The petrology, geochronology and significance of Granite Harbour Intrusive Complex xenoliths and outcrop sampled in western McMurdo Sound, Southern Victoria Land, Antarctica

Abstract

Granite Harbour Intrusive Complex xenoliths in McMurdo Volcanic Group rocks and in situ outcrops have been studied from Mount Morning, western McMurdo Sound, Antarctica. Calc-alkalic samples have whole rock signatures and normative compositions similar to the Dry Valleys 1b suite, and zircon grains in one specimen yield a 545.2 ± 4.4 Ma crystallisation age. This supports subduction-related magmatism initiating in Southern Victoria Land by 545 Ma. A second group of xenoliths is alkalic, with titanite grains in one xenolith from this group dated at 538 ± 8 Ma. Whole rock chemistry, normative compositions and geochronology of the alkalic group are comparable to the Koettlitz Glacier Alkaline Suite (KGAS). The position of a proposed lower crustal discontinuity that may form a significant basement suture in the McMurdo Sound region is newly constrained to the east of Mount Morning, perhaps along the trace of the Discovery Glacier. The boundary between East and West Antarctica may also pass along the trace of the Discovery Glacier if, as previously hypothesised, its location is controlled by the basement suture. A significant basement suture may also have provided the necessary egress for the (regionally) early and sustained magmatic activity observed at Mount Morning over the last 24 million years.

Keywords:

• Granite Harbour Intrusive Complex
• Koettlitz Glacier Alkaline Suite
• Mount Morning
• Ross Orogen
• Transantarctic Mountains
• U-Pb geochronology

Introduction

Antarctica's Ross Orogen formed during Neoproterozoic and early Paleozoic Gondwanaland amalgamation, creating part of one of the world's longest orogenic belts, the Terra Australis orogen (Cawood 2005). In the McMurdo Sound region of Antarctica, the stretched and thinned lithosphere underlying the Ross Sea, associated with the West Antarctic Rift System, sits adjacent to the up-thrown Ross Orogen Transantarctic Mountains, to provide dramatic topographic relief of up to 5000 m between East and West Antarctica (Fig. 1A–C). The boundary, or front (Transantarctic Mountains Front or TMF), between East and West Antarctica is at the eastern edge of the Transantarctic Mountains. In the McMurdo Sound region the position of the TMF has been seismically mapped north of a prominent, approximately east-west trending lineament (the Blue Lineament; Wilson 1999) at approximately 78ºS (McGuiness et al. 1985; Barrett et al. 1995), but south of this the location of the TMF is debated (Fig. 2A). In the McMurdo Sound region a major and significant discontinuity in the lower crust has been proposed based upon geophysics (Smithson 1972) and isotopic and geochemical studies of lower crustal xenoliths entrained in Cenozoic magmas (Kalamarides et al. 1987; Berg et al. 1989; Berg 1991; Kalamarides & Berg 1991). It has also been suggested that this discontinuity may mark a difference in age, although no supporting geochronological evidence has been presented (Kalamarides et al. 1987). The discontinuity is thought to occur at the TMF between East and West Antarctica, and Kalamarides et al. (1987) proposed that ‘the isotopic differences in the lower crust across the boundary suggest that the current faulting and rifting may coincide with an older crustal suture.’ They drew the TMF position to coincide with the seismically mapped boundary north of the Blue Lineament and south of the Blue Lineament they defined an approximate position for the TMF using information obtained from lower crustal xenolith information (Fig. 2). One limitation of their study was the availability of lower crustal xenolith localities close to the approximate TMF position south of Brown Peninsula (Fig. 2). A slightly different position for the TMF is suggested by Wilson et al. (2007) in their figure 1 (Fig. 2B of this study), where they infer the TMF to continue south of the Blue Lineament to intersect a second, approximately east-west, prominent lineament (the Radian Lineament; Wilson 1999). Wilson et al. (2007) infer a c. 10 km dextral offset of the TMF at the Radian Lineament. To the south of the Radian Lineament, the TMF is inferred to follow a third, approximately north-south trending, prominent structural feature: the Pyramid Lineament (Wilson 1999).

Figure 1 Location. A, Outline of ice-covered Antarctic continent. B, Regional setting. The dashed line indicates the western boundary of the West Antarctic Rift System (after LeMasurier 2008). C, Victoria Land. Cenozoic volcanic provinces in Victoria Land include Hallett volcanic province (Hvp), Melbourne volcanic province (Mvp) and Erebus volcanic province (Evp) (Harrington 1958; Kyle & Cole 1974). The pluton names are italicised and are referenced in the text, Fig. 10, and the online supplementary data.

In this study a suite of 46 crustal xenoliths and two in situ outcrops of basement from Mount Morning, a Cenozoic eruptive centre in McMurdo Sound, have been examined. Mount Morning is located immediately south of the Radian Lineament and east of the Pyramid Lineament in the area of the inferred position of the TMF (Figs 1–2). The crustal sample suite has been examined using petrography and whole rock geochemistry and a sub-set of samples has been analysed for mineral geochemistry, thermometry, and geochronology to distinguish between an East Antarctica (Transantarctic Mountain-like) or West Antarctica (Ross Sea-like) affinity for the crust beneath Mount Morning. The results are significant because (i) they refine the position of the lower crustal discontinuity first proposed by Kalamarides et al. (1987); (ii) they can be used to assess the effect of refining the position of the lower crustal discontinuity on the inferred position of the TMF; (iii) they can advance understanding of the petrogenesis of the crust beneath Mount Morning as it relates to either the Transantarctic Mountains or McMurdo Sound; and iv) they can be used to discuss the control, if any, of the pre-existing basement boundary suture (Kalamarides et al. 1987) on the distribution of Cenozoic volcanism.

Figure 2 Structures and eruptive centres in Southern Victoria Land. A, Diagram of Southern Victoria Land showing the location of eruptive centres in the Erebus volcanic province including Mount Morning (MM), Mason Spur (MS), Gandalf Ridge (GR), Mount Discovery (MD), Minna Bluff (MB), Brown Peninsula (BP), Black Island (BI), White Island (WI), and Ross Island (RI). The Koettlitz Glacier Alkaline Suite (KGAS) is marked (thin dashed outline) along with the approximate area of the extension of Transantarctic Mountain (TAM)-like basement identified in this study (thick dashed ellipse). The Discovery Accommodation Zone is defined by Wilson (1999) as a structural corridor where the axis of the Transantarctic Mountains is displaced between the Dry Valleys block and the Mulock Glacier (Fig. 1), incorporating both the Royal Society and Skelton blocks. The locations of lower crustal granulite xenoliths (Kalamarides et al. 1987) are shown as circles; those with tholeiitic (Ross Sea-like) compositions are light fills with black rims; those with calc-alkalic (Transantarctic Mountain-like) compositions are dark fills with black rims. The Transantarctic Mountain Front (TMF) is seismically mapped (after McGuiness et al. 1985; Barrett et al. 1995). B, A restricted section of the image in panel A. showing the potential extension of the TMF, and the approximate trend of the lower crustal discontinuity, south of the Blue Lineament. The TMF may follow the proposed lower crustal discontinuity trend inferred from lower crustal granulites.

Geological setting

Mount Morning is a stratovolcano cropping out in the Ross Sea c. 100km south-west of the active Mount Erebus volcano, within the West Antarctic Rift System (Fig. 1). Mount Morning has been active for at least 18.7 Ma and probably as far back as at least 24.1 Ma (Martin et al. 2010). This is longer lived than any other Cenozoic eruptive centre or province in Antarctica (LeMasurier & Thompson 1990; Martin et al. 2010). The volcano consists of the strongly alkalic Mason Spur Lineage and the mildly alkalic Riviera Ridge Lineage (Martin et al. 2013). The crustal xenoliths were collected exclusively in Riviera Ridge Lineage rocks that erupted between at least 6.13 and 0.02 Ma (Martin et al. 2010). Faults, interpreted as the onshore continuation of part of the TMF fault array, have been mapped on the north-east flank of Mount Morning (Kyle & Muncy 1989; Martin & Cooper 2010). Based upon thermobarometry of plagioclase-bearing mantle xenoliths, the crust beneath Mount Morning is thought to be < 20 km thick (Martin et al. 2014). To the east of Mount Morning the Ross Sea crust is estimated to be 18–20 km thick and it has been suggested that to the west of Mount Morning, beneath the Transantarctic Mountains, the crustal thickness increases to around 40 km (Berg et al. 1989; Bannister et al. 2003).

