Isotopic and trace element geochemistry of alkalic–mafic–ultramafic–carbonatitic complexes and flood basalts in NE India: Origin in a heterogeneous Kerguelen plume

Available online at www.sciencedirect.com

Geochimica et Cosmochimica Acta 115 (2013) 46–72 www.elsevier.com/locate/gca

Isotopic and trace element geochemistry of alkalic–mafic–ultramafic–carbonatitic complexes and flood basalts in NE India: Origin in a heterogeneous Kerguelen plume Arundhuti Ghatak 1, Asish R. Basu ⇑ Department of Earth and Environmental Sciences, University of Rochester, Rochester, NY 14627, USA Received 27 June 2012; accepted in revised form 4 April 2013; available online 12 April 2013

Abstract The Archean East Indian cratonic margin was affected by the Kerguelen plume (KP) 117 Ma, causing flood-basalt eruptions of the Rajmahal–Bengal–Sylhet Traps (RBST). The RBST cover one million km2 in and around the Bengal Basin as alkalic– ultrabasic intrusives in the west and Sikkim in the north, and Sylhet basalts and alkalic–carbonatitic–ultramafic complexes in the Shillong plateau – Mikir hills farther east of the Rajmahal–Bengal Traps. We provide new Nd–Sr–Pb-isotopic and trace element data on 21 unreported discrete lava flows of the Rajmahal Traps, 56 alkalic–carbonatitic–mafic–ultramafic rocks from four alkalic complexes, and three dikes from the Gondwana Bokaro coalfields, all belonging to the RBST. The data allow geochemical correlation of the RBST with some contemporaneous Kerguelen Plateau basalts and KP-related volcanics in the southern Indian Ocean. Specifically, the new data show similarity with previous data of Rajmahal group I–II basalts, Sylhet Traps, Bunbury basalts, and lavas from the southern Kerguelen Plateau, indicating a relatively primitive KP source, estimated as: eNd(I) = +2, 87 Sr/86Sr(I) = 0.7046, with a nearly flat time-integrated rare earth element (REE) pattern. We model the origin of the uncontaminated RBST basalts by 18% batch melting with a 2 chondritic KP source in the spinel-peridotite stability depths of 60–70 km in the mantle. The new geochemical data similar to the Rajmahal group II basalts indicate a light REE enriched average source at eNd(I) = 5, 87Sr/86Sr(I) = 0.7069. Our geochemical modeling indicates these lavas assimilated granulites of the Eastern Ghats, reducing the thickness of the continental Indian lithosphere. Lack of an asthenospheric MORB component in the RBST province is indicated by various trace element ratios as well as the Nd-Sr isotopic ratios. Three alkalic complexes, Sung, Samchampi, and Barpung in NE India, and one in Sikkim to the north are of two groups: carbonatites, pyroxenites, lamproites, nephelinites, sovites, melteigite in the first group and syenites and ijolites in the second. The Nd–Sr–Pb-isotopic and trace element geochemistry of the first group of carbonatitic–ultrabasic rocks are consistent with similar data of the RBST lavas of the present and previous studies, and are modeled as derived from a relatively primitive carbonated garnet peridotite source in the KP. In contrast, the syenites and ijolites of the second group show a wide range of Nd–Sr–Pb isotopic compositions, modeled by low-degree melts of an ancient recycled carbonated eclogite also in the KP. The KP thus reflects heterogeneities in the lower mantle-derived plume with carbonated components yielding ultrabasic melts at greater depths with low-degree melting, followed by rise of the plume at shallower depths causing tholeiitic flood basalt volcanism. Collectively, these data imply a zone of influence of the plate-motion-reconstructed KP head for 1000 km around the Bengal Basin, as represented by the widely scattered and diverse rock types of the RBST. Ó 2013 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Current address: Department of Earth and Environmental Sciences, University of Texas at Arlington, Arlington, TX 76019, USA. E-mail address: [email protected] (A.R. Basu). 1 Current address: Department of Earth and Environmental Sciences, Indian Institute of Science Education and Research Bhopal, Gas (ITI Rahat) Building, Govindpura, Bhopal 462023, Madhya Pradesh, India.

0016-7037/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.gca.2013.04.004

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

1. INTRODUCTION Large volume basaltic volcanism that erupted in the Early Cretaceous on the eastern Indian continental margin, southwestern Australia (Bunbury-Naturaliste Plateau), and Antarctica (Fig. 1) are considered to have caused the opening of the Indian Ocean (e.g. Mahoney et al., 1983; Frey et al., 2000). This widespread volcanism is attributed to a large plume, the remnant of which is a hot spot beneath the Kerguelen Plateau (e.g. Storey et al., 1989; Weis et al., 1989; Kent et al., 1997). The Kerguelen plume is believed to have created the Ninetyeast Ridge (NER), Broken

Fig. 1. Map of part of the Indian Ocean and surrounding continents with physiographic features, after Ingle et al. (2002) and Frey et al. (2000), showing locations of the Sylhet and Rajmahal Traps in northeastern India. Also shown in gray is the extended Eastern Ghats – Shillong orogenic belt (Yin et al., 2010) along the east coast of India. Basalt provinces attributed to the Kerguelen plume (Frey et al., 2002) include Kerguelen Plateau, Broken Ridge, Ninetyeast Ridge, Bunbury basalts and Rajmahal Traps. Abbreviations used for relevant fields of comparison are: BB – Bunbury basalt drill core sites; NP – Naturaliste Plateau (Site 264); NKP – North Kerguelen Plateau; CKP – Central Kerguelen Plateau; SKP – South Kerguelen Plateau; CG – Chilka Granulites (Chakrabarti et al., 2011). Black crosses are ODP sites. Sites 253, 254, 756, 757, 214, 216, and 758 are from the Ninetyeast Ridge and are grouped as NER in subsequent Nd–Sr–Pb isotopic plots.

47

Ridge, Bunbury basalts, Naturaliste and Kerguelen Plateaus in the southern Indian Ocean, and the Comei igneous Province in southern Tibet (Fig. 1) (e.g. Weis and Frey, 1991; Frey et al., 2000; Zhu et al., 2009; Ghatak and Basu, 2011). The Kerguelen hotspot with high 3He/4He ratios (18 R/RA) belongs to the same group of hotspots as Hawaii and Iceland (Ingle et al., 2004). Based on geochronological, geochemical data and plate reconstructions, the largest episode of Kerguelen volcanism seem related to the eastern Indian volcanic province of the RBST (Fig. 2a) of 116 ± 3.5 Ma age (Pantulu et al., 1992; Baksi, 1995; Kent et al., 2002; Ray et al., 2005; Ghatak and Basu, 2011). The RBST (Fig. 2) occupy 2  105 km2 (Baksi, 1995; Kent et al., 2002). The early geochemical studies of the Rajmahal Traps did not consider a connection between the Kerguelen basalts and the volcanism in eastern India although the Kerguelen plume was implicated as the heat source for the basaltic traps from a compositionally ‘normal’ asthenosphere (Mahoney et al., 1983; Baksi et al., 1987; Storey et al., 1992). It is being recognized now that Rajmahal-age volcanic rocks are wide-spread in and around the Bengal Basin as diverse groups of alkalic, ultrabasic, carbonatitic rocks and basalts over an area of 1 million km2 (Fig. 2a); from the Rajmahal hills in the west, these volcanic suites are believed to extend beneath the Tertiary sediments of the Bengal Basin in West Bengal (e.g. Ghatak and Basu, 2011). These alkalic rocks include lamproite dikes in the Gondwana sediments to the west, carbonatite–alkalic complexes in the Shillong plateau and Mikir hills to the northeast, and lamproite dikes of the Sikkim Himalayas to the north (Fig. 2a). The alkalic intrusives are similar to those found in the Deccan and Siberian and other flood basalt provinces in an association commonly found with mantle plumes (e.g. Basu et al., 1993, 1995; Bell, 2001). It has been recognized that in some flood basalt provinces alkalic and carbonatitic magmatism occurred before and after the main pulse of tholeiitic volcanism (Basu et al., 1993, 1995; Bell, 2001; Heaman et al., 2002). We report here the trace element and Nd–Sr–Pb isotopic analyses of 20 basalts and andesites from four different locations of the Rajmahal Traps, three mafic–ultramafic dikes of the Bokaro coal fields southwest of the Rajmahal Traps, and 56 alkalic, ultramafic, and carbonatitic rocks from four alkalic complexes around the Bengal Basin (Fig. 2). We use the geochemical data to document this widespread volcanism in eastern India by the Kerguelen plume that may approach an area of one million km2 (Fig. 1). Using the geochemical results and their interpretation we evaluate the petrogenesis of the tholeiitic traps and associated alkalic–mafic–ultramafic and carbonatitic magmatism in the context of Kerguelen plume volcanism. 2. GEOLOGICAL BACKGROUND OF ROCKS ANALYZED 2.1. Rajmahal Traps The Rajmahal basalts exposed in the Rajmahal hills (Figs. 2b and A1–A4) of eastern India cover an area of

48

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Fig. 2. (a) Geological map showing structural features and locations of the Rajmahal and Sylhet Traps in and around the Bengal Basin including borehole sites where Rajmahal age basalts have been encountered (Sengupta, 1966; Baksi et al., 1987; Ray et al., 2005). Alkalic and ultrabasic rocks in Samchampi, Sung, Sikkim, and the Bokaro dykes are related to the Rajmahal–Sylhet Traps. Associated volcanic rocks are also reported from the Bangladesh part of the basin (personal communications from Prof. A.T.M. Fazlul Haq, Dhaka University, Department of Geology, 2010). (b) Map showing the distribution of volcanic and sedimentary rocks in the Rajmahal Hills and surrounding areas (Kent et al., 1997), and locations of sample sites.

4100 km2, with exposed thickness of up to 230 m over Gondwana Supergroup sediments to the west. To the east, down faulting of the basement causes the sequence to be beneath the thick sediments of the Bengal Basin shed from the Himalayas. Exploration drilling indicated at least 332 m thick basaltic lavas underlying much of the Bengal Basin (Sengupta, 1966). 40Ar/39Ar data in basalts from these drillings suggest ages of 117 Ma (Baksi, 1995). More recent 40 Ar/39Ar results (Kent et al., 2002) from the Rajmahal hills and the Sylhet basalts (Ray et al., 2005) are consistent with an 118 Ma age for the magmatic activity, contemporaneous with the final stage of volcanism in ODP site 1136 in the southern Kerguelen Plateau (119–118 Ma). Geochemical data from four different regions of the Rajmahal hills (Figs. 2b and A1–A4) either excluded or sparsely sampled in previous studies are presented here. There are six individual lava flows in the Sahibganj area (Fig. A1) and the samples analyzed of these flows are from the Ambadihi, Rangamatia and Adro Bedo traverses of this area. Four basaltic flows of the Tinpahari location (Fig. A2) and two samples of an andesitic dike intrusive into these flows were also analyzed. The Harinsingha location (Fig. A3) in the southernmost location of the Rajmahal hills show two tholeiitic flows, separated by an inter-trappean bed, in contact with the Gondwana (Dubrajpur) sandstone. Four samples

were analyzed from this location, two of which are in contact with the sediments. The Taljhari location (Fig. A4) has three recognizable basaltic–andesitic and andesitic–dacitic flows from which five samples were analyzed. 2.2. Bokaro dikes The Gondwana coal-bearing sediments of Bihar–West Bengal (Fig. 2) were intruded by ultrabasic dikes and sills (Sarkar et al., 1980). We had determined a 40Ar/39Ar age of phlogopites from a lamproite dike in the eastern Bokaro coal field (Courtesy of P.R. Renne, Berkeley Geochronology Center) that gave a precise age of 114.4 ± 0.1 Ma indicating this intrusive event to be younger than the 118 Ma tholeiitic flood volcanism in the Rajmahal Hills. Mineralogically, these dikes and sills contain olivine, phlogopite, apatite, aegirine, amphibole (usually K-richterite), carbonate, spinel, and perovskite, characteristic of lamproites. These sills show varying mineralogy by differentiation, and may be termed minette lamprophyre and lamproite in gradational contact (Rock et al., 1992). In addition to the ultrabasic intrusions, the Gondwana sediments of the region host many diabase dikes and sills; three diabase dikes and a lamprophyre of the Bokaro field (Fig. 2) were analyzed in this study.

Table 1 Trace element concentrations of the Rajmahal Trap basalts and Bokaro dikes of this study. Analytical uncertainties are less than 5% for all the elements analyzed and usually less than 2% for the REEs. Sample

1.5

2.8

2.10

3.8

3.12

3.14

5.3

5.4

5.6

5.8

5.10

5.11

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Andesite dike

Andesite dike

Rb Ba Sr Pb La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Y Th U Zr Hf Nb Ta Sc

16.7 470 349 5.53 17.7 39.5 5.19 22.5 5.73 1.89 6.12 0.97 5.84 1.24 3.33 0.49 3.11 0.45 33.0 3.19 0.43 158 4.02 9.33 0.64 32.2

16.5 253 312 4.29 14.6 32.5 4.41 19.7 5.01 1.76 5.31 0.84 5.18 1.10 2.90 0.42 2.72 0.39 29.1 2.37 0.34 134 3.41 7.85 0.56 31.7

18.4 239 367 3.91 12.9 31.4 4.26 18.8 5.15 1.75 5.56 0.86 5.14 1.09 2.86 0.43 2.75 0.39 30.2 2.16 0.33 133 3.31 8.73 0.54 33.1

22.0 166 332 3.76 12.0 27.0 3.65 16.6 4.41 1.60 4.72 0.75 4.54 0.98 2.57 0.38 2.40 0.35 27.0 2.04 0.29 114 2.87 6.56 0.46 30.7

17.6 186 332 4.47 12.2 27.8 3.76 16.7 4.41 1.58 4.70 0.75 4.56 0.98 2.60 0.38 2.46 0.35 26.5 2.14 0.31 119 2.95 6.98 0.47 31.4

17.6 186 334 3.75 12.1 27.4 3.72 16.4 4.33 1.58 4.68 0.74 4.48 0.96 2.56 0.37 2.41 0.35 26.4 2.04 0.29 116 2.89 6.88 0.47 31.0

25.1 257 335 5.21 15.7 34.9 4.56 19.8 5.10 1.77 5.76 0.91 5.61 1.17 3.14 0.46 2.94 0.42 31.1 2.65 0.38 141 3.57 7.89 0.52 32.4

7.04 68 230 1.49 7.05 17.1 2.65 12.9 4.36 1.61 5.06 0.85 5.24 1.12 2.98 0.44 2.84 0.40 29.9 0.69 0.14 94 2.34 6.16 0.43 37.1

4.97 65 231 1.89 6.64 16.9 2.53 12.2 3.97 1.52 5.05 0.83 5.18 1.12 2.95 0.43 2.73 0.40 30.0 0.64 0.14 95 2.43 5.54 0.40 38.7

32.3 312 330 6.16 18.8 40.9 5.32 22.8 5.88 1.96 6.36 0.99 6.06 1.30 3.45 0.50 3.13 0.46 35.1 3.38 0.47 160 3.91 9.56 0.66 33.0

11.3 202 353 4.73 14.5 31.0 4.24 18.4 5.01 1.80 5.14 0.83 5.04 1.08 2.86 0.43 2.75 0.40 29.9 2.43 0.34 123 2.94 8.23 0.50 33.5

19.3 234 353 5.15 16.4 35.9 4.78 21.2 5.39 1.83 5.93 0.93 5.70 1.21 3.25 0.47 3.02 0.44 32.9 2.70 0.39 142 3.48 8.45 0.56 35.4

Sample

25

1

6a

Dubrajpur sst.

