600 700 800 900 1000 Temperature (°C)

(d) 12

XSps at Rensselaer Polytechnic Institute on November 14, 2013

8 0.04 0.04

0.06

0.08 0.04 0.06 0.06 6 0.08 0.10 0.08 0.10 0.12 0.12

0.10

0.16 0.16

0.04 Pressure (kbar) 0.06 0.14 0.08 4 0.100.12

grt in 2

0.08 0.06

0.04 0 400 500 600 700 800 900 1000 Temperature (°C) Fig. 16. Continued.

25 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

(e) 12 0.20 0.32 WLN 0.22 0.34 0.40

XAn 0.24 0.36 0.26 0.38 0.42 10 0.28

0.30 0.44

0.32

8 0.34 0.46 0.48 0.36 0.48

0.40 0.38 6 0.50 pl in Downloaded from

0.42 Pressure (kbar) 0.52 4

0.40 0.46

0.44 0.58

0.42 http://petrology.oxfordjournals.org/

0.60 0.62 0.54 2 0.56 0.64 0.68 0.66

0.42 0.58 0.44

0.46 0 400 500 600 700 800 900 1000 Temperature (°C) Fig. 16. Continued. are consistent with plagioclase growth during a higher to the equilibrium assemblage diagram for a WLN rock at Rensselaer Polytechnic Institute on November 14, 2013 pressure metamorphic phase in the subsolidus region (Fig. 16a) and an LBB rock (Fig. 17a). Reaction numbers (Fig. 16e), overgrown by plagioclase during retrograde presented in the text are linked to Figs 16aand17 a. melt crystallization following melt extraction or loss. Garnet in the EH49 (LBB) model is inferred to have Winchell Lake nappe been almost completely resorbed in the presence of stauro- In the WLN graphitic schist (EH06, EH09, EH10), some lite, suggesting that garnet may have re-nucleated during garnets contain a relict core, in some cases rich in quartz staurolite breakdown (Fig. 17a). Modeled isopleths of Fe/ inclusions, which grew during prograde garnet zone meta- (Fe þ Mg) in garnet (Fig. 17b) are relatively low at esti- morphism, most probably by the breakdown of chlorite mated peak conditions, suggesting that wholesale changes and chloritoid owing to the garnet-in isograd reaction in the measured Fe/(Fe þ Mg) are dramatically affected chlorite þ chloritoid þ quartz ¼ garnet þ H O: ð2Þ by diffusion on the retrograde path. The proposed pro- 2 grade P^T path through modeled plagioclase compos- Garnet cores do not appear to be resorbed, as euhedrally itions spans a wide range in XAn (Fig. 17c), which is shaped trace element (Y, Sc) zoning is observed, probably difficult to interpret without knowing whether plagioclase representing the disappearance of xenotime from the rock was initially detrital or grew during a phase of prograde (e.g. Pyle & Spear, 1999). With further heating, some metamorphism. It is likely that measured matrix plagio- garnet growth would have continued via the staurolite- clase compositions (XAn ¼ 0·18^0·22 for EH49) entirely re- producing reaction flect growth during retrograde melt crystallization, following extraction of a significant volume of melt. chloritoid þ muscovite þ quartz ¼ garnet þ staurolite þ biotite þ H2O ð3Þ MAJOR PHASE REACTION In lower aluminum WLN pelites, garnet growth probably HISTORY began by the reaction Reactions and textures deduced from petrological analysis þ þ ¼ þ þ are consistent with the results of thermodynamic modeling. chlorite muscovite quartz garnet biotite H2O: An interpreted reaction history is presented, with reference ð4Þ

26 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

12 (a) (C) LBB (A) Grt Kfs Ky Pl Qtz L Bt Grt (B) Bt Chl Ky Ms 10 Grt Ms Pl Qtz Grt Kfs Qtz Chl Grt Pl Qtz Ms Qtz effective solidus Sil L

-St Bt 8 Chl Bt Grt Ms Grt 9 Grt Bt Grt Pl Sil Pl Qtz Ms Pl +Ky Chl Grt Qtz Kfs Pl L - Pl Ms Pl St Qtz Sil Qtz L Crd Grt Bt Grt Crd Grt Pl Sil 6 Ms Pl Crd +St (D) Kfs Pl L

Grt Downloaded from Qtz Sil Qtz Sil -St Kfs Pl 4 L Sil L Crd Pl +Bt Chl Ms Sil L Pressure (kbar) -Grt Pl Qtz (E) Crd Kfs 4 Pl Sil (G) (F) Crd Kfs L Pl Qtz Sil L

(H) Crd Pl http://petrology.oxfordjournals.org/ And Bt Crd Kfs L 2 Crd Ms Pl L Pl Qtz (I) (J) And Crd Kfs Pl Qtz +H2O 0 400 500 600 700 800 900 1000 Temperature (°C)