West of the West Antarctic Rift System are the Transantarctic Mountains (Fig. 1B–C), a product of rift flank uplift in Southern Victoria Land since the Palaeogene (Fitzgerald 1992). The Transantarctic Mountains comprise late Precambrian to Cambrian Ross Orogen lithologies of the Skelton Group (Cooper et al. 2011) that include various greywackes, schists, metaconglomerates, argillites, quartzites, marbles, and metavolcanic rocks (Borg & DePaolo 1991; Stump 1992, 1995; Cook & Craw 2001; Cox et al. 2012). These are unconformably overlain by Devonian to Triassic aged sediments of the Beacon Supergroup (Barrett 1981, 1991; Collinson 1991; Woolfe & Barrett 1995), into which the Ferrar Dolerite was injected during the Jurassic (Elliot 1992; Heimann et al. 1994). The development of the Ross Orogen was punctuated by voluminous magmatism, with the emplacement of plutons that are collectively referred to as the Granite Harbour Intrusive Complex (GHIC; Gunn & Warren 1962; Findlay 1985; Stump 1995; Cox et al. 2012). The GHIC rocks were emplaced into the Skelton Group between the late Precambrian and early Paleozoic (e.g. Encarnacion & Grunow 1996; Cox et al. 2000; Allibone & Wysoczanski 2002; Tulloch & Ramezani 2012) and the history is summarised in Cox et al. (2012) and Goodge et al. (2012). Despite the size of the Ross Orogen, its study is challenged by remoteness, inclement weather, long periods of darkness, extreme relief, and obscuring ice cover. In the Transantarctic Mountains, the GHIC comprises four distinct suites of plutonic rocks. Three of these (Dry Valleys 1a–DV1a; Dry Valleys 1b–DV1b and Dry Valleys 2–DV2) are calc-alkalic or alkalic-calcic and consist mainly of granitoids (Smillie 1992; Allibone et al. 1993a,b; Cox et al. 2000; Allibone & Wysoczanski 2002). Both the DV1a and DV1b suites have chemistry consistent with their generation and emplacement in a convergent margin setting. The DV1a Suite plutons have calc-alkalic compositions with relatively low Sr/Y ratios (LoSY; Tulloch & Kimbrough 2003), whereas DV1b Suite plutons have alkalic-calcic chemistry and high Sr/Y ratios (HiSY), which has also been referred to as ‘adakite-like’. The DV2 suite is alkalic-calcic and associated with high-K and shoshonitic dyke swarms (e.g. Keiller 1991). The fourth plutonic suite is the alkalic, Precambrian to Cambrian Koettlitz Glacier Alkaline Suite (KGAS) that crops out between the Walcott and Darwin Glaciers (Cooper et al. 1997; Read et al. 2002; Cottle & Cooper 2006a; Fig. 1C). Rock types include syenites, carbonatites, and A-type granite intrusives (Rowell et al. 1993; Hall et al. 1995; Read et al. 2002; Read 2010). Despite the overall compressional regime in this section of Antarctica during the Precambrian to early Palaeozoic (Gunn & Warren 1962; Borg & DePaolo 1991; Stump 1992, 1995), the KGAS alkalic chemical associations are considered to have been emplaced in an extensional tectonic setting (Read et al. 2002) that may have been part of an extensional jog associated with oblique compression during the Ross Orogeny (Encarnacion & Grunow 1996; Read et al. 2002) or represent a period of more widespread extension prior to the onset of convergence (Allibone 1988).

The lower crustal discontinuity in the McMurdo Sound region is based upon the study of mainly granulites (Kalamarides et al. 1987; Berg et al. 1989; Berg 1991; Kalamarides & Berg 1991) the location of which is shown in Fig. 2. The granulite xenoliths from the Transantarctic Mountains are, relative to granulite from the Ross Sea area, characterised by higher δ¹⁸O, ⁸⁷Sr/⁸⁶Sr and SiO₂ content, and by retention of original calc-alkalic major and trace element signatures despite granulite facies metamorphism (Kalamarides et al. 1987; Berg et al. 1989; Berg 1991; Kalamarides & Berg 1991). The granulite xenoliths from the Ross Sea area are reported to contain lower silica and are more basic relative to granulite from the Transantarctic Mountains, and they may have originated from alkalic or tholeiitic parents. It has also been speculated that an age difference may exist between Transantarctic Mountain-like lower crust and lower crust beneath the Ross Sea (Kalamarides et al. 1987; Kalamarides & Berg 1991), although there is little geochronological evidence published to support this hypothesis.

Methods

Whole rock chemistry

Representative bulk rock samples from Mount Morning were analysed by X-ray fluorescence (XRF) at the University of Otago, New Zealand, on a Philips PW2400 XRF spectrometer, following analytical procedures described in Martin et al. (2013) and Norrish & Chappell (1977) and using the SuperQ qualitative software control program (3^rd ed.). Samples were crushed using a TEMA swing mill fitted with a WC head. Martin et al. (2013) have demonstrated that for samples prepared in WC carbide, significant contamination is limited to W and Co. Fused discs were produced for major element analyses (and subsequently used for some trace element analyses) and pressed powder pellets for trace elements. The lower limit of detection is better than 0.004 wt% (weight percent) oxide (for Al₂O₃). Loss on ignition is at 1100 °C. In representative, selected samples, whole rock trace elements were analysed by laser-ablation inductively coupled mass spectrometry (LA-ICPMS) in the Otago Community Trust Centre for Trace Element Analysis at the University of Otago. After major element analysis, the XRF discs were cut, mounted in epoxy resin and polished. Laser ablation was conducted with a pulsed NewWave 213 nm laser operated at 50%, 10 Hz, a 100 µm spot size and a tracking rate of 10 µm/s. Ablated material was carried by He gas from the sample cell, mixed with Ar and inlet into an Agilent 7500 ICP-MS. Background (laser off) data were acquired for 20 s followed by 40 s of analyses with the laser on giving about 100 mass scans from a 400 µm track approximately 5 µm deep. A purge time of at least 120 s was allowed between each spot analysis. Each set of tracks on ten unknowns was bracketed by analyses of standard glass NIST 610 (Pearce et al. 1997). Trace-element concentrations were obtained by normalising count rates for each element to those for Si in the sample and standard using known SiO₂ and trace element concentrations in NIST 610 and the SiO₂ in the fused discs obtained previously by XRF analysis. The J-series rock standards, by comparison with results reported in Orihashi et al. (2003), show precision (1σ relative) of ≤ 5% for ≥ 10 ppm, ≤ 20% for 1–10 ppm and ≤ 50% for 0.1–1 ppm. In terms of accuracy, all results are within 25% (and most within 10%) of values reported by Orihashi et al. (2003).

Mineral chemistry

The JEOL JXA-8600 electron microprobe at the University of Otago was used to determine mineral major element chemistry in wavelength dispersive mode, operated with accelerating voltage 15 kV, current 1 μA and beam diameter 20 μm. A standard ZAF correction was applied to all data. Natural silicates and oxides were used as standards. A typical lower limit of detection is shown in the online supplementary material. Zircon mineral rare earth element (REE) and Ti chemistry were collected from specimen OU78656 at the same time as U-Pb data was collected for geochronology (see below for instrument configuration) at the Australian National University. Count rates for each element were normalised to Si, assuming SiO₂ to be stoichiometric in zircon with a concentration of 32.8 wt%, and multiplying by a correction factor based on measurement of standard glass NIST 610 with concentration values recommended by Pearce et al. (1997). The low abundance ⁴⁹Ti was chosen to avoid interference with ⁹⁶Zr on the major ⁴⁸Ti peak.

