Abstract

New in situ U-Th-Pb monazite ages from metapelites within the upper Greater Himalayan sequence in the Manaslu–Himal Chuli Himalaya provide insight into the tectonometamorphic history of this portion of the mountain range. Monazite grains record a protracted evolution characterized by a three-stage growth history at lower structural levels (ca. 30 Ma, ca. 26–21 Ma, and ca. 18–15 Ma) and a single growth stage between ca. 21 and 16 Ma, following complete resorption of earlier monazite at higher structural levels. All specimens record monazite growth during protracted early to middle Miocene melt crystallization, with monazite at higher structural levels significantly overlapping in age with the young phase of the Manaslu pluton. The metamorphic and anatectic ages reported in this study are consistent with a middle Miocene transition from deep hinterland-style deformation characterized by extending flow in the upper Greater Himalayan sequence to shallower foreland-style deformation characterized by compressing flow in the lower Greater Himalayan sequence. This change outlines a deformational continuum between lateral midcrustal flow beneath Tibet and thrust wedge taper deformation along the mountain front as has been put forth in recent thermo-mechanical and conceptual models.

INTRODUCTION

The conceptual model of large-scale lateral, ductile crustal flow beneath Tibet was first developed by Bird (1991) in an examination of the isostatic limitations of supporting high topography. The concept of lateral flow was revisited in subsequent geophysical studies, which outlined a structural geometry that connected coeval and opposite shear-sense structures, the Main Central thrust and the South Tibetan detachment system (Fig. 1), in the Himalayan front to prominent seismic reflectors that dip northward below the plateau toward an anomalously electrically conductive, low-seismic-velocity zone interpreted to be partially molten middle crust (Nelson et al., 1996). The interpreted geophysical framework is consistent with field-based studies, which conclude that the exhumed midcrustal core of the orogen was extruded laterally between the Main Central thrust and the South Tibetan detachment system (e.g., Grujic et al., 1996). The southward extrusion of the metamorphic core of the orogen between the two faults has been numerically modeled in thermo-mechanical simulations within the constraints of published geologic data (e.g., Beaumont et al., 2001) and is now commonly known as the channel-flow hypothesis. Ideas about the midcrustal evolution of the Himalaya and the models that reflect them have evolved as more data have been gathered and published from across the orogen (e.g., Beaumont et al., 2004, 2006).

Models of lateral midcrustal flow, however, are not the only hypotheses put forward that account for the features observed in the Himalaya. There are researchers who favor a thrust wedge model (e.g., DeCelles et al., 1998; Robinson et al., 2001; Robinson, 2008; Kohn, 2008; McQuarrie et al., 2008; Sachan et al., 2010) in which the structures and pressure-temperature-time (P-T-t) paths observed in the Himalaya are the result of critical taper wedge failure (e.g., Dahlen, 1990) characterized by thrust fault development and the juxtaposition of discrete thrust sheets. As such, there has been much recent debate about the validity of lateral midcrustal flow versus thrust wedge taper models within the evolution of the Himalayan system.

The separation of Himalayan orogenic models into two distinct end members, channel flow or thrust wedge, which has occurred of late within Himalayan literature, may be a false dichotomy (Beaumont and Jamieson, 2010). Studies that have examined different portions of the metamorphic core of the Himalaya have led to differing conclusions about the deformation processes that affected those rocks. Examination of P-T-t paths recorded in the lower and middle portions of the metamorphic core in the Langtang region of west-central Nepal have been interpreted to be consistent with thrust wedge taper processes (Kohn, 2008), while studies of the timing of metamorphism, melting, and deformation in the upper portion of the metamorphic core near the Everest region are interpreted to be consistent with channel flow (Cottle et al., 2009a). As Beaumont and Jamieson (2010) noted, however, channel-flow and thrust wedge taper models are not mutually exclusive. Rather, the dominant processes occurring in any one place within an orogen are expected to change through time as the orogen evolves and variables such as erosion and crustal temperature and thickness change. Examination of thermo-mechanical models of the Himalayan orogen commonly cited as “channel-flow models” reveals that while ductile midcrustal flow occurs at depth within the middle portion of the orogen, cooler material in front of the active channel develops through processes approximating a thrust wedge (Beaumont et al., 2001; Jamieson et al., 2004, 2006; Beaumont and Jamieson, 2010). Furthermore, recent research in central Nepal has outlined a transition within the exhumed metamorphic core of the orogen between rocks that record deformation characteristic of lateral midcrustal flow and rocks that record thrust wedge–style deformation (Larson et al., 2010a). The transition between the two likely evolved spatially and temporally, dependent primarily on changes in rheology. Therefore, a rock that was initially part of a midcrustal channel could have moved across the transition and become part of a thrust wedge (Larson et al., 2010a). These observations support a deformational continuum between lateral ductile flow in the deep hinterland and thrust wedge–style deformation in the shallow foreland.