4.1

4.2

4.11

4.14

4.15

kJ/1

kJ/24

II/L3 S.2

Basaltic flow

Basaltic flow

Basaltic flow

Gondwana sandstone

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Andesitic flow

Lamprophyre dike

Diabase dike

Diabase dike

6.35 107 247 2.73 8.02 19.6

3.19 139 294 1.73 6.64 16.3

2.16 126 240 1.94 7.68 19.4

31.0 831 82 7.97 9.62 16.3

32.7 357 319 6.97 21.1 46.1

23.3 515 414 7.26 19.0 41.9

7.49 161 246 4.61 9.50 21.1

9.76 135 246 3.33 9.05 22.1

1189 939 672 25.1 92 205

103 3965 1652 42.6 427 807

16.2 317 203 2.86 15.8 38.2

12.2 150 326 2.11 8.69 20.2

Sahibganj

Tinpahar

Rb Ba Sr Pb La Ce

Taljhari

Bokaro

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Harinsingha

49

5.36 24.3 6.71 2.18 7.79 1.23 7.52 1.57 4.13 0.60 3.75 0.53 41.3 2.13 0.48 232 5.59 14.9 1.03 42.7 80 278 39.1 9.75 22.9 2.65 11.2 1.81 4.12 0.50 2.88 0.32 49.4 32.3 6.54 1403 25.8 161 7.59 22.7 23.5 90 15.3 3.76 10.2 1.28 6.65 1.24 2.92 0.37 2.21 0.25 26.7 13.8 3.12 1064 23.8 67 3.66 7.90 3.32 16.0 5.06 1.82 6.33 1.03 6.38 1.34 3.54 0.51 3.25 0.47 35.1 0.90 0.20 123 3.24 7.17 0.59 38.5 3.32 15.9 4.85 1.82 5.75 0.95 5.78 1.24 3.22 0.47 2.99 0.43 32.6 0.86 0.20 117 3.11 6.99 0.53 38.2 5.51 23.8 5.92 1.93 6.38 1.01 6.16 1.30 3.46 0.51 3.28 0.49 34.3 3.69 0.49 160.3 4.04 10.3 0.73 33.5 6.34 26.9 6.91 2.15 7.41 1.19 6.97 1.47 3.95 0.59 3.78 0.52 39.8 4.36 0.52 172 4.32 12.1 0.69 34.7 2.24 7.87 1.38 0.39 0.99 0.15 0.97 0.20 0.56 0.09 0.61 0.09 4.82 1.86 0.64 74 1.96 3.15 0.23 1.95 2.95 14.2 4.68 1.73 5.73 0.94 5.98 1.29 3.40 0.50 3.20 0.48 34.4 0.89 0.18 118 2.99 5.82 0.43 41.3 2.41 11.7 3.73 1.47 4.64 0.76 4.82 1.03 2.74 0.39 2.48 0.37 26.8 0.67 0.14 91 2.38 4.95 0.36 34.6 2.87 14.1 4.53 1.65 5.65 0.92 5.85 1.25 3.34 0.48 3.05 0.45 32.3 0.87 0.17 115 3.02 5.93 0.43 37.1 Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Y Th U Zr Hf Nb Ta Sc

Petrological studies inducing petrography and mineralogy were conducted both at the University of Rochester and the Geological Survey of India. Rock names were assigned to the samples based on these studies.

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

3.06 14.5 4.70 1.71 5.33 0.88 5.41 1.13 2.99 0.44 2.78 0.39 29.6 1.09 0.20 188.3 4.25 6.85 0.43 35.2

50

2.3. Sikkim alkalic intrusives The Rangit tectonic window is in the Main Central Thrust in the Himalayan foothills, near Darjeeling (Fig. A5). This window is made up mostly of high-grade crystalline rocks. At the southern end of the window is a linear east–west exposure of the Gondwana sediments with lamprophyre and lamproite dikes cutting across these sediments. These dikes are exposed in the central portion of the window hosting a series of syenitic dikes. Recent kinematic modeling showed that the Rangit window experienced shortening of 125 km since the collision of the Indian and Asian plates 50 million years ago (Mitra et al., 2010). This shortening implies that these dikes were emplaced at least 125 km to the north of the current site, extending the initial aerial extent of Rajmahal volcanism to the north (Figs. 1 and 2). We have selected 13 of these dike samples from the Sikkim–Darjeeling area and seven ultra-potassic syenites that intruded into the Daling group metasediments now found in a nappe overlying the Gondwana rocks (Fig. A5). These rocks are ultra-potassic syenites, minettes, nephelinites and lamproites comprising unusual alkalic rock associations of any flood basalt province, including the Parana, Deccan and Siberian provinces. 2.4. Sung Valley alkalic complex The Sung Valley Alkaline Igneous Complex, 26 Km2 in size is intrusive into the Proterozoic quartzites and phyllites of a block in the Shillong plateau in northeast India (Fig. A6), in an uplifted Precambrian basement, bordered by the Dauki fault to the South and the Brahmaputra graben to the north. A N–S trending lineament cuts across the Shillong plateau and contains several alkalic intrusive bodies including the Sung complex (Kumar et al., 1996). This complex gives a Pb–Pb age of 134 ± 20 Ma (Veena et al., 1998) and a more precise U–Pb perovskite age of 115 ± 5.1 Ma (Srivastava et al., 2005). An oval pyroxenite body forms the main rock type of the complex, intruded by peridotites, ijolites, carbonatites and syenites (Veena et al., 1998). The pyroxenites are composed of coarse grained diopsidic augite with minor amounts of phlogopite, titanite and apatite. A medium grained variety contains diopside and aegirine augite with minor K-feldspar, titanite and apatite. Stock-size peridotites are emplaced in the northern part of the complex. Ijolites are the third most abundant rocks, intruding the pyroxenites as a ring dike. Carbonatites of the complex occur mainly as dikes and cone sheets. Minor felspathic syenite veins and dikes cut pyroxenites, ijolites and the Precambrian quartzites (Fig. A6). We analyzed 18 samples of various lithologies from this complex. 2.5. Samchampi alkalic complex The Samchampi complex (Fig. A7) shows similarities to the Sung complex. It occurs in the Mikir hills, northeast of the Shillong plateau as a relatively circular stock-like body, cored by a titano–magnetie–perovskite rock. Syenites of

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

variable mineralogies occur as discrete lenses including nepheline syenites. Pyroxenites occur as small patches, composed of aegirine–augite and interstitial potassium feldspars (Hoda et al., 1997). An ijolite–melteigite suite occurs to the north in a ring dike pattern, much like the Sung complex. Aegirine–augite, nepheline, biotite and carbonate minerals constitute the primary minerals of the ijolite– melteigites with apatite and sphene as accessory phases. Fifteen samples were analyzed, including alkali pyroxenites, syenites, ijolites, and sovites. 2.6. Barpung alkalic complex The Barpung alkalic complex is very similar to Samchampi and is located to the southeast of Samchampi (Fig. A8). This body, intrusive into amphibolites is composed of pyroxenites and aegirine-bearing syenites. The Barpung complex contains the only monomineralic potassium-feldspar syenites among all the syenites associated with the Rajmahal–Sylhet volcanism (Weaver, 2000). Ten pyroxenites and syenites were analyzed in our study. 3. ANALYTICAL METHODS Whole rock samples were powdered using a spex alumina ball, 0.5 kg size rock samples were broken into chips, washed in an ultrasonic bath with mildly acidified (0.2 M HCl) deionized water, and dried, and finally 20 g of these chips were powdered in an alumina ball mill. All the trace element and isotopic analyses reported here were carried out at the University of Rochester. Trace element concentrations were measured using an ICPMS (Thermo elemental X-7) at the University of Rochester using established procedures in our laboratory (Hannigan et al., 2001). BCR-2 was used as a standard, AGV-2 and BHVO-2 rock standards were run as unknowns to estimate error (Appendix Table T1), usually less than 5% (2r error) for most of the trace elements, commonly less than 2% for the rare earth elements (REEs). Nd–Sr–Pb-isotopic ratios were measured using a multicollector thermal ionization mass spectrometer (VG-Sector) for which 100–200 mg of the powdered rock samples were dissolved in HF–HNO3 and HCl. Nd–Sr-isotopes were measured using the procedures established for our laboratory (Basu et al., 1990). Measured 87Sr/86Sr ratios were normalized to 86Sr/88Sr = 0.1194. Uncertainties for the measured 87Sr/86Sr ratios were less than ±0.00004 (2r of the mean). The SRM-987 Sr standard analyzed during the course of this study yielded 87Sr/86Sr = 0.71024 ± 0.00002 (2r, n = 6). Measured 143Nd/144Nd ratios were normalized to 146Nd/144Nd = 0.7219 with uncertainties less than ±0.00003 (2r of the mean). La Jolla Nd-standard yielded 143 Nd/144Nd = 0.51186 ± 0.00003 (2r, n = 5). Initial eNd values were calculated using present day Bulk Earth 143 Nd/144Nd = 0.512638 and 147Sm/144Nd = 0.1967 (Jacobsen and Wasserburg, 1984). Pb-isotopes were measured using a silica-gel technique of our laboratory (Sharma et al., 1992). Filament temperature during Pb-isotope measurements was monitored; raw ratios were calculated as weighted averages of the ratios measured at 1150, 1200

51

and 1250 °C, respectively. The reported Pb-isotopic data were corrected for mass fractionation of 0.12 ± 0.03% per a.m.u. based on replicate analyses of the NBS-981 Pb standard. Estimated errors are less than 0.05% per a.m.u. Laboratory procedural blanks were 400 pg for Sr and 200 pg for Nd and Pb. 4. RESULTS AND DISCUSSION In this section we present the geochemical results of our study of basaltic lavas from the Rajmahal Traps, dikes from Bokaro, and alkaline rocks from the Sung, Samchampi, Barpung and Sikkim alkaline complexes. The data comprise multiple trace element concentrations including the rare earths, and the isotopic compositions of Nd, Sr, and Pb. The analytical results are presented in Tables 1–4 and Figs. 3–11. Basaltic lavas from four locations of the Rajmahal Traps are compared with data obtained from previous work on Early Cretaceous (117 Ma) volcanic rocks related to the Kerguelen plume activity. From the discussions we conclude that the volcanics may have 80–100% of the primitive Kerguelen plume component with up to 20% contaminants from the lower continental crust and mantle lithospheric source that we identified as granulites of the Eastern Ghats Belt in India (Fig. 1). We also provide a discussion and analyses of the geochemical data of the ultrabasic–carbonatitic rocks (consisting of pyroxenites, nephelinites, lamproites, sovites, melteigite, uncompahgrites and carbonatites), and syenites-ijolites from four different complexes distributed north, northeast and northwest of the main Rajmahal volcanic province that were possibly synchronous with the main pulse of flood volcanism. We propose a petrologicevolutionary model for all these varied rock types found in a disparate but large area that occur as vestiges of the Kerguelen plume volcanism in northeastern India. 4.1. Trace Element Geochemistry of Rajmahal lavas and Bokaro dikes Elemental abundances normalized to primitive mantle (Evensen et al., 1978) for the four newly analyzed Rajmahal sections (21 samples) and the Bokaro dikes are compared to the two compositional groups of Rajmahal basalt recognized by Kent et al. (1997) (Fig. 3). Rajmahal lavas from Harinsingha, some basalts from Tinpahar and Taljhari, and two diabase dikes from Bokaro in the west show remarkably flat primitive mantle-normalized trace element patterns, except for a few incompatible elements, such as Ba and Pb (Fig. 3). The RJ I lavas were shown to be similar to lavas from ODP site 1138 in the Indian Ocean (Neal et al., 2002). Consideration of the other trace elements of these Harinsingha, Tinpahar and Taljhari rocks (Fig. 3) that show relatively low Rb, U and Pb and slightly higher Ba, and Sr also indicate their similarity with RJ I basalts, as well as the least contaminated Sylhet Traps lavas (Ghatak and Basu, 2011). Overall, these rocks are distinctly different from both NMORB and upper continental crust in trace element patterns (Fig. 3).

52

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Fig. 3. Multiple trace element concentrations normalized to primitive mantle for the Rajmahal Trap and Bokaro sections: (a) Sahibganj; (b) Tinpahar; (c) Taljhari; (d) Harinsingha; and (e) Bokaro dikes. All the lavas and dikes are compared to Rajmahal Group I (RJ I) and Group II (RJ II) basalts (Kent et al., 1997). The lava with flat REE patterns (a–e) display small variations in their elemental concentration patterns. In contrast, the lavas with LREE enrichment (a–e) also display enrichment in incompatible elements relative to primitive mantle, characteristically in Rb, Ba, Sr, and Pb, while being depleted in general in Nb–Ta. (f) Average REE patterns from various Kerguelen Plateau ODP sites (Frey et al., 2000; 2002; Ingle et al., 2002; Neal et al., 2002), and continental clasts from Site 1137 (Ingle et al., 2002) are shown for comparison.

Fig. 4. (a and b) Primitive mantle normalized Nb/Zr, and La/Nb vs. Th/Nb for the Rajmahal lava flows and dikes of this study compared with various Rajmahal Group I and II basalts (RJI and RJII respectively; Kent et al., 1997), Bunbury basalts (Storey et al., 1992; Frey et al., 1996), South and Central Kerguelen Plateau lavas (sites 747, 749, 750, 1136, 1138, 1141, 1142) (Mahoney et al., 1995; Frey et al., 2000, 2002; Neal et al., 2002), and mantle reservoirs NMORB, lithospheric mantle (LM), upper crust (UC) and average continental crust (CC) (Neal et al., 2002). Also shown here are lavas from the Sylhet Traps (Ghatak and Basu, 2011).

The relatively flat REE patterns of the group I Rajmahal Traps (Fig. 3) rocks has been attributed by previous workers (e.g. Baksi et al., 1987; Kent et al., 1997) to have formed by decompressional melting of the asthenosphere and its passive upwelling through the rifted margin of eastern India. We suggest an alternate scenario in Section 5.1, based on our geochemical modeling. Based on this model the relatively primitive nature of these lavas with flat REE patterns and eNd(I) values of 0.6–2.8 are explained by 18% melting of a relatively primitive component of the Kerguelen plume. The primitive nature of these lavas is also seen is other trace elemental plots (Fig. 4, dotted field) where they cluster tightly around or near the primitive mantle fields. Most of the lavas with flat REEs also overlap with or lie close to the field of 1138 basalts and the most primitive Sylhet basalts (Fig. 4) that are considered to be sourced in the Kerguelen plume-head (Neal et al., 2002; Ghatak and Basu, 2011). In contrast to the basalts with flat REE as discussed above, the remaining lavas from Tinpahar, Taljhari and Sahibganj show considerably different REE and multiple element patterns (Fig. 3) with higher LREE enrichment, steeper HREE slopes (Fig. 3; LaN/SmN = 1.8–2.0; LaN/ YbN = 3.6–4.1) and negative eNd(I) ( 2.4 to 7.7) values

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

53

Fig. 5. (a) Ce/Nb vs. Th/Nb plot for the Rajmahal Traps samples are shown in comparison with Iceland plume related basalts (He´mond et al., 1993), and Chilka Granulites (CG, Chakrabarti et al., 2011). Average primitive mantle, depleted mantle, E-MORB and continental crust are also shown for comparison (Sun and McDonough, 1989; Rudnick and Fountain, 1995; Taylor and McLennan, 1995). Note some of the Sylhet lavas show a trend towards E-MORB. This E-MORB contamination is distinctly absent for the Rajmahal lavas. (b) Nb/Y vs. Zr/Y ratios of the Rajmahal lavas and dikes of this study and the Sylhet Traps of our previous study (Ghatak and Basu, 2011) compared to various ODP sites related to the Kerguelen Plume. The DNb line distinguishes plume and non-plume sources well (Fitton et al., 1997). Notice that all the Rajmahal–Sylhet Traps lavas fall within the field of “plume sources” with a trend from primitive mantle (PM) towards ancient continental fragments from 1137. Data sources are as in Fig. 4. (c) Plot of eNd(I) vs. Ce/Pb, with modeled mixing lines between continental crust (CC; Taylor and McLennan, 1995) and bulk silicate Earth (BSE). Also plotted are fields of depleted mantle (DMM; Salters and Stracke, 2004), Indian Ocean E-MORB (Mahoney et al., 2002), early differentiated Earth (EDR; Boyet and Carlson, 2006), and primitive Kerguelen plume (Ghatak and Basu, 2011). Mixing line is modeled using mixing equations of DePaolo (1988).