U-Pb chronology

Zircon grains from sample OU78656 were analysed by LA-ICPMS at the Australian National University. Grains were separated by conventional techniques, handpicked and mounted in polished epoxy mounts. Cathodoluminescence images were used to target specific grains and zones within grains for analysis. Instrument settings were set at a 35 µm spot diameter at 5 Hz with 70% power, with fluence values c. 2.47 J/cm² typical, and further details are given in Ballard et al. (2001). Data reduction and uncertainty propagation methods are described in Scott & Palin (2008). Standard data from NIST-610 were collected and zircon reference material TEMORA-1 (Black et al. 2003) was used as the primary standard: the reproducibility of the reference material was < 5% (2σ) for both ²⁰⁷Pb/²⁰⁶Pb and ²⁰⁶Pb/²³⁸U.

Uranium-Pb geochronology of titanite was determined at the NERC Isotope Geosciences Laboratory (NIGL), Keyworth, Nottingham, using a polished thin section of sample OU78658. Analyses were made in situ with a Nu Instruments AttoM single collector ICPMS and a New Wave Research 193FX laser ablation system. The full method and instrument parameters employed are described in Thomas et al. (2013). Laser ablation was conducted with a 35μm static spot at a fluence of c. 2.2 J/cm² for 30 seconds. Sample data were normalised to Ontario-2 titanite reference material (Spencer et al. 2013), and Khan and Fish Canyon Tuff titanite reference materials were also analysed during the session to check on precision and accuracy. Analyses of Khan titanite gave an age of 516 ± 7 Ma 2σ (ID-TIMS age: 522 Ma; unpublished NIGL data) and analyses of Fish Canyon Tuff titanite gave an age of 30.1 ± 1.7 Ma 2σ (ID-TIMS age: 28.5 Ma; Schmitz & Bowring 2001).

Results

Petrography and sample description

Forty-eight samples were selected from a range of localities across Mount Morning. The most prolific crustal xenolith site occurs in an isolated cinder cone on the upper flanks of Mount Morning (78.26.044 south; 164.26.479 east; altitude 1118 m). All sample locations are given in the online supplement Table S1. Xenolith diameters range between < 1 cm and 50 cm (Fig. 3A–C). In rare cases crustal xenoliths comprise 20% of the outcrop of the host basanite (Paulsen 2008 and see Fig. 3A) but abundances are commonly much lower (Fig. 3B). In addition, two in situ outcrops of rocks belonging to the GHIC (Kyle & Muncy 1989; Martin & Cooper 2010) were also sampled; the petrographic characteristics of these samples have been described previously (Martin & Cooper 2010). Based upon macroscopic texture the 48 samples have been divided into metamorphic (granulite) and igneous (plutonic) sub-types. The petrographic details are summarised in Table 1. Of the thirteen granulites collected, only 1 is mafic, with the remaining igneous sub-types representing a full range of felsic to intermediate lithologies.

Table 1 Summary of sample petrography.

Rock type	Classification	Protolith	OU#	Colour	Texture	grain size (cm)	Minerals
Metamorphic	Granulite	Granite	OU78656	leucocratic	foliated	0.75	Qtz	Afs	Pl	Cpx
Metamorphic	Granulite	Granodiorite	OU78491	leucocratic	foliated	0.75	Qtz	Afs	Pl	Opx	Zr-tr	Ap-tr
Metamorphic	Granulite	Ulvöspinel Gabbro	OU78496	mesocratic	foliated	0.75	Pl	Cpx	Usp	Opx	Hbl	Spl
Plutonic	Granitoid	-	OU78713	leucocratic	equigranular	1	Qtz	Afs	Pl	Cpx
Plutonic	Nph Alkali Feldspar Syenite	-	OU78682	leucocratic	hiatal	3–1.5	Afs-p	Nph	Amp	Cpx	Pl
Plutonic	Nepheline Syenite	-	OU78657	leucocratic	equigranular	1	Nph	Afs	Aeg
Plutonic	Syenite	-	OU78683	leucocratic	equigranular	1	Afs	Aeg-Aug	Spl
Plutonic	Monzodiorite	-	OU78685	melanocratic	equigranular	1.5	Pl	Sp	Qtz	Afs	Aug
Plutonic	Gabbroic diorite	-	OU78454	melanocratic	equigranular	1	Pl	Cpx	Qtz	Spl
Plutonic	Anorthosite	-	OU78690	leucocratic	equigranular	1	pl	Opx-tr
Plutonic	Kaersutite Gabbro	-	OU78699	mesocratic	equigranular	1	krs	pl	Usp-tr
Plutonic	Gabbro	-	OU78457	mesocratic	equigranular	1	Pl	Cpx	Spl	Qtz
Plutonic	kaersutite Gabbronorite	-	OU78674	mesocratic	porphyritic	2 & 0.1	Pl-p	Opx-p	Cpx-p	Krs-g	Spl-g
Plutonic	Norite	-	OU78663	mesocratic	equigranular	0.5	Pl	Opx	Spl	Krs-tr
Plutonic	Ulvöspinel Gabbro	-	OU78659	mesocratic	equigranular	0.75	Pl	Cpx	Usp	Opx	Hbl	Spl

OU#, Otago University catalogue number. All mineral abbreviations follow Whitney & Evans (2010): Aeg, aegerine; Afs, alkali feldspar; Amp, amphibole; Ap, apatite; Aug, augite; Cpx, clinopyroxene; Hbl, hornblende; Krs, kaersutite; Nph, nepheline; Pl, plagioclase; Qtz, quartz; Spl, spinel; Usp, Ulvöspinel; Zr, zircon. Other abbreviations: g, groundmass; p, phenocryst; tr, trace. Classification of kaersutite is from Martin (2009) and aegerine is from Paulsen (2008). Other minerals are optically determined.

Figure 3 Photographs of field outcrops and thin sections. A, Abundant crustal xenoliths entrained in a phonolitic air fall deposit (163.37.771E; 78.27.245 S). B, A felsic granulite xenolith with characteristic foliation and leucocratic – mesocratic banding (163.49.675; 78.23.901). C, An alkalic syenite xenolith enclosed in volcanic breccia (164.06.094 E; 78.28.286 S). D, Dated specimen OU78656 in hand specimen. E, Thin section photograph in cross-polarised light (XPL) showing the typical texture observed in dated specimen OU78656. F, Thin section photograph in XPL showing the typical texture observed in dated specimen OU78658.

The granulite xenoliths have a foliation defined by the common orientation of the long axis of mineral grains and alternation of leucocratic and melanocratic bands (2 to 20 mm-thick; Fig. 3B), with the banding defined by the abundance of pyroxene. No garnet was identified in any granulite specimen analysed as part of this study or observed in the field. One felsic granulite specimen (OU78656) has been isotopically dated and is described here. The sample is leucocratic and equigranular (0.75 mm) with a foliation defined (Fig. 3D–E) by the common orientation of the long-axes of quartz, alkali feldspar, plagioclase, orthopyroxene, and deformed FeTi-oxide grains and the alternation of mm-scale light and dark bands (pyroxene-poor and -rich respectively). Accessory minerals are zircon (Fig. S1), apatite, and titanite. One igneous-textured specimen (OU78658) has also been dated. This sample has an equigranular (1 mm), coarse-grained texture (Fig. 3F) and is composed of labradorite, microcline, clinopyroxene, ferropargasitic amphibole, titanite, and quartz, plus trace amounts of biotite, apatite, and magnetite.

Whole rock chemistry and normative QAP compositions

The whole rock data are summarised in Table 2 with full details provided in the online supplementary Table S1. Whole rock compositions are plotted onto the plutonic TAS (total alkalis versus silica) diagram of Middlemost (1994) in Fig. 4A. The samples that plot above the alkalic–sub-alkalic division line of either Macdonald & Katsura (1964) or Kuno (1966) are alkalic (black stars on Fig. 4A). The remaining samples are sub-alkalic and plot in the calc-alkalic field of Irvine & Baragar (1971; online supplementary figure S2). All the metamorphic textured granulites are calc-alkalic with the igneous textured xenoliths being both calc-alkalic (the majority) and alkalic. None of the samples has a tholeiitic composition.

Table 2 Whole rock geochemistry.