In the present study area, the U-Th-Pb characteristics of the lower portion of the Himalayan metamorphic core are well constrained (e.g., Kohn et al., 2001; Catlos et al., 2001) and are interpreted to be consistent with thrust wedge processes (Kohn, 2008; Larson et al., 2010a). In contrast, the only U-Th-Pb geochronologic data from the upper portion of the metamorphic core in the study area, which has been interpreted to be part of an exhumed midcrustal ductile channel (Larson et al., 2010a), comes from monazite ages from the Manaslu leucogranite. These ages indicate an episodic emplacement of the Manaslu pluton, with two dominant phases crystallizing at 23 Ma and 19 Ma, respectively (Harrison et al., 1999; Fig. 1B), but they provide no direct constraints on the timing of metamorphism in the midcrustal rocks or crystallization of migmatite. This study presents new monazite U-Th-Pb dates from across the sillimanite-grade, migmatitic rocks that comprise the upper Himalayan metamorphic core in order to better constrain its tectonometamorphic evolution. These new dates are used to outline the geologic history of the Manaslu–Himal Chuli Himalaya and assess the current models, if any, that are consistent with the results.

GEOLOGIC SETTING

In this study, we follow Larson et al. (2010a) and Searle et al. (2008) in including the entire inverted metamorphic package that characterizes the lower part of the exhumed metamorphic core as part of the Greater Himalayan sequence in the hanging wall of the Main Central thrust of the Manaslu–Himal Chuli region of west-central Nepal. The Greater Himalayan sequence in the study area is composed of an ∼35-km-thick panel of metasedimentary and meta-igneous rocks that has been deformed and metamorphosed at midcrustal levels (Fig. 1). It consists of a structurally lower portion dominated by locally calcareous phyllitic schist intercalated with quartzite, marble, and augen orthogneiss and an upper portion characterized by intercalated quartzite and pelitic schist, sillimanite- and anatectite-bearing paragneiss, augen orthogneiss, leucogranite, and calc-silicate schist (Colchen et al., 1986; Larson et al., 2010a). The entire Greater Himalayan sequence cooled through mica and hornblende argon retention temperatures during the Cenozoic (Hodges et al., 1988; Copeland et al., 1990; Larson et al., 2010a) while rocks in the footwall of the Main Central thrust yield Paleozoic to Proterozoic 40Ar/39Ar ages (Wobus et al., 2003; Larson et al., 2010a).

Peak metamorphic pressure-temperature conditions recorded in the lower portion of the Greater Himalayan sequence increase from 542 ± 25 °C and 0.70 ± 0.12 GPa near its base to 624 ± 25 °C and 1.14 ± 0.12 GPa near the middle of the Greater Himalayan sequence (Kohn et al., 2001; Larson et al., 2010a). Monazite U-Th-Pb geochronology of these rocks shows an increase in age with structural distance above the Main Central thrust fault (ca. 8 Ma near the fault to ca. 21 Ma near the middle of the Greater Himalayan sequence) such that younger rocks record lower pressure and temperature conditions at structurally lower positions (Kohn et al., 2001). This relationship has been interpreted to reflect the coeval downward migration of shear along the Main Central thrust and exhumation of the Greater Himalayan sequence (Larson et al., 2010a). The downward migration of the Main Central thrust and the associated progressive subcretion of material to the base of the Greater Himalayan sequence are consistent with vertical thickening and horizontal shortening, or compressing flow deformation—compression in the direction of flow—features characteristic of thrust wedge deformation. In contrast, the sillimanite-grade, migmatitic upper portion of the Greater Himalayan sequence records metamorphic isograds that define a conspicuously steep upward-decreasing metamorphic pressure field gradient of ∼0.062 GPa per kilometer (∼62 MPa/km; Larson et al., 2010a). This apparent pressure gradient is consistent with a 50% postpeak-metamorphic vertical thinning of the upper Greater Himalayan sequence (Larson et al., 2010a). Alternatively, a similar apparent gradient could be achieved through differential translation during lateral ductile flow and the juxtaposition of material that was originally far-separated. Both interpretations are consistent with gravity-driven, midcrustal lateral flow–type models and extending flow deformation (extension in the direction of flow) (Price, 1972).