(b) 12 0.64

LBB at Rensselaer Polytechnic Institute on November 14, 2013 FM Grt 0.68

0.72 10 0.84

0.88 0.76

0.92

0.96 8 0.80

0.64 0.62 0.84 0.68 0.72 6 0.76 0.80 grt in 0.84

0.88 Pressure (kbar) 4

0 400 500 600 700 800 900 1000 Temperature (°C) Fig. 17. Equilibrium assemblage diagrams in the MnNCKFMASH system. L, liquid (melt). Dashed thick gray or black line with arrow represents inferred P^T path. (See Electronic Appendix 5 for activity models and Table 2 for bulk compositions used.) (a) Assemblage stability fields for prograde composition based on LBB sample EH49 with reintegrated melt. Subsolidus region is entirely vapor saturated, using an H2O-enriched bulk composition. (See text for discussion.) Dot pattern indicates peak assemblage. Shading darkness correlates to assemblage variance. P^T path is labelled with reaction numbers from text. Field labels: (A) Bt þ Grt þ Ms þ Qtz; (B) Bt þ Grt þ Kfs þ Ky þ Pl þ Qtz þ L; (C) Grt þ Kfs þ Qtz þ Sil þ L; (D) Bt þ Ms þ Pl þ Qtz þ St; (E) Bt þ Chl þ Ms þ Pl þ Qtz þ St; (F) And þ Bt þ Grt þ Ms þ Pl þ Qtz; (G) Bt þ Crd þ Kfs þ Pl þ Qtz þ Sil þ L; (H) And þ Bt þ Crd þ Kfs þ Pl þ Qtz; (I) And þ Crd þ Kfs þ Pl þ Qtz þ L; (J) And þ Crd þ Kfs þ Pl þ L. (b) As in (a) with Fe/(Fe þ Mg) isopleths for garnet. (c) As in (a) with isopleths of XAn in plagioclase. Some isopleths are omitted for clarity. (continued)

27 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

0.16

10 0.26

pl in 0.28 0.18

0.34 0.30 0.34 8 0.36

0.20 0.38

0.32

0.24 0.42 6 0.46 Downloaded from

0.46 0.38 0.34 0.50 0.44 0.22

Pressure (kbar) 0.42 4

0.26 http://petrology.oxfordjournals.org/

0.20 2 0.40 out pl

0.22

0 400 500 600 700 800 900 1000 Temperature (°C) Fig. 17. Continued. at Rensselaer Polytechnic Institute on November 14, 2013 Garnet growth would be continuous via reaction (4). No It is possible that most kyanite was completely consumed chloritoid inclusions are found in garnet in rocks for by reaction (7) to grow garnet. Alternatively, kyanite may which abundant aluminosilicates reflect an aluminous have been replaced by sillimanite during a decompression bulk composition (i.e. Fig. 16). segment of the prograde path that passed into the silliman- Staurolite growth most probably led to the removal of ite field. chlorite by the reaction At the highest pressure recorded in these rocks, reaction (7) was active along with the GASP reaction garnet þ chlorite þ muscovite ¼ staurolite ð5Þ þ biotite þ quartz þ H2O 3 anorthite ¼ kyanite þ quartz þ grossular ð8Þ which was probably the main staurolite-producing reac- which together resulted in significant XGrs-rich garnet tion. Significant garnet breakdown by this reaction may growth. have occurred in some rocks, although it is not reflected Decompression, starting in the kyanite field, resulted in in chemical zoning profiles. Garnet consumption probably the partial resorption of garnet via the retrograde forms took place above xenotime stability, based on a euhedral of reactions (7) and (8). Theoretically, kyanite and biotite shape to the high-Y zone (Fig. 8d) that shows no evidence growth would have occurred during this initial decompres- of resorption. sion phase, prior to reaching dehydration melting reactions Garnet growth resumed during the breakdown of and/or the sillimanite field. Although nearly all matrix staurolite by a reaction that may have produced large kyanite in samples EH05^EH10 is partly replaced by silli- kyanite porphyroblasts that are present in samples EH09 manite, it is uncertain whether the relict kyanite had and EH10: grown during initial decompression. The samples with staurolite þ muscovite ¼ garnet þ biotite þ kyanite relict kyanite also contain abundant large knots of fibrous ð6Þ sillimanite and prismatic sillimanite grains. Some severely þ quartz þ H O: 2 resorbed relict kyanite is completely rimmed by plagio- Garnet growth probably continued to peak pressure condi- clase with no evidence of replacement by sillimanite (see tions above the staurolite field by the continuous reaction Fig. 6a).This rim may have formed as a result of increased Ca2þ and Naþ intergranular diffusion at near peak meta- biotite þ kyanite þ quartz ¼ garnet þ muscovite: ð7Þ morphic conditions, perhaps related to growth of higher

28 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

Table 2: Bulk compositions used in thermodynamic modeling and equilibrium assemblage diagrams (Figs 16 and 17)(see text for discussion)

SiO2 Al2O3 FeO MnO MgO CaO Na2OK2OH2O

EH10 raw composition 58·559 23·942 8·940 0·343 2·937 1·989 1·011 1·614 0·665 melt at solidus 69·692 14·049 0·436 0·000 0·110 0·581 2·518 3·370 9·244 with 13% melt added* 59·386 23·255 8·342 0·319 2·738 1·891 1·118 1·739 1·212 EH49 raw composition 59·791 26·195 6·161 0·059 2·472 1·033 1·449 2·064 0·775 melt at solidus 70·476 14·030 0·563 0·000 0·124 0·323 3·006 2·937 8·541

with 10% melt addedy 60·827 25·016 5·619 0·053 2·245 0·964 1·600 2·149 1·528 Downloaded from