OU#	OU78656	OU78658	OU78713	OU78714
Field ID	3G	151E	221b	225b
Association	calc-alkalic	alkalic	calc-alkalic	calc-alkalic
Classification	Metamorphic	Igneous	Igneous	Igneous
Longitude (east)	163.47.373	163.23.352	164.07.112	164.08.080
Latitude (south)	78.22.605	78.29.678	78.20.961	78.20.885
SiO2 wt%	73.34	61.71	74.52	73.68
TiO2	0.26	0.53	0.07	0.09
Al2O3	13.50	13.99	14.69	14.87
Fe2O3 T	2.13	4.07	0.34	0.8
MnO	0.02	0.07	bdl	0.01
MgO	0.46	2.47	0.12	0.23
CaO	1.91	6.39	0.92	0.92
Na2O	2.73	1.56	3.54	3.54
K2O	5.05	8.99	5.83	5.83
P2O5	0.07	0.19	0.04	0.04
Loi%	0.35	0.56	0.5	0.5
Total	99.94	100.76	100.57	100.57
Fe Ratio	0.50	0.50	0.50	0.50
Mg#	43.5	68.4	58.3	53.3
Q	33	8	-	-
A	32	23	-	-
P	30	69	-	-
Sc ppm	3.89	9.29	63.32	3.29
V	22.1	64.0	300	15.4
Cr	20.9	33.2	406	75.1
Co	15.8	34.4	110	26.4
Ni	20.2	28.7	207	48.6
Ga	14.3	13.1	19.5	35.7
Rb	132	264	4.89	108
Sr	117	455	638	400
Y	5.56	42.0	11.4	5.87
Zr	106	78.7	47.6	41.6
Nb	3.20	11.6	2.15	3.92
Ba	415	1200	51.6	1458
La	36.9	29.9	7.27	40.3
Ce	52.7	88.7	18.4	33.4
Pr	4.65	10.5	2.14	2.90
Nd	15.3	44.7	9.25	13.6
Sm	2.56	10.1	4.55	3.80
Eu	0.73	1.08	4.09	0.80
Gd	1.45	8.13	bdl	3.01
Tb	0.23	1.38	1.50	0.71
Dy	1.23	8.31	2.85	2.17
Ho	0.23	1.61	1.06	bdl
Er	0.55	4.27	bdl	bdl
Tm	0.09	0.59	bdl	bdl
Yb	0.80	4.77	bdl	bdl
Lu	0.11	0.58	0.81	0.67
Hf	3.28	2.28	bdl	bdl
Ta	0.11	1.86	bdl	0.59
Pb	27.8	48.7	5.91	32.4
Th	5.39	9.52	bdl	6.25
U	0.87	5.98	bdl	1.16

OU#, Otago University catalogue number.

Figure 4 Whole rock chemistry and normative mineralogy. A, Total alkalis versus silica diagram with field boundaries after Middlemost (1994). The dashed lines represent various alkalic (high alkalis)–sub-alkalic division lines. B, Normative QAP compositions of Mount Morning granulite and alkalic xenoliths compared to the Dry Valley suite (DV) fields of Smillie (1992) and the Koettlitz Glacier Alkaline Suite (KGAS) of Read (2010).

The normative compositions of the granulite xenoliths have been calculated and plotted onto the quartz-alkali feldspar-plagioclase feldspar (QAP; Streckeisen 1976) diagram where they plot in the granodiorite and monzogranite fields (Fig. 4B). The majority of monzogranite data, including dated specimen OU78656, overlap with the field defined for DV1 suite rocks defined by Smillie (1992). One monzogranite specimen (OU78442) and the two samples from in situ outcrop overlap with the field defined for DV2 rocks, though they also plot very close to the DV1 field. The granulite specimens with granodiorite protolith compositions do not overlap with either of the DV1 or DV2 fields defined by Smillie (1992). The normative compositions of the crustal xenoliths with alkalic whole rock compositions are also plotted on Fig. 4B. Four specimens, including dated specimen OU78658, lie in the field defined for KGAS by Read (2010). One specimen plots with a monzonite composition slightly outside of the KGAS field and five samples (not shown) have nepheline normative compositions. The Mount Morning granulite xenoliths have high SiO₂ contents between c. 65 and 75 wt%, and Fe₂O₃ less than about 5 wt% (Fig. 5A). The whole rock compositions have around 15 wt% Al₂O₃, variable P₂O₅ between 0.05 and 0.15 wt% (Fig. 5B) and up to 5000 ppm Ba (Fig. 5C).

Figure 5 Whole rock chemistry of Mount Morning granulite xenoliths. Field for comparison is the Transantarctic Mountain-like lower crustal granulite xenoliths (TM) and grey circles show Ross Sea-like data from lower crustal granulite xenoliths, after Kalamarides & Berg (1991). A, wt% SiO₂ versus Fe₂O₃. B, wt% Al₂O₃ versus P₂O₅. C, ppm Ba versus Ni.

One alkalic igneous (OU78658) specimen, one calc-alkalic metamorphic (OU78656) specimen, and one calc-alkalic igneous (OU78714) specimen are plotted on a normal mid ocean ridge basalt (N-MORB) normalised extended element plot (Fig. 6A), with other whole rock crustal xenolith patterns shown in grey lines for comparison. All patterns have decreasing enrichment levels with increasing element compatibility and marked negative anomalies (relative to the element immediately before and after) of Nb, P, and Ti and enrichments in K, Pb, and Nd. For comparison, a representative alkalic Riviera Ridge Lineage volcanic specimen (OU78540) from Mount Morning is plotted in Fig. 6A. This sample lacks a negative Nb or Ti anomaly and does not have a significant K, Pb, or Nd anomaly. Whole rock analyses from representative GHIC samples in the Transantarctic Mountains are shown in Fig. 6B for further comparison and are discussed below.

Figure 6 Whole rock multi element diagrams. A, normal mid ocean ridge basalt, normalised (to the values of Sun & McDonough 1989) extended element plot of calc-alkalic metamorphic sample OU78656, calc-alkalic (labelled as granite) sample OU78714 and alkalic sample OU78658. Other Mount Morning crustal sample patterns are depicted with grey lines. Shown for comparison is a representative analysis of the volcanic host for crustal xenoliths at Mount Morning, OU78540 (Martin et al. 2013). B, A representative analysis from the Fontaine Pluton (Cottle & Cooper 2006b), analyses from the Dry Valleys suites (DV1a, DV1b and DV2; Cox et al. 2000) and a representative analysis of the alkalic (part of the KGAS) A-Type Mulock Glacier Granite (Cottle & Cooper 2006a).

The felsic granulite xenoliths and the two in situ Gandalf Ridge granites have been plotted onto granite discrimination diagrams in Fig. 7A–B and lie within the volcanic arc granite and the I- and S-type granite fields. On a Y versus Sr/Y plot (Fig. 7C) the granites and felsic granulites have low Y and high Sr and almost all plot within the adakite field defined by Defant et al. (2002) and may be described as HiSY.

Figure 7 Felsic granulite discrimination plots. Plotted for comparison are the McMurdo Volcanic Group granulite xenolith from Cox et al. (2000) and the Dry Valleys suites of granites (Cox et al. 2000) A, Y+Nb versus Rb plot after Pearce et al. (1984). The Mount Morning data all plot within the volcanic arc granite (VAG) field. B, 10⁴Ga/Al versus Zr plot after Whalen et al. (1987). The Mount Morning data all plot within the I- and S-type field. C, Y versus Sr/Y plot with adakite fields from Defant et al. (2002) and Defant & Drummond (1990) plotted for comparison. Several of the Mount Morning data plot within the adakite field. Shown for comparison are the DV1b adakite-like rocks (Cox et al. 2000), Fontaine Pluton (FP) rocks (Cottle & Cooper 2006b), Fontaine Adakite (FA) rocks (Cottle 2002) and Panorama Pluton rocks (Mellish et al. 2002; Fig. 1C).

Mineral chemistry and thermometry

Feldspar and zircon mineral chemistry was collected from dated specimen OU78656 with representative compositions shown in Table 3 and the complete analyses given in the online supplement Table S2. Two feldspar compositions are present: plagioclase with a typical andesine composition of An₃₂Ab₅₉Or₉; and orthoclase with a typical composition of An₂Ab₂₇Or₇₁ (Fig. 8A). In zircon, REEs and Ti have been measured. The mantle normalised zircon REE patterns have a steeply positive trend and Nd and Eu depletions relative to the elements with similar compatibility (Fig. 8B). The measured Ti abundance varies between 9 and 43 ppm, with the median value being 16 ppm (n = 8).