U-Th-Pb MONAZITE GEOCHRONOLOGY

Monazite grains from five anatectite-bearing sillimanite paragneiss specimens were dated by in situ U-Th-Pb laser ablation–multicollector–inductively coupled plasma–mass spectrometry (LA-MC-ICP-MS) in order to provide temporal constraints on the evolution of the upper Greater Himalayan sequence. The specimens were selected because they represent a traverse through the dominant lithology of the upper Greater Himalayan sequence in the study area.

Metamorphic Conditions

The rocks collected for geochronologic analyses all share a similar metamorphic assemblage of quartz + muscovite + biotite + plagioclase + garnet + alumino-silicate. Each specimen records metamorphic temperatures in excess of 590 °C (Larson et al., 2010a; Table 1; Fig. 1B). Metamorphic pressure estimates vary across the sampled region between 0.84 GPa at the lowest structural level and 0.39 GPa at the highest structural level (Larson et al., 2010a; Table 1; Fig. 1B). Garnet grains in the specimens are characterized by relatively flat compositional profiles, which may reflect thermal reequilibration after initial growth. This is consistent with the multiphase metamorphic history commonly recorded across the orogen in which initial prograde high-P, moderate-T assemblages are succeeded by later high-T, moderate-P conditions (e.g., Godin et al., 2001; Catlos et al., 2002; Cottle et al., 2009a; Larson et al., 2010b; Streule et al., 2010).

Methods

Monazite grains in polished thin sections were identified using backscattered electron (BSE) imaging to find potential grains, which were then confirmed using energy dispersive X-ray spectroscopy analyses.

Between 6 and 12 monazite grains from each specimen were compositionally mapped using Cameca SX-100 electron probe microanalyzers at the University of California, Santa Barbara, and at the Saskatchewan Research Council Advanced Microanalysis Centre. Th, Nd, Y, and U elemental maps for each grain were constructed using an accelerating voltage of ∼15 kV, a beam current of ∼200 nA, step size of ∼0.5 mm, and rastering of the electron beam to produce the highest possible quality images. Furthermore, spot analyses of yttrium concentrations were also performed on the monazite to allow dates to be linked to yttrium domains. Low-magnification BSE images and compositional maps were obtained to ascertain the spatial and/or textural context of monazite within the specimen.

Grains were dated using a LA-MC-ICP-MS system housed at the University of California, Santa Barbara (UCSB). Instrumentation consisted of a Nu Plasma MC-ICP-MS (Nu Instruments, Wrexham, UK) and a 193 nm ArF laser-ablation system equipped with a two-volume “HelEx” ablation cell that facilitates rapid transfer and washout of ablated material (Photon Machines, San Diego, USA). Analytical protocol is similar to that described by Cottle et al. (2009a, 2009b, 2011) with the modification that the collector arrangement on the Nu Plasma at UCSB allows for simultaneous determination of 232Th and 238U on high-mass side Faraday cups equipped with 1011 ohm resistors and 208Pb, 207Pb, 206Pb, and 204Pb on four low-mass side ETP discrete dynode secondary electron multipliers.