*Bulk composition used to construct Fig. 16. yBulk composition used to construct Fig. 17. http://petrology.oxfordjournals.org/ XAn rims on matrix plagioclase during melt crystalliza- the presence of melt occurs as embayed garnet rims show- tion. Other kyanite grains are overgrown by sillimanite, ing diffusional Fe/(Fe þ Mg) zoning, as well as artificially suggesting a direct polymorphic phase transformation high garnet core þ biotite Fe/Mg temperatures owing to (Fig. 6b). Muscovite, which is not present and was probably melt-involving retrograde net transfer reactions that pro- not part of the peak temperature assemblage, must have duced XPhlogopite-rich biotite. Alkali feldspar is not present been completely consumed by the primary melting reac- in these rocks, suggesting that it was entirely consumed by tion that produced the leucosomes: the retrograde form of reaction (10): muscovite þ plagioclase þ quartz ¼ sillimanite garnet þ alkali feldspar þ melt ¼ biotite ð9Þ ð10rÞ þ alkali feldspar þ melt: þ plagioclase þ sillimanite þ quartz: at Rensselaer Polytechnic Institute on November 14, 2013 If this reaction occurred in the kyanite field, significant This retrograde net transfer reaction significantly altered kyanite growth would occur during partial melting. the bulk mineralogy of these metapelitic rocks. The ab- Instead, it appears that sillimanite was a product of melt- sence of alkali feldspar and muscovite in these rocks con- ing, requiring initial melting to have occurred at or below firms that (1) significant melt was separated from the the intersection of the muscovite dehydration melting reactive portion of these rocks above the solidus and (2) curve [reaction (10)] with the kyanite^sillimanite join. all of the remaining melt and alkali feldspar after melt sep- Depending on the peak pressure experienced by these aration reacted with garnet to grow the products of reac- rocks, this observation suggests that significant decompres- tion (10r) that dominate the melanosomes of these sion had to have occurred prior to and probably during de- migmatites. Continuous biotite dehydration melting, reac- hydration melting. No evidence is observed that these tion (10), acting to grow anatectic garnet may have been rocks cooled and/or reheated prior to melting. minimal in WLN rocks, although melt crystallization With any increase in temperature above the nearly uni- owing to reaction (10r) consumed significant quantities of variant muscovite dehydration melting curve, continuous garnet, including nearly all anatectic garnet, to produce melting proceeded by the reaction retrograde biotite (Fig. 6f). The progress of this reaction was probably limited by the amount of melt and alkali biotite þ plagioclase þ sillimanite þ quartz ¼ garnet feldspar available to react, and hence P^T conditions þ alkali feldspar þ melt: under which reaction (10r) ceased are related to the bulk ð10Þ composition of the rock. Evidence for this reaction having occurred in the prograde Lizzies Basin Block sense is subtle in samples EH09 and EH10, and generally In the LBB, no evidence is found for the early phase of limited to minor low-XGrs garnet rims. higher pressure, kyanite-grade metamorphism. One inter- In sample EH21, lower Ca garnet, interpreted to have pretation of this observation is that the LBB did not ex- grown during anatexis, indicates that partial melting and perience the higher pressure conditions and hence the garnet growth in this rock must have taken place after sig- emplacement of the WLN is an important phase that nificant decompression. Although felsic bands in this mylo- acted to juxtapose rocks of distinctly different parageneses. nitic rock are composed of exclusively quartz, evidence of Alternatively, if the LBB experienced the same higher

29 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

pressure metamorphism, complete overprinting and/or dif- Therefore, the function of this retrograde reaction represents fusional resetting of the original mineral zoning would a continued component of decompression below the effective have been required to erase all record of this event. If this solidus. Reaction (11), in conjunction with Fe/Mg exchange were the case one might expect similar zoning patterns with biotite on the subsolidus retrograde path, results in for trace elements and XGrs in garnet from both the WLN steep diffusional increases in Fe/(Fe þ Mg) near LBB garnet and LBB, although garnet preserves different trace elem- rims, suggesting that low-temperature retrograde cooling ent and XGrs zoning for the two blocks. may have been relatively slow. Initial garnet growth in migmatitic pelites from the LBB probably occurred during prograde reactions involving the breakdown of hydrous phases [chlorite þ biotite; e.g. DECOMPRESSION OF THE reaction (3)]. Although no evidence is preserved for WINCHELL LAKE NAPPE garnet resorption, thermodynamic modeling results sug- Structural evidence for the early phase of decompression Downloaded from gest that it is possible that some garnet was resorbed from the inferred WLN P^T path (Fig. 13)isnotrecog- during staurolite growth. Partial melting in the sillimanite nized in the northern East Humboldt Range. Several pos- field broke down muscovite by reaction (9) to produce sig- sible explanations exist for the tectonic cause of this nificant melt (Fig. 7a) as well as abundant fibrolitic silli- decompression event: (1) erosion; (2) vertical ductile thin- manite and alkali feldspar present in melanosomes.