Table 3 Mineral chemistry from granulite specimen OU78656.

Category	OU#	–	SiO2	Al2O3	CaO	Na2O	K2O	Total	O	Si	Al	Ca	Na	K	Total	Or	Ab	An	T °C^1	T °C^2
Feldspar
Granulite	OU78656	wt%	59.97	24.36	7.06	7.25	0.70	99.4	32	10.771	5.158	1.359	2.525	0.161	19.984	4.0	62.4	33.6	871	887
Granulite	OU78656	wt%	60.07	23.77	6.68	6.34	2.03	99.0	32	10.866	5.069	1.295	2.223	0.468	19.932	11.8	55.8	32.5	–	–
Granulite	OU78656	wt%	60.08	23.95	6.79	6.51	1.88	99.3	32	10.840	5.094	1.313	2.277	0.433	19.966	10.8	56.6	32.6	–	–
Granulite	OU78656	wt%	60.40	24.05	6.88	6.45	1.98	99.9	32	10.841	5.088	1.323	2.244	0.453	19.964	11.3	55.8	32.9	–	–
Granulite	OU78656	wt%	60.40	24.07	6.76	6.59	1.83	99.8	32	10.846	5.095	1.301	2.294	0.419	19.965	10.4	57.2	32.4	–	–
Granulite	OU78656	wt%	60.42	24.06	6.84	6.35	2.00	99.8	32	10.849	5.093	1.316	2.211	0.458	19.932	11.5	55.5	33.0	–	–
Granulite	OU78656	wt%	60.54	23.98	6.73	6.70	1.70	99.7	32	10.865	5.073	1.294	2.331	0.389	19.958	9.7	58.1	32.2	–	–
Granulite	OU78656	wt%	60.62	24.12	6.77	7.11	1.32	100.1	32	10.838	5.084	1.297	2.465	0.301	19.997	7.4	60.7	31.9	–	–
Granulite	OU78656	wt%	60.63	24.19	6.81	7.70	0.36	99.8	32	10.831	5.094	1.303	2.667	0.081	19.990	2.0	65.8	32.2	–	–
Granulite	OU78656	wt%	60.65	24.07	6.66	7.10	1.23	99.9	32	10.848	5.075	1.276	2.462	0.281	19.971	7.0	61.3	31.8	–	–
Granulite	OU78656	wt%	61.22	23.07	5.76	7.26	1.93	99.7	32	11.004	4.888	1.109	2.530	0.443	20.030	10.8	62.0	27.2	–	–
Granulite	OU78656	wt%	65.03	18.50	0.53	3.13	12.09	99.4	32	11.947	4.007	0.105	1.115	2.834	20.018	69.9	27.5	2.6	–	–
Granulite	OU78656	wt%	65.06	18.37	0.41	3.02	12.18	99.1	32	11.983	3.988	0.080	1.078	2.862	19.997	71.2	26.8	2.0	–	–
Granulite	OU78656	wt%	65.52	18.17	0.33	2.95	12.38	99.4	32	12.023	3.930	0.065	1.050	2.898	19.970	72.2	26.2	1.6	–	–
Category	OU#	–	La	Ce	Pr	Nd	Pm	Sm	Eu	Gd	Tb	Dy	Ho	Er	Tm	Yb	Lu		Ti	T °C^3
Zircon
Granulite	OU78656	ppm	0.07	4.66	nd	1.48	nd	2.28	0.30	12.6	3.80	64.6	15.1	71.2	13.5	126	20.3		23	821
Granulite	OU78656	ppm	0.17	3.80	nd	0.20	nd	1.02	bdl	11.3	4.89	89.2	36.0	216	50.6	558	121		11	747
Granulite	OU78656	ppm	0.68	14.7	nd	3.42	nd	9.36	0.30	69.4	24.7	406	131	688	144	1448	278		21	812
Granulite	OU78656	ppm	bdl	19.1	nd	1.43	nd	2.71	0.13	15.4	4.86	72.5	20.8	101	19.7	187	33.9		43	887
Granulite	OU78656	ppm	0.36	16.4	nd	0.92	nd	3.46	bdl	32.2	12.7	227	80.5	455	101	1068	215		25	830
Granulite	OU78656	ppm	0.08	6.83	nd	0.75	nd	2.46	bdl	20.7	7.81	134	45.5	250	53.9	558	110		10	737
Granulite	OU78656	ppm	bdl	8.59	nd	0.87	nd	2.33	bdl	17.1	6.08	100	31.8	167	34.9	351	66.8		9	732
Granulite	OU78656	ppm	bdl	21.3	nd	2.85	nd	6.07	0.25	37.8	12.4	193	57.0	284	56.8	550	110		11	747

bdl, below detection limit; nd, not determined; OU#, Otago University catalogue number.

¹ Fuhrman & Lindsley (1988) two-feldspar thermometer; ²Elkins & Grove (1990) two feldspar thermometry; ³Watson et al. (2006) Ti in zircon thermometer.

Figure 8 Mineral chemistry from dated specimen OU78656. A, Feldspar ternary diagram. Black line represents the isotherm at 887 °C and 15 kb, based upon the feldspar data with filled symbols. B, Zircon rare earth element data normalised to chondrite (McDonough & Sun 1995).

Two-feldspar geothermometry is based upon the exchange between An-Ab-Or components of co-existing orthoclase and plagioclase feldspars such as seen in OU78656. Problems exist with the application of two-feldspar geothermometry to metamorphic rocks, as noted by Elkins & Grove (1990) and Fuhrman & Lindsley (1988) and reviewed by Kroll et al. (1993). The major problem is that often the two feldspar pairs do not plot on a single isotherm, which invariably leads to error in the temperature calculation (Kroll et al. 1993). The two-feldspar analyses in OU78656 do, however, plot on a single isotherm (Fig. 8A), thus in this instance the two-feldspar geothermometry method is applicable. A maximum pressure of 15 kb is thought reasonable for the geothermometry calculations based upon the thickness of crust beneath the adjacent Transantarctic Mountains (Berg et al. 1989; Bannister et al. 2003). Application of two different geothermometers returns very similar results of 871 °C (Fuhrman & Lindsley 1988) and 887 °C (Elkins & Grove 1990). Natural and experimental results have defined a log-linear relationship between temperature and Ti content in zircon (Watson & Harrison 2005; Watson et al. 2006; Ferry & Watson 2007) that has been extensively tested using natural suites of samples, for example see the reviews of Fu et al. (2008), Ewing et al. (2013) and references there-in. Using the Ti values from zircons in sample OU78656, a range of temperatures is calculated between 732 and 887 °C with a median temperature of 779 °C (n = 8).

Geochronology

The petrography of the dated samples is described above and the geochronology results for these are presented in Table 4. Zircon grains separated from felsic granulite specimen OU78656 are elongate, optically clear grains ≤ 200 µm in maximum dimension. Oscillatory concentric zoning, typical of magmatic zircon, is discernible on cathodoluminescence images (supplementary Fig. S3). Eighteen zircons were analysed by LA-ICPMS, yielding concordant (± 4%; Fig. 9A) ²⁰⁶Pb/²³⁸U results between c. 565 and 531 Ma and a weighted average ²⁰⁶Pb/²³⁸U age at 545.2 ± 4.4 Ma (MSWD = 1.2, probability of fit = 0.23; Fig. 9B). For alkalic specimen OU78658, 24 spots were measured across three titanite crystals. These yield a regression towards common lead giving a ²⁰⁶Pb/²³⁸U age of 538 ± 8 Ma (2σ; MSWD = 0.94; Fig. 9C).

Table 4 U-Th-Pb isotope data for grains analysed by LA-ICPMS from sample OU78656 and OU78658.