U-Th-Pb analyses were conducted for 30 s each using spot diameters ranging from 7 to 20 mm in diameter, a frequency of 3 Hz, and 0.75 J/cm2 fluence (equating to crater depths of ∼7–8 mm). U-Th-Pb data from five samples were collected over four analytical sessions. A primary reference material, “Managotry” monazite (554 Ma Pb/U isotope dilution–thermal ionization mass spectrometry [ID-TIMS] age; Paquette et al., 1994), was employed to monitor and correct for mass bias, as well as Pb/U and Pb/Th down-hole fractionation. To monitor data accuracy, two secondary reference monazites “FC-1” (55.7 Ma Pb/U ID-TIMS age; Horstwood et al., 2003) and “44096” (424 Pb/U Ma ID-TIMS age; Aleinikoff et al., 2006) were analyzed concurrently (once every 3–5 unknowns) and mass bias– and fractionation-corrected based on measured isotopic ratios of the primary reference material. During the analytical period, repeat analyses of FC-1 gave a weighted mean 206Pb/238U age of 54.9 ± 0.4 Ma, mean square of weighted deviates (MSWD) = 0.9, and a weighted mean 208Pb/232Th age of 57.2 ± 0.5 Ma, MSWD = 1.4 (2σ). Data reduction was performed with in-house software at UCSB. All uncertainties are quoted at 2σ and include contributions from the external reproducibility of the primary reference material for the 207Pb/206Pb, 206Pb/238U, and 208Pb/232Th ratios. The full data set is presented in Table 2.

Results

Cenozoic monazites commonly display reverse discordance such that apparent 206Pb/238U ages are older than measured 207Pb/235U and 208Pb/232Th ages (in the case of these samples, the difference is as much as 3 m.y.). This discrepancy is inferred to be a result of initial disequilibrium in the 238U → 206Pb decay scheme due to incorporation of excess 230Th during monazite crystallization. This in turn leads to an “excess” of 206Pb (Schärer, 1984). Because the 232Th → 208Pb decay system is unaffected by this issue, we use the dates derived from this ratio as the best estimate of the monazite crystallization age.

All monazite grains analyzed in this study occur in the matrix and are commonly associated with biotite (Fig. 2). Only in one specimen, MS-31, are monazite inclusions observed in garnet, and then only rarely (Fig. 2). Furthermore, all monazite grains are compositionally zoned with respect to yttrium and thorium (Fig. 3), and the rocks in which these monazite grains occur contain up to 40% anatectite by volume (Fig. 3E of Larson et al., 2010a). The specimens analyzed will be discussed next in order of structural position from low to high. The relative positions of the specimens are shown in Figure 1. See Table 2 for analytical data.

MS-62

In MS-62, the structurally lowest specimen collected (Fig. 1), 66 analyses were performed on 9 monazite grains. The grains are generally symmetric and range in diameter from ∼30 mm to 100 mm, with modal abundances tending toward the higher end of that range. The results of these U-Th-Pb analyses are presented in Figure 4 as 208Pb/232Th–206Pb/238U concordia and 208Pb/232Th age probability diagrams, which highlight three dominant age populations, a peak at ca. 30 Ma that can be correlated to a moderate yttrium core (∼7000–9000 ppm), an intermediate peak at ca. 23 Ma that that is characteristic of an yttrium-poor outer core (∼3000 ppm), and a prominent peak at ca. 16 Ma reflecting spots analyzed in the yttrium-rich rim (∼15–25,000 ppm; Fig. 3A).

MS-31

In total, 36 analyses on six monazite crystals were performed on specimen MS-31, which was collected 4 km structurally higher than MS-62. Grains in this specimen are typically elongate (Fig. 3B) with a long axis of ∼40–70 mm and a short axis of ∼10–30 mm. The probability age distribution for monazite crystallization in this specimen is characterized by a dominant peak at ca. 16.5 Ma and a secondary peak at ca. 13–12 Ma (Fig. 4). Most of the monazite grains in MS-31 are characterized by moderate yttrium cores (∼6000–9000 ppm), which yielded ages that fall within the prominent age peak of ca. 16.5 Ma. Analysis of two yttrium-rich rim (∼14,000 ppm) domains gave ages of 12–13 Ma (Fig. 3B).

MS-34, MS-47, and MS-37

Specimens MS-34, MS-47, and MS-37 are the structurally highest specimens analyzed (Fig. 1). The monazite grains contained in these specimens have very similar characteristics and as such will be discussed together. The monazite grains in all three specimens are typically equant and £30 mm in diameter. From these specimens, 36 analyses from 11 grains in MS-34 (Fig. 3C) define a continuous spectrum of dates between 18.0 and 15.8 Ma (Fig. 4), 23 analyses from eight grains in MS-47 (Fig. 3D) define a single peak distribution between 19.6 and 17.5 Ma, and 106 analyses from 12 grains in MS-37 (Fig. 3E) define a distribution between 21.5 and 17 Ma, with a subpeak of ages at 20.5 and 19.5 (Fig. 4). Unlike the structurally lower specimens, yttrium zonation in these specimens does not appear to correlate strongly with age. The ages yielded from relatively yttrium-poor monazite cores (∼6000–12,000 ppm) and relatively yttrium-rich rims (∼16,000–24,000 ppm) are not statistically different (Table 2) and define a single dominant age distribution (Fig. 4).