ning; (3) normal faulting; or potentially some combination http://petrology.oxfordjournals.org/ Biotite dehydration melting with garnet and alkali feldspar of these three (see Ring et al., 1999). Erosion as a mechan- growth then continued and additional melt was produced ism for decompression is possible, perhaps as a response to by continuous reaction (10). Some garnet shows a rim- thrusting, although Late Cretaceous to Eocene strata are ward decrease in the inclusion density (Fig. 7b), interpreted relatively sparse across the Sevier hinterland to represent the start of a new phase of garnet growth, pos- (Vandervoort & Schmitt, 1990), suggesting that erosion sibly during continuous melting reaction (10). played only a minor or supporting role in decompression. Based on (1) the relatively large estimated modal amount Vertical ductile thinning in the lower to middle crust is of melanosome or mesosome material on hand-sample another mechanism of decompression that is compatible scale (80^90%), (2) the relatively low modal amount of with significant crustal thickening by thrusting and nappe muscovite and its cross-cutting nature with respect to bio- formation in the Sevier hinterland (e.g. Ring & Kassem, at Rensselaer Polytechnic Institute on November 14, 2013 tite and sillimanite (Fig. 7d), and (3) the observation of 2007). Subhorizontal foliations in the WLN (and LBB) are interconnected web-like melt textures (Fig. 4f), melt was consistent with vertical ductile thinning as a result of de- most probably removed from the reactive bulk composition formation during high-grade metamorphism (Kassem & by some combination of isolation and extraction during de- Ring, 2004). However, decompression by vertical ductile formation near peak conditions. As discussed above, and thinning alone is not compatible solely with the phase of interpreted by previous workers (Wright & Snoke, 1993; emplacement of the WLN onto the LBB for two reasons. Batum, 1999; McGrew et al., 2000; McGrew & Snoke, First, the LBB records somewhat higher peak temperatures 2010), some proportion of 1^10m scale leucogranitic (308 higher than the WLN), which, based on inferred dikes may represent coalesced melt produced during pro- retrograde P^T paths, occurred prior to juxtaposition of grade metamorphism and effectively separated from the the two blocks, suggesting that some cooling and exhum- metapelitic layers, driving them toward a ‘restitic’ ation of the LBB predated WLN emplacement. Second, composition. the record of nearly isothermal loading followed by 3^5 Although a significant portion of the melt phase was lost kbar decompression with minor heating of the WLN is not as a result of mechanical removal, much of the remaining observed in the LBB, suggesting that these two blocks melt back-reacted with the surrounding rock to break were well separated during WLN loading and early WLN down prograde garnet and alkali feldspar (Fig. 7b, c and decompression. e) via reaction (10r). This is observed as embayed garnet Normal faulting in the middle to upper crust is a possible grains and biotite þ sillimanite pseudomorphs after garnet explanation for significant decompression. The nearby (Fig. 7f). Further retrograde garnet breakdown below the Pequop fault has been interpreted to have accommodated effective solidus probably continued via the anhydrous 10km of crustal thinning in the Late Cretaceous reaction (Camilleri & Chamberlain, 1997). This structure was rooted beneath the RM^EHR and therefore could not garnet þ muscovite ¼ biotite þ sillimanite þ quartz ð11Þ have accommodated the WLN decompression. Although which, in addition to melt crystallization reactions, helped no preserved Late Cretaceous normal fault system is produce abundant biotite and sillimanite in these rocks. observed in the present configuration of the RM^EHR, it Isopleths of garnet abundance as a result of this reaction are is perhaps possible that a localized system similar to the relatively shallow in slope (ÁP/ÁT; e.g. Spear et al., 1999). Pequop fault could have been active at shallower crustal