	Signals			Ratios									Isotopic ages
	Th/	Pb	U	^238U/	1σ%	^207Pb/	1σ%	^207Pb/	1σ%	^206Pb/	1σ%	Rho	^207Pb/	2σ abs	^206Pb/	2σ abs	^207Pb/	2σ abs	% conc.
Grain	U	(ppm)	(ppm)	^206Pb		^206Pb		^235U		^238U			^206Pb		^238U		^235U		‡
OU78656
z01	0.43	48	536	11.55	0.41	0.0586	1.43	0.6995	2.64	0.0866	1.06	0.72	552.2	31.2	538.5	14.2	535.3	5.7	100.0
z02	0.48	159	1676	11.05	0.30	0.0585	0.83	0.7295	2.35	0.0905	1.03	0.89	547.7	18.0	556.3	13.1	558.3	5.7	101.0
z04	0.76	68	707	11.55	0.61	0.0616	2.36	0.7351	3.27	0.0865	1.16	0.65	660.2	50.6	559.5	18.3	535.1	6.2	96.0
z07	0.49	122	1309	11.29	0.35	0.0588	1.40	0.7180	2.62	0.0885	1.04	0.81	560.1	30.6	549.5	14.4	547.0	5.7	100.0
z08	0.39	15	166	11.43	0.63	0.0582	2.70	0.7023	3.53	0.0875	1.17	0.54	537.7	59.0	540.1	19.0	540.7	6.3	100.0
z09	0.48	74	792	11.20	0.61	0.0582	2.18	0.7159	3.14	0.0893	1.16	0.60	536.0	47.7	548.3	17.2	551.2	6.4	101.0
z11	0.55	283	3047	11.47	0.36	0.0586	1.12	0.7045	2.48	0.0872	1.05	0.91	552.5	24.5	541.4	13.4	538.8	5.6	100.0
z14	0.60	45	471	11.34	0.53	0.0571	1.92	0.6946	2.96	0.0882	1.12	0.68	496.7	42.4	535.6	15.8	544.8	6.1	102.0
z16	0.76	67	668	11.32	0.39	0.0572	2.13	0.6968	3.07	0.0883	1.06	0.77	499.5	46.9	536.9	16.5	545.7	5.8	102.0
z18	0.62	49	496	11.06	0.50	0.0598	4.49	0.7455	5.03	0.0904	1.14	0.74	596.4	97.3	565.6	28.4	558.0	6.4	99.0
z20	0.47	40	432	11.41	0.59	0.0578	3.78	0.6987	4.41	0.0877	1.16	0.50	522.2	82.9	538.0	23.7	541.7	6.3	101.0
z21	0.36	167	1882	11.40	0.32	0.0588	0.82	0.7107	2.35	0.0877	1.03	0.87	558.2	17.8	545.2	12.8	542.1	5.6	100.0
z22	0.42	87	948	11.23	0.35	0.0591	1.08	0.7255	2.46	0.0890	1.04	0.81	570.3	23.6	553.9	13.6	549.9	5.7	100.0
z23	0.40	21	237	11.57	0.67	0.0580	2.26	0.6905	3.21	0.0864	1.19	0.58	528.2	49.6	533.1	17.1	534.2	6.3	101.0
z24	0.41	225	2499	11.40	0.37	0.0578	0.90	0.6990	2.39	0.0877	1.05	0.89	522.0	19.7	538.2	12.8	542.0	5.7	101.0
z26	0.61	45	477	11.49	0.63	0.0574	2.82	0.6885	3.62	0.0870	1.17	0.52	505.8	62.0	531.9	19.3	538.0	6.3	102.0
z28	0.57	199	2153	11.59	0.47	0.0608	2.95	0.7233	3.70	0.0863	1.11	0.91	632.2	63.6	552.6	20.5	533.5	5.9	97.0
z29	0.58	70	726	11.17	0.36	0.0587	1.42	0.7253	2.62	0.0895	1.04	0.77	557.8	30.9	553.8	14.5	552.8	5.8	100.0
OU78658
t1_01	-	13	397	11.521	1.95	0.0706	1.15	0.8442	2.27	0.0868	1.95	0.86	945	24	537	20	621	21	43
t1_02	-	6	194	11.239	2.07	0.0797	1.22	0.9769	2.40	0.0890	2.07	0.86	1189	24	549	22	692	24	54
t1_03	-	6	190	10.952	2.04	0.0799	1.41	1.0052	2.48	0.0913	2.04	0.82	1194	28	563	22	706	25	53
t1_04	-	6	185	10.888	2.07	0.0777	1.19	0.9833	2.39	0.0918	2.07	0.87	1139	24	566	22	695	24	50
t1_05	-	8	221	10.905	1.91	0.0806	1.13	1.0186	2.22	0.0917	1.91	0.86	1212	22	566	21	713	23	53
t1_06	-	13	389	11.446	1.93	0.0659	1.13	0.7936	2.24	0.0874	1.93	0.86	803	24	540	20	593	20	33
t1_07	-	11	340	11.302	1.83	0.0690	1.00	0.8413	2.09	0.0885	1.83	0.88	898	21	547	19	620	19	39
t1_08	-	13	393	10.975	1.83	0.0675	1.15	0.8473	2.16	0.0911	1.83	0.85	852	24	562	20	623	20	34
t3_01	-	7	223	11.324	2.03	0.0751	1.34	0.9141	2.43	0.0883	2.03	0.83	1071	27	546	21	659	23	49
t3_02	-	7	222	11.426	1.96	0.0799	1.11	0.9637	2.25	0.0875	1.96	0.87	1194	22	541	20	685	22	55
t3_03	-	8	221	10.783	2.25	0.0785	1.47	1.0029	2.69	0.0927	2.25	0.84	1159	29	572	25	705	27	51
t3_04	-	8	224	11.213	2.16	0.0762	1.43	0.9370	2.59	0.0892	2.16	0.83	1101	29	551	23	671	25	50
t3_05	-	8	224	11.224	1.98	0.0766	1.35	0.9408	2.39	0.0891	1.98	0.83	1111	27	550	21	673	23	50
t3_06	-	8	229	11.315	1.79	0.0758	1.00	0.9229	2.05	0.0884	1.79	0.87	1089	20	546	19	664	20	50
t3_07	-	7	226	11.557	2.08	0.0790	1.17	0.9420	2.38	0.0865	2.08	0.87	1172	23	535	21	674	23	54
t3_08	-	7	222	11.151	2.14	0.0780	1.41	0.9645	2.56	0.0897	2.14	0.84	1148	28	554	23	686	25	52
t4_01	-	7	208	11.403	1.99	0.0869	1.25	1.0498	2.36	0.0877	1.99	0.85	1357	24	542	21	729	24	60
t4_02	-	6	183	11.095	2.10	0.0830	1.21	1.0315	2.43	0.0901	2.10	0.87	1270	24	556	22	720	25	56
t4_03	-	7	193	11.062	1.98	0.0841	1.19	1.0472	2.31	0.0904	1.98	0.86	1294	23	558	21	728	24	57
t4_04	-	6	173	11.310	1.94	0.0787	1.07	0.9592	2.22	0.0884	1.94	0.87	1165	21	546	20	683	22	53
t4_05	-	8	208	10.048	2.21	0.1489	3.81	2.0421	4.40	0.0995	2.21	0.50	2333	65	612	26	1130	58	74
t4_06	-	7	185	10.464	2.07	0.1045	1.38	1.3764	2.48	0.0956	2.07	0.83	1706	25	588	23	879	29	66
t4_07	-	7	197	10.981	1.98	0.0818	1.18	1.0267	2.31	0.0911	1.98	0.86	1241	23	562	21	717	23	55
t4_08	-	6	186	11.100	2.00	0.0797	1.47	0.9899	2.48	0.0901	2.00	0.81	1190	29	556	21	699	25	53

‡Concordance percentage (% conc.) calculated by dividing the ²⁰⁶Pb-²³⁸U date by the ²⁰⁶Pb/²⁰⁷Pb date and multiplying by 100.

Figure 9 U-Pb geochronology of Mount Morning xenoliths. A, Conventional concordia plot for zircons (OU78656) analysed by the LA-ICPMS method. B, A ²⁰⁶Pb-²³⁸U plot of data from Fig. 9A with numbers (z1, z2, etc.) linked to the data in Table 4. C, Tera-Wasserburg plot for titanite (OU78658) analysed by the LA-ICPMS method.