DISCUSSION

Monazite Growth, Resorption, and Associated Implications

Previous geochronologic studies have noted a correlation between monazite chemical zonation and U-Th-Pb age (e.g., Foster et al., 2002; Gibson et al., 2004; Kohn et al., 2005; Cottle et al., 2009a; Kellett et al., 2010). These studies have shown that individual grains are capable of recording evidence of multiple growth and resorption events and that the relative abundance of yttrium within the different zones can be linked to processes controlling the yttrium budget of the rock. As demonstrated by Kohn et al. (2005), under subsolidus conditions, prograde monazite is likely to have a relatively high initial yttrium concentration that decreases following the onset of garnet growth during continued prograde metamorphism; garnet effectively acts as an yttrium sink. Subsequent monazite growth, therefore, will have relatively lower yttrium content, with the lowest-yttrium monazite reflecting the time at which garnet (Kohn et al., 2005) and/or xenotime was most abundant. Under dehydration melting conditions, however, high phosphorus solubility in Al-rich melts (Wolf and London, 1994), such as those in the Himalaya, may cause monazite to dissolve (Spear and Pyle, 2002; Kohn et al., 2005). The resultant melt will be rich in yttrium, as it is liberated from the breakdown of both monazite and garnet, and phosphorus. If the yttrium is not transported out of the system, it can upon crystallization precipitate high-yttrium monazite as new grains and/or rim overgrowths.

An examination of the monazite analyzed in this study in the context of the aforementioned growth framework sheds light on the geologic history of the Manaslu–Himal Chuli Himalaya. The oldest reported ages, ca. 30 Ma in specimen MS-62, are measured in a moderate-yttrium core (Fig. 3A). That core is interpreted to reflect prograde metamorphic growth during initial Eocene-Oligocene crustal thickening (Fig. 5; e.g., Godin et al., 1999, 2001). A peak of ages spanning ca. 26 to 21 Ma measured in specimen MS-62 comes from an yttrium-poor outer core (Fig. 3A). These dates likely represent a second metamorphic episode (Fig. 5; e.g., Godin et al., 2001), while the low-yttrium content indicates that this episode corresponds to further monazite growth in the presence of garnet and/or xenotime, though no xenotime is observed in this specimen (Fig. 2). This metamorphic event overprinted the earlier event and is reflected in the anatectite-bearing sillimanite-grade assemblages preserved throughout the upper Greater Himalayan sequence in the Manaslu–Himal Chuli Himalaya. Such assemblages are commonly associated with voluminous Miocene melt production in the Himalaya (Deniel et al., 1987; Searle et al., 1997; Streule et al., 2010).

Structurally higher specimens show little internal variation in monazite ages (Fig. 4). The restricted range of ages yielded by specimens MS-34, MS-47, and MS-37 is interpreted to reflect the resorption of any monazite present during early Miocene anatexis into the melt phase (Fig. 5). Growth of these monazite grains is interpreted to have initiated when melt began to crystallize. This growth is characterized by yttrium-rich monazite, which likely reflects increased availability of heavy rare earth elements (HREEs) likely due to garnet resorption during anatexis. Partial resorption of garnet is evident in thin Mn-enriched rims in these specimens (Larson et al., 2010a). Increased yttrium concentration in structurally higher rocks is also reflected by the presence of a few small xenotime grains (Fig. 2). Alternatively, the lack of old monazite ages from structurally higher specimens may indicate that the specimens were not subjected to earlier metamorphism. This is unlikely, however, because early Oligocene–late Eocene ages are commonly reported from upper structural levels within the Greater Himalayan sequence from across the orogen (e.g., Godin et al., 1999, 2001; Lee and Whitehouse, 2007; Cottle et al., 2009a). Furthermore, the garnet within these specimens is characterized by flat major-element profiles (Larson et al., 2010a). Similar flat garnet chemical zoning has been noted within the Greater Himalayan sequence of the Everest region, where it is interpreted to represent thermal overprinting subsequent to original growth (Jessup et al., 2008). This latter interpretation is more consistent with resorption and recrystallization of monazite grains than those rocks escaping an earlier tectonic event.