30 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

levels above a decompressing WLN. Syncontractional conditions in the WLN involved widespread partial melt- normal faulting is a very important exhumation mechan- ing, it is not straightforward to link the abundant leucogra- ism for the High Himalayan Crystalline complex nite dikes and sills interpreted as Late Cretaceous to (Burchfiel & Royden, 1985; Burchfiel et al., 1992; Grujic evidence of melting in metapelitic rocks. Although et al., 1996; Ring & Glodny, 2010). Although no evidence McGrew et al. (2000) argued that leucogranite intrusions exists for a major structure such as the South Tibet are associated with WLN pelites, it cannot be demon- Detachment system in the Sevier orogen, gravitational col- strated that all leucogranite magmatism was contemporan- lapse of the Sevier hinterland may have been accommo- eous with local in situ melt crystallization. There is a dated by localized extensional deformation synchronous possibility that some (or most) leucogranite intrusion and with foreland shortening. Normal faulting in the shallow melt migration from deeper crustal levels during deform- crust could have a significant effect on the exhumation his- ation may pre-date in situ melting related to decompres- tory of the middle to lower crustal section. Alternatively, sion, and acted as a ‘heat source’ to push the WLN P^T Downloaded from the formation of an extensional allochthon in the lower to path into the melt field. middle crust (e.g. Hodges & Walker, 1992) could accom- In metapelitic rocks from the LBB, partial melting modate some decompression, although this creates a sig- occurred during prograde metamorphism, directly in the nificant space problem with no evidence for extrusion to sillimanite field. The absence of relict kyanite and stauro- the surface. lite plus differences in the XGrs and trace element zoning The driving force for, and the significance of, normal in garnet porphyroblasts in migmatitic rocks from each http://petrology.oxfordjournals.org/ faulting to accommodate syncontractional exhumation in block supports this interpretation. The different P^T his- a portion of the crustal section is a topic of much debate tory for LBB rocks indicates that the emplacement of the in the literature (e.g. Hodges & Walker, 1992; McGrew WLN is an important event in the tectonic history of the et al., 2000; Wells & Hoisch, 2008; Druschke et al., 2009b; RM^EHR core complex that acted to juxtapose two Long, 2012; Miller et al., 2012; We l l s et al., 2012). A model blocks with different petrogenetic and melting histories. proposed by Wells & Hoisch (2008) and We l l s et al.(2012) The two blocks share a similar structural style and similar argues that mantle delamination occurred in the Late constraints on retrograde metamorphic re-equilibration. Cretaceous (75^67 Ma) beneath the Sevier hinterland, The LBB does not show an episode of nearly isothermal resulting in cryptic Late Cretaceous syncontractional

loading, which would have occurred had the WLN been at Rensselaer Polytechnic Institute on November 14, 2013 exhumation and leucogranite magmatism, in part as a re- emplaced prior to peak temperatures (see Spear et al., sponse to localized changes in gravitational forces in an 2002). Therefore, juxtaposition of these metamorphosed orogenic wedge. Although this model explains anatexis blocks must have occurred after, but probably near, the and a potential change in the thermal structure of the thermal peak of metamorphism, consistent with the inter- lower crust to explain heating during WLN decompres- pretations of the timing of the WLN formation of sion, it also predicts (1) significant extension in the upper McGrew et al. (2000). The results of this study imply a crust, and (2) a juvenile component to peraluminous gran- single metamorphic episode in the WLN (and Clover itoid magmatism (We l l s et al., 2012). As mentioned above, Hill) resulting in continued heating through peak pressure Late Cretaceous extensional structures are poorly pre- and then peak temperature phases of metamorphism. served. Late Cretaceous magmatism in the RM^EHR in- In situ anatexis in the WLN and the subsequent emplace- cludes no juvenile component, and ages for peraluminous ment of the WLN must have occurred after a phase of de- leucogranites span a much larger range (Howard et al., compression with heating that is recorded only by WLN 2011) than the proposed delamination event, suggesting dif- rocks. In contrast, anatexis of LBB metapelites occurred ferent or additional triggers for crustal anatexis. during a pressure increase along a P^T segment with a positive slope across dehydration melting reactions. CONCLUSIONS: THE Despite melting at similar pressures, the difference in these two P^T histories emphasizes the fact that a range SIGNIFICANCE OF PARTIAL of tectonic processes can lead to partial melting. In the MELTING exhumed High Himalayan Crystalline (HHC) complex, In the WLN, peak temperature conditions (7608C) are recent studies of anatexis and leucogranite magmatism estimated to be higher than the temperature at peak pres- have highlighted different P^T paths for anatectic rocks sure (57408C), suggesting that heating continued during exposed at similar structural levels. Different P^T histories early decompression. The source of heat during decom- are documented for melts and migmatites derived from pression is difficult to pinpoint, and may be related to ad- crustal anatexis containing (both peritectic and cotectic) vection of heat from the lower crust by leucogranite andalusite (Visona' et al., 2012), cordierite (Groppo et al., magmatism in the Late Cretaceous (Wright & Snoke, 2013), and sillimanite (Searle et al., 2010; Groppo et al., 1993; McGrew et al., 2000). Although estimated peak 2012) at similar structural levels. These petrological