Discussion

Association of Mount Morning lower crust granulite samples

A calc-alkalic composition is evident in Mount Morning lower crustal xenoliths despite granulite facies metamorphism (Fig. S2). This is one of the characteristics Kalamarides et al. (1987) used to define lower crust associated with the Transantarctic Mountains. The Mount Morning granulite data overlap with the Transantarctic Mountain-like field and plot discretely from the Ross Sea-like field – with these discriminatory fields defined by Kalamarides & Berg (1991) – on whole rock major element variation diagrams (Fig. 5). On a QAP diagram, granulite xenoliths representing monzogranite protoliths overlap with the fields for the DV 1 and 2 suites of rocks (Fig. 4B). The granulite and in situ granite whole rock multi-element diagrams have negative Nb, P, Ti, and Ce? anomalies and positive K, Pb, and Nd anomalies that are very similar to the whole rock multi-element patterns of the Fontaine Pluton and DV intrusives (Fig. 6) and the Panorama Pluton (not shown). The calc-alkalic felsic granulite data and in situ granite data plot in the volcanic arc granite field of the Y+Nb versus Rb diagram and overlap with the field for DV1b suite rocks (Fig. 7A). The bulk rock chemistry of several granulite xenoliths, and the in situ granite samples, features relatively low Y values and high Sr/Y ratios (Fig. 7C) that have been described as either adakite-like or HiSY. The age of the dated granulite specimen, OU78656, overlaps with that of the Fontaine Pluton (and is discussed in detail below). We interpret the petrology, geochemistry and geochronology to indicate that the lower crustal granulite xenoliths from Mount Morning have the defining characteristics provided by Kalamarides et al. (1987) for Transantarctic Mountain-like lower crust and that at least some samples can be specifically related to the DV1b and DV2 suites of rocks (and the Fontaine Pluton).

A possible adakite-like signature in the granulite samples

The bulk rock chemistry of several granulite xenoliths, and the in situ granite samples, features relatively low Y values and high Sr/Y ratios (Fig. 7C). These characteristics, along with SiO₂ ≥ 56 wt%, Al₂O₃ ≥ 15 wt%, MgO < 3 wt% are typical of adakite-like rocks (Castillo 2006). Defant & Drummond (1990) identified rocks with low Y and Yb values and high Sr/Y and La/Yb ratios in a Cenozoic arc and described them as adakites, and rocks elsewhere in the world with comparable geochemistry are commonly referred to as ‘adakite-like’ (Castillo 2006; Wang et al. 2007). Much controversy exists over the origin(s) of rocks with an adakitic geochemical signature (see Castillo (2006) for a comprehensive review). The geochemistry is generally interpreted to reflect the involvement of garnet in melting or fractionation processes; Y and HREE such as Yb are strongly partitioned into garnet relative to melt. It has been variously argued that adakite-like rocks are: (a) possible examples of partial melts of subducted oceanic crust (Defant & Drummond 1990; Kay et al. 1993; Stern & Kilian 1996; Defant et al. 2002; Martin et al. 2005); (b) formed via fractional crystallisation and assimilation processes acting on parental basaltic magmas in an arc setting (Castillo et al. 1999); or (c) derived by partial melting of thickened lower crust or a basaltic underplate (Atherton & Petford 1993; Wang et al. 2007). Mount Morning is underlain by a basement whose age and chemistry are comparable to those of rocks formed during the Ross Orogeny. Ross Orogen plutons have long been convincingly argued to be the product of subduction processes (Borg et al. 1990; Encarnacion & Grunow 1996; Cox et al. 2000; Cottle & Cooper 2006b). The close geochemical affinity between Mount Morning basement xenoliths and Transantarctic Mountain plutons (Figs 6–7) can be interpreted to suggest that at least part of the Mount Morning basement originated through magmatic processes taking place in a Cambrian to Precambrian arc. While this does not rule out the possibility that the adakite-like basement rocks at Mount Morning are the product of partial melting of a thickened lower crust, it is considered more likely that they formed from the partial melting of subducted oceanic crust. Martin et al. (2013, 2014) have also argued that the metasomatic fluids responsible for mantle refertilisation in the lithosphere beneath Mount Morning were derived from melts formed during early Palaeozoic subduction.

Age of the granulite samples and their significance to Ross Orogen magmatism

Ross Orogen magmatism is thought to have been initiated around 590 Ma as defined by zircon ages in glacial granitoid clasts from the central Transantarctic Mountains (Goodge et al. 2012). This age estimate is in agreement with the detrital zircon record preserved in syn- to post-orogenic molasse sediments; these have an age estimated at c. 580 Ma (Goodge et al. 2002). Magmatism is thought to have continued to around 480 Ma (Goodge et al. 2012). In the central Transantarctic Mountains the magmatism has been divided by Goodge et al. (2012) into three phases. The first phase (I) represents early igneous activity within a developing arc between 590 and 555 Ma. This was followed by a phase of syn-tectonic magma generation (phase II) represented by plutons that were emplaced within deep parts of the Ross Orogen between 555 and 515 Ma, and the third phase (III) of mostly post-tectonic granitoid bodies was emplaced within a rapidly growing arc between 515 and 480 Ma. In Southern Victoria Land Ross Orogen magmatism has been divided into four suites (Cox et al. 2012 and see Introduction). The age ranges of various plutons within each suite are summarised on Fig. 10. The early KGAS has distinctive alkaline and A-type affinities with a range of ages reported between 579 to 495 Ma, whereas the DV1a suite is interpreted to have formed between 521 Ma and 485 Ma. The DV1b suite has distinctive adakite-like or HiSY chemistry and has ages reported to be between 536 and 486 Ma. The DV2 suite is interpreted to have been emplaced post-subduction and has ages reported to be between 505 and 492 Ma (Fig. 10).

Figure 10 Summary compilation of ages from key plutons in the Southern Victoria Land segment of the Transantarctic Mountains. The numbers link to Table S3 with the description of the pluton and appropriate reference. Included for comparison are the two new data from this study (see Fig. 9). The specific references are: Rowell et al. (1993); Turnbull et al. (1994); Hall et al. (1995); Encarnacion & Grunow (1996); Cooper et al. (1997); Cox et al. (2000); Allibone & Wysoczanski (2002); Mellish et al. (2002); Read et al. (2002); Cottle & Cooper (2006b); Veevers et al. (2006); Rocchi et al. (2009); Read (2010); Hagen-Peter et al. (2011); Tulloch & Ramezani (2012).

Calc-alkalic granulite specimen OU78656 (this study) is chemically similar to DV1b suite rocks. Both two-feldspar thermometry and Ti in zircon thermometry provide a maximum estimated temperature of 887 °C for this sample (Table 3), which is below the zircon closure temperature at about 1000 °C (e.g. Flowers et al. 2005). There is some debate as to whether Ti in zircon thermometry represents a crystallisation temperature or cooling temperature (e.g. Ewing et al. 2013) but the concentric zoning (Fig. S3) and the REE pattern (Fig. 8C) in the zircon are typical of igneous zircon (Belousova et al. 2002; Fu et al. 2009). The agreement between different thermometers is also evidence of rock equilibration and consequently the radiometric data are interpreted to indicate a magmatic crystallisation age. In the context of the regional geology this also makes sense, as the age of sample OU78656 at 545.2 ± 4.4 Ma is at least 5 Ma older than the oldest age reported from the DV1b suite (datum 27 on Fig. 10), which agrees more with a crystallisation rather than a cooling age. This age is comparable to the Fontaine Pluton some 150 km south of Mount Morning (Cottle & Cooper 2006b; Fig. 10) and further supports the hypothesis that Ross Orogen magmatism was initiated in Southern Victoria Land at around 545 Ma. This age falls within the range of phase II (555 to 515 Ma) Ross Orogen magmatism in the central Transantarctic Mountains and is consistent with development of an early magmatic arc during this period.