While record of metamorphism has been lost in monazite at higher structural levels in the study area, in the adjacent Nar region (Fig. 1), on the west side of the Manaslu pluton, rocks at the same structural position as MS-37 (Fig. 1) yield a hornblende 40Ar/39Ar cooling age of 24.9 ± 0.6 Ma (Godin et al., 2006). This implies that peak metamorphic temperatures occurred by ca. 25 Ma at the highest structural level within the Greater Himalayan sequence in the region. This is consistent with the early Miocene monazite dates from MS-62, which is structurally lower in the section. In contrast to the diachronous metamorphism in the lower Greater Himalayan sequence, it appears that a relatively synchronous late Oligocene–early Miocene metamorphic event affected the entire upper Greater Himalayan sequence.

The most prominent peaks in the age probability diagrams in Figure 4 are associated with the yttrium-rich rim in specimen MS-62, moderate-yttrium cores in MS-31, and both yttrium-rich rims and cores in specimens MS-34, MS-47, and MS-37. In all specimens, the main age probability peaks are interpreted to record monazite growth during melt crystallization following metamorphism (Fig. 5). In MS-34, MS-47, and MS-37, the differences in yttrium content between cores and rims may reflect a two-stage crystallization history. The monazite dated in these rocks does not generally conform to a single growth event, either within specimens or between specimens. The range recorded in the main peak within each specimen is larger than would be expected for a single age population, with uncertainties that are normally distributed about a mean value, i.e., the MSWD for each group of analyses is always significantly greater than 1 (Fig. 4). There are three possible explanations for this: (1) The uncertainties assigned to individual isotopic measurements are underestimated; (2) the spread in ages may be attributed to Pb-loss; or (3) they reflect protracted monazite recrystallization and/or growth. We rule out the first scenario because repeat analyses of secondary reference materials (see methods section) result in ages that are within uncertainty of the published values and have MSWD values consistent with a single population, suggesting that the data are accurate and that the uncertainty propagation is appropriate. The second scenario is also considered unlikely due to extremely low rates of Pb diffusion in monazite over geologic time scales (Cherniak et al., 2004). We therefore prefer to interpret the range of ages recorded in each sample to record semicontinuous growth associated with several short thermal pulses or one continuous event.

The new data presented herein also show that the upper structural levels within the Greater Himalayan sequence might have been the source rocks for the young 19 Ma phase of the Manaslu pluton (Fig. 4; Harrison et al., 1999). There is no overlap, however, between the ages of high-structural-level monazite and the older 23 Ma phase of the pluton (Fig. 4). This may be because the monazite that recorded an early Miocene thermal pulse was subsequently resorbed in the country rock, or it may indicate a different melt source for that phase of the pluton, perhaps from lower structural levels. The early Miocene peaks recorded in specimen MS-62 overlap with the 23 Ma phase of magmatism in the Manaslu pluton (Fig. 4).

Insight on Orogenic Models

The new U-Th-Pb dates and previously published 40Ar/39Ar thermochronology for the present study area can be compared to model predictions for the evolution of the migmatitic upper Greater Himalayan sequence. It is important to note that while thrust wedge and channel-flow models may not be mutually exclusive at the orogenic scale, each makes testable predictions about the P-T-t paths recorded in rocks that have been deformed and metamorphosed during either process. In thermo-mechanical channel-flow models, peak temperatures in the Greater Himalayan sequence occur in the early Miocene (ca. 20 Ma) and remain high until the late Miocene, when rapid cooling initiates, cooling through ∼350 °C in the latest Miocene–earliest Pliocene (Jamieson et al., 2004). In contrast, thrust wedge models predict peak temperatures in the early Miocene, followed by immediate and gradual cooling reaching temperatures below 350 °C by ca. 10 Ma (Kohn, 2008).