(Table 3). These geochemical characteristics are also commonly observed in Rajmahal Group II basalts (Fig. 3), some Sylhet lavas that are inferred to be contaminated by the Eastern Ghat granulites of India (Ghatak and Basu, 2011), and the granitoid rocks recovered from drill core site 1137 in the Indian Ocean (Ingle et al., 2002) on the Kerguelen Plateau (Fig. 1). Interestingly, these rocks show strong negative Nb–Ta anomalies (Fig. 3) characteristic of continental crustal rocks. An andesitic lava flow from Taljhari and a sandstone from Harinsingha are distinctly different in trace element

geochemistry from the rest of the Rajmahal and Bokaro lavas (Fig. 3c, d); the sandstone from Harinsingha, in contact with the basaltic flows shows depleted heavy REE with enrichment (10 Chondrite) in La, Ce, Pr and Nd, indicating that these Lower-Middle Jurassic Gondwana sediments are not likely contaminants of the lavas. The Taljhari andesite (Fig. 3c), has a relatively higher LREE (LaN/ SmN = 3.8) and a much steeper overall REE slope (LaN/ YbN = 28), possibly due to garnet in its residual parent source. The diabase dikes are strikingly similar to the nearly flat REE patterns in Group I basalts from the Rajmahal

54

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Fig. 6. (a) 208Pb/204Pb(I) vs. 206Pb/204Pb(I) and (b) 207Pb/204Pb(I) vs. 206Pb/204Pb(I) plots for the Rajmahal Traps and Bokaro dikes compared with Kerguelen Plateau basalts, Bunbury basalts, Naturaliste Plateau lavas, Rajmahal Traps and Chilka Granulites at the 117 Ma age of eruption of these traps. l values of 8.3, 8.4, and 8.5 are also shown in (b) where l = 238U/204Pb. The field of ancient continental crustal clasts at ODP site 1137 (Ingle et al., 2002) uses present day values. Abbreviations used: BB-C – Bunbury, Casuarina; BB-G – Bunbury, Gosselin; N – Naturaliste Plateau; CG – Chilka Granulites. Data sources: Indian MORB (Mahoney et al., 1992); Sylhet Basalts (Ghatak and Basu, 2011); Naturaliste Plateau (Mahoney et al., 1995); Elan Bank 1137, and ancient continental fragment from 1137 (Ingle et al., 2002); Chilka Granulites (Chakrabarti et al., 2011). The field of Indian MORB also includes the Southeast Indian Ridge (Mahoney et al., 2002). Also shown are the domains of the BSE – bulk silicate Earth; NHRL – Northern Hemispheric Reference Line; DM – depleted mantle (Hart and Zindler, 1989); and upper and lower continental crust (Zartman and Doe, 1981). Other abbreviations and data sources are as in Fig. 4.

Traps (LaN/SmN = 1.2; LaN/YbN = 2.7) with LREEs 20 times enriched than chondrites, with no Nb–Ta anomaly and a slightly steeper HREE slope (Fig. 3e). For comparison, average primitive mantle normalized trace element patterns of various ODP sites from the Kerguelen Plateau and continental clasts from ODP site 1137 are shown in Fig. 3f. Some HFSE and REE ratios are plotted in Figs. 4 and 5a, b. Abundance ratios of HFSE and REE are usually not affected by alteration, therefore these elements and their ratios are considered to be excellent petrogenetic indicators of source and degree of contamination. These plots are also important in distinguishing between enriched mantle (EMI, EM-II), Depleted mantle (DM), and primitive mantle sources. The continental crust is almost always depleted

in Nb, making Nb important in distinguishing between continental crust and other sources. Primitive mantle normalized trace element ratios of Hf, Zr, Nb, Th, and La for the Rajmahal basalts and Bokaro dykes are compared with some ODP site rocks, Bunbury basalts, Rajmahal–Sylhet lavas, average continental crustal estimates, and various mantle reservoirs in Fig. 4a, b. In all the plots of Fig. 4, the Rajmahal lavas and Bokaro dikes fall in two clusters, one tightly grouped near the Primitive mantle and enclosed by the dotted line, and the other trending towards the continental crust. This clustering of the Rajmahal and Bokaro rocks is also seen in Ce/Nb vs. Th/Nb (Fig. 5a) and Nb/Y vs. Zr/Y plots (Fig. 5b). The similarity of primitive mantle like basalts with RJI and the remaining flows with RJII is

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

55

Fig. 7. (a) Initial eNd vs. 206Pb/204Pb(I) and (b) 87Sr/86Sr(I) vs. 206Pb/204Pb(I) for the Rajmahal lavas and Bokaro dikes compared with Indian MORB, including Southeast Indian Ridge data of Mahoney et al. (2002), Bunbury basalts, Rajmahal basalts, Kerguelen Plateau basalts, and possible crustal contaminants such as Chilka Granulites (CG) and ancient continental clasts from ODP site 1137. Group II kimberlites from the Rajmahal flood basalt province (Km) are from Kumar et al. (2003). Other data sources as in Figs. 4 and 6. Additional data sources: ODP sites 738 (Alibert, 1991).

Fig. 8. Initial eNd vs. 87Sr/86Sr at 117 Ma for the Rajmahal Traps and Bokaro dikes of this study with fields of Rajmahal Groups I and II lavas, Bunbury basalts, Kerguelen Plateau basalts, and Eastern Ghat Belt (EGB) granulites (Rickers et al., 2001). Sylhet Traps data from the CH and MB sections (Ghatak and Basu, 2011) are also plotted here. Fields of Rajmahal Groups I and II (Kent et al., 1997) overlap with Nd–Sr data of Rajmahal basalts reported by Mahoney et al. (1983), Storey et al. (1992), and Baksi (1995). The field of Indian MORB also includes the Southeast Indian Ridge (Mahoney et al., 2002). Data sources for other fields and abbreviations are as in Figs. 4, 6 and 7. Partial mixing curves resulting from the modeling of the Nd–Sr data of a relatively primitive Kerguelen plume (P) as elaborated in Fig. 13 are also shown here, with mixing between a modeled plume-melt and two EGB granulites as end members.

again indicated in Fig. 4b. A strong trend towards the lower crustal/lithospheric mantle derived Chilka Granulites (CG) is observed for the basalts in their Ce/Nb vs. Th/Nb relationship (Fig. 5a). Lower continental crust as a major contaminant has been shown for continental flood basalts (e.g. Peng et al., 1994) and for several ODP site lavas of the South and Central Kerguelen Plateau (Frey et al., 2002). Rajmahal lavas are shown in an eNd(I) vs. Ce/Pb plot (Fig. 5c) and compared with end member reservoirs of continental crust (CC; Taylor and McLennan, 1995), depleted mantle (DMM; Salters and Stracke, 2004), average Indian Ocean E-MORB (Mahoney et al., 2002), early differentiated bulk Earth (EDR) as defined by Boyet and Carlson (2006), bulk silicate chondritic Earth, and the primitive Kerguelen plume-head composition proposed by Ghatak and Basu (2011). Ce/Pb ratios are distinct indicators of bulk silicate Earth and continental crustal sources (Hofmann et al., 1986). It is interesting to note that the Rajmahal lavas fall along the mixing line between bulk silicate Earth and continental crust and show 20% crustal contamination. The absence of N-MORB or E-MORB components for the Rajmahal basalts is also indicated in Fig. 5c where none of the lavas trends towards these reservoirs. It is interesting that some of the most primitive Rajmahal lavas with flat REE patterns and relatively low positive eNd(I) values fall close to Ce/Pb = 11, very near the primitive mantle value of 10. In the trace element plots discussed above (Figs. 3–5) the absence of an N-MORB contaminant

56

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Fig. 9. Multiple trace element concentrations normalized to Primitive Mantle for the ultrabasic–carbonatitic rocks (left column) and the syenites and ijolites (right column) from the four alkalic complexes of this study.

is strongly indicated. An E-MORB type contaminant was suggested for some of the Sylhet basalts based on their LREE enrichment and high eNd(I) values of >+3 (Ghatak and Basu, 2011). E-MORB contamination for some of the Sylhet-CH basalts is possible in consideration of various trace element ratios in Figs. 4 and 5. However, the Rajmahal lavas do not show this trend towards E-MORB (Figs. 4 and 5). The Rajmahal lavas of this study also lack higher than Sylhet eNd(I) values (Fig. 8 and Table 3) and LREE enrichment (Fig. 3) seen in the Sylhet lavas, suggested to be contaminated by E-MORB (Ghatak and Basu, 2011). Thus N-MORB and E-MORB can be ruled out as possible contaminants for these Rajmahal lavas and the most likely contaminant is the lower crust/mantle lithosphere-derived Chilka Granulites of the Eastern Ghats Belt (Fig. 1). This will be further discussed in Section 4.2.

4.2. Nd–Sr–Pb isotopes in Rajmahal lavas and Bokaro dikes Nd–Sr–Pb isotopic ratios of all the basaltic and andesitic lavas and Bokaro dikes are plotted in Figs. 6–8 along with the relevant fields as indicated in the figure captions. The Nd–Sr–Pb data for the basalts and andesites are estimated at 117 Ma from the 40Ar–39Ar age for these tholeiites (Kent et al., 2002). Estimates of initial isotope ratios due to age corrections are described in Note 2 in Supplementary materials. Initial 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb of the Rajmahal lavas and Bokaro dikes (Fig. 6) show ranges of 17.4–18.3, 15.5–15.7, and 37.1–40.4, respectively (Table 3). The Pb-isotopic ratios of the Rajmahal lavas, the diabase and lamprophyre dikes, and the fields of south and central Kerguelen Plateau sites and Rajmahal Groups I and II are

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

57

Fig. 10. (a) 208Pb/204Pb(I) vs. 206Pb/204Pb(I) and (b) 207Pb/204Pb(I) vs. 206Pb/204Pb(I) plots for the alkalic rocks of this study compared with Bunbury basalts and the Rajmahal Traps at the 117 Ma. l values of 8.3, 8.4, and 8.5 are also shown in (b) where l = 238U/204Pb. Also shown are the domains of the BSE, NHRL, DM and average lower continental crust. Data sources and abbreviations are as in Figs. 4 and 6.

shown along with the 4.45 Ga Geochron (at l = 8.3–8.5) in the 206Pb/204Pb vs. 207Pb/204Pb plot (Fig. 6). Most of the Rajmahal lavas are clustered around the fields of the Naturaliste Plateau, Bunbury basalts, Rajmahal–Sylhet Traps and Kerguelen Plateau drill core sites 1136, 1138, 1141, and 1142 (Fig. 6) and have a nearly vertical trend along the bulk silicate Earth (BSE) field (Fig. 6b). The correspondence of the Rajmahal lavas and Bokaro dikes in Pb-isotopes with Rajmahal Group I and II basalts is consistent with inferences from their trace element patterns as discussed above in Section 4.1. A few lavas fall close to the field of average lower crust in Fig. 6, close to the field of Site 750 on the Kerguelen Plateau which is considered contaminated by the lower continental crust (Frey et al., 2002; Neal et al., 2002). Other lavas analyzed in our study show higher values of both 207Pb/204Pb(I) (Fig. 6b) and 208Pb/204Pb(I) (Fig. 6a) falling closer to the fields of Chilka Granulites (CG), considered as derived from the subcontinental Indian mantle lithosphere (Chakrabarti et al., 2011). These lavas with higher 207Pb/204Pb(I) and 208Pb/204Pb(I) ratios also lie near the most contaminated Kerguelen Plateau basalts from drill core 738 (Fig. 6, Frey et al., 2002). The field for the ancient continental crustal clasts from Site 1137 of the Elan Bank in Fig. 6 represents conglomerates, sand-

stones and granulites that are of unequivocal continental origin (Ingle et al., 2002). A few samples falling close to the field of Indian MORB (Fig. 6a) overlap with Kerguelen Plateau basalts, derived from the Kerguelen plume (e.g. Mahoney et al., 1992; Frey et al., 1996, 2000; Ingle et al., 2002; Neal et al., 2002). Furthermore, significant N-MORB source mantle contribution in these rocks is already eliminated by the trace element data (Section 5.1) and on the basis of Nd–Sr–Pb isotopic data as discussed below. We also note from Fig. 6 that the Rajmahal lavas plot far removed from the average upper continental crust in their Pb-isotopic compositions. An important conclusion that can be drawn from the Pb isotopes (Fig. 6) is that the lithospheric contaminants in the RBST basalts as well as the Kerguelen Plateau basalts are clearly of lower crustal – lithospheric mantle affinity without any upper crustal component. Rb–Sr and Sm–Nd isotope systematics data are reported in Table 3. Initial eNd and 87Sr/86Sr vs. 206Pb/204Pb(I) at 117 Ma are shown in Fig. 7a, b along with possible crustal and lithospheric contaminants. A general correspondence of our new Rajmahal data with the southern and central Kerguelen Plateau, Bunbury and Rajmahal–Sylhet Traps basalts is notable (Fig. 7a, b). The distinction between the

58

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Fig. 11. Initial eNd vs. 87Sr/86Sr at 117 Ma for the alkalic rocks of (a) Sung, (b) Samchampi, (c) Barpung, and (d) Sikkim compared with the primitive Kerguelen plume component (Neal et al., 2002; Ghatak and Basu, 2011), Indian MORB (Mahoney et al., 2002), Groups I and II kimberlites (Chakrabarti et al., 2011), Krishna lamproites (Chakrabarti et al., 2007), Gaussberg lamproites (Murphy and Collerson, 2002), and Australian lamproites (Mitchell and Bergman, 1991). Additional data sources and abbreviations as in Figs. 4, 6 and 11.