31 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

observations highlight the complex polyphase melting and Caddick, M. J., Konopa¤ sek, J. & Thompson, A. B. (2010). Preservation structural evolution of the HHC, and other orogens such of garnet growth zoning and the duration of prograde metamorph- as the Sevier that may have experienced syncontractional ism. Journal of Petrology 51, 2327^2347. Camilleri, P. A. & Chamberlain, K. R. (1997). Mesozoic tectonics normal faulting and exhumation. Recognizing different and metamorphism in the Pequop Mountains and Wood Hills petrogenetic histories for discrete blocks within an orogen region, northeast Nevada: Implications for the architecture and provides the keys for unraveling the tectonic significance evolution of the Sevier orogen. Geological Society of America Bulletin of deep crustal melting and metamorphism in an exhumed 109, 74^94. orogenic belt. Connolly, J. & Cesare, B. (1993). C^O^H^S fluid composition and oxygen fugacity in graphitic metapelites. Journal of Metamorphic Geology 11, 379^388. ACKNOWLEDGEMENTS Dallmeyer,R.D.,Snoke,A.W.&McKee,E.H.(1986).The Mesozoic^Cenozoic tectonothermal evolution of the Ruby We thank Jonathan Price and Daniel Ruscitto for assist- Mountains, East Humboldt Range, Nevada: A Cordilleran meta- Downloaded from ance in the laboratory and with the electron microprobe. morphic core complex.Tectonics 5,931^954. Also, thanks go to Allen McGrew for engaging discussion Davis, G. H. & Coney, P. J. (1979). Geologic development of the about rocks of the East Humboldt Range. This contribu- Cordilleran metamorphic core complexes. Geology 7, 120^124. tion was vastly improved by careful reviews by U. Ring, de Capitani, C. & Brown, T. H. (1987). The computation of chemical G. Droop, and A. Snoke. equilibrium in complex systems containing non-ideal solutions.

Geochimica et Cosmochimica Acta 51, 2639^2652. http://petrology.oxfordjournals.org/ de Capitani, C. & Petrakakis, K. (2010). The computation of equilibrium assemblage diagrams with Theriak/Domino software. FUNDING American Mineralogist 95, 1006^1016. This research was supported by the Department of Earth Dokka, R. K., Mahaffie, M. J. & Snoke, A. W. (1986). and Environmental Sciences at Rensselaer Polytechnic Thermochronologic evidence of major tectonic denudation asso- Institute, grants from the National Science Foundation ciated with detachment faulting, northern Ruby Mountains-East (EAR-1019768 EAR-0337413, and EAR-0409622) to F. Humboldt Range, Nevada.Tectonics 5, 995^1006. Spear, the Edward P. Hamilton Distinguished Chair for Druschke, P., Hanson, A. D., Wells, M. L., Rasbury, T., Stockli, D. F. & Gehrels, G. (2009a). Synconvergent surface-breaking normal Science Education, and a graduate student research grant faults of Late Cretaceous age within the Sevier hinterland, east^ from the Geological Society of America. central Nevada. Geology 37, 447^450. at Rensselaer Polytechnic Institute on November 14, 2013 Druschke, P., Hanson, A. D. & Wells, M. L. (2009b). Structural, stratigraphic, and geochronologic evidence for extension predating SUPPLEMENTARY DATA Palaeogene volcanism in the Sevier hinterland, east^central Supplementary data for this paper are available at Journal Nevada. International Geology Review 51, 743^775. Foster, D. A., Doughty, P. T., Kalakay, T. J., Fanning, C. M., of Petrology online (http://www.petrology.oxfordjournals. Coyner, S., Grice, W. C. & Vogl, J. (2007). Kinematics and timing org). of exhumation of metamorphic core complexes along the Lewis and Clark fault zone, northern Rocky Mountains, USA. In: Till,A.B.,Roeske,S.M.,Sample,J.C.&Foster,D.A.(eds) REFERENCES Exhumation Associated with Continental Strike-Slip Fault Systems. Armstrong, R. L. (1982). Cordilleran metamorphic core complexesç Geological Society of America, Special Papers 434,207^232. from Arizona to southern Canada. Annual Review of Earth and Groppo, C., Rolfo, F. & Indares, A. (2012). Partial melting in the Planetary Sciences 10, 129^154. Higher Himalayan Crystallines of Eastern Nepal: the effect of de- Arzi, A. A. (1978). Critical phenomena in the rheology of partially compression and implications for the ‘channel flow’ model. Journal melted rocks.Tectonophysics 44, 173^184. of Petrology 53, 1057^1088. Batum, M. A. (1999). Petrology of Late Cretaceous and Cenozoic Groppo, C., Rolfo, F. & Mosca, P. (2013).The cordierite-bearing ana- granitic rocks. East Humboldt Range. MS thesis, Texas Tech, tectic rocks of the higher Himalayan crystallines (eastern Nepal): Lubbock, 167 pp. low-pressure anatexis, melt productivity, melt loss and the preser- Bohlen, S. R., Wall, V. J. & Boettcher, A. L. (1983). Experimental in- vation of cordierite. Journal of Metamorphic Geology 31,187^204. vestigations and geological applications of equilibria in the system Grujic, D., Casey, M., Davidson, C., Hollister, L. S., Ku« ndig, R., Pavlis, T. & Schmid, S. (1996). Ductile extrusion of the Higher Fe O ^ Ti O 2^Al2O3^SiO2^H2O. American Mineralogist 68,1049^1058. Brown, M. (1994). The generation, segregation, ascent and emplace- Himalayan Crystalline in Bhutan: evidence from quartz microfab- ment of granite magma: the migmatite-to-crustally-derived granite rics.Tectonophysics 260, 21^43. connection in thickened orogens. Earth-Science Reviews 36,83^130. Harris, N. & Massey, J. (1994). Decompression and anatexis of Burchfiel, B. C. & Royden, L. H. (1985). North^south extension within Himalayan metapelites.Tectonics 13, 1537^1546. the convergent Himalayan region. Geology 13, 679^682. Harris, N. B. W., Caddick, M., Kosler, J., Goswami, S., Vance, D. & Burchfiel, B. C., Chen Zhiliang, Liu Yuping, Royden, L. H., Deng Tindle, A. G. (2004).The pressure^temperature^time path of mig- Changrong, & Xu Jiene (eds) (1992). The SouthTibetan detachment matites from the Sikkim Himalaya. Journal of Metamorphic Geology systemHimalayan orogen: Extension contemporaneous with and parallel to 22, 249^264. shortening in a collisional mountain belt. Geological Society of Henry, D. J., Guidotti, C. V. & Thomson, J. A. (2005). The Ti-satur- America, Special Papers 269, pp. 1^41. ation surface for low-to-medium pressure metapelitic biotites:

32 HALLETT & SPEAR P^T HISTORY, EAST HUMBOLDT RANGE

implications for geothermometry and Ti-substitution mechanisms. Lee, S., Barnes, C. G., Snoke, A.W., Howard, K. A. & Frost, C. D.(2003). American Mineralogist 90, 316^328. Petrogenesis of Mesozoic, peraluminous granites in the Lamoille Hickmott, D. D., Shimizu, N., Spear, F. S. & Selverstone, J. (1987). Canyon area, Ruby Mountains, Nevada, USA. Journal of Petrology 44, Trace-element zoning in a metamorphic garnet. Geology 15, 713^732. 573^576. Long, S. P. (2012). Magnitudes and spatial patterns of erosional exhum- Hodges, K.V. & Crowley, P. D. (1985). Error estimation and empirical ation in the Sevier hinterland, eastern Nevada and western Utah, geothermobarometry for pelitic systems. American Mineralogist 70, USA: Insights from a Paleogene paleogeologic map. Geosphere 8, 702^709. 881^901. Hodges, K. V. & Spear, F. S. (1982). Geothermometry, geobarometry Lush, A. P., McGrew, A. J., Snoke, A. W. & Wright, J. E. (1988).

and the Al2SiO5 triple point at Mt. Moosilauke, New Hampshire. Allochthonous Archean basement in the northern East Humboldt American Mineralogist 67, 1118 ^ 1134. Range, Nevada. Geology 16, 349^353. Hodges, K. V. & Walker, J. D. (1992). Extension in the Cretaceous MacCready,T.,Snoke,A.W.,Wright,J.E.&Howard,K.A.(1997).Mid- Sevier orogen, North American Cordillera. Geological Society of crustal flow during Tertiary extension in the Ruby Mountains core America Bulletin 104, 560^569. complex, Nevada. GeologicalSocietyof AmericaBulletin109,1576^1594. Downloaded from Hodges, K. V., Snoke, A. W. & Hurlow, H. A. (1992). Thermal evolu- McGrew, A. J. (1992). Tectonic evolution of the northern East tion of a portion of the Sevier hinterland: the northern Ruby Humboldt Range, Elko County, Nevada. PhD thesis, University of Mountains^East Humboldt Range and Wood Hills, northeastern Wyoming, Laramie, 191 p. Nevada. Tectonics 11, 154^164. McGrew, A. J. & Snee, L. W. (1994). 40Ar/39Ar thermochronologic Hoisch, T. D., Wells, M. L. & Hanson, L. M. (2002). Pressure^tem- constraints on the tectonothermal evolution of the Northern East perature paths from garnet-zoning: Evidence for multiple episodes Humboldt range metamorphic core complex, Nevada. http://petrology.oxfordjournals.org/ of thrust burial in the hinterland of the Sevier orogenic belt. Tectonophysics 238, 425^450. American Mineralogist 87,115^131. McGrew,A.J.&Snoke,A.W.(2010).SHRIMP-RGU^Pbisotopic Holland, T. & Powell, R. (1998). An internally consistent thermo- systematics of zircon from the Angel Lake Orthogneiss, East dynamic data set for phases of petrological interest. Journal of Humboldt Range, Nevada; is this really Archean crust?: Metamorphic Geology 16, 309^344. Comment. Geosphere 6, 962^965. Hollister, L. S. (1969). Contact metamorphism in the Kwoiek area of McGrew, A. J., Peters, M. T. & Wright, J. E. (2000). British Columbia: an end member of the metamorphic process. Thermobarometric constraints on the tectonothermal evolution of Geological Society of America Bulletin 80, 2465^2494. the East Humboldt Range metamorphic core complex, Nevada. Hollister, L. S. & Crawford, M. L. (1986). Melt-enhanced deform- Geological Society of America Bulletin 112, 45^60. ation: a major tectonic process. Geology 14, 558^561. Miller, E. L., Konstantinou, A. & Strickland, A. (2012). Comment on Howard, K. A. (1980). Metamorphic infrastructure in the northern ‘Geodynamics of synconvergent extension and tectonic mode Ruby Mountains, Nevada. In: Crittenden, M. D., Coney, P. J. & switching: Constraints from the Sevier^Laramide orogen’ by at Rensselaer Polytechnic Institute on November 14, 2013 Davis, G. H. (eds) Cordilleran Metamorphic Core Complexes. Michael L. Wells et al.Tectonics 31,TC4015. Geological Society of America, Memoirs 153, pp. 335^347. Ohmoto, H. & Kerrick, D. (1977). Devolatilization equilibria in graph- Howard, K. A., Wooden, J. L., Barnes, C. G., Premo, W. R., Snoke, A. itic systems. AmericanJournal of Science 277, 1013^1044. W. & Lee, S. Y. (2011). Episodic growth of a Late Cretaceous and Paquet, J., Francois, P. & Nedelec, A. (1981). Effect of partial melting Paleogene intrusive complex of pegmatitic leucogranite, Ruby on rock deformation: experimental and natural evidences on rocks Mountains core complex, Nevada, USA. Geosphere 7, 1220^1248. of granitic compositions.Tectonophysics 78, 545^565. Hudec, M. R. (1990).The structural and thermal evolution of the cen- Patin‹ o Douce, A. E. & Harris, N. (1998). Experimental constraints on tral Ruby Mountains, Elko County, Nevada. PhD thesis, Himalayan anatexis. Journal of Petrology 39, 689^710. University of Wyoming, Laramie, 272 p. Peters, M. T. & Wickham, S. M. (1994). Petrology of upper amphibo- Hudec, M. R. (1992). Mesozoic structural and metamorphic history of lite facies marbles from the East Humboldt Range, Nevada, USA; the central Ruby Mountains metamorphic core complex, Nevada. evidence for high-temperature, retrograde, hydrous volatile fluxes Geological Society of America Bulletin 104, 1086^1100. at mid-crustal levels. Journal of Petrology 35, 205^238. Hurlow, H. A., Snoke, A. W. & Hodges, K. V. (1991).Temperature and Peters, M. T. & Wickham, S. M. (1995). On the causes of 18O-deple- pressure of mylonitization in a Tertiary extensional shear zone, tion and 18O/16O homogenization during regional metamorphism; Ruby Mountains^East Humboldt Range, Nevada: Tectonic impli- the East Humboldt Range core complex, Nevada. Contributions to cations. Geology 19, 82^86. Mineralogy and Petrology 119,68^82. Indares, A.,White, R.W. & Powell, R. (2008). Phase equilibria model- Powell, R. & Holland, T. J. B. (1988). An internally consistent dataset ling of kyanite-bearing anatectic paragneisses from the central with uncertainties and correlations: 3. Applications to geobarome- Grenville Province. Journal of Metamorphic Geology 26, 815^836. try, worked examples and a computer program. Journal of Jones, J. V. (1999a). Deformational, magmatic and metamorphic his- Metamorphic Geology 6, 173^204. tory of the central Ruby Mountains, Elko county, Nevada. MS Premo, W. R., Castin‹ eiras, P. & Wooden, J. L. (2008). SHRIMP-RG thesis, University of Wyoming, Laramie, 166 p. U^Pb isotopic systematics of zircon from the Angel Lake orthog- Jones, J.V. (1999b). Jurassic and Cretaceous magmatism, deformation, neiss, East Humboldt Range, Nevada: Is this really Archean crust? and metamorphism in the central Ruby Mountains, NE Nevada. Geosphere 4, 963^975. Geological Society of America, Abstracts with Programs 31,116. Premo, W. R., Castin‹ eiras,P.&Wooden,J.L.(2010).SHRIMP-RG Kassem, O. & Ring, U. (2004). Underplating-related finite-strain pat- U^Pb isotopic systematics of zircon from the Angel Lake terns in the Gran Paradiso massif, Western Alps, Italy: heteroge- Orthogneiss, East Humboldt Range, Nevada; is this really neous ductile strain superimposed on a nappe stack. Journal of the Archean crust?; Reply. Geosphere 6, 966^972. Geological Society, London 161, 875^884. Pyle, J. M. & Spear, F. S. (1999).Yttrium zoning in garnet: coupling of Kretz, R. (1983). Symbols for rock-forming minerals. American major and accessory phases during metamorphic reactions. Mineralogist 68, 277^279. Geological Materials Research 1,1^49.