The data from this study are consistent with an early Miocene metamorphism/thermal pulse, as is the generation of the initial phase of the Manaslu pluton (e.g., Harrison et al., 1999). The data are also consistent with temperatures remaining elevated for a significant period, with monazite growth occurring throughout the middle Miocene (Fig. 4). In addition, 40Ar/39Ar thermochronology from the study area indicates that these rocks cooled through argon diffusion closure temperatures in muscovite predominantly in the Pliocene (Copeland et al., 1990; Larson et al., 2010a). The geochronologic and thermochronologic data from the migmatitic upper Greater Himalayan sequence in the Manaslu–Himal Chuli Himalaya, therefore, appear to be more consistent with thermal-mechanical channel flow (e.g., Jamieson et al., 2004) than with models that only evoke thrust wedge deformation. Furthermore, the ages of the primary probability peaks for Th-Pb monazite dates in this study appear to decrease down structural section (Fig. 4). While this crystallization pattern may be explained by cooling that migrated from upper structural levels to lower structural levels through time, it can also be explained by postcrystallization juxtaposition of formerly laterally separated material by differential translation within a ductilely extruded midcrustal channel. A similar pattern is noted in the adjacent Annapurna region, though only two of the specimens reported yielded melt crystallization-related peaks (Corrie and Kohn, 2011). The absence of similar peaks in other specimens from the Annapurna region may reflect the relative paucity of anatectite in that region.

As previously mentioned, in the inverted metamorphic package of lower Greater Himalayan sequence rocks, monazite dates generally decrease in age from 21 Ma to 8 Ma with deeper structural position (Kohn et al., 2001). This is interpreted to record diachronous metamorphism, the downward migration of shear along the Main Central thrust, and subcretion of overridden rocks to the base of the Greater Himalayan sequence (Larson et al., 2010a). These are characteristics similar to those expected in a thrust wedge model (e.g., Kohn, 2008) and in fact are entirely consistent with predictions from within the “Main Central thrust zone” in thermo-mechanical models of Jamieson et al. (2004), which describe the evolution of the lower portion of the metamorphic core as the progressive incorporation of material from outside the original channel that was buried at increasingly shallow depths. The exhumed metamorphic sequence as a whole, therefore, records thrust wedge–style deformation in its lower portion and ductile channel flow in its migmatitic upper portion. This is entirely consistent with the thermal-mechanical model of Jamieson et al. (2004) and the field-based findings of Larson et al. (2010a), which demonstrate a continuum between channel-flow processes in the upper Greater Himalayan sequence and thrust wedge processes in the lower Greater Himalayan sequence. Midcrustal lateral flow and thrust wedge taper are, therefore, compatible processes integral to the contemporaneous accommodation of convergence at different locations and structural levels within large, hot orogenic systems.

CONCLUSIONS

Zoned monazite grains in upper Greater Himalayan sequence migmatitic paragneiss record evidence of metamorphic events in the Oligocene (ca. 30 Ma) and early Miocene (ca. 23 Ma). Prominent peaks associated with high-yttrium monazite domains dominate age probability plots of the data collected for each specimen. These domains, which range between 20 and 17 Ma, are interpreted to have grown during melt crystallization. Older monazite material at higher structural levels in the Greater Himalayan sequence appears to have been consumed during partial melting and regrown during subsequent crystallization; there are no old yttrium-poor cores preserved as there are in structurally lower specimens. These new data indicate that melt was present throughout the upper Greater Himalayan sequence at approximately the same time, allowing for lateral ductile flow in response to a horizontal pressure gradient. These new ages are consistent with thermo-mechanical models and field-based studies that outline a deformational continuum between lateral midcrustal flow in the orogenic hinterland and thrust wedge in the orogenic foreland, and they demonstrate that the two processes are not mutually exclusive as they have been commonly characterized in recent literature.

Funding from a Natural Sciences and Engineering Research Council Discovery Grant to L. Godin, a Canada Graduate Scholarship to K. Larson, and a U.S. National Science Foundation grant (EAR-1119380) to J. Cottle supported this study. S. Creighton is thanked for his analytical support at the Saskatchewan Research Council. Logistical support was provided by Pradap Tamang, Ralesh Tamang, Sure Tamang, Singgi Tamang, Yo Singgi, Forpa Tamang, Kami Tamang, and Teke Tamang from Peke Peak Trekking. This manuscript benefited from discussions with D. Kellett, M. Searle, and R. Price. Careful reviews by R. Jamieson, M. Williams, and K. Mahan and editorial guidance from J. Goodge helped improve this paper.