lavas and dikes with flat RJ I like REE patterns and those with LREE enriched RJ II-like REE patterns (Fig. 3) is well differentiated in the plot of eNd(I) against 206Pb/204Pb(I) (Fig. 7). A kimberlite dike (Km field in Fig. 7) analyzed from the Bokaro area (Kumar et al., 2003) is distinctly different from the bulk of the Rajmahal lavas with flat REE patterns, and ODP sites 1138, 1141, and 1142 basalts that are considered the least contaminated lavas of the Rajmahal Traps and the Kerguelen Plateau, respectively (Fig. 7a, b). Sites 1138, 1141, and 1142 basalts are likely the most representative of the main component of the Kerguelen plume in the middle Cretaceous (Neal et al., 2002). These lavas are geochemically similar to the RJ I-like lavas of this study, showing higher eNd (0–1) than those ( 0.2 to 2.9) proposed by Weis et al. (1993) to be the “pristine” Kerguelen plume. In contrast with the RJ I like lavas, the LREE-enriched RJ II-like lavas lie closer to the CG and contaminated basalts from Kerguelen site 738 in the Nd–Sr–Pb isotopic plots (Fig. 7), falling distinctly away from the uncontaminated fields of RJ I, 1138, 1141, and 1142. An elongate array, which we interpret as a contamination or mixing trend, is seen clearly in Fig. 7b, with the LREE enriched basalts showing the most radiogenic Sr-isotopic ratios. The initial Nd, Sr isotopes of the Bokaro dikes and Rajmahal basalts range from +2.8 to 7.7 and 0.70347 to 0.70804, respectively. In Nd–Sr isotopic plot (Fig. 8) all the basaltic rocks and the Bokaro diabase and lamprophyre dikes show affinity with the Rajmahal Groups I and II lavas, Sylhet Traps, Kerguelen Plateau basalts, and the field of the primitive Kerguelen plume component (Neal et al., 2002; Ghatak and Basu, 2011). Based on these correlations, a Rajmahal–Sylhet–Kerguelen Plateau connection is strongly indicated. All the basalts with flat REE patterns corresponding with Rajmahal Group I lie in the quadrant with less-radiogenic Sr and positive eNd values in the Nd– Sr plot (Fig. 8). In contrast, the LREE-enriched lavas fall near fields of RJ II, drill core 738, and ancient continental

fragments in core 1137, both from the Kerguelen Plateau (Fig. 8 which are all considered as contaminated by the lower continental crust (Frey et al., 2002; Ingle et al., 2002; Ghatak and Basu, 2011). We suggest a relatively primitive Kerguelen plume source for the lavas with nearly flat REEs, that was also responsible for the Sylhet, Bunbury-Casuarina, Rajmahal Group I and the least contaminated Kerguelen Plateau basalts. Our suggestion is in contrast with the previous proposal by several workers (Kumar et al., 2003) for a generally enriched end-member for the Kerguelen plume with eNd(I) = 2 to 4 and 87Sr/86Sr(I) of 0.7058 (747, Km in Fig. 8). Although Weis et al. (1993, 1998) also suggested an enriched end-member for the Kerguelen plume, these authors were discussing the Kerguelen Archipelago alkalic lavas representing the young Kerguelen plume stem. Our data indicate the primitive plume source to have the same geochemical composition as the uncontaminated Sylhet Traps (Ghatak and Basu, 2011) and Kerguelen drill core 1138 (Neal et al., 2002). The remaining LREE enriched Rajmahal basalts are contaminated by a lower crustal or lithospheric mantle, similar in geochemical signatures of the Eastern Ghats granulites. 4.3. Trace element geochemistry of the alkalic complexes: Sung, Sikkim, Samchampi and Barpung The trace elements of the pyroxenites, lamproites, carbonatites, meltiegites, and sovites of the four alkaline complexes indicate high concentrations of Ba (average 900 ppm), Sr (average 1900 ppm), Zr (average 400 ppm) and La (average 125 ppm) (Table 2), conforming to the chemical characteristics of global lamproites (e.g. Mitchell and Bergman, 1991). The concentrations of these elements are lower in the corresponding syenites and ijolites of these complexes, except for Sikkim where the ultra potassic syenites are very similar in their trace element concentrations to the lamproites and nephelinites

Table 2 Trace element concentrations of rocks from the four alkalic complexes of this study. Analytical uncertainties are less than 5% for all the elements analyzed and usually less than 2% for the REEs. Sample

SKM-62

SKM-56

SKM-59

B-33

B-48

B-52

SKM-8

SKM-9

SKM-10

SKM-24

SKM-44

L-21

B-54

Sikkim

Sample

Nephelinite

Lamproite

Lamproite

Lamproite

Lamproite

LamProite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

66 1340 1161 11.9 95 179 24.3 91 15.6 4.22 10.7 1.27 5.72 0.96 2.14 0.26 1.50 0.16 25.7 11.4 1.68 580 11.3 67 3.47 17.1

2.97 423 1605 10.3 129 318 38.1 1578 26.6 6.87 15.4 1.64 6.40 0.95 1.89 0.21 1.16 0.12 24.5 8.02 0.89 328 6.28 65 3.52 14.2

72 3232 10250 73 622 1253 151 545 100 24.5 63 6.37 25.2 3.85 138 0.88 4.68 0.59 88 76 5.17 329 8.19 171 7.08 28.0

115 2835 1712 29.2 157 380 42.1 159 27.5 7.56 19.8 2.43 11.2 1.88 4.31 0.53 3.06 0.34 51 22.0 3.87 1262 25.0 120 6.50 23.6

104 1576 1771 19.3 126 248 31.6 119 18.2 4.84 12.9 1.50 6.71 1.13 2.62 0.31 1.70 0.18 31.5 18.0 2.69 647 13.3 82 4.33 13.2

102 3926 3377 17.2 131 322 38.0 154 26.0 7.01 17.7 2.04 8.62 1.41 3.06 0.35 1.94 0.23 38.6 9.29 1.63 540 12.4 70 3.43 17.6

126 3360 2101 81 370 702 75 257 37.4 10.6 26.7 3.29 16.2 2.91 6.68 0.78 3.77 0.39 75 79 6.61 417 10.8 392 18.6 17.4

141 4377 1380 71 211 426 46.64 162 24.9 6.95 17.0 2.06 9.79 1.66 3.68 0.42 2.05 0.21 40.0 31.2 3.79 201 6.08 152 8.97 13.1

139 4340 1275 65 199 404 44.6 162 25.1 6.93 16.5 1.99 9.19 1.54 3.37 0.38 1.86 0.20 37.3 36.4 3.95 205 5.81 122 7.84 10.7

108 2900 1135 132 461 882 94 307 41.2 10.2 21.9 2.39 10.0 1.52 3.27 0.39 2.30 0.31 32.9 93 2.67 268 9.63 304 9.11 13.3

142 1372 3953 103 465 890 95 309 43.7 11.4 26.3 3.06 13.2 1.93 3.72 0.39 2.05 0.26 43.0 74 2.38 98 5.01 193 6.85 12.1

148 315 4760 75 427 869 85 288 45.6 12.2 28.7 3.66 15.9 2.34 4.50 0.46 2.41 0.26 57 57 1.69 389 13.8 309 15.1 11.2

148 587 54 45.0 104 162 23.7 81 12.2 3.03 6.29 0.86 4.62 0.92 2.40 0.34 2.10 0.24 16.6 23.4 6.74 778 19.4 70 2.28 1.78

SAM71

SAM179A

SAM180

SAM-70

SAM-112

SAM-M/ 1J

SAM63

SAM161

SAM59

SAM77

SAM84

SAM169

SAM185

SAM-SG/ 68

SAM-S/ 92

Syenite

Samchampi

Rb Ba Sr Pb La Ce Pr

Sovite

Sovite

Sovite

Pyroxenite

Pyroxenite

Melteigite

Ijolite

Ijolite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

11.7 567 4550 2.16 391 918 119

2.87 173 3348 1.88 209 488 55

33.3 346 3348 1.39 205 472 53

22.3 242 990 2.53 96 236 40.2

124 842 896 7.17 66 165 22.8

3.12 152 623 3.35 149 422 48.6

5.86 220 739 4.71 176 532 57

20.5 234 819 2.29 50 125 17.4

122 2660 1075 46.5 166 377 41.0

182 443 488 12.1 36.2 79 9.35

124 1536 1438 7.58 94 200 26.6

246 1790 441 7.76 64 138 18.5

188 1700 508 30.4 134 245 28.0

98 141 497 1631 835 1235 1.74 5.02 61 82 142 188 17.4 23.7 (continued on next page)

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Rb Ba Sr Pb La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Y Th U Zr Hf Nb Ta Sc

59

60

Table 2 (continued) Sample

SAM71

SAM179A

SAM180

SAM-70

SAM-112

SAM-M/ 1J

SAM63

SAM161

SAM59

SAM77

SAM84

SAM169

SAM185

SAM-SG/ 68

SAM-S/ 92

Samchampi

Sample

Sovite

Sovite

Sovite

Pyroxenite

Pyroxenite

Melteigite

Ijolite

Ijolite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

441 85 21.1 61 6.49 26.9 4.15 147 0.90 4.37 0.52 101 93 4.68 684 4.35 185 53 4.22

208 30.8 8.37 20.5 2.12 9.43 1.51 3.15 0.34 1.80 0.20 39.2 1.49 0.44 546 14.6 13.6 1.53 14.5

195 27.9 7.60 18.5 1.91 8.42 1.33 2.70 0.28 1.41 0.18 36.0 0.49 0.51 59 1.02 15.8 2.08 1.93

157 28.8 7.76 23.3 2.99 14.9 2.62 6.48 0.82 5.00 0.72 66 17.0 3.60 310 5.94 261 21.4 16.0

89 15.8 4.21 11.0 1.34 6.65 1.14 2.73 0.38 2.70 0.41 29.1 22.5 1.52 666 12.2 364 20.2 18.1

191 32.2 8.39 21.7 2.71 12.6 2.11 5.00 0.63 4.03 0.57 50 45.5 2.63 561 11.2 224 13.9 37.6

221 37.1 9.51 25.0 3.06 13.8 2.31 5.33 0.67 4.09 0.56 56 52 4.71 557 11.8 192 9.98 25.9

69 12.7 3.30 9.36 1.18 6.08 1.12 2.90 0.48 3.99 0.74 27.7 13.1 0.45 523 9.32 97 2.66 29.7

144 23.1 4.90 15.3 1.90 9.81 1.84 4.60 0.63 3.81 0.53 47.9 14.5 0.55 168 4.25 41.7 1.64 21.2

32.3 5.55 1.46 4.08 0.56 2.93 0.54 23.9 0.21 1.39 0.18 14.1 29.0 17.9 343 5.34 1488 14.6 0.73

97 13.5 3.15 8.41 1.02 4.75 0.86 2.21 0.32 2.31 0.38 20.3 8.36 0.93 231 5.34 222 3.56 22.4

69 11.6 2.93 7.75 0.97 4.43 0.77 1.80 0.24 1.60 0.24 18.5 9.28 2.86 226 5.35 238 6.78 6.76

97 14.7 3.25 10.6 1.33 7.07 1.35 3.29 0.43 2.52 0.35 35.7 37.4 1.11 89 2.45 63.1 2.82 10.4

65 10.9 2.84 7.73 1.02 5.20 0.97 2.53 0.37 2.53 0.38 23.3 10.5 1.40 478 9.3 412 9.96 11.1

86 11.3 2.70 6.61 0.71 3.42 0.59 1.45 0.22 1.65 0.29 13.7 8.45 1.52 212 4.65 192 2.93 19.6

K2

K4

K5

K8

K12

K1

K6

K10

K11

K13

S-2

Pyroxenite

Pyroxenite

Pyroxenite

Pyroxenite

Pyroxenite

Syenite

Syenite

Syenite

Syenite

Syenite

1.68 29.7 455 10.1 51 126 17.9 73 13.6 3.85 10.4 1.27 5.88 1.01 2.31 0.30

60 674 1992 5.34 124 243 34.4 113 15.5 4.04 10.9 1.23 5.52 0.93 2.23 0.26

50 821 1182 3.76 87 138 15.6 55 10.5 3.26 9.52 1.29 6.95 1.30 3.15 0.42

11.1 123 1433 1.93 92 147 16.0 55 9.03 2.60 7.73 0.93 4.84 0.89 2.23 0.30

25.0 2692 1075 3.63 66 87 10.1 36.3 7.72 2.48 8.22 1.08 5.87 1.13 2.70 0.35

11.7 86 137 15.6 22.3 52 6.70 28.4 7.07 2.16 7.95 1.24 7.23 1.52 3.97 0.58

210 1560 227 3.08 10.0 17.5 2.40 8.80 1.51 0.38 1.06 0.15 0.84 0.16 0.42 0.06

266 935 473 17.0 47.4 93 9.66 33.7 5.72 1.06 4.81 0.68 3.90 0.79 2.20 0.34

60 933 391 21.9 35.2 74 9.23 35.0 6.71 1.63 5.64 0.80 4.69 0.97 2.58 0.38

332 495 263 21.9 53 124 8.67 26.8 4.28 1.14 3.44 0.49 3.20 0.72 2.15 0.34

Barpung

Rb Ba Sr Pb La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm

S-102

S-111

Carbonatite

Carbonatite

Carbonatite

0.02 36.14 2710 1.24 21.5 47.1 5.64 19.5 3.07 0.86 2.08 0.24 1.08 0.17 0.39 0.04

0.16 162 3705 1.65 124 325 35.9 134 23.6 6.84 18.1 2.18 9.71 1.57 3.41 0.37

0.01 307 6825 4.21 146 384 39.8 142 23.4 6.76 18.9 2.43 11.2 1.88 4.36 0.53

Sung

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Y Th U Zr Hf Nb Ta Sc

Yb Lu Y Th U Zr Hf Nb Ta Sc

1.48 0.20 23.5 13.9 3.70 315 7.45 116 7.13 17.8

2.63 0.37 36.2 2.84 2.00 330 7.38 140 4.44 36.7

1.92 0.29 28.1 4.49 0.85 96 1.69 18.7 0.45 9.07

2.21 0.35 33.4 4.32 0.56 93 2.74 6.71 0.70 39.5

3.72 0.52 41.3 5.46 0.91 63 1.63 15.8 1.14 27.8

0.41 0.06 3.89 1.11 0.64 70 1.32 20.5 0.76 0.85

2.36 0.36 23.0 14.9 2.52 239 6.55 14.7 2.29 8.81

2.44 0.35 27.5 9.86 1.22 176 4.14 9.05 0.62 40.8

2.13 0.26 19.7 21.1 3.31 309 5.55 83 1.14 0.86

0.26 0.04 4.08 0.38 0.02 2.01 0.09 1.72 0.01 5.62

1.96 0.25 41.0 1.15 0.08 42.7 1.12 11.2 3.25 14.4

3.04 0.42 48.7 1.06 0.40 0.37 0.05 57 1.13 7.36

S-114

S-115

S-24

S-28

S-15

S-69

S-8

S-34

S-35

S-46

S-50

S-56

S-65

S-94

S-96

Carbonatite

Carbonatite

Uncompahgrite

Uncompahgrite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

0.96 204 3416 2.15 172 436 53 203 35.8 10.3 26.5 3.06 14.0 2.25 4.66 0.51 2.67 0.32 54 3.94 0.31 416 6.63 18.4 4.89 22.0

0.02 105 4208 2.07 119 330 37 143 25.1 7.16 18.2 2.17 9.60 1.56 3.31 0.37 1.92 0.24 39.9 2.10 1.46 11.5 0.30 39.1 25.7 15.7

1.34 23.3 341 0.53 4.61 9.97 1.38 5.76 1.21 0.36 0.97 0.13 0.62 0.11 0.26 0.03 0.22 0.03 2.70 0.06 0.00 20.2 0.87 0.53 0.07 123