33 JOURNAL OF PETROLOGY VOLUME 0 NUMBER 0 MONTH 2013

Ring, U. & Glodny, J. (2010). No need for lithospheric extension for Northeast Section Meeting of the Geological Society of America Short Course exhuming (U)HP rocks by normal faulting. Journal of the Geological 504. Society, London 167, 225^228. Spear, F. S., Kohn, M. J., Cheney, J. T. & Florence, F. (2002). Ring, U. & Kassem, O. M. (2007).The nappe rule: why does it work? Metamorphic, thermal, and tectonic evolution of central New Journal of the Geological Society, London 164, 110 9 ^ 1112. England. Journal of Petrology 43, 2097^2120. Ring, U., Brandon, M. T., Willett, S. D. & Lister, G. S. (1999). Sullivan, W. A. & Snoke, A. W. (2007). Comparative anatomy of core- Exhumation processes. In: Ring, U., Brandon, M. T., Willett, S. D. complex development in the northeastern Great Basin, U.S.A. & Lister, G. S. (eds) Exhumation Processes: Normal Faulting, Ductile Rocky Mountain Geology 42,1^29. Flow and Erosion. Geological Society, London, Special Publications 154, Tomkins, H., Powell, R. & Ellis, D. (2007). The pressure dependence 1^27. of the zirconium-in-rutile thermometer. Journal of Metamorphic Rosenberg, C. & Handy, M. (2005). Experimental deformation of par- Geology 25, 703^713. tially melted granite revisited: implications for the continental Tracy, R. J., Robinson, P. & Thompson, A. B. (1976). Garnet compos- crust. Journal of Metamorphic Geology 23, 19^28. ition and zoning in the determination of temperature and pressure of metamorphism, central Massachusetts. American Mineralogist 61, Rushmer, T. (1996). Melt segregation in the lower crust: how have ex- Downloaded from periments helped us? In: Brown, M., Candela, P. A., Peck, D. L., 762^775. Stephens, W. E., Walker, R. J. & Zen, E. (eds) The Third Hutton Vanderhaeghe, O. (2009). Migmatites, granites and orogeny: Flow Symposium on the Origin of Granites and Related Rocks. Geological Society modes of partially-molten rocks and magmas associated with of America, Special Papers 315,73^83. melt/solid segregation in orogenic belts.Tectonophysics 477,119^134. Searle, M. P., Cottle, J. M., Streule, M. J. & Waters, D. J. (2010). Vandervoort, D. S. & Schmitt, J. G. (1990). Cretaceous to early Tertiary paleogeography in the hinterland of the Sevier thrust Crustal melt granites and migmatites along the Himalaya: melt http://petrology.oxfordjournals.org/ source, segregation, transport and granite emplacement mechan- belt, east^central Nevada. Geology 18, 567^570. isms. Earth and Environmental ScienceTransactions of the Royal Society of Visona' , D., Carosi, R., Montomoli, C., Tiepolo, M. & Peruzzo, L. (2012). Miocene andalusite leucogranite in centralêast Himalaya Edinburgh 100, 219^233. (Everest^Masang Kang area): low-pressure melting during heat- Sicard, K. R., Snoke, A. W. & Swapp, S. M. (2011).The metamorphic ing. Lithos 144^145, 194^208. and structural history of Clover Hill, Nevada, part of the Ruby^ Watson, E., Wark, D. & Thomas, J. (2006). Crystallization thermom- East Humboldt Core Complex. Geological Society of America, eters for zircon and rutile. Contributions to Mineralogy and Petrology Abstracts with Programs 43,15. 151, 413^433. Snoke, A. W. (1992). Clover Hill, Nevada: Structural link between the Wells, M. L. & Hoisch, T. D. (2008).The role of mantle delamination Wood Hills and East Humboldt Range. In: Wilson, J. R. (ed.) in widespread Late Cretaceous extension and magmatism in the Field Guide to Geologic Excursions in Utah and Adjacent Areas of Nevada, Cordilleran orogen, western United States. Geological Society of Idaho and Wyoming. Geological Society of America, pp. 107^122. America Bulletin 120, 515^530. at Rensselaer Polytechnic Institute on November 14, 2013 Snoke, A. W. & Miller, D. M. (1988). Metamorphic and tectonic his- Wells, M. L., Hoisch, T. D., Cruz-Uribe, A. M. & Vervoort, J. D. tory of the northeastern Great Basin. In: Ernst, W. G. (ed.) (2012). Geodynamics of synconvergent extension and tectonic Metamorphism and Crustal Evolution of the Western United States: Rubey mode switching: Constraints from the Sevier^Laramide orogen. Vo l u m e 7. Prentice^Hall, pp. 607^648. Tectonics 31,TC1002. Snoke, A. W., McKee, E. H. & Stern, T. W. (1979). Plutonic, meta- White, R.W., Powell, R. & Holland,T. J. B. (2001). Calculation of par- morphic, and structural chronology in the northern Ruby tial melting equilibria in the system Na2O^CaO^K2O^FeO^ Mountains, Nevada; a preliminary report. Geological Society of MgOÂl2O3^SiO2^H2O (NCKFMASH). Journal of Metamorphic America, Abstracts with Programs 11, 520^521. Geology 19, 139^153. Spear, F. S. (1995). Metamorphic Phase Equilibria and Pressure^ Wickham, S. M. & Peters, M. T. (1990). An oxygen isotope discontinu- Temperature^Time Paths. Mineralogical Society of America Monograph 1, ity in high-grade rocks of the East Humboldt Range, Nevada. 799 p. Nature 345,150^153. Spear, F. S. & Florence, F. P. (1992). Thermobarometry in granulites: Wickham, S. M. & Peters, M. T. (1992). Oxygen and carbon isotope pitfalls and new approaches. Precambrian Research 55, 209^241. profiles in metasediments from Lizzies Basin, East Humboldt Spear, F. S. & Menard, T. (1989). Program GIBBS; a generalized Range, Nevada: constraints on mid-crustal metamorphic and mag- Gibbs method algorithm. American Mineralogist 74,942^943. matic volatile fluxes. Contributions to Mineralogy and Petrology 112, 46^65. Spear, F. S. & Selverstone, J. (1983). Quantitative PT paths from Wright, J. E. & Snoke, A. W. (1993).Tertiary magmatism and myloniti- zoned minerals: theory and tectonic applications. Contributions to zation in the RubyÊast Humboldt metamorphic core complex, Mineralogy and Petrology 83, 348^357. northeastern Nevada: U^Pb geochronology and Sr, Nd, and Pb iso- Spear, F. S., Kohn, M. J. & Cheney, J. T. (1999). P^T paths from ana- tope geochemistry. Geological Society of America Bulletin 105, 935^952. tectic pelites. Contributions to Mineralogy and Petrology 134, 17^32. Zack, T., Moraes, R. & Kronz, A. (2004).Temperature dependence of Spear,F.S.,Pyle,J.M.&Storm,L.C.(2001).Thermodynamic Zr in rutile: empirical calibration of a rutile thermometer. modeling of mineral reactions: An introduction to Program Gibbs. Contributions to Mineralogy and Petrology 148, 471^488.