Association of Mount Morning alkalic crustal xenoliths

The alkalic xenolith whole rock trace element patterns in Fig. 6A have Nb, K, Pb, P, and Ti anomalies similar to those observed in samples of Transantarctic Mountain plutons (including KGAS rocks; Fig. 6B); characteristics not shared with typical Cenozoic alkalic whole rock trace element patterns obtained for Mount Morning volcanic rocks. This is an indication that the alkalic xenoliths are not the plutonic equivalent of Cenozoic volcanic rocks seen at Mount Morning but are in fact chemically similar to GHIC alkalic rocks in the Transantarctic Mountains. In particular the alkalic whole rock chemistry is comparable to the KGAS (Figs 4 & 6). As part of this study, titanite from an alkalic xenolith (OU78658) has been dated at 538 ± 8 Ma, which indicates that this sample is also temporally associated with the GHIC (Fig. 10). Ages derived from titanite grains in KGAS rocks are usually interpreted to reflect cooling ages (e.g. Read 2010). Some workers have proposed that U content in titanite is a discriminator between metamorphic and igneous titanite. For example Aleinikoff et al. (2002) suggest metamorphic titanite contains < 100 ppm U; the titanite grains in sample OU78658 contain hundreds of ppm U (Table 4) as expected in igneous titanite. The euhedral and unzoned morphology of the titanite grains is also consistent with an igneous origin. Although we cannot rule out that the 538 ± 8 Ma date is a cooling age rather than a crystallisation age, the overall conclusion that sample OU78658 is associated with the KGAS (Fig. 10) remains valid. By inference the chemistry and chronology are consistent with the suggestion that all of the alkalic xenoliths considered in this study are most appropriately assigned to the KGAS. This interpretation suggests the boundary of the KGAS extends eastwards to include the basement area beneath Mount Morning and this extension is drawn (with approximate borders) in Fig. 2. Even though the broad regional tectonic regime in the TAM at the time of emplacement of the KGAS is compressional, the KGAS chemistry is thought to reflect a localised extensional jog that has been comprehensively discussed in a number of studies (Encarnacion & Grunow 1996; Read et al. 2002; Read et al. 2003; Read 2010).

Location of the lower crustal discontinuity

The Mount Morning lower crustal granulites described in this study are geochemically and temporally associated with the DV1b and DV2 rocks and the alkalic xenoliths with the KGAS, and both may be associated with the Transantarctic Mountain-like lower crust described by Kalamarides et al. (1987). Following Kalamarides et al. (1987) this allows the geochemical discontinuity between the Transantarctic Mountain basement and the Ross Sea basement to be drawn east of Mount Morning, reducing the possible uncertainty in the basement discontinuity location to < 25 km. Based upon this interpretation the lower crustal discontinuity probably passes east of Mount Morning and west of Mount Discovery and Minna Bluff, perhaps along the line of the Discovery Glacier (Fig. 2). Constrained by the southern-most extent of the seismically mapped section of the TMF and the occurrences of granulite xenoliths at Mount Morning, the lower crustal discontinuity is a straight line closely aligned with the position of the TMF inferred by Wilson et al. (2007) between the Blue and Radian Lineaments (Fig. 2B). However, the new position of the TMF inferred from this study, east of Mount Morning, indicates that little offset of this boundary occurs along the Radian Lineament. The location of the lower crustal boundary to the south of Mount Morning is unknown.

The possible trace of the Transantarctic Mountain front fault and melt migration

Kalamarides et al. (1987) have speculated that the position of the TMF in the McMurdo Sound region is controlled by an ancient lower crustal suture, which they have identified from markedly different isotopic ratios, geochemistry, and possibly chronology in lower crustal granulites collected either side of the TMF. If this line of argument is applied to the observations and conclusions derived from the present study, with the lower crustal geochemical discontinuity lying to the east of Mount Morning, the TMF must also lie to the east of Mount Morning, possibly along the trace of the Discovery Glacier (Fig. 2). The presence of post 3.9 Ma onshore faulting on the north-eastern flank of Mount Morning (Martin & Cooper 2010), inferred to form part of the TMF fault array, may support this interpretation. Mantle xenolith geothermobarometry implies that the lithosphere beneath Mount Morning is < 20 km thick (Martin et al. 2014), indicating that the crust has been thinned relative to the approximately 40 km-thick crust associated with the adjacent Transantarctic Mountains. This abrupt gradient in crustal thickness may indicate that significant faults associated with the TMF occur to the west of Mount Morning. The final word on the location of the TMF south of the Blue Lineament awaits further geophysical work. However, if Kalamarides et al. (1987) are correct and a lower crustal discontinuity can be mapped in granulite xenoliths in the McMurdo Sound region, then it seems plausible that Pliocene, or earlier, onshore faulting at Mount Morning is closely following the trend of a pre-existing lower crustal suture. In this scenario it is unsurprising that magmatic activity is known earliest at Mount Morning (since 24.1 Ma) out of all the Southern Victoria Land volcanic centres and that magmatism is known to have occurred to at least 0.02 Ma (Martin et al. 2010).

Conclusions

A suite of crustal xenoliths, including lower crustal granulites, and two in situ outcrops have been analysed from Mount Morning, Southern Victoria Land, Antarctica. At least some show affinities with Granite Harbour Intrusive Complex rocks from the adjacent Transantarctic Mountains. In particular, the lower crustal granulite xenolith petrography, whole rock chemistry, normative mineralogy, and geochronology are comparable to DV1b suite rocks in the adjacent Transantarctic Mountains. Furthermore, a sub-set of these rocks has adakite-like (or HiSY) compositions that can be explained by partial melting of, or modification by fluids derived from, subducted oceanic crust. Zircon grains from one granulite sample yield a 545.2 ± 4.4 Ma age that is interpreted as a crystallisation age, comparable to the emplacement age of the Fontaine Pluton, and this supports the proposal for the existence of a magmatic arc during this period. The Mount Morning alkalic xenoliths are geochemically comparable to the Koettlitz Glacier Alkaline Suite (KGAS). Titanite grains in one alkalic xenolith yield a 538 ± 8 Ma lower intercept age that may reflect cooling but is within the range of crystallisation and cooling ages reported for KGAS rocks. Based on the study of granulite xenoliths, Kalamarides et al. (1987) have proposed a lower crustal discontinuity in the McMurdo Sound region. The Transantarctic Mountain-like compositions in the Mount Morning granulite suite are consistent with the proposal that this lower crustal discontinuity occurs east of Mount Morning, perhaps along the trace of the Discovery Glacier. The discontinuity could be a significant suture in the lower crust. The Transantarctic Mountain range bounding fault may also pass along the trace of the Discovery Glacier if its location is controlled by the lower crustal suture, as hypothesised by Kalamarides et al. (1987). The extended Cenozoic magmatic history at Mount Morning (more than 24 Myr duration) could be the result of magma exploiting this lower crustal suture.

Supplementary data

Supplementary file 1: Figure S1. Characteristics of dated specimen. Calc-alkalic felsic granulite specimen OU78656 in hand specimen with an in situ zircon in crossed polarised light shown in the inset.

Supplementary file 2: Figure S2. Whole rock discrimination diagrams. A, Alkalic–sub-alkalic division line of Irvine & Baragar (1971). B, The samples that plot in the sub-alkalic field of diagram A. All plots in the calc-alkalic field of Irvine & Baragar (1971).

Supplementary file 3: Figure S3. Images of zircon and titanite grains used for dating. A, Cathodoluminescence images of zircon grains from felsic granulite sample OU78656 used in U-Pb geochronology. Red circles mark approximate areas of analysis and zircon number (z3, z11, etc.) correlating with analysis numbers in Table 4. B, Plane polarised light photographs of titanite (Ttn) from alkalic sample OU78658.

Supplementary file 4: Tables S1–S3. Table S1. Whole rock chemistry; Table S2. Mineral chemistry; Table S3. Age constraints.

Acknowledgements

Adam Martin was supported by an Antarctica New Zealand (New Zealand Post) Antarctic scholarship and a University of Otago award from the Department of Geology. Antarctica New Zealand is thanked for logistic support. Brent Pooley and Damion Walls are thanked for help with sample preparation and analysis. JM Palin is thanked for his assistance. Andrew Allibone and an anonymous reviewer are thanked for their helpful reviews and Richard Jongens for editorial handling.