3.89 45.8 1227 1.07 13.0 25.3 3.02 11.3 1.90 0.58 1.52 0.17 0.80 0.13 0.26 0.03 0.14 0.02 3.19 0.03 0.00 8.41 0.17 0.52 0.07 1.94

8.12 65 468 9.06 39.8 71 10.8 43.3 7.73 1.99 6.75 0.83 4.08 0.72 1.64 0.20 1.15 0.17 19.1 3.70 0.38 85 2.41 11.8 4.32 27.3

5.04 8.55 278 0.83 228 501 56 202 34.3 9.04 23.4 3.04 13.9 2.29 5.05 0.59 3.41 0.50 43.1 38.9 6.71 568 8.15 553 33.4 13.4

123 1116 603 5.87 6.34 13.5 1.55 5.84 1.07 0.31 1.02 0.15 0.90 0.19 0.54 0.08 0.46 0.05 4.93 5.04 0.15 17.4 0.87 32.9 2.15 0.27

56 543 232 1.25 5.77 13.6 1.50 5.61 1.11 0.34 1.04 0.16 1.10 0.24 0.76 0.16 1.59 0.32 5.69 3.19 0.25 293 6.61 8.51 0.84 0.49

81 866 522 10.1 21.3 43.5 5.03 18.6 3.61 1.08 3.36 0.51 3.08 0.64 1.78 0.28 2.11 0.36 16.5 4.42 0.55 221 5.09 36.8 2.42 1.66

78 1040 595 9.26 21.8 44.0 4.92 17.4 3.26 0.98 2.90 0.44 2.59 0.55 1.50 0.24 1.79 0.31 13.8 4.32 0.44 175 4.09 29.5 1.77 1.38

8.83 83 365 0.31 5.25 14.1 2.14 9.24 2.07 0.63 1.72 0.23 1.12 0.20 0.45 0.06 0.36 0.05 4.98 0.23 0.01 42.1 0.81 2.79 0.56 6.85

155 1131 290 4.98 24.2 54 7.30 27.3 6.26 1.17 5.57 0.92 5.40 1.06 42.7 0.40 2.70 0.41 27.9 3.34 1.31 179 3.69 78 3.05 11.2

49.2 11.6 325 0.34 37 76 13.8 55 10.6 2.78 9.13 1.23 6.22 1.11 2.75 0.37 2.55 0.42 27.1 7.32 0.45 569 10.1 147 12.5 1.22

118 376 267 25.4 43.5 89 12.2 50 12.3 3.35 15.9 2.47 14.2 2.85 7.44 1.05 6.37 0.87 72 9.27 1.13 47.7 1.83 39.0 2.35 29.5

311 285 117 8.75 11.7 26.1 3.50 14.7 3.87 1.20 11.1 1.74 9.84 1.94 5.04 0.70 4.22 0.58 57 5.02 0.52 12.3 0.55 28.5 1.58 35.6

Sung

Rb Ba Sr Pb La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Y Th U Zr Hf Nb Ta Sc

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Sample

1.91 0.27 25.9 4.47 0.58 237 8.41 73 7.65 36

61

1.5

2.8

2.10

3.8

3.12

3.14

Sahibganj

5.3

5.4

5.6

5.8

5.10

62

Table 3 Present day (0) and initial (I) at 117 Ma for Nd, Sr and Pb isotope data of Rajmahal lavas and Bokaro dikes and their corresponding Rb/Sr, Sm/Nd, U/Pb and Th/Pb ratios. 5.11

Tinpahar

Basaltic flow Basaltic flow Basaltic flow Basaltic flow Basaltic flow Basaltic flow Basaltic flow Basaltic flow Basaltic flow Basaltic flow Andesitic dike Andesitic dike 147

0.16 0.51244 0.51233 3.4

0.17 0.51239 0.51227 4.5

0.17 0.51240 0.51229 4.2

0.17 0.51233 0.51222 5.5

0.17 0.51232 0.51221 5.7

0.16 0.51235 0.51224 5.1

0.21 0.51278 0.51263 2.4

0.21 0.51272 0.51258 1.5

0.16 0.51224 0.51213 7.3

0.17 0.51223 0.51210 7.7

0.16 0.51234 0.51223 5.2

87

Rb/86Sr Sr/86Sr(0) 87 Sr/86Sr(I)

0.14 0.70777 0.70756

0.15 0.70657 0.70633

0.14 0.70586 0.70563

0.19 0.70656 0.70625

0.15 0.70703 0.70679

0.15 0.70665 0.70641

0.21 0.70768 0.70734

0.09 0.70360 0.70347

0.06 0.70406 0.70397

0.28 0.70770 0.70725

0.09 0.70760 0.70745

0.15 0.70758 0.70733

206

18.019 15.694 39.280 4.97 0.04 38.13 17.931 15.690 38.869

17.987 15.633 38.949 5.11 0.04 36.36 17.897 15.629 38.558

17.944 15.615 38.718 5.43 0.04 36.18 17.848 15.610 38.328

17.997 15.628 38.951 4.83 0.04 35.58 17.912 15.624 38.568

18.412 15.659 39.018 4.40 0.03 31.68 18.334 15.656 38.677

18.010 15.658 39.010 5.00 0.04 35.77 17.922 15.653 38.625

17.996 15.697 39.136 4.71 0.03 33.49 17.913 15.693 38.776

17.932 15.568 38.095 6.10 0.04 30.08 17.824 15.563 37.771

17.624 15.488 37.636 4.66 0.03 21.72 17.541 15.484 37.402

18.027 15.738 40.800 4.99 0.04 37.00 17.939 15.734 40.401

17.901 15.689 38.964 4.58 0.03 33.70 17.820 15.685 38.601

17.996 15.711 39.200 4.83 0.04 34.65 17.910 15.707 38.827

4.2

4.11

4.14

4.15

kJ-1

87

Pb/204Pb(0) Pb/204Pb(0) 208 Pb/204Pb(0) 238 U/204Pb 235 U/204Pb 232 Th/204Pb 206 Pb/204Pb(I) 207 Pb/204Pb(I) 208 Pb/204Pb(I) 207

25

1

6a

Dubrajpur sst.

Harinsingha

4.1 Tinpahar

kJ-24

II-L3

Bokaro

Basaltic flow

Basaltic flow

Basaltic flow

Gondwana sandstone

Basaltic flow

Basaltic flow

Basaltic flow

Basaltic flow

Andesitic flow

Lamprophyre dike

Diabase dike

Diabase dike

Sm/144Nd Nd/144Nd(0) 143 Nd/144Nd(I)

0.20 0.51278 0.51264

0.20 0.51267 0.51253

0.21 0.51271 0.51257

0.11 0.51228 0.51220

0.16 0.51238 0.51227

0.16 0.51228 0.51217

0.19 0.51278 0.51264

0.20 0.51272 0.51258

0.11 0.51240 0.51232

0.09 0.51244 0.51238

0.17 0.51275 0.51263

0.20 0.51267 0.51253

eNd(I) 87 Rb/86Sr 87 Sr/86Sr(0) 87 Sr/86Sr(I)

2.8 0.07 0.70442 0.70430

0.6 0.03 0.70478 0.70473

1.3 0.03 0.70469 0.70465

-5.8 1.06 0.72138 0.71968

-4.5 0.29 0.70850 0.70804

-6.5 0.16 0.70800 0.70774

2.8 0.09 0.70395 0.70381

1.6 0.11 0.70465 0.70447

-3.5 0.50 0.70602 0.70522

-2.4 0.18 0.70583 0.70555

2.5 0.22 0.70544 0.70508

0.6 0.11 0.70670 0.70653

206

18.107 15.598 38.343 3.99 0.03 20.81 18.036 15.595 38.119

17.759 15.558 38.012 5.20 0.04 24.87 17.667 15.553 37.744

17.881 15.572 38.292 5.67 0.04 29.75 17.781 15.567 37.972

18.460 15.754 38.832 5.10 0.04 15.44 18.370 15.749 38.665

18.053 15.677 39.082 4.79 0.03 41.24 17.969 15.673 38.638

17.872 15.675 38.282 4.22 0.03 33.10 17.798 15.671 37.926

17.407 15.510 37.244 2.65 0.02 11.83 17.360 15.507 37.116

17.553 15.567 37.607 3.65 0.03 17.27 17.489 15.563 37.422

18.136 15.579 38.197 7.85 0.06 35.80 17.997 15.572 37.812

17.720 15.523 37.710 9.53 0.07 48.59 17.551 15.515 37.187

17.331 15.416 38.230 10.42 0.08 47.91 17.147 15.407 37.715

18.051 15.546 38.304 6.03 0.04 33.61 17.944 15.541 37.942

147 143

Pb/204Pb(0) Pb/204Pb(0) 208 Pb/204Pb(0) 238 U/204Pb 235 U/204Pb 232 Th/204Pb 206 Pb/204Pb(I) 207 Pb/204Pb(I) 208 Pb/204Pb(I) 207

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Sm/144Nd 0.16 Nd/144Nd(0) 0.51241 143 Nd/144Nd(I) 0.51229 eNd(I) 4.0 143

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

(Fig. 9). In general, the ultrabasic–carbonatitic rocks and some of the syenites of the four alkaline complexes are similar to global lamproites and Group II kimberlites (e.g. Mitchell and Bergman, 1991; Chakrabarti et al., 2007). Notable absence of Eu-anomaly is ubiquitous in all the ultrabasic–carbonatitic rocks. The extreme enrichment of the LREEs (Fig. 9), in the mafic rocks of the alkalic complexes precludes significant contamination by continental crust. The ultrabasic–carbonatitic rocks also show characteristic depletion in HREE relative to both continental crust and MORB (Fig. 9), excluding contamination with by sources. Therefore, we consider the ultrabasic–carbonatitic rocks were derived by direct melting of the plume source, without significant contamination of continental crust or MORB. The syenites of the alkalic complexes, in contrast, display variable REE patterns (Fig. 9) with strong LREE enrichment and HREE depletion (Sikkim) to relatively small LREE enrichment with nearly flat HREE patterns (Sung, Barpung). Two syenites from Sung and Samchampi, and an ijolite from Samchampi show strong concave up HREE enrichment (Lu 10–20 chondrite) that may be due to the presence of zircons (which show characteristic U-shaped REE patterns) in these rocks (Hoskin and Schaltegger, 2003). A few of the syenites exhibit a small negative Eu anomaly due to small amount of plagioclase fractionation. Strong incompatible element enrichment, except Rb, with progressive depletion of the less incompatible elements in the different lithologies of the ultrabasic–carbonatitic rocks compared to the primitive mantle (Fig. 9) characterizes the four alkaline complexes. These trace element patterns are likely due to low-degree partial melting of a metasomatized garnet-peridotite source, as suggested for the origin of lamproites, kimberlites, and carbonatites world-wide (e.g. Mitchell and Bergman, 1991). We suspect Rb is variable based on the presence or absence of Rb-bearing minerals these alkalic rocks. Positive Nb–Ta anomalies in Sung, Samchampi and Barpung alkalic rocks, we suggest, indicate variable modal mineralogies with Ti-bearing minerals, including perovskite. Similarly negative Nb–Ta anomalies in some of these rocks are interpreted due to removal of Ti-bearing minerals by fractional crystallization. The extremely negative Zr–Hf anomalies in a few ultrabasic–carbonatitic rocks (Fig. 9) are likely caused by the removal of zircons. The multiple trace element patterns of the syenitic rocks of the four alkalic complexes are similar to the corresponding ultrabasic–carbonatitic rocks in the same complex (Fig. 9), suggesting common parental sources for these rocks and their host complexes. It is interesting that the syenitic rocks occur as veins or pods within the ultrabasic–carbonatitic hosts (see Appendix figures A5– A8). 4.4. Nd–Sr–Pb-isotopes: Sung, Samchampi, Barpung, and Sikkim The Pb isotopic ratios of the alkalic rocks are shown in Fig. 10a, b with the relevant basalt fields and crust–mantle reservoirs, as in Fig. 6. The Sikkim lamproites show high initial 207Pb/204Pb compared to 206Pb/204Pb values, falling

63

near the 4.45 Ga Geochron (Fig. 10b). Komatiites with lower silica and depleted HREE patterns (e.g. Nesbitt et al., 1979) are suggested in the sources of other global lamproites including the Proterozoic (1.2 Ga) Krishna lamproites, interestingly, in the eastern margin of the Southern Indian craton (Chakrabarti et al., 2007). We will evaluate a likely subducted ancient basaltic-komatiite (eclogitized) source for the Sikkim lamproites in the peridotitic source of the Kerguelen plume. The carbonatites and alkalic ultrabasic rocks, except the Sikkim lamproites, show a wide range in initial Pb isotopic space (Fig. 10) with many samples clustering above and parallel to the NHRL in continuity with the Sikkim data (Fig. 10). Five syenites show significantly higher 207 Pb/204Pb ratios. We believe these high 207Pb/204Pb ratios may be inherited from some ancient reservoirs. Interestingly, as discussed previously, these rocks show a strong Pb-depletion, when normalized to the primitive mantle (Fig. 9), a feature inherited from the mantle source. Post emplacement processes likely have not affected their Pb-isotopic ratios that are different from the upper continental crust in trace element (Fig. 9) and in Nd–Sr plots (Fig. 11). Sr and Nd isotope data are reported in Table 4 for the four alkalic complexes and plotted in Fig. 11 along with fields of primitive Kerguelen plume component (Ghatak and Basu, 2011), Indian MORB, Groups I and II kimberlites, and lamproites from India, Australia, and Gaussberg. Carbonatites from Sung and Samchampi sovites fall in a tight cluster (Fig. 11) and overlap with Group I kimberlites, Rajmahal Group I basalts, and the primitive Kerguelen plume component (Neal et al., 2002; Ghatak and Basu, 2011) in the upper left quadrant of the Nd–Sr plot, essentially at the bulk silicate Earth (BSE) composition (Fig. 11). We note, however, the carbonatites of the Sung Valley complex previously analyzed by Srivastava et al. (2005) and those reported by Veena et al. (1998) plot in the upper right quadrant of the Nd–Sr diagram. These authors suggested the cause of Sr enrichment in these rocks due to an enriched mantle source involvement (Veena et al., 1998; Srivastava et al., 2005), specifically, an EM-2 source (Veena et al., 1998). However, none of the 15 samples of various lithologies analyzed in this study fall in the upper right hand quadrant (Fig. 11). Note that several alkalic mafic–ultramafic rocks from Sung, Samchampi and Barpung form a cluster very close to the BSE (Fig. 11). It is interesting that these samples belonging to the shaded region representing the Kerguelen plume component show low Rb/Sr ratios from 0.01 to 0.1 causing little reduction of 87Sr/86Sr for the initial ratio estimate. The rest of the samples of this study show a wide scatter originating from the upper left hand quadrant near the BSE and going to the lower right quadrant with more radiogenic 87Sr/86Sr ratios and strongly negative initial eNd values. The lack of isotopic coherence among different lithologic units of each of the four alkalic complexes is an important characteristic that questions the consanguinity of the various rock types in these complexes. Traditionally such inhomogeneities are explained by assimilation of country rocks by the initial magmas of the intrusives. Because the trace

SKM-62

SKM-56

SKM-59

B-33

B-48

B-52

SKM-8

SKM-9

SKM-10

SKM-24

SKM-44

B-54

64

Table 4 Present day (0) and initial (I) at 117 Ma for Nd, Sr and Pb isotope data of ultrabasic–carbonatitic rocks and the syenites and ijolites from the four alkalic complexes of this study, including their corresponding Rb/Sr, Sm/Nd, U/Pb and Th/Pb ratios. L-21

Sikkim Lamproite

Lamproite

Lamproite

Lamproite

Lamproite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Syenite

Sm/144Nd 143 Nd/144Nd(0) 143 Nd/144Nd(I) eNd(I)

0.11 0.51230 0.51222 5.4

0.11 0.51238 0.51231 3.8

0.12 0.51231 0.51223 5.2

0.11 0.51234 0.51226 4.6

0.10 0.51224 0.51217 6.4

0.11 0.51237 0.51229 4.0

0.09 0.51219 0.51212 7.4

0.10 0.51226 0.51219 6.0

0.10 0.51218 0.51211 7.6

0.08 0.51224 0.51218 6.2

0.09 0.51224 0.51217 6.4

0.10 0.51221 0.51215 6.9

0.10 0.51214 0.51207 8.4

87

Rb/86Sr Sr/86Sr(0) 87 Sr/86Sr(I)

0.16 0.70684 0.70658

0.01 0.71146 0.71145

0.02 0.70608 0.70605

0.19 0.70910 0.70879

0.16 0.70809 0.70782

0.09 0.70733 0.70720

0.17 0.70652 0.70625

0.29 0.70675 0.70629

0.31 0.70723 0.70673

0.27 0.70667 0.70624

0.10 0.70794 0.70778

7.71 0.71529 0.70291

0.09 0.70848 0.70834

206

18.209 15.575 38.525 8.93 0.06 62.49 18.052 15.567 37.852

18.068 15.568 38.207 5.45 0.04 50.67 17.972 15.563 37.662

17.909 15.517 38.375 4.48 0.03 67.72 17.829 15.514 37.646

18.225 15.674 38.786 8.44 0.06 49.69 18.075 15.666 38.251

17.777 15.546 38.440 8.75 0.06 60.40 17.622 15.538 37.789

17.921 15.599 39.418 6.05 0.04 35.68 17.814 15.594 39.034

17.744 15.566 38.490 5.11 0.04 63.28 17.654 15.562 37.809

17.726 15.526 38.393 3.33 0.02 28.40 17.667 15.523 38.088

17.690 15.468 38.189 3.79 0.03 36.20 17.623 15.465 37.799

17.871 15.622 38.551 1.28 0.01 45.97 17.848 15.621 38.056

17.839 15.548 38.369 1.45 0.01 46.60 17.813 15.547 37.867

18.090 15.605 38.601 9.49 0.07 34.11 17.923 15.597 38.234

17.793 15.528 38.308 1.41 0.01 48.66 17.768 15.527 37.784

SAM-71

SAM-179A

SAM-180

SAM-70

SAM-112

SAM M-IJ

SAM-63

SAM-161

SAM-59

SAM-77

SAM-84

SAM-169

SAM-185

87

Pb/204Pb(0) Pb/204Pb(0) 208 Pb/204Pb(0) 238 U/204Pb 235 U/204Pb 232 Th/204Pb 206 Pb/204Pb(I) 207 Pb/204Pb(I) 208 Pb/204Pb(I) 207

Samchampi Sovite

Sovite

Sovite

Pyroxenite

Pyroxenite

Melteigite

Ijolite

Ijolite

Syenite

Syenite

Syenite

Syenite

Syenite

147

Sm/144Nd 143 Nd/144Nd(0) 143 Nd/144Nd(I) eNd(I)

0.12 0.51260 0.51252 0.3

0.09 0.51258 0.51252 0.3

0.09 0.51263 0.51256 1.2

0.12 0.51244 0.51236 2.7

0.11 0.51245 0.51237 2.5

0.11 0.51263 0.51256 1.1

0.11 0.51249 0.51242 1.6

0.12 0.51235 0.51227 4.5

0.10 0.51182 0.51175 14.7

0.11 0.51249 0.51242 1.7

0.09 0.51242 0.51236 2.7

0.11 0.51200 0.51193 11.1

0.10 0.51166 0.51159 17.8

87

Rb/86Sr Sr/86Sr(0) 87 Sr/86Sr(I)

0.01 0.70472 0.70471

0.00 0.70473 0.70472

0.03 0.70482 0.70477

0.06 0.70617 0.70606

0.39 0.70971 0.70909

0.01 0.70660 0.70658

0.02 0.70669 0.70666

0.07 0.71139 0.71128

0.32 0.71387 0.71336

1.05 0.70820 0.70651

0.24 0.70807 0.70768

1.57 0.70391 0.70138

1.04 0.72901 0.72734

206

22.810 15.810 61.426 189.32 1.37 3867.65 19.462 15.648 19.793

19.727 15.739 40.299 15.38 0.11 54.25 19.455 15.726 39.715

19.892 15.713 41.518 25.01 0.18 24.51 19.450 15.692 41.254

21.275 15.816 44.683 101.92 0.74 497.59 19.473 15.729 39.327

18.934 15.708 40.912 14.07 0.10 214.98 18.686 15.696 38.598

19.931 15.757 44.295 54.96 0.40 983.68 18.959 15.710 33.707

20.179 15.773 44.398 70.32 0.51 800.65 18.935 15.713 35.780

19.157 15.724 41.331 13.16 0.10 359.55 18.925 15.713 37.073

17.724 15.613 38.943 0.75 0.01 20.44 17.711 15.613 38.723

20.745 15.787 40.729 99.74 0.72 167.46 18.981 15.702 38.927

19.299 15.795 40.567 8.10 0.06 75.48 19.155 15.788 39.754

19.093 15.785 41.086 24.47 0.18 82.18 18.661 15.764 40.201

18.521 15.785 42.287 2.45 0.02 85.37 18.478 15.782 41.368

87

Pb/204Pb(0) Pb/204Pb(0) 208 Pb/204Pb(0) 238 U/204Pb 235 U/204Pb 232 Th/204Pb 206 Pb/204Pb(I) 207 Pb/204Pb(I) 208 Pb/204Pb(I) 207

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Nephelinite 147

147

0.11 0.51225 0.51217 6.4

0.08 0.51242 0.51236 2.8

0.12 0.51247 0.51239 2.1

0.09 0.51256 0.51249 0.1

0.12 0.51237 0.51229 4.2

0.10 0.51251 0.51244 1.2

0.13 0.51239 0.51230 3.9

0.16 0.51237 0.51226 4.6

0.11 0.51234 0.51227 4.6

0.11 0.51204 0.51196 10.5

0.12 0.51195 0.51187 12.3

0.10 0.51236 0.51229 4.0

0.10

87

Rb/86Sr Sr/86Sr(0) 87 Sr/86Sr(I)

0.33 0.71155 0.71102

0.32 0.70729 0.70677

0.01 0.70688 0.70686

0.10 0.70646 0.70630

0.12 0.70762 0.70743

0.02 0.70595 0.70592

0.07 0.70836 0.70825

0.24 0.71555 0.71516

2.60 0.70921 0.70503

1.58 0.71563 0.71309

0.43 0.71369 0.71300

3.55 0.71160 0.70589

0.00 0.70537 0.70537

206

19.603 15.699 41.550 54.03 0.39 419.81 18.648 15.653 37.031

19.159 15.642 39.437 3.73 0.03 29.64 19.093 15.639 39.118

19.797 15.732 40.183 45.91 0.33 178.46 18.986 15.693 38.262

19.611 15.744 39.559 34.91 0.25 51.12 18.993 15.715 39.009

19.635 15.713 40.232 29.09 0.21 159.43 19.121 15.688 38.516

18.748 15.694 39.427 9.95 0.07 79.70 18.573 15.686 38.570

18.304 15.739 39.169 3.76 0.03 23.22 18.238 15.736 38.919

19.328 15.635 39.439 13.57 0.10 24.21 19.088 15.623 39.179

19.681 15.762 40.630 9.89 0.07 60.34 19.506 15.754 39.980

17.660 15.575 38.547 3.50 0.03 29.27 17.598 15.572 38.232

19.638 15.674 40.077 9.96 0.07 65.52 19.462 15.665 39.372

19.390 15.666 41.086 1.01 0.01 21.25 19.372 15.665 40.857

87

Pb/204Pb(0) Pb/204Pb(0) 208 Pb/204Pb(0) 238 U/204Pb 235 U/204Pb 232 Th/204Pb 206 Pb/204Pb(I) 207 Pb/204Pb(I) 208 Pb/204Pb(I) 207

S-102

S-111

S-114

S-115

S-24

S-28

S-15

S-69

S-8

S-34

S-35

S-46

Carbonatite

Carbonatite

Carbonatite

Carbonatite

Uncompahgrite

Uncompahgrite

Pyroxenite

Pyroxenite

Syenite

Syenite

Syenite

Syenite

Sm/ Nd Nd/144Nd(0) 143 Nd/144Nd(I) eNd(I)

0.11 0.51262 0.51254 0.7

0.10 0.51265 0.51258 1.5

0.11 0.51262 0.51254 0.8

0.11 0.51265 0.51257 1.3

0.13 0.51254 0.51247 0.7

0.11 0.51262 0.51197 0.7

0.12 0.51251 0.51242 1.5

0.11 0.51254 0.51246 0.7

0.12 0.51197 0.51189 11.9

0.13 0.51192 0.51183 13.1

0.12 0.51205 0.51197 10.4

0.12 0.51203 0.51195 10.8

87

Rb/86Sr Sr/86Sr(0) 87 Sr/86Sr(I)

0.00 0.70456 0.70456

0.00 0.70447 0.70447

0.00 0.70443 0.70443

0.00 0.70442 0.70442

0.01 0.70566 0.70564

0.01 0.70562 0.70561

0.05 0.70676 0.70668

0.05 0.70571 0.70563

0.58 0.71165 0.71072

0.68 0.71159 0.71050

0.44 0.71113 0.71043

0.37 0.71090 0.71030

206

19.505 15.608 39.648 3.06 0.02 47.09 19.451 15.606 39.142

19.627 15.718 39.804 6.26 0.05 17.22 19.516 15.713 39.619

19.624 15.627 40.026 9.41 0.07 124.88 19.457 15.618 38.682

20.370 15.668 40.060 46.94 0.34 69.92 19.540 15.628 39.307

18.974 15.596 38.821 0.58 0.00 7.11 18.964 15.595 38.744

19.361 15.678 36.429 0.14 0.00 1.84 19.359 15.678 36.410

18.416 15.646 38.517 2.63 0.02 26.83 18.369 15.644 38.228

32.393 16.327 59.705 768.46 5.57 4598.63 18.804 15.671 10.205

18.719 15.759 39.704 1.65 0.01 57.65 18.690 15.757 39.083

18.996 15.758 40.780 13.26 0.10 174.06 18.762 15.747 38.907

18.661 15.689 39.309 3.54 0.03 29.19 18.598 15.686 38.995 (continued

Sung 147

144

143

87

Pb/204Pb(0) Pb/204Pb(0) 208 Pb/204Pb(0) 238 U/204Pb 235 U/204Pb 232 Th/204Pb 206 Pb/204Pb(I) 207 Pb/204Pb(I) 208 Pb/204Pb(I) 207

18.761 15.700 39.458 3.07 0.02 31.23 18.707 15.697 39.122 on next page)

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Sm/144Nd Nd/144Nd(0) 143 Nd/144Nd(I) eNd(I) 143

65

18.103 15.624 38.705 3.76 0.03 37.66 18.036 15.620 38.299

7.49 0.80253 0.79051

S-96

0.17 0.51224 0.51213 7.3

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72 Syenite

66

element concentrations are extremely high in the lithologic members of these complexes, crustal assimilation is unlikely. Therefore, we recognize the isotopic inhomogeneities as inherited from the mantle source for these complexes. 5. PETROGENESIS OF THOLEIITES AND ALKALICMAFIC–ULTRAMAFIC-CARBONATITE COMPLEXES IN NORTHEAST INDIA IN A HETEROGENEOUS KERGUELEN PLUME

18.144 15.655 38.661 2.82 0.02 23.98 18.094 15.653 38.403 20.984 15.783 48.752 97.19 0.70 1643.57 19.265 15.700 31.061 19.242 15.675 39.274 17.13 0.12 45.02 18.939 15.660 38.789 19.428 15.660 39.689 1.37 0.01 49.69 19.404 15.659 39.154 Pb/204Pb(0) Pb/204Pb(0) 208 Pb/204Pb(0) 238 U/204Pb 235 U/204Pb 232 Th/204Pb 206 Pb/204Pb(I) 207 Pb/204Pb(I) 208 Pb/204Pb(I) 207

206

1.25 0.72571 0.72370 0.46 0.70906 0.70832 0.71558 0.71397 0.07 0.70545 0.70534 Rb/86Sr Sr/86Sr(0) 87 Sr/86Sr(I) 87

87

0.15 0.51227 0.51216 6.6 0.13 0.51249 0.51240 2.0 0.15 0.51229 0.51219 6.0

Syenite Syenite Syenite

Sung

0.14 0.51208 0.51198 10.1 Sm/ Nd Nd/144Nd(0) 143 Nd/144Nd(I) eNd(I) 143

144

S-65 S-56 S-50 Table 4 (continued)

147

S-94

Syenite

5.1. Flood basalt genesis: REE and Nd–Sr isotopes Results of partial melting modeling using the batchmelting equation for a spinel lherzolite parent mantle with 50% olivine, 30% orthopyroxene, 18% clinopyroxene and 2% spinel, are shown in Fig. A9 with the starting composition for the REE in sources as 2 Chondrite (see Table T4 and Note 2 in Supplementary materials). Our new analyses of Rajmahal lavas representing 14 separate flows that show a relatively flat REE pattern are compared to an 18% partial melt of the above parent lherzolite in the model. There is strong correspondence of the Rajmahal Group I basalts (RJ I in Fig. A9), with our new results from lavas that extend geographically the Rajmahal province, matching approximately the 18% batch melt pattern of this model. The similarity of the Group I population basalts among the Sylhet Traps, recently established (Ghatak and Basu, 2011) with a 18% batch melt-modeling strongly support the ubiquitous presence of these Group I basalts over a much larger area of the Rajmahal–Sylhet Traps flood basalt province, having a common eNd(I) signature of +0.6 to 2.8. Our modeling in Fig. A9 requiring a spinel peridotite source also suggests Group I basalt generation in the 60–70 km depth in the mantle. This depth constraint has important implications for the nature of the lithospheric roots as discussed below. It is instructive to compare (Fig. 12) initial eNd and 87 Sr/86Sr isotopic ratios of the Rajmahal basalts with available data for the Proterozoic massif anorthosites and granulites of the Chilka Complex (Chakrabarti et al., 2011) and data from mafic granulites from the Eastern Ghats Belt (EGB) (Rickers et al., 2001). Two mixing curves are constructed with the average of the primitive Sylhet basalts considered the most primitive Kerguelen plume component (Ghatak and Basu, 2011) as an end member P, and two end-members of the EGB field, granulites 1 and 2, as the second end members (Fig. 12 and Table T2) corresponding to a tonalite-gneiss and a mafic granulite, respectively, (Rickers et al., 2001). The granulites of the Eastern Ghats Belt are generally believed to constitute the lower continental crust (Rickers et al., 2001; Chakrabarti et al., 2011) in the eastern Indian continental margin (Fig. 1) as they are widely exposed along the eastern margin of the Indian subcontinent (e.g. Yin et al., 2010). Our modeling (Fig. 12) indicates less than 20% of these granulite contaminants in the Rajmahal lavas of this and previous studies (Mahoney et al., 1983; Storey et al., 1989; Baksi, 1995; Kent et al., 1997; Ghatak and Basu, 2011). The presence of lower crustal and continental lithospheric geochemical signatures in the Rajmahal–Sylhet la-

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

67

other similarly contaminated rocks from the Sylhet lavas, some 500 km east of the Rajmahal Traps, Kerguelen Plateau lavas in the Indian Ocean, and lavas in drill hole 738 and 1137 as evidence for the contamination-erosion of the “lost” Indian lithosphere (Kumar et al., 2007), caused by the impact of the Kerguelen plume head beneath the eastern Indian margin. 5.2. Petrogenesis of the alkalic rocks

Fig. 12. Initial eNd and 87Sr/86Sr isotopic ratios of the Rajmahal lavas of this study and the Sylhet Traps of a previous study (Ghatak and Basu, 2011) compared to massif anorthosites-granulites from Chilka (Chakrabarti et al., 2011) and mafic granulites of the Eastern Ghat Belt (EGB), India (Rickers et al., 2001). Four mixing curves are shown resulting by modeling of a common end member P, representing the most primitive CH lavas, and granulites 1 and 2 from the EGB, representing the second end member. This modeling indicates less than 20% contamination for all the Rajmahal lavas of this study by an EGB-like component.

vas would have implications for the local subcontinental lithospheric mantle that was isolated from the convecting asthenosphere for billions of years, evolving with isotopic ratios different from crustal or asthenospheric Nd–Sr isotopic signatures. This residual subcontinental reservoir cannot yield large volume magmas but can contaminate local continental flood basalts (e.g. Neal et al., 2002) as in the Rajmahal–Sylhet flood basalts (Mahoney et al., 1983; Storey et al., 1989; Baksi, 1995; Kent et al., 1997; Ghatak and Basu, 2011). These locally contaminated basalts are thus an important source of knowledge for these geochemically isolated subcontinental lithospheric roots. The Rajmahal Traps and Bokaro dikes are divided into one group with flat REE patterns and mantle-like Nd–Sr– Pb isotopic ratios, and another with LREE enrichments and time-integrated effects in Nd–Sr–Pb isotopes. The samples with LREE enrichment show a clear mixing between a plume end member and a lower crustal-lithospheric end member, similar to the Eastern Ghat granulites (Fig. 12). These traps reside on a late Archaean eastern Indian craton which once had deep lithospheric roots as evidenced by the presence of diamond bearing Proterozoic kimberlites (Basu and Tatsumoto, 1979; Rao et al., 2004). The thickness of the present day Indian lithosphere has recently been estimated from receiver function analyses to be 100 km (Kumar et al., 2007) which is half to one-third the thickness of the Gondwana Supercontinent lithosphere, comprising South Africa, Australia and Antarctica. We interpret the geochemical contamination of the lavas of this study and

7Petrogenetic models commonly invoked for the origin of carbonatites found in association with silicate magmatic rocks include: (i) fractional crystallization of primary carbonated nephelinites; (ii) immiscible liquid that separated from a fractionated silicate magma of nephelinitic composition; and (iii) derivation directly from low-degree melting of a carbonated mantle peridotite (Bailey, 1993; Bell and Simonetti, 1996; Bell and Tilton, 2001; Collerson et al., 2010). The isotopic similarity of the ultramafic–carbonatitic rocks of the alkalic provinces of this study and their contemporaneity and field association with the flood volcanism of the RBST (Srivastava et al., 2005; Ghatak and Basu, 2011) are similar to those in the Siberian and Deccan flood basalt provinces where lower-mantle derived plume origins were suggested on the basis of high 3He evidence (Basu et al., 1993, 1995). Experimental phase equilibria studies of peridotite–CO2 system demonstrate that carbonatite magmas can be generated at depths greater than 70 km (e.g. Wyllie and Huang, 1976; Lee and Wyllie, 1997). These primary carbonated magmas generated at such depths may release CO2 vapor on rising to shallower depths (70 km), increasing the solubility of pyroxenes in the magma from the ambient lithosphere or in a decarbonating plume and thinning the lithosphere by high CO2 pore pressure. The decarbonation converts mantle lherzolites into wehrlite that can coexist with carbonated magmas, dissolving adequate olivine and pyroxene to provide Al, Fe, and Si necessary for the generation of nephelinitic magma composition (e.g. Wyllie and Lee, 1998). In another scenario, Dasgupta et al. (2004) have experimentally shown that carbonated eclogites could be the potential source of continental carbonatites. Based on the geochemical data presented here and in light of experimental phase equilibria studies as mentioned above, we suggest a model for the generation of the alkalic rocks associated with the RBST by invoking melting of carbonated recycled eclogites in the mantle source of a relatively primitive carbonated garnet peridotite plume. A similar scenario was recently suggested by Srivastava et al. (2005) for the Sung Valley carbonatites. This eclogite-derived melt might dissolve enough olivine and pyroxene from the host peridotite for the generation of syenites and ijolites. Pyroxenites and lamproites may result from the reaction of the silicates with the carbonated melts derived both from the eclogite and peridotite. In our batch melting modeling discussed below, carbonatities, sovites, uncompahgrites and some pyroxenites from the Sung, Samchampi and Barpung complexes, based on their trace element abundances and initial Nd-Sr isotopic ratios

68

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

(Fig. 11), are suggested to be derived from a carbonated primitive garnet lherzolite in the Kerguelen plume source (Fig. 14). A carbonate-metasomatized recycled eclogite component is also modeled as a contaminant residing in the Kerguelen plume-head, as required by the Nd–Sr–Pb isotopic signatures (Figs. 10 and 11) of many of the alkalic rocks. The isotopic arrays seen for each of the four alkalic complexes (Figs. 10 and 11) are the result of mixing of the melts in various proportions, derived from these eclogitic and peridotitic sources in the Kerguelen plume-head. Modeling of the REE patterns for Sung carbonatites, Samchampi melteigite, Barpung pyroxenites, and Sikkim lamproites and nephelinites by batch melting of a carbon-

ated peridotite and eclogite are shown in Fig. 13. For Sung carbonatites, we consider the starting composition of the REE to be 1.2 chondrite in a garnet peridotite. This source is assumed to be enriched in LREEs by carbonate metasomatism as a precursor to melting and resulting in modal abundances of the source as 55% olivine, 22% opx, 12% cpx, 7% garnet, and 4% dolomite. Low-degree batch melting (0.2%) of this carbonated garnet peridotite gives an REE pattern similar to the Sung carbonatities and similarly 2% batch melting of such a source gives the REE pattern similar to the Samchampi melteigite. Slightly different modal mineralogy is used to match the REE patterns of the Barpung pyroxenites by 2% batch melting of a carbon-

Fig. 13. Modeling of REE patterns for Sung carbonatites, Samchampi melteigite, Barpung pyroxenites, and Sikkim lamproites and nephelinites for batch melting of carbonated peridotites and eclogites. The source in the mantle is shown to be carbonate metasomatized peridotite or eclogite starting with 1.2 chondritic abundances of the REEs. D-values are from Table T4 (Supplementary materials) discussion.

Fig. 14. A proposed model for the formation of the RBST flood basalt province. The mantle plume ascends with recycled eclogitized basalts now carbonated (Archean) within it. The top of the plume head undergoes high degree of batch melting of the relatively primitive part of the plume in the stability field of spinel to form flood basalts. Lower degree melting of the carbonated relatively primitive peridotite at greater depths of garnet stability also occur to form some of the carbonatites and alkalic rocks. Contamination of melts from carbonated eclogites are needed to account for the variable Nd, Sr, and Pb isotopes of the rest of the alkalic rocks with the lower degree melts in the Kerguelen plume. See text for details. Diagram modified from Takahahshi et al. (1998).

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

ated garnet peridotite (55% olivine, 20% opx, 10% cpx, 11% garnet, 4% dolomite). Sikkim lamproites and nephelinites are modeled by the necessity of different source mineralogy to match their REE patterns by 1% batch melting of a carbonated eclogite (70% clinopyroxene, 25% pyrope garnet, 5% dolomite). Dolomite is considered in the source as the carbonate in these models instead of magnesite, due to lack of REE partitioning data for magnesite. In light of the above discussion we have constructed a schematic diagram (Fig. 14), taking into account the inferences from the geochemical characteristics of each of the lithologies presented here, and describing their petrogenesis from a lower-mantle derived, high 3He/4He (e.g. Ingle et al., 2004) bearing Kerguelen plume-head. In this model the relatively primitive garnet lherzolites are carbonated (dolomitized) by deep mantle CO2 metasomatism causing enrichment in the incompatible elements of the garnet peridotite source. The alkalic–mafic–ultramafic and some of the carbonatitic rocks of this study, as discussed above, are the result of partial melting of this source peridotite. The other important components that are suggested by our geochemical data in the source peridotite are carbonated recycled eclogite of variable ages that produced both chemical and isotopic heterogeneities in the Kerguelen plume source. These recycled components provided the contaminant sources for many of the alkalic rocks, most notably the syenites. The flat REE pattern bearing voluminous tholeiitic flood basalts of the RBST are sourced in the spinel peridotite part of the primitive plume at relatively shallower depths. Some of these basalts are contaminated by the lower crustal granulites of the Eastern Ghats affinity in this model. 6. CONCLUSIONS In light of previous geochemical studies of the Kerguelen plume activity, our study with new results from the RBST province of the Rajmahal flood basalts and from alkaliccarbonatitic volcanics in the Shillong plateau, Mikir hills, and in Sikkim and Bihar leads to the following conclusions: 1. New data of the Rajmahal Traps allow petrogenetic links with previous data of the Sylhet Traps 550 km to the east, and with basalts of comparable ages from the Kerguelen Plateau. Collectively, the data show: (i) the least contaminated tholeiites and some of the alkalic–ultramafic and carbonatitic rocks of this study are similar to the least contaminated KP-derived basalts, sourced from the relatively primitive part of the plume (Neal et al., 2002; Ghatak and Basu, 2011); (ii) a second group of basalts with time-integrated light REE enrichements are considered contiminated by lower crustal Eastern Ghats granulites as shown in our study of the Sylhet Traps (Ghatak and Basu, 2011). The geochemical results, in general, indicate absence of MORB or asthenospheric component in the source of the RBST. 2. The Rajmahal lavas of our study are modeled by 18% batch-melting of a relatively primitive KP at spinel peridotite stabilization depths of 60–70 km, followed by assimilation by melt-erosion of the granulites of the eastern Indian continental lithosphere.

69

3. The Sung Valley carbonatities and pyroxenites, nephelinites, sovites, melteigites, and uncompahgrites from the other three alkalic complexes are modeled, on the basis of Nd–Sr–Pb isotopes and trace elements, by low-degree partial melting products of the primitive KP with a carbonated garnet peridotite parent. These low-degree batch melts are in contrast with a wider range of Nd– Sr–Pb isotopic ratios found in the synenites and ijolites of all four complexes that are modeled by low-degree melting of recycled carbonated eclogites in the KP, requiring chemical heterogeneities of the plume. 4. The diverse rock-types analyzed in our study, in addition to indicating the plume heterogeneities, also document a dynamically ascending KP yielding low-volume ultrabasic melts at greater depths with lower degree melting, followed by the final arrival of this plume at shallower depths in the upper mantle causing flood volcanism by larger degree of melting. 5. The reconstructed Rangit window, shortened by 150 km during the Himalayan orogeny, the Comei igneous complex of Kerguelen age in southeastern Tibet, and the other vestiges of the KP in the RBST as documented and reviewed here suggest at least a 1000 km diameter region was affected by the KP activity in NE India.

ACKNOWLEDGMENTS We are much indebted to the officers of the Geological Survey of India, Calcutta, India, specifically to Dr. S.Sengupta, for collecting and providing the samples analyzed in this study. The geological contexts of these rocks were also provided by the Indian geologists. This study was made possible by an NSF Grant (Instrumentation and Facilities EAR- 9034833) that allowed the upgrade and maintenance of the mass spectrometer at the University of Rochester. Correspondence with and reviews of our previous paper on the Sylhet Traps by Drs.F.A.Frey, the late J.Mahoney and C.Neal helped to improve the interpretations of the combined geochemical data of the Kerguelen-related volcanism in north-eastern India. We are grateful to Drs. M.Norman (Editor), S.Huang(Associate Editor), S. Mallick and two anonymous reviewers for their critical reviews and comments that greatly improved this paper.

APPENDIX A. SUPPLEMENTARY DATA Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.gca.2013.04.004. REFERENCES Alibert C. (1991) Mineralogy and geochemistry of a basalt from Leg 119, Site 738: Implications for the tectonic history of the most southern part of the Kerguelen Plateau. In Proceedings of the Ocean Drilling Program, Scientific Results (eds. J. Barron and B. Larsen) College Station, Texas. pp. 293–298. Bailey D. K. (1993) Carbonate magmas. J. Geol. Soc. 150, 637–651. Baksi A. K., Barman T. R., Raul D. K. and Farrar E. (1987) Widespread early Cretaceous flood basalt volcanism in Eastern India: Geochemical data from the Rajmahal–Bengal–Sylhet Traps. Chem. Geol. 63, 133–141.

70

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Baksi A. K. (1995) Petrogenesis and timing of volcanism in the Rajmahal flood basalt province, northeast India. Chem. Geol. 121, 73–90. Basu A. R. and Tatsumoto M. (1979) Sm–Nd systematics in kimberlites and in the minerals of garnet lherzolite inclusions. Science 205, 398–401. Basu A. R., Sharma M. and DeCelles P. G. (1990) Sr-isotopic provenance and trace element geochemistry of Amazonian foreland basin fluvial sands, Bolivia and Peru; implications for ensialic Andean orogeny. Earth Planet. Sci. Lett. 100, 1–17. Basu A. R., Renne P. R., DasGupta D. K., Teichmann F. and Poreda R. J. (1993) Early and Late alkali igneous pulses and a high-3He plume origin for the Deccan flood basalts. Science 261, 902–906. Basu A. R., Poreda R. J., Renne P. R., Teichmann F., Vasilev Y. R., Sobolev N. V. and Turrin B. D. (1995) High 3He plume origin and temporal–spatial evolution of the Siberian flood basalts. Science 269, 822–825. Bell K. (2001) Carbonatites: Relationships to mantle-plume activity. In Mantle Plumes: Their Identification Through Time (eds. R. E. Ernst and K. L. Buchan). Geological Society of America Special Paper. Bell K. and Simonetti A. (1996) Carbonatite magmatism and plume activity: Implications from the Nd, Pb and Sr isotope systematics of Oldoinyo Lengai. J. Petrol. 37, 1321–1333. Bell K. and Tilton G. R. (2001) Nd, Pb and Sr isotopic compositions of east African carbonatites: Evidence for mantle mixing and plume inhomogeneity. J. Petrol. 42, 1927–1945. Boyet M. and Carlson R. W. (2006) A new geochemical model for the Earth’s mantle inferred from 146Sm–142Nd systematics. Earth Planet. Sci. Lett. 250, 254–268. Chakrabarti R., Basu A. R., Bandyopadhyay P. K. and Zou H. (2011) Age and Origin of the Chilka Anorthosites, Eastern Ghats, India: Implications for Massif Anorthosite Petrogenesis and break-up of Rodinia. In Topics in Igneous Petrology (ed. J. S. Ray et al.). Springer Science, pp. 355–382. Chakrabarti R., Basu A. R. and Paul D. K. (2007) Nd–Hf–Sr–Pb and trace element geochemistry of Proterozoic lamproites from southern India: Subducted komatiite in the source. Chem. Geol. 236, 291–302. Collerson K. D., Williams Q., Ewart A. E. and Murphy D. T. (2010) Origin of HIMU and EM-1 domains sampled by ocean island basalts, kimberlites and carbonatites: The role of CO2fluxed lower mantle melting in thermochemical upwellings. Phys. Earth Planet. Inter. 181, 112–131. Dasgupta R., Hirschmann M. M. and Withers A. C. (2004) Deep global cycling of carbon constrained by the solidus of anhydrous, carbonated eclogite under upper mantle conditions. Earth Planet. Sci. Lett. 227, 73–85. DePaolo D. J. (1988) Neodymium Isotope Geochemistry: An Introduction (ed. P. J. Wyllie). Springer-Verlag, pp. 1–187. Evensen N. M., Hamilton P. J. and O’Nions R. K. (1978) Rareearth abundances in chondritic meteorites. Geochim. Cosmochim. Acta 42, 1199–1212. Fitton J. G., Saunders A. D., Norry M. J., Hardarson B. S. and Taylor R. N. (1997) Thermal and chemical structure of the Iceland plume. Earth Planet. Sci. Lett. 153, 197–208. Frey F. A., McNaughton N. J., Nelson D. R., deLaeter J. R. and Duncan R. A. (1996) Petrogenesis of the Bunbury Basalt, Western Australia: Interaction between the Kerguelen plume and Gondwana lithosphere? Earth Planet. Sci. Lett. 144, 163– 183. Frey F. A., Coffin M. F., Wallace P. J., Weis D. and Zhao X. (2000) Origin and evolution of a submarine large igneous province: The Kerguelen Plateau and Broken Ridge, southern Indian Ocean. Earth Planet. Sci. Lett. 176, 73–89.

Frey F. A., Weis D., Borisova A. Y. and Xu G. (2002) Involvement of continental crust in the formation of the Cretaceous Kerguelen Plateau: New perspectives from ODP leg 120 sites. J. Petrol. 43, 1207–1239. Ghatak A. and Basu A. R. (2011) The Sylhet Traps: Vestiges of the Kerguelen plume in northeastern India. Earth Planet. Sci. Lett. 308, 52–64. Hannigan R. E., Basu A. R. and Teichmann F. (2001) Mantle reservoir geochemistry from statistical analysis of ICP-MS trace element data of equatorial Mid-Atlantic MORB glasses. Chem. Geol. 175, 397–428. Hart S. and Zindler A. (1989) Constraints on the nature and development of chemical heterogeneities in the mantle. In Mantle Convection: Plate Tectonics and Global Dynamics (ed. W. R. Peltier). Gordon and Breach, New York, pp. 261–387. Heaman L. M., Srivastava R. K. and Sinha A. K. (2002) A precise U–Pb zircon/baddeleyite age for the Jasra igneous complex, Karb-Analong district, Assam, NE India. Curr. Sci. 82, 744– 748. He´mond C., Arndt N. T., Lichtenstein U. and Hofmann A. W. (1993) The heterogeneous Iceland plume: Nd–Sr–O isotopes and trace element constraints. J. Geophys. Res. 98, 15833– 15850. Hoda S. Q., Rawat T. P. S., Krishnamurthy P. and Dwivedy K. K. (1997) Geology and economic resources of the Samchampi alkaline carbonatite complex, Mikir hills, Assam, India. Explor. Res. At. Miner. 10, 79–86. Hofmann A. W., Jochum K. P., Seufert H. M. and White W. M. (1986) Nb and Pb in oceanic basalts: new constraints on mantle evolution. Earth Planet. Sci. Lett. 79, 33–45. Hoskin P. W. O. and Schaltegger U. (2003) The composition of zircon and igneous and metamorphic petrogenesis. Rev. Mineral. Geochem. 53, 27–62. Ingle S., Weis D. and Frey F. A. (2002) Indian continental crust recovered from Elan Bank, Kerguelen Plateau (ODP Leg 183, Site 1137). J. Petrol. 43, 1241–1257. Ingle S., Scoates J. S., Weis D., Brugmann G. and Kent R. W. (2004) Origin of Cretaceous continental tholeiites in southwestern Australia and eastern India: Insights from He and Os isotopes. Chem. Geol. 209, 83–106. Jacobsen S. B. and Wasserburg G. J. (1984) Sm–Nd isotopic evolution of chondrites and achondrites, II. Earth Planet. Sci. Lett. 67, 137–150. Kent R. W., Saunders A. D., Kempton P. D. and Ghose N. C. (1997) Rajmahal Basalts, eastern India: mantle sources and melt distribution at a volcanic rifted margin. In Large Igneous Provinces: Continental, Oceanic, and Planetary Flood Volcanism (eds. J. Mahoney and M. F. Coffin). Geophysical Monograph (AGU), Washington, DC, pp. 145–182. Kent R. W., Pringle M. S., Muller R. D., Saunders A. D. and Ghose N. C. (2002) 40Ar/39Ar geochronology of the Rajmahal Basalts, India, and their relationship to the Kerguelen Plateau. J. Petrol. 43, 1141–1153. Kumar A., Dayal A. M. and Padmakumari V. M. (2003) Kimberlite from Rajmahal magmatic Province: Sr–Nd–Pb isotopic evidence for Kerguelen plume derived magmas. Geophys. Res. Lett. 30, 2053–2056. Kumar D., Mamallan R. and Dwivedy K. K. (1996) Carbonatite magmatism in north India. J. South Asian Earth Sci. 13, 145– 158. Kumar P., Yuan X., Ravi Kumar M., Kind R., Li X. and Chadha R. K. (2007) The rapid drift of the Indian tectonic plate. Nature 449, 894–897. Lee W.-J. and Wyllie P. J. (1997) Liquid immiscibility between nephelinite and carbonatite from 2.5 to 1.0 GPa compared with mantle melt compositions. Contrib. Miner. Petrol. 127, 1–16.

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72 Mahoney J. J., Jones W. B., Frey F. A., Salters V. J. M., Pyle D. G. and Davies H. L. (1995) Geochemical characteristics of lavas from Broken Ridge, the Naturaliste Plateau and southernmost Kerguelen Plateau: Cretaceous plateau volcanism in the southeast Indian Ocean. Chem. Geol. 120, 315–345. Mahoney J. J., Graham D. W., Christie D. M., Johnson K. T. M., Hall L. S. and VonderHaar D. L. (2002) Between a hotspot and a cold spot: isotopic variation in the Southeast Indian Ridge asthenosphere, 86°E–118°E. J. Petrol. 43, 1155–1176. Mahoney J. J., Macdougall J. D., Lugmair G. W. and Gopalan K. (1983) Kerguelen hotspot source for Rajmahal and Ninetyeast Ridge? Nature 303, 385–389. Mahoney J. J., Roex A. P., Peng Z., Fisher R. L. and Natland J. H. (1992) Southwestern limits of the Indian Ocean Ridge mantle and the origin of low 206Pb/204Pb mid-oceanic ridge basalt: Isotopic systematics of the central Southwest Indian Ridge (170–500 E). J. Geophys. Res. 97, 19771–19790. Mitchell R. H. and Bergman S. C. (1991) Petrology of Lamproites. Plenum Press, New York. Mitra G., Bhattacharyya K. and Mukul M. (2010) The lesser Himalayan Duplex in Sikkim: Implications for variations in Himalayan shortening. J. Geol. Soc. India 75, 289–301. Murphy D. T. and Collerson K. D. (2002) Lamproites from Gaussberg, Antarctica: Possible transition zone melts of Archean subducted sediments. J. Petrol. 43, 981–1001. Neal C. R., Mahoney J. J. and Chazey, III, W. J. (2002) Mantle sources and the highly variable role of continental lithosphere in basalt petrogenesis of the Kerguelen Plateau and Broken Rigde LIP: Results from ODP leg 183. J. Petrol. 43, 1177–1205. Nesbitt R. W., Sun S. S. and Purvis A. C. (1979) Komatiites: Geochemistry and Genesis. Canadian Mineralogist 17, 165–186. Pantulu G., Macdougall J. D., Gopalan K. and Krishnamurthy P. (1992) Isotopic and chemical compositions of Sylhet Traps basalts: Links to the Rajmahal Traps and Kerguelen hotspot. American Geophysical Union, EOS Transactions 72. Peng Z. X., Mahoney J., Hooper P., Harris C. and Beane J. (1994) A role for lower continental crust in flood basalt genesis? Isotopic and incompatible element study of the lower six formations of the western Deccan Traps. Geochim. Cosmochim. Acta 58, 267–288. Rao N. V. C., Gibson S. A., Pyle D. M., Dickin A. P. and Day J. (2004) Petrogenesis of Proterozoic lamproites and kimberlites from the Cuddapah Basin and Dharwar Craton, southern India. J. Petrol. 45, 907–948. Ray J. S., Pattanayak S. K. and Pande K. (2005) Rapid emplacement of the Kerguelen plume-related Sylhet Traps, eastern India: Evidence from 40Ar–39Ar geochronology. Geophys. Res. Lett. 32, 1–4. Rickers K., Mezger K. and Raith M. M. (2001) Evolution of the continental crust in the Proterozoic Eastern Ghats Belt, India and new constraints for Rodinia reconstruction: Implications from Sm–Nd, Rb–Sr and Pb–Pb isotopes. Precambr. Res. 112, 183–210. Rock N. M. S., Griffin B. J., Edgar A. D., Paul D. K. and Hergt J. M. (2012) A spectrum of potentially diamondiferous lamproites and minettes from the Jharia coalfield, eastern India. Journal of Volcanology and Geothermal Research 50, 55–83. Rudnick R. L. and Fountain D. M. (1995) Nature and composition of the continental crust: A lower crustal perspective. Rev. Geophys. 33, 267–309. Salters V. and Stracke A. (2004) Composition of the depleted mantle. Geochem. Geophys. Geosyst. 5. http://dx.doi.org/ 10.1029/2003GC000597. Sarkar A., Paul D. K., Balasubrahmanyan M. N. and Sengupta N. R. (1980) Lamprophyres from the Indian Gondwanas. J. Geol. Soc. India 37, 547–559.

71

Sengupta S. (1966) Geological and geophysical studies in Western part of Bengal Basin, India. Bull. Am. Assoc. Petrol. Geol. 50, 1001–1017. Sharma M., Basu A. R. and Nesterenko G. V. (1992) Temporal Sr, Nd-, and Pb-isotopic variations in the Siberian flood basalts; implications for the plume-source characteristics. Earth Planet. Sci. Lett. 113, 365–381. Srivastava R. K., Heaman L. M., Sinha A. K. and Shihua S. (2005) Emplacement age and isotope geochemistry of Sung Valley alkaline–carbonatite complex, Shillong Plateau, northeastern India: Implications for primary carbonate melt and genesis of the associated silicate rocks. Lithos 81, 33–54. Storey M., Kent R. W., Saunders A. D., Hergt J., Salters V. J. M., Whitechurch H., Sevigny J. H., Thirlwall M. F., Leat P., Ghose N. C. and Gifford M. (1992) Lower cretaceous volcanic rocks on continental margins and their relationship to the Kerguelen Plateaus. In Proceedings of the Ocean Drilling Program, Scientific Results (eds. S. W. Wise and R. Schlich), Ocean Drilling Program, College Station, Texas. pp. 33–53. Storey M., Saunders A. D., Tarney J., Gibson I. L., Norry M. J., Thirwall M. F., Leat P., Thompson R. N. and Menzies M. A. (1989) Contamination of Indian Ocean asthenosphere by the Kerguelen-Heard mantle plume. Nature 338, 574–576. Sun S.-S. and McDonough W. F. (1989) Chemical and isotopic systematics of oceanic basalts: Implications for mantle composition and processes. In Magmatism in the Ocean Basins (eds. A. D. Saunders and M. J. Norry). Geological Society of London Special Publications, United Kingdom, pp. 313–345. Takahahshi E., Nakajima K. and Wright T. L. (1998) Origin of the Columbia River basalts: Melting model of a heterogeneous plume head. Earth Planet. Sci. Lett. 162, 63–80. Taylor S. R. and McLennan S. M. (1995) The geochemical evolution of the crust. Rev. Geophys. 33, 241–265. Veena K., Pandey B. K., Krishnamurthy P. and Gupta J. N. (1998) Sr and Nd isotopic systematics of the carbonatites of Sung Valley, Meghalaya, Northeast India: Implications for contemporary plume-related mantle source characteristics. J. Petrol. 39, 1875–1884. Weaver K. L. (2000). , p. 241. Isotopic and Trace Element Geochemistry of Cretaceous Continental Volcanism, Northeast India: A Link to the Kerguelen Plume. Weis D., Bassias Y., Gautier I. and Mennessier J.-P. (1989) Dupal anomaly in existence 115 Ma ago: Evidence from isotopic study of the Kerguelen Plateau (South Indian Ocean). Geochim. Cosmichim. Acta 53, 2125–2131. Weis D. and Frey F. A. (1991) Isotope geochemistry of the ninetyeast ridge basalts: Sr, Nd, and Pb evidence for the involvement of the Kerguelen hotspot. In Proceedings of the Ocean Drilling Program, Scientific Results (eds. J. Weissel, J. Peirce, E. Taylor and J. Alt). College Station, Texas. pp. 591–610. Weis D., Frey F. A., Leyrit H. and Gautier I. (1993) Kerguelen Archipelago revisited: Geochemical and isotopic study of the Southeast Province lavas. Earth Planet. Sci. Lett. 118, 101–119. Weis D., Frey F. A., Giret A. and Cantagrel J.-M. (1998) Geochemical characteristics of the youngest volcano (Mount Ross) in the Kerguelen Archipelago: Inferences for magma flux, lithosphere assimilation and composition of the Kerguelen plume. J. Petrol. 39, 937–994. Wyllie P. J. and Huang W. L. (1976) Carbonation and melting reactions in the system CaO–MgO–SiO2–CO2 at mantle pressures with geophysical and petrological applications. Contrib. Miner. Petrol. 54, 79–107. Wyllie P. J., Lee W.-J. 1998. Kimberlites, carbonatites, peridotites and silicate–carbonate liquid immiscibility explained in parts of the system CaO–(Na2O + K2O)–(MgO + FeO)–(SiO2+

72

A. Ghatak, A.R. Basu / Geochimica et Cosmochimica Acta 115 (2013) 46–72

Al2O3)–CO2. Extended Abstracts, 7th International Kimberlite Conference, Cape Town, University of Cape Town 974–976. Yin A., Dubey C. S., Webb A. A. G., Kelty T. K., Grove M., Gehrels G. E. and Burgess W. P. (2010) Geologic correlation of the Himalayan orogen and Indian craton: Part 1. Structural geology, U–Pb zircon geochronology, and tectonic evolution of the Shillong Plateau and its neighboring regions in NE India. Geol. Soc. Am. Bull. 122, 336–359.

Zartman R. E. and Doe B. R. (1981) Plumbotectonics – The model. Tectonophysics 75, 135–162. Zhu D.-C., Chung S.-L., Mo X.-X., Zhao Z.-D., Niu Y., Song B. and Yang Y.-H. (2009) The 132 Ma Comei-Bunbury large igneous province: Remnants identified in present day southeastern Tibet and southwest Australia. Geology 37, 583–586. Associate editor: Shichun